Lu pour vous : rapport du PNUD « Développement Humain en 2019 »

Human Development Report 2019

Beyond income, beyond averages, beyond today: Resilientnations. The 2019 Human Development Report is the latest in the series of global Human Development Reports published by the United Nations Development Programme (UNDP) since 1990 as independent, analytically and empirically grounded discussions of major development issues, trends and policies.

Additional resources related to the 2019 Human Development Report can be found online at http://hdr.undp.org, including digital versions and translations of the Report and the overview in more than 10 languages, an interactive web version of the Report, a set of background papers and think pieces commissioned for the Report, interactive data visualizations and databases of human development indicators, full explanations of the sources and methodologies used in the Report’s composite indices, country profiles and other background materials as well as previous global, regional and national Human Development Reports. Cor- rections and addenda are also available online.

The cover conveys the inequalities in human development of a changing world. The dots in different colors represent the com- plex and multidimensional nature of these inequalities. The shad- ow of the climate crisis and sweeping technological change, evoked by the color of the cover background that suggests heat, will shape progress in human development in the 21st century.

Copyright @ 2019 By the United Nations Development Programme 1 UN Plaza, New York, NY 10017 USA

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted, in any form or by means, electronic, mechanical, photocopying, recording or otherwise, without prior permission.

Sales no.: E.20.III.B.1 ISBN: 978-92-1-126439-5 eISBN: 978-92-1-004496-7 Print ISSN: 0969-4501 eISSN: 2412-3129

A catalogue record for this book is available from the British Library and Library of Congress

General disclaimers. The designations employed and the presentation of the material in this publication do not imply the expression of any opinion whatsoever on the part of the Human Development Report Office (HDRO) of the United Nations Development Programme (UNDP) concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. Dotted and dashed lines on maps represent approximate border lines for which there may not yet be full agreement.

The findings, analysis, and recommendations of this Report, as with previous Reports, do not represent the official position of the UNDP or of any of the UN Member States that are part of its Executive Board. They are also not necessarily endorsed by those mentioned in the acknowledgments or cited.

The mention of specific companies does not imply that they are endorsed or recommended by UNDP in preference to others of a similar nature that are not mentioned.

Where indicated, some figures in the analytical part of the Report were estimated by the HDRO or other contributors and are not necessarily the official statistics of the concerned country, area or territory, which may be based on alternative methods. All the figures used to calculate the human development composite indices are from official sources. All reasonable precautions have been taken by the HDRO to verify the information contained in this publication. However, the published material is being distributed without warranty of any kind, either expressed or implied.

The responsibility for the interpretation and use of the material lies with the reader. In no event shall the HDRO and UNDP be liable for damages arising from its use.

Printed in the USA, by AGS, an RR Donnelley Company, on Forest Stewardship Council certified and elemental chlorine-free papers. Printed using vegetable-based ink. Human Development Report 2019

Published for the United Nations Programme (UNDP)

Human Development Report 2019 Team

Director and lead author

Pedro Conceição

Research and statistics

Jacob Assa, Cecilia Calderon, George Ronald Gray, Nergis Gulasan, Yu-Chieh Hsu, Milorad Kovacevic, Christina Lengfelder, Brian Lutz, Tanni Mukhopadhyay, Shivani Nayyar, Thangavel Palanivel, Carolina Rivera and Heriberto Tapia

Production, communications, operations

Botagoz Abdreyeva, Oscar Bernal, Andrea Davis, Rezarta Godo, Jon Hall, Seockhwan Bryce Hwang, Admir Jahic, Fe Juarez Shanahan, Sarantuya Mend, Anna Ortubia, Yumna Rathore, Dharshani Seneviratne, Elodie Turchi and Nu Nu Win

External contributors

Chapter 3 (by the World Inequality Lab): Lucas Chancel, Denis Cogneau, Amory Gethin, Alix Myczkowski and Thomas Piketty Boxes and spotlights: Elizabeth Anderson, Michelle Bachelet, Bas van Bavel, David Coady, James Foster, Nora Lustig and Ben Philips

ii | HUMAN DEVELOPMENT REPORT 2019 HANGING. PROVING. ELIFVoreEwoRrdING.

The wave of demonstrations sweeping across further entrench inequalities and consolidate

countries is a clear sign that, for all our pro- the power and political dominance of the few.

A REMARgKrAesBs, LsoEmYeEthAinRg FinOoRurUgNloDbaPli.zed sTohcieetyyriesinforcWehaatrwaiet Iahreasveeeicnog mtoedatyoias dthmeircereisnt othfisa

rst full yeanorDtowifffoeirmrkeinpntglet.rmiggeenrstianrge barinngeinwg peooprlegaonntiosatiodwnoa:wvietnsotpfoioncteehqonuitacilaei.tlytJ.ouWsttahakasteihncaehpqapuenanlgisteynetboxetgsciconams leaets.

an ­ a pltahne sbtrueeilts:tothehceolspt ocfoautrnatirnietiscket, Itthiesparipceotenbtiirathl ,wdeefinneesetdhetofrfeueldfiollmtoansudpoppoportrtucnoiutines-

he SustainoafAbpleectrooDnln,epevocetliiltnoigcpatlmhdreeemandat,nGtdhsoofaoulrgshi.n,disepdterenieedpsenaicnned.meeothtfioncsgheiltodhfreethnae,mandbeuxitltitosgneannoedfrtaehtliedoen2r,s0,s3oan,0tdoAopg,eerpmnoldeiacati.eess

rising frustration with inequalities. to prevent inequalities can follow the lifecycle. change forUtnhdeerUstnanitdeindgNhoawtiotonsa,dwdreess todInay’2s 0di1s8- , weFrdomemproen­slatbraouter dmatrhkeattinwveestmareentsuinpthtoe

h our parqtnuieetrsreqtouirheselolpokpinego”BpeleyongdeInt comet,hBeeyocnhdallehnegaleth, awnditnhuttrhiteionhoifgyhoeunsgt cphirlodrgernatmo min­e

Averages and Beyond Today,” as this Human and post­labour market investments around

et and staDyetvheleopreme­ntmReepeotritnsgetssohuot rtot-do. delivery inacfciveses tyoecaarpsit,arl,emstionirmedumfinwaagnecsiaanl dstsaobcilail-

s while layinTgoothoeftenfo, uinneqduaatliiotyniss ffroamr ead arotuyn,dineccor-easseerdviceefsf,icpioelnitcicyianasnadndapogleicoymgraakeprhs ihcaavlelya- nomics, fed and measured by the notion that gbfoaertnttehdreyecoro-fbnctaheloxaticnoecfseetahdcaht, ciloef ucaondrtrreeyrcsotlhryigpcroomuptbe,inwaemidll

nfident futmuarkei.ng money is the most importantdtihviengrsien,

life. ­ even astrawneslatpeoinwtoeraelidfelUonNg inrevefostmrmentainnde,qunaoliwty,

and the cLoatthoknBiksetuieantesCupssuoehpmcaeitepnodttniieoibserneslao,rgiaencnicoktdrhnewtea,hikhretiinhplegoaopcruuketnegoedtprhse-l,reptmhoewaasysetgtprereraiposinnttechobsietfeasckiamnneMfdtrnhosateumktsictonaogictnuatoahcnbokotilslrriiedinteycgi.nshtoawhitceienescgsostemaUrrpvtnlseeiwx.teiittdyh aNcoamtimonits- fragility of human

ess, we sparwotaagonneiswt owf athyisosftowroy:rkthinegpower of the development—to pushing the boundaries to

umanitarianpfeowawn;etdrhoedfpteohvweeeprleoloepspsmlneeetsosnodtfeammcaatnnodyr;scahnadngcAeo.lsleActdivme inihSsuetrlsptaaticonoaru,bniltetrwDieaesvsaenlaodptmrcuoementmpGuleonaailtssi.eusrereatloizeletahde

ridging life-sGaoviinnggbereyosnpdoinnscoemwe iwthillrree-quirethtaecklin#gNextGTheins iUs tNheDmPissitornanatstfhoerhmeaarttiofnthedUunritnegd

developmenenontrrtme.nscehmedbeindtdeerdesdtse—epthweitshoicniaal annadti2opn0o’ls1i8toi,cradal isruNtopagtteiitnohngesrhwDoietwhvetlhwoepe1m7te0hncitnoPkurn,otigrnirevasmeasnmtde, ,tmewrarointrokarigineegs

group’s history and culture. and deliverwtoe saecrvcee.lerate sustainable development.

to Fiji, we saLwooakisnug rbgeeyoonfdintondoavy,athioen2019 Human Some 40 years ago the founding father of

people wDithevtehloepmseenrvtiRceepsotrht earytincueleatdes theTroisde aoyf ,a UNhDumPa’sn dmevieslsopiomnenht,aPsrofnesesovreAr mbaerteyanSeans, new generation of inequalities. asked a deceptively simple question: equality of povertyJu, stshasatphienggapgionvbeasrincalnivcineg stancdlearadrs: iswe oafrwehahte? rHee taonswheerlepd wthiteh e17qu0alcsoimupnlitcriiteys:

f the futunraer.roAwinygo, uwtihth eanmupnlporeycmedeenntted nuamnbderteofrritoorfiethse tihninwgshwicehcarwe eabocuutrtroebnutillyd twheofruktutroe

people in the world escaping poverty, hunger we aspire to.

reneurshiapndinditiisaeatisvee, thtehaabtilsittiaerstepedopilne willreneaecdhtotheir PdreofveessloorpSmene’s nwtorpdrsioherliptieuss tosotaktehaatfrensoh

ack in 201c3omispenteoinwthienim10mecdoiautenfutrtiuerse haveoenvoelvoedn. thilsoopkla; ntoegt oisbleeyfotnbdeghrionwdt.h and markets to A new gap has opened, such as in tertiary understand why people take to the streets in be scaledetuudnpuictcaietoisononnticanenedcnoantc-swcideisedsreetdoblbuyrxoutahrdieebsatnhda—tWaoreepnplooowro- k n. foprrIowtweaostru,dladntldiokweythooattuhlreaanpdkaearrlsltncthaenorsdseohwiahpbooahuantvidet.tackoenl-

considered critical to compete andlabbeloornagt,ionthoins jothurenejyooufrenxeplyo.ration with us over the past

pages yaponauritniwccuriellaalrslisynegiennuasmkobnmeorweolfeodyogf eutnhegceopneoomplye,awreheedre- 12 months, and I encourage you to read on.

ts we achuiecavteedd, cionnn2e0ct1e8d.aTndhestyucrkewinit-h no ladder of

UNDP is ucnhiAoqitucteehseltyosamdmeoevsteiigmunpe,.ecdlimtoatehcehalpnge, gender in-

plex deveeqloupalmityeanntd vpiorolebntlecomnsflicitncoantinue to drive

s, integrataAendsdthaenentHdreuninmchnanobavDsaiectviavenleodpwnmeaewnyti.nReeqpuoarlittiseestsaloikuet,. Achim SteAincehr im Steiner AUdnmiteindisNtAraatdtioomrnsinDisetvrealotpomr ent Programme failure to address these systemic challenges will United Nations Development Programme

Foreword | iii Acknowledgements

Producing a Human Development Report Carol Graham, Kenneth Harttgen, Homi is truly a collective endeavour. It reflects the Kharas, Michèle Lamont, Santiago Levy, Ako formal and informal contributions of many Muto, Ambar Nayaran, Alex Reid, Carolina people and institutions. What ultimately is Sánchez-Páramo, Paul Segal, Amartya Sen, included in these pages cannot fully capture Juan Somavia, Yukio Takasu, Senoe Torgerson the richness of ideas, interactions, partnerships and Michael Woolcock. and collaborations associated with the effort. These acknowledgements are an imperfect Appreciation is also extended for the writ- attempt at recognizing those who generously ten contributions by Lucas Chancel and our gave their time and energy to help produce colleagues at the World Inequality Lab, who the 2019 Human Development Report—with contributed chapter 3 of the Report. Boxes apologies for the many that contributed and and spotlights were contributed by Elizabeth that we have failed to include here. As authors, Anderson, Michelle Bachelet, Bas van Bavel, we hope that the content lives up to the out- David Coady, James Foster, Nora Lustig, standing contributions that were received and Ben Philips, the International Lesbian, Gay, that the Report adds to what the UN General Bisexual, Trans and Intersex Association Assembly has recognized as “an independent and the Peace Research Institute in Oslo. intellectual exercise” that has become “an Background papers and written inputs were important tool for raising awareness about prepared by Fabrizio Bernardi, Dirk Bezemer, human development around the world.” Matthew Brunwasser, Martha Chen, Sirianne Dahlum, Olivier Fiala, Valpy FitzGerald, James Our first word of thanks goes to the K. Galbraith, Jayati Ghosh, John Helliwell, members of our Advisory Board, energet- Martin Hilbert, Patrick Kabanda, Emmanuel ically led by Thomas Piketty and Tharman Letouze, Juliana Martínez, Håvard Mokleiv, Shanmugaratnam in their Co-Chair roles. The José Antonio Ocampo, Gudrun Østby, Inaki other members of the Advisory Board were Permanyer, Ilze Plavgo, Siri Aas Rustad, Diego Olu Ajakaiye, Kaushik Basu, Haroon Bhorat, Sánchez-Ancochea, Anya Schiffrin, Jeroen Francisco Ferreira, Janet C. Gornick, David P.J.M. Smits, Eric Uslaner, Kevin Watkins and Grusky, Ravi Kanbur, Enrico Letta, Chunling Martijn van Zomeren. We are thankful to all Li, Nora Lustig, Laura Chinchilla Miranda, of them. Njuguna Ndung’u and Frances Stewart. A number of consultations with thematic Complementing the advice from our and regional experts were held between March Advisory Board, the Report’s Statistical and September 2019, including in Beirut, Advisory Panel provided guidance on several Bonn, Buenos Aires, Cairo, Doha, Geneva, methodological and data aspects of the Report, Marrakech, Nairobi, Nursultan, Paris, Rabat in particular related to the calculation of the and Tokyo. For their inputs during these con- Report’s human development indices. We sultations, we are especially grateful to Touhami are grateful to all the panel members: Oliver Abdelkhalek, Touhami Abi, Hala Abou Ali, Chinganya, Albina A. Chuwa, Ludgarde Laura Addati, Shaikh Abdulla bin Ahmed Al Coppens, Marc Fleurbaey, Marie Haldorson, Khalifa, Ibrahim Ahmed Elbadawi, Asmaa Friedrich Huebler, Dean Mitchell Jolliffe, Al Fadala, Abdulrazak Al-Faris, Najla Ali Yemi Kale, Steven Kapsos, Robert Kirkpatrick, Murad, Facundo Alvaredo, Yassamin Ansari, Jaya Krishnakumar, Mohd Uzir Mahidin, Max Kuralay Baibatyrova, Alikhan Baimenov, Roser and Pedro Luis do Nascimento Silva. Radhika Balakrishnan, Carlotta Balestra, Luis Beccaria, Debapriya Bhattacharya, Roberto Many others provided generous sugges- Bissio, Thomas Blanchet, Sachin Chaturvedi, tions without any formal advisory role, in- Alexander Chubrik, Paulo Esteves, Elyas cluding Sabina Alkire, Sudhir Anand, Amar Felfoul, Cristina Gallach, Amory Gethin, Battacharya, Sarah Cliffe, Miles Corak, Angus Sherine Ghoneim, Liana Ghukasyan, Manuel Deaton, Shanta Devarajan, Vitor Gaspar, Glave, Xavier Godinot, Heba Handoussa, and Social Commission for Western Asia Gonzalo Hernández-Licona, Ameena (ESCWA); Roger Gomis, Damian Grimshaw, Hussain, Hatem Jemmali, Fahmida Khatun, Stefan Kühn and Perin Sekerler from the Alex Klemm, Paul Krugman, Nevena Kulic, International Labor Organization (ILO); Christoph Lakner, Tomas de Lara, Eric Livny, Astra Bonini, Hoi Wai Jackie Cheng, Elliott Paul Makdisi, Gordana Matkovic, Rodrigo Harris, Ivo Havinga, Marcelo Lafleur, Shantanu Márquez, Roxana Maurizio, Marco Mira, Mukherjee, Marta Roig, Michael Smedes Cielo Morales, Salvatore Morelli, Rabie and Wenyan Yang from the United Nations Nasr, Heba Nassar, Andrea Villarreal Ojeda, Department of Economic and Social Affairs Chukwuka Onyekwena, Andrea Ordonez, (UNDESA); Manos Antoninis, Bilal Fouad Magued Osman, Mónica Pachón, Emel Memi Barakat and Anna Cristina D’Addio from the Parmaksiz, Maha El Rabbat, Racha Ramadan, United Nations Educational, Scientific and Hala El Saeed, Ouedraogo Sayouba, Sherine Cultural Organization (UNESCO); Lakshmi Shawky, André de Mello e Souza, Paul Stubbs, Narasimhan Balaji, Laurence Chandy and Hamid Tijani, René Mauricio Valdés, Peter Mark Hereward from the United Nations Van de Ven, Ngu Wah Win, Xu Xiuli, Cai Children’s Fund (UNICEF); Shams Banihani, Yiping, Sabina Ymeri and Stephen Younger. Jorge Chediek and Xiaojun Grace Wang from Further support was also extended by other the United Nations Office for South-South individuals who are too numerous to men- Cooperation (UNOSSC); Paul Ladd from tion here (consultations are listed at http:// the United Nations Research Institute for hdr.undp.org/en/towards-hdr-2019 with Social Development (UNRISD); Rachel more partners and participants mentioned Gisselquist, Carlos Gradin and Kunal Sen from at http://hdr.undp.org/en/acknowledge- the United Nations University World Institute ments-hdr-2019). Contributions, support for Development Economics Research (UNU- and assistance from partnering institutions, WIDER); Margaret Carroll and Emma including UNDP regional bureaus and coun- Morley from the UN Volunteers (UNV); try offices, are also acknowledged with much Shruti Majumdar, Shahrashoub Razavi and gratitude. Silke Staab from the United Nations Entity for Gender Equality and the Empowerment The Report also benefited from peer re- of Women (UN Women); and Theadora Swift views of each chapter by Paul Anand, Carlos Koller from the World Health Organization Rodriguez Castelan, Lidia Ceriani, Daniele (WHO). Checchi, Megan Cole, Danny Dorling, Csaba Feher, Oliver Fiala, Maura Francese, Aleksandr Many colleagues in UNDP provided ad- V. Gevorkyan, Leonard Goff, Didier Jacobs, vice and encouragement. Luis Felipe López- Silpa Kaza, Jeni Klugman, Anirudh Krishna, Calva, Michele Candotti, Joseph D’Cruz and Benoit Laplante, Max Lawson, Marc Morgan, Abdoulaye Mar Dieye gave guidance not only Teresa Munzi, Brian Nolan, Zachary Parolin, on the content of the Report but also towards Kate E. Pickett, Sanjay Reddy, Pascal Saint- the evolution of the Human Development Amans, Robert Seamans, Nicholas Short and Report Office over the coming years. We Marina Mendes Tavares. are grateful, in addition, to Marcel Alers, Fernando Aramayo, Gabriela Catterberg, We are grateful to many colleagues in the Valerie Cliff, Esuna Dugarova, Mirjana United Nations family that supported the Spoljaric Egger, Almudena Fernández, Cassie preparation of the report by hosting consulta- Flynn, Stephen Gold, Nicole Igloi, Boyan tions or providing comments and advice. They Konstantinov, Raquel Lagunas, Marcela include Prosper Tanyaradzwa Muwengwa Meléndez, Ruben Mercado, Ernesto Pérez, and Thokozile Ruzvidzo from the Economic Kenroy Roach, Renata Rubian, Narue Shiki, Commission for Africa (ECA); Alberto Arenas, Ben Slay, Mourad Wahba, Douglas Webb, Alicia Bárcena, Mario Cimoli and Nunzia Haoliang Xu and Diego Zavaleta. Saporito from the Economic Commission for Latin America and the Caribbean (ECLAC); We were fortunate to have the support of Khalid Abu-Ismail, Oussama Safa, Niranjan talented interns—Farheen Ghaffar, Michael Sarangi and Saurabh Sinha from the Economic Gottschalk, Xiao Huang, Sneha Kaul and

Acknowledgements | v Adrian Pearl—and fact checkers—Jeremy always challenging us to aim higher, while Marand, Tobias Schillings and Emilia giving us the space to be bold. He called for Toczydlowska. a Report that would speak to the public, to policymakers and to experts—because that is The Human Development Report Office also the only way to advance the cause of human extends its sincere gratitude to the Republic of development. We hope we have lived up to Korea for its financial contribution. Their on- those expectations. going support and dedication to development research and this Report is much appreciated. Pedro Conceição

We are grateful for the highly professional ed- Director iting and layout by a team at Communications Human Development Report Office Development Incorporated—led by Bruce Ross-Larson, with Joe Caponio, Nick Moschovakis, Christopher Trott and Elaine Wilson.

We are, to conclude, extremely grateful to the UNDP Administrator Achim Steiner for

vi | HUMAN DEVELOPMENT REPORT 2019 Contents

CHAPTER 4

Foreword iii Gender inequalities beyond averages: Between social norms and Acknowledgements iv power imbalances Overview 147

PART I Gender inequality in the 21st century 148 Beyond income 1 Are social norms and power imbalances shifting? 152

Restricted choices and power imbalances over the lifecycle 158

Empowering girls and women towards gender equality: A template to

reduce horizontal inequalities 164

CHAPTER 1 PART III

Inequality in human development: Moving targets in the 21st century 29 Beyond today 171

Understanding inequality in capabilities 30

Dynamics of inequality in human development: Convergence in basic CHAPTER 5

capabilities, divergence in enhanced capabilities 32 Climate change and inequalities in the Anthropocene 175

Convergence in the basics is not benefiting everyone: Identifying How climate change and inequalities in human development are

those furthest behind 48 intertwined 178

Towards enhanced agency 51 Environmental inequalities and injustices are pervasive—a global

Moving targets and 21st century inequalities 57 snapshot of waste, meat consumption and water use 186

A break from the past: Making new choices for people and planet 192

CHAPTER 2

Inequalities in human development: Interconnected and persistent 73 CHAPTER 6

How inequalities begin at birth— and can persist 74 Technology’s potential for divergence and convergence: Facing a

How inequalities interact with other contextual determinants of human century of structural transformation 199

development 82 Inequality dynamics in access to technology: Convergence in basic,

Inequalities can accumulate through life, reflecting deep power divergence in enhanced 200

imbalances 93 Technology is reshaping the world: How will it shape inequality in

human development? 205

Harnessing technology for a Great Convergence in human development 208

PART II

Beyond averages 97 CHAPTER 7

Policies for reducing inequalities in human development in the 21st

CHAPTER 3 century: We have a choice 223

Towards convergence in capabilities beyond income: From basic to

Measuring inequality in income and wealth 103 enhanced universalism 225

Tackling inequality starts with good measurement 103 Towards inclusive income expansion: Raising productivity and

The elephant curve of global inequality and growth 109 enhancing equity 233

How unequal is Africa? 116 Postscript: We have a choice 245

Inequality in BRIC countries since the 2000s 119

Inequality and redistribution in Europe and the United States 120 Notes 257

Global wealth inequality: Capital is back 127 References 268

Afterword: Data transparency as a global imperative 132

Contents | vii STATISTICAL ANNEX 5.1 Household income, inequality and greenhouse gas emissions 175

Readers guide 295 5.2 From Holocene to Anthropocene: Power—and who wields it—at the brink

Statistical tables of a new era 177

1. Human Development Index and its components 300 5.3 When history is no longer a good guide 187

2. Human Development Index trends, 1990­2018 304 5.4 The impacts of a global dietary shift on sustainable human development 189

3. Inequality-adjusted Human Development Index 308 6.1 Mobile technology promotes financial inclusion 203

4. Gender Development Index 312 6.2 Digital technologies for the Sustainable Development Goals: Creating the

5. Gender Inequality Index 316 right conditions 209

6. Multidimensional Poverty Index: developing countries 320 6.3 Artificial intelligence and the risk of bias: Making horizontal inequalities worse? 212

Human development dashboards 6.4 The United Kingdom’s Data Ethics Framework principles 213

1. Quality of human development 323 6.5 Intellectual property rights, innovation and technology diffusion 217

2. Life-course gender gap 328 7.1 Enhancing capabilities in China: Tackling inequality at its roots 227

3. Women’s empowerment 333 7.2 Unlocking the potential of preprimary education for advancing human

4. Environmental sustainability 338 development in Ethiopia 227

5. Socioeconomic sustainability 343 7.3 The persistence of health gradients even with universal health coverage 228

Developing regions 348 7.4 Girls’ coding choices and opportunities 230

Statistical references 349 7.5 Gender equality in the labour market 235

7.6 How market concentration can disproportionately affect poor people 240

SPECIAL CONTRIBUTION 7.7 The power of fiscal redistribution 241

S7.1.1 Being right is not enough: Reducing inequality needs a movement from below 248

A new look at inequality—Michelle Bachelet 25

FIGURES

BOXES 1 The share of the population stating that income should be more equal

1 A new take on the Great Gatsby Curve 11 increased from the 2000s to the 2010s 2

I.1 The capabilities approach and the 2030 Agenda for Sustainable Development 25 2 Children born in 2000 in countries with different incomes will have very

1.1 Inequality of capabilities 31 unequal paths to 2020 2

1.2 Article 25 of the Universal Declaration of Human Rights: The right to a basic 3 Beyond income, beyond averages and beyond today: Exploring inequalities in

standard of living 37 human development leads to five key messages 3

1.3 Inequality in healthy life expectancy 38 4 Thinking about inequalities 5

1.4 Divergence in life expectancy at older ages in Chile 43 5 Human development, from basic to enhanced capabilities 6

1.5 Crises and divergence 52 6 Across countries the world remains deeply unequal in both basic and

1.6 Social exclusion of lesbian, gay, bisexual, trans and intersex people 54 enhanced capabilities 8

1.7 Inequality in human security in Japan: The role of dignity 55 7 Slow convergence in basic capabilities, rapid divergence in enhanced ones 9

1.8 Horizontal inequalities in India: Difference dynamics in basic and enhanced 8 Education and health along the lifecycle 10

capabilities 56 9 Inequalities, power asymmetries and the effectiveness of governance 12

1.9 A social-psychological perspective on inequality 58 10 Bias against gender equality is on the rise: The share of women and men worldwide with no gender social norms bias fell between 2009 and 2014 13 S1.3.1 Income poverty reduction scenarios to 2030 67

2.1 Key competencies of social and emotional learning 79 11 Between 1980 and 2017 post-tax incomes grew close to 40 percent for the

2.2 How perceived relative deprivations affect health outcomes 80 poorest 80 percent of the European population, compared with more than

180 percent for the top 0.001 percent 14

2.3 The power of perceived inequalities in South Africa 86 12 A framework for designing policies to redress inequalities in human development 15

2.4 The power of your neighbour 87 13 Redistributive direct taxes and transfers explain nearly all the difference in disposable income inequality between advanced and emerging economies 16 2.5 Economic inequality and human development 89

2.6 Internal armed conflict and horizontal inequalities 92 14 Strategies for practical universalism in unequal developing countries 16

3.1 Investigative journalism uncovering inequality 106 15 Ecological footprints expand with human development 18

3.2 What income concepts are we measuring? 109 16 Technology can displace some tasks but also create new ones 19

3.3 What about consumption? 110 I.1 The share of the population stating that income should be more equal

3.4 Where do you stand in the global distribution of income? 114 increased from the 2000s to the 2010s 23

3.5 Income growth of the bottom 40 percent—higher than the national average? 119 1.1 Children born in 2000 in countries with different incomes will have severely

4.1 Practical and strategic gender interests and needs 151 different capabilities by 2020 29

4.2 Overlapping and intersecting identities 153 1.2 Still massive inequality in human development across the world, 2017 30

4.3 The multidimensional gender social norms index—measuring biases, 1.20 School dropout rates converge with human development, but not for the

prejudices and beliefs 155 poorest 20 percent 51

4.4 The man box 159 1.3 Human development, from basic to enhanced capabilities 33

4.5 Climate change and gender inequality 163 1.4 The world remains deeply unequal in key areas of human development in

4.6 Better data are needed on gender inequalities 165 both basic and enhanced capabilities 34

viii | HUMAN DEVELOPMENT REPORT 2019 1.5 In all regions of the world the loss in human development due to inequality 3.14 The average pretax income of the top 10 percent in the United States was

is diminishing, reflecting progress in basic capabilities 35 about 11 times higher than that of the bottom 40 percent in 1980 and 27

1.6 Convergence in basic capabilities, divergence in enhanced capabilities 36 times higher in 2017, while in Europe the ratio rose from 10 to 12 127

1.7 Inequalities persist in life expectancy and mortality 38 3.15 Between 1981 and 2017 the average top corporate tax rate in the European

1.8 The changing inequality in life expectancy, 2005­2015: Low human Union fell from about 50 percent to 25 percent, while the average value

development countries catching up in life expectancy at birth but lagging added tax rate rose from about 18 percent to more than 21 percent 127

behind in life expectancy at older age 40 3.16 Net private wealth in Western European countries rose from 250­400 percent of national income in 1970 to 450­750 percent in 2016 129 1.9 Infant mortality rates, an important determinant of life expectancy at birth,

have been declining everywhere, but significant gradients remain 41 3.17 Countries are getting richer, but governments are becoming poor 130

1.10 Mortality: Convergence in basic capabilities, divergence in enhanced capabilities 41 3.18 Trends in wealth inequality 132

1.11 The lower a country’s human development, the larger the gap in access to 3.19 If current trends continue, by 2050 the global top 0.1 percent could end up

education 44 owning as much of the world’s wealth as the middle 40 percent of the

1.12 Gaps in access to education among children and youth are also large within world’s population 133

countries 44 3.2 Income inequality based on the top 10 percent’s income share has risen

1.13 Inequality in primary and secondary education has been falling over the past since 1980 in most regions but at different rates 111

decade 45 3.3 The elephant curve of global inequality and growth 112

1.14 Dynamics of education attainment, 2007­2017 46 3.4 In 2010 the top 10 percent of income earners received 53 percent of global income, but if there had been perfect equality in average income between 1.15 Inequalities in postsecondary education within countries are growing 47 countries, the top 10 percent would have received 48 percent of global income 113

1.16 Widening inequalities in the availability of physicians between countries 48 3.5 The ratio of the average income of the top 10 percent to that of the middle

1.17 Harmonized test scores across human development groups 48 40 percent increased by 20 percentage points between 1980 and 2016, but

1.18 Child mortality converges with human development, but not for the poorest the ratio of the average income of the middle 40 percent to that of the

20 percent 50 bottom 50 percent decreased by 27 percentage points 114

1.19 Some 846,000 of 3.1 million child deaths are preventable if the bottom 3.6 The geographic breakdown of each percentile of the global distribution of

20 percent converge to the country average 50 income evolved from 1990 to 2016 115

S1.1.1 Description of the stages in the development of the historical market economies 61 3.7 Between 1995 and 2015 the income share of the top 10 percent in North

S1.1.2 Linking the hazard of high water to flood disasters: Economic and political Africa and West Africa remained relatively stable, while the share of the

equality enhances the chance of institutions becoming adjusted to bottom 40 percent in Southern Africa declined 117

circumstances and preventing disaster 62 3.8 The income share of the top 1 percent has significantly increased in China,

S1.1.3 Sub-Saharan countries have the most overlapping deprivations 69 India and the Russian Federation since the early 1980s 120

S1.2.1 Transmitting inequalities in human development across the lifecycle 65 3.9 The pretax income share of the top 10 percent in the United States rose

S1.2.2 Distribution of subjective well-being across the world (measured by people’s from around 35 percent in 1980 to close to 47 percent in 2014 123

overall satisfaction with their lives) 66 S3.1.1 Contiguous human development patterns, cutting across national borders:

S1.3.1 Some 600 million people live below the $1.90 a day poverty line 68 The Gulf of Guinea 134

S1.3.2 Poverty at the $1.90 a day level is tied to multidimensional poverty 69 S3.1.1 Lorenz curve 136

2.1 Intergenerational mobility in income is lower in countries with more S3.1.2 Adult female malnutrition and child stunting can be high in nonpoor

inequality in human development 74 households 135

2.2 Education and health along the lifecycle 76 4.1 Remarkable progress in basic capabilities, much less in enhanced capabilities 147

2.3 Intergenerational persistence of education is higher in countries with higher 4.10 Countries with higher social norms biases tend to have higher gender inequality 157

inequality in human development 76 4.11 Biases in social norms show a gradient 158

2.4 Skill gaps emerge in early childhood, given parents’ education 77 4.12 Contraceptive use is higher among unmarried and sexually active adolescent girls, but so is the unmet need for family planning, 2002­2014 160 2.5 Socioeconomic status affects specific areas of health later in the lifecycle 81

2.6 The hollowing out of the middle in South Africa 83 4.13 The gap in unpaid care work persists in developing economies 161

2.7 The effectiveness of governance: An infinity loop 90 4.14 A large proportion of employed women believe that choosing work implies

3.1 Dozens of countries have almost no transparency in inequality data 105 suffering for their children, while a large proportion of female homemakers

3.10 Between 1980 and 2017 the share of post-tax national income received by feel that by staying home they are giving up a career or economic

the top 10 percent rose from 21 percent to 25 percent in Northern Europe, independence, 2010­2014 162

while the share received by the bottom 40 percent fell from 24 percent to 4.15 The percentage of women with an account at a financial institution or with

22 percent 124 a mobile money-service provider is below 80 percent in all developing

3.11 Between 1980 and 2017 the post-tax incomes of the poorest 80 percent of country regions in 2018 163

the European population grew close to 40 percent, while those of the top 4.16 Girls and women of reproductive age are more likely to live in poor

0.001 percent grew more than 180 percent 125 households than boys and men 164

3.12 Between 1980 and 2017 the pretax income share of the bottom 40 percent 4.2 Gender inequality is correlated with a loss in human development due to

in the United States fell from about 13 percent to 8 percent, while the share inequality 149

of the top 1 percent rose from about 11 percent to 20 percent 126 4.3 Progress towards gender equality is slowing 150

3.13 Between 1980 and 2017 the average pretax income of the bottom 4.4 The greater the empowerment, the wider the gender gap 151

40 percent grew 36 percent in Europe, while it declined 3 percent in 4.5 The percentage of informal employment in nonagricultural employment in

the United States 126 developing countries is generally higher for women than for men 152

4.6 How social beliefs can obstruct gender and women’s empowerment 154

Contents | ix 4.7 Only 14 percent of women and 10 percent of men worldwide have no 7.2 Higher labour productivity is associated with a lower concentration of

gender social norms biases 156 labour income at the top 233

4.8 The share of both women and men worldwide with no gender social norms 7.3 The relationship between labour productivity and concentration of labour income appears to hold over time at most levels of human development 233 bias fell between 2005­2009 and 2010­2014 156

4.9 Progress in the share of men with no gender social norm bias from 7.4 Minimum wage: a tool to share the fruit of progress? 236

2005­2009 to 2010­2014 was largest in Chile, Australia, the United States 7.5 Unpaid family workers, industrial outworkers, home workers and casual

and the Netherlands, while most countries showed a backlash in the share workers are predominantly women with low earnings and a high risk of

of women with no gender social norms bias 157 poverty, while employees and regular informal workers with higher

S4.1.1 About a third of women ages 15 and older have experienced physical or wages and less risk of poverty are more often men 237

sexual violence inflicted by an intimate partner, 2010 166 7.6 The rising market power of firms in recent decades has been led by firms

S4.1.2 Female members of European parliaments experience high rates of acts of at the top 10 percent of the markup distribution 238

political violence against women, 2018 167 7.7 Top personal income tax rates have declined around the world 242

S4.1.3 Traditional social norms encourage different forms of violence against 7.8 Offshore wealth is bigger than the value of top corporations or of billionaires 244

women 168 S7.1.1 Strategies for practical universalism in (unequal) developing countries 246

5.1 Per capita ecological footprints increase with human development 176 S7.1.2 Power of the economic elite and action mechanisms 247

5.10 Richer countries generate more waste per capita 188 S7.3.1 Fiscal redistribution in European countries, 2016 251

5.11 Developing countries will drive most of the rise in meat production to 2030 190 S7.3.2 Fiscal progressivity and fiscal effort in European countries, 2016 252

5.12 In some countries basic water and sanitation coverage for the wealthiest S7.3.3 Market income inequality and variation in fiscal redistribution 252

quintile is at least twice that for the poorest quintile 192

5.2 Today’s developed countries are responsible for the vast majority of SPOTLIGHTS

cumulative carbon dioxide emissions 179

5.3 Of the top 10 percent of global emitters of carbon dioxide equivalent 1.1 Power concentration and state capture: Insights from history on consequences

emissions, 40 percent are in North America, and 19 percent are in the of market dominance for inequality and environmental calamities 60

European Union 179 1.2 Rising subjective perceptions of inequality, rising inequalities in perceived

5.4 Within-country inequality in carbon dioxide equivalent emissions is now as well-being 64

important as between-country inequality in driving the global dispersion of 1.3 The bottom of the distribution: The challenge of eradicating income poverty 67

carbon dioxide equivalent emissions 180 3.1 Looking within countries and within households 134

5.5 Economic damages from extreme natural hazards have been increasing 181 3.2 Choosing an inequality index 136

5.6 Human development crises are more frequent and deeper in developing 3.3 Measuring fiscal redistribution: concepts and definitions 139

countries 184 4.1 Women’s unequal access to physical security—and thus to social and

5.7 The lower the level of human development, the more deadly the disasters 185 political empowerment 166

5.8 In El Salvador and Honduras people in the lower quintiles of the income 5.1 Measuring climate change impacts: Beyond national averages 194

distribution were more likely to be affected by floods and landslides 185 5.2 Climate vulnerability 195

5.9 Fewer deaths in the 2000s than in the 1960s and 1970s despite more 7.1 Addressing constraints in social choice 246

occurrences of natural disasters 186 7.2 Productivity and equity while ensuring environmental sustainability 249

6.1 Digital divides: Groups with higher development have greater access, and 7.3 Variation in the redistributive impact of direct taxes and transfers in Europe 251

inequalities are greater for advanced technologies, 2017 202

6.10 Income and productivity are strongly correlated, and the higher the

productivity, the greater the share of productivity that the median worker TABLES

receives as compensation 218 1.1 Limited convergence in health and education, 2007­2017 49

6.11 A significant decoupling of emissions from development has allowed some S1.1.1 Certain and possible cases of market economies 60

countries to reduce their carbon dioxide emissions, reflecting more efficient 3.1 Main data sources for inequality measurement 107

forms of production 219

6.2 Dynamics of access to technology 204 3.2 Difference between income growth of the bottom 40 percent and average income growth in Africa’s five subregions, 1995­2015 (percentage points) 118 6.3 The bandwidth gap between high-income and other countries fell from 3.3 Difference between income growth of the bottom 40 percent and average 22-fold to 3-fold 205 income growth in selected African countries, 1995­2015 (percentage points) 118

6.4 The distribution of mobile subscriptions is converging to the distribution of 3.4 Inequality and growth in the BRIC countries 120

population by region, but installed bandwidth potential is not 206

6.5 From 1987 to 2007 little changed in the global ranking of installed 3.5 Post-tax average and bottom 40 percent growth in Europe and the

bandwidth potential, but at the turn of the millennium things started to United States, 1980­2017 and 2007­2017 121

change, with the expansion of bandwidth in East and North Asia 206 S3.1.1 Targeting errors of inclusion and exclusion: Proxy means tests 135

6.6 Market power is on the rise, particularly for firms intensive in information S3.2.1 Statistics most frequently published in 10 commonly used international

and communication technology 208 databases 138

6.7 Technology can displace some tasks but also reinstate new ones 210 S3.3.1 Comparison of income concepts in databases with fiscal redistribution

6.8 Workers in medium and high wage jobs are more likely to participate in indicators 141

adult learning 214 4.1 Gender Inequality Index: Regional dashboard 149

6.9 There are enormous asymmetries in research and development across 6.1 Different tasks have different potential for being replaced by artificial

human development groups 217 intelligence 211

7.1 A framework for designing policies to redress inequalities in human development 224

x | HUMAN DEVELOPMENT REPORT 2019

Inequalities in human development in the 21st century Inequalities in human development

In every country many people have little prospect for a better future. Lacking hope, purpose or dignity, they watch from society’s sidelines as they see others pull ahead to ever greater prosperity. Worldwide many have escaped extreme poverty, but even more have neither the opportunities nor the resources to control their lives. Far too often gender, ethnicity or parents’ wealth still determines a person’s place in society.

Inequalities. The evidence is everywhere. So people’s sense of fairness and can be an affront is the concern. People across the world, of all to human dignity. political persuasions, increasingly believe that income inequality in their country should be Such inequalities in human development reduced (figure 1). hurt societies, weakening social cohesion and people’s trust in government, institutions and Inequalities in human development are more each other. Most hurt economies, wastefully profound. Consider two children born in preventing people from reaching their full po- 2000, one in a very high human development tential at work and in life. They often make it country, the other in a low human development harder for political decisions to reflect the as- country (figure 2). Today the first has a more pirations of the whole of society and to protect than 50-50 chance of being enrolled in higher the planet, as the few pulling ahead flex their education: More than half of 20-year-olds in power to shape decisions primarily in their very high human development countries are interests today. In the extreme, people can take in higher education. In contrast, the second is to the streets. much less likely to be alive. Some 17 percent of children born in low human development These inequalities in human development countries in 2000 will have died before age 20, are a roadblock to achieving the 2030 Agenda compared with just 1 percent of children born for Sustainable Development.4 They are not in very high human development countries. just about disparities in income and wealth. The second child is also unlikely to be in higher They cannot be accounted for simply by using education: In low human development coun- summary measures of inequality that focus on tries only 3 percent are. Circumstances almost a single dimension.5 And they will shape the entirely beyond their control have already set prospects of people that may live to see the them on different and unequal—and likely 22nd century. Exploring inequalities in human irreversible—paths.1 The inequalities are like- development thus has to go beyond income, wise high within countries—both developing beyond averages and beyond today, leading to and developed. In some developed countries five key messages (figure 3). the gaps in life expectancy at age 40 between the top 1 percent of the income distribution First, while many people are stepping and the bottom 1 percent have been estimated above minimum floors of achievement in to be as high as 15 years for men and 10 years human development, widespread dispar- for women.2 ities remain. The first two decades of the 21st century have seen remarkable progress Inequalities do not always reflect an unfair in reducing extreme deprivations, but gaps world. Some are probably inevitable, such as remain unacceptably wide for a range of the inequalities from diffusing a new tech- capabilities—the freedoms for people to be nology.3 But when these unequal paths have and do desirable things such as go to school, little to do with rewarding effort, talent or get a job or have enough to eat. And progress entrepreneurial risk-taking, they may offend is bypassing some of the most vulnerable even on the most extreme deprivations—so much

Overview | 1 FIGURE 1

The share of the population stating that income should be more equal increased from the 2000s to the 2010s

Change in the share of 35 out 33 out 32 out population stating that of 39 of 39 of 39 income should be more countries countries countries

0

­40

Leaning left Center Leaning right

Population in selected countries by political self-identification

Note: Each dot represents one of 39 countries with comparable data. The sample covers 48 percent of the global population. Based on answers on a 1­5 scale, where 1 is “income should be more equal” and 5 is “we need larger income differences.” Source: Human Development Report Office calculations based on data from the World Values Survey, waves 4, 5 and 6.

FIGURE 2

Children born in 2000 in countries with different incomes will have very unequal paths to 2020

In higher Children born education 55 in 2000 in very high 3 human development

Children born Not in development countries 80 higher

Note: These are estimates (using median values) for a typical individual from a country with low human development and from a country with very high human development. Data for participation in higher education are based on household survey data for people ages 18­22, processed by the United Nations Educational, Scientific and Cultural Organization Institute for Statistics in www.education-inequalities.org (accessed 5 November 2019). Percentages are with respect to people born in 2000. People that died before age 20 are computed based on births around 2000 and estimated deaths for that cohort between 2000 and 2020. People in higher education in 2020 are computed based on people estimated to be alive (from cohort born around 2000), and the latest data of participation in higher education. People not in higher education are the complement. Source: Human Development Report Office calculations based on data from the United Nations Department of Economic and Social Affairs and the United Nations Educational, Scientific and Cultural Organization Institute for Statistics.

2 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 3 Beyond income, beyond averages and beyond today: Exploring inequalities in human development leads to five key messages

Exploring inequalities in five key messages

Disparities in human A new generation Inequalities accumulate Assessing and responding We can redress development remain of inequalities is through life, often to inequalities in human inequalities if we act widespread, despite emerging, with divergence reflecting deep development demands now, before imbalances achievements in reducing in enhanced capabilities, power imbalances a revolution in metrics in economic power are extreme deprivations despite convergence politically entrenched

so that the world is not on track to eradicate Third, inequalities in human development them by 2030, as called for in the Sustainable can accumulate through life, frequently Development Goals. heightened by deep power imbalances. They are not so much a cause of unfairness as a con- Second, a new generation of severe inequal- sequence, driven by factors deeply embedded ities in human development is emerging, even in societies, economies and political structures. if many of the unresolved inequalities of the Tackling inequalities in human development 20th century are declining. Under the shadow means addressing these factors: Genuine im- of the climate crisis and sweeping technological provement will not come from trying to fix dis- change, inequalities in human development parities only when people are already earning are taking new forms in the 21st century. very different incomes—because inequalities Inequalities in capabilities are evolving in dif- start at birth, often even before, and can ac- ferent ways. Inequalities in basic capabilities— cumulate over people’s lives. Or from looking linked to the most extreme deprivations—are back and simply trying to reinstate the policies shrinking. In some cases, quite dramatically, and institutions that held inequalities in check, such as global inequalities in life expectancy at times and in some countries, during the 20th at birth. Many people at the bottom are now century. It was under those very conditions that reaching the initial stepping stones of human power imbalances deepened, in many cases ac- development. At the same time, inequalities centuating the accumulation of advantage over are increasing in enhanced capabilities—which the lifecycle. reflect aspects of life likely to become more im- portant in the future, because they will be more Fourth, assessing inequalities in hu- empowering. People well empowered today man development demands a revolution appear set to get even farther ahead tomorrow. in metrics. Good policies start with good

Overview | 3 measurement, and a new generation of ine- This is not to say that redistribution does not qualities requires a new generation of measure- matter—quite the opposite. But long-lasting ment. Clearer concepts tied to the challenges change in both income and the broader range of current times, broader combinations of data of inequalities in human development depends sources, sharper analytical tools—all are need- on a wider and more systemic approach to ed. Ongoing innovative work suggests that policies. income and wealth may be accumulating at the top in many countries much faster than What to do? The approach proposed in this one could grasp based on summary measures Report outlines policies to redress inequalities of inequality. Making these efforts more in human development within a framework systematic and widespread can better inform that links the expansion and distribution of public debates and policies. Metrics may not both capabilities and income. The options span seem a priority, until one considers the contin- premarket, in-market and postmarket policies. uing hold of such measures as gross domestic Wages, profits and labour participation rates product since its creation in the first half of the are typically determined in markets, which 20th century. are conditioned by prevailing regulations, in- stitutions and policies (in-market). But those Fifth, redressing inequalities in human de- outcomes also depend on policies that affect velopment in the 21st century is possible—if people before they become active in the econo- we act now, before imbalances in economic my (premarket). Premarket policies can reduce power translate into entrenched political disparities in capabilities, helping everyone dominance. Improvements in inequality for enter the labour market better equipped. In- some basic capabilities show that progress is market policies affect the distribution of in- possible. But the record of progress in basic come and opportunities when individuals are capabilities in the past will not respond to peo- working, shaping outcomes that can be either ple’s aspirations for this century. And doubling more or less equalizing.7 Postmarket policies down on reducing inequalities in basic capabil- affect inequalities once the market along with ities further, while needed, is not enough. If en- the in-market policies have determined the hanced capabilities are indeed associated with distribution of income and opportunities. more empowerment, ignoring the gaps that are These sets of policies interact. For instance, opening up in them can alienate policymakers the provision of public services premarket may from people’s agency—their ability to make depend in part on the effectiveness of postmar- choices that fulfil their aspirations and values. ket policies (taxes on market income to fund Only by turning attention towards tackling a health and education, for instance), which new generation of inequality in enhanced capa- matter in mobilizing government revenue to bilities, many of which are only just beginning pay for those services. And taxes, in turn, are to emerge, will it be possible to avoid further informed by how much society is willing to entrenchment of inequalities in human devel- redistribute income from those with more to opment over the course of the 21st century. those with less.

How? Not by looking at policies in isolation The future of inequalities in human devel- or thinking that a single silver bullet will solve opment in the 21st century is in our hands. everything. The redistribution of income, But we cannot be complacent. The climate which often dominates the policy debate on in- crisis shows that the price of inaction com- equality, is sometimes seen as that silver bullet. pounds over time, as it feeds further ine- Yet, even a full redistributive package of four quality, which can in turn make action on ambitious policies—higher and more progres- climate more difficult. Technology is already sive income taxes, earned income discounts at changing labour markets and lives, but not low income levels, taxable benefits paid out for yet locked-in is the extent to which machines each child and a minimum income for all indi- may replace people. We are, however, ap- viduals—would be insufficient to fully reverse proaching a precipice beyond which it will be the increase in income inequality in the United difficult to recover. We do have a choice, and Kingdom between the late 1970s and 2013.6 we must exercise it now.

4 | HUMAN DEVELOPMENT REPORT 2019 Beyond income, beyond Even understanding income disparities averages and beyond today requires examining other forms of inequality. Disadvantages in health and education (of This Report builds on a new framework of one’s parents and one’s own) interact and often analysis that looks at inequalities by going compound over a lifetime. Gaps open before beyond income, beyond averages and beyond birth, starting with the “birth lottery” of where today (figure 4). children happen to be born, and can widen over the years. Children from poor families Beyond income may not be able to afford an education and are at a disadvantage when they try to find work. Any comprehensive assessment of inequality These children are likely to earn less than those must consider income and wealth. But it must in higher income families when they enter the also go beyond dollars and rupees to under- labour market, when penalized by compound- stand differences in other aspects of human ing layers of disadvantage. development and the processes that lead to them. There is economic inequality, of course, Beyond averages but there are also inequalities in key elements of human development such as health, edu- Too often the debate about inequality is over- cation, dignity and respect for human rights. simplified, relying on summary measures of And these might not be revealed by consid- inequality and incomplete data that provide a ering income and wealth inequality alone. A partial—sometimes misleading—picture, both human development approach to inequality in the sorts of inequality to consider and the takes a people-centred view: It is about peo- people affected. The analysis must go beyond ple’s capabilities to exercise their freedoms to averages that collapse information on distribu- be and do what they aspire to in life. tion to a single number and look at the ways

FIGURE 4

Thinking about inequalities

A comprehensive assessment of inequality Beyond income must consider income and wealth. But it must also understand differences in other aspects of human development and the processes that lead to them.

Exploring inequalities Beyond averages The analysis of inequalities in in human development: human development must go beyond summary measures of inequality a new framework that focus on only a single dimension.

Inequalities in human development Beyond today will shape the prospects of people that may live to see the 22nd century.

Overview | 5 inequality plays out across an entire popula- capabilities—their freedoms to make life choic- tion, in different places and over time. For every es—are fundamental.9 Capabilities are at the aspect of human development, what matters is heart of human development. This Report the entire inequality gradient (the differences follows the same path and explores inequalities in achievements across the population accord- in capabilities. ing to different socioeconomic characteristics). Capabilities evolve with circumstances as Beyond today well as with values and with people’s changing demands and aspirations. Today, having a set of Much analysis focuses on the past or on the basic capabilities—those associated with the ab- here and now. But a changing world requires sence of extreme deprivations—is not enough. considering what will shape inequality in the Enhanced capabilities are becoming crucial for future. Existing—and new—forms of inequal- people to own the “narrative of their lives.”10 ity will interact with major social, economic and environmental forces to determine the Enhanced capabilities bring greater agency lives of today’s young people and their children. along people’s lives. Given that some capabili- Two seismic shifts will shape the 21st century: ties build over a person’s life, achieving a basic Climate change and technological transforma- set—such as surviving to age 5 or learning to tions. The climate crisis is already hitting the read—provides initial stepping stones to form- poorest hardest, while technological advances ing enhanced capabilities later in life (figure 5). such as machine learning and artificial intelli- gence can leave behind entire groups of people, A similar evolution from basic to enhanced even countries—creating the spectre of an un- capabilities is reflected in the use of technology certain future under these shifts.8 or in the ability to cope with environmental shocks, from frequent but low-impact hazards Evolving human aspirations: From to large and unpredictable events. The distinc- basic to enhanced capabilities tion is also important when it comes to under- standing inequalities across groups, such as the When Amartya Sen asked what kind of progression from women being able to vote in inequality we should ultimately care about elections (a basic capability) to participating in (“Equality of what?”), he argued that people’s politics as national leaders (an enhanced capa- bility). The evolution in ambition from basic to enhanced capabilities mirrors the evolution from the Millennium Development Goals to the Sustainable Development Goals.

FIGURE 5 Human development, from basic to enhanced capabilities

Enhanced Examples of achievements

• Access to quality health at all levels
• High-quality education at all levels
• Effective access to present-day technologies
• Resilience to unknown new shocks

Examples of achievements Basic

• Early childhood survival capabilities
• Primary education
• Entry-level technology
• Resilience to recurrent shocks

6 | HUMAN DEVELOPMENT REPORT 2019 Key message 1: Disparities treatment, child mortality rates in the poorest in human development households in the world’s poorest countries remain widespread, despite remain high. The highest rates are in low and achievements in reducing medium human development countries, but extreme deprivations there are vast disparities within countries: The poorest 20 percent in some middle-income The 21st century has witnessed great progress countries can have the same average mortality in living standards, with an unprecedented rate as children from a typical low-income number of people around the world making country. a “great escape”11 from hunger, disease and poverty—moving above minimum subsist- Key message 2: A new ence. The Human Development Index shows generation of inequalities is impressive improvement on average, reflecting emerging, with divergence in dramatic improvements in achievements such enhanced capabilities, despite as life expectancy at birth, driven largely by convergence in basic capabilities sharp declines in infant mortality rates. As we enter the 2020s, a new set of capabilities Still, many people have been left behind, is becoming fundamental to 21st century life. and inequalities remain widespread across all Inequalities in these enhanced capabilities capabilities. Some refer to life and death, oth- show strikingly different dynamics from those ers to access to knowledge and life-changing in basic capabilities. They are at the root of a technologies. new generation of inequalities.

Despite having shrunk considerably, the Inequalities for some basic capabilities are difference in life expectancy at birth between slowly narrowing across most countries, even low and very high human development coun- if much remains to be done. Life expectancy tries is still 19 years. There are differences in at birth, percentage of the population with expected longevity at every age. The differ- a primary education and mobile-cellular ence in life expectancy at age 70 is almost 5 subscriptions all show narrowing inequalities years. Some 42 percent of adults in low hu- across human development groups (figure 7). man development countries have a primary The people at the bottom are progressing education, compared with 94 percent in very faster than those at the top. The gain in life high human development countries. There expectancy at birth between 2005 and 2015 for are gaps at all education levels. Only 3.2 per- low human development countries was almost cent of adults in low human development three times that for very high human develop- countries have a tertiary education, compared ment countries, driven by a reduction in child with 29 percent in developed countries. In mortality rates in developing countries. And access to technology developing countries countries with lower human development are have 67 mobile phone subscriptions per 100 catching up in access to primary education and inhabitants, half the number in very high access to mobile phones. human development countries. For access to broadband, low human development This good news comes with two caveats. countries have less than 1 subscription per First, despite progress, the world is not on track 100 inhabitants, compared with 28 per 100 to eradicate extreme deprivations in health and inhabitants in very high human development education by 2030, when 3 million children countries (figure 6). under age 5 are still expected to die every year (at least 850,000 above the Sustainable The furthest behind include the 600 million Development Goal target), and 225 million people still living in extreme income poverty— children are expected to be out of school. and that jumps to 1.3 billion when measured by Second, gaps are falling in part because those the Multidimensional Poverty Index.12 Some at the top have little space to keep moving up. 262 million children are out of primary or secondary school, and 5.4 million children do In contrast, inequalities in enhanced capa- not survive their first five years of life. Despite bilities are widening. For instance, despite data greater access to immunizations and affordable

Overview | 7 FIGURE 6 Across countries the world remains deeply unequal in both basic and enhanced capabilities

Basic Enhanced

Life expectancy at birth, 2015 Life expectancy at age 70, 2015

72.9 78.4 14.6 59.4 66.6 Health 11.1

Low Medium High Very high Low Medium High Very high Human development group Human development group

Population with a primary education, 2017 Population with a tertiary education, 2017 (percent) (percent)

93.5 28.6 66.5 13.7

Low Medium High Very high 3.2

Mobile-cellular subscriptions, 2017 Fixed broadband subscriptions, 2017 (per 100 inhabitants) (per 100 inhabitants)

90.6 Access to 67.0 technology

Low Medium High Very high Low Medium High Very high Human development group Human development group

Source: Human Development Report Office calculations based on data from the International Telecommunication Union, the United Nations Educational, Scientific and Cultural Organization Institute for Statistics and the United Nations Department of Economic and Social Affairs.

8 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 7 Slow convergence in basic capabilities, rapid divergence in enhanced ones

Basic Enhanced

Declining inequality Increasing inequality Change between 2005 and 2015 (years) Life expectancy at age 70 5.9 Change between 2005 and 2015 (years)

4.9 1.2

2.7 2.4 0.5

Low Medium High Very high Low Medium High Very high Human development group Human development group

Share of the population with a primary education Share of the population with a tertiary education Change between 2007 and 2017 (percentage points) Change between 2007 and 2017 (percentage points)

5.3 Education

3.0 1.1 Human development group Human development group

Mobile-cellular subscriptions Fixed broadband subscriptions Change between 2007 and 2017 (per 100 inhabitants) Change between 2007 and 2017 (per 100 inhabitants)

59.5 12.3 49.3 49.3 Access to 26.1 Low Medium High Very high

Low Medium High Very high Human development group

Source: Human Development Report Office calculations based on data from the International Telecommunication Union, the United Nations Educational, Scientific and Cultural Organization Institute for Statistics and the United Nations Department of Economic and Social Affairs.

Overview | 9 challenges, estimates suggest that the gain in Key message 3: Inequalities life expectancy at age 70 from 1995 to 2015 in accumulate through life, very high human development countries was often reflecting deep more than twice that in low human develop- power imbalances ment countries.13 Understanding inequality—even income ine- There is evidence for the same pattern of quality—means homing in on the underlying divergence across a wide range of enhanced processes that lead to it. Different inequalities capabilities. Indeed, divergences in access to interact, while their size and impact shift over more advanced knowledge and technology a person’s lifetime. The corollary is that policies are even starker. The proportion of the adult to tackle economic inequality require much population with tertiary education is growing more than a mechanistic transfer of income. more than six times faster in very high human They often need to address social norms, poli- development countries than in low human cies and institutions formed deep in history. development countries, and fixed broadband subscriptions are growing 15 times faster. Lifelong disadvantage

These new inequalities—both between and Inequalities can start before birth, and many within countries—are hugely consequential. of the gaps may compound over a person’s Shaping 21st century societies, they are pushing life. When that happens, it can lead to persis- the frontiers in health and longevity, knowl- tent inequalities. This can happen in several edge and technology. These are the inequalities ways, especially in the nexus among health, that will likely determine people’s ability to education and parents’ socioeconomic status seize the opportunities of the 21st century, (figure 8). function in a knowledge economy and cope with climate change.

FIGURE 8

Education and health along the lifecycle

Child’s Early Assortative health childhood mating health Education

Note: The circles represent different stages of the lifecycle, with the orange ones resenting final outcomes. The rectangle represents the process of assortative mating. The dashed lines refer to interactions that are not described in detail. A child’s health affects early childhood development and prospects for education. For example, an intellectually disabled child will not be able to benefit from early childhood development and education opportunities in the same way as a healthy child. Education can also promote a healthy lifestyle and convey information on how to benefit from a given health care system if needed (Cutler and Lleras-Muney 2010). Source: Human Development Report Office, adapted from Deaton (2013a).

10 | HUMAN DEVELOPMENT REPORT 2019 Parents’ incomes and circumstances affect BOX 1 their children’s health, education and incomes. A new take on the Great Gatsby Curve Health gradients—the disparities in health across socioeconomic groups—often start The positive correlation between higher income inequality and lower intergenerational mobil- before birth and can accumulate at least up ity in income is well known. This relation, known as the Great Gatsby Curve, also holds true to adulthood, if not counteracted. Children using a measure of inequality in human development instead of income inequality alone (see born to low-income families are more prone to figure). The greater the inequality in human development, the lower the intergenerational poor health and lower education. Those with mobility in income—and vice versa. lower education are less likely to earn as much as others, while children in poorer health are These two factors go hand in hand, but that does not imply that one causes the other. more likely to miss school. And when children In fact, it is more likely that both are driven by underlying economic and social factors, so grow up, if they partner with someone who has understanding and tackling these drivers could both promote mobility and redress inequality. similar socioeconomic status (as often happens in assortative mating), inequalities across gen- Intergenerational mobility in income is lower in countries with more inequality in erations can persist. human development

The cycle can be difficult to break, not least Intergenerational Colombia because of the ways in which inequality in income elasticity Rwanda income and political power co-evolve. When India wealthy people shape policies that favour them- 1.2 selves and their children—as they often do— that can sustain the accumulation of income Ecuador and opportunity at the top. Unsurprising, then, 1.0 that social mobility tends to be lower in more unequal societies. Still, some societies have Latvia more mobility than others—so institutions and 0.8 policies matter—in part because what tends to reduce inequality can also boost social mobility Albania (box 1). Slovakia Power imbalances 0.4 Pakistan

China Ethiopia

0.2 Singapore

0 Finland

0 10 20 30 40

Inequality in human development (percent)

Income and wealth inequalities are often trans- Note: Inequality in human development is measured as the percentage loss in Human Development Index value due to inequality lated into political inequality, in part because in three components: income, education and health. The higher the intergenerational income elasticity, the stronger the association inequalities depress political participation, between parents’ income and their children’s income, reflecting lower intergenerational mobility. giving more space to particular interest groups Source: Human Development Report Office using data from GDIM (2018), adapted from Corak (2013). to shape decisions in their favour. Those priv- ileged can capture the system, moulding it to (the exchange of political support for personal fit their preferences, potentially leading to even gain), people tend to withdraw from political more inequalities. Power asymmetries can even processes, amplifying the influence of elites. lead to breakdowns in institutional functions, weakening the effectiveness of policies. When One way of understanding the interplay institutions are captured by the wealthy, citi- between inequality and the dynamics of power zens are less willing to be part of social contracts is to draw on a framework that explores the (the sets of rules and expectations of behaviour process through which inequalities are gener- that people voluntarily conform to that un- ated and perpetuated. At its core, this process derpin stable societies). When that translates is often referred to as governance—or the way into lower compliance with paying taxes, it in which different actors in society bargain to diminishes the state’s ability to provide quality reach agreements (policies and rules). When public services. That can in turn lead to greater these agreements take the form of policies, inequalities in health and education. When the they can directly change the distribution of overall system is perceived as unfair, possibly resources in society (the bottom arrow in the due to systematic exclusions or clientelism right loop of figure 9, “outcome game”). For

Overview | 11 example, policies on taxation and social spend- inequality may undermine the effectiveness of ing determine who pays into the fiscal system governance) or pave the way to more equalizing and who benefits from it. These policies directly and inclusive dynamics. influence development outcomes such as eco- nomic inequality (and growth). However, by Gender inequality redistributing economic resources, these poli- cies are also redistributing de facto power (the Some groups of people are systematically dis- top arrow in the right loop of figure 9). This advantaged in many ways. These groups might can generate (or reinforce) power asymmetries be defined by ethnicity, language, gender or between actors bargaining in the policy arena, caste—or simply by whether they live in the which can in turn adversely affect the effective north, south, east or west of a country. There are implementation of policies. For example, power many examples of such groups, but undoubt- asymmetries can manifest in the capture of pol- edly the largest worldwide is women. Gender icies by elite actors—undermining the ability of disparities are among the most entrenched governments to commit to achieving long-term forms of inequality everywhere. Because these goals. Or they may manifest in the exclusion disadvantages affect half the world, gender ine- of certain population groups from accessing quality is one of the greatest barriers to human high-quality public services—undermining development. cooperation by harming the willingness to pay taxes. This can lead to a vicious cycle of inequal- Gender inequality is complex, with differing ity (inequality traps) in which unequal societies progress and regress from place to place and begin to institutionalize the inequality. This issue to issue. Awareness has increased through loop plays out in prevailing institutions and so- the #MeToo movement, or the #NiUnaMenos cial norms (the outcome game) and can lead to movement, which shined a spotlight on vio- actors deciding to change the rules of the game lence against women. And girls around the (the bottom arrow in the left loop of figure 9). world have been catching up on some of the In this way, de jure power is also redistributed. basics, such as enrolment in primary school. This can be far more consequential because it not only changes current outcomes but also sets But there is less to celebrate about progress the conditions that shape actors’ behaviour in beyond these fundamentals. Inequality is still the future. Once again, the way in which power sharp in the power men and women exercise at asymmetries play out in the policy arena can home, in the workplace or in politics. At home exacerbate and entrench inequalities (clearly, women do more than three times as much un- paid care work as men. And although in many countries women and men vote equally in

FIGURE 9 Inequalities, power asymmetries and the effectiveness of governance

De jure power De facto power

Rules Policy Development arena outcomes

Rules game Outcome game

Note: Rules refer to formal and informal rules (norms). Development outcomes refer to security, growth and equity.

12 | HUMAN DEVELOPMENT REPORT 2019 elections, there are differences in higher levels · There are inequalities among groups (hori- of political power. The higher the power, the zontal inequalities) and among individuals larger the gap from parity, rising to 90 percent (vertical inequalities). in the case of heads of state and government.

• There are inequalities between and within

Social and cultural norms often foster be- countries, which can follow different dynamics. haviour that perpetuates such inequalities. Norms—and a lack of power—both have an · There are intrahousehold inequalities (for in- impact on all forms of gender inequality, from stance, in 30 Sub-Saharan countries roughly violence against women to the glass ceiling. three-quarters of underweight women and This Report presents a new social norms index undernourished children are not in the poor- that looks at the links between social beliefs est 20 percent of households, and around and gender equality in multiple dimensions. half are not in the poorest 40 percent).14 Globally only 1 man in 10 (and 1 woman in 7) A new generation of metrics is needed to fill did not show some form of clear bias against gender equality. The biases follow a pattern: the many data gaps to measure these different They tend to be more intense in areas where inequalities and, more generally, to go sys- more power is involved. And there is backlash, tematically beyond averages. This starts with as the proportion of people biased against gen- gaps in some of the most basic statistics, with der equality has grown over the last few years many developing countries still lacking in vital (figure 10), even though there are different registration systems. For income and wealth patterns across countries. inequality the progress over the past few years has been remarkable. But data remain scarce, in Key message 4: Assessing and part because of the lack of transparency and the responding to inequalities in low availability of information. On a new index human development demands presented in this Report, 88 countries score 1 a revolution in metrics or less (on a 20-point scale) for availability of information on income and wealth inequal- Existing standards and practices for measuring ity—meaning that they have 5 percent or less inequality are inadequate to inform public de- of what would be an ideal level of transparency. bate or to support decisionmaking. Innovative work—some experimental—is Part of the challenge is the sheer number of unfolding, led by academics, multilateral or- different ways to understand inequality. To ganizations and even a few governments, to highlight a few: make more systematic and comparable use of statistics on income inequality. But data sources remain only partially integrated, and coverage remains very limited.

FIGURE 10

Bias against gender equality is on the rise: The share of women and men worldwide with no gender social norms bias fell between 2009 and 2014

Percent of surveyed population responding 2005­2009 with biases towards gender equality 2010­2014

Indicated bias in one or Female 40.1 43.3 fewer questions from the Male

Indicated bias in two or Female 56.7 59.9 World Values Survey Male 69.7 70.4

Note: Balanced panel of 32 countries and territories with data from both wave 5 (2005­2009) and wave 6 (2010­2014) of the World Values Survey, accounting for 59 percent of the world population. Gender biases in social norms are measured through people’s views about gender roles in politics (from political rights to the ability to serve as leader), education (importance of a university degree), the economy (from the right to have jobs to the ability to work as business executive) and the physical integrity of women (from intimate partner violence to reproductive health). Source: Based on data from the World Values Survey.

Overview | 13 The distributional national accounts method- may not even reflect society’s views. To under- ology is still in its infancy, and many of its as- stand any single aspect of inequality—and there sumptions have been challenged. Still, as long as are many—one needs to look across the entire it remains fully transparent and improvements population, going beyond averages. What pro- continue to be made, it could integrate, in an portions of people survive to certain ages, reach overarching agenda, the combination of data key education levels or earn certain amounts? from the System of National Accounts, house- And how likely is it that the relative position in hold surveys and administrative data to pro- society of an individual, a family or a particular vide new perspectives on the evolution of the group changes over time? Summary measures distribution of income and wealth. This would remain important—when they reflect sound encompass some of the main recommendations properties to assess distributions—but are only of the Commission on the Measurement of a small window onto a wider discussion about Economic Performance and Social Progress, inequalities in human development. including an integrated focus on income and wealth inequality.15 This Report presents results Key message 5: We can redress based on the methodology that reveal dynamics inequalities if we act now, before of income inequality that are masked when us- imbalances in economic power ing summary measures that rely on a single data are politically entrenched source. To give an illustration, the results sug- gest that the top of the income distribution in Nothing is inevitable about many of the most Europe has been the main beneficiary of income pernicious inequalities in human development. growth since 1980 (figure 11). This is the single most important message of this Report. Every society has choices about the levels Summary measures of inequality aggregate and kinds of inequalities it tolerates. That is not complex information into one number. They are to say that tackling inequality is easy. Effective based on implicit judgements about what forms action must identify drivers of inequality, which of inequality are—or are not—important. Those judgements are rarely transparent and

FIGURE 11

Between 1980 and 2017 post-tax incomes grew close to 40 percent for the poorest 80 percent of the European population, compared with more than 180 percent for the top 0.001 percent

Total income Top 1 percent captured growth (percent) 13 percent of growth

0

10 20 30 40 50 60 70 80 90 99 99.9 99.99 99.999

Note: After the 90th percentile the scale on the horizontal axis changes. The composition of income groups changes from 1980 to 2017, so the estimates do not represent the changes in income of the same individuals over time. Source: Blanchet, Chancel and Gethin (2019); World Inequality Database (http://WID.world).

14 | HUMAN DEVELOPMENT REPORT 2019 are likely complex and multifaceted, often relat- inequalities. They raise revenue to improve key ed to prevailing power structures that the people public services (health care and schools) and to currently holding sway may not wish to change. provide social insurance—benefiting both poor people and people in the middle of the income But what to do? Much can be done to redress distribution. inequalities in human development with a dual policy objective. First is to accelerate convergence Income inequality is lower after taxes and in basic capabilities while reversing divergences government transfers, but the impact of redis- in enhanced capabilities and eliminating gen- tribution varies. In a selection of developed der- and other group-based (or horizontal) ine- countries, taxes and transfers led to a 17-point qualities. Second, to jointly advance equity and reduction in the Gini coefficient, when com- efficiency in markets, increasing productivity that paring pretax and post-tax incomes. But in translates into widely shared growing incomes— developing countries the reduction was just 4 redressing income inequality. The two sets of points (figure 13). policies are interdependent, with those that ad- vance capabilities beyond income often requiring Equally important, however, is to go beyond resources to fund public health or education, taxation and transfers (postmarket policies) by which are financed by taxes. And the overall re- also addressing inequalities while people are sources available are, in turn, linked to productiv- working (in-market policies) and before they ity, which is linked in part to people’s capabilities. start working (premarket policies). The two sets of policies can thus work together in a virtuous policy cycle (figure 12). In-market policies can level the economic playing field. Policies related to market power It is often possible to make progress in eq- (antitrust), inclusive access to productive cap- uity and efficiency at the same time. Antitrust ital, and collective bargaining and minimum policies are an example. They curb firms’ ability wages affect how the benefits from production to use market power, levelling the playing field are distributed. Equally relevant are premarket and increasing efficiency. And they lead to policies aimed at equalizing opportunities dur- more equitable outcomes by reducing econom- ing childhood in health and education—and ic rents that concentrate income. postmarket policies, such as income and wealth taxes, public transfers and social protection. An integrated battery of policies One clear role for premarket policies is in beyond any single silver bullet early childhood, where inequality-reducing interventions can support health, nutrition Taxes—whether on income, wealth or and cognitive development and produce a big consumption—can do much to redress return on investment. That is not to say that every good policy can reduce inequality and

FIGURE 12

A framework for designing policies to redress inequalities in human development

Redressing inequalities in basic and enhanced capabilities

Policies to: Premarket Premarket In-market Policies for

• Accelerate convergence Postmarket inclusive expansion
• Reverse divergence in in incomes

enhanced capabilities (productivity and equity)

Overview | 15 FIGURE 13 is ultimately a societal and political choice. History, context and politics matter. Social Redistributive direct taxes and transfers explain norms that can lead to discrimination are hard nearly all the difference in disposable income to change. Even with legislation setting equal inequality between advanced and emerging rights, social norms may prevail in determin- economies ing outcomes. This Report’s analysis of gender inequality shows that reactions become more Income inequality Advanced economies intense in areas where more power is involved, (absolute reduction in Emerging markets which can culminate in a backlash towards the Gini coefficient) and developing countries very principles of gender equality. Explicit poli- cies for tackling stereotypes and the stigmatiza- 0.48 0.49 tion of excluded groups are an important part of the toolkit to reduce inequalities. The political economy of tackling inequality 0.31 can be particularly challenging. For public ser- vices, change can happen from the top down, by Before After extending benefits enjoyed by those at the top to others (figure 14). But those already benefit- Source: Based on IMF (2017a). ing may have little incentive to extend services if that might be perceived to reduce quality. increase welfare—as noted, processes such as Change can also happen from the bottom up, the diffusion of new technology and human increasing the income below which a family development achievements in large segments of qualifies for free public or subsidized services, society may increase inequality. What matters for example. But higher income groups might is whether the process that generates that ine- resist this if they seldom use such services. A quality is, in itself, somehow biased or unfair. third approach is to build out from the mid- dle—when a system covers those who are not Creating incentives for change the poorest but who are vulnerable, such as for- mal workers earning low wages. Here, coverage Even if resources are available to undertake can be expanded both upward and downward. an agenda for convergence in both basic and As the quality of services improves, higher in- enhanced capabilities, reducing inequalities come groups are likely to want to participate, broadening the support to expand services to poor people.

FIGURE 14

Strategies for practical universalism in unequal developing countries

Top-down Bottom-up Lower middle-up and trajectory trajectory

Wealthy and high Low High

Hard to expand, as it would Effective to address urgent needs. Relative high quality can attract compromise quality. But hard to expand because of high-income groups to join middle resource constraints and because class. This might be used to finance low quality does not attract expansion to the poor (interclass participation of middle class. alliance).

Source: Human Development Report Office based on the discussion in Martínez and Sánchez-Ancochea (2016).

16 | HUMAN DEVELOPMENT REPORT 2019 In developed countries one challenge for be translated into political dominance. And sustaining social policies is to ensure that they that in turn can lead to more inequality. At benefit a broad base, including the middle that stage interventions are far harder and less classes. Yet such benefits may be eroding. In sev- effective than if they had been taken earlier eral Organisation for Economic Co-operation on. Of course, action is context specific. The and Development countries, members of the nature and relative importance of inequalities middle class perceive themselves as being pro- vary across countries—and so should policies gressively left behind in income, security and to address them. In much the same way that affordable access to quality health care and there is no silver bullet to address inequalities education. within a country, there is no one-size-fits-all basket of policies to address inequalities across In developing countries the challenge is often countries. Even so, policies in all countries will to solidify social policies for a still vulnerable have to confront two trends that are shaping in- middle. In some of these countries members equalities in human development everywhere: of the middle class pay more for social services climate change and accelerating technological than they receive, and they often perceive the progress. quality of health care and education to be poor. So they turn to private providers: The share of Climate change and inequalities students going to private schools for primary in human development education in some of these countries rose from 12 percent in 1990 to 19 percent in 2014. Inequality and the climate crisis are interwo- ven—from emissions and impacts to policies A natural response would be to take resourc- and resilience. Countries with higher human es from those at the top. But the richest, though development generally emit more carbon per few in number, can be an obstacle to expanding person and have higher ecological footprints services. And they can frustrate action in mul- overall (figure 15). tiple ways, through lobbying, donating to polit- ical campaigns, influencing the press and using Climate change will hurt human develop- their economic power in other ways in response ment in many ways beyond crop failures and to decisions they dislike. natural disasters. Between 2030 and 2050 climate change is expected to cause an addi- Globalization means national policy is often tional 250,000 deaths a year from malnutrition, circumscribed by entities, rules and events malaria, diarrhoea and heat stress. Hundreds beyond the control of national governments, of millions more people could be exposed to with pervasive downward pressures on corpo- deadly heat by 2050, and the geographic range rate income tax rates and labour standards. Tax for disease vectors—such as mosquitoes that evasion and avoidance are made easier by insuf- transmit malaria or dengue—will likely shift ficient information, by the rise of large digital and expand. companies operating across tax jurisdictions and by inadequate interjurisdictional cooper- The overall impact on people will depend ation. In these policy domains international on their exposure and their vulnerability. Both collective action must complement national factors are intertwined with inequality in a vi- action. cious circle. Climate change will hit the tropics harder first, and many developing countries are Where next? tropical. Yet developing countries and poor communities have less capacity than their rich- A human development approach opens new er counterparts to adapt to climate change and windows on inequalities—why they matter, severe weather events. So the effects of climate how they manifest themselves and what to do change deepen existing social and economic about them—helping move towards concrete fault lines. action. But the opportunities to address ine- qualities in human development keep narrow- There are also effects in the other direction, ing the longer that inaction prevails because with evidence that some forms of inequality imbalances in economic power can eventually may make action on climate harder. High in- come inequality within countries can hinder

Overview | 17 FIGURE 15

Ecological footprints expand with human development

Ecological footprint, 2016 Medium human High human Very high human (global hectare per person) development development development

8

6

4

2 Biocapacity per person, world average (1.7 global hectares) 0 0.4 0.5 0.6 0.7 0.8 0.9 1

Note: Data cover 175 countries in the Global Ecological Footprint Network database (www.footprintnetwork.org/resources/data/; accessed 17 July 2018). As used here, the ecological footprint is a per capita measure of how much area of biologically productive land and water a country requires, domestically and abroad, to produce all the resources it consumes and to absorb the waste it generates. Each bubble represents a country, and the size of the bubble is proportional to the country’s population. Source: Cumming and von Cramon-Taubadel 2018.

the diffusion of new environmentally friendly is not the only variable that matters. It is also technology. Inequality can also influence the important to consider a broader set of social balance of power among those arguing for policy packages that address inequalities and and against curbing carbon emissions. Income climate together while facilitating the reali- concentration at the top can coincide with the zation of human rights. There are choices for interests of groups that oppose climate action. countries and communities as they raise their ambitions for inclusive and sustainable human Inequalities in human development are fun- development. damental to the climate crisis in another way. They are a drag on effective action because Harnessing technological higher inequality tends to make collective ac- progress to reduce inequalities tion, key to curbing climate change both within in human development and across countries, more difficult. Scientific progress and technological innova- Yet there are options to address economic tion—from the wheel to the microchip—have inequalities and the climate crisis together, driven improvements in living standards which would move countries towards inclu- throughout history. And technological change sive and sustainable human development. will likely continue to be the fundamental Carbon pricing is one. Some of the unavoida- driver of prosperity, pushing increases in pro- ble distributional impacts of carbon prices can ductivity and hopefully enabling a transition be addressed by providing financial support to to more sustainable patterns of production and poorer people, hardest hit by higher energy consumption. bills. But such strategies have faced challenges in practice, because the distribution of money

18 | HUMAN DEVELOPMENT REPORT 2019 But what will be the magnitude of future Divergence, dividing the few societies that changes and how will the gains from innova- industrialized from the many that did not. tion be distributed? Concern is growing about What is different now is that—perhaps for the how technological change will reshape labour first time in history—much of the technology markets, particularly in how automation and behind the current transformation could be artificial intelligence might replace tasks now accessed anywhere. Yet the gaps in countries’ performed by humans. abilities to harness the new opportunities are very large, with massive implications for both Technological change has been disruptive be- inequality and human development. fore, and much can be learned from the past. One key lesson is to ensure that major innovative dis- Technological change does not occur in a ruptions help everyone, which requires equally vacuum but is shaped by economic and social innovative policies and perhaps new institutions. processes. It is an outcome of human action. The current wave of technological progress will Policymakers can shape the direction of tech- require other changes, including stronger anti- nological change in ways that enhance human trust policies and laws to govern the ethical use development. For instance, artificial intelli- of data and artificial intelligence. Many of these gence might replace tasks performed by people, will require international cooperation to succeed. but it can also reinstate demand for labour by creating new tasks for humans, leading to a The Industrial Revolution set humanity on net positive effect that can reduce inequalities a path towards unprecedented improvements (figure 16). in well-being. But it also triggered the Great

FIGURE 16 Technology can displace some tasks but also create new ones

Displacement Productivity Reinstatement effect effect effect

(tasks related to accounting + (cyber security and bookkeeping, experts, digital travel agents) Net change in transformation specialists, demand for data scientists) - labour +

Overview | 19 Towards reducing inequalities in human primary and secondary enrolment rates. Many development in the 21st century of these aspirations are already reflected in the 2030 Agenda for Sustainable Development. This Report argues that tackling inequalities is possible. But it is not easy. It requires clarifying Power imbalances are at the heart of many which inequalities matter to the advancement inequalities. They may be economic, political of human development and better understand- or social. For example, policies might need to ing the patterns of inequality and what drives reduce a particular group’s disproportionate them. This Report urges everyone to recognize influence in politics. They might need to level that the current, standard measures to account the economic playing field through antitrust for inequality are imperfect and often mis- measures that promote competition for the leading—because they are centred on income benefit of consumers. In some cases, addressing and are too opaque to illuminate the under- the barriers to equality mean tackling social lying mechanisms generating inequalities. So, norms embedded deep with a country’s history this Report argues for the value of looking at and culture. Many options would enhance both inequalities beyond income, beyond averages equity and efficiency—and the main reason —a nd summary measures of inequality—and they are not pursued often has to do with the beyond today. power of entrenched interests who stand not to gain much from change. There should be a celebration of the remark- able progress that has enabled many people Thus, while policies matter for inequalities, around the world to reach minimum standards inequalities also matter for policies. The human of human development. But continuing the development lens—placing people at the heart policies that have led to these successes alone is of decisionmaking—is central to open a new insufficient. Some people have been left behind. window on how to approach inequality, asking At the same time, many people’s aspirations are why and when it matters, how it manifests itself changing. It is short-sighted for societies to and how best to tackle it. This is a conversation focus only on inequality in the most basic capa- that every society must have. It is also a con- bilities. Looking beyond today means scanning versation that should begin today. True, action ahead to recognize and tackle the new forms may carry a political risk. But history shows of inequality in enhanced capabilities that are that the risks of inaction may be far greater, growing in importance. Climate change and with severe inequalities eventually propelling technological transformations are adding to a society into economic, social and political the urgency. tensions.

Tackling these new inequalities can have a There is still time to act. But the clock is profound impact on policymaking. This Report ticking. What to do to address inequalities does not claim that any one set of policies will in human development is ultimately for each work everywhere. But it does argue that poli- society to determine. That determination cies must get beneath the surface of inequality will emerge from political debates that can be to address their underlying drivers. Addressing charged and difficult. This Report contributes some of these drivers will mean realigning to- to those debates by presenting facts on inequal- day’s policy goals: emphasizing, for instance, ities in human development, interpreting them high-quality education at all ages, including through the capabilities approach and propos- preprimary levels, rather than focusing on ing ideas to reduce them over the course of the 21st century.

20 | HUMAN DEVELOPMENT REPORT 2019 Part I PART I.

Inequality of what? In addressing this deceptively simple question, Amartya Sen developed the approach that has informed Human Development Reports since the first one was published in 1990.1 Sen posed that question because celebrating human diversity calls for reflecting on the kind of inequality we should ultimately care about. The answer to Sen’s question “inequality of what?” is the “inequality of capabilities.”

As the second decade of the 21st century world—now takes second place to interest in Many more people comes to an end, the questions about inequali- global inequality.3 around the world, ty that motivated Sen in the late 1970s have re- across political surfaced with a vengeance. Now, however, the Reducing inequality was enshrined in the orientations, feel conversation is not only about understanding 2030 Agenda for Sustainable Development, strongly that income what kind of inequality should be measured; with several Sustainable Development Goals inequality should be it is also about how to cope with them.2 Many (SDGs) speaking to the aspiration to reduce reduced, a preference more people around the world, across political inequality across multiple dimensions. In line that has intensified orientations, feel strongly that income ine- with the 2030 Agenda, part I of the Report since the 2000s quality should be reduced, a preference that argues that we need to go beyond income in has intensified since the 2000s (figure I.1). exploring inequality—and especially in con- Indeed, some evidence suggests that interest fronting the new inequalities of the 21st cen- in global growth—often equated with broader tury. It advances the view that the capabilities improvements in development around the approach is well suited to understanding and confronting these new inequalities.4

FIGURE I.1 The share of the population stating that income should be more equal increased from the 2000s to the 2010s

Change in the share of 35 out 33 out 32 out population stating that of 39 of 39 of 39 income should be more countries countries countries

0

­40 Center Leaning right Leaning left

Population in selected countries by political self-identification

Note: Each dot represents one of 39 countries with comparable data. The sample covers 48 percent of the global population. Based on answers on a 1­5 scale, where 1 is “income should be more equal” and 5 is “we need larger income differences.” Source: Human Development Report Office calculations based on data from the World Values Survey, waves 4, 5 and 6.

To equality in capabilities | 23 Despite improvement Why, after all, should concerns about in- social injustices are seen, much less acknowl- and convergence in equality be rising today—at a time of great edged, by social institutions, and this is often progress in living standards, with an unprec- the case for indigenous or ethnic groups; mi- the capabilities central edented number of people around the world grants; lesbian, gay, bisexual, transgender and to the Millennium making a “great escape”5 from hunger, disease intersex people; and other socially stigmatized and poverty?6 Even though many are still being groups that suffer abuse and discrimination.8 Declaration of 2000 left behind, the Human Development Index Such inequality also—in too many places— and the Millennium (HDI) shows, on average, impressive improve- affects the situation of women, who—even Development Goals, ment—even convergence—in the capabilities when they share a home with a man giving some gaps remain included in the HDI. Yet, chapter 1 shows that them access in principle to similar goods and along with convergence in the basic capabilities services—are subject to imposed roles and stark, and new that were the focus of Human Development often violence. The #MeToo movement has ones are opening in Reports in the early 1990s, divergences are shown how systematic abuse and humiliation capabilities that will opening in other indicators, both within and are widespread and not defined by income or increasingly determine across countries: Life expectancy at older ages social status.9 differences between is becoming more unequal, as is access to ter- those who can and tiary education. In short, despite improvement To be sure, income and wealth inequalities and convergence in the capabilities central to can be significant and central to policymakers’ those who cannot the Millennium Declaration of 2000 and the thinking about inequality in human develop- take full advantage Millennium Development Goals, some gaps ment. Such economic inequalities, narrowly of the 21st century’s remain stark, and new ones are opening in considered, can be perceived as unfair or can new opportunities capabilities that will increasingly determine actually constrain people’s well-being (through differences between those who can and those several channels, as explored in chapter 2). who cannot take full advantage of the 21st Analysis of income and wealth inequalities is century’s new opportunities. Time and again, thus necessary and is considered throughout the analysis shows that countries and people at the Report, but focusing exclusively on income the bottom are catching up in basic capabilities and wealth inequalities would be too reductive while those at the top pull away in enhanced by failing to acknowledge the full scope of ine- capabilities.7 quality in human development.

Convergence in basic capabilities gives the Chapter 2 documents how inequalities in direction of change but does not mean that the capabilities emerge, showing how they are often gaps are fully closed. In fact, those furthest be- interconnected and persistent. Even as differ- hind are making little to no progress. Chapter 1 ences in basic capabilities are reduced, as more thus shows that the world is expected to reach and more people acquire the basic capabilities 2030 with preventable gaps in infant mortality, towards meeting minimum achievements in out-of-school children and extreme income health and education, gradients—meaning poverty. Drawing on granular data to zoom in that individuals who are better off have better on geographic areas, it documents overlapping health and education outcomes than those deprivations and intersectional exclusions. who are worse off—persist or become more Finally, the chapter zooms out on the dynam- pronounced. ics of risk—health, natural disaster or conflict shocks that expose groups or individuals to The mechanisms accounting for the emer- added vulnerability. Behind these patterns lie gence of inequalities in capabilities are described the stubborn challenge of strengthening the in chapter 2 at two levels. First, by taking a capabilities of those furthest behind. lifecycle approach that traces how parents’ ad- vantages in income, health and education shape The persistent and increasing inequalities their children’s path over time, often leading to in enhanced capabilities matter more than for persistent “hoarding” of opportunities across their instrumental value. Chapter 1 also looks generations. Second, by noting that these at how they have a bearing on human dignity. mechanisms do not occur in a vacuum and that Individuals or groups of people might have context, including economic inequality, shapes access to resources—but not equal treatment opportunity through multiple channels, such through formal law or social norms. Not all as how policies are designed and implemented. The distribution of resources and opportunities

24 | HUMAN DEVELOPMENT REPORT 2019 in a society depends heavily on the distribution development. It shows that focusing on raising Focusing on of power. Power concentration creates imbal- people above minimums is insufficient, given raising people ances and can lead to the capture of both gov- that gradients of inequality in capabilities con- above minimums is ernment and markets by powerful elites—which tinue to open up and persist. insufficient, given that can further drive income and wealth inequality, gradients of inequality in a cycle that weakens responsiveness to the as- Part I of the Report opens our view about in capabilities continue pirations of the general population. This pattern inequalities in human development. But to open up and persist appears to have already happened in history (see this is just the first step. As United Nations spotlight 1.1 at the end of chapter 1).10 These High Commissioner for Human Rights dynamics can in turn erode governance, hurting Michelle Bachelet points out in her Special human development.11 Contribution, “Diagnosis is not enough—we must push for public policies that tackle these Part I of the Report takes the inequality forms of injustice.” These findings, inspired discussion beyond income towards capabilities, by the human development approach, will be broadening the range of data considered in critical to support efforts to implement the the inequality debate and uncovering patterns 2030 Agenda for Sustainable Development of convergence and divergence in human (box I.1).12

BOX I.1 The capabilities approach and the 2030 Agenda for Sustainable Development

The dimensions of inequality in human development 3), quality education and lifelong learning opportunities considered in this report are reflected in the 2030 (SDG 4), gender equality and empowerment for all wom- Agenda for Sustainable Development and its accompa- en and girls (SDG 5), sustainable water and sanitation nying Sustainable Development Goals (SDGs). (SDG 6), sustainable reliable energy (SDG 7), decent jobs (SDG 8) and access to justice (SDG 16). Other goals The global consensus around the SDGs represents an aim to advance the provision of global public goods evolution from what the Millennium Development Goals (such as climate stability). considered “basic” or essential for developing countries by the end of the 20th century. This report is inspired by As with any global approach, considering a specific that evolution and considers dimensions of inequality that set of dimensions has limitations. It does not address all are universally relevant and go beyond the basic. dimensions of unfairness and injustice that might be im- portant in particular places. However, the Report com- The SDGs seek to reduce inequality in many forms. plements and cross-checks globally defined measures of They not only aim to reduce inequality between and inequality—based on objective data—with information within countries (SDG 10) but also envision an abso- on perceptions of inequality, with measures of inequal- lute end to some deprivations: poverty in all its forms ity in subjective well-being and with some nationally (SDG 1) and hunger (SDG 2). They also seek to extend defined measures. some basic conditions to all people: healthy lives (SDG

To equality in capabilities | 25 SPECIAL CONTRIBUTION

A new look at inequality

As every year, the 2019 Human Development Report of the United Nations together, in the face of these new scenarios, and achieve greater well-being Development Programme invites us to take a look at ourselves in the mirror. for people? It is a path that we must learn to tread together. In systematically integrating information about the development of our so- cieties, we are confronted with the evidence of what we have achieved and Access to health, education, new technologies, green areas and spaces where we are failing. free of pollution are increasingly an indicator of the way in which oppor- tunities and well-being are distributed among groups of people and even This evidence is much more than a compilation of numbers and figures. between countries. Because it is all about people’s well-being: Each gap that persists or grows is a call to respond to the injustice of inequality with effective policies. Explaining and understanding the dimensions of inequalities most What can we expect when a girl is born in poverty, with no proper health critical to people’s well-being helps in choosing the best lines of action. coverage and in an environment where it is harder and harder to access Diagnosis is not enough—we must push for public policies that tackle these drinking water due to climate change? How much longer can our societies forms of injustice. keep getting it wrong when what they do breaches basic human rights? These are the issues with which inequality faces us. Therefore, all countries have a job to do. But over many years we have found that individual efforts are not enough; many challenges demand a We know that inequality takes many forms. Many, such as inequalities collective approach. of income or gender inequalities, have been around us for a long time. It should be a matter of pride that considerable progress has been made in In the United Nations System, we believe that the 2030 Agenda for these issues in much of the globe. This Report highlights that inequalities Sustainable Development and the Sustainable Development Goals (SDGs) in the basic capabilities reflecting extreme deprivations are going down. For are the kind of response needed in these modern times: They take an all- instance, the world is moving towards average gender parity in access to round look at the phenomena and the solutions; they seek convergence primary and secondary education. However, at the same time, inequalities between the actions of governments and international agencies; and they reflecting greater levels of empowerment and more important for the future are based on transparent and comparable measurements. With their inter tend to be higher and, in some cases, increasing. Here, we have the example sectoral approach and the commitment of all governments, the SDGs put us of women’s representation at the top political level. all at the service of a single endeavour.

Although we still have a long way to go, we have accumulated experi- The best example of what we hold in our own hands is the enormous ence about what works in social protection, financial instruments and path- challenge of limiting the rise in the global temperature to 1.5°C. Our United ways of social mobility. There are success stories of better representation Nations Office for Human Rights has said it clearly: Climate change directly of women, more equitable participation in the labour market or driving out and indirectly affects a range of human rights which must be guaranteed. discrimination against sexual diversity. The paradox of having such long- We view with satisfaction that science, governments, business and civil so- standing inequalities is that we, as a society, have found pathways for posi- ciety are starting to coalesce around concrete targets. Thus, little by little, tive change. What is needed in many cases is the political will. sectoral isolation and arguments are breaking down.

Yet there are inequalities which face us with even greater challenges. It It is the path we must insist on. We have a duty to eradicate old and is precisely on these that the Report seeks to shed light: These are inequal- new forms of inequality and exclusion which every day breach the rights of ities which stem from new phenomena and global conflicts. These inequali- millions of people living on our planet. ties are more challenging as they respond to complex and dynamic processes still to be well understood. Are we fully aware of the impact of migrations, It would be a mistake to think that there have not been successes, that the effects of climate disasters or the new epidemiological threats to our injustice in the world has not been driven back. But so long as there is pain coexistence? Because that is what it is about; how do we manage to live and suffering due to inequality, we have a duty to face up to what we are doing wrong and which we can put right.

We have more future than yesterday: This is the invitation that we must all make our own.

Michelle Bachelet Jeria United Nations High Commissioner for Human Rights

26 | HUMAN DEVELOPMENT REPORT 2019 Chapter 1

Inequality in Moving targets 1.

Inequality in human development: Moving targets in the 21st century

This chapter considers two main questions: Where do human development inequalities stand today and how are they changing? Many inequalities in human development embody unfairness. To see how, take two babies, both born in 2000— one in a low human development country, the other in a very high human development country (figure 1.1). What do we know about their prospects for adult life today? We know that they are vastly different. The first is very likely to be enrolled in higher education, along with the majority of 20-year-olds in more developed countries today. She or he is preparing to live in a highly globalized and competitive world and has chances do so as a highly skilled worker.

In contrast, the child from the low human education.1 Both of these young people are just development country is much less likely to be beginning their adult lives, but circumstances al- alive. Some 17 percent of children born in low most entirely beyond their control have already human development countries in 2000 will have set them on different and unequal paths in terms died before age 20, compared with just 1 percent of health, education, employment and income of children born in very high human develop- prospects—a divergence that can be irreversible. ment countries. And those who survive have an expected lifespan 13 years shorter than their Some inequalities within countries—wheth- counterparts in the group of more developed er developing or developed—are no less countries. The child born in the low human extreme than those in the between-country development country is also unlikely to still example above. In the United States average life be in education: Only 3 percent are in higher expectancy at age 40 between the top 1 percent of the income distribution and the bottom

FIGURE 1.1

Children born in 2000 in countries with different incomes will have severely different capabilities by 2020

In higher Children born education 55 in 2000 in very high 3 human development

Children born Not in development countries 80 higher

Note: These are estimates (using median values) for a typical individual from a country with low human development and from a country with very high human development. Data for participation in higher education are based on household survey data for people ages 18­22, processed by the United Nations Educational, Scientific and Cultural Organization Institute for Statistics in www.education-inequalities.org (accessed 5 November 2019). Percentages are with respect to people born in 2000. People that died before age 20 are computed based on births around 2000 and estimated deaths for that cohort between 2000 and 2020. People in higher education in 2020 are computed based on people estimated to be alive (from cohort born around 2000), and the latest data of participation in higher education. People not in higher education are the complement. Source: Human Development Report Office calculations based on data from the United Nations Department of Economic and Social Affairs and the United Nations Educational, Scientific and Cultural Organization Institute for Statistics.

Moving targets in the 21st century | 29 1 percent differs by 15 years for men and 10 challenge because they are dynamic, complex years for women.2 Such disparities are widening. and multidimensional. Which to include? How to measure them? How to aggregate The 21st century presents an unprecedented- them? How to analyse them? And at what ly broad range of human experiences. See, for level: globally, nationally, subnationally, within instance, how the distribution of nonincome social groups or even in the household? Amid indicators of the Human Development Index this complexity, however, it might be possible for subnational areas covers a huge spectrum to discern broad patterns of evolution in ine- of outcomes in health and education. Extreme qualities that are widely shared. This is the task deprivations still exist, not only among low that the rest of this chapter explores. human development countries (figure 1.2). Global elites, including people in low human Understanding inequality development countries, enjoy more knowledge, in capabilities more years of healthy life and more access to life-changing technologies. Human development means expanding the substantive freedoms to do things that people Why do striking inequalities persist? Partly value and have a reason to value.5 What people because of social structures—many with histor- actually choose to be and do—their achieved ical roots—that remain entrenched in formal functioning—is enabled by income and wealth and informal institutions, adamantly resisting but is distinct from it. And while the achieved change.3 To shift the curve of human develop- functioning matters, human development is ment inequalities, it is not enough to improve not defined merely by the choices that people just one or two particular indicators. Instead, actually make; it is also defined by “the freedom the social structures that perpetuate inequity that a person has in choosing from the set of need to change.4

Portraying the scope of inequalities in human development and their evolution is a daunting

FIGURE 1.2 Still massive inequality in human development across the world, 2017

Life expectancy at birth Mean years of schooling Expected years of schooling (frequency) (frequency) (frequency)

Years Years Years 15 25

5

5

50 0 0 Low Medium High Very high Low Medium High Very high Low Medium High Very high Human development group Human development group Human development group

Source: Human Development Report Office based on calculations of subnational Human Development Index values by Permanyer and Smits (2019).

30 | HUMAN DEVELOPMENT REPORT 2019 feasible functionings, which is referred to as but also with values and with people’s changing Inequalities we care the person’s capability.”6 Thus, the analysis of demands and aspirations. about may indeed inequality in this chapter considers inequality be moving targets of capabilities (box 1.1). The capabilities approach is thus open-ended, which some observers see as a shortcoming.9 But what capabilities to consider? Sen argued One objection is that it does not lend itself to that one must adjust in response to evolving specifying a standard and fixed goal for evalu- social and economic conditions. For example, ating social welfare because capabilities are con- in India at the time of independence in 1947, it tinuously moving targets. This Report takes a was reasonable to concentrate “on elementary different view: It considers that the inequalities education, basic health, […] and to not worry we care about may indeed be moving targets and too much about whether everyone can effec- thus aims to identify patterns and dynamics of tively communicate across the country and be- inequality in a wider set of capabilities that may yond.”7 Later, however—with the internet and be increasingly relevant during the 21st century. its applications, as well as broader advances in information and communication technology— Another challenge is how to measure capa- access to the internet and freedom of general bilities—that is, how to move from concepts communication became an important capabil- to the empirical assessment of how capabilities ity for all Indians. Whereas one relevant aspect are distributed. Here the Report follows the of this insight is strictly linked to capabilities approach taken when the Human Development (access to the internet), another intersects with Index (HDI) was introduced and identifies a human rights and specifically with the right to few observable achieved functionings to capture freedom of opinion and expression.8 Moreover, broader capabilities (for instance, in the HDI, capabilities evolve not only with circumstances having the option to live a long and healthy life is associated with the indicator of life expectancy

BOX 1.1 Inequality of capabilities

In keeping with previous Human Development Reports, free choice or whether the person wanted to travel but this Report assumes, from a normative perspective, either could not afford it or was denied entry.3 that the inequalities that matter intrinsically are ine- qualities in capabilities. Capabilities—broadly defined The first Human Development Reports used the as people’s freedom to choose what to be and do— capabilities approach to intervene in the development cannot be reduced to income and wealth alone, be- discourse of the time, when debates centred on ba- cause these are instrumental.1 Nor can they be defined sic needs,4 leading to the introduction of the Human as utility and measured by people’s actual choices, for Development Index (HDI)—measuring the capability to that would obscure real differences in how individuals live a long and healthy life, to acquire knowledge and to use income for achievements that they value.2 Instead, earn income for a basic standard of living.5 The HDI was capabilities are people’s freedoms to choose what meant to be a metric of a very minimal list of capabili- they want to be and do—regardless of whether they ties, “getting at minimally basic quality of life.”6 It was actually make those choices. Thus, capabilities are never a statistic to be maximized, as in aggregate utility. closely related to the concept of opportunities: It is not It was computed at the country level, mostly because of enough to know that someone has not travelled to a data availability, and was meant to enrich the assess- foreign country; we need to know whether that was a ment of countries’ development performance.7

1. Sen (1980) went further than Rawls’s social primary commodities, with essentially the same argument—that these are, at best, instrumental. 2. More precisely, Sen (1980) was showing the limitations of utilitarianism as a normative principle to adjudicate welfare. In utilitarianism, social welfare is assessed based on the actual choices that people make. People are assumed to maximize their individual utility—an increasing function of income, but one that yields less utility the higher the income. So achieving the ideal social welfare implies maximizing the sum total of utility in a society. That, in turn, can happen only if income is distributed so that individual marginal utility is equalized. Sen used a well known and compelling illustration to show how this principle could result in outcomes that violate our sense of fairness. Consider two individuals: One, who lives with a disability, is not very efficient in turning an additional dollar of income into utility; another, in contrast, derives satisfaction from every single additional dollar. Utilitarianism would dictate giving more income to the second person, an outcome that violates our sense of fairness. 3. Basu and Lopez- Calva 2011. 4. Stewart, Ranis and Samman 2018. 5. Sen (2005) credits joint work with Mahbub Ul Haq to develop a general index for global assessment and critique, going beyond gross domestic product (GDP). 6. Sen 2005. 7. Perhaps more important, quoting Klasen (2018, p. 2), “Many of the battles of the 1990s that came to define the Human Development Reports have been won. Today, the entire development community accepts that development is more than increasing per capita gross domestic product (GDP)… The HDI has been canonized in all standard textbooks on development economics or development studies … and is considered the most serious and comprehensive alternative to GDP per capita. […]”

Moving targets in the 21st century | 31 Initial stepping stones, at birth). To motivate the empirical information capabilities can be framed in the context of a li- such as surviving to considered, a lifecycle approach is used, given fecycle analysis (which is also used in chapter 2 that achievements in human development when analysing the mechanisms leading up to age 5, learning to read build over a person’s life through a sequence of the emergence of inequalities in capabilities). and doing basic math observable and measurable indicators. Initial Later in the Report the same patterns will be are crucial to further stepping stones, such as surviving to age 5, learn- illustrated in two other dimensions: human development: These ing to read and doing basic math are crucial to security in the face of shocks linked to trends further development: These basic achievements on climate change (chapter 5) and technology basic achievements present some of the necessary conditions for cre- (chapter 6).13 These drivers of the distribution present some of the ating further capabilities in life.10 The enhanced of capabilities in the 21st century are consid- necessary conditions achievements that follow, such as a long and ered without implying that others, such as for creating further healthy adult life or tertiary education, reflect demographic changes, are unimportant or that more advanced access to opportunities. they are the only two that matter, but to allow capabilities in life. for a treatable elaboration of the arguments The enhanced While these observable achievements are what showing the relevance of analysing the ine- can be measured (and compared across countries quality dynamics in both basic and enhanced achievements that in a global report), they are taken to represent a capabilities. follow, such as a long wider set of capabilities that also range from ba- and healthy adult life sic to enhanced. Emphasis should be placed on Admittedly, constraining the analysis to these or tertiary education, the underlying concept of basic and enhanced four dimensions is arbitrary. And in no way reflect more advanced capabilities over the specific measurements, should these aspects be regarded as the most access to opportunities which can evolve and change from country to important or have any normative meaning. But country. Here the inspiration is Amartya Sen’s it is plausible to claim that the distribution and definition of a basic capability as “the ability to evolution of capabilities across these four dimen- satisfy certain elementary and crucially impor- sions will be paramount in determining people’s tant functionings up to certain levels.”11 Basic agency over the 21st century—that is, “the abil- capabilities thus refer to the freedom to make ity to decide on and the power to achieve what choices necessary for survival and to avoid or they want.”14 These capabilities, while essential escape poverty or other serious deprivations. for agency, are not their sole determinants be- cause human motivations are not driven exclu- The differentiation between basic and en- sively by improvements in one’s own well-being; hanced capabilities is valid also for other human “people’s sense of fairness and concern that they development dimensions that are not necessarily and others be treated fairly”15 also matter. While tied to an individual lifecycle—for example, in a full treatment of the implications of these the progression from basic to frontier technolo- broader determinants of agency is beyond the gies and in the ability to cope with environmental scope of the Report, this chapter concludes with shocks, from perhaps frequent but low-impact a section that looks at perceptions of inequality events to large and unpredictable hazards. (which could indicate how a sense of fairness, or lack thereof, is evolving) as well as some of the This distinction between basic and enhanced social and psychological underpinnings of how capabilities resembles the analysis of practical these perceptions may emerge and how they needs and strategic needs in the context of connect with human dignity. gender empowerment, pioneered by Caroline Moser.12 Associated with the distinction is a Dynamics of inequality in human cautionary message: While investment in basic development: Convergence in needs is essential, to focus on them exclusively basic capabilities, divergence is to neglect inequalities in strategic aspects in enhanced capabilities of life, those that change the distribution of power. On each of the four dimensions considered in the Report, it is possible to identify a differen- Thus, the next section presents a stylized tiation in capabilities, from basic to enhanced analysis along two key dimensions beyond in- (figure 1.3): come: health and access to knowledge—both core dimensions of the human development approach since the first Human Development Report. The sequence from basic to enhanced

32 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 1.3 Human development, from basic to enhanced capabilities

Enhanced Examples of achievements

• Access to quality health at all levels
• High-quality education at all levels
• Effective access to present-day technologies
• Resilience to unknown new shocks

Examples of achievements Basic

• Early childhood survival
• Primary education Inequalities and
• Entry-level technology unfairness persist.
• Resilience to recurrent shocks Human development

inequalities remain Source: Human Development Report Office. widespread. Convergence appears

• Health. From, for example, the ability to death, and others to access to knowledge and in basic capabilities.

survive the first years of life to the prospect of to life-changing technologies. Across coun- Those at the bottom tries the world remains deeply unequal in key are catching up in the enhanced healthy longevity. areas of human development in both basic basics. Divergence

• Education and knowledge. From, for example, and enhanced capabilities (figure 1.4). There appears in enhanced

is a difference of 19 years in life expectancy at capabilities. Gaps in having basic primary education to accessing a birth between low and very high human devel- enhanced capabilities opment countries, reflecting gaps in access to exceed those in the high-quality learning experience at all levels. health. That represents a quarter of a lifespan basic ones or are rising

• Human security in the face of shocks. From lost just for being born in a poor country. The

differences tend to remain over the lifecycle. the daily lack of freedom from fear where The differences in life expectancy at age 70, is almost 5 years, representing a third of the re- interpersonal violence is rampant to facing maining lifespan lost. The percentage of adults with a primary education is 42 percent in low the consequences of conflict. The ability to human development countries, compared with 94 percent in very high developing countries. face recurrent shocks and the capabilities to Again, gaps remain through the lifecycle: Only 3 percent of adults have a tertiary education in deal with uncertain events linked to climate low human development countries, compared with 29 percent in developed countries. In change are addressed in chapter 5. access to technology, there are 67 mobile-cel-

• Access to new technologies. From entry-level to lular subscriptions per 100 inhabitants in

developing countries, half the amount in very more advanced ones (discussed in more de- high human development countries. In more advanced technologies, such as access to fixed tail in chapter 6, with some results presented broadband, there is less than one subscription per 100 inhabitants, compared with 28 in very in this chapter). high human development countries.

Cutting across key human development di- The same is true within countries. One way to capture within-country inequalities in key mensions are the section’s three main findings: areas of human development is through the

• Inequalities and unfairness persist. Human

development inequalities remain widespread.

• Convergence appears in basic capabilities.

Those at the bottom are catching up in the

basics.

• Divergence appears in enhanced capabilities.

Gaps in enhanced capabilities exceed those

in the basic ones or are rising (or in some

cases, both).

First, inequalities persist and are widespread.

Across all dimensions considered there are

significant inequalities in constitutive areas of

human development: Some refer to life and

Moving targets in the 21st century | 33 FIGURE 1.4

The world remains deeply unequal in key areas of human development in both basic and enhanced

Basic Enhanced

Life expectancy at birth, 2015 Life expectancy at age 70, 2015

72.9 78.4 14.6 59.4 66.6 Health 11.1

Low Medium High Very high Low Medium High Very high Human development group Human development group

Population with a primary education, 2017 Population with a tertiary education, 2017 (percent) (percent)

93.5 28.6 66.5 13.7

Low Medium High Very high 3.2

Mobile-cellular subscriptions, 2017 Fixed broadband subscriptions, 2017 (per 100 inhabitants) (per 100 inhabitants)

90.6 Access to 67.0 technology

Low Medium High Very high Low Medium High Very high Human development group Human development group

Source: Human Development Report Office calculations based on data from the International Telecommunication Union, the United Nations Educational, Scientific and Cultural Organization Institute for Statistics and the United Nations Department of Economic and Social Affairs.

34 | HUMAN DEVELOPMENT REPORT 2019 Inequality-adjusted Human Development have been “escaping” from the imprisonment While there is catching Index (IHDI), which adjusts the HDI value for of extreme deprivations, to use Angus Deaton’s up in the basics, this inequality within countries in each of its compo- expression.21 This chapter also documents that is happening years nents (health, education and income). According this is an unfinished business, as the challenge of after the wealthier to the IHDI, the global average loss in human reaching those furthest behind persists. segments of society development due to inequality is 20 percent. exhausted the space to While there is catching up in the basics, this is make further progress Second, on average, there is convergence in happening years after the wealthier segments of on the same fronts basic capabilities. Inequality in the basic capa- society exhausted the space to make further pro- bilities of human development included in the gress on the same fronts. People at the top of the HDI is falling. This can be seen in the evolution distribution typically have reached the limit of of the IHDI, where indicators representing basic progress in basic capabilities: Universal coverage capabilities have very high implicit weights.16 in primary education and secondary education, In all regions of the world the loss in human very low infant mortality rates and access to ba- development due to inequality is diminishing sic technology are now taken for granted among (figure 1.5). This trend is repeated in many better-off segments of most societies. They are subnational HDI values17 and has happened looking towards more advanced goals. What is against a backdrop of aggregate development happening in these enhanced areas? progress across achievements representing basic capabilities on multiple fronts.18 The global ex- Third, there is divergence in enhanced ca- treme poverty rate fell from 36 percent in 1990 pabilities. Inequality is typically higher across to 9 percent in 2018.19 Infant mortality rates enhanced capabilities, and when it is not, it have been falling consistently. Primary school is growing. In each of the key dimensions of enrolment rates have seen great strides, with uni- human development considered—health, ed- versal coverage in most countries, and secondary ucation, living standards, access to technology education is making rapid progress (though the and security—groups converging in basic capa- substantive significance of these achievements bilities lag behind in access to enhanced capa- needs to be seen in the context of an imped- bilities. Greater ambitions are defining moving ing “learning crisis,” as discussed later in the targets. Yet this set of enhanced achievements chapter).20 The number of people living in low will increasingly determine people’s lives in this human development countries is 923 million century, in part because they are linked to some today, down from 2.1 billion in 2000. People of the most consequential change drivers of our time: technology and climate change.

FIGURE 1.5

In all regions of the world the loss in human development due to inequality is diminishing, reflecting progress in basic capabilities

Loss in human development due to inequality 2010 2018 35.1 (percent) 30.5 25.3 22.3 29.6 27.4 25.9 23.4 World 24.5 20.2 21.9

16.6 16.1

11.7

Arab States East Asia and Europe and Latin America and South Asia Sub-Saharan the Pacific Central Asia the Caribbean Africa

Source: Human Development Report Office calculations.

Moving targets in the 21st century | 35 FIGURE 1.6 Convergence in basic capabilities, divergence in enhanced capabilities

Basic Enhanced

Declining inequality Increasing inequality Change between 2005 and 2015 (years) Life expectancy at age 70 5.9 Change between 2005 and 2015 (years)

4.9 1.2

2.7 2.4 0.5

Low Medium High Very high Low Medium High Very high Human development group Human development group

Population with a primary education Population with a tertiary education Change between 2007 and 2017 (percentage points) Change between 2007 and 2017 (percentage points)

5.3 Education

3.0 1.1

Low Medium High Very high Low Medium High Very high Human development group Human development group

Mobile-cellular subscriptions Fixed broadband subscriptions Change between 2007 and 2017 (per 100 inhabitants) Change between 2007 and 2017 (per 100 inhabitants)

59.5 12.3 49.3 49.3 Access to 26.1 Low Medium High Very high

Low Medium High Very high Human development group

Source: Human Development Report Office calculations based on data from the International Telecommunication Union, the United Nations Educational, Scientific and Cultural Organization Institute for Statistics and the United Nations Department of Economic and Social Affairs.

36 | HUMAN DEVELOPMENT REPORT 2019 Figure 1.6 summarizes the emerging human BOX 1.2 People born in development divide with pairs of indicators, very high human measuring progress over the last decade in one Article 25 of the Universal Declaration of Human development countries basic and one enhanced indicator for each of Rights: The right to a basic standard of living are expected to three key human development dimensions: live almost 19 more health, education and access to technologies. “Everyone has the right to a standard of living ade- years (or almost a Across human development groups there are quate for the health and well-being of himself and of third longer) than two opposing trends in gradients for basic and his family, including food, clothing, housing and med- people in low human enhanced capabilities. Inequalities are falling in ical care and necessary social services, and the right development countries basic capabilities because lower human devel- to security in the event of unemployment, sickness, opment countries are making larger progress disability, widowhood, old age or other lack of liveli- on average. When the ones that are behind hood in circumstances beyond his control. grow faster, there is convergence. By contrast, inequalities are growing in enhanced capabili- “Motherhood and childhood are entitled to spe- ties because high and very high human devel- cial care and assistance. All children, whether born opment countries are getting ahead, leading to in or out of wedlock, shall enjoy the same social divergence. The Report documents later that protection.” these trends are also observed within countries. Source: www.un.org/en/universal-declaration-human-rights/. The basic indicators in the figure all reflect narrowing inequalities between countries in Inequalities in health outcomes are different human development groups. For widespread instance, in life expectancy at birth (driven mainly by survival to age 5), in access to prima- Life expectancy at birth is a helpful indicator ry education and in access to mobile phones, to track health inequalities. As one of the lower human development countries are mak- three components of the HDI, it has been ing faster progress. They are catching up with used as a proxy for long and healthy life since higher human development countries. the first Human Development Report in 1990. In contrast, the more advanced indicators in the figure reveal widening inequalities. Higher Here, the analysis extends life expectancy human development countries start with an beyond that at birth to that at different ages advantage in life expectancy at age 70, in ter- in order to identify the dynamics of health tiary education enrolment and in broadband through the lifecycle. This lifecycle approach access—and they are increasing their lead in makes it possible to capture changes in both these areas. The effect of these widening gaps— the demographic and the socioeconomic representing just few examples of enhanced transitions. And it shows how, across various capabilities—will be revealed over the 21st indicators, not only do deep inequalities century. And that effect will impact those born persist, but new gaps are also opening. Life today, many of whom will see the 22nd century. expectancies—both at birth and at older The remainder of this section considers the ages—are considerably higher in countries dynamics of convergence and divergence in with higher income or higher human devel- health and education in more detail. opment (figure 1.7)—this is often called a health gradient. People born in very high Health: The well-off are living healthier human development countries are expected and longer in the 21st century to live almost 19 more years (or almost a third longer) than people in low human de- Health inequalities can be a clear manifestation of velopment countries.22 People at age 70 in social injustice (see chapter 2 for a more detailed very high human development countries are discussion). These inequalities also reflect short- expected to live almost 5 more years (around comings in meeting basic human rights, such 50 percent longer) than people in low human as those defined by article 25 of the Universal development countries. The gaps are also very Declaration of Human Rights (box 1.2). large when the quality of health is considered (box 1.3).

Moving targets in the 21st century | 37 FIGURE 1.7 Inequalities persist in life expectancy and mortality

Inequalities in life expectancy

Life expectancy at birth Life expectancy at age 70 13.414.6 (years) (years)

70.3 72.9 76.0 78.4 12.6 11.8 59.4 61.8 66.6 10.411.1 53.6 9.2 9.8

Low Medium High Very high Low Medium High Very high

Human development group Human development group

2005 2015

Inequalities in mortality

Probability of death by age 5 Probability of death at ages 70­79 (percent) (percent)

13.2 58.354.7 51.4 8.8 47.5 43.8 Low 7.5 39.7 35.2 29.8

3.2 1.0 0.6 2.1

Medium High Very high Low Medium High Very high

Human development group Human development group

Source: Human Development Report Office calculations based on data from the United Nations Department of Economic and Social Affairs.

BOX 1.3 Inequality in healthy life expectancy

While the length of life is important for human de- United Arab Emirates but 724 per 100,000 in Lesotho. velopment, equally essential is how those years are The HIV prevalence rate among adults is 27.2 percent lived. Are they enjoyable? Does health remain good? in the Kingdom of Eswatini but only 0.1 percent in many The indicator of healthy life expectancy suggests large very high human development countries, among them discrepancies. Healthy life expectancy for very high Australia, Bahrain, Kuwait and Romania.2 Malaria has human development countries is about 68 years, com- been defeated in Sri Lanka and is projected to be de- pared with only about 56 years for low human develop- feated in 2020 in Argentina, Belize, Costa Rica, Ecuador, ment countries.1 El Salvador, Mexico, Paraguay and Suriname.3 But prev- alence is still high in Mali, with 459.7 cases per 1,000 A look at some specific diseases can shed some people at risk, and Burkina Faso, with 423.3.4 In May light on causes of inequalities in life expectancy and 2019, 1,572 people in the Democratic Republic of Congo healthy life expectancy. The prevalence of tuberculo- suffered from Ebola.5 sis, for example, is only 0.8 per 100,000 people in the

1. See Statistical table 8 at http://hdr.undp.org/en/human-development-report-2019. 2. UNDP 2018a. 3. WHO 2017. 4. UNDP 2018a. 5. WHO 2019.

38 | HUMAN DEVELOPMENT REPORT 2019 Catching up in the basics: Global both at young ages (ages 0­5) and at older ages Low human convergence in life expectancy at birth, (ages 70­79) (figure 1.10). While the level of development countries especially through reduced infant mortality inequality in mortality rates is much higher at gained almost 6 years young ages than at older ages, the changes in of life expectancy at The increase in life expectancy at birth—from a mortality rates reflect different patterns. Child birth between 2005 weighted average of 47 years in the 1950s to 72 mortality rates converge—dropping faster for and 2015, compared years around 2020—portrays the extraordinary lower human development countries—just as with 2.4 years for progress in health.23 In 2000 several countries still mortality rates at older ages diverge. very high human had life expectancy at birth below 50 years, a cat- development countries egory expected to disappear from every country If the countries performing poorly in 2005 are average by 2020.24 The improvement has been the ones with greater progress over 2005­2015, observed across human development groups (see there is catching up or convergence. But if the figure 1.7). Moreover, low human development countries with worse performance in 2005 are countries gained almost 6 years of life expectan- the ones with less improvement over 2005­ cy at birth between 2005 and 2015, compared 2015, there is divergence. Different patterns with 2.4 years for very high human development can be observed with different definitions of life countries (figure 1.8, left panel). This is consist- expectancy: going from clear convergence in life ent with a reduction of more than 4 percentage expectancy at birth to clear divergence in life ex- points in under-five mortality rates in low human pectancy at age 70 (see figure 1.8, right panel).27 development countries. Another area with signif- icant reduction is maternal mortality, which fell Inequalities in life expectancy at older ages 45 percent between 1990 and 2013.25 are an emerging form of inequality in human development in the 21st century. Divergence A detailed look at the situation within develop- in life expectancy at older ages is much stronger ing countries confirms these trends. To facilitate today than during the second half of the 20th meaningful comparability, figure 1.9 groups the century.28 And since the turn of the century, within-country results (information per quintile life expectancy at older ages has been increasing in 54 countries), according to their human de- much faster in very high human development velopment level. Consider infant mortality rates, countries than elsewhere. During 2005­2015 an important determinant of life expectancy at life expectancy at age 70 increased 0.5 year in birth. They have been declining everywhere, but low human development countries and 1.2 years significant gradients remain: Children born in in very high human development countries. poorer quintiles have a much higher probability of dying during the first year of life than those Improvements in technologies, enhanced born in wealthier quintiles. This is the case across social services and healthy habits are moving the all human development groups. frontiers of survival at all ages. While the space for reducing mortality under age 5 is shrinking The convergence in mortality rates at younger fast, it remains large at older ages (under age ages is also confirmed within countries: Infant 80).29 An important factor behind different mortality appears to be falling for all segments mortality rates at older ages are variations in of the population, and in most countries the noncommunicable disease rates across different greatest reductions in infant mortality are in the groups. People with lower socioeconomic status poorest three quintiles. This result is consistent or living in more marginalized communities are with the decline in the dispersion of life expec- at higher risk of dying from a noncommunicable tancy at birth documented in an analysis of more disease.30 than 1,600 regions in 161 countries, covering more than 99 percent of the world population.26 The world is getting older fast. People over age 60 are the fastest growing age segment of Growing inequalities in enhanced the global population. By 2050, one person capabilities: Divergence in life in five worldwide is expected to be in this expectancy at older ages age group; in more developed regions the proportion is expected to be one in three.31 Consider the levels and the evolution of average Therefore, the relevance of inequalities linked mortality rates for different groups of countries, to older people will grow.

These between-country results are consistent with emerging evidence from within-country

Moving targets in the 21st century | 39 FIGURE 1.8

The changing inequality in life expectancy, 2005­2015: Low human development countries catching up in life expectancy at birth but lagging behind in life expectancy at older age

Declining inequality Increasing inequality Change between 2005 and 2015 (years) Life expectancy at age 70 5.9 Change between 2005 and 2015 (years)

4.9 1.2

0.7 0.8

2.7 2.4 0.5

Low Medium High Very high Low Medium High Very high

Human development group Human development group

Basic Enhanced

Convergence Divergence

Life expectancy at birth Life expectancy at age 70 Change in life expectancy at birth, between 2005 and 2015 (years) Change in life expectancy at age 70 8 between 2005 and 2015 (years) 6 De ve lo ped

4 South Asia East Asia and the Pacific Latin America

East Asia and the Pacific De ve lo ped

2 Arab States 0.8 South Asia Latin America Sub-Saharan

0 0.4

50 55 60 65 70 75 80 9 11 13 15

Life expectancy at birth, 2005 (years) Life expectancy at age 70, 2005 (years)

Note: Convergence and divergence are tested for in two ways: by using the slope of an equation that regresses the change over 2005­2015 with respect to the initial value in 2005 (with ordinary least squares, robust and median quantile regressions) and by comparing the gains of very high human development countries and the gains of low and medium human development countries. For life expectancy at birth there is convergence according to both metrics (p-values below 1 percent). For life expectancy at age 70 there is divergence according to both metrics (p-values below 1 percent). Source: Human Development Report Office calculations based on data from the United Nations Department of Economic and Social Affairs.

40 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 1.9 Infant mortality rates, an important determinant of life expectancy at birth, have been declining everywhere, but significant gradients remain

Deaths per 1,000 births 2007 2017

103.6 102.0 93.4 85.0 62.4 81.2 79.0 72.7 63.9 54.0 55.3 ­33.0 ­33.0 ­26.9 ­22.8 ­14.4 ­21.4 ­27.7 ­25.5 ­24.1 ­20.8 ­18.4 70.6 68.9 66.5 59.8 51.3 47.2 39.8 33.2 36.9 62.2 48.0 48.0 ­19.0 40.1 34.1 27.5 29.0 ­15.8 ­11.3 ­10.3 24.3 22.8 17.2

Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5

Low human development Medium human development High/very high human development

Note: Data for 2007 refer to the most recent year available during 1998­2007, and data for 2017 refer to the most recent year available for 2008­2017. Data are simple averages across human development groups. Only one very high human development country (Kazakhstan) is included in the sample. Quintiles reflecting within-country distribution of assets are grouped by human development groups. Source: Human Development Report Office calculations based on data from the United Nations Department of Economic and Social Affairs.

FIGURE 1.10 Mortality: Convergence in basic capabilities, divergence in enhanced capabilities

Concentration curves (2015)

Probability of death by age 5, 2015 Probability of death at ages 70­79, 2015

Cumulative Cumulative outcome proportion outcome proportion

1.0 1.0

0.8 0.8

0.6 0.6

0.4 0.4

Perfect equality

0.2 0.2

0.0 0.0 20 40 60 80 100 0 Percent of population

Probability of death by age 5 Probability of death at ages 70­79 Change between 2005 and 2015 (percentage points) Change between 2005 and 2015 (percentage points)

Human development group Human development group Low Medium High Very high Low Medium High Very high

­0.3 ­1.1

­2.6 ­3.6 ­3.9 ­4.0 ­4.5

­5.5

Source: Human Development Report Office calculations based on data from ICF Macro Demographic and Health Surveys and United Nations Children’s Fund Multiple Indicator Cluster Surveys.

Moving targets in the 21st century | 41 Education is expanding studies. In the United States higher income is Inequalities in education are widespread in most countries, associated with greater longevity. And inequality across all levels of of life expectancy has increased in recent years. Education is expanding in most countries, development. But Between 2001 and 2014 individuals in the top across all levels of development. But inequality inequality remains 5 percent of the income distribution gained remains in both enrolment among younger in both enrolment more than 2 years of life expectancy at age 40, generations and adults’ education attainment. among younger while lifespans in the bottom 5 percent re- On average, the lower a country’s human mained nearly unchanged.32 The importance of development, the larger the gap in access to generations and adults’ socioeconomic factors is highlighted by the fact education (figure 1.11).44 For low and very education attainment that life expectancy at age 40 among low-income high human development countries the gaps in people (the bottom quartile) varies by about 4.5 enrolment ratios range from 20 points for pri- years across cities: Low-income individuals in mary education to 58 points for secondary and affluent cities with highly educated populations tertiary education to 61 points for preprimary and high government expenditures, such as education. New York and San Francisco, tend to live longer (and to have healthier lifestyles) than elsewhere. Gaps in access to education among children Those cities also experienced the largest gains and youth are also large within countries (fig- in life expectancy among poor people during ure 1.12). Across levels of human development, the 2000s. Finally, differences in life expectancy the bottom income quintiles nearly always limit redistribution because low-income individ- have less access to education, except for pri- uals obtain benefits from social programmes for mary education in high and very high human fewer years than high-income individuals do.33 development countries, where access is already universal. Other studies show increasing inequalities in life expectancy in Canada,34 Denmark,35 Catching up in the basics: Convergence Finland,36 Japan,37 the United Kingdom,38 the in primary education but not fast enough United States39 and some Western European countries.40 The literature on developing and Inequality is usually smaller in primary and emerging countries is very limited.41 In Chile the secondary education, and most countries increase in inequalities in life expectancy at older are on track to achieve universal primary ages between 2002 and 2017 is linked to the education, which represents the potential socioeconomic status of municipalities (box 1.4). acquisition of basic capabilities. Enrolment in secondary education is nearly universal These emerging inequalities reflect how ad- in very high human development countries, vances in longevity are leaving broad segments while in low human development countries of people behind. More detailed analyses are only about a third of children are enrolled. necessary to identify determinants and policy The success in reducing inequality is captured actions to ensure that the fruits of progress are by concentration curves, showing equality as within reach of everyone. But if these trends proximity to the diagonal (figure 1.13, top are not reversed, they will lead to increased in- panel). Inequality in primary and secondary equality in the progressivity of public policies education has been falling over the past dec- focused on supporting older citizens.42 ade. People in countries with initially low enrolment (predominantly low and medium Education: Increasing access but with human development countries) have seen the widening inequality in capabilities highest increases on average (see figure 1.13, bottom panel). Trends in education attain- Through education students from disadvan- ment are similar: There is a strong reduction taged backgrounds can improve their chances in gaps in primary education (figure 1.14). of social mobility. But for children who leave But these are averages, and convergence is not the school system early or do not receive a equally strong in all contexts because some high-quality education, gaps in learning can groups are being left behind (as discussed later become a trap with lifetime and even inter in this chapter). generational implications.43

42 | HUMAN DEVELOPMENT REPORT 2019 BOX 1.4 Divergence in life expectancy at older ages in Chile

Chile has historically been an unequal country in terms unevenness of progress in health across the country. of income, with a Gini coefficient of 0.50 in 2017 (of- Advances in healthy life are taking place, but they are ficial figures from the CASEN Survey). For life expec- not reaching all social groups and territories equitably. tancy at older ages, inequality is significant as well. Second, there are potentially regressive distributive In Santiago Metropolitan Region, people living in the effects through the pension system, which ties retire- wealthier comunas (municipalities) have a higher life ment benefits to the amount of money accumulated in expectancy at age 65—more than 2 years on average an individual savings account and to the life expectancy (those at the upper right in the figure). There has been after retirement—that is currently common across so- generalized improvement in life expectancy over the cial groups. last 15 years (between the 2002 and 2017 censuses). However, the differences between comunas are per- This example shows the importance of comprehen- sistent and, indeed, have increased. Today, in terms sive analysis of inequalities using the human develop- of life expectancy at older ages, there is little overlap ment lens, going beyond income (assessing the health between the situation of the wealthier comunas and dimension), beyond averages (looking at disaggregated the rest. data in different areas) and beyond today (covering in- equalities expected to become more important in the There are multiple implications of the divergence years to come). This new look at emerging inequalities in life expectancy at older ages. First, they reflect the is essential for the design of policies.

People living in the wealthiest comunas in the Santiago Metropolitan Region have, on average, increased their already higher life expectancy at older ages more than people living in poorer comunas have

Life expectancy at age 65, 2002 Life expectancy at age 65, 2017

24 24

22 Las Condes Vitacura Las Condes 22

Providencia Vitacura Ñuñoa La Reina Providencia 20 Ñuñoa

La Reina

18 18

16 16

12.5 13.0 13.5 14.0 14.5 12.5 13.0 13.5 14.0 14.5

Log of household income per capita, 2017 Log of household income per capita, 2017

Source: Based on Hsu and Tapia (2019).

Moving targets in the 21st century | 43 FIGURE 1.11 The lower a country’s human development, the larger the gap in access to education

Preprimary Primary 84.3 86.5 92.3 93.3 Enrolment ratio Enrolment ratio 95.2 95.4 (net) (net)

60.5 77.9 68.7 75.3 50.9 70.1

41.6 33.1

10.316.4

Low Medium High Very high Low Medium High Very high

Human development group Human development group

2007 2017

Secondary Tertiary

Enrolment ratio Enrolment ratio (net) (gross)

72.478.8 87.9 91.0

59.4 66.8 48.8 55.1

32.8 44.6 22.4 33.4

5.3 8.3 20.1 Low 12.7

Low Medium High Very high Medium High Very high

Human development group Human development group

Note: Data are simple averages of country-level data. Source: Human Development Report Office calculations based on data from the United Nations Educational, Scientific and Cultural Organization Institute for Statistics.

FIGURE 1.12

Gaps in access to education among children and youth are also large within countries

Attendance rate Primary Secondary Tertiary (percent) (net) (net) (gross)

0

Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5

Low Medium High/very high Low Medium High/very high Low Medium High/very high

Human development group Human development group Human development group

Note: Only one very high human development country (Montenegro) is included in the sample. Data are for 2016 or the most recent year available. Quintiles are based on distribution of ownership of assets within countries. Source: Human Development Report Office calculations based on data from ICF Macro Demographic and Health Surveys, United Nations Children’s Fund Multiple Indicator Cluster Surveys and the World Bank.

44 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 1.13 Inequality in primary and secondary education has been falling over the past decade

Inequalities in basic capabilities are lower and falling (convergence): But inequalities in enhanced capabilities are large and This is the case of enrolment ratios in primary and secondary education. growing (divergence): Low human development countries are catching up with high and very Inequalities in enrolment ratios in preprimary education and tertiary high human development countries. education are high or growing.

Concentration curves (2017)

Primary Secondary Preprimary Tertiary

Cumulative Cumulative Cumulative Cumulative outcome proportion outcome proportion outcome proportion outcome proportion 1.0 1.0 1.0 1.0 Perfect equality 0.8 0.8 0.8 0.6 0.6 0.6 0.8 0.4 0.4 0.4 0.6 0.2 0.2 0.2 0.0 0.0 0.0 0.2 0 20 40 60 80 100 0 20 40 60 80 100 0 20 40 60 80 100 Percent of population Percent of population Percent of population

Basic Enhanced

Change in enrolment ratio between 2007 and 2017 (percentage points)

Primary Secondary Preprimary Tertiary 11.2 11.6 6.6 10.4 10.5 9.5 6.4 7.4 2.2 3.1 8.5 7.8 1.0 0.2 3.0 6.0 Human development group Low Medium High Very high Low Medium High Very high Low Medium High Very high Human development group Human development group Human development group

Note: Concentration curves are ordered by Human Development Index value. Source: Human Development Report Office calculations based on country-level data from the United Nations Educational, Scientific and Cultural Organization Institute for Statistics.

Moving targets in the 21st century | 45 FIGURE 1.14 Increasing inequality Dynamics of education attainment, 2007­2017 Percent of population with a tertiary education Declining inequality Change between 2007 and 2017 (percentage points) Percent of population with a primary education Change between 2007 and 2017 (percentage points) 7.1

9.2 6.2 5.9

5.3

1.1

Low Medium High Very high Low Medium High Very high

Human development group Human development group

Basic Enhanced

Convergence Divergence

Percent of population with a primary education Percent of population with a tertiary education Change between 2007 and 2017 Change between 2007 and 2017 (percentage points) (percentage points) 12 14

10 Latin America East Asia and the Pacific Sub-Saharan Africa and the Caribbean 12 8 Europe and Central Asia South Asia East Asia and the Pacific 10

6 8

Latin America De ve lo ped

6 and the Caribbean

4 4 Arab States

2 2 Sub-Saharan Africa Developed

0 0

40 50 60 70 80 90 100 0 5 10 15 20 25

Percent of population with a primary education, 2007 Percent of population with a tertiary education, 2007

Note: Convergence and divergence are tested for in two ways: by using the slope of an equation that regresses the change over 2007­2017 with respect to the initial value in 2007 (with ordinary least squares, robust and median quantile regressions) and by comparing the gains of very high human development countries and the gains of low and medium human development countries. For attainment of primary education there is convergence according to both metrics (p-values below 1 percent in all regressions and below 5 percent in the comparison between human development groups). For attainment of tertiary education there is divergence according to both metrics, with different significance levels in regressions: the parameter is positive in all cases; it is not statistically significant in the ordinary least squares regression, but it is statistically significant in the robust regression (p-value below 10 percent) and the median quintile regression (p-value below 1 percent) and for the comparison between human development groups (p-value below 5 percent). Source: Human Development Report Office calculations based on country-level data from the United Nations Educational, Scientific and Cultural Organization Institute for Statistics.

46 | HUMAN DEVELOPMENT REPORT 2019 Growing inequalities in enhanced The unevenness in distribution has conse- Data for 47 developing capabilities: Gaps in tertiary education quences for human development. The largest countries show and in preprimary education are wide gaps appear in the formation of enhanced ca- divergence in and increasing pabilities, which are the areas with the highest the acquisition of returns: in preprimary education, with the enhanced capabilities: Inequalities in preprimary education and post- highest social returns,45 and in tertiary educa- Quintiles with secondary education are high and, in many tion, with the highest private returns.46 This higher access to places, growing. Concentration curves reflect analysis considers preprimary education an postsecondary how these achievements are more unevenly dis- enhanced achievement because of its impor- education 10 years tributed for preprimary and tertiary education tance and because societies have come to ac- ago have seen the (see figure 1.13, right side). Moreover, the gaps knowledge its importance only in recent years. largest gains are growing on average: Low human develop- The inequalities in the formation of enhanced ment countries—already behind—tend to have capabilities pave the way to future inequality slower progress. throughout the lifecycle, particularly in access to work opportunities and income.47 These trends—of convergence in basic education and divergence in enhanced educa- The distinction between basic and enhanced tion—are not destiny; there is heterogeneity, capabilities in education depends on the effect reflecting space for policies. Taking information of various achievements on what people can about attainment, for instance, East Asia and do. The large and widening gaps not only show the Pacific and Europe and Central Asia have differentiated access to tertiary education and made notable progress in expanding tertiary ed- its direct impact on access to learning; they also ucation, closing in on developed countries (see determine inequalities in the availability of pro- figure 1.14). However, the other regions follow fessionals between and within countries, with the overall trend, with Sub-Saharan Africa effects on multiple areas of human development. catching up very strongly in primary education For instance, the inequalities in the availability and lagging behind in tertiary education. of physicians are widening between countries. High and very high human development coun- Data for 47 developing countries show tries had significantly more physicians per capita divergence in the acquisition of enhanced ca- in 2006 and have, on average, increased the gaps pabilities: Quintiles with higher access to post- between themselves and low and medium hu- secondary education 10 years ago have seen the man development countries (figure 1.16). largest gains (figure 1.15).

FIGURE 1.15 Inequalities in postsecondary education within countries are growing

Mean change in gross attendance ratio

8

4

0

Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5

Low Medium High/very high

Note: Data are simple averages for each human development group. Only one very high human development country (Montenegro) is included in the sample. Quintiles are based on distribution of ownership of assets within countries. Source: Human Development Report Office calculations based on data from Demographic and Health Surveys and Multiple Indicator Cluster Surveys processed by the World Bank.

Moving targets in the 21st century | 47 FIGURE 1.16 status and conditions at home (such as access to books)—remains the strongest predictor of Widening inequalities in the availability of learning outcomes.50 physicians between countries The learning gradient compounds inequality Physicians per 1,000 people over inequality: Those from disadvantaged 2016 groups not only have fewer opportunities to re- 2006 ceive education; they also learn much less once in the classroom (figure 1.17). These socio- 0.3 3.1 economic inequalities have remained high and 2.7 stable over the past two decades in countries with a longer history of standard data.51 0.2 1.8

0.2 0.0 0.5 Convergence in the basics 0.1 0.5 is not benefiting everyone: 0.0 Identifying those furthest behind

Low Medium High Very high This chapter has documented convergence across basic capabilities. But does that imply While more than Human development group that the rising tide is lifting all boats? This 90 percent of section shows that, despite convergence, many Note: Data are simple averages for each human development group. people are excluded and remain stuck at the children in the world Source: Human Development Report Office calculations based on country-level very bottom of society. Convergence in basic today receive some data from the World Bank’s World Development Indicators database. capabilities is not absolute—advances in health schooling, fewer than and education within countries continue to half of those in school Growing inequalities in more empowering leave many behind. areas: The learning crisis achieve minimum Average convergence is not a sufficient con- proficiency in reading Education should mean ensuring that schooling dition to leave no one behind. Convergence and mathematics by the leads to learning. But the great education ex- can be characterized into four cases, from the end of primary school pansion has not translated into commensurate point of view of a particular group: gains in learning, where large inequalities exist. And much remains to be done—in many coun- FIGURE 1.17 tries achievement in learning is disturbingly low. While more than 90 percent of children in Harmonized test scores across human development the world today receive some schooling, fewer groups than half of those in school achieve minimum proficiency in reading and mathematics by the Harmonized test end of primary school.48 scores 600 The rapid expansion of education in devel- oping countries has led to the enrolment of 500 millions of first-generation learners, who lack support from their families when they fall 400 behind in the curriculum. Students who fall behind may struggle if the level of classroom 300 Medium High Very high instruction (based on textbooks that follow Low ambitious curricular standards) is considerably above their learning level.49 These problems are Human development group exacerbated at higher grades, if students are automatically promoted to the next grade with- Note: Each box plots the middle 50 percent of the distribution; the central line out having acquired foundational skills. Low is the median; the extreme lines are the approximate minimum and maximum of skills continue to undermine career opportu- the distribution. nities—and earnings—long after students leave Source: Human Development Report Office calculations based on country-level school. data from World Bank (2018b).

In nearly all countries, family background— including parent education, socioeconomic

48 | HUMAN DEVELOPMENT REPORT 2019

• Absolute convergence: the group catches up children are out of school at either the primary Today, 5.4 million

with respect to all those above. or secondary level; and nearly 600 million children, more than people around the globe still live on less than half of them newborns,

• Weak convergence: a group catches up on $1.90 a day.53 This suggests that those with low do not survive their

average with those at the top. human development face a double challenge. first five years of life; Part of the population has not met the basic at current rates of

• Simple divergence: a group records very slow set of human development capabilities in their progress there will

progress, so the average gap with those at the life expectancy, schooling and income. And a be around 3 million top increases. larger part is also falling behind the enhanced child deaths in 2030 capability set that revolves around higher

• Full divergence: there is a setback, with an thresholds of educational achievement, labour

increasing gap with respect to the rest and and digital skills. the initial situation. Two indicators from the HDI that are more Despite greater access to immunizations and affordable medical technologies, child mortali- linked to basic capabilities (life expectancy at ty rates in the poorest households of the world’s birth and mean years of schooling) can illus- poorest countries remain high (figure 1.18). trate the limits of average convergence. The The highest rates are concentrated in low and analysis is based on the share of the population medium human development countries. And in low, medium and high human development there are vast disparities within countries: countries converging (or not) to very high The poorest 20 percent in middle income human development achievements (table 1.1). Guatemala have the same average mortality rate Over 2007­2017 there was significant con- as in low income Senegal. vergence, but it was partial (only half the population) and mostly weak (only 0.3 percent At current rates of progress there will be achieved absolute full convergence). The differ- around 3 million child deaths in 2030. Most ence between absolute convergence and weak would be the result of eminently preventable convergence was consequential: the “lost” pro- causes rooted in poverty and unequal access gress in terms of life expectancy at birth was 2.8 to quality health care. Around 850,000 will years and in terms of mean years of schooling reflect the gap between the SDG target and was 0.7 year. By contrast, 36 percent of the pop- the outcomes on the current trajectory. Given ulation was in a mixed zone, with convergence that the ratio of deaths between the poorest in one variable and divergence in the other and the richest is more than 5 to 1, accelerating (yellow cells in table 1.1). And 14 percent of progress for the poorest children would act as the population was in the divergence zone (red a powerful catalyst for overall progress—and cells in table 1.1). this illustrates the power of convergence by moving up those at the bottom, which would The partial and weak convergence has impli- save 4.7 million lives between 2019 and 2030 cations for the future and for the achievement (figure 1.19). of the SDGs. Today, 5.4 million children, more than half of them newborns, do not sur- vive their first five years of life;52 262 million

TABLE 1.1

Limited convergence in health and education, 2007­2017 (percent of population in low, medium and high human development countries)

Mean years of schooling

Full Divergence Weak Absolute divergence convergence convergence

Life Full divergence 0.1 3.5 2.7 0.2 expectancy Divergence 0.2 1.7 Weak convergence 1.0 10.6 16.4 4.3 at birth Absolute convergence 0.0 0.3 12.9 42.8

1.4 1.7

Note: Estimates are population weighted with respect to performance of very high development countries. Source: Human Development Report Office calculations based on subnational data from Permanyer and Smits (2019).

Moving targets in the 21st century | 49 FIGURE 1.18 Child mortality converges with human development, but not for the poorest 20 percent

Under-five mortality rate, 2017 (per 1,000 live births)

Low human Medium human High human Very high human development development development development Poorest 20%

120

100 Mali

60 Richest 20% Senegal About 262 million

children and 40

youth were out of Guatemala

school in 2017

0

0.4 0.5 0.6 0.7 0.8 0.9 1.0

Human development index, 2018 (value)

Note: Colours represent human development thresholds. Each bubble represents a country, and the size of the bubble is proportional to the country’s population.

FIGURE 1.19 The leading causes of death among children under age 5 remain unaddressed. They include Some 846,000 of 3.1 million child deaths are preterm birth complications (18 percent of the preventable if the bottom 20 percent converge to global total), pneumonia (16 percent), compli- the country average cations during birth (12 percent), with congen- ital anomalies, diarrhoea, neonatal sepsis and Under-five mortality rate Average rate of reduction malaria each accounting for a further 5­10 per- (deaths per 1,000 live births) needs to accelerate cent.54 Targeted interventions in tuberculosis, from 2.7 percent to pneumonia and diarrhoea have some of the Rate of reduction 4.5 percent, saving highest return for reducing under-five mortal- 4.7 million lives ity in the developing world. And three-quarters 150 for poorest 20 percent of deaths among those ages 0­14 are from communicable, perinatal and nutritional con- needs to accelerate ditions.55 Lack of data is also an issue. Targeted interventions benefit from real-time record from 2.8 percent keeping, using home-based records to supple- ment health provider registries. Early adopters to 5.5 percent of electronic medical records—Peru, Kenya, Malawi and Haiti—show how information 100 systems can help with micro-targeting of those furthest behind. Staying in school remains a challenge at 846,000 the bottom of the global distribution. About deaths 262 million children and youth were out of Sustainable Development Goal minimum threshold target 2.1 million deaths

0 2000 2005 2010 2015 2020 2025 2030

Average Poorest 20 percent Richest 20 percent

Source: Fiala and Watkins 2019.

50 | HUMAN DEVELOPMENT REPORT 2019 school in 2017, 64 million of primary school was faster than the improvement of the total On current trends the age, 61 million of lower secondary school age population. This suggests overall convergence. out-of-school rate will and 138 million of upper secondary age.56 However, the situation is heterogeneous when drop from 18 percent Sub-Saharan Africa has the highest rates of looking beyond the averages. While in India in 2017 to 14 percent in exclusion. And simply attending school does the territories that were lagging behind were 2030. A deviation from not guarantee that children are learning. able to catch up quite significantly—notably the target, representing Over half the world’s children cannot read Bihar and Jharkhand—in Ethiopia some of the 225 million children and understand a simple story by age 10.57 As poorer territories were the ones with the slow- with mortality rates, there are wide disparities est progress, notably Oromia.60 within countries, showing that being at the bottom of the national income distribution Lack of human security in a broad sense is sharply increases the chance of dropping out one of the factors behind divergence in particu- (figure 1.20).58 lar territories (box 1.5). Human development for those at the bottom of the distribution is On current trends the out-of-school rate will thwarted by shocks—income, health, conflict drop from 18 percent in 2017 to 14 percent in or disaster—that make already vulnerable 2030. A deviation from the target, representing households more vulnerable. Risks refer to 225 million children59 starting their life with a events possibly occurring that can damage wel- hardly reversible disadvantage. fare, and vulnerability can be understood as the (ex ante) magnitude of the threat to human de- The mixed picture of progress can velopment outcomes.61 Individuals and house- be seen through the lens of the Global holds can reduce their vulnerability—that is, Multidimensional Poverty Index, produced by they can strengthen their ability to deal with the United Nations Development Programme shocks when they happen—by having access to and the Oxford Poverty and Human assets that can soften the blow. Development Initiative. Today 1.3 billion people in developing countries are multi- The stakes at the bottom are high. Shocks dimensionally poor. In a detailed study of 10 can affect people’s actions in ways that dimin- countries with comparable data over time, nine ish human development potential over the saw a reduction in the multidimensional pov- long run (for instance taking children out of erty rate in recent years. And in nine of them school), but they can also push individuals and the improvement of the bottom 40 percent households into extreme deprivation without much notice. FIGURE 1.20

School dropout rates converge with human Towards enhanced agency development, but not for the poorest 20 percent

Children out of The preceding section presented some stylized primary school (percent) facts about inequalities in human develop- ment—going beyond income. But the analysis 100 of a few dimensions using a limited set of stand- ard indicators is far from exhaustive. Relevant 80 inequalities in human development likely vary across geographies, cultures and time. Indeed, 60 the people-centred human development ap- proach is pluralist—admitting different valua- 40 tions and priorities—and open-ended.

20 How best to manage this complexity—the multidimensional and changing nature of ine- 0 Lowest Highest qualities—to explore the inequalities emerging quintile quintile in the 21st century? Lowest Highest Lowest Highest quintile quintile quintile quintile High/very high This section addresses this question by looking at two aspects that bear on people’s Low Medium agency, supplementing the aspects linked to

Note: Each box plots the middle 50 percent of the distribution; the central line is the median; the extreme lines are the approximate minimum and maximum of the distribution.

Moving targets in the 21st century | 51 BOX 1.5 Crises and divergence

Economic crises are an important factor behind divergence 3,400 classrooms had been destroyed or damaged in in economic and social conditions. Countries suffering re- Mozambique, with close to 305,000 children losing cessions often take several years to recover.1 Moreover, out on lessons at school after the floods.4 Malaria cas- within countries, crises tend to hurt the most vulnerable. es rose to 27,000, and cholera cases to almost 7,000. In a study of Latin American countries all economic crises About 1.6 million people received food assistance, and were followed by an increase in the poverty rate, and most close to 14,000 people had to live in displacement cen- were followed by an increase in inequality.2 tres. The cumulative effects of the storms will be fully understood only over the next few years. Disasters linked to natural hazards can have dev- astating impacts and harm human development, as Conflicts are also devastating for human devel- discussed in chapter 5. And such disasters will be- opment. Before the escalation of conflict in Yemen in come more common as the climate crisis worsens. The 2015, the country ranked 153 in human development, effects can be truly devastating. On 14 March 2019 138 in extreme poverty, 147 in life expectancy and 172 tropical Cyclone Idai made landfall at the port of Beira, in education attainment. The conflict has reversed the Mozambique, before moving across the region. Millions pace of development—with nearly a quarter of a million of people in Malawi, Mozambique and Zimbabwe were people killed directly by fighting and indirectly through hit by Southern Africa’s worst natural disaster in at least lack of food, infrastructure and health services. Some two decades.3 Six weeks later Cyclone Kenneth made 60 percent of those killed are children under age 5. The landfall in northern Mozambique—the first time in re- long-term impacts make it among the most destructive corded history that two strong tropical cyclones hit the conflicts since the end of the Cold War (see figure) and country in the same season. The cyclones left around have already set back human development in the coun- 1.85 million people in Mozambique alone in urgent need try by 21 years. If the conflict continues through 2022, of humanitarian assistance. development would be set back 26 years—more than one generation. If the conflict persists through 2030, the The cyclones were only the beginning of what impact grows to nearly 40 years. has become an education and health disaster. Around

Conflict has already set back Yemen’s human development by 21 years

Human Development Index (years set back at the end of the conflict)

1990 2000 2010 2020 2030

2019 1998 (21-year setback)

2022 1996 (26-year setback)

2030 1991 (39-year setback)

Source: Moyer and others 2019.

1. Unemployment takes more than four years to recover; output, around two years (Reinhart and Rogoff 2009) and in many cases even more (Cerra and Saxena 2008). 2. Lustig 2000. 3. UNICEF 2019b. 4. See UNICEF (2019b).

inequalities in capabilities discussed until now. the fruits of progress, with perverse effects on As noted, capabilities are determinants of social mobility and long-term social progress well-being and are required for agency—but (chapter 2), and because they erode human are not the sole determinants. Thus, this section dignity—and with it social recognition and first considers how inequality, often in the form respect, which may limit agency. Second, since of discrimination, deprives people of dignity. inequality is a social and relational concept, it Inequalities hurt because they restrict access to responds to comparisons across social groups

52 | HUMAN DEVELOPMENT REPORT 2019 and between individuals. So, social perceptions open social interaction and personal realiza- The search for can bring information about the social differ- tion (box 1.6). dignity is crucial ences that matter to people, given that human in defining the actions are also shaped by perceptions of fair- Dignity as equal treatment and non constitutive aspects ness towards what happens to one’s own and to discrimination can be even more important of development in others. than imbalances in the distribution of income. the 21st century In Chile, with its very unequal income distri- Inequalities and the search for dignity bution, inequality in income appeared high in the ranking of people’s concerns (53 per- The search for dignity is crucial in defining cent of people said they were bothered by the constitutive aspects of development in the income inequality) in a 2017 United Nations 21st century. This is true for both basic and Development Programme survey.65 But they enhanced capabilities and achievements, and expressed even more discontent with unequal it is a powerful insight to explore emerging access to health (68 percent), unequal access sources of exclusion—sources hard to cap- to education (67 percent) and unequal respect ture through indicators typically reported and dignity in the way people are treated by national statistical offices. The search for (66 percent). Of the 41 percent of people who dignity is explicit in the “central capabilities” said they had been treated with disrespect of Martha Nussbaum.62 Amartya Sen, in turn, over the last year, 43 percent said it was be- emphasizes that, in defining minimally re- cause of their social class, 41 percent said it quired freedoms, what matters is not only the was because they are female, 28 percent said it effect of directly observable outcomes—such was because of where they live and 27 percent as income—but also the potential restric- said it was because of how they dress. In this tions in the capability to function in society context, progress in policies to advance agency without shame.63 He follows Adam Smith’s and reduce shame and discrimination appear Wealth of Nations, highlighting the role of as important as those to increase material relative deprivations—with symbolic social conditions.66 In Japan the concept and meas- relevance, even if not essential for biological urement of dignity also signal inequalities subsistence—as defining basic necessities. that other material indicators cannot capture This is one of the roots of moving targets in (box 1.7). development. And indeed, human dignity has been a central element in the evolution of Lack of equal treatment and nondiscrimina- the global consensus about universally shared tion are also reflected in inequalities between ambitions, from the Universal Declaration groups, which are known as horizontal ine- of Human Rights in 1948 to the Sustainable qualities.67 Horizontal inequalities are unfair, Development Goals in 2015. as they are rooted in people’s characteristics, beyond their control. The SDGs encourage The search for dignity can also be crucial for examining horizontal inequalities through dis- policymaking, particularly when recognition aggregation that spotlights priority groups— (in the sense of equal treatment) is required those traditionally disadvantaged by income, to complement other pro-equity policies, in- gender, age, race, ethnicity, migratory status, cluding redistribution.64 One example is pro- disability, geographic location and other char- gress in recognition and rights of lesbian, gay, acteristics relevant in national contexts.68 bisexual, transgender and intersex (LGBTI) people. The ability to appear in public with- Horizontal inequalities can reflect delib- out shame is severely undermined when a erate discrimination in policies, laws and person’s identity is socially penalized. The actions—or hidden mechanisms embedded exclusion of LGBTI people takes the form of in social norms, unconscious biases or the discrimination in work and in communities. functioning of markets. Often the cultural An environment hostile to LGBTI people currents that drive horizontal inequality are forces individuals to choose between facing deep enough to perpetuate it despite policies oppression and hiding their sexual identity to ban or reduce it, as in India (box 1.8). In and preferences, limiting their possibilities of Latin America horizontal inequalities appear connected to a culture of privilege, with roots in colonial times.69

Moving targets in the 21st century | 53 BOX 1.6 International Lesbian, Gay, Bisexual, Social exclusion of lesbian, gay, bisexual, trans and intersex people Trans and Intersex Association

Across the globe, lesbian, gay, bisexual, trans and intersex life. There is abundant research showing how LGBTI (LGBTI) people continue to face social exclusion in differ- people suffer from erasure, negation, discrimination ent spheres of life on the basis of their sexual orientation, and violence:6 A spiral of rejection may start at a very gender identity, gender expression and sex characteristics. young age within the family and continue in school,7 Restrictive legal frameworks, discrimination and violence employment,8 health care facilities and public spaces.9 based on those qualities (perpetrated by state and non- State officials can be the main perpetrators of violence state actors) and the lack of effective public policy are and abuse against LGBTI people, carrying out arbitrary among the main causes behind the exclusion of LGBTI arrests, blackmail, humiliation, harassment and even people.1 forced medical examinations. LGBTI people also face exclusion when seeking access to justice, which con- Restrictive legal frameworks tributes to under-reporting of violence against LGBTI Criminalization is a major barrier for LGBTI people’s de- people and low rates of prosecution of perpetrators of velopment. As of May 2019, 69 UN Member States still such violence because LGBTI people are often isolated criminalize consensual same-sex sexual acts between from state institutions for fear of self-incrimination and adults, and at least 38 of them still actively arrest, pros- further abuses.10 ecute and sentence people to prison, corporal punish- ment or even death based on these laws.2 Moreover, Lack of effective public policy many UN Member States also have laws criminalizing The third main group of causes of social exclusion of diverse forms of gender expression and cross-dressing, LGBTI people has to do with state inaction on public pol- which are used to persecute trans and gender-diverse icy issues of sexual and gender diversity.11 As with other people.3 social groups that have been subjected to protracted discrimination, full social inclusion of LGBTI people The lack of legal gender recognition4 is one of the requires more than removing discriminatory legislation most challenging barriers to trans and gender-diverse and enacting legal protections. Effective public policies people’s social inclusion. When personal documents do designed and implemented to tackle, reduce and even- not match the holder’s appearance, it becomes a huge tually eradicate social prejudice and stigma are required obstacle to carry out common activities in daily life, such to counter the effects of systemic exclusion, especially as opening a bank account, applying for a scholarship, among those living in poverty. Affirmative action may finding a job and renting or buying property. It also expos- also be necessary. es trans people to the scrutiny of strangers, distrust and even violence. In many countries legal gender recognition Intersex people also face particular forms of ex- is granted only under pathologizing requirements such as clusion that differ from those experienced by lesbian, surgeries, invasive treatments/inspections or third-party gay, bisexual and trans people. In particular, they are submissions.5 Furthermore, when antidiscrimination laws often subjected to unnecessary medical interventions do not expressly protect people based on their sexual at birth, characterized as intersex genital mutilation.12 orientation, gender identity, gender expression and sex These interventions are often conducted in accord- characteristics, LGBTI people are unable to seek justice ance with medical protocols that allow health profes- against acts of discrimination that may prevent them from sionals to mutilate intersex bodies without consent accessing vital services, including health care, education, to modify atypical sex characteristics, usually when housing and social security, and employment. victims are infants. Such traumatizing and intrusive experiences can extend throughout childhood and Discrimination and violence based on sexual adolescence and can cause severe mental, sexual orientation, gender identity, gender expression and and physical suffering.13 This is usually compounded sex characteristics by the total secrecy about intersex conditions, lack Suffering violence and discrimination can deeply af- of information among family members and societal fect a person’s ability to lead a productive and fulfilling prejudice.14

Source: International Lesbian, Gay, Bisexual, Trans and Intersex Association and United Nations Development Programme. 1. ILGA 2019; OutRight Action International 2019. 2. ILGA 2019. 3. Greef 2019; ILGA 2019. 4. Legal gender recognition refers to the right of trans people to legally change their gender markers and names on official documents. For a survey of the legislation in force with regard to legal gender recognition in more than 110 countries, see Chiam, Duffy and Gil (2017). 5. Chiam, Duffy and Gil 2017. 6. Harper and Schneider 2003. 7. Almeida and others 2009. 8. Pizer and others 2012; Sears and Mallory 2011. 9. Eliason, Dibble and Robertson 2011. 10. ILGA 2019. 11. Oleske 2015. 12. Wilson 2012. 13. WHO Study Group on Female Genital Mutilation and Obstetric Outcome 2006. 14. Human Rights Watch 2017.

54 | HUMAN DEVELOPMENT REPORT 2019 BOX 1.7 Inequality in human security in Japan: The role of dignity

In Japan the Sustainable Development Goals present an was measured through 26 indicators: 7 about the sit- opportunity to revisit the country’s development prior- uation of children and women, 6 about trust in the ities with a people-centred perspective. What defines public sector, 2 about life satisfaction and 11 about deprivation after most material shortages have been community, civic engagement and sound absorption of overcome? The Human Security Index has three dimen- migrants. sions of human security: life, livelihood and dignity. Life and livelihood are linked to peace of mind and feelings Early results show significant inequalities in of safety. Dignity aims for a society where every person Japan across all three main dimensions. But the dig- can be proud of himself or herself. nity subindex shows a lower average than the life and livelihood subindices. From this perspective the In Japan data were collected on 47 prefectures, greatest space for improvement would be in promot- using a battery of 91 indicators. The dignity dimension ing dignity.

Source: Based on Takasu (2019). want more redistribution when informed of Subjective measures their true ranking.74 consistently Uncovering what is behind perceptions indicate that many of inequalities in the 21st century The way societies process inequalities is com- people around the plex. Studies in behavioural economics have world find current The proportion of people desiring more in- quantified how much people tend to underes- inequality too high come equality has risen over the past decade timate inequalities (see spotlight 1.2 at the end (see figure 1.1). Inequality is considered a ma- of the chapter). And social psychology has in- jor challenge in 44 countries surveyed by Pew vestigated the mechanisms and sociostructural Research. A median of 60 percent of respond- conditions that determine the perception of ents in developing countries and 56 percent inequalities, the perception of inequalities as in developed countries agree that “the gap be- unfair outcomes and the response to those per- tween the rich and poor is a very big problem” ceptions. This literature gives new insights into facing their countries.70 Remarkably, these feel- why people come to terms with very high ine- ings are shared across the political spectrum. quality from a social perspective. First, people might accept or even contribute to inequality Similarly, according to the latest perception through self-segregation following a desire for surveys in the European Union, an overwhelm- harmony. Second, motivational narratives can ing majority think that income differences are justify inequality, and stereotypes and social too great (84 percent) and agree that their gov- norms have enormous influence (box 1.9). This ernments should take measures to reduce them is a consistent and powerful complement to the (81 percent).71 In Latin America the perception theory of adaptive preferences—based on the of unfairness in the distribution of wealth has individual’s tendency to underestimate depri- increased since 2012, returning to levels of the vations to make them bearable—now from a late 1990s, with only 16 percent of respond- social point of view. ents assessing the distribution as fair.72 This is not to suggest that this is the only, or even the In summary, subjective measures consistently most important, issue that people are worried indicate that many people around the world about—but it is clear evidence of the great, and find current inequality too high. Perceptions increasing, desire for more equality. data—when the limitations are well under- stood—can complement objective indicators. These perceptions matter and may depend Indeed, some of the frontier measures of capa- on whether the broader context is one of bilities and agency are subjective indicators.75 stagnating or expanding incomes. Perceptions Perceptions of inequality tend to underesti- of inequality—rather than actual levels of mate the actual situation, so at high levels, they inequality—drive society’s preferences for have particular value as a red flag. Some of the redistribution.73 In Argentina people who be- objective indicators of inequality—such as the lieved themselves to be higher in the income distribution than they actually were tended to

Moving targets in the 21st century | 55 BOX 1.8 Horizontal inequalities in India: Different dynamics in basic and enhanced capabilities

India is a fast-growing economy. Its gross national income stigma and exclusion for centuries. Modern India has per capita has more than doubled since 2005. Thanks to tried to constitutionally redress the disparities through a mix of fast economic growth and social policies, there affirmative action, positive discrimination and reserva- has been a sharp reduction in multidimensional pov- tion policies for these groups.3 erty. Between 2005/2006 and 2015/2016 the number of multidimensionally poor people in India fell by more Second, since 2005/06 there has been a reduction than 271 million. On average, progress was more intense in inequalities in basic areas of human development. among the poorest states and the poorest groups.1 For example, there is a convergence of education at- tainment, with historically marginalized groups catching Despite progress on human development indica- up with the rest of the population in the proportion of tors, horizontal inequalities persist, and their dynamics people with five or more years of education. Similarly, follow the same pattern described in the context of there is convergence in access to and uptake of mobile vertical inequalities in human development: significant phones. gaps, convergence in basic capabilities and divergence in enhanced capabilities. Third, there has been an increase in inequalities in enhanced areas of human development, such as access First, the Scheduled Castes, Scheduled Tribes and to computers and to 12 or more years of education: Other Backward Classes underperform the rest of so- Groups that were more advantaged in 2005/2006 have ciety across human development indicators, including made the most gains, and marginalized groups are mov- education attainment and access to digital technologies ing forward but in comparative terms are lagging further (box figures 1 and 2).2 These groups have suffered from behind, despite progress.

Box figure 1 India: Horizontal inequality in education of working-age people (ages 15­49)

Population with 5 or more Population with 12 or more years of education, 2015 (percent) years of education, 2015 (percent)

78.7 83.1 87.1 38.7 60.7 66.3 77.9 68.8 30.4 50.2 29.4

23.2 21.2

17.5 15.7

10.6

Scheduled Scheduled Other Other Scheduled Scheduled Other Other

Tribe Caste Backward Class Tribe Caste Backward Class

Men Women

Change in population with 5 or more years of education Change in population with 12 or more years of education between 2005 and 2015 (percentage points) between 2005 and 2015 (percentage points)

20.6 20.0 10.811.4 11.1 10.5 18.0 17.8 9.8 9.4

8.8

13.3 11.3 6.7 10.4 7.3

Scheduled Scheduled Other Other Scheduled Scheduled Other Other

Tribe Caste Backward Class Tribe Caste Backward Class

Source: Human Development Report Office calculations based on Demographic and Health Survey data.

56 | HUMAN DEVELOPMENT REPORT 2019 BOX 1.8 (CONTINUED) Horizontal inequalities in India: Different dynamics in basic and enhanced capabilities

Box figure 2 India: Horizontal inequality in access to technology

Households with access (percent)

Mobile, 2015 Computer, 2015 16.7 87.8 92.0 94.7

77.6

8.0

4.8

Scheduled Scheduled Other Other Scheduled Scheduled Other Other A shift in people’s aspirations as a result Tribe Caste Backward Class Tribe Caste Backward Class of individual and social achievements can be Mobile, change between 2005 and 2015 Computer, change between 2005 and 2015 a natural part of the (percentage points) (percentage points) development process

79.1 77.4 10.2

72.6 6.0

66.2 4.0 2.3

Scheduled Scheduled Other Other Scheduled Scheduled Other Other

Tribe Caste Backward Class Tribe Caste Backward Class

Source: Human Development Report Office calculations based on Demographic and Health Survey data.

1. See UNDP and OPHI 2019. 2. See IIPS and Macro International (2007) and IIPS and IFC International (2017). 3. Mosse 2018.

Gini coefficient in developing countries—do Moving targets and 21st not yet capture this reality, and it is plausible century inequalities that those indicators might be missing part of the story.76 The empirical discussion in this A shift in people’s aspirations as a result of indi- report provides numerous examples showing vidual and social achievements can be a natural how going beyond income, beyond averages part of the development process. This moving (and summary measures such as the Gini co- target is inherently relative and, therefore, re- efficient) and beyond today in measurement quires a more flexible way to assess inequality. A (capturing elements expected to become more definition of inequality from a few decades ago important) makes a difference in uncovering may no longer be relevant. In a world without the growing inequalities that might be behind extreme poverty, for example, the poverty line those perceptions. will inevitably rise—indeed, poverty in devel- oped countries is usually measured in relative Finally, increasing demand for equality in terms. For human development a shift in focus perception surveys has concrete consequences from basic to enhanced capabilities may be for society. No matter the degree of subjectivity relevant. And what is considered enhanced is and potential distortion, these opinions have bound to change over time: Think of how the the chance to become part of the political dis- access to electricity and sanitation infrastructure cussion and to stimulate action. There is an ur- moved from ambitions to basic during the 20th gent need for evidence-based policy approaches to respond to new demands.

Moving targets in the 21st century | 57 BOX 1.9 A social-psychological perspective on inequality

This box is grounded in an emerging social-psychological perspective on be motivated to deny or justify the existence of inequality to uphold beliefs people as relational beings, motivated to regulate their network of social about the fairness of the broader system.7 Income inequality may be viewed relationships. This perspective, which moves beyond more individualistic as fair by those who endorse a meritocratic belief system (affirming a level perspectives, suggests that social embeddedness (the experience of social playing field for everyone). Indeed, stereotypes are often used to acknowl- connection within social networks and through group identities) and relative edge inequalities in order to maintain them and thus the broader system in deprivation (the experience of being unfairly worse off than others, based in which they are embedded.8 social comparisons with others) have important consequences. Against this backdrop, a social-psychological perspective offers an- Humans are an ultra-social species, with a need to belong. The psycho- swers to questions such as why people do or do not act against inequality logical bonds that individuals develop with others through social interac- (such as the gender pay gap) and why they often appear to act irrationally tion reflect sources of social support and agency and offer targets for social (as in voting for a party that does not protect their interests). Such a perspec- comparison (subjective assessments of whether others are doing better or tive helps move beyond general correlations in aggregated data (such as be- worse than oneself).1 This is key to understanding the consequences of ine- tween-country indicators of income inequality and public health) and zooms quality because a social-psychological perspective focuses on whether and in on the part of the broader relationship that can be explained through such how individuals subjectively perceive and feel about inequality depending psychological processes as embeddedness and relative deprivation.9 on their network of relationships. A social-psychological perspective of inequality also goes beyond in- But even when individuals perceive inequality, they may not perceive come inequality. Many health inequalities have social antecedents in vari- it as unfair.2 Social networks tend to be homogeneous because individuals ous forms of inequality, including gender, ethnicity and race.10 Reference and tend to self-segregate (“birds of a feather flock together”).3 Individuals often social comparison groups suggest that it is important to know whom people compare themselves with those around them, the ones forming a “bubble,” compare themselves with and thus who is in their network and which group who are thus likely to affirm their opinions about inequality. Contact with identities they value—and which specific forms of inequality they are likely others—for instance, between members of advantaged and disadvantaged to perceive as unfair and feel relatively deprived in. These psychological di- groups—may increase people’s awareness of inequality,4 but research also mensions can be easily lost as the level of analysis and aggregation goes up. suggests that such contact is often characterized by a desire to maintain harmony rather than to discuss the uncomfortable truth of inequality be- Take education. It is not just an objective factor that affords or inhibits tween groups (the “irony of harmony”).5 As such, social embeddedness of- opportunities for social mobility. It can also be a potential bubble and identi- ten implies a sedative effect when it comes to perceiving inequality—one ty factor in political participation.11 For example, making people aware of the cannot act on what one cannot see within one’s bubble.6 status differences between different education groups only reinforces those differences, likely based in confirmation of the competence stereotypes as- There is also a motivational explanation for why inequality, even when sociated with the lower and higher educated.12 This is reminiscent of how perceived, is not necessarily perceived as unfair. Specifically, individuals can beliefs in meritocracy can justify inequalities.13

1. Festinger 1954; Smith and others 2012. 2. Deaton 2003; Jost 2019; Jost, Ledgerwood and Hardin 2008; Major 1994. 3. Dixon, Durrheim and Tredoux 2005. 4. MacInnis and Hodson 2019. 5. Saguy 2018. 6. Cakal and others 2011. 7. Jost, Ledgerwood and Hardin 2008; Major 1994. 8. Jost, Ledgerwood and Hardin 2008; Major 1994. 9. Corcoran, Pettinicchio and Young 2011; Green, Glaser and Rich 1998. 10. Marmot 2005. 11. Spruyt and Kuppens 2015. 12. Spruyt, Kuppens, Spears and van Noord forthcoming. 13. Jost 2019. Source: Based on van Zomeren (2019).

century. For development-induced gaps, reduc- goals and aspirations, while people at the top are tions in inequality are desirable and expected, enhancing their advantages in those relevant for not from restricting gains of those taking the the 21st century. Between the bottom and the lead, but from broadly diffusing the newer more top of the human development distribution is advanced dimensions of development.77 the most diverse global middle class in history. It is diverse in its cultural composition, geographic This chapter has argued for measuring human location and relative position in the dynamics of development based on the formation of capabil- convergence and divergence. It is also a middle ities, step by step, from basic to enhanced. It has class increasingly fragmented within countries in documented large gaps in human development access to goods and services, as documented in in all dimensions. But the evolution of inequal- developed countries.78 ities shows two distinct patterns. Overall, the global bottom is catching up in basic capabili- It can be argued that some of the new in- ties, and inequality appears to be falling. But the equalities are a natural result of progress.79 global top is pulling ahead in enhanced capabil- Progress has to start somewhere, so some ities, and here inequality is growing. People at groups go first. Based on gradual progress, the the bottom are catching up with 20th century evolution of inequality might follow the shape

58 | HUMAN DEVELOPMENT REPORT 2019 of an inverted-U over time, a version of the These simultaneous patterns of convergence But moving targets Kuznets curve.80 When very few people achieve and divergence are likely to play a prominent could also be a a “target” (say, access to a new technology), role in the 21st century. Both trends are im- challenge for human inequality is low: Most people perform equally portant, not only because of their separate development if poorly. Subsequently, as more people obtain ac- effects—reducing extreme deprivations in more efforts and cess, inequality starts to increase, reflecting the the first case and concentrating power in the accomplishments division between the haves and the have-nots. second—but also because of their political are needed to get the Later on, once a large proportion of people have implications. Progress might not mean as much same capabilities reached access, inequality starts to decrease: if combined with increased inequality in areas The majority of people are performing equally people care deeply about, because of the con- well. This shows that there are different types nections with empowerment and agency. of inequalities. There are multiple processes of divergence and convergence taking place at the Once most of the population has attained same time—overlapping Kuznets curves81—so certain goals, other elements become more rel- the same person could be catching up with ba- evant for how people see themselves in relation sic capabilities and, simultaneously, being left to others and how others perceive them. They behind in the building up of enhanced capabil- begin to focus on their place in society and the ities. When these patterns are not random, and associated rights, responsibilities and opportu- some groups tend to be in the lead, while others nities. Emerging inequalities can trigger per- are consistently behind, this process is bound ceptions of unfairness to the extent that there is to be perceived as unfair. no or slow catching up.

Thus, even if transitory inequality goes along But moving targets could also be a challenge with some forms of progress, that inequality for human development if more efforts and can be unfair if subsequent progress does not accomplishments are needed to get the same spread out widely and fast enough. Inequalities capabilities. People are likely to feel themselves in enhanced capabilities that were already constantly falling behind.82 high 10 years ago have been increasing since. This can be changed, and it is a motivation for These dynamics83 pose new and difficult policies that specifically address equality in challenges that will affect development paths capabilities. in the coming decades. Chapter 2 turns to a description of the mechanisms that underpin these dynamics.

Moving targets in the 21st century | 59 Spotlight 1.1

Power concentration and state capture: Insights from history on consequences of market dominance for inequality and environmental calamities

Bas van Bavel, Distinguished Professor of Transitions of Economy and Society, Utrecht University, The

The organization of markets, their functioning, not an arbitrary set, but these are all known cas- their interaction with the state and their broader es of economies with dominant markets, which effects on an economy and society develop slow- can be followed over a long period. This allows a ly. While debates on inequality are dominated better understanding of how market economies by developments spanning a few decades, and develop, something that theoretical and formal often even a few years, observing and analysing work and short-period cases studies cannot do. how inequality emerges, how it concentrates power and how it can lead to the capture of All six market economies displayed a similar markets and the state call for a much longer, his- evolution. In each of three cases analysed in torical perspective. Such a long-term approach depth—Iraq, Italy and the Low Countries2— may have seemed irrelevant for issues pertaining markets emerged in an equitable setting and to the market economy, since it was widely held became dominant, with an institutional or- that the market economy was a modern phe- ganization that allowed easy market access to nomenon, having developed only from the 19th broad groups within society. The opportunities century on, closely associated with moderniza- that market exchange offered further pushed tion. Recent economic historical work, however, up economic growth and well-being, with the has changed this idea, by identifying several fruits of growth fairly evenly distributed. As market economies much earlier in history.1 markets became dominant, and especially the markets for land, labour and capital, inequality Nine market economies, from antiquity to also grew in a slow process as ownership of land the modern era, have been identified with cer- and capital became more concentrated. Wealth tainty, and six of them have sufficient data to inequality in these cases grew to Gini index of investigate them well (table S1.1.1). This is thus 0.85 or higher3 from substantially lower levels.

TABLE S1.1.1 Certain and possible cases of market economies

Location Period Date Note Babylonia Possible case Babylonia Ur III / old-Babylonian period c. 1900­1600 BCE Limited data Athens/Attica Possible case Italy Neo-Babylonian period c. 700­300 BCE Limited data Iraq Lower Yangtse Classical period c. 600 BCE­300 BCE Limited data Italy (Center and North) Low Countries (especially the West) Roman period c. 200 BCE­200 CE England United States (North) Early Islamic period c. 700­1000 CE Northwestern Europe Song period c. 1000­1400 CE

c. 1200­1600 CE

c. 1500­1900 CE

c. 1600­

c. 1825­

c. 1980­

Source: Bas van Bavel (Utrecht University, The Netherlands).

60 | HUMAN DEVELOPMENT REPORT 2019 As inequality grew, economic growth initial- them to break existing inequities and forms of ly continued, but it became ever less translated coercion and to obtain a more equitable distri- into broad well-being. With the stagnating bution of wealth and resources. They also won purchasing power of large shares of the pop- the freedom to exchange their land, labour and ulation, lagging demand and the declining capital without restraints by elite power, thus profitability of economic investments, owners opening the opportunity to use the market to of large wealth increasingly shifted their in- this end. Their struggles and forms of self-or- vestments to financial markets. They used their ganization were thus at the base of the rise of wealth to acquire political leverage through factor markets—and the rise happened within patronage and buying political positions or by a relatively equal setting, ensuring that large acquiring key positions in the fiscal regime, groups could access the market and benefit bureaucracy and finance and through their from market exchange. dominance in financial markets and their role as creditors of the state. Over the course of This formative, positive phase was also found 100­150 years markets became less open and in the more familiar, modern cases of market equitable, through both large wealth owners’ economies: England, where the market became economic weight and their ability to skew the dominant in the 17th century, and the north- institutional organization of the markets.4 As ern United States, in the first half of the 19th a result, productive investments declined, the century. Both were the most equitable societies economy started to stagnate and economic of the time, with large degrees of freedom, inequality rose further, coupled with growing good access to decisionmaking and relatively political inequity and even coercion. equal distribution of land and other forms of wealth.6 Market economies were thus not the Each of the market economies started from base of freedom and equity, as some theories a very equitable situation, with relatively equal would have it, but rather developed on the distribution of economic wealth and political basis of earlier-won freedom and equity. The decisionmaking. This was the result of a long market subsequently replaced the associations preceding period of smaller and bigger revolts and organizations of ordinary people as the and forms of self-organization of ordinary allocation system, a process that sped up when people—in guilds, fraternities, associations, market elites and state elites came to overlap corporations, commons and companies and jointly, and often deliberately, marginal- (figure S1.1.1).5 Their organization enabled ized these organizations. This reduced ordinary people’s opportunities to defend freedom, their FIGURE S1.1.1 access to decisionmaking power and their grip over land and resources. Description of the stages in the development of the historical market economies The allocation systems that prevailed before the rise of the market, whether the commons or Well-being other associations, had mostly included long- term security and environmental sustainability Dominance Dominance in their functioning, as ensured by their rules. of markets of market But the market does not do so explicitly.7 And in these other systems, cause and effect, and elites actor and affected person, were more closely linked, because of their smaller scale. In mar- Rise of markets kets they are less so. This poses a risk, since in a market economy, owners of land, capital Social movements and natural resources are often far detached from those affected by damage from exploiting Inequality resources. They also face fewer constraints on exploitation than systems with more divided Source: van Bavel 2016. property rights.

In coastal Flanders, a mature market econ- omy in the 14th­16th centuries, land was

Moving targets in the 21st century | 61 accumulated by investors who did not live in Groningen) over the very long run in confront- the area. These absentee investors changed the ing the hazard of high water tables.9 Growing logic of coastal flood protection from long- material inequality increased the incidence of term security to low cost and high risk, increas- serious floods, not directly, but through the ing the flood risk and further marginalizing institutional framework for water management. the local population.8 More generally, all cases Only where this institutional organization was of market economies in their later, downward adapted in line with growing material inequality phases experienced grave ecological problems, were disastrous effects avoided (figure S1.1.2). from the salinization and breakdown of essen- This adaptation did not happen automatically tial irrigation systems (medieval Iraq) to in- or inevitably, however, even when a society was creasing floods and famines (Renaissance Italy) confronted with major floods.10 When both to malaria and floods (coastal Low Countries), property and decisionmaking rights were widely even though the later, modern market econo- distributed, chances were best that institutions mies increasingly avoided the negative effects for water management were adapted and ad- of ecological degradation by acquiring resourc- justed to changing circumstances to reduce the es overseas. risk of flood disaster. When wealthy actors and interest groups controlled property rights over To see the interaction among market econ- the main resources and held decisionmaking omies, material inequality and vulnerability power, however, they upheld the prevailing ar- to natural shocks, look at three of the most rangements to protect their particular interests, market-dominated parts of the Low Countries even if this actually weakened a society’s coping (coastal Flanders, the Dutch river area and

FIGURE S1.1.2

Linking the hazard of high water to flood disasters: Economic and political equality enhances the chance of institutions becoming adjusted to circumstances and preventing disaster

Equality Inequality Successful

Failed

Hazard Poorly adjusted institutions Disaster (high water) Well adjusted institutions (major flood)

No disaster

Source: Adapted from van Bavel, Curtis and Soens (2018). 62 | HUMAN DEVELOPMENT REPORT 2019 capacity. And if some adaptation in these cases developments are not aberrations or accidental did take place, it was often aimed at increasing events. And perhaps they require broader and the capacity of the economic system to recover deeper consideration of a wider range of policy production levels after a shock—but at the ex- actions to curb the concentration of econom- pense of segments of the population that were ic and political power. The concentration of no longer included in decisionmaking.11 The economic power (wealth), the first stage, is risk of these negative outcomes happening and easiest to curb. But after the establishment of of institutions being poorly adjusted to ecologi- economic power and its translation to political cal and social circumstances was high in market dominance, this is far harder to do. economies with high wealth inequality, where the grip of a small group of private owners over Notes natural resources was strongest and decision- making power became concentrated in their 1 This is true even if the market economy is defined in a very hands. strict way—that is, as an economy where not only goods, products and services, but also inputs (land and natural How relevant are these observations for de- resources, labour and capital) are predominantly allocated by velopments today? The historical cases where way of the market. markets emerged as the dominant allocation system for factors of production (land, labour 2 van Bavel 2016. For an analysis of long, cyclical patterns of rising and capital) all showed an accumulation of and declining inequality see also Turchin and Nefedov (2009). wealth in the hands of a small group, which then also concentrated political power, shaping 3 van Bavel 2016 (see pp. 72­73 on Iraq, 128 on Florence in incentives in markets that increased inequality 1427 and 194­195 on Amsterdam in 1630). and environmental calamities. Today, even in parliamentary democracies, economic wealth 4 This is true even in (relatively) inclusive political systems, in again seems to be translated into political lev- contrast to the argument by Acemoglu and Robinson (2012), erage—through lobbying, campaign financing where they are assumed to form a virtuous cycle. and owning media and information—whereas mobile wealth owners can easily isolate them- 5 van Bavel 2019. selves, for say, social disruption or environmen- 6 For the United States, see Acemoglu and Robinson (2012) and tal degradation.12 History shows that these Larson (2010). To be sure, a position obtained at the expense of the native population. 7 On the nonembeddedness of market outcomes, see Gemici (2007). 8 Soens 2011. 9 van Bavel, Curtis and Soens 2018. 10 See also Rohland (2018). 11 Soens 2018. 12 Gilens and Page 2014; Schlozman 2012.

Moving targets in the 21st century | 63 Spotlight 1.2

Rising subjective perceptions of inequality, rising inequalities in perceived well-being

Subjective perceptions of inequality are at to consider subjective perceptions of well-being odds with the decline in extreme deprivations and their distribution. They change from region in objective data. Surveys have revealed rising to region (figure S1.2.1). First, both the ability perceptions of inequality, rising preferences for to enjoy life and the ability to assess experiences greater equality and rising global inequality in through well-being play a paramount role in subjective perceptions of well-being. All these providing direct well-being and “evidential trends should be bright red flags—especially merit” to inform individual decisionmaking.7 given the tendencies of subjective views to Second, subjective indicators can provide val- underestimate income and wealth inequality uable information to cover some of the blind in some countries and to understate global in- spots in objective data. equalities in well-being. To be sure, subjective measures of well-be- Downward bias in perceived ing must be handled with care—but the very income and wealth inequality reasons for doubt strengthen the case for at- tending to rising perceptions of inequality. In On average, people misperceive actual income Amartya Sen’s theory of adaptive preferences, and wealth inequality. Underestimating ine- people adapt preferences to their circumstanc- quality is common in some countries, such as es.8 In data on self-reported happiness, people the United Kingdom and the United States.1 facing deprivations moderate their preferences In one survey Americans believed that the top to make their condition more bearable. In wealth quintile owned about 59 percent of the contrast, the affluent report lower happiness total wealth; the actual number was closer to than their wealth might seem to warrant, be- 84 percent.2 And ideal wealth distributions are cause their high satiation has reduced the space significantly more equal than respondents’ esti- for adding to personal satisfaction.9 For both mates. All demographic groups desired a more reasons subjective measures of happiness can equal distribution of wealth than the status quo.3 understate inequalities in well-being. And the actual wage ratio of chief executive of- ficers to unskilled workers (354:1) far exceeded Remarkably, self-reported happiness shows the estimated ratio (30:1), which in turn was increasing inequality in subjective well-being substantially higher than the ideal ratio (7:1).4 around the world—a trend that has steepened sharply since 2010 (figure S1.2.2). This has Other studies have asked respondents to esti- been an increasing trend during 2006­2018 in mate their position in the income or wealth dis- all regions except Europe.10 Inequality in the tribution. In Argentina only about 15 percent Commonwealth of Independent States was of respondents placed their household income stable at first but has been rising since 2013. in the correct decile.5 A significant portion of Inequality was steady in Latin America until poorer individuals overestimated their rank, 2014 and has risen since and rose until 2010 in while a significant proportion of richer indi- the US-dominated North America, Australia viduals underestimated theirs. Similar biases and New Zealand region but has been constant emerged in a randomized survey experiment in since. Inequality has been rising since 2010 in eight countries.6 Southeast Asia but has not risen as much in the rest of Asia. In Sub-Saharan Africa inequality Rising global inequality in subjective has followed a steep post-2010 path, similar to perceptions of well-being that in Southeast Asia. And in the Middle East and North Africa inequality rose from 2009 to In assessing inequalities, one way to look be- 2013 but has been stable since. yond income—a wholly objective measure—is The trend towards greater inequality in subjective well-being poses a challenge. First,

64 | HUMAN DEVELOPMENT REPORT 2019 FIGURE S1.2.1 Transmitting inequalities in human development across the lifecycle

Western Europe Central and Eastern Europe

Commonwealth of Independent States Southeast Asia

South Asia East Asia

Latin America and the Caribbean North America, Australia and New Zealand

Middle East and North Africa Sub-Saharan Africa

1 2 3 4 5 6 7 8 9 10 11 0 1 2 3 4 5 6 7 8 9 10

Note: Respondents’ answers to life evaluation questions asked in the poll on a scale from 0 (the worst possible) to 10 (the best possible). Source: Helliwell 2019.

Moving targets in the 21st century | 65 FIGURE S1.2.2 remarkably the World Happiness Reports, show to be correlated strongly with life evaluations— Distribution of subjective well-being across the namely income, social support, healthy life ex- world (measured by people’s overall satisfaction pectancy at birth, freedom to make life choices, with their lives) generosity and corruption—are all dimensions of human development.12 So if a society is more Global inequality in subjective well-being unequal in its experience of life satisfaction, it is (index, 2006 = 100) likely more unequal in its experience of life and 160 human development. 140 Second, higher inequality in subjective 130 well-being is associated with lower subjective 120 well-being.13 In other words, greater inequality 110 in happiness makes everyone less happy. 90 Source: Human Development Report Office.

2006 2008 2010 2012 2014 2016 2018

Ratio of happiness between people: Notes in the 95th percentile and people in the 5th percentile in the 90th percentile and people in the 10th percentile 1 Hauser and Norton 2017. in the 75th percentile and people in the 25th percentile 2 Norton and Ariely 2011. 3 Norton and Ariely 2013. Source: Human Development Report Office calculations based on Helliwell 4 Kiatpongsan and Norton 2014. (2019), using Gallup data. 5 Cruces, Pérez-Truglia and Tetaz 2013. 6 Bublitz 2016. These countries include Brazil, France, Germany, people’s overall life satisfaction is in many ways a barometer of everything else in their the Russian Federation, Spain, Sweden, the United Kingdom and lives. There are strong links between higher the United States. With regard to own estimated income posi- life evaluations and several key measures of tion, individuals in the bottom income quintile have a positive human development—including higher job income position bias, whereas individuals in higher income quin- satisfaction and more effective government— tiles have a negative income position bias (except individuals in and moderately strong links between higher the second quintile, who show almost no income position bias). life satisfaction and greater freedom of choice 7 See Sen (2008a). and lower inequality.11 Moreover, the variables 8 See, for instance, Sen (1999, pp. 62­63). that the literature on subjective well-being, and 9 Graham 2012. 10 Helliwell 2019. 11 See Hall (2013). 12 See Hall (2013). 13 Helliwell 2019.

66 | HUMAN DEVELOPMENT REPORT 2019 Spotlight 1.3

The bottom of the distribution: The challenge of eradicating income poverty

Today, about 600 million people live on less in poverty are increasing. If current trends con- than $1.90 a day.1 There has been considerable tinue, nearly 9 of 10 people in extreme poverty progress in the fight against poverty in recent will be in Sub-Saharan Africa in 2030.2 decades. The extreme income poverty rate fell from 36 percent in 1990 to 8.6 percent in 2018. Income poverty is only one form of poverty. Despite this progress, the number of people liv- Those furthest behind suffer from overlapping ing in extreme poverty globally is unacceptably deprivations, discriminatory social norms high, and poverty reduction may not be fast and lack of political empowerment. Risks and enough to end extreme poverty by 2030, as the vulnerabilities only enhance the fragility of Sustainable Development Goals demand. After achievements—as explained in the United decades of progress, poverty reduction is slow- Nations Development Programme’s framework ing (box S1.3.1). on Leaving No One Behind.3

Overall, extreme poverty rates tend to be Among countries that are off track, most are higher in low human development countries, in Africa and more than one third exhibit high but poor people can be found in countries at levels of conflict or violence.4 Together they all levels of development (figure S1.3.1). While pose some of the world’s most severe develop- poverty rates have declined in all regions, pro- ment challenges. They also share characteristics gress has been uneven, and more than half of of low tax effort and low health and education people in extreme poverty live in Sub-Saharan spending. They are hampered by weak private Africa, where absolute numbers of people living sector development in the nonagricultural service sector and share a high dependence on

BOX S1.3.1 Box figure 1 Poverty headcount by track classification, 2017 and 2030 Income poverty reduction scenarios to 2030 Poverty headcount 104 countries Today, 70 people escape poverty every minute, but once based on SSP 2 23 million most countries in Asia achieve the poverty target, the (millions) rate of poverty reduction is projected to slow to below 50 people per minute in 2020. The projected global 600 24 countries 104 countries 24 countries poverty rate for 2030 ranges from 4.5 percent (around 10 million 10 million 375 million people) to almost 6 percent (over 500 million 207 million people) (see figure). Even the most optimistic projec- tions show more than 300 million people living in ex- 400 40 countries treme poverty in Sub-Saharan Africa in 2030. 40 countries 131 million 202 million According to the benchmark scenario, 24 countries are on track to reach the poverty target, with 207 mil- 200 20 countries lion people expected to move out of poverty before 20 countries 290 million 2030. In 40 off-track countries, even though poverty 242 million headcounts will fall, 131 million people are expected 2030 to remain in poverty by 2030. In 20 countries the num- 0 ber of people living in poverty is projected to increase 2017 from 242 million to 290 million (see figure). However, the benchmark scenario is a relatively optimistic view Rising poverty Off track of future economic development, especially in Sub- Saharan Africa. On track No extreme poverty

Note: The Intergovernmental Panel on Climate Change’s Shared Socioeconomic Pathways reflect different degrees of climate change mitigation and adaptation. SSP2 corresponds to the benchmark scenario and assumes the continuation of current global socioeconomic trends. Source: Cuaresma and others 2018.

Moving targets in the 21st century | 67 FIGURE S1.3.1 Some 600 million people live below the $1.90 a day poverty line

Population living below PPP $1.90 a day Medium human High human Very high human income poverty line, 2007­2017 (percent) development development development

0

0.4 0.5 0.6 0.7 0.8 0.9 1

Note: Each bubble represents a country, and the size of the bubble is proportional to the country’s population in income poverty.

natural resources. Increasing labour income is Multidimensional poverty indices can shed critical for those at the very bottom.5 Access to further light on the people furthest behind by physical and financial assets is also important— capturing overlapping deprivations in house- land, capital and other inputs for production holds and clusters of households in a geograph- or services help as income-generating streams ic area. These are linked to income poverty, and buffers against shocks.6 Social protection, but with significant variations (figure S1.3.2). in the form of a noncontributory minimum Some people might be multidimensionally payment, providing for the most vulnerable is poor even if they live above the monetary pov- also important.7 erty line. The global Multidimensional Poverty Index (MPI) covers 101 countries, home to Human development progress involves the 77 percent of the world’s population, or 5.7 bil- capacity to generate income and translate it lion people. Some 23 percent of these people into capabilities, including better health and (1.3 billion) are multidimensionally poor. The education outcomes. This process plays out MPI data illustrate the challenge of addressing throughout the lifecycle. Each person’s devel- overlapping deprivations: 83 percent of all opment starts early—even before birth, with multidimensionally poor live in South Asia nutrition, cognitive development and educa- and Sub-Saharan Africa, 67 percent in middle tion opportunities for infants and children. It income countries, 85 percent in rural areas and continues with formal education, sexual health 46 percent in severe poverty.8 Poor people in and safety from violence before entering the rural areas tend to have deprivations in both labour market. For the poorest people the education and access to water, sanitation, elec- lifecycle is an obstacle course that reinforces tricity and housing. But the challenges extend deprivations and exclusions.

68 | HUMAN DEVELOPMENT REPORT 2019 FIGURE S1.3.2 Poverty at the $1.90 a day level is tied to multidimensional poverty

Population living below PPP $1.90 a day Human development group income poverty line, 2007­2017 (percent) Very high High Medium Low

0 0.1 0.2 0.3 0.4 0.5 0.6 0

Multidimensional Poverty Index value, 2007­2018

to urban areas, too: Child mortality and mal- FIGURE S1.1.3 nutrition are more common in urban areas.9 Sub-Saharan Africa has the most overlapping Sub-Saharan countries have the most overlapping MPI deprivations—with more than half the deprivations populations of Burundi, Somalia and South Sudan experiencing severe multidimensional People living with poverty, with 50 percent or more of overlap- multidimensional deprivations ping deprivations (figure S1.3.3). (billions)

As countries develop, people tend to leave 1.0 0.88 1.32 poverty, but the process is neither linear nor 0.72 mechanic. It comprises both an upward motion 0.8 (moving out) and a risk of downward motion 0.60 (falling back in). The very definition of a mid- 0.6 dle-class threshold can be computed by thinking of the threshold as a probability rather than an 0.4 absolute line. That is, a person might be consid- ered middle class when he or she is not poor and 0.2 is at very little risk of becoming poor. For dozens of countries that have reduced poverty, the stakes 0 Nonsevere Severe of not losing the progress of the past 15­20 years Vulnerable are significant. As Anirudh Krishna points out nonpoor Poor in his analysis of the life stories of 35,000 house- holds in India, Kenya, Peru, Uganda and North Arab States Latin America and the Caribbean Carolina (United States), many low-income East Asia and the Pacific South Asia

Europe and Central Asia Sub-Saharan Africa

Note: Vulnerable nonpoor population refers to people with 20 percent or more and less than 33 percent of overlapping deprivations. nonsevere poor population refers to people with 33­50 percent of overlapping deprivations, and severe poor population refers to people with 50 percent or more of overlapping deprivations. Source: Human Development Report Office calculations based on the methodology to compute the Multidimensional Poverty Index in HDRO and OPHI (2019).

Moving targets in the 21st century | 69 individuals are just one illness away from pov- Horizontal inequalities also have dynamic erty.10 Even relatively well-off households can effects. Between 2002 and 2005 ethnicity re- drop below the poverty line after personal (such duced the probability of transitioning out of as severe health problems) or communal shocks poverty in Mexico by 12 percentage points and (such as a disaster or the termination of the main increased the probability of falling back into source of employment). Another study shows poverty from vulnerability by 10 percentage that just 46 percent of Ugandans who were in points.14 the bottom quintile in 2013 had been there two years before.11 In Indonesia 52 percent of house- Notes holds with children were new to the bottom quintile from one year to the next.12 1 See World Bank (2018a) and the World Poverty Clock (https:// worldpoverty.io). Between 2003 and 2013, tens of millions of people moved out of poverty in Latin America. 2 See www.worldbank.org/en/topic/poverty/overview. Yet, large numbers of people remain vulnerable 3 UNDP 2018b. See also UNSDG 2019. to falling back in poverty. In Peru having the 4 Based on the classification by Gert and Kharas (2018). head of the household covered by a pension 5 See Azevedo and others (2013). increased the probability of exiting poverty by 6 See López Calva and Castelán (2016). 19 percentage points and reduced the probabil- 7 See ILO (2017). ity of falling back into poverty by 7 percentage 8 OPHI and UNDP 2019. points. By contrast, access to remittances re- 9 Aguilar and Sumner 2019. duced the probability of falling back into pov- 10 Krishna 2010. erty by 4 percentage points.13 11 Kidd and Athias 2019. 12 This analysis follows Martínez and Sánchez-Ancochea (2019a). 13 Abud, Gray-Molina and Ortiz-Juarez 2016. 14 See Abud, Gray-Molina and Ortiz-Juarez (2016).

70 | HUMAN DEVELOPMENT REPORT 2019 Chapter 2

Inequalities in Interconnected and persistent 2.

Inequalities in human development: Interconnected and persistent

“Inequality is not so much a cause of economic, political, and social processes as a consequence. […] Some of the pro- cesses that generate inequality are widely seen as fair. But others are deeply and obviously unfair, and have become a legitimate source of anger and disaffection.”1

How do the patterns of inequalities in human lifecycle perspective, similar to the one that Addressing development emerge? Where are the opportu- inspired the analysis of capabilities linked deprivations in one nities to redress them? Much of the debate on to health and education in chapter 1 (with dimension not only these questions has centred on the thesis that climate change and technology addressed at has benefits in and income inequality, in and of itself, has detri- length in part III of the Report), and considers of itself but can mental effects on human development. So re- what happens to children from birth, and even also support the ducing income inequality—primarily through before birth, and how families, labour markets amelioration of others redistribution using taxes and transfers— and public policies shape children’s oppor- would also enhance capabilities and distribute tunities.4 Parents, through their actions and them more equally. decisions, pass on to their children the quali- ties that the labour market values or devalues, Yet, this is far too reductionistic and mech- explaining in part how family background anistic a formulation of the links between determines personal income. Children’s edu- income inequality and capabilities. As in cation attainment depends on their parents’ chapter 1, it is crucial to go beyond income socioeconomic status, which also determines and lay out the mechanisms through which children’s health, starting before birth, and inequalities in human development emerge— cognitive ability, in part through early child- and often persist. hood stimuli. That status also determines the neighbourhood they grow up in, the schools This chapter’s approach follows Amartya they attend and the opportunities they have Sen’s argument in Development as Freedom in the labour market, in part through their that addressing deprivations in one dimension knowledge and networks. not only has benefits in and of itself but can also support the amelioration of others.2 For While this lifecycle approach is helpful to instance, deprivations in housing or nutrition illuminate mechanisms at the individual and may hinder health and education outcomes. household levels, the determinants of the While income is also a factor, deprivations distribution of capabilities cannot be fully are not necessarily tied to household ability to accounted for by behaviour at these levels. buy goods and services in markets. That is the Policies, institutions, and the rate of growth motivation for the global Multidimensional and change in the structure of the economy, Poverty Index, the nonmonetary measure among other factors, also matter a great deal. of deprivation published in the Human Thus, the chapter follows a second approach to Development Report since 2010.3 Being in consider how income inequality interacts with poor health and having low education achieve- institutions and balances of power, the way soci- ments, in turn, can hinder the ability to earn eties function and even the nature of economic income or participate in social and political growth. Going beyond income does not imply life. These deprivations can reinforce each excluding income inequality. Instead, it means other and accumulate over time— d riving and that income inequality should, in the words of even amplifying disparities in capabilities. Angus Deaton, not be considered some sort of “pollution” that directly harms human devel- The difficulty with this approach, however, is opment outcomes.5 It is crucial to spell out the similar to the one in chapter 1: where to start? mechanisms through which income inequality

This chapter addresses the question by following a dual approach. The first takes a

Chapter 2 Inequalities in human development: Interconnected and persistent | 73 interacts with society, with politics and with the FIGURE 2.1 economy in ways that can both beget more ine- qualities and harm human development. Intergenerational mobility in income is lower in countries with more inequality in human One example is how income inequality, development institutions and balances of power co-evolve. When elite groups can shape policies that fa- Intergenerational Colombia vour themselves and their children, that drives income elasticity Rwanda further accumulation of income and opportu- India nity at the top. High income inequality is thus 1.2 related to lower mobility—individuals’ ability to improve their socioeconomic status. Ecuador 1.0 Intergenerational income mobility—the extent to which parents’ income accounts for Latvia their children’s income—is persistently low in 0.8 some societies.6 When that happens, the skills and talent in an economy are not necessarily Albania allocated in the most efficient way, reducing Slovakia economic growth from a counterfactual that 0.6 allocates resources to earn the greatest returns. The point to emphasize is less the precision of 0.4 Pakistan cross-country econometric estimates and more the identification of a plausible mechanism that China Ethiopia runs from high inequality through opportuni- ty (key for human development) to economic 0.2 Singapore growth—and back. The nature of inequalities also matters. For 0 example, horizontal inequalities—which, as 0 10 20 30 40 highlighted in chapter 1, refer to disparities among groups rather than among individuals— Inequality in human development (percent) seem to matter for conflict. Once again, spell- ing out the mechanism is crucial: In this case, Note: The measure of inequality used is the percentage loss in Human horizontal inequalities not only lead to shared Development Index (HDI) value due to inequality in three components: income, grievances within a group but can also interact education and health. The loss can be understood as a proxy for inequality in with political inequality to mobilize collective capabilities. The correlation coefficient is .6292. Inequality in income is the strongest action for that group to take up arms. correlate among the three components (with a correlation coefficient of .6243), followed by inequality in education (.4931) and inequality in life expectancy (.4713). How inequalities begin at Source: Human Development Report Office using data from GDIM (2018), birth—and can persist adapted from Corak (2013).

In countries with high In countries with high income inequality the The greater the inequality in human develop- income inequality the association between parents’ income and their ment, the greater the intergenerational income association between children’s income is stronger— that is, intergen- elasticity— that is, the lower the mobility. This parents’ income and erational income mobility is lower. This relation relation does not imply direct causality in either their children’s income is known as the Great Gatsby Curve,7 often direction and can be accounted for by a number portrayed in a cross-plot of country data with of mechanisms running in both directions.8 is stronger— that is, income inequality on the horizontal axis and a This section explores how “the adult outcomes intergenerational measure of the correlation between parents’ in- of children reflect a series of gradients between income mobility come and their children’s income on the vertical their attainments at specific points in their lives is lower axis. The Great Gatsby Curve also holds using and the prevailing socioeconomic inequalities a measure of inequality in human development to which they are exposed.”9 instead of income inequality alone (figure 2.1): The underlying mechanisms of this relation can be understood, departing from inequal- ity (because it is possible to account for the relationship also in the direction running from low mobility to high inequality), as fol- lows: “Inequality lowers mobility because it shapes opportunity. It heightens the income consequences of innate differences between individuals; it also changes opportunities, in- centives, and institutions that form, develop, and transmit characteristics and skills valued in the labour market; and it shifts the balance of power so that some groups are in a position

74 | HUMAN DEVELOPMENT REPORT 2019 to structure policies or otherwise support between the rich and the poor— u ntil health People with a their children’s achievement independent of innovations around the 18th century made it certain income and talent.”10 Opportunities are thus shaped by possible for the richest to start having access education tend to incentives and institutions that interact as driv- to health technologies: “Power and money are marry (or cohabit with) ers behind the Great Gatsby Curve. In more useless against the force of mortality without partners with similar unequal countries it tends to be more difficult weapons to fight it.”19 In the second half of the socioeconomic status to move up because opportunities to do so are 19th century health gradients were carefully unequally distributed among the population.11 documented in Britain and elsewhere, with But what factors constitute inequality of op- their persistence remaining an enduring area of portunity? There are several, including—but policy and academic debate.20 not limited to—family background, gender, race, or place of birth—all crucial in explaining How do health and education gradients income inequality.12 The above hypothesis is evolve to opportunity? Some interactions supported by a negative association between can describe what happens over the lifecycle a measure of inequality in opportunity and (figure 2.2). mobility in education, finding that the share of income inequality that is attributable to A key channel for a potential vicious cycle of circumstances is higher in countries with lower low mobility is an education loop. Education education mobility.13 A similar relation was mobilizes individuals to improve their lot, found between inequality in opportunity and but when low education is passed on from mobility in income.14 parents to children, those opportunities for improvement are not fully seized. To break the Inequality in opportunity is thus a link be- cycle requires understanding how these loops tween inequality and intergenerational mobil- operate, pointing to opportunities for interven- ity: If higher inequality makes mobility more tions, considered in the next section. Another difficult, it is likely because opportunities for significant loop relates to health status, starting advancement are more unequally distributed at birth and evolving through life depending among children. Conversely, the way lower on family choices and health policies.21 The mobility may contribute to the persistence of unequal distribution of health conditions can inequalities is by making opportunity sets very contribute to inequalities in other areas of life, different among the children of the rich and such as education and the possibility to gener- the children of the poor.15 These opportunities ate income.22 The relation also goes the other not only affect the level of welfare that will be way, with health gradients in income suggesting attained; they also determine the efforts that that higher income “protects” health, which in will have to be invested to achieve certain out- turn enables people to be less prone to losing comes.16 A measure of inequality that assesses income as a result of being sick (with a vicious only outcomes will thus never be able to fully cycle in reverse potentially happening to those assess the fairness of a certain allocation of with lower income). resources.17 Inequalities in key areas of human devel- But relative mobilityi s not alone in being im- opment are thus interconnected and can be portant for human development. Without ab- persistent from one generation to the next. solute mobility, education and income would Many aspects of children’s outcomes can be not increase from one generation to the next, carried through to other stages of the lifecycle, which is important for progress, especially for where they affect adults’ ability to generate low human development countries that need to income. The resulting socioeconomic status catch up in capabilities (see chapter 1).18 shapes mating behaviours among adults.23 People with a certain income and education As introduced in chapter 1, a gradient de- tend to marry (or cohabit with) partners with scribes how achievements along a dimension similar socioeconomic status (assortative mat- (say, health or education) increase with socio- ing).24 When these couples have children, the economic status. A vast literature describes how feedback loop can start from the top again,25 gradients emerge and persist. Angus Deaton with parents’ socioeconomic status shaping described how health gradients were flat— their children’s health and early childhood with very little difference in health outcomes development.26

Chapter 2 Inequalities in human development: Interconnected and persistent | 75 FIGURE 2.2

Education and health along the lifecycle

Child’s Early Assortative health childhood mating health Education

Note: The circles represent different stages of the lifecycle, with the orange ones resenting final outcomes. The rectangle represents the process of assortative mating. The dashed lines refer to interactions that are not described in detail in this chapter. A child’s health affects early childhood development and prospects for education. For example, an intellectually disabled child will not be able to benefit from early childhood development and education opportunities in the same way as a healthy child. Education can also promote a healthy lifestyle and convey information on how to benefit from a given health care system if needed (Cutler and Lleras-Muney 2010). Source: Human Development Report Office, adapted from Deaton (2013b).

Countries with Education: how gaps can FIGURE 2.3 higher inequality in emerge early in life human development Intergenerational persistence of education is Similar to the Great Gatsby Curve and to higher in countries with higher inequality in human see higher figure 2.1, countries with higher inequality in development intergenerational human development see higher intergenera- tional persistence of education (a coefficient Intergenerational persistence of that estimates the impact of one additional year persistence of education education of parents’ schooling on respondents’ years of schooling).27 This means that education levels 0.9 across generations are stickier (that is, there is less relative mobility) in more unequal coun- Romania India Guatemala Benin tries (figure 2.3). The component with the Mali strongest correlation coefficient is education, meaning that intergenerational persistence in 0.6 Hungary China education is higher the more unequally distrib- uted the mean years of schooling in a given so- 0.3 Comoros ciety are. As above, no direct causation should be inferred without looking at the mechanisms United Philippines Lesotho behind the correlation, which requires exam- ination at the individual level rather than the Kingdom Maldives country level. The questions are how parents’ socioeconomic status (most importantly their 0.0 education levels) and health status (see the next section) are related to their children’s 0 10 20 30 40 50

Inequality in human development (percent)

Note: The measure of inequality used is the percentage loss in Human Development Index (HDI) value due to inequality in three components: income, education and health. The loss can be understood as a proxy for inequality in capabilities. The correlation coefficient is .4679. Inequality in education is the strongest correlate among the three components (with a correlation coefficient of .5501), followed by inequality in life expectancy (.4632) and inequality in income (.1154). Source: Human Development Report Office using data from GDIM (2018).

76 | HUMAN DEVELOPMENT REPORT 2019 education, and what role do institutions play in 13.7 percent return on investment for compre- Inequalities in the relationship? hensive, high-quality, birth-to­age 5 early ed- education start during ucation, which is even higher than previously infancy, because Inequalities in education start during infancy. estimated.31 However, children from families parents are unequally Exposure to stimuli and the quality of care, both with different socioeconomic status also have able to exploit the in the family and in institutional environments, unequal access to these programmes, nation- opportunity to nurture. are crucial for expanding children’s choices in ally and globally. Enrolment in preprimary But institutions can later life and for helping them develop their full programmes (age 3 to school entrance age) play a crucial role in potential.28 Parents provide stimuli for young ranges from 21 percent in low human devel- fostering mobility children, and families can be nurturing. Parents’ opment countries to 31 percent in medium education shapes the nurturing care provided to human development countries to 74 percent a child from conception to early childhood: a in high human development countries and to home environment that is responsive, emotion- 80 percent in very high human development ally supportive, conducive to children’s health countries.32 and nutrition needs, and developmentally stimulating and appropriate, including oppor- But even if children attend preprimary tunities for play, exploration and protection programmes, disparities in learning abilities from adversity.29 But parents are unequally are often already apparent for the reasons ex- able to exploit the opportunity to nurture. For plained above. Consider the relation between example, children in US professional families average achievement test scores by a child’s age are exposed to more than three times as many and levels of parents’ education in Germany words as children in families receiving welfare (a proxy for socioeconomic status; figure 2.4). benefits.30 This has effects on early learning and The differences in age-specific scores are sub- later achievement test scores, leading to inter- stantial, and they increase enormously during generational persistence in education. the first five years of a child’s life and persist throughout childhood. This does not mean Institutions can play a crucial role in that children do not learn in school (as the tests fostering mobility. For example, there is a

FIGURE 2.4

Skill gaps emerge in early childhood, given parents’ education

Composite Stage score (z) K1 K2 G1 G2 G3 G4 G5 G6 G7 G8 G9 G9

0.5 Preschool Primary Lower secondary 0.3 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 0.2 Age -0.2 -0.3 -0.4 -0.5

1

Low Medium High

Parents’ level of education

Note: Dashed vertical lines emphasize the temporal dynamics of achievement gaps from preschool through lower secondary school. The composite index (z) involves multiple measures at all measurement occasions except 7 months of age, which includes only one assessment (sensorimotor skills), and age 4, which also includes only one assessment (mathematical competence). Predictions are based on life stage­specific regression models. The vertical lines on each dot are 95 percent confidence intervals for predictions. K refers to kindergarten, and G refers to grade level in school. Long-dashed black lines connect data from the same National Educational Panel Study cohort. Source: Skopek and Passaretta 2018.

Chapter 2 Inequalities in human development: Interconnected and persistent | 77 Interventions need to become more difficult), nor does it mean that France, Germany and other European countries consider both how to schooling contributes nothing to help disad- as well as in different institutional and political finish closing the gap vantaged children (because the gaps could, and contexts, such as Soviet Leningrad in the late probably would, significantly increase through- 1960s and the United States in the late 1970s.36 in basic education out childhood were it not for the equalizing Parents with high socioeconomic status can achievements and effect of schooling). But it does highlight the provide direct help, pay for private tutoring, substantial influence of parents’ education computers and travel or move their children to how to stem the on the education achievements of their chil- remedial school or to a less demanding school persistent—or even dren—even in a very high human development and thus give them a second chance.37 country with low inequality in human develop- the increasing— ment and low intergenerational persistence in Another potential source of divergence is so- divergence i n more education.33 Therefore, universal participation cial and emotional learning, which is critical for advanced education in early childhood development programmes, creating productive adults (box 2.1).38 Social even before preprimary education, has the po- and emotional learning is conducive not only achievements tential to reduce inequality in education as well for productivity but also for peaceful social in- as increase education mobility. teraction in cohesive societies.39 Modern forms of education increasingly take such learning In many lower human development countries into account when designing curricula, but it is unequal early childhood stimuli are not the an additional challenge for many low and me- only barrier to mobility in education. Children dium human development countries that are from lower socioeconomic status families may undertaking substantial efforts to provide uni- be unable to attend school because they have versal basic education. There is thus potential duties around the house or on the farm or be- for even more divergence between countries. cause they need to earn income for the family.34 But even if all children had the same grade at- This illustrates a crucial point consistent with tainment, the gap in universal numeracy would the evidence of chapter 1: While much attention close by only 8 percent in India and 25 percent has been paid to raising people above a certain in Pakistan, and the gap in universal literacy “floor,” that does not eliminate the persistence would close by only 8 percent in Uganda and —a nd in some cases the generation—o f steeper 28 percent in Pakistan. So, even if a child from gradients in achievement. Policies geared to a poor household completed as many grades as raising people above a floor fail to boost young a child from a rich household, the likelihood of people’s opportunities to move on to higher becoming numerate or literate would still not education. Interventions thus need to consider be the same for both children. Children from both how to finish closing the gap in basic the poorest 40 percent of households usually education achievements and how to stem the show lower abilities in numeracy and literacy persistent—or even the increasing—divergence at each grade. If these children had the same in more advanced education achievements. learning profiles—that is, the same relation be- tween years of schooling and a measure of skills The effect of the gradient is also carried on or learning— as children from rich families, to the labour market. Someone with high the gap in universal numeracy would close by socioeconomic status but low final education 16 percent in Pakistan and Uganda and 34 per- attainment— such as a member of a privileged cent in India, and the gap in universal literacy family who lacks a university degree or an upper would close by between 13 percent (Uganda) secondary diploma—h as a much higher chance and 44 percent (India).35 Hence, in addition to than a less privileged person of working at a well expanding access to education, gaps in learning paid job and avoiding manual labour. People ability have to be reduced, the earlier the better, from families with high socioeconomic status as the example from Germany shows. often manage to avoid downward occupational mobility relative to their parents, even with poor Early childhood stimuli are not the only education performance.40 A crucial role in this advantage children from high socioeconomic can be attributed to social networks and family status families have. Even if they perform poor- networking activity.41 In some countries improve- ly in school, they are still much more likely to ments in mobility in education have not had the move on to higher education, as evidenced in expected equalizing effect on income because of the increasing importance of networks and

78 | HUMAN DEVELOPMENT REPORT 2019 BOX 2.1 Key competencies of social and emotional learning

Five key social and emotional competencies have been European countries, especially for at-risk children such identified as essential: self-awareness, self-manage- as children from ethnic and cultural minorities, children ment, social awareness, relationship skills and respon- from deprived socioeconomic backgrounds and children sible decisionmaking (see figure). They are interrelated, experiencing social, emotional and mental health chal- synergistic and integral for children’s and adults’ growth lenges.2 Social and emotional learning can thus flatten and development.1 Including and strengthening learning the education gradient by expanding capabilities, with material that teaches social and emotional compe- the potential to reduce inequalities in human develop- tencies in core curricula have been highly effective in ment and promote equity and social inclusion.

Five key social and emotional competencies and how to obtain them

Homes and communities Schools

Classrooms

Self- Self- awareness management

Social Social and Responsible awareness emotional decision- learning making

Soc Relationship tion

ial and skills instruc FeSamcmhoiotliyoolanwnaidldleceoapmrrnamicntuignceictsuyraprniacdrutnlpueomrlischaieinpsds

Source: Jagers, Rivas-Drake and Borowski 2018.

1. Jagers, Rivas-Drake and Borowski 2018. 2. Cefai and others 2018.

networking activities that may at times be more accessible. Moreover, networks are crucial for In today’s job markets, effective than higher levels of education in the entering the labour market. So, important op- which are subject to labour market.42 portunities to redress inequalities exist at three constant technological main points in the lifecycle: early childhood, advances and thus In sum, children start on an unequal footing school age and youth (especially during the reskilling, substantial because of their experiences before entering transition from school to the labour market). investments are the formal education system— particularly, the Additionally, there is a need for lifelong learning. needed at every early education and stimuli that their parents Especially in today’s job markets, which are sub- stage of life provide. Together with differences in the access ject to constant technological advances and thus to and quality of education (see chapter 1), this reskilling, substantial investments are needed at accounts for intergenerational persistence in every stage of life. This is both an economic and education within countries. Children from low a social strategy, in the search for ways to expand socioeconomic status families are less likely to continue education, even if it is available and

Chapter 2 Inequalities in human development: Interconnected and persistent | 79 Parents’ income capabilities throughout life.43 (Part III elaborates status; and whether the mother received prena- and education have on concrete ideas of interventions.) tal health care.49

profound effects Health: How unequal outcomes both And parents’ health behaviour also shapes on their children’s drive and reflect unequal capabilities children’s health after the child is born. For health, which in turn example, child obesity is a result of both na- affects the children’s Parents’ income and education have profound ture and nurture, depending partly on genes education achievement effects on their children’s health, which in turn and partly on family eating and living pat- affects the children’s education achievement terns.50 For adolescents the mechanism of the (and health in (and health in adulthood) and thus future socioeconomic status health gradient works adulthood) and thus income, if not counteracted.44 Hence, health differently. Subjective social status is more gradients— d isparities in health across socio- important for self-reported health than is future income, if economic groups—start at birth, or even before, parent-reported household income and assets, not counteracted and can accumulate over the lifecycle. Higher even when parents’ education is controlled for. socioeconomic status families invest in health, This is either because subjective social status consume more healthily and are mostly able to and self-reported health feed into each other avoid physically and psychosocially demanding due to their bidirectional causal relation or work conditions. This in turn increases the gap because other factors that are more important between low and high socioeconomic status at this stage of the lifecycle weigh strongly on individuals, even resulting in differences in life the subjective social status evaluation (doing expectancy.45 well in school, having friends).51 Even adults’ health outcomes can sometimes be affected by Health conditions at birth, or even before, perceived socioeconomic status (box 2.2). strongly influence health throughout the lifecy- cle.46 And when affected adults become parents The debate around the relationship between themselves, the socioeconomic status health income inequality and health outcomes has gradient can be carried on to future genera- used mainly the proxies of life expectancy at tions, because health inequality starts very early birth and infant mortality.52 But the effects in life—indeed, with the foetus.47 For example, of the socioeconomic status health gradient parents’ occupational status and home postal may not always be fatal, and they may also not code indicate a baby’s health at birth for several be immediate. A nuanced look at different reasons:48 the mother’s eating and other health types of health outcomes reveals how socio- behaviour (smoking), which are closely related economic status affects some specific areas to education; the mother’s exposure to pollu- of health later in the lifecycle (figure 2.5). A tion, which is related to parents’ socioeconomic summary calculation shows that in selected middle-income countries the probability

BOX 2.2 How perceived relative deprivations affect health outcomes

Perceived relative deprivation—how people perceive A potentially mitigating factor for this mechanism their situation compared with others’—leads to poorer is social embeddedness—social connections in in- health outcomes.1 Why is this so? One answer is that terpersonal relationships within social networks and perceived relative deprivation is experienced as an emo- group identities.4 Social embeddedness acts as a buffer, tional state. People feel worse off than others, which dubbed the “social cure,” reducing stress and anxiety.5 causes feelings of anger and resentment.2 Even people Social embeddedness also promotes health because who are objectively well off may feel this, while those socially integrated people exercise more, eat better, who are objectively worse off may not. These emotional smoke less and adhere to medical regimes, unless they states, not always related to actual average inequality in engage in toxic networks that foster risky behaviours.6 a country, cause poorer health outcomes such as greater Health and social embeddedness thus reinforce each self-reported stress and mental and physical illness.3 other.

1. Mishra and Carleton 2015; Sim and others 2018; Smith and others 2012. 2. Smith and others 2012. 3. Van Zomeren 2019. 4. Van Zomeren 2019. 5. Jetten and others 2009. 6. Uchino 2006.

80 | HUMAN DEVELOPMENT REPORT 2019 of poor health outcomes in some aspects of reflected in longevity, low drug use and vac- It is not enough to raise health is two to almost four times higher for cination at all ages.54 Hence, it is not enough people above a certain those in the lowest socioeconomic status group to raise people above a certain floor to ensure floor to ensure that than for those in the highest socioeconomic that gradients do not persist. gradients do not persist status group— a pattern that is similar in the United Kingdom and the United States.53 The Socioeconomic status thus influences health, gradients in m iddle-income countries can be which in turn is pivotal for other opportunities partially related to urbanization (the steepest in life. Policies that redistribute income cannot gradients are in urban areas). They could also break this cycle without addressing the under- reflect deficiencies in the countries’ public lying mechanisms. Universal health coverage is health systems. But even in Sweden, a country needed so that people can use the preventive, well served through universal health coverage, curative, palliative and rehabilitative health ser- gradients in health achievements persist and vices they need (see Sustainable Development sometimes increase throughout the lifecycle. Goal target 3.8). The available services need to Most significantly, having medical experts in be communicated and promoted to the public the family benefits family members’ health as together with information on healthful lifestyles so that people can make educated choices. Still,

FIGURE 2.5 Socioeconomic status affects specific areas of health later in the lifecycle

Bogotá, Colombia Mexico, urban

Poor cognition 0.40 0.75 1.00 1.25 1.75 2.25 3.00 Depression 0.40 0.75 1.00 1.25 1.75 2.25 3.00 Function Function

Depression Poor cognition Stroke Poor SRH Poor SRH Diabetes Hypertension Heart disease Diabetes Hypertension Obesity Heart disease Obesity Stroke

Chance of poor health outcome Chance of poor health outcome

South Africa, urban United States

Poor SRH 0.40 0.75 1.00 1.25 1.75 2.25 3.00 Poor cognition 0.40 0.75 1.00 1.25 1.75 2.25 3.00 Poor cognition Poor SRH Hypertension Walking

Depression Depression Heart disease Function Diabetes Stroke Obesity Function Diabetes Hypertension Obesity Stroke Heart disease

Chance of poor health outcome Chance of poor health outcome

Low socioeconomic status High socioeconomic status

SRH is self-reported health. Note: The chance of poor health outcome was calculated with the odds ratio (log scale). Data for Colombia are from the Survey on Health, Well-being and Aging), data for Mexico and South Africa are from the Study on Global Ageing and Adult Health and data for the United States are from the Health and Retirement Study. Values greater than 1 (the vertical line) indicate a greater chance of a particular health outcome compared with people with mid-socioeconomic status, and values less than 1 indicate a lower chance. For example, in Bogotá, Mexico and the United States the chance of poor cognition is nearly two times higher for people with low socioeconomic status than it is for people with mid-socioeconomic status but much lower for people with high socioeconomic status. Source: Adapted from McEniry and others (2018).

Chapter 2 Inequalities in human development: Interconnected and persistent | 81 What came to be tackling gradients in health cannot be achieved eventually be reached, and inequality would known as the Kuznets simply by gearing policies towards providing a start to fall (given the very low weight of the hypothesis predicted minimum level of access to health services to all. agricultural and rural sector). an inverse-U relation Other social determinants are also relevant. What came to be known as the Kuznets hy- (or curve) between How inequalities interact with pothesis thus predicted an inverse-U relation income levels and other contextual determinants (or curve) between income levels and income income inequality, of human development inequality, with structural change as the main mechanism accounting for the relation. This with structural This section moves beyond the individual-level, became the most enduring legacy of Simon change as the main lifecycle analysis and considers how inequalities Kuznets’s 1955 article, but it was by no means mechanism accounting interact with other contextual determinants of the only contribution of that work. human development. Not intended to be com- for the relation prehensive, it considers four dimensions that Simon Kuznets analysed other mechanisms are crucial for human development: the econ- that he thought influenced the interplay among omy (how inequalities interact with patterns of growth, structural change and inequality. These economic growth), the society (how inequali- ranged from demographic changes (includ- ties affect social cohesion), the political arena ing the economic paths of immigrants into (how political participation and the exercise fast-growing modernizing economies) to the of political power are influenced by inequali- influence of political processes in determining ties) and peace and security (how inequalities the distribution of income: “In democratic interact with violence, which is influenced by societies the growing political power of the economic, social and political factors). urban lower-income groups led to a variety of protective and supporting legislation, much Income and wealth inequalities, of it aimed to counteract the worst effects of economic growth and rapid industrialization and urbanization and structural change to support the claims of the broad masses for more adequate shares of the growing income of There are longstanding debates on the rela- the country.”56 The more nuanced and sophis- tion among structural change in an economy, ticated analysis in Kuznets’s original article has economic growth, and income and wealth in- been lost over time, replaced almost exclusively equality. Sustained economic growth typically by a description of a mechanistic relation be- happens with structural shifts in the economy tween growth and inequality.57 And perhaps (with employment and value added moving the Kuznets hypothesis can be best understood from agriculture to both manufacturing and as describing the evolution of income during services). But the relation with income distri- major phases of structural change, in “Kuznets bution is more ambiguous. Simon Kuznets waves,” as opposed to a deterministic “once and was the first to take up the issue systematical- for all” pathway for inequality as economies ly, putting forward the hypothesis that with develop.58 economic growth, as labour moved away from the agricultural and rural sector to nonagri- In addition, structural change, growth and cultural and urban economic activities (with a inequality can interact through mechanisms higher mean wage than agriculture and a more other than the changes in sectoral composition widespread distribution of earnings), there highlighted by Simon Kuznets. The nature would be two stages in the evolution of overall of technological change and how it interacts income distribution.55 During the initial stage with labour markets is a particularly important economywide inequality would increase with channel. Jan Tinbergen posited that if tech- economic growth as the relative weight of the nological change is skill-biased—that is, if it nonagricultural sector expanded from very demands higher skilled workers— then a “race” low levels. But as the share of labour in the ag- between technology and skill supply would be ricultural sector shrank, a tipping point would expected.59 With technology forging ahead, if skill supply lags, then a wage premium would be expected for higher skills, increasing wages at the top of the skill/income distribution and thus inequality, as lower skilled workers fail to

82 | HUMAN DEVELOPMENT REPORT 2019 keep up with the race. There is some evidence to globalization (nontradable, in more techni- With technology that is consistent with this hypothesis for some cal terms, such as personal and social care, for forging ahead, if skill developed economies in the latter part of the instance) can be in high demand, even if they supply lags, then 20th century,60 but Tinbergen’s “race” does not correspond to low skills.62 It is in the middle a wage premium seem to account fully for more recent develop- of the skill distribution, with several tasks in would be expected ments in labour markets this century. the manufacturing sector, that there is higher for higher skills, vulnerability to offshoring or technology re- increasing wages at Rather than a steep gradient, many labour placement, which explains the hollowing out of the top of the skill/ markets in developed economies have polar- the middle.63 These factors seem to be at play in income distribution ized. This polarization is sometimes manifested some developing countries as well.64 Over the and thus inequality, as with an increase in the labour shares both at the course of this century there has been a hollow- lower skilled workers bottom and the top of the skill distribution and ing out of the middle, in this case measured by fail to keep up a hollowing out at the middle.61 Jan Tinbergen’s changes across the wage distribution in South race model, therefore, needs to be adjusted to Africa (figure 2.6).65 This can be accounted for account for wage growth at the bottom— in part by these mechanisms, along with the assuming that the same mechanism can explain fact that labour market institutions such as the either wage increases or gains in employment minimum wage do not protect those in the shares at the top. A large literature has emerged middle and that trade unions have been cap- to account for job polarization, premised on tured in part by those at the top. The relation the concept that not only technology but also between polarization and inequality is still con- other factors—i ncluding trade—determine the tested, with the impact on aggregate inequality demand for skills. measures ambiguous.66

The most influential approach in this field The debate has ebbed and flowed on the considers tasks and assesses the extent to which empirical validity of the Kuznets hypothesis, they can be easily replaced by either technolo- its interpretation, alternative mechanisms, di- gy or globalization (with production moving rections of causality and the relation between to lower labour cost economies). With this economic growth and income inequality.67 framework, some tasks that are nonroutine Assessing the weight of the empirical evidence (thus difficult to automate) and more immune

FIGURE 2.6 The hollowing out of the middle in South Africa

Annual average growth rate of real earnings, 2001­2015 (percent) 8 7 6 5 4 3 2 1 0

0 6 12 18 24 30 36 42 48 54 60 66 72 78 84 90 96 -1

Wage percentile

Source: Bhorat and others 2019.

Chapter 2 Inequalities in human development: Interconnected and persistent | 83 What matters is to is particularly challenging, given the range of whether income is accruing to the middle class identify policies that income inequality measures in the literature or to the bottom of the distribution. Moreover, as well as the difficulty of disentangling meas- since at least Simon Kuznets’s 1955 article, it can lead both to urement error from plausible causal relations.68 has been well understood that growth processes growth and to more Further compounding the analysis are factors can at times be unequalizing. What matters is that, at some point in history and in some to identify policies that can lead both to growth inclusive sharing contexts, have a greater bearing on inequality and to more inclusive sharing of the gains from of the gains from than either growth or structural change. This is expanding income. expanding income at the heart of Thomas Piketty’s critique of the Kuznets hypothesis, which argues that inequal- Identifying these more inclusive growth ity dynamics depend primarily on institutions patterns matters in particular for those at the and policies.69 And Walter Scheidel argues that bottom of the income distribution. In this violence and major epidemics have historically case, the redistribution of productive capacity been the greatest downward drivers of inequali- (leading to the accumulation of assets, access ty, not structural change or policies.70 to markets and connection of returns to asset use at the bottom) can lead to both growth Beyond the more secular and longer term and income gains at the bottom, reducing in- structural approach explored by Simon Kuznets equality.77 More mechanically, interactions be- and the subsequent debate is the related ques- tween growth and inequality affect how much tion of whether there are tradeoffs between income flows to poor people.78 As a matter of growth and inequality over shorter time spans. pure arithmetic decomposition, the impact of Concerns with efficiency, or how much in- expanding mean income on poverty depends come is growing, have traditionally dominated on the growth rate as well as on how much concerns with equity, or how it is distributed. additional income flows to the bottom of the Arthur Okun has suggested a tradeoff between distribution.79 Redistribution to the bottom economic efficiency and equality, arguing can create more than a one-off reduction in that more equality could weaken economic poverty and inequality— it can change the pov- growth by harming incentives to work, save erty elasticity of income, which would make and invest.71 And because income growth has growth more impactful on poverty reduction such an overwhelming impact over the longer over time.80 A recent simulation exercise run in improving living standards, the impact quantifies how reducing inequality could help of redistributing production would pale in reduce poverty using those direct relationships. comparison with the “apparently limitless The number of extremely poor people would potential of increasing production.”72 Yet re- remain above 550 million in 2030 if GDP per cent empirical studies have found that higher capita were to grow according to International income inequality can be associated with Monetary Fund forecasts and inequality were lower and less durable growth,73 including in held constant. But reducing the Gini index by developing countries.74 But both the data and 1 percent a year in each country would cut the techniques used in some of these econometric global poverty rate to about 5 percent in 2030, studies remain contested, casting a shadow of which would bring 100 million more people uncertainty over claims that inequality is either out of extreme poverty. 81 “bad” or “good” for economic growth.75 In the spirit of understanding further pos- Ultimately, it is less relevant to explore sible mechanisms for the interaction between whether inequality is harmful to growth (in inequality and growth, one hypothesis is that a mechanistic way) than to understand the if high inequality reduces mobility, that would impact of policies on income distribution and lead to an inefficient allocation of resources economic growth.76 And the evaluation of the (talent, skills and capital) that, compared with impact of policies on distribution, in turn, a counterfactual in which the resources are depends on the weights that society and poli- allocated efficiently, would hurt growth. If this cymakers attribute to different segments of the mechanism holds, there would be a negative im- population. Thus, blanket statements on the pact of income inequality on economic growth, effect of inequality on growth are not helpful, with the channel running through inequalities in part because they do not enable insights into

84 | HUMAN DEVELOPMENT REPORT 2019 in opportunity.82 Yet, once again, the empirical in strangers is not seen as risky.90 But higher When horizontal support for this channel is ambiguous.83 inequality may cause the less wealthy to feel inequalities are high, or powerless and less trusting in a society generally perceived to be high, Another hypothesis is that the relation perceived as unfair, while people at the top may people may withdraw works through efficiency: Productivity, and not feel that they share the same fate as people from certain social hence GDP, increase most when resources are at the bottom or that they should strive towards interactions, which efficiently used and the potential for technolog- a common goal.91 can diminish trust ical learning is fully exploited.84 This has been and social cohesion shown historically by the East Asian growth Empirical evidence shows that in developed model. Investments in education, among countries the higher the income inequality, the others, have contributed to economic growth lower the level of trust within society.92 And through productivity increases.85 Productivity in European countries with higher income is lower in most countries with high income inequality, people are less willing to improve inequality than in countries with low in- the living conditions of others, independent of come inequality.86 One reason could be that household income, while there is probably less inequality reduces incentives for innovation solidarity and people are less likely to support and investment through various supply-side redistributive institutions.93 The interaction mechanisms.87 between inequalities and solidarity may thus go in both directions. The relation could also work in reverse: Slow economic growth could increase inequality When horizontal inequalities are high, or under certain circumstances. For instance, perceived to be high, people may withdraw when rates of return are higher than economic from certain social interactions (box 2.3), which growth, especially for large wealth portfolios, can also diminish trust and social cohesion.94 In wealth inequality tends to increase.88 Together highly unequal countries people from differ- with other mechanisms contributing to the rise ent social strands are also less likely to mingle of top-end bargaining power and high incomes and interact.95 They probably live in different (including top executive compensation), this neighbourhoods, their children attend different dynamic could create a vicious circle of slow schools, they read different newspapers and growth and high inequality. they are in different groups on social media (box 2.4). Their worldviews likely differ, and Trust and social interaction they know little about the fate of their fellow in unequal societies citizens. People who do not meet and interact do not directly see the concerns and needs of Income inequality can damage social cohesion others (see box 1.9 in chapter 1),96 which may in societies. Trust, solidarity and social interac- reduce support for equalizing policies. tion can be diminished by large income gaps, impairing the social contract (sets of rules and A comparison between Canada and the expectations of behaviour with which people United States at the subnational level shows voluntarily conform that underpin stable the effect of segregation on intergenerational societies). But does income inequality simply income mobility. On average, mobility is lower damage social cohesion, or is the relation two- in the United States than in Canada, but at the way— does low social cohesion block redistrib- subnational level the southern United States is utive policies? least mobile, like northern Canada. One reason for low mobility in the southern United States Important features of social cohesion include is the history of exclusion of African Americans, the strength of social relationships, shared many of whom have not been fully integrated values, feelings of identity and the sense of into the economic mainstream.97 Some parts belonging to a certain community.89 One of of northern Canada also have lower mobility the most common measures of social cohesion than the rest of the country, due most likely to is the level of trust among society. Trusting the remote geographic locations of some indig- people means accepting strangers as part of the enous peoples, which make their integration community and sharing with them the under- into the economy challenging. However, their lying commonality of values. Trust is based on proportion of the population is much smaller senses of optimism and control: Putting faith

Chapter 2 Inequalities in human development: Interconnected and persistent | 85 BOX 2.3 The power of perceived inequalities in South Africa

South Africa is an interesting case study of social cohe- even increased over time are less likely to participate sion and inequalities, given its history of racial segre- in interracial socialization than those who perceive gation and related vertical and horizontal inequalities. that inequality is declining. Across race groups, interra- According to multidimensional living standards meas- cial socialization and the desire to interact increase as ures, inequality has declined significantly among indi- perceived inequality declines (see figure). The desire to viduals and among races since 2008. And yet interracial interact is crucial here, as it varies from the actual inter- interactions—measured by actual interracial social in- actions due to circumstances. The finding remains sig- teractions, the desire to interact and the desire to know nificant even after a multidimensional Living Standards about the customs of people of other races—have also Measure, race, education, trust and other measures are declined since 2010. While interracial interaction is just controlled for. one part of social cohesion, it is crucial in South Africa. These findings are thus counterintuitive and run contrary These findings are important because interracial to the empirical findings of other countries. interaction is crucial for social cohesion in South Africa. Social cohesion in turn increases the possibility of con- One possible explanation is that perceived trends in sensus on equalizing policies that reduce inequality. inequality, which are substantially different from actu- There is also weak evidence for reduced objective ine- al trends, are more important for predicting interracial quality improving social cohesion. This opens an oppor- socialization. The roughly 70 percent of South Africans tunity to create a virtuous cycle of social cohesion and who feel that inequality has not changed much or has low inequalities.

More interracial interaction with lower perceived inequalities

Actual interracial interaction Perceived Desired interracial interaction inequality Percent Percent 0 20 40 60 80 Improved 0 20 40 60 80

Talk Desire to talk Socialize Learn custom

Talk Stayed Desire to talk Socialize the same Learn custom

Talk Worsened Desire to talk Socialize somewhat Learn custom

Talk Worsened a Desire to talk Socialize great deal Learn custom

Talk Don’t Desire to talk Socialize know Learn custom

Source: David and others 2018. White Asian/Indian Coloured Black Source: David and others 2018.

86 | HUMAN DEVELOPMENT REPORT 2019 BOX 2.4 The power of your neighbour

Human beings do not act in isolation—their behaviour depends partly on 30 percent brings segregation down to about 75 percent (see figure).4 Only the behaviour of people in their cognitive neighbourhood.1 An example from lowering the preference to the single digits results in very low emergent agent-based models demonstrates the emergent nature of human inequal- segregation (for example, 9 percent leads to 52 percent). This means that ities.2 A model of neighbourhood segregation along ethnic lines—which people of similar ethnic characteristics automatically move closer together. can be thought of as a form of geographic inequality—shows that even These behavioural patterns can accelerate inequalities due to the power when there are few individual prejudices, segregation can nonetheless arise of the neighbourhood effect—an expression used to describe the impact merely from the interaction of individuals.3 of neighbourhood on the possibility of an individual moving up the social ladder, especially through the influence of peers and role models. In most The segregation model has two types of agents—red and green—in developing countries neighbourhood effects are likely to be even stronger equal numbers, each occupying one “patch” of the model’s environment given the vast differences in the provision of public goods and services, es- (equivalent to a house). On average, each agent begins with an equal pecially between rural and urban areas.5 number of green and red neighbours. A key parameter is the average per- centage of same-colour neighbouring agents wish to live near (such as However, public policy interventions can help shape human behaviour, 30 percent or 70 percent). If an agent does not have enough neighbours providing counterincentives to mitigate the power of the neighbourhood ef- of its own colour (according to the preference parameter), they move to fect. In the United States inequality in housing prices limits workers’ ability a spot nearby. to move to a location with higher earning potential.6 Similarly, the quality of public services such as schools can differ across neighbourhoods, fur- The results of the simulation are dramatic. Starting from a preference ther heightening inequalities. Government subsidies for housing or equal- for perfect equality (having 50 percent of one’s neighbours the same col- ly good quality public schools could help offset this effect. The Moving to our), agents’ individual movements give rise to an aggregate segregation of Opportunity experiment showed the effectiveness of these policies by of- around 86 percent (in other words, roughly 86 percent of one’s neighbours fering randomly selected families housing vouchers to move into better off end up being the same colour despite each person wishing to have a 50 per- neighbourhoods. The move increased college attendance and earnings for cent level of diversity). Reducing the preference to 40 percent results in the people who moved during childhood.7 overall rate of segregation dropping to around 83 percent; reducing it to

How segregation can arise from interaction After interaction between agents Starting point with equal number of green and red neighbours

Source: Wilensky 1997. 1. Iversen, Krishna and Sen 2019. 2. Agent-based models have been used to predict human behaviour. Using a variety of software tools, agent-based models typically create a group of agents (people, firms, trees, animals, societies, countries and so on), design simple behavioural rules (either for all agents or for subgroups), place the agents in a given simulated environment (usually consisting of time and space dimensions) and then set the agents free to interact based on the behavioural rules. The point of the simulation is to see what emergent phenomena and aggregate properties arise from the interactions based on these basic settings, with no ex ante determination of equilibrium or any other goal. 3. Schelling 1978. 4. The exact numbers depend on the specific run of the simulation and on the density parameter (that is, the proportion of the neighbourhood that is occupied; in this case 95 percent). 5. Iversen, Krishna and Sen 2019. 6. Bayoumi and Barkema 2019. 7. Chetty, Hendren and Katz 2016.

Chapter 2 Inequalities in human development: Interconnected and persistent | 87 Government policy than the African American population in the becomes constrained because political decisions space to address southern United States.98 reflect the balance of power in society. This is often referred to as elite capture of institutions.104 inequalities becomes When more incentives for interaction are constrained because directed towards diversity (including people Power asymmetries can even lead to break- from all ethnicities, religions and social strands) downs in institutional functions, constraining political decisions interaction, trust, networks and social cohesion the effectiveness of policies. When institutions reflect the balance can be built.99 Ethnicity quotas and subsidies are afflicted by clientelism and captured by of power in society. for cultural activities, civic associations, schools elites, citizens may be less willing to cooperate This is often referred and the like could be an effective way of facil- on social contracts. When that translates into, to as elite capture itating interaction in the long run. Initially for instance, lower compliance with paying tax- people may resist interaction, and there could es, the state’s ability to provide quality public of institutions be a temporary decline in trust, but in the long services is diminished. This, in turn can lead to run intergroup interaction counters these in- higher and more persistent inequalities—for itial negative effects, increasing trust and even instance, in health and education. As the overall improving the perceived quality of life.100 system will be perceived as unfair, people tend to withdraw from political processes, which The cycle of social cohesion and inequalities further strengthens the influence of elites.105 is strongly connected to the cycle of education and inequalities, which, again, is connected to In a world in which information becomes the cycle of health gradients. Education can cre- more and more accessible and important, ate strong social bonds among different groups media is a decisive channel through which the in a society by teaching people about different imbalances of power can be further amplified. cultures and bringing them into contact with Different stakeholders “create, tap, or steer people of different backgrounds. Likewise, it information flows in ways that suit their goals can teach norms and values and promote par- and in ways that modify, enable, or disable ticipatory and active citizenship. But schools the agency of others, across and between a can also act as a flattener for the health gradi- range of older and newer media settings.”106 ent, teaching children healthy habits and how Even though information is easily accessible to follow a balanced and nutritious diet.101 The for many people, not everyone is equally well convergence in primary and secondary educa- informed. In countries with high internet pen- tion (see chapter 1) thus gives hope for creating etration, income inequality correlates positively virtuous cycles of equity in the future. with both information inequality (measured by the Gini coefficient estimated over the How inequalities are transferred number of news sources individuals use) and into political inequality— a nd back information poverty (the probability of using zero or only one news source). In Australia, Most of the literature has found that in high hu- the United Kingdom and the United States, man development countries inequalities depress where income and information inequality are political participation, specifically the frequency high, 1 individual in 10 uses zero or only one of political discussion and participation in news source (information poverty).107 Less well elections among all citizens but the richest.102 informed voters become more susceptible to Economic elites (or sometimes even the upper the above described political influence by the middle class) and organized groups representing few media sources they consume. Depending business interests thus shape policies substantially on how these sources are financed, they may more than average citizens or mass-based interest promote and protect the interests of a specific groups do. Additionally, mechanisms through group. This form of biased reporting has been which this can happen include opinion making, referred to as media power.108 A combination lobbying and clientelism.103 Income and wealth of high information poverty and media power inequalities are thus transferred into political ine- can weaken democratic processes109 because it quality (box 2.5), with privileged groups mould- can influence voters’ behaviour, which is espe- ing the system according to their needs and cially delicate with fake news.110 preferences, leading to even more inequalities. Government policy space to address inequalities Inequalities can also increase both the demand for and supply of populist and authoritarian

88 | HUMAN DEVELOPMENT REPORT 2019 BOX 2.5 Elizabeth Anderson, Arthur F. Thurnau Professor and John Dewey Economic inequality and human development Distinguished University Professor of Philosophy and Women’s Studies at the University of Michigan

How does inequality matter for human development? It limits the prospects top 1 percent of income and wealth distributions3 as well as by a small or for development of the less advantaged. It undermines the ability of untar- stagnant middle class. geted pro-growth policies to reduce poverty because most of the growth will be appropriated by the better-off. And it reduces social mobility by enabling The independent normative significance of inequality suggests that advantaged groups to hoard opportunities and close ranks against those abolishing poverty and deprivation should not be the only aim; the concen- beneath them. tration of income and wealth at the top should also be limited.4 In 2019 the richest 26 individuals in the world owned as much wealth as the bottom Beyond these concerns, political theorists have drawn attention to the half of the world’s population.5 There is no normative justification for such relational aspects of inequality, beyond the bare facts of distributive ine- extreme inequality. The wealth of the ultra-rich has not always been accu- quality: Distributive inequalities reflect, reproduce and sometimes consti- mulated legally—given the vast scale of global corruption, organized crime, tute oppressive social relations of domination, esteem and standing.1 It is financial manipulation, money laundering and tax evasion. But even when not simply the material injury of wage theft or of being physically beaten by it has, that would only call into question the justification of laws so heavily a domestic partner but the fact of living in subjection to others who wield tilted towards the interests of the rich. It is absurd to credit such inequality the power to inflict harm with impunity and who feel free to sacrifice one’s to differences in merit, given the rising capital share of income, which re- vital interests to their own greed or vanity that not only deprives but also wards mere ownership, and the large impact of chance on outcomes. Nor oppresses. It is not simply the bare fact of lacking adequate clothing but the can such extreme inequality be rationalized as necessary for poverty reduc- stigma others attach to such deprivation that makes poverty sting. It is not tion or as socially advantageous in any other way. Extreme wealth does not simply the physical difficulty the disabled have of navigating public spaces even enhance the consumption possibilities of the ultra-rich, who cannot but also the little account public architects and public policy have given to personally consume all of their wealth or even a significant fraction of it. their interests that not only inconveniences but constitutes their diminished standing in the eyes of others. Indeed, most of what the ultra-rich do with their wealth is exercise power over others. If they own, direct or manage a firm, they deploy their Across the world, inequality tracks differences of social identity such wealth to control their workers and their working conditions. If they hold a as gender, race, ethnicity, religion, caste, class and sexual orientation— monopoly or monopsony position, they may dominate consumers, suppliers arbitrarily marking some social groups as superior to others in the oppor- and the communities where they operate. If they lobby or donate money to tunities they enjoy, the powers they command and the respect others owe politicians, they capture the state. The ultra-rich also have disproportionate them. Under such conditions members of subordinated groups lack effective clout in global institutions, particularly regarding the rules of global finance, means to vindicate their human rights, even in states that legally acknowl- which have contributed to systemic financial risks and to the instability ex- edge these rights. Groups targeted for sexual harassment and assault can- perienced by many countries around the world. not vindicate their rights if social or legal norms systematically disparage the credibility of their testimony. Groups subject to disproportionate siting The current era is witnessing global democratic backsliding, following of toxic waste dumps and polluting industries cannot vindicate their rights a surge of democratization in the 1990s and early 2000s. Freedom House if they are disenfranchised or if state decisionmakers are otherwise unac- reports that 22 of 41 democracies have become less free in the last five countable to them. Groups denied effective access to education cannot vin- years.6 While the causal connections between distributive inequality (in- dicate their rights if they do not know what their rights are or lack the ability cluding extreme concentrations of wealth at the top and declining prospects to navigate the judicial and bureaucratic processes needed to secure them. for the global middle) and the decay of democratic norms and institutions have yet to be fully explored, what is already known should raise alarms. Distributive inequality for social relations undermines trust among While the ultra-rich might escape the worst of unmitigated global climate members of society as well as trust in institutions. It depresses political, civ- change, what will happen to the billions left homeless, sick or stateless ic, social and cultural participation. It spurs communal violence and crime. It by rising sea levels, extreme floods, droughts, heat waves and attendant undermines democracy by enabling the rich to capture the state and thereby social conflict and civil war? The great inequalities defined by citizenship appropriate a disproportionate share of public goods, shift tax burdens in status threaten the freedom of environmental and wartime refugees, while a regressive direction, enforce fiscal austerity and avoid accountability for politicians in receiving states attack democratic institutions in the name of predatory and criminal behaviour. Even the laws and regulations that consti- closing their borders. Just at the point where meeting the challenges of tute the basic economic infrastructure of markets, property and firms have climate change is demanding ever-greater international cooperation, states been designed under the influence of powerful groups to rig purportedly are retreating from global institutions. Greater attention to the case for neutral rules in their interests.2 equality, both within and between states and in the governance of global institutions, is needed to promote human development and cope with the These effects occur in states at all levels of human development, even greatest challenge humanity faces in the 21st century. those with low poverty. They are exacerbated by extreme inequalities in the

1. Anderson 1999; Fourie, Schuppert and Wallimann-Helmer 2015. 2. Harcourt 2011; Pistor 2019. 3. Piketty 2014. 4. Robeyns 2019. 5. Oxfam 2019. 6. Freedom House 2019.

Chapter 2 Inequalities in human development: Interconnected and persistent | 89 The way in which leaders. When higher inequalities lead to an outcomes such as economic inequality (and power asymmetries enhanced sense of systemic unfairness, it can growth). However, by redistributing economic play out in the policy raise the public’s openness to nonmainstream resources, these policies are also redistributing arena can exacerbate political movements.111 In some contexts polit- de facto power (the top arrow in the right loop ical participation increases under high income of figure 2.7). This can generate (or reinforce) and entrench inequality, when populist leaders trigger griev- power asymmetries between actors bargaining inequalities or pave the ances by explicitly connecting political and in the policy arena, which can in turn adversely way to more equalizing socioeconomic exclusion.112 More generally, affect the effective implementation of devel- and inclusive dynamics populist leaders use economic anxiety, public opment policies. For example, power asym- anger and the reduced legitimacy of status quo metries can manifest in the capture of policies parties to build narratives that exploit one of the by elite actors—undermining the ability of following two cleavages: Right-wing populism governments to commit to achieving long-term thrives on cultural cleavages, including religious, goals. Or they may manifest in the exclusion ethnic or national differences, while left-wing of certain population groups from accessing populism emphasizes economic differences be- high-quality public services—undermining tween the wealthy elite and the lower classes.113 cooperation by harming tax morale. This can Both divide society and weaken social cohesion. lead to a vicious cycle of inequality (inequality traps) in which unequal societies begin to insti- One way of understanding the interplay tutionalize the inequality. This loop plays out between inequality and the dynamics of power in prevailing institutions and social norms (the is to draw on a framework that explores one of outcome game) and can lead to actors deciding the processes through which inequalities are to change the rules of the game (the bottom ar- generated and perpetuated. At its core, this row in the left loop of figure 2.7). In this way, de process is often referred to as governance—or jure power is also redistributed. This can be far the way in which different actors in society more consequential because it not only changes bargain to reach agreements (policies and current development outcomes but also sets rules). When these agreements take the form of the conditions that shape actors’ behaviour in policies, they have the power to directly impact the future. Once again, the way in which power the distribution of resources in society (the asymmetries play out in the policy arena can ex- bottom arrow in the right loop of figure 2.7, acerbate and entrench inequalities or pave the “outcome game”). For example, policies on tax- way to more equalizing and inclusive dynamics. ation and social spending determine who pays This is one clear way in which inequality may into the fiscal system and who benefits from it. undermine the effectiveness of governance.114 These policies directly influence development

FIGURE 2.7 The effectiveness of governance: An infinity loop

De jure power De facto power

Rules Policy Development arena outcomes

Rules game Outcome game

Note: Rules refer to formal and informal rules (norms). Development outcomes refer to security, growth and equity.

90 | HUMAN DEVELOPMENT REPORT 2019 Violence and inequalities: The distinguishing itself from the others by its his- Political disturbances cruellest vicious cycle tory, religion, language, race, region, class or the —including violent like.122 Group differences appear in all societies, conflict and civil This last section expounds on what can be but they are only likely to lead to conflict and war—can a rise from considered the two cruellest vicious cycles: the violence when social, economic and political horizontal inequalities relations between inequalities and homicides inequalities are exacerbated by politically ex- and violent conflict. There are more homicides cluding certain groups.123 in countries with higher income inequality across all categories of human development. A condition for horizontal inequalities to For high and very high human development lead to conflict is that leaders or elites have countries the association is strong: Income an interest in mobilizing groups and initiat- inequality explains almost a third of the overall ing a conflict. That interest often arises from variation in homicide rates, even after years of horizontal political inequalities among the schooling, GDP per capita, democratization elite.124 Added to this are more determinants and ethnic fractionalization are accounted of conflict: the nature of the state, the role for.115 Education has a moderating effect on of local institutions, the presence of natural this relation, but only in high and very high hu- resources125 and the struggle between some man development countries: 1.8 more years of groups for access to power, resources, services average schooling more than halves the associa- and security.126 tion between income inequality and homicide rates.116 Findings from a study of Mexico’s drug Shocks can also interact with horizontal war are in line with the hypothesis that income inequalities and contribute to outbreaks of inequality is associated with more violence. A instability. One example is the contribution 1 point increase in the Gini coefficient between of the drought that affected Syria prior to the 2006 and 2010 translated into an increase uprisings of 2011, showing how shocks and of more than 10 drug-related homicides per horizontal inequalities (primarily between the 100,000 inhabitants.117 rural population affected by the drought and the population in urban areas) can interact to The mechanism behind this relation is less trigger instability.127 clear. Some suggest that the feeling of shame and humiliation in unequal societies drives vi- While only 9 percent of armed conflict olence, predominantly by young men pressured outbreaks between 1980 and 2010 coincided to ensure status.118 Others suggest a psychosocial with disasters such as droughts or heatwaves, explanation: Income inequality intensifies so- the proportion increases to 23 percent in eth- cial hierarchies, causing social anxiety and class nically fractionalized settings, where disruptive conflict, damaging trust and social cohesion.119 events seem to play out in a particularly tragic This is empirically supported by data showing a way.128 Droughts also significantly increase the negative correlation between trust and income likelihood of sustained violent conflict in low- inequality— at least in developed countries (see income settings where ethnically or politically above). Societies with low trust and weak social excluded groups depend on agriculture. This cohesion have lower capacity to create safe com- leads to a vicious cycle between violent conflict munities, and this, together with high pressure and environmental shocks, with the groups’ for status, may increase violence. vulnerability to one increasing their vulnerabil- ity to the other.129 On a macro level, evidence about the rela- tion between inequalities and violent conflict Comparisons of civil and communal con- is mixed. Some studies find that income in- flicts among 155 politically relevant ethnic equality triggers instability that may lead to groups in Africa show that both political and violence.120 Others find no relation between economic horizontal inequalities can lead income inequality and violent conflict.121 to conflict. But the targets of violence differ. More recently, Frances Stewart has argued Political exclusion leads to violence that targets that political disturbances— including violent the central government. Horizontal income conflict and civil war— arise from horizontal or wealth inequalities act more broadly as a inequalities between different groups, each determinant of organized political violence, increasing the risk of civil and communal con- flicts. Communal conflicts appear to be driven

Chapter 2 Inequalities in human development: Interconnected and persistent | 91 Some forms of mostly by politically included groups with less social unrest increases when individuals per- horizontal inequalities reason to fear government intervention.130 ceive their group as disadvantaged. Support for violence is highest when included groups increase before, Afrobarometer perceptions data suggest that enjoying high political status perceive that the during and in the not only real horizontal inequalities but also government treats them unfairly. But the effect immediate years after perceived inequalities and exclusion matter of exclusion on support for violence can also the onset of conflict for conflict (see box 2.3). The likelihood of

BOX 2.6

Internal armed conflict and horizontal inequalities Peace Research Institute Oslo

The impact of internal armed conflict on horizontal ine- no longer imposes direct costs (on some areas).3 Yet, qualities can play out in several ways. In some cases it postconflict redistributions of power and resources can reduce horizontal inequalities,1 while in others it can may depend on the outcome of the conflict. Patterns exacerbate them. First, if the costs of internal conflict are of inequality in the aftermath of conflict may be con- greatest for those who are already poorest,2 horizontal in- tingent on whether the outcome is a postconflict equalities may increase. Many countries and areas expe- agreement securing the interests of both the losers riencing armed conflicts had high horizontal inequalities and the winners. prior to the conflict, and such inequalities are exacerbat- ed when the most disadvantaged groups are dispropor- In the years prior to armed conflict, regional ine- tionately affected by it. Second, internal armed conflict is quality in infant mortality rates—used here as a proxy often restricted to or focused largely in certain areas of a for one dimension of horizontal inequalities—increases country. These areas, and the groups that reside in them, (see figure). This increase continues in the immediate may be cut off from the rest of society and the economy. years (1­5) after the onset of conflict, which is con- Some areas will also suffer disproportionally from the de- sistent with the argument that horizontal inequality in- struction of facilities, buildings and human lives. creases during conflict. But this acceleration wears off after 5­10 years. Hence, some evidence suggests that In the postconflict phase these outcomes may the postconflict phase is associated with a decrease in wear off, as the economy picks up and the conflict a measure of one dimension of horizontal inequalities.

Regional inequality in infant mortality rates prior to and after conflict onset

Regional inequality in infant mortality rates (deviation from country-mean) 0.03

Outbreak of war

0.02

0.01

0

­0.01

­0.02

­0.03

­0.04

­4 ­2 0 2 4 6 8 10

Phases of war onset (years)

Note: The x-axis is the number of years prior to and after the onset of conflict. Conflict is defined here as armed conflict with at least 1,000 battle deaths. The y-axis is the global average of countries’ deviation from their mean level of horizontal inequality. In other words, it captures whether countries have higher or lower horizontal inequality than usual. Regional inequality is measured using the ratio between best- and worst-performing region in infant mortality rates. Source: Dahlum and others forthcoming.

1. Women’s political participation, for instance, often increases in postconflict settings (World Bank 2017b). 2. Gates and others 2012. 3. Bircan, Brück and Vothknecht 2017. Source: Dahlum and others forthcoming.

92 | HUMAN DEVELOPMENT REPORT 2019 be attenuated by subjective perceptions (on economic redistribution, offer opportunities to Income inequality perceptions of inequalities, see spotlight 1.2 in prevent recurrence.137 increases during chapter 1).131 violent conflict Inequalities can accumulate and during the Horizontal inequalities can drive violent through life, reflecting first five years of conflict, and in some cases they may increase deep power imbalances typical postwar even more before, during and in the immediate reconstruction. But years after the onset of conflict (box 2.6). Even This chapter has taken a dual approach in violent conflicts can though major conflicts such as World War I revealing the mechanisms through which in- also widen inequalities and World War II can bring income inequality equalities in key areas of human development in other areas of down (essentially by increasing the bargaining emerge, reproduce and persist across genera- human development, power of labour, when there is a need for mass tions. It has also shown how these areas of hu- such as health mobilization),132 empirical evidence from recent man development are connected and how they and education (internal) conflicts shows that income inequality interact, transferring inequalities in one area of increases during violent conflict and during the human development to another. first five years of typical postwar reconstruction. The rise in income inequality associated with The first part took a lifecycle perspective, violent conflict is not permanent—but it takes arguing that parents’ socioeconomic status 19­22 years for inequality to fall again, and it strongly influences children’s health and early may take up to 40 years to return to prewar levels childhood development, both of which shape of income inequality if peace is sustained.133 the way children benefit from universal prima- ry and secondary education. Their education Violent conflicts can also widen inequalities attainment in turn constitutes the stepping in other areas of human development, such as stone for a successful start in the labour market. health and education. This is because violent But parents’ socioeconomic status is relevant at conflicts disproportionately affect poor peo- this stage of the lifecycle as well. Depending on ple: They increase undernourishment, infant parents’ knowledge and networks, adolescents mortality and the number of people deprived may receive a jump start for better opportuni- from access to potable water.134 Given that ties in the labour market. Assortative mating social spending often declines as a consequence then closes the feedback loop by creating fami- of rising military expenditure,135 public service lies in which both parents come from a similar provision is also weakened— a nother potential socioeconomic status. source for increasing inequalities in human development. The second approach transcended individual outcomes and looked at the macro framework Preventing violence at the early stage of for these mechanisms. It considered how in- conflict is without a doubt the best approach equalities affect institutions and balances of to avoid suffering, deaths and other costs of power, how societies function and whether violent conflict. Violence is path dependent: inequalities nurture economic growth. One Once it starts, incentives and systems work in key point was that the nature of inequality a way that sustain it. Group grievances have matters as well: Inequalities between groups to be recognized early so that patterns of ex- can determine war or peace—a pivotal decision clusion and institutional weaknesses can be for any desired expansion of capabilities at the addressed.136 When prevention is ineffective, individual and societal levels. postconflict settlements, which often involve political power sharing and could also include

Chapter 2 Inequalities in human development: Interconnected and persistent | 93 Part II PART II.

Part I of the report focuses on inequalities of capabilities, going beyond income. In parallel, part I points out that, even within segments of the population the disparities are large, particularly for those at the bottom. The evolution of indicators such as the poverty headcount ratio fail to account for what happens to those who are left behind, as well as to those who, having escaped or not even having been deprived, fall into destitution.1 Part I also highlights that a consequential aspect of inequality has to do with group—or horizontal—inequalities. Some groups get ahead, while others are in practice blocked—sometimes insidiously—from full economic and social participation. Even so, information on group inequality is often ignored, and sometimes is simply not available, despite the strong call in the Sustainable Development Goals to collect such data.

These aspects have one thing in common: They people and so may need to be complemented hide behind average patterns of inequality that with more information. harm progress in human development.2 Part II tackles this issue head-on. It goes beyond av- In fact, summary measures of inequality are erages3 to report on what is happening across sensitive to different parts of the distribution. entire distributions of income and wealth, Every summary measure implies judgements uncovering patterns in the evolution of these about how much to value the income shares of distributions.4 And it zooms in on horizontal poorer and richer people. Sometimes these are inequality’s most systematic and widespread called “weights” in a social welfare function. manifestation—inequality across gender—of- Each summary statistic assigns these weights ten obscured because biases in data collection implicitly—and, for most people, not that and analysis hurt women in a world “designed transparently. Some may even be using social for men.”5 Spotlight 3.1 at the end of chapter 3 weights that do not reflect social values. Tony illustrates the importance of looking within Atkinson, writing in the late 1960s, asserted: countries and even within households to bet- “[In examining] the problem of measuring in- ter identify those farthest behind, who may equality […] at present this problem is usually have been hidden by averages. approached through the use of such summary statistics as the Gini coefficient […]. This con- Tackling inequality starts with good meas- ventional method of approach is misleading urement and good data. Indeed, a major weak- [because the] examination of the social welfare ness of today’s public discourse on inequality functions implicit in these measures shows is its reliance on summary measures, whose that in a number of cases they have proper- choice is far from trivial (see spotlight 3.2 at ties which are unlikely to be acceptable, and the end of chapter 3). This is not an academic in general there are no grounds for believing issue—it is critical for policy. that they would accord with social values. […] I hope that these conventional measures will Conventional summary measures of inequal- be rejected.”6 In other words, the concept of ity can fail to identify what truly concerns peo- inequality one uses, and its implied ethical ple about the distribution of income, wealth judgements, will determine the conclusion one and other human development outcomes. For reaches about it.7 instance, income share ratios are insensitive to regressive transfers within the poor (as noted As it happens, the Gini coefficient is more in spotlight 3.1), a matter of importance for sensitive to transfers of income in the middle policymaking. Income inequality is often of the distribution than at the bottom or the described using the Gini coefficient. True, top—while in many countries most of the the Gini coefficient is sensitive to regressive action on income and wealth dynamics is transfers throughout the distribution and is precisely at the ends of the distribution (chap- frequently used in this Report—as it is in pol- ter 3). In particular, much of the inequality icy and much of the inequality research. But action occurs at the very top, so that measures it may not fully express what is of concern to looking at the top 10 percent—even, in some

PART II Beyond averages | 97 cases, the top 1 percent—lack the resolution to often too coarse. Thanks to innovative efforts fully capture the accumulation of income and combining information from various sources wealth. on income and wealth distribution, however, it is now feasible to estimate at a more gran- In addition, concepts and measurement in- ular level how income is distributed and how teract, each shaping how the other evolves. It is this distribution changes over time for various historically inaccurate to assume that the com- population segments. Meeting the growing de- plete axiomatic foundation of all inequality mand for comparable cross-country inequality measures was developed before these measures estimates, a number of databases with regional were used. The Human Development Index, or global coverage provide estimates for a range which Human Development Reports issue of countries and years. Although there is much regularly, is a good illustration. As Amartya agreement across different databases, there are Sen said, it was introduced as a “rough and differences across the concepts of income that ready” measure of basic capabilities, and several are used, with important implications for con- aspects of it—including changes introduced clusions, such as the extent to which fiscal re- over the years—remain controversial.8 But the distribution affects inequality (see spotlight 3.3 same can be said of national accounts estimates at the end of chapter 3).11 and the origin of macroeconomic aggregates such as gross domestic product (GDP). In To go beyond averages, part II has two the edifice of statistical manuals agreed to by chapters. Chapter 3 presents recent findings the United Nations Statistical Commission, on inequality levels and trends in global pre- national accounts may seem an unassailable tax incomes and wealth, pointing out that, as construction—but they are no more than just things stand, the wealthiest 1 percent of the that: a construction. population is on track to capture 35 percent of global wealth by 2030. The chapter breaks out Tracing the history of national accounts and these trends across regions, using recent data GDP, Diane Coyle recounts the 1940s debate and new methods to survey income inequali- in the United States on whether to include ty. It then delves into the dynamics of wealth government spending in GDP.9 The Commerce concentration. Department at the time argued that govern- ment spending should be included. But a The use of innovative methods to account for founding father of GDP measurement, Simon the evolution of income and wealth inequality Kuznets, argued for leaving it out (partly be- across the distribution has captured previously cause he viewed some government spending as hidden patterns of accumulation at the very top not necessarily enhancing welfare). Ultimately, in many countries. The drivers of this accumu- Coyle argues, the decision to include it had lation need to be understood in depth and are profound implications for the government’s likely to vary by country. (For instance, recent perceived role in the economy as another agent analysis has shown that the typical top earners alongside private actors (the same approach in the United States derive their high incomes advocated by John Maynard Keynes). Hugh from founding or managing their businesses Rockoff goes further, showing how economic rather than from financial capital).12 The inno- statistics such as price indices and unemploy- vative methods in this chapter are still evolving, ment rates originated “in bitter debates over requiring assumptions that are contested in the economic policy, ultimately debates over the literature.13 distribution of income.”10 Chapter 3 is transparent about assumptions Clearly, measurement influences policy. Yet and decisions in dealing with data challenges the issue is more complex that just measure- to encourage the type of scrutiny that, over ment. It is one thing to agree to look beyond time, will improve data and information on summary measures of income inequality, and inequality. It bears recalling that even the another to have the data to do this. To be best-established economic statistics have sure, summary measures are constructed from some uncertainty. The chapter argues that information on the very distribution that today’s innovations in measuring economic they collapse into a single summary statistic, inequality can open the way for the more sys- although the data on that distribution are tematic measuring and reporting of income

98 | HUMAN DEVELOPMENT REPORT 2019 and wealth distribution. Such reporting would catch up, targets move, and the enhanced ca- complement the aggregate measures that tend pabilities that bring strategic empowerment all to dominate literature and policy at present, too often elude them. The chapter documents whether GDP growth rates or changes in the that gender inequalities are multidimensional, Gini coefficient. pervading life in varying degrees across devel- oping and developed countries alike. That is Chapter 4 considers gender inequality. because they are cultural and rooted in social While there are signs of progress, the chapter norms—biases and gender discrimination are points out that it may be slowing. In fact, there endemic to our social institutions.14 The chap- are troubling signs of reimposing inequality— ter discusses how the challenge of reducing linked to backlash in social norms observed in gender inequalities ranges from how to create half the countries with data. It is true that most enabling conditions for cultural change to how girls around the world are catching up in the to avert societal reactions against progress to- basics, such as primary education. These prac- wards gender equity. tical achievements are evident. But as women

PART II Beyond averages | 99 Chapter 3

Measuring inequality in income and wealth 3.

Measuring inequality in income and wealth

A contribution by the World Inequality Lab

Measuring income inequality is a key step to properly address it. Public debates grounded in facts are critical for societies to determine to what extent they accept inequality, what policies they should implement to tackle it and what taxation they will use—a particularly difficult decision.

Transparency in income and wealth dynamics Tackling inequality starts Publishing timely, is also essential to evaluate public policies with good measurement standardized and track government progress towards more and universally inclusive economies. Sound data on income Publishing timely, standardized and universally recognized statistics and wealth are also required to fight (legal) tax recognized statistics is key to properly address in- is key to properly avoidance and (illegal) evasion, made possible equality. Indeed, the production of standardized address inequality in part by the built-in opacity of the global GDP statistics from the 1950s onwards,4 thanks financial system.1 Greater transparency would to the United Nations Systems of National thus support the highest return to tax policy, Accounts, has had huge impacts on framing pol- part of the policy package to reduce inequality icy debates and policymaking over the past seven and to finance investments for the Sustainable decades. A new generation of growth statistics Development Goals.2 distributed across income groups (distributional national accounts5) is also likely to shape these The secrecy surrounding ownership of assets policy debates. Moving towards developing around the globe—particularly financial as- and publicizing such indicators requires efforts sets—currently makes it impossible to properly from all actors: policymakers, academia and civil track capital accumulation, just as it makes it society. The synergies among different actors impossible to ensure that top earners and wealth committed to transparency become apparent holders pay their fair share of taxes. Some pro- when, for example, information on evaded gress on financial transparency has been made taxes is released by journalists and subsequently since the 2008 financial crisis, but it has been analysed by researchers, including some at the too slow and limited in relation to the challenge. World Inequality Lab.6 The share of global wealth hidden in tax havens is an estimated 8 percent of global GDP.3 This chapter discusses challenges and recent advances in methodology and data collection to The current lack of transparency on income fill a crucial gap in data on human development. and wealth dynamics is a political choice. While It first introduces a new inequality data trans- most governments have (or can find, if they parency index. Then, based on data from the wish) detailed information on the top incomes World Inequality Database and analysis from and wealth, they do not disclose it. This is a dig- the World Inequality Report, it presents recent ital age paradox: Multinationals have detailed findings on inequality in global incomes. It information on individuals’ lives and can trade also surveys income inequality in three country it in the global marketplace. Yet people struggle groups, assessing the evolution of inequality by to get basic information about how growth in comparing the rate of income growth of the income and wealth is shared across the popula- bottom 40 percent with that of the entire pop- tion. Public statistics still rarely move beyond ulation—a target for Sustainable Development reporting averages. This weakness applies to Goal 10. The first country group is African economic inequality and to other forms of countries—where new inequality estimates inequality—particularly inequality related to have recently become available. The second pollution—which are not scrutinized by most is for Brazil, China, India and the Russian statistical institutions today (see chapter 5).

Chapter 3 Measuring inequality in income and wealth | 103 On a new inequality Federation. And the third is European countries On a new inequality data transparency index data transparency and the United States, noting the relative impact that ranges from 0 to 20, no country scores of different policies on income distribution. above 15, and dozens have a score of 0 (see fig- index that ranges from 0 Finally, the chapter turns to the measurement of ure 3.1). Data are particularly scarce in Africa to 20, no country scores wealth inequality around the world. and Central Asia. This simple index is prelim- inary and will be improved as more informa- above 15, and dozens Measuring the transparency gap tion is released on income and wealth taxes have a score of 0 and availability of survey data. But it already Data for tracking income and wealth inequal- provides an overview of the efforts required to ity remain scarce across the globe (figure 3.1). supply transparent data on inequality. To measure inequality in a country, national statistical authorities ideally would produce Though the availability of official data is low, rich annual household surveys of individuals’ the triangulation of different sources has shed living conditions. And the tax administration new light on income and wealth inequality. would publish income and wealth administra- Investigative journalism has played a critical role, tive tax each year. To track income and wealth providing new information that has influenced inequality, survey data and tax data would be public discussions and decisionmaking (box 3.1). linked so that it would be possible to know the fiscal income reported in the tax data by Where to look for global an individual who participated in the living income inequality data conditions survey. But linked survey and tax data are an exception across the globe, done by Several global income inequality databases only a few countries: for example, Sweden and have been constructed over the past decades.9 other Nordic countries. And even there, the They include the World Bank’s PovcalNet, ability to measure inequality has deteriorated which provides inequality data from house- in recent decades, partly because of the large hold surveys; the World Inequality Database, wealth hidden in offshore financial assets with- which produces distributional national out a proper international registration system accounts based on tax, survey and national to follow them.7 accounts; the LIS Cross-National Data Center in Luxembourg (LIS),10 which harmonizes In many countries tax data are not available to to a high level of detail income and wealth the public. The production of administrative tax concepts in rich countries using household data has historically been closely related to the surveys; the Organisation for Economic existence of an income or wealth tax in a country. Co-operation and Development’s Income It was the introduction of the income tax in the Distribution Database,11 which contains United States in 1913, and in India in 1922, that distributional survey data for advanced econ- led public administrations to publish income omies; the University of Texas Inequality tax statistics. Such information is critical for tax Project Database,12 which uses industrial and administrations to properly administer taxes sectoral data to measure inequality; and the and for legislators and taxpayers to be informed Commitment to Equity Data Center,13 which about tax policy. But governments are sometimes provides information on fiscal incidence—the unwilling to publicly release the data.8 impact of taxes and transfers on different in- come groups. The United Nations University While some countries have released new World Institute for Development Economics tax data over the past decade, others have Research’s World Income Inequality Database actually stopped producing them. And when provides a range of statistics on income ine- governments repeal income or wealth taxes, the quality for several countries.14 There are also statistical tools to measure inequality also dis- detailed regional databases such as the Socio- appear. The deterioration of administrative tax Economic Database for Latin America and the data thus raises serious concerns, since proper Caribbean,15 the harmonized regional statistics information on wealth and income is key to maintained by the Economic Commission track inequality and inform public debates. But for Latin America and the Caribbean16 and the situation is worsening in several countries the European Union Statistics on Income and rather than improving.

104 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 3.1 Dozens of countries have almost no transparency in inequality data

Inequality data quality index

LO W H NOR CHE IRL HKG DEU ISL AUS SSGPWE

ISRLUXAUJTPBNEULSGABNRZCLFAIDNNNNLKD 20 10 9 10 VERY HIGH HUMAN DEVELOPMENT

UMAN DEVELOPMENT NER20 CAF18 TCD CROWNMGATUZAGAMMARDTBGELNSOCSIVETNGSODNHTI SSD16 KOSRVNESFPRCAZMELITTEASCTYGPPRLOCTALSURAEU 14 SVK BDI12 LVA MLI10 PRT BFEARI8 QAT CHL PNG SLE6 HUN SYR MCOOGZDNYEBMLBGRGINMEBTMHWDI JAIFG BRN

SLB BHR PAK CMR 4 HRV OMN ZWE 2 ARG RUS AGO KAZ BLR KEN MNE BGR NPL 0 LKAGEOALCBRPIAMNUIRSTNTSORMBBYTHSUUSKRRRWYOTU

KHM MMR

GNQ ZMB CUBBIHMETXHBARCAOARLMMPKEDDRZAGHA ECCAHUZNELAO UKR DOMSCWBOGZBGHDTNNNTDLASMINCDGPVTMNICTJSKLVGUY TUN MNG LBN BWA VEN JAM PRY

TKLBIMYDZNABFGOLAEBGVYN PMSIERMQARKGZ MUDZAB PHL MDV BLZ JOR SUR FJI

About the Inequality data quality index: The data quality and availability index measures the current availability of inequality data around the globe. The index ranges from 0 (a country with no survey or tax data to track inequality available at all) to 20 (an ideal case where there are income and wealth surveys and income and wealth tax data, and the sets of information are linked with one another). Currently, no country has a score above 15, and dozens of countries have a score of 0. Data are particularly scarce in Africa and Central Asia.

Note: The index presents the level of availability and quality of data on income and wealth inequality. Source: World Inequality Lab (http://wid.world/transparency); accessed 17 July 2019.

Chapter 3 Measuring inequality in income and wealth | 105 BOX 3.1 Investigative journalism uncovering inequality

Investigative journalism can shed light and generate Journalistic exposure of corruption can also protect data on aspects of inequality for which no measurement public finances.5 standards exist or that have remained opaque because of asymmetries in the distribution of power (see chap- In a globalized world, internationally coordinated ter 2). New and widespread protocols to assess who efforts to find and disclose information can catch up is being left behind or extreme wealth concentration with actors that operate strategically in different coun- might take years or even decades to generate, with con- tries, taking advantage of transparency blind spots. straints ranging from corruption to pressure by interest The Global Investigative Journalism Network and the groups. International Consortium of Investigative Journalism are two prominent examples of this approach.6 These Investigative journalism has played a remarkable networks have the potential to develop and defend role in informing the public of important dimensions of standards of responsible reporting and diversify the inequality. Today, we know more about the globalization risks of pressure from interest groups. of hidden wealth because of disclosures such as those in the Panama Papers and the Paradise Papers.1 On the Quality journalism tends to face financial, polit- other side of the distribution, decentralized reporting ical and safety challenges. When journalism and me- based on investigative journalistic research routinely dia produce information and knowledge that has the uncovers abuse towards disadvantaged groups: When characteristics of a public good, indirect and direct all other mechanisms that give voice to excluded groups subsidies remain fundamental to avoid underprovision.7 fail, journalism is often their last hope.2 Journalists can be subject to pressures, intimidation and attacks, which appear to be on the rise in many Amartya Sen has argued that a free press and countries,8 highlighting the importance of protecting an an active political opposition constitute an effective independent, plural and diverse media. early warning systems against famines because infor- mation and political pressure push for action.3 By the Investing in quality investigative journalism has same token, the media has played an important role high social returns, deterring and correcting corruption, in thwarting behaviours that impede human develop- protecting those left behind and informing public poli- ment—human trafficking and, in the worst instances, cies. One area to explore is an enhanced role for inter- slavery; child labour; child marriages; genital muti- national cooperation: Currently only around 0.3 percent lation; and malnutrition, especially among children, of official development assistance is spent in media de- which can cause stunting that has lifelong effects.4 velopment, a small fraction of which is clearly linked to investigative journalism.9

1. In additional to the increase in public awareness and accountability, these data have been used as part of academic research. See, for instance, analysis of the relation of tax evasion and inequality by Alstadsæter, Johannesen and Zucman (2019). 2. See examples and discussion in Brunwasser (2019). 3. Sen 1982, 1999. 4. Schiffrin 2019. 5. Brunwasser 2019; Schiffrin 2019. 6. Brunwasser 2019; Schiffrin 2019. 7. Schiffrin 2019. 8. In resolution 33/2, the United Nations Human Rights Council expressed “deep concern” at the increased number of journalists and media workers who had been killed, tortured, arrested or detained in recent years as a direct result of their profession (UNHRC 2018). 9. Over 2010­2015, $32.5 million appears to be clearly linked to investigative journalism. See annex 1 of Myers and Juma (2018). This is a small amount compared with the net benefits associate with individual investigative journalism projects. See examples in Hamilton (2016) and Sullivan (2016). Source: Human Development Report Office based on Brunwasser (2019) and Schiffrin (2019).

Living Conditions database (see spotlight 1.3 and its interactions with other dimensions of at the end of the chapter for more sources).17 welfare in an international perspective. Regional databases, such as the Socio-Economic Database These databases have helped researchers, for Latin America and the Caribbean and the policymakers, journalists and the general public European Union Statistics on Income and Living focus on the evolution of inequality over the Conditions database, enable detailed regional past decades. There is no one perfect database on analyses of inequality, while the Commitment inequality, and there will never be: The different to Equity Data Center can be used to analyse the datasets support complementary insights on impact of tax and transfer policies. inequality, and whether to use one or another depends largely on the specific issues to be stud- Most of these databases rely almost exclusively ied.18 Some, such as PovcalNet have been used to on one type of information source—household compute global poverty measures. Others, such surveys with face-to-face or virtual interviews as the LIS database, have been used by genera- that ask individuals about their consumption, tions of researchers to study economic inequality income, wealth and other socioeconomic aspects

106 | HUMAN DEVELOPMENT REPORT 2019 of their lives. Surveys, like any other data source, households report an income share of around have pros and cons in the measurement of ine- 10 percent, suggesting that household survey quality (table 3.1). One way of overcoming the data starkly underestimate incomes at the top of limitations of each data source is to combine data the distribution. The extent to which they do so from different types of sources, particularly com- varies by country but is likely to be substantial. bining administrative tax data with survey data. In addition, surveys may also miss important evolutions. In Brazil, household surveys indicate For example, consider the level and evolution the income share of the top 10 percent has fallen of inequality in Brazil and India. In Brazil house- over the past two decades.19 But revised estimates hold surveys show that the richest 10 percent based on additional sources of information from received just over 40 percent of total income in national accounts and tax data suggest that the 2015, but when all forms of income are consid- income share has been fairly stable. Household ered—not just income reported in surveys—the surveys captured fairly well the increase in wage revised estimates suggest that the top 10 percent income across most of the distribution, which actually received more than 55 percent of total has indeed taken place in Brazil since the 2000s, income. In India estimates based on adminis- but failed to fully capture the dynamics of top trative tax data show that the top 1 percent may incomes—particularly capital incomes. have an income share close to 20 percent. But

TABLE 3.1 Main data sources for inequality measurement

Data source Pros Cons

Household survey data · Survey data gather information about income or assets as well as · Limited sample size is a problem. Given the small number of social and demographic dimensions, key for human development. extremely rich individuals and of some vulnerable groups, the likelihood that they will be included in surveys is typically very

• Households surveys support a better understanding of the small. These are called sampling errors.

determinants of income and wealth inequality and allow income and wealth inequality to be analysed in combination with · Self-reported information about income and wealth is erratic. other dimensions—such as racial, spatial, education or gender Generally, it largely underestimates the income share of the inequality. top. Oversampling cannot correct this bias. These are called nonsampling errors.

• Concepts and scope may vary widely across countries and over

time, rendering international and historical comparisons difficult. Surveys may be administered with uneven frequency.

• Income and wealth totals generally do not match national

accounts totals, so growth rates are typically lower in surveys than in macroeconomic growth statistics.

Administrative (tax) data · In countries with sound enforcement of taxes, tax data capture the · Tax data have limited coverage of the lower tail of distribution. income and wealth of those at the top of the wealth distribution. Particularly in developing countries, they typically cover only a small share of the population.

• Tax data also cover longer periods than surveys. Administrative

data are usually available annually starting at the beginning of · Tax avoidance and evasion affect tax data. Tax data tend to the 20th century for income taxes and in some countries as far underestimate income and wealth at the top. In most cases back as the early 19th century for inheritance taxes. inequality estimates based on these data should be viewed as lower-bound estimates.

• Tax data are subject to changes in fiscal concepts over time and across

countries, making historical and international comparisons difficult.

National accounts data (gross · National accounts data follow internationally standardized · National accounts do not provide information on the extent to national product, national definitions for measuring the economic activity of countries, which different social groups benefit from growth of national income, national wealth) so they allow for a more consistent comparison over time and income and gross domestic product. across countries than fiscal data. National account definitions, in particular, do not depend on local variations in tax legislation or · National accounts are heterogeneous across countries, determined other parts of the legal system. by quality of national data and country-specific assumptions.

Source: Based on Alvaredo and others (2018).

Chapter 3 Measuring inequality in income and wealth | 107 The World Inequality World Inequality Database and accounts data are used as the overarching Database project distributional national accounts framework, since they provide the most uni- versally recognized concepts of income and seeks to combine data Studying inequality in a context of extreme wealth to date. sources transparently data opacity is difficult, and results are neces- sarily imperfect and preliminary. Yet, income The World Inequality Database project em- and consistently in and wealth dynamics must be tracked as sys- phasizes the distribution of national income order to estimate tematically as possible. The World Inequality and national wealth equally. There are two Database project seeks to combine data sources main reasons for this. First, it is impossible the distributions of transparently and consistently in order to esti- to properly track income inequality, particu- national income and mate the distributions of national income and larly at the top of the distribution, without a national wealth. In doing so, the project’s main sound measure of wealth inequality dynamics. national wealth objective is to reconcile the macroeconomic Indeed, where there has been a recent rise study of income and wealth (which deals with in income inequality, it has often been due economic growth, public debt or interna- largely to the surge in capital income (rents, tional capital flows) with the microeconomic dividends, retained earnings and so on) among study of inequality (which considers how the the wealthy.26 Second, rates of return on wealth income and wealth growth rates actually expe- have been much higher than macroeconomic rienced by individuals in a single country differ income growth over the past four decades, depending on their position in the income implying that wealth is taking an increasingly distribution). important place in 21st century economies.27 How the fast growth of wealth is distributed The World Inequality Database project across the population becomes a pressing began with renewed interest in using tax data question. Unfortunately, available official data to study the long-run dynamics of inequality, are even scarcer for wealth than for income, so following the pioneering work on income and distributional national accounts estimates for wealth inequality series by Simon Kuznets and wealth inequality cover only a few countries at by Tony Atkinson and A.J. Harrison.20 Top in- this stage. come shares, based on fiscal data, were initially produced for France21 and the United States22 For transparency, the distributional national and rapidly expanded to dozens of countries accounts project releases distributional nation- thanks to the contribution of more than 100 al accounts estimates and the methods used researchers.23 These series had a large impact on to compute them. Technical details and the the global inequality debate because they made computer codes used to produce the estimates it possible to compare the income shares of top (including those presented in this chapter) groups (say, the top 1 percent) over long peri- are published online on the World Inequality ods of time, revealing new facts and refocusing Database website.28 This level of transparency the discussion on long-run historical evolutions should become the norm for existing economic of income and wealth inequality. statistics databases.

More recently, the World Inequality Inequality series published online should Database project has sought to go beyond the also be as comprehensive as possible, given the top income shares based on tax data to pro- limitations of summary measures of inequality duce distributional national accounts, relying (as discussed in the introduction to part II of on a consistent and systematic combination of the Report), which can mask relevant inequal- fiscal, household survey, wealth and national ity dynamics behind a veil of stability. Beyond accounts data sources.24 The objective of the offering summary measures and a limited set of distributional national accounts is to make decile shares, the World Inequality Database the most of all data sources (see table 3.1). Tax project publishes average income and wealth data are used to track the top of the distribu- levels for each 1 percent of the population in tion properly—and when available, informa- a given country or region (that is, income and tion on tax evasion is also used.25 Survey data wealth percentiles). Given the importance are used to obtain information not available of the very top groups in income and wealth from administrative records. And national growth, the project decomposes the top 1 per- cent itself into smaller subgroups (up to the

108 | HUMAN DEVELOPMENT REPORT 2019 top 0.001 percent) and estimates income and The elephant curve of global wealth levels for each. inequality and growth

Currently, the United Nations System of The release of new tax data and the recent National Accounts includes standards and methodological developments by research- guidance only for aggregate indicators.29 The ers collaborating with the World Inequality next revision, due sometime in 2022­2024, Database and at the World Inequality Lab make might consider how to cover distribution of it possible to produce new inequality estimates income and wealth growth across the popu- (see boxes 3.2 and 3.3 for definitions of income lation, in line with the recommendations of and consumption concepts used throughout the 2008 Report of the Commission on the the Report).31 A starting place in tracking the Measurement of Economic Performance and evolution of income inequality over time and Social Progress.30 Such an evolution would across countries is to estimate the share of total represent significant progress for global public income received by the richest 10 percent of the statistics and global public debates on growth population. But such an indicator should be and inequality. The distributional national complemented by others—ideally, the income accounts framework considered in this chapter level or growth of each percentile, or 1 percent provides a concrete model of how this shift of the population, as below. beyond averages could work.

BOX 3.2 What income concepts are we measuring?

This chapter focuses on the distribution of national artificially high pretax inequality (because retired indi- income, which is the sum of all income received by viduals would have no pretax income and would appear individuals in an economy. This corresponds to gross as “virtual poor” before taxes), while a country with domestic product, to which are added net income from private pensions would have positive pretax income for abroad (when a Brazilian citizen owns a company in the elderly (because they would benefit from pretax in- India, the income from the capital of the company is come from their pension plans). Differences in inequal- counted in Brazil) and from which are subtracted the ity measures between the countries would not reflect amounts required to replace any productive apparatus differences in income concentration or the effectiveness (roads, machines, computers) that has become obsolete. of pension systems but simply different choices made for organizing the pension system. There are two broad ways to measure income received by individuals in a country: before taxes and In the end, pretax income is similar to the taxable government transfers (pretax income) and after taxes income of many countries, but its definition is usually and government transfers (post-tax income). There are broader and more comparable across countries. Several different ways to define pre- and post-tax incomes, and variants of pretax income should be looked at, and the definitions can affect the results substantially. In the distributional national accounts guidelines discuss them World Inequality Lab’s distributional national accounts in more detail. Unless stated otherwise, the income framework, pretax national income is defined as the concept in this chapter is pretax income.1 sum of all personal income flows, before taking into ac- count the tax and transfer system but after taking into Post-tax national income equals pretax income after account pension and unemployment insurance systems. subtracting all taxes and adding all forms of government This concept adjusts traditional computations of “mar- transfers. In line with the distributional national accounts ket income,” as explained in spotlight 3.3. Contributions methodology, all forms of government spending are allo- to pension and unemployment insurance schemes are cated to individuals, so that post-tax income sums to na- considered deferred income and therefore deducted, but tional income. Not doing so would make countries with the corresponding benefits are included. a stronger provision of public goods appear mechanically poorer. By definition, at the aggregate or macroeconomic The adjustment is crucial for good comparability of level—when summing all income of all individuals in a pretax inequality across countries. Otherwise a country country—post-tax national income is exactly equal to with a public pension system would appear to have pretax national income and to national income.

1. See Alvaredo and others (2016) for a technical description of income concepts and methods used for this chapter.

Chapter 3 Measuring inequality in income and wealth | 109 BOX 3.3 What about consumption?

Income inequality For the distributional national accounts project of the make it possible to systematically relate income, wealth based on the top World Inequality Lab and its network of partners, the and eventually consumption (income minus savings). In 10 percent’s income objective is a fully integrated representation of the our view, however, it would be a mistake to overempha- share has risen since economy. It would link the microeconomic study of size the consumption perspective, as the literature on 1980 in most regions income and wealth inequality (typically focusing on poverty has sometimes done. Consumption obviously is but at different rates household wages, transfers and poverty or inequality) a very important indicator of wealth, particularly at the with macroeconomic issues such as capital accumula- bottom of the distribution. The problem is that house- tion, the aggregate structure of property and privatiza- hold surveys routinely used to measure consumption tion or nationalization policies. Too often, “micro” and tend to underestimate income, consumption and wealth “macro” issues have been treated separately. at the top.

To be clear, however, a lot of progress is needed In addition, consumption is not always well defined before it will be possible to publish a fully integrated ap- for top income groups, which generally save a very large proach to these issues, analysing the joint evolution of share of their income, choosing to consume more in lat- inequality of income and wealth in all countries. Indeed, er years, but more generally to consume the prestige or that approach requires careful measurement not only of economic or political power conferred by wealth owner- pretax and post-tax income inequality but also of the ship. To develop a consistent and global perspective on distribution of savings rates across different income economic inequality—one that views economic actors groups. not only as consumers and workers but also as owners and investors—requires putting equal emphasis on in- The production of such series—pretax inequality, come and wealth. post-tax inequality and savings rate inequality—will

Source: Extracted from Alvaredo and others (2018).

The European Union stands out as the most levels in low- and middle-income countries equal region based on the top 10 percent’s share also deserve particular attention.34 of pretax income, with 34 percent. The Middle East is the most unequal, with the top 10 percent The diversity of patterns across countries holding 61 percent of pretax income.32 In be- since 1980 shows that the extreme rise in tween are a variety of inequality levels that do not inequality in some parts of the world was not appear to be correlated with average income. The inevitable but resulted from policy choices. top 10 percent received an estimated 47 percent Openness to trade and the digitalization of of income in the United States, 41 percent in the economy are often put forward to explain China and 55 percent in India.33 the rise in inequality in a country, but such arguments fail to fully account for the diversity Income inequality based on the top 10 per- of trajectories just presented. The radical diver- cent’s income share has risen since 1980 gence of the United States and Europe—de- in most regions but at different rates (fig- spite similar exposures to technological change ure 3.2). The rise was extreme in the Russian and trade openness—shows that other factors Federation, which was one of the most equal were at play—specifically, factors related to na- countries in 1990 (at least by this measure) tional policies. Differences between the United and became one of the most unequal in just States and Europe were due less to direct five years. The rise was also pronounced in taxes and transfers and more to other policy India and the United States, though not as mechanisms, particularly health, education, sharp as in the Russian Federation. In China, unemployment and pensions systems, as well after a sharp rise, inequality stabilized in the as labour market institutions.35 Fiscal redistri- mid-2000s. The rise in inequality in Europe bution and monetary transfers to the worse-off was more moderate than in other regions. indeed helped low-income groups in Europe Inequality in Sub-Saharan Africa, Brazil but did not play the main role in restraining the and the Middle East stayed extremely high, increase in income inequality.36 with the 10 percent’s income share around 55­60 percent. These extreme inequality What happened to inequality among indi- viduals globally—treating the world as just one

110 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 3.2

Income inequality based on the top 10 percent’s income share has risen since 1980 in most regions but at different rates

Share of national 2016 income (percent) value Middle East 61 Sub-Saharan 57 India 55 Brazil 55

North America 47

Russian Federation 46

China 41

Europe 34

20 The global top 1 percent, the 1980 1985 1990 1995 2000 2005 2010 2015 economic elite of rich and poor countries, Source: Based on Alvaredo and others (2018), with data from the World Inequality Database (http://WID.world). made huge gains over 1980­2016 single country? Branko Milanovic pioneered groups in Europe and North America. In the such analysis, arguing for its relevance in a United States the situation was even worse: The more integrated and globalized world. bottom 50 percent was almost entirely left out of economic growth. A graph of income growth from 1980 to 2016 for the world population, ranked from At the very top of the global income distribu- the poorest to the richest,37 presents the sil- tion, growth rates were extremely high—more houette of an elephant with a raised trunk than 200 percent. The global top 1 percent, (figure 3.3).38 At the bottom of the global in- the economic elite of rich and poor countries, come distribution (the left side), the low- and made huge gains over 1980­2016. In China middle-income emerging countries had high and India, for instance, growth rates at the growth: above 100 percent—for a doubling top of the income ladder reached triple digits. of income per adult since 1980. In some coun- These results, based on new and more precise tries, such as China, the bottom 50 percent of data (combining tax, survey and national ac- the population saw growth of around 400 per- counts data), magnify the results of previous cent—incomes quintupled.39 studies using fewer sources of data.42

The dynamics illustrate how hundreds of The top 1 percent alone received 27 percent millions of individuals were lifted out of income of income growth over the period, compared poverty and saw improvements in their living with the 12 percent received by the bottom standards. Note that the figure represents relative 50 percent. A huge share of global growth gains, which for the bottom of the distribution thus benefited the top of the global income are from very low levels—a figure representing distribution. absolute gains looks essentially flat except for a spike for people at the very top.40 In India the Was such a concentration of global growth in absolute poverty rate was more than halved over the hands of a fraction of the population neces- the period, and at the global level the share of sary to trigger growth among bottom income people living in absolute poverty was reduced by groups? Country and regional case studies a factor of more than three.41 In the upper half of provide very little empirical support to the the distribution, however, incomes grew much trickle-down hypothesis over recent decades.43 less rapidly, with less than 50 percent since 1980. Higher income growth at the top of the distri- That segment of the global income distribution bution are not correlated with higher growth corresponds to the bottom and middle-income at the bottom. The comparison between the United States and Europe is an illustration. As

Chapter 3 Measuring inequality in income and wealth | 111 FIGURE 3.3 Top 1 percent captured 27 percent The elephant curve of global inequality and growth Real income growth per adult (percent)

Bottom 50 percent captured 12 percent

0

10 20 30 40 50 60 70 80 90 99 99.9 99.99 99.999

Note: On the horizontal axis the world population is divided into 100 groups of equal population and sorted in ascending order from left to right by each group’s income. The top 1 percent group is divided into 10 groups, the richest of which is also divided into 10 groups of equal population and the richest of that group is again divided into 10 groups of equal population. The vertical axis shows the total income growth of an average individual in each group between 1980 and 2016. For percentile group p99p99.1 (the poorest 10 percent among the world’s richest 1 percent), growth was 74 percent between 1980 and 2016. The top 1 percent captured 27 percent of total growth over this period. Income estimates account for differences in the cost of living between countries. Values are net of inflation. The composition of each group evolved between 1980 and 2016. Source: Based on Alvaredo and others (2018), with data from the World Inequality Database (http://WID.world).

noted, growth at the top was much higher in the mid-2000s to 52 percent in 2016 (fig- the United States than in Europe, but the bot- ure 3.4). Consider two counterfactual scenar- tom 50 percent benefited little from growth, ios. The first is a world with no differences in while Europe was more successful at triggering average income across countries (all countries growth for the majority of its people, despite have converged to the same average income) lower growth at the top. but with within-country inequality matching the levels observed in reality since 1980. The Between-country convergence second is a world with no within-country versus within-country divergence inequality (all individuals in a country have the same income) but with countries’ average To understand the dynamics of global income incomes differing exactly as observed in reality inequality over the past four decades, it is also since 1980.45 useful to decompose global inequality into two components.44 One is the evolution of In the first counterfactual the income share global inequality between countries, driven by of the top 10 percent increases significantly the rise in productivity in emerging countries over the period because of the rise of income and the technological catch-up with countries inequality in most countries. In the second sce- at the frontier. The other is inequality within nario the income share of the top 10 percent in- countries. Both forces have been at play over creases slightly, falls then recovers in the recent the past four decades, but the latter appears to period to its 1980 level. Since the mid-2000s have dominated. the reduction in between-country inequality has dominated but not enough to bring global The share of global income held by the top inequality back to its early 1980s level. 10 percent rose from less than 50 percent in 1980 to 55 percent in 2000 and slipped from Another way to look at the relative im- portance of within- and between-country

112 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 3.4

In 2010 the top 10 percent of income earners received 53 percent of global income, but if there had been perfect equality in average income between countries, the top 10 percent would have received 48 percent of global income

Share of global

55 Global top 10 percent share

45 … perfect equality between countries Global top 10 percent 40 share assuming…

35 The decline in between-country 30 … perfect equality inequality has not been enough to counter within countries the rise of within- country inequality 25 since 1980 or 1990

1980 1985 1990 1995 2000 2005 2010 2015

Source: Based on Alvaredo and others (2018), with data from the World Inequality Database (http://WID.world).

inequalities is to focus on the Theil index, measure of inequality thus masks the catch-up which provides a measure of inequality that can of low-income groups with the middle of the be decomposed into a between-country and global income (reduction in between-country a within-country component. The two com- inequality) as well as the relative decrease of ponents sum to an overall measure of global the middle compared with the top (rising with- inequality. The decomposition confirms and in-country inequality in rich countries). From amplifies the results above: The decline in be- 1980 to 2016 the income gap between the top tween-country inequality has not been enough 10 percent and the middle 40 percent increased to counter the rise of within-country inequality by 20 percentage points (figure 3.5). But the since 1980 or 1990. Global inequality accord- gap between the middle 40 percent and the ing to the Theil Index rose from 0.92 in 1980 bottom 50 percent fell by more than 20 per- to 1.07 in 2016, peaking in 2007 before a centage points. In short: The Gini coefficient slight decline and then a plateau since the early masks a lot of movement. 2010s.46 The changing geography of Going beyond summary global income inequality measures of inequality Understanding the dynamics of global inequal- The dynamics of global income inequality over ity also entails looking at the changing geo- the past decades are the result of the dynam- graphic distribution (box 3.4). The geographic ics of between-country and within-country breakdown of each percentile of the global inequalities. These are not well captured by distribution of income has evolved. In 1990 an oft-used measure of inequality: the Gini Asians were mostly absent from top global in- coefficient. Since 1980 the Gini coefficient for come groups, and massively represented at the global income has hovered around 0.65, with bottom of the global distribution (figure 3.6), a peak of 0.68 in 2005­2006. This summary while Americans and Canadians were the

Chapter 3 Measuring inequality in income and wealth | 113 FIGURE 3.5

The ratio of the average income of the top 10 percent to that of the middle 40 percent increased by 20 percentage points between 1980 and 2016, but the ratio of the average income of the middle 40 percent to that of the bottom 50 percent decreased by 27 percentage points

Index, 100 = 1980

Top 10 percent to

middle 40 percent

120 average income

Gini coefficient

Middle 40 percent to

80 bottom 50 percent

average income

1980 1985 1990 1995 2000 2005 2010 2015

Source: Based on Alvaredo and others (2018), with data from the World Inequality Database (http://WID.world).

BOX 3.4 Where do you stand in the global distribution of income?

Who is part of the global top 1 percent? And how much for instance, an adult individual is part of the top 8 per- does one need to make to belong to the global middle cent of earners in Côte d’Ivoire (see table). The same 40 percent? It is not always clear how much income one income would place an individual in the top 33 percent needs to belong to different income groups discussed in in China and in the bottom 22 percent in the United academic or public debates on inequality. States. At the world level, that individual belongs to the top 33 percent. The global top 1 percent entry threshold The World Inequality Database’s online simulator is $11,990 per adult per month. allows anyone to position their income relative to that of others throughout the world. With $1,000 a month,

On different rungs in different countries

Monthly income China United States World per adult (PPP $) Côte d’Ivoire Bottom 5 percent Bottom 8 percent Bottom 22 percent Top 33 percent $100 Bottom 20 percent Bottom 7 percent Bottom 42 percent Top 18 percent Top 24 percent Top 5 percent $1,000 Top 8 percent Top 33 percent Top 5 percent Top 1 percent $2,0000 Top 3 percent Top 12 percent

$5,000 Top 1 percent Top 4 percent

$12,000 Top 1 percent Top 1 percent

Source: World Inequality Database website (http://WID.world/simulator).

114 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 3.6 The geographic breakdown of each percentile of the global distribution of income evolved from 1990 to 2016

In 1990, 33 percent of the population of the world’s top 0.001 percent income group were residents of the United States and Canada.

Population share within each global income group (percent)

0

1 10 20 30 40 50 60 70 80 90 99 99.9 99.99 99.999

India Other Asia China Sub-Saharan Latin America Middle East Russian Europe United States

Africa Federation and Canada

In 2016, 5 percent of the population of the world’s top 0.001 percent income group were residents of the Russian Federation.

0

1 10 20 30 40 50 60 70 80 90 99 99.9 99.99 99.999

Source: Based on Alvaredo and others (2018), with data from the World Inequality Database (http://WID.world).

largest contributors to global top income earn- among the very top global groups, as they each ers and almost absent at the very bottom of the made up about 20 percent of the population of distribution. Europe was well represented in the top 0.001 percent earners. the upper half of the global distribution but less so in the very top groups. Middle Eastern and The situation had changed considerably by Latin American elites were disproportionately 2016. Chinese earners are now present through- out the income distribution. Indians remain

Chapter 3 Measuring inequality in income and wealth | 115 Available global and concentrated at the bottom. Russian earners are New estimates combining survey, fiscal and African evidence also stretched throughout, from the poorest to national accounts data suggest that inequality the richest income groups, in contrast to 1990. remains very high in most African countries. shows that the average Africans, present throughout the bottom half The income received by the top 10 percent income of the top of the distribution, are now even more concen- ranges from 37 percent in Algeria to 65 percent trated in the bottom quarter, due to Africa’s slow in South Africa, while that received by the bot- 1 percent of earners growth relative to Asian countries. At the top tom 40 percent is at most 14 percent in Algeria is typically 1.5­2 times of the distribution, both North America’s and and about 4 percent in South Africa. Europe’s shares fell (leaving room for their Asian higher than what is counterparts), Europe’s share fell much more. The Regional differences across Africa are sig- reported in surveys reason? Most large European countries followed nificant.53 Southern Africa is clearly the most a more equitable growth trajectory over the past unequal. The share of national income received decades than the United States and Asian giants. by the top 10 percent is highest in South Africa (65 percent in 2014) and Namibia (64 percent How unequal is Africa? in 2015), while the bottom 40 percent received 4 percent of national income in both countries. Based on survey data for African countries,47 the income share of the top 10 percent is On average, income inequality is lower in typically around 30­35 percent (except in Central Africa but still very high. For instance, Southern African countries), compared with in 2011 the top 10 percent of income earners in 34 percent in Europe, 45­55 percent in North Congo received 56 percent, while the bottom and South America and 40­55 percent in 40 percent received 7 percent. East African Asia.48 The comparison could thus suggest that countries are a bit less unequal, especially at the most African countries have low inequality.49 bottom. In Kenya in 2015 the top 10 percent received 48 percent of national income, while But there are good reasons to think that the bottom 40 percent received 9 percent. survey-based data significantly underestimate inequality across Africa. First, the concepts to Income inequality tends to decrease towards measure inequality and growth (at times con- the north and the west of the continent. In sumption, at times income) are often compared Sierra Leone in 2011 the top 10 percent re- indiscriminately, even though using consump- ceived 42 percent of national income, while the tion typically underestimates inequality by bottom 40 percent received 12 percent, and its 25­50 percent compared with using income.50 neighbours show similar income shares. The Second, individuals at the top of the distribu- lowest inequality is in North Africa: In Algeria, tion are largely under-represented in surveys, the least unequal country in Africa for which particularly in developing countries.51 Available estimates are available, the top 10 percent of global and African evidence shows that the earners received 37 percent of national income average income of the top 1 percent of earners in 2011, while the bottom 40 percent received is typically 1.5­2 times higher than what is 14 percent. reported in surveys.52 Heterogeneous trajectories: So, are African countries characterized Inequality trends from 1995 to 2015 by low or high inequality? The question, as simple as it may be, is difficult to answer due There is no single African trend in inequality, to the dissimilarity of data sources. Applying, not even clear regional trends. Income distri- to the extent possible, distributional national butions evolved in a wide variety of ways across accounts methods to Africa yields estimates countries, which underlines the role of national that are more in line with recent ones for devel- institutions and policies in shaping inequality. oped and emerging countries. Such estimates, Given the important differences in data quality however, are still far from perfect and will be across African countries, the lack of harmoniza- greatly improved as more administrative data tion of data collection instruments and welfare are released, as has occurred with Côte d’Ivoire, concepts, and the irregularity of survey availa- Senegal, South Africa and Tunisia. bility, comparing inequality trends is a perilous exercise, and the results must be interpreted with great caution. (In this section, countries

116 | HUMAN DEVELOPMENT REPORT 2019 with an asterisk [] have data available only (box 3.5). For Botswana, Lesotho, Eswatini Inequality, as from 1995 to 2005, and countries with two as- and Namibia** inequality fell: The incomes of measured by the share terisks [] have data available only after 2005.) the bottom 40 percent grew at different paces: of income going to the from 10 percentage points to 88 percentage top 10 percent and to On average, it appears that inequality, as points more than the average. the bottom 40 percent, measured by the share of income going to the increased in Southern top 10 percent and to the bottom 40 percent, In East Africa the income share of the top Africa but fell in East increased in Southern Africa but fell in East 10 percent fell significantly from 1995 to 2000, Africa in the late 1990s Africa in the late 1990s before stabilizing in and the incomes of the bottom 40 percent grew before stabilizing in the the 2000s and stagnated in North, Central more than the average. Since the beginning 2000s and stagnated and West Africa, despite small fluctuations of the 2000s, however, the distribution has in North, Central (figure 3.7). remained rather stable: Income shares fell only and West Africa slightly at the top and grew slightly at the bot- In Southern Africa the dramatic rise of the tom (see figure 3.7). income share of the top 10 percent occurred at the expense of both the middle and the bot- This general trend can be explained by tom of the distribution, whose income shares the decline of inequality in two of the most fell. Indeed, Southern Africa’s performance populous countries, Ethiopia and Kenya. The between 1995 and 2015 was highly nega- overall decline was drastic in Ethiopia, where tive (on average, the incomes of the bottom the incomes of the bottom 40 percent grew 40 percent grew 70 percentage points less than 48 percentage points more than the average. the average) and is the worst among African Inequality rose in most other countries in subregions (table 3.2). This trend was very the subregion. The increase was modest in much driven by South Africa (by far the most Madagascar and more significant in Djibouti, populous country in Southern Africa), which Tanzania and Uganda, where the incomes of saw a strong increase in income inequality the bottom 40 percent grew 6­15 percentage (table 3.3)—despite declining poverty rates.54 points less than the average. In Mozambique** Based on these estimates, it is possible to pres- the incomes of the bottom 40 percent grew ent evidence on the evolution of inequality, 40 percentage points less than the average, and comparing the growth in income of the bottom in Zambia they grew 60 percentage points less. 40 percent with that of the entire population

FIGURE 3.7

Between 1995 and 2015 the income share of the top 10 percent in North Africa and West Africa remained relatively stable, while the share of the bottom 40 percent in Southern Africa declined

Share of total Top 10 percent Share of total Bottom 40 percent income (percent) income (percent)

70 20

65

55 10

5

40 2000 2005 2010 2015 0 2000 2005 2010 2015 1995 1995

North Africa West Africa East Africa Central Africa Southern Africa

Note: Data are weighted by population. Estimates combine survey, fiscal and national accounts data. Source: Chancel and others (2019), based on data from the World Inequality Database (http://WID.world).

Chapter 3 Measuring inequality in income and wealth | 117 TABLE 3.2 grew 33 percentage points more than the aver- age, and in Tunisia, where the incomes of the Difference between income growth of the bottom 40 percent and average income bottom 40 percent grew 54 percentage points growth in Africa’s five subregions, 1995­2015 (percentage points) more than the average. The decline of the in- come share of the top was driven much more Subregion 1995­2015 1995­2005 2005­2015 by the very top of the distribution in Tunisia, East Africa 47.2 40.5 ­4.9 while inequality stagnated in Morocco and Central Africa 11.4 increased modestly in Egypt. North Africa 18.3 7.8 8.0 Southern Africa ­70.3 ­19.2 ­54.8 In West Africa the incomes of the bottom West Africa 0.6 40 percent grew 25 percentage points more 25.0 18.8 than the average. But this hides a wide diversity of trajectories. Inequality rose in Côte d’Ivoire, Note: Estimates combine survey, fiscal and national accounts data. Estimates combine survey, fiscal and national accounts data and are Ghana and Guinea-Bissau, with the incomes of derived from panregional distributions; they are not averages of national indicators. Green (red) cells indicate where the income growth the bottom 40 percent growing 20 percentage rate of the bottom 40 percent was higher (lower) than the average. points less than the average, and even more so Source: Chancel and others (2019), based on data from the World Inequality Database (http://WID.world). in Benin**, with the incomes of the bottom 40 percent growing 30 percentage points less TABLE 3.3 than the average.

Difference between income growth of the bottom 40 percent and average income Inequality declined elsewhere in the subre- growth in selected African countries, 1995­2015 (percentage points) gion. In Senegal the improvement was mild (the incomes of the bottom 40 percent grew Country 1995­2015 1995­2005 2005­2015 only 2 percentage points more than the aver- Algeria 32.5 19.6 9.6 age). In Mauritania the incomes of the bottom Angola 40 percent grew 21 percentage points more Botswana 56.4 ­26.1 71.8 than the average. In Nigeria* the incomes of the Cameroon ­9.8 ­19.3 bottom 40 percent grew 19 percentage points Côte d’Ivoire ­21.2 more than the average. In Niger inequality fell Egypt ­7.1 ­22.1 8.2 substantially, as the incomes of the bottom Ethiopia 48.3 ­5.5 ­0.6 40 percent grew 35 percentage points more Gabon 75.1 ­46.8 than the average. Ghana ­24.1 10.4 a Kenya 12.6 ­13.7 ­4.5 Inequality fell in Gambia, Guinea and Mali*, Madagascar ­0.0 ­8.6 25.7 where the incomes of the bottom 40 percent Mali 10.4 a ­8.4 grew 60­80 percentage points more than the Nigeria ­74.4 70.6 average. The largest inequality declines were in South Africa ­59.6 19.2 ­57.8 Burkina Faso, where the incomes of the bottom Zambia ­22.7 ­20.9 40 percent grew 93 percentage points more than ­24.7 the average, and Sierra Leone, where they grew 117 percentage points more than the average. Note: Estimates combine survey, fiscal and national accounts data. Green (red) cells indicate where the income growth rate of the bottom 40 percent was higher (lower) than the average. Data for Central Africa are scarce and cover a. Average income fell. a short time span. No country showed a strong Source: Chancel and others (2019), based on data from the World Inequality Database (http://WID.world). trend in inequality, up or down, especially at the top. For most countries the data cover only In North Africa the incomes of the bottom 2000 and 2010. In Cameroon**, Chad** and 40 percent grew 18 percentage points more Congo** inequality increased, as the incomes than the average from 1995 to 2015. The de- of the bottom 40 percent grew 13­19 per- cline in inequality resulted from two opposite centage points less than the average. Inequality trends. Inequality fell significantly in Algeria, stagnated in Sao Tome and Principe** and where the incomes of the bottom 40 percent decreased markedly in Gabon**, where the av- erage income fell: the incomes of the bottom 40 percent grew around 12 percentage points more than the average. The two countries with data for 1995 and 2005 are Angola* and

118 | HUMAN DEVELOPMENT REPORT 2019 BOX 3.5 Income growth of the bottom 40 percent—higher than the national average?

Sustainable Development Goal target 10.1 reads, “By Ensuring that the bottom 40 percent sees growth In China the incomes of 2030 progressively achieve and sustain income growth that is at least as high as the average may not be enough the bottom 40 percent of the bottom 40 percent of the population at a rate to contain rising inequalities. Take another example: At grew at an impressive higher than the national average.”1 the global level, average annual pretax income increased 263 percent between 95 percent (net of inflation) for the bottom 40 percent, 2000 and 2018, which Including that inequality target in the list of from 1,300 in 1980 to 2,500 in 2017, but increased contributed to the Sustainable Development Goals was not straight- 40 percent overall, from 11,100 to 16,600. Thus, the fast reduction of forward. Several countries initially opposed it, arguing global bottom 40 percent saw growth that was 45 per- extreme poverty that only poverty reduction mattered.2 Its inclusion thus centage points higher than the global average. marks an important shift in how countries think about sustainable development. At the other end of the distribution, the top 0.1 per- cent’s average annual pretax income increased 117 per- What is the income inequality target about? It seeks cent, from 671,600 to 1,462,000. Despite its small size, to ensure that people in the bottom income groups see the 0.1 percent saw a larger share of total growth than growth that is at least as high as the average. While the the bottom 40 percent of the population—about 12 per- target is meant to be achieved by 2030, it is useful to cent versus about 8.5 percent. Indeed, it is mathematical- look at the past to consider how countries have fared on ly impossible for all groups to see growth that is higher the indicators relevant to the target. The United States, than the average. At the global level, those who lost despite high overall economic growth, the bottom were the middle 40 percent, whose average income rose 40 percent of the population has seen pretax income per just over 33 percent, from 11,900 in 1980 to 15,600 in adult fall by 2 percent, from $13,700 in 1980 to $13,400 2016. So, their share in global income was reduced. This in 2017.3 During the same period the average income shows that ensuring that the bottom 40 percent grows in the United States grew 66 percent, from $41,900 to at the same rate as the average may be insufficient for $61,400. If the bottom 40 percent’s income had grown tackling inequality at all segments of the distribution. as fast as the average, it would be $22,600 today.

1. www.un.org/sustainabledevelopment/inequality/. 2. For a discussion of the debates surrounding inclusion of the income inequality target, see Chancel, Hough and Voituriez (2018). 3. All figures are net of inflation. Since distributional national accounts data for 2014­2016 are not yet available, it was assumed that since 2014, the bottom 40 percent has seen growth that is at least as high as the average—a very optimistic assumption since that occurred only six times between 1980 and 2014, two of which were recessions. Source: World Inequality Lab.

Central African Republic*. In Angola inequali- growth rates led to a rise in income inequality ty increased at both ends of the distribution. In in China. From 2007 to 2018, however, the Central African Republic inequality fell, but so 135 percent growth rate of the bottom 40 per- did average incomes. cent and the 138 percent average in China were much closer, and the rise of inequality halted Inequality in BRIC countries (this stabilization could partly reflect data lim- since the 2000s itations). The more recent period in China is also characterized by wages growing more than This section presents the income growth of output, to the benefit of low-income groups. the bottom 40 percent and the top 1 percent compared with average income growth for In India the income growth of the bottom the four BRIC countries—Brazil, the Russian 40 percent—58 percent between 2000 and Federation, India and China (table 3.4). 2018—was significantly below the average. At the other end of the spectrum the top 1 percent In China the incomes of the bottom saw their incomes grow significantly more than 40 percent grew at an impressive 263 percent the average since 2000 and since 2007. between 2000 and 2018, which contributed to the fast reduction of extreme poverty and to In Brazil the incomes of the bottom 40 per- the decline of the global extreme poverty rate. cent grew 14 percentage points more than the But that growth was significantly below the average between 2000 and 2018. But the top average for China (361 percent) and just half 1 percent also saw higher growth than the av- the rate of the top 1 percent. Such different erage. Since all groups cannot grow more than the average, this means that middle-income groups (between the bottom 40 percent and

Chapter 3 Measuring inequality in income and wealth | 119 TABLE 3.4 Inequality and growth in the BRIC countries

2000­2018 2007­2018

Country Average Bottom Difference Top Average Bottom Difference Top income growth 40 percent between 1 percent income growth 40 percent between 1 percent income growth growth income growth growth (percent) growth of the bottom (percent) (percent) growth of the bottom (percent) (percent) 40 percent (percent) 40 percent and average 16 ­3 and average ­2 income growth income growth (percentage 518 138 (percentage 117 points) points) 213 68 78 Brazil 5 20 14 3 6 68 6 ­20 China 361 263 ­97 135 ­3

India 122 58 ­64 41 ­27

Russian Federation 72 121 49 35 29

Note: Distribution of per adult pretax national income growth. See http://wid.world/methodology for country-level information on the series. Income growth between 2016 and 2018 is assumed to be distribution neutral (all groups benefit from average national income growth). Green (red) cells indicate where the income growth rate of the bottom 40 percent was higher (lower) than the average. Source: Based on data from the World Inequality Database (http://WID.world).

the top 1 percent) were squeezed with lower FIGURE 3.8 than average growth. The income share of the top 1 percent has In the Russian Federation the incomes of the significantly increased in China, India and the bottom 40 percent grew more than the average Russian Federation since the early 1980s between 2000 and 2018, while the incomes of the top 1 percent grew at a rate close to the aver- Top 1 percent share of national age. The top 1 percent actually saw their incomes income (percent) fall between 2007 and 2018. Between 1980 and 2018 the top 0.01 percent saw four-digit income 35 growth rates. Income and wealth inequality to- day remain extreme by global standards, and the 30 recent decline of the top 1 percent has not gone Brazil nearly far enough to reverse this.55 25 A rapid review of growth and inequality trajectories in the BRIC countries shows that 20 India the evolution of the indicators underpinning Russian Sustainable Development Goal target 10.1 Fed. must be interpreted with care. Complementing 15 the bottom 40 percent target with other indica- China tors (such as the income growth rate of the top 1 percent) more fully accounts for the dynamics 10 of growth in a given country. Assessing dynam- ics over various timeframes is also enriching. 5 Good performance over a short time may mask a huge increase in income and wealth inequality 0 in the longer run. The income share of the top 1980 1986 1992 1998 2004 2010 2016 1 percent has significantly increased in China, India and the Russian Federation since the early Note: Distribution of per adult pretax national income growth. See http://wid. 1980s (figure 3.8). In Brazil the income share of world/methodology for country-level information on the series. Income growth the top 1 percent has been broadly stable since between 2016 and 2018 is assumed to be distribution neutral. the early 2000s but at a high level. Source: Based on data from the World Inequality Database (http://WID.world).

Inequality and redistribution in Europe and the United States

Income inequality in European countries and the United States has risen to varying degrees and at different speeds.56 Inequality, both at the top and at the bottom of the distribution, varies widely across developed countries. These

120 | HUMAN DEVELOPMENT REPORT 2019 heterogeneous dynamics are linked to different averages between 1980 and 2007, and richer Driving the rising institutional trajectories, policy choices and people have benefited from a disproportionate inequalities in the patterns of inclusive growth. share of income growth, although the income United States since growth of the bottom 40 percent has been high- the 1980s has been a By combining surveys, tax data and national er than the national average for several countries surge in top incomes accounts, it has become possible to produce since 2007, especially in Eastern Europe. combined with little estimates tracking inequality dynamics across or no pretax income individuals from the bottom to the top Income inequality has risen more in growth among 0.001 percent in a way fully consistent with the United States than in any other poorer individuals national accounts.57 How have European coun- developed country since 1980 tries and the United States performed in pro- moting inclusive growth in the past decades? Driving the rising inequalities in the United States since the 1980s has been a surge in top in- Since the beginning of the 1980s almost no comes combined with little or no pretax income country considered in the analysis has seen the growth among poorer individuals. The current incomes of the bottom 40 percent grow more income inequality in the United States is vastly than the average (table 3.5). Growth has been different from the levels seen at the end of World either distributionally neutral or associated with War II. Indeed, changes in inequality since 1945 rising inequality. In Norway, Spain, France and can be split into two phases (figure 3.9). From Croatia the difference is close to zero: The bot- 1946 to 1980 inequality fell. During that period tom 40 percent saw their incomes grow at a rate the average incomes of the bottom 50 percent similar to that of the average income. In Norway more than doubled. By contrast, the 1980­2014 and France, however, the top 1 percent of in- period coincided with lower and much more comes grew more than the average, meaning skewed growth, with the average income of that the income share of the groups in between the bottom half essentially stagnating (it grew was squeezed. In all other countries, especially less than 2 percent, while that of the bottom in Eastern Europe and the United States, poorer individuals have lagged far behind national

TABLE 3.5 Post-tax average and bottom 40 percent growth in Europe and the United States, 1980­2017 and 2007­2017

1980­2017 2007­2017

Difference Difference

between between

income growth income growth

of the bottom of the bottom

40 percent 40 percent

Income growth and average Income growth Income growth and average Income growth

Average of the bottom income growth of the top Average of the bottom income growth of the top

income growth 40 percent (percentage 1 percent income growth 40 percent (percentage 1 percent

Country (percent) (percent) points) (percent) (percent) (percent) points) (percent)

Eastern Europe

Albania 17.8 20.0 2.2 5.4

Bosnia and Herzegovina 318.7 229.8 ­89.0 475.5 16.7 15.4 ­1.3 16.8

Bulgaria 102.2 39.6 ­62.6 583.3 36.6 30.1 ­6.6 51.9

Croatia 3.8 2.2 ­1.6 77.5 0.8 5.0 4.2 ­2.2

Czechia 37.3 17.6 ­19.7 382.5 10.3 9.5 ­0.9 21.0

Estonia 88.1 44.4 ­43.6 202.7 7.4 8.3 0.9 ­18.8

Hungary 47.1 2.3 ­44.8 426.0 11.8 6.4 ­5.3 2.9

Latvia 48.0 10.4 ­37.7 212.2 12.5 15.2 2.8 19.8

Lithuania 66.9 15.1 ­51.8 318.4 20.8 12.1 ­8.7 31.5

Chapter 3 Measuring inequality in income and wealth | 121 TABLE 3.5 (CONTINUED) Post-tax average and bottom 40 percent growth in Europe and the United States, 1980­2017 and 2007­2017

1980­2017 2007­2017

Difference Difference

between between

income growth income growth

of the bottom of the bottom

40 percent 40 percent

Income growth and average Income growth Income growth and average Income growth

Average of the bottom income growth of the top Average of the bottom income growth of the top

income growth 40 percent (percentage 1 percent income growth 40 percent (percentage 1 percent

Country (percent) (percent) points) (percent) (percent) (percent) points) (percent)

Moldova (Republic of) 36.5 54.6 18.1 23.7

Montenegro ­20.1 ­33.4 ­13.4 16.7 16.2 17.2 1.0 22.3

North Macedonia ­0.2 ­19.3 ­19.1 16.0 22.3 39.1 16.8 10.5

Poland 94.8 33.6 ­61.2 551.2 30.8 28.0 ­2.8 18.0

Romania 69.9 21.0 ­48.9 242.0 30.6 43.0 12.4 ­3.2 Serbia ­8.1 ­27.1 ­19.0 44.4 10.5 9.0 ­1.5 40.6

Slovakia 69.1 57.7 ­11.4 198.0 19.1 19.7 0.6 7.3 12.4 ­7.3 ­19.7 127.7 ­1.1 ­5.6 ­4.5 35.3

Southern Europe ­15.5 ­19.1 ­3.6 ­6.8

Greece ­31.3 ­43.8 ­12.5 5.9 16.5 ­3.5 ­20.0 69.5 ­10.6 ­16.3 ­5.7 ­16.6

Malta 28.8 13.4 ­15.3 183.2 60.1 34.1 ­26.0 54.4 ­0.3 4.3 4.6 ­14.7

Spain 61.1 68.5 7.4 60.0 3.1 1.1 ­2.0 31.0

Western Europe

Austria 53.2 45.6 ­7.7 118.2 ­0.1 ­2.2 ­2.1 20.8 51.3 43.1 ­8.2 79.1 1.6 ­0.6 ­2.2 ­2.5

France 42.3 42.9 0.6 71.0 0.6 1.0 0.5 ­5.5 40.9 21.2 ­19.7 97.9 9.8 3.7 ­6.0 10.7

Ireland 182.0 141.3 ­40.7 323.3 2.9 0.6 ­2.2 4.3 Luxembourg 93.4 63.4 ­30.0 163.5 ­32.6 ­35.9 ­3.3 ­33.0

Netherlands 36.1 26.8 ­9.3 90.6 ­0.6 ­4.2 ­3.7 ­17.6 Switzerland 26.2 21.0 ­5.2 58.4 0.7 4.7 4.0 1.8

United Kingdom 77.9 75.7 ­2.2 136.8 1.3 10.7 9.4 ­23.0

Northern Europe

Denmark 64.7 43.1 ­21.6 263.2 2.4 ­8.6 ­11.0 60.3 68.0 58.7 ­9.4 179.7 ­6.7 ­9.5 ­2.8 ­7.7

Iceland 6.9 15.4 8.6 ­41.4 Norway 84.9 91.9 7.1 158.4 ­2.1 ­0.2 1.9 ­9.6

Sweden 95.5 70.2 ­25.2 172.6 10.5 4.8 ­5.7 ­0.9

United States 63.2 10.8 ­52.4 203.4 3.1 ­0.1 ­3.2 7.6

Note: Green cells indicate countries that achieved Sustainable Development Goal target 10.1 over the period considered and red cells indicate countries that did not. Source: Blanchet, Chancel and Gethin (2019), based on data from the World Inequality Database (http://WID.world).

122 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 3.9

The pretax income share of the top 10 percent in the United States rose from around 35 percent in 1980 to close to 47 percent in 2014

Share of national Pretax

35 Rising inequalities Post-tax in the United States coincide with a

25 gradual decrease in

1917 1927 1937 1947 1957 1967 1977 1987 1997 2007 2017 the progressiveness of

Source: Piketty, Saez and Zucman 2018. the US tax system over the past few decades

40 percent fell 5 percent), and the income of the individuals (total taxes paid as a share of total top 1 percent more than tripled. Accordingly, income) have become more compressed. In the the share of pretax national income received by 1950s the top 1 percent of income earners paid the top 10 percent grew from 34 percent to more 40­45 percent of their pretax income in taxes, than 45 percent, and that received by the top while the bottom 50 percent of earners paid 1 percent grew from 10 percent to 20 percent. 15­20 percent. Today the gap is much smaller. Top earners paid about 30­35 percent, while Accounting for the redistributive effects the poorest half paid around 25 percent. of taxes and transfers does not change the dynamics. Between 1980 and 2014 the share Inequality has increased in a of post-tax national income received by the majority of European countries top 10 percent grew from 30 percent to about 40 percent. During the same period the post-tax Although inequalities remain lower in Europe income of the bottom 50 percent grew a meagre than in the United States, European countries 20 percent, driven entirely by Medicare and have also seen increases in the concentration of Medicaid. Only through in-kind health trans- income at the top. In 1980 income disparities fers and collective expenditures did the incomes were generally higher in Western Europe than in of the bottom half of the distribution rise. Scandinavia and Eastern Europe (figure 3.10). The gap increased between 1980 and 1990 as Rising inequalities in the United States co- income inequality rose in Germany, Portugal incide with a gradual decrease in the progres- and the United Kingdom. In 1990­2000, by siveness of the US tax system over the past few contrast, top income inequality rapidly in- decades, a trend present in many other coun- creased in Finland, Norway and Sweden and in tries (see chapter 7). The country’s share of to- Eastern European countries. As a result, income tal taxes in national income, including federal, inequality is higher today in nearly all European state and local taxes, increased from 8 percent countries than at the beginning of the 1980s. in 1913 to 30 percent in the late 1960s, where it has remained since. Effective tax rates paid by

Chapter 3 Measuring inequality in income and wealth | 123 FIGURE 3.10

Between 1980 and 2017 the share of post-tax national income received by the top 10 percent rose from 21 percent to 25 percent in Northern Europe, while the share received by the bottom 40 percent fell from 24 percent to 22 percent

Share of national Top 10 percent Share of national Bottom 40 percent income (percent) income (percent) 31 27 26 19

15 18 1980 1985 1990 1995 2000 2005 2010 2015 1980 1985 1990 1995 2000 2005 2010 2015

European countries Eastern Europe Northern Europe Southern Europe Western Europe have also seen increases in the Source: Blanchet, Chancel and Gethin (2019), based on data from the World Inequality Database (http://WID.world).

concentration of In 2017 the top 10 percent of income earners top 10 percent had risen to 34 percent, while income at the top. received more than 30 percent of national in- the poorest half of the population received The incomes of the top come in most Western European countries and only a fifth. In the past 37 years the incomes of 0.1 percent of earners 25­35 percent in East European countries.58 the poorest 40 percent of Europeans increases more than doubled 30­40 percent (figure 3.11). The European during the period, The income share of the top 10 percent in middle class benefited only slightly more from and the incomes of Southern Europe was slightly higher than in growth than the poorer groups, as the incomes the top 0.001 percent other regions in the 1980s but increased less of those between percentiles 40 and 90 in- (see figure 3.10). Income gaps widened in Italy creased 40­50 percent. For the more advan- nearly tripled and Portugal, for instance, but remained stable taged sections of society, however, total growth in Spain and fluctuated in Greece. In Northern rates are markedly higher. The incomes of the Europe and Western Europe, by contrast, top 0.1 percent of earners more than doubled income inequality increased more linearly. during the period, and the incomes of the top Eastern Europe is the area where income ine- 0.001 percent nearly tripled. quality has risen the most, due to increases at the top of the distribution in the 1990s and the While income inequality has increased sig- early 2000s.59 Today post-tax income inequality nificantly in Europe, poverty has more or less remains, on average, slightly lower in Northern stagnated. Some 20 percent of Europeans lived Europe than in other regions of the continent. on less than 60 percent of the European median income in 1980, compared with 22 percent in Top income earners have thus been the pri- 2017. In recent years moderate convergence mary beneficiaries of income growth in Europe across countries, due to higher growth in since the 1980s. And between 1980 and 2017 Eastern Europe, has slightly reduced the per- the at risk of poverty rate remained stable or centage of people at risk of becoming poor in rose in most countries.60 Europe as a whole, but the trend has been fully offset by rising percentages in other European Inequality has risen in countries, particularly in Southern Europe. Europe as a whole Convergence would be insufficient to address the percentage of people at risk of poverty in Taking the European countries as a whole, the Europe: If all countries fully converged to the top 10 percent pretax income earners in Europe same average national income, the European- received 29 percent of total regional income in wide percentage would remain as high as 1980, while the bottom 50 percent received 17 percent. 24 percent. In 2017 the income share of the

124 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 3.11

Between 1980 and 2017 the post-tax incomes of the poorest 80 percent of the European population grew close to 40 percent, while those of the top 0.001 percent grew more than 180 percent

Total income Top 1 percent captured growth (percent) 13 percent of growth

50 The combined operation of all the 0 mechanisms acting on pretax incomes 10 20 30 40 50 60 70 80 90 99 99.9 99.99 99.999 enabled Europe to contain the rise of Income group (percentile) the ratio of the top 10 percent to the Note: After the 90th percentile the scale on the horizontal axis changes. The composition of income groups changes from 1980 to 2017, so the estimates do not represent bottom 40 percent the changes in income of the same individuals over time. Source: Blanchet, Chancel and Gethin (2019), based on data from the World Inequality Database (http://WID.world).

The US­Europe comparison points differences between Europe and the United to predistribution and redistribution States. These differences are due mainly to a rise policies to address inequalities in pretax inequality (income measured before direct taxes and transfers, see box 3.3), which has Since 1980 the United States and Europe have been much more marked in the United States. In experienced diverging inequality trajectories. In 1980 the average income of the top 10 percent 2017 the share of national income received by was 10 times higher than that of the bottom the top 1 percent in the United States was more 40 percent in the United States. In 2017 this than twice as large as that received by the poor- multiple jumped above 26. In Europe the same est 40 percent. In Europe, by contrast, the share indicator rose from 10 to 12 over the same period. received by the bottom 40 percent exceeded that received by the top 1 percent (figure 3.12). For post-tax inequality the ratio rose from 7 This was not always the case: In 1980 the share to 14 in the United States between 1980 and of the bottom 40 percent in the two regions was 2017 and from 8 to 9 in Europe (figure 3.14). similar, about 13 percent (figure 3.13). So, the national systems of taxation (which include taxes on income and wealth) and the The divergence in trajectories cannot be systems of social transfers (such as disability accounted for by either trade or technology, benefits or housing support) have therefore not which are often invoked to explain the evolu- enabled the rise in inequalities to be contained tion in inequality in developed countries, given either in the United States or in Europe. that all countries under analysis have been sim- ilarly exposed to both. Instead, the difference The combined operation of all the mechanisms in inequality dynamics appears to be more the acting on pretax incomes enabled Europe to outcome of policy choices and institutional contain the rise of the ratio of the top 10 percent arrangements. to the bottom 40 percent. Social spending—in- cluding mainly public spending on education, The findings reported here allow for a bet- health and retirement pensions—plays an im- ter understanding of the determinants of the portant role. In particular, quality and affordable

Chapter 3 Measuring inequality in income and wealth | 125 FIGURE 3.12

Between 1980 and 2017 the pretax income share of the bottom 40 percent in the United States fell from about 13 percent to 8 percent, while the share of the top 1 percent rose from about 11 percent to 20 percent

Share of national United States Share of national Europe income (percent) income (percent)

25 25

20 20

15 15

10 10

5 5 1980 1985 1990 1995 2000 2005 2010 2015 1980 1985 1990 1995 2000 2005 2010 2015

Bottom 40 percent Top 1 percent

Source: Blanchet, Chancel and Gethin (2019), based on data from the World Inequality Database (http://WID.world).

FIGURE 3.13

Between 1980 and 2017 the average pretax income of the bottom 40 percent grew 36 percent in Europe, while it declined 3 percent in the United States

Bottom 40 percent average Bottom 40 percent average income (relative to 1980) income (relative to 1980)

1.6 Pretax 1.6 Post-tax

1.5 Europe bottom 40 percent growth: +36 percent 1.5 Europe bottom 40 percent growth: +44 percent

1.4 1.4

1.3 1.3

1.2 1.2

1.1 1.1

1.0 1.0

0.9 0.9 US bottom 40 percent growth: +10 percent

0.8 US bottom 40 percent growth: ­3 percent 0.8

0.7 0.7 1980 1985 1990 1995 2000 2005 2010 2015 1980 1985 1990 1995 2000 2005 2010 2015

United States Europe

Source: Blanchet, Chancel and Gethin (2019), based on data from the World Inequality Database (http://WID.world).

education and health systems are key to ensure States.61 Furthermore, access to health and that individuals from low-income backgrounds education is usually more egalitarian in Europe can access economic opportunities. than in the United States, particularly through free or low-cost health care and vocational Social spending remains markedly higher in training in Europe, which contributes to a less Europe than in the United States and the rest unequal distribution of pretax incomes. of the world. It amounts to 25­28 percent of GDP in most countries of continental Europe, Other important dynamics help account for compared with 19 percent in the United higher income growth at the bottom of the

126 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 3.14

The average pretax income of the top 10 percent in the United States was about 11 times higher than that of the bottom 40 percent in 1980 and 27 times higher in 2017, while in Europe the ratio rose from 10 to 12

Ratio of top 10 percent to bottom Ratio of top 10 percent to bottom 40 percent of pretax income 40 percent of post-tax income

30 Pretax 16 Post-tax

26 14

18

10 8

6 6 Still, there has been 1980 1985 1990 1995 2000 2005 2010 2015 1980 1985 1990 1995 2000 2005 2010 2015 a reduction in tax progressiveness in United States Europe Europe in recent decades, with the top Source: Blanchet, Chancel and Gethin (2019), based on data from the World Inequality Database (http://WID.world). corporate tax rate having fallen from distribution in Europe. For example, between FIGURE 3.15 almost 50 percent 1980 and 2017 the minimum wage fell from at the beginning 42 percent of average earnings to 24 percent in Between 1981 and 2017 the average top corporate of the 1980s to the United States. In many European countries tax rate in the European Union fell from about 25 percent today movement has been in the opposite direction, 50 percent to 25 percent, while the average value with the minimum wage maintained at a high added tax rate rose from about 18 percent to more level (as in France, where it is about 50 percent than 21 percent of the average wage) or introduced (as in the United Kingdom in the 1990s and more re- Average top Average standard consumption cently in Germany).62 corporate tax rate (percent) tax rate (percent) Still, there has been a reduction in tax pro- gressiveness in Europe in recent decades, with 50 Top corporate tax rate Standard value 22 the top corporate tax rate having fallen from al- added tax rate 21 most 50 percent at the beginning of the 1980s to 25 percent today—this is part of a global 40 20 trend common to developed and developing countries (see chapter 7). The top marginal in- 30 19 come tax rate has also fallen in most European countries. And the value added tax, which dis- 18 proportionately hits those with low incomes, has risen on average by more than 3 percentage 20 17 points since the beginning of the 1980s. While Europe as a whole has been able to have more 1980 1985 1990 1995 2000 2005 2010 2015 moderate increases in inequality than the United States, these developments may eventu- Source: Eurostat (standard VAT rate) and Organisation for Economic Co-operation ally limit the capacity of governments to get the and Development (top corporate tax rate). winners in European growth to contribute to fi- nancing public services, which have been so key Global wealth inequality: to sustain incomes at the middle and bottom of Capital is back the distribution (figure 3.15). To properly track the dynamics of economic inequality, focusing on income alone is not enough.63 It is also necessary to track the dynam- ics of wealth concentration. Although wealth data remain particularly scarce (even more than income data), recent research has unveiled findings on the evolution and composition of

Chapter 3 Measuring inequality in income and wealth | 127 The globalization of countries’ national wealth. Analysing the com- sheets with information on the total stock of wealth management position of an economy’s national wealth, assets wealth and its evolution. In many emerging and since the 1980s raises that are both privately and publicly owned, is developing countries there is no macroeconomic new challenges, with a prelude to understanding the dynamics of wealth information. Lack of wealth data is an wealth inequality among individuals. issue in itself, since precise information on wealth a growing amount dynamics can prove critical to preventing finan- of world wealth The renewed effort in studying wealth in- cial crises or to fine-tuning tax policies. Lack of held in offshore equality is crucial because it is linked to the data also makes it impossible to properly track the financial centres increase in income inequality at the top of the dynamics of wealth at the micro level—among distribution observed since 1980, since capital individuals. So, macroeconomic discussion of income tends to be concentrated among wealth- wealth is limited to developed economies and a ier people. The prominence of wealth in driving few emerging economies with wealth data. the income distribution is linked to its relative importance in many economies, with national Ratios of private wealth to national wealth as an aggregate having grown significant- income have risen sharply in ly more than income in many countries.64 all countries since 1970, with substantial regional variations Because most countries do not tax wealth directly, producing reliable estimates of wealth Country trajectories in Western Europe have inequality requires combining different data been roughly similar: Net private wealth rose sources, such as billionaire rankings and from 250­400 percent of national income in income tax and inheritance tax data.65 The 1970 to 450­750 percent in 2016 (figure 3.16). globalization of wealth management since The highest increases were in Italy and the the 1980s raises new challenges, with a grow- United Kingdom, where the ratios more than ing amount of world wealth held in offshore doubled. The private wealth­income ratio also financial centres. Indeed, offshore assets are increased greatly in Canada (from 250 per- disproportionately owned by the wealthiest, cent to more than 550 percent) and a bit less so accounting for these offshore assets has (but still substantially) in Australia. It rose large implications for measuring wealth at the by half in the United States (from less than very top of the distribution.66 More generally, 350 percent to around 500 percent) and almost measuring the inequality of income and wealth doubled in Japan (from 300 percent to almost from a global perspective, and not simply at the 600 percent). country level, is becoming critical. China and the Russian Federation had the Understanding the evolution of the level largest increases. In China private wealth and structure of national capital (or national rose from 110 percent of national income in wealth)67 and its relationship to national in- 1978 (when the opening-up policy started) come is key to addressing several economic and to 490 percent of national income in 2015. public policy issues. Wealth is a “stock” concept: In the Russian Federation the ratio tripled It is the sum of all assets accumulated in the past between 1990 and 2015 (from 120 percent to (particularly housing, business and financial 370 percent). assets) net of debt. Private wealth is always more concentrated than income, while public wealth, Note that the 2008 financial crisis did owned by a government, greatly affects the gov- not significantly disturb this trend: Though ernment’s capacity to implement redistributive wealth­income ratios dipped following the policies. This is why looking at the evolution crash, they recovered, at various speeds and to of national wealth-to-income ratios and at the various extents. partition of wealth between the private and the public sectors can help in understanding But public wealth to national income ratios the evolution of economic inequality. Keep in underwent a strong and steady decline almost mind, though, that the definitions of public and everywhere. Public wealth became negative in private property vary across countries.68 the United Kingdom and the United States and now amounts to only 10­20 percent of Reliable macroeconomic data on wealth national income in France, Germany and Japan. are scarce across the globe. Only in 2010 did By contrast, in China the value of public wealth Germany start to publish official national balance

128 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 3.16

Net private wealth in Western European countries rose from 250­400 percent of national income in 1970 to 450­750 percent in 2016

Value of net private wealth

700 China

Russian Fed.

600 Germany

500 France

400 Italy

Japan

United States Public wealth to 300 national income ratios underwent a strong 200 and steady decline almost everywhere

1970 1980 1990 2000 2010

Source: Alvaredo and others (2018), based on data from the World Inequality Database (http://WID.world).

remained fairly constant relative to national stocks) and nonfinancial assets (such as roads) income (250 percent in 1978 and 230 percent but would still be indebted. Taxpayers would in 2015), and in the Russian Federation it fell thus have to continue to pay taxes to reimburse dramatically from more than 230 percent of owners of the debt, and citizens would also national income in 1990 to around 90 percent have to pay a rent to the new owners of the in 2015. stock of capital that was formerly public (roads, energy or water systems, or health or education These two trends have radically modified the infrastructure). Such a situation arguably leaves structure of national wealth in most countries. government with little room to invest in the In the late 1970s the value of public wealth was future (in, say, education or environmental about 50­100 percent of national income in protection) and thus tackle current and future developed countries; it is now negative in the income and wealth inequality. United Kingdom and the United States and only marginally positive in France, Germany A combination of factors accounts for these and Japan. This domination of private wealth trends. The reduction in the share of public in national wealth is a marked change from the wealth accounts for a part of the rise of private 1970s (figure 3.17). wealth. The decline in net public wealth is also due largely to the rise of public debt. The Zero or negative public wealth is exceptional ratio of public assets to national income has by historical standards. Governments tend remained fairly stable because a significant to adopt various strategies to recover positive chunk of public assets was privatized (particu- public wealth levels, such as inflation, debt larly shares in public or semipublic companies) cancellation or progressive wealth taxes—as and the market value of the remaining assets after World War II in Europe (France and increased. But the long-run decline in the share Germany). To understand what a zero or of public wealth in total wealth, in no way in- negative net public wealth situation implies, evitable, is the result of public policy choices consider the following: A government with (privatizing public assets, expanding public negative public wealth willing to repay its debts debt or running fiscal deficits). would have to sell all its financial assets (such as

Chapter 3 Measuring inequality in income and wealth | 129 FIGURE 3.17

Countries are getting richer, but governments are becoming poor

Value of net public and private wealth Private capital

700

600 Spain 500 France 400 Japan

300

High wealth­income 200 ratios imply that wealth Public capital

inequality is going to 100 play a growing role in the overall structure of 0

economic inequality ­100

1970 1980 1990 2000 2010

Source: Alvaredo and others (2018), based on data from the World Inequality Database (http://WID.world).

Overall, the evolution of national wealth wealth was the exclusive driver for the rapid (public and private) to national income ratios is rise of national wealth, at the expense of public determined by the interplay between national wealth. By contrast, China’s privatization of savings, economic growth (quantity factor) and public assets was much more gradual, enabling relative asset prices (price factor). The higher public wealth to remain constant while private the savings rate, the larger the accumulation wealth was increasing. In addition, savings rates of assets. And the higher the economic growth were markedly higher in China. And Chinese rate, the lower the accumulation of assets rel- savings financed mostly domestic capital in- ative to national income. Relative asset prices vestment (leading to more domestic capital depend on institutional and policy factors (rent accumulation), whereas about half of Russian control, for instance) and on the patterns of savings financed foreign investments. Relative saving and investment strategies. In developed asset prices also increased more in China. countries quantity effects contributed to about 60 percent of wealth accumulation between In the long run the low ratios of the mid-20th 1970 and 2010 and price effects to about century may have been due to very special circum- 40 percent, with cross-country variations. stances, perhaps unlikely to recur.69 So savings and growth rates, the main long-run determinants of The differences in privatization strategies and these ratios, will matter greatly in the near future. in price and volume factors also explain the And given their current levels, national wealth to widely divergent patterns of national wealth national income ratios may be returning to those accumulation in the Russian Federation and in the 19th century’s Gilded Age. High wealth­ China. Indeed, Russia’s national wealth in- income ratios imply that wealth inequality is go- creased weakly, from 400 percent of national ing to play a growing role in the overall structure income in 1990 to 450 percent in 2015, while of economic inequality. Because wealth tends to China’s doubled from 350 percent of national be very concentrated, this raises new issues about income in 1978 to 700 percent in 2015. capital taxation and regulation. These issues emerge in a context where the ability of govern- The Russian Federation opted to transfer ments to regulate and redistribute income may be wealth from the public to the private sector as limited by the decline of public wealth. quickly as possible. So the increase in private

130 | HUMAN DEVELOPMENT REPORT 2019 Global wealth inequality This rise was more than offset at the top by Wealth is substantially between individuals the rise in within-country wealth inequality more concentrated everywhere, so wealth increased much faster than income: In The dynamics of wealth inequality between in- at the top of the global distribution: While 2017 the global top dividuals are linked to the evolution of income the average wealth growth was 2.8 percent a 10 percent (the richest inequality and the evolution of public and pri- year per adult over 1987­2017, it was 3.5 per- 10 percent in the vate capital inequality. In the long run wealth cent for the top 1 percent, 4.5 percent for the United States, Europe inequality between individuals also depends top 0.1 percent and 5.7 percent for the top and China) owned on the inequality of savings rates across income 0.01 percent. more than 70 percent and wealth groups, the inequality of labour in- of the total wealth, comes and rates of returns to wealth—and on The factors affecting wealth inequality (in- and the top 1 percent the progressiveness of income and wealth taxes. come inequality, inequality of savings rates owned 33 percent, and asset rates of return) are affected by public while the bottom How have these factors affected the process policies. For example, progressive taxation in- 50 percent owned of wealth concentration in the past, and what fluences income and savings inequality, while less than 2 percent can they tell about potential future dynamics? financial regulation and innovation can have Recent research has shown that relatively small an impact on asset rates of return. Privatization changes in savings behaviours, returns to wealth can also play a role when it benefits mostly a or tax progressiveness can have rather large specific part of the distribution, as in many impacts on wealth inequality.70 This instability countries since the 1980s and particularly in reinforces the need for better data quality to emerging countries. So there is nothing inevi- properly study and understand the dynamics of table about the rise of wealth inequality within income and wealth. countries.

Given the low availability of data on wealth In the Russian Federation and China the con- inequality among individuals, estimates of the centration of wealth increased since the 1990s. global distribution of wealth come from only a The share of the top 1 percent doubled (from handful of countries: France, Spain, the United 22 percent in 1995 to 43 percent in 2015 in Kingdom and the United States and to less ex- the Russian Federation and from 15 percent to tent China. Less certain estimates are also avail- 30 percent in China, although with some vol- able for the Russian Federation and countries atility; figure 3.18). The divergences between in the Middle East. the two countries come from the differences between their privatization strategies: The fast Wealth is substantially more concentrated pace of privatizing public assets in the Russian than income: In 2017 the global top 10 percent Federation favoured the wealthiest even more (the richest 10 percent in the United States, than in China. In the Russia Federation hous- Europe and China) owned more than 70 per- ing had a small dampening effect on the rise of cent of the total wealth, and the top 1 percent inequality. In China housing wealth was privat- owned 33 percent, while the bottom 50 percent ized through a very unequal process, whereas owned less than 2 percent.71 These estimates are the approach was more gradual and equitable a lower bound, since inequality would probably in the Russia Federation. be higher if Africa, Latin America and the rest of Asia were included. The United States has had a less abrupt but no less significant rise of wealth inequality since Wealth inequality has been increasing the mid-1980s, after a considerable decline in since 1980, unaffected by the 2008 crisis. The the 1930s and 1940s, then due particularly to evolution of the global distribution of wealth the policies of the New Deal (see figure 3.18). depends on the disparity of average wealth The share of wealth owned by the top 1 percent between countries and within countries. Since grew from a historic low of 22 percent in 1978 1980 the rise of average private wealth has been to almost 39 percent in the 2010s. The key faster in large emerging economies, such as driver of this increase was the upsurge of very China,72 than in developed countries, because top incomes, enabled by financial deregulation of faster economic growth and massive wealth and lower top tax rates. Inequality of savings transfers from the public to the private sector. rates and of asset return rates amplified the phe- This has greatly increased the wealth of the bot- nomenon in a snowballing trend. Meanwhile, tom 75 percent of the global distribution.

Chapter 3 Measuring inequality in income and wealth | 131 FIGURE 3.18 China Trends in wealth inequality France Russian Federation Share of personal wealth held by the top 1 percent United Kingdom (percent) United States

Wealth inequality 30 has been increasing since 1980, unaffected 20

by the 2008 crisis 10 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010

Source: Alvaredo and others (2018), based on data from the World Inequality Database (http://WID.world).

the income of the middle and the bottom of current trends in savings, income and return the distribution stagnated, and household rate inequality persist, within-country wealth debt (mortgages, student loans and credit card inequality could be returning to 19th century debt, among others) sharply increased. This led Gilded Age levels in the coming decades. On to a substantial fall of the wealth share of the a global scale, if current trends continue, by middle 40 percent—from a historic high of 2050 the global top 0.1 percent could end up 37 percent in 1986 to 28 percent in 2014. owning as much of the world’s wealth as the middle 40 percent of the world’s population In France and the United Kingdom wealth (figure 3.19). inequality also increased after a historical decline, but at a much slower pace than in the Afterword: Data transparency United States. The top 1 percent share rose as a global imperative from 16 percent in both countries in 1985 to 20 percent in the United Kingdom in 2012 and This chapter has discussed recent advances in 23 percent in France in 2015. This was due to methodology and data collection to fill a public greater earnings disparities, amplified by a fall debate data gap. Such information is necessary in tax progressiveness, the privatization of for- for peaceful and deliberative debates over in- merly state-run industries and, most important, come inequality and growth. Worryingly, in the growing inequality of asset return rates, as the few years of the digital age the quality of the returns on financial assets, disproportion- publicly available economic data on these issues ately owned by the wealthy, increased. has been deteriorating in many countries, par- ticularly for fiscal data on capital income, wealth Small changes in savings rate differentials and inheritance. across wealth groups, or in progressive taxation patterns, can have a very large impact on wealth To provide historically and internationally inequality, though it may take several decades comparable estimates of income and wealth for the impacts to play out. This raises many issues for the future of wealth inequality: If the

132 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 3.19

If current trends continue, by 2050 the global top 0.1 percent could end up owning as much of the world’s wealth as the middle 40 percent of the world’s population

Share of global Top 1 percent

35

25 Middle 40 percent

“Global middle class”

20 Today’s knowledge of global income and Top 0.1 percent wealth inequality 15 remains limited and unsatisfactory. Much 10 more data collection Top 0.01 percent lies ahead to expand the geographical 5 coverage of inequality data—and to provide 0 more systematic representations of 1980 1990 2000 2010 2020 2030 2040 2050 pretax and post- tax income and Source: Alvaredo and others (2018), based on data from the World Inequality Database (http://WID.world). wealth inequality

inequality, new efforts require the use of the data limitations, the rise of income and wealth best available data sources from household sur- inequality observed across the world over the veys, administrative tax data, national accounts past decades is not destiny. It arises from eco- or financial leaks. nomic and institutional policy choices. As part III shows, different pathways can be followed To be sure, today’s knowledge of global in- in the coming decades—if there is political come and wealth inequality remains limited will. For the policies of tomorrow to reflect a and unsatisfactory. Much more data collection sound debate on national and global economic lies ahead to expand the geographical coverage inequalities clearly requires the continuing of inequality data—and to provide more sys- publication of transparent and timely data on tematic representations of pretax and post-tax inequalities in income and wealth. income and wealth inequality. Despite these

Chapter 3 Measuring inequality in income and wealth | 133 Spotlight 3.1

Looking within countries and within households

Understanding inequality beyond averages im- cross national borders (see figure S3.1.1 for an plies looking at what is happening subnationally: example with a group of countries in the Gulf within a nation, within a group or even within of Guinea). Clusters of low subnational HDI households. It is particularly important to have values exist in Latin America, including parts a better grasp of who and where those furthest of Central America. In Central­South Asia behind and at the very bottom of the income subnational areas stretch from Tajikistan and distribution are. One way of looking within Kyrgyzstan to most of Afghanistan, and in countries is to identify the hotspots, the subna- Southeast Asia, sections of Cambodia and Viet tional districts, states or provinces set not to have Nam. Not all in a hotspot are necessarily poor, a GDP per capita of $4,000 or more in 2005 of course. Within any area the next step implies purchasing power parity terms in 2030.1 There identifying households most in need of social are 840 such poverty hotspots globally, among assistance. Most countries apply some sort of test more than 3,600 districts, states and provinces. to decide who is eligible for assistance, tests that Moreover, 102 countries have at least one region generally are flawed. A critical challenge for the that qualifies. In other words, people are being tests is their high exclusion errors (not including left behind in a large, diverse group of countries. individuals or households who are eligible but do not receive a benefit) and their high inclusion But there is considerable variation within errors (of individuals or households who are not countries. Over half of low-income countries eligible but do receive a benefit). The inclusion have at least one region that is not a pover- and exclusion errors for a set of African econ- ty hotspot; 36 of 46 lower-middle-income omies are striking (table S3.1.1). For instance, countries have at least one region that is. Even Ghana has an estimated inclusion error of among upper middle-income countries some 35 percent (35 percent of the identified poor 30 percent of regions are hotspots.2 households are nonpoor) and an exclusion error of 63 percent (63 percent of the poor are not Another way of identifying diversity within identified as poor using the proxy means test). countries is to consider the Human Development Index (HDI) at a subnational level.3 By this Finally, it is important to go even deeper to look measure, there are “clusters” of hotspots that within households. As noted, many countries try to identify poor and vulnerable households. FIGURE S3.1.1 There are good reasons for using households as a general proxy. One reason is that data on income Contiguous human development patterns, cutting and consumption are often better collected—and across national borders: The Gulf of Guinea understood—at the household level. A second is that the average well-being of a household is cor- (0.65,0.69] related with individual well-being among those (0.64,0.65] within it. And so while household identification (0.60,0.64] inevitably comes with inclusion and exclusion (0.60,0.60] errors, it has been the standard for decades. (0.59,0.60] (0.53,0.59] The outliers to this pattern are significant (0.49,0.53] and often comprise people with disabilities, or- (0.44,0.49] phans and widows, migrants and mobile popu- (0.41,0.44] lations, and the homeless. The numbers of such cases are considerable. In 30 Sub-Saharan coun- Source: Permanyer and Smits 2019. tries roughly three-quarters of underweight women and undernourished children are not in the poorest 20 percent of households, and around half are not in the poorest 40 percent

134 | HUMAN DEVELOPMENT REPORT 2019 TABLE S3.1.1 Targeting errors of inclusion and exclusion: Proxy means tests

Inclusion Exclusion Inclusion Exclusion Targeting Targeting error rate error rate error rate error rate error error

Fixed poverty line Fixed poverty rate

Country z = F­1 (0.2) z = F­1 (0.4) H = 0.2 H = 0.4 Burkina Faso Ethiopia 0.401 0.751 0.304 0.375 0.522 0.329 Ghana Malawi 0.515 0.945 0.396 0.362 0.621 0.413 Mali Niger 0.354 0.628 0.257 0.350 0.428 0.288 Nigeria Tanzania, United Republic of 0.431 0.880 0.333 0.451 0.353 0.373 Mean 1.000 1.000 0.348 0.485 0.553 0.375

0.539 0.875 0.384 0.340 0.584 0.362

0.332 0.348 0.247 0.243 0.392 0.244

0.396 0.822 0.323 0.291 0.513 0.314

0.357 0.663 0.350 0.294 0.455 0.335

0.481 0.807 0.309 0.359 0.505 0.319

Note: F­1 (x) indicates the poverty line consistent with fixing the poverty rate at x. H = x means headcount poverty rate of x. Source: Brown, Ravallion and van de Walle 2018.

FIGURE S3.1.2 Adult female malnutrition and child stunting can be high in nonpoor households

Proportion of underweight women found Proportion of stunted children found in poor households in wealth-poor households

0.7 0.7

0.6 0.6

0.5 Poorest Poorest

40 percent 0.5 40 percent (r=­0.31) 0.4 0.4 (r=­0.65)

0.3 Poorest

20 percent 0.3 Poorest

(r=­0.31) 0.2 20 percent 0.2 (r=­0.61)

0.1 0.1

0.0 0.0

0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.00 0.10 0.20 0.30 0.40 0.50 0.60

Source: Brown, Ravallion and van de Walle 2017.

(figure S3.1.2). Countries with higher rates falling below the national poverty line is less than 10 percent of undernutrition tend to have a higher share (Lopez-Calva and Ortiz-Juarez 2014). of undernourished individuals in nonpoor 2 Cohen, Desai and Kharas 2019. households.4 3 Permanyer and Smits (2019). 4 New individual consumption data reveal that within-household Notes inequality accounts for nearly 16 percent of total inequality in Senegal. One of the consequences of such unequal repartition 1 This threshold of $4,000 represents twice the ceiling for a of resources within households is the potential existence of low-income country, as defined by the World Bank in 2015. It “invisible poor” in households classified as nonpoor. As many corresponds roughly to a daily income where the probability of as 12.6 percent of poor individuals live in nonpoor households. The evidence from Senegal suggest that the more complex the household structure and the bigger the household size, the more inequality is likely to be underestimated when computed using

standard consumption surveys (Lambert and de Vreyer 2017).

Chapter 3 Measuring inequality in income and wealth | 135 Spotlight 3.2

Choosing an inequality index

James Foster, Professor of Economics and International Affairs at the George Washington University, and Nora Lustig, Samuel Z. Stone Professor of Latin American Economics and Director of the Commitment to Equity Institute at Tulane University

A useful way to describe the distribution of to the right of L1, so an inequality index would income is the Lorenz curve, constructed as be expected to indicate greater inequality in follows.1 First, the population is ranked ac- the L2 case. Another way to see this is that the cording to income (or consumption, wealth or poorest x percent of the population will always another measure of resources) from the lowest have an equal or greater share of income under to the highest. Then the cumulative shares L1 than under L2, regardless of what x is. This of individuals in the population are plotted is called the Lorenz dominance criterion or against their respective cumulative share Lorenz criterion for short. in total income. The curve drawn is called the Lorenz curve. The horizontal axis of the What constitutes a “good” inequality index? Lorenz curve shows the cumulative percent- One approach is to require the measure to be ages of the population arranged in increasing consistent with the Lorenz criterion: that is, order of income. The vertical axis shows the to be Lorenz consistent. For a measure to be percentage of total income received by a Lorenz consistent the following two conditions fraction of the population. For example, the must hold: First, inequality rises (declines) (80 percent, 60 percent) point on the Lorenz when the Lorenz curve lies everywhere below curve means that the poorest 80 percent of (above) the original Lorenz curve as with L2 the population receives 60 percent of total compared with L1 (L1 compared to L2) in the income while the richest 20 percent receives figure. Second, inequality is the same when 40 percent of total income.2 Lorenz curves are identical. For a measure to be Weakly Lorenz Consistent, condition 1 Figure S3.2.1 shows two Lorenz curves: L1 becomes the following: 1’. inequality rises (de- and L2. If everybody has the same income, the clines) or stays the same when the Lorenz curve Lorenz curve will coincide with the 45-de- lies everywhere below (above) the original gree line. The greater the level of inequality, Lorenz curve. the farther the Lorenz curve will be from the 45-degree line. In the figure, L2 lies below and A second approach is to require the inequal- ity index to fulfil the following four principles: FIGURE S3.1.1 1 Symmetry (or anonymity). If two people

Lorenz curve switch incomes, the index level should not change. Cumulative income 2 Population invariance (or replication in- variance). If the population is replicated or L1 “cloned” one or more times, the index level L2 should not change. 3 Scale invariance (or mean independence). If Cumulative population all incomes are scaled up or down by a com- mon factor (for example, doubled), the index Source: Authors’ creation. level should not change. 4 Transfer (or the Pigou-Dalton Transfer Principle). If income is transferred from one person to another who is richer, the index level should increase. In other words, in the face of a regressive transfer, the index level must rise.

136 | HUMAN DEVELOPMENT REPORT 2019 It can be shown that indices satisfying these top income shares focus on limited ranges of four principles are Lorenz Consistent and vice incomes and thus violate the transfer prin- versa. ciple (and thus violate Lorenz consistency). The latter means that transfers entirely within These indices include: or entirely outside the relevant ranges have no

• Summary indices based on relatively com- effect on measured inequality. For example,

the 10/40 ratio is insensitive to regressive plex formulas designed to capture inequal- transfers that stay within the poorest 40 per- ity along the entire distribution. The most cent, within the richest 10 percent or within commonly used are (in alphabetical order): the remaining 50 percent in the middle, while the Atkinson, Gini and Theil measures (and the income share of the top 1 percent is in- the generalized entropy measures, more sensitive to transfers within the top 1 percent generally). and within the bottom 99 percent. Despite While inequality measures that satisfy the disagreeing with the transfer principle, and transfer principle are in common use, there are thus the Lorenz criterion, these partial indi- also simpler indices that do not satisfy 1­4 but ces are useful for conveying easily understood are popular. These include: information about the extent of inequality.

• Partial indices based on simple formulas that Importantly, they satisfy the weak transfer

focus on inequality across certain parts of principle and thus guarantee that in the face the distribution. These include the Kuznets of a regressive transfer anywhere along the ratios expressed as the income share of top distribution, inequality measured by any of x percent over the income share of bottom these indices will never decline but, notably, y percent. There are, of course, many possible it can stay the same. Kuznets ratios. The one proposed by the Nobel Laureate Simon Kuznets was 20/40.3 In contrast, other common inequality indices Partial indices also include the top income do not even fulfil the weak transfer principle shares, expressed as the income share of the (transfer principle 4’). Examples include the top x percent. Common examples include quantile ratios (such as the income of percen- the income share of the top 1 percent or of tile 90 to the income of the 10th percentile also the top 10 percent.4 The top income shares known as the p90/p10 ratio) and the variance are, in fact, limiting cases of Kuznets ratios of logarithms. For example, a transfer from the obtained by setting the “bottom” income 5th percentile to the 10th would reduce the share to cover the entire population: that is, p90/p10 ratio despite the fact that the trans- by setting y percent = 100 percent.5 fer is clearly regressive because it redistributes Such partial Indices satisfy the following income from the very poor to the less poor. principle: Regressive transfers at the upper end of the dis- 4’ Weak transfer principle: If income is trans- tribution can lower the variance of logarithms ferred from one person to another who is and lead to extreme conflicts with the Lorenz richer (or equally rich), the index level should criterion.6 increase or remain unchanged. In other words, in the face of a regressive Finally, the mean to median ratio (mean di- transfer, the inequality index can never decline, vided by the median) is a measure of skewness but it may remain unchanged. It can be shown that can also be interpreted as a partial index of that indices satisfying 1­3 and 4’ principles are inequality. Virtually every inequality measure is weakly Lorenz consistent and vice versa. a ratio of two “income standards” that summa- In sum, the summary indices of Atkinson, rize the size of the income distributions from Gini and Theil (and the whole family of two perspectives: one that emphasizes higher Generalized Entropy Indices) satisfy princi- incomes and a second that emphasizes lower ples 1­3 and 4 and thus are Lorenz consistent incomes.7 So long as only distributions that are (and vice versa). This guarantees that in the skewed to the right are considered, the mean face of a regressive (progressive) transfer exceeds the median, and the mean to median anywhere along the distribution, inequality ratio takes on this form. This index satisfies the measured by any of these indices will rise first three principles but can violate the weak (decline). In contrast, the Kuznets ratios and transfer principle when the regressive transfer

Chapter 3 Measuring inequality in income and wealth | 137 raises the median income. Like the other partial TABLE S3.2.1 indices, it is weaker in terms of the properties it satisfies but has the advantage of simplicity and Statistics most frequently published in 10 is often used in political economy.8 commonly used international databases

How to apply the above in practice? When Statistic Frequency making pairwise comparisons, first graph the Gini 9 Lorenz curves. If the Lorenz curves do not Quantile ratio 90/10 4 cross, an unambiguous Lorenz comparison can Theil 3 be made. One can conclude from this that any Top 10 percent 3 reasonable (that is, Lorenz consistent) measure would agree that inequality has unambiguously Source: Authors’ creation. increased or declined, according to what the Lorenz curves indicates. However, it is also pos- Thus, the most frequently reported ine- sible that the Lorenz curves cross, in which case quality measures include two that are Lorenz reasonable inequality measures can disagree. consistent (the Gini and Theil measures), What can be done when Lorenz curves cross? one that is weakly Lorenz consistent (the top One approach is to narrow the set of reasonable 10 percent) and one that is neither (the 90/10 inequality measures using an additional crite- quantile ratio). In addition to inequality meas- rion. For instance, transfer-sensitive measures ures, international datasets report other sta- are Lorenz consistent measures that emphasize tistics. Among those, the most frequent is the distributional changes at the lower end over distribution of income by decile.10 those at the upper end. The Atkinson class and the two Theil measures (including the mean log Notes deviation) are transfer-sensitive measures. By contrast, the coefficient of variation (standard 1 Named after Max Otto Lorenz, a US economist who developed deviation divided by the mean) is neutral with the idea of the Lorenz curve in 1905. respect to where transfers occur, while many other generalized entropy measures emphasize 2 Often, especially with historical data, we only have distributional changes at the upper end and grouped-data or information on equal-sized population groups thus are not in the set of transfer-sensitive such as quintiles or deciles (5 or 10 groups, respectively). measures. The resulting Lorenz curve is an approximation of the actual Lorenz curve where inequality within each group has been When do all transfer-sensitive measures suppressed. agree? As a subset of Lorenz-consistent meas- ures, they agree when Lorenz curves do not 3 Some international databases report the 20/20 (sometimes cross as well as in many cases when they do called S80/S20) and 10/40 ratios. cross. For example, suppose that Lorenz curves cross once and that the first Lorenz curve is 4 The top 1 percent has been the focus of the recent literature higher at lower incomes than the second. There on top incomes. See, for example, Atkinson, Piketty and Saez is a simple test: The first has less inequality than (2011). the second, according to all transfer-sensitive measures exactly when the coefficient of varia- 5 By definition, 100 percent of the population receives 100 per- tion for the first is no higher than that for the cent of the income so the denominator of the Kuznets ratio second.9 An even simpler approach is to select becomes 100/100 = 1, and thus the 1/100 Kuznets ratio equals a (finite) set of particularly relevant inequality 1 percent. measures for making inequality comparisons. If all agree on a given comparison, the result is 6 Foster and Ok 1999. robust. If not, the conclusion is ambiguous for 7 Foster and others (2013, p. 15). For example, one Atkinson that set of measures, with inequality ranked one way for some measures and reversed for others. measure compares the higher arithmetic mean to the lower geometric means; the 1 percent income share effectively Table S3.2.1 shows the statistics most fre- compares the higher 1 percent mean to the lower arithmetic quently published in commonly used interna- mean. tional databases.9 8 The mean to median ratio is the inequality measure used by Meltzer and Richards (1981) in their model to predict the size of government. The greater the ratio, the higher the taxes and redistribution. 9 For details, see Shorrocks and Foster (1987). See also Zheng (2018), who presents additional criteria for making compari- sons when Lorenz curves cross. 10 The complete set of measures reported in international databases and their properties can be found in supplemental material for this spotlight available at http://hdr.undp.org/ en/2019-report.

138 | HUMAN DEVELOPMENT REPORT 2019 Spotlight 3.3

Measuring fiscal redistribution: concepts and definitions

A number of databases publish indicators of postfiscal income concepts. There are important the extent of income redistribution due to differences, and some can have significant impli- taxes and transfers. For example, they publish cations for the scale of redistribution observed. prefiscal and postfiscal Gini coefficients and other indicators of inequality and poverty. In The following table compares the definitions alphabetical order, the multicountry and mul- of income used by the six databases mentioned tiregional databases most frequently used are above. the Commitment to Equity Institute’s (CEQ) Data Center on Fiscal Redistribution (Tulane There are five important differences: University), the Organisation for Economic · While all six databases start out with similar Co-operation and Development’s (OECD) Income Distribution Database, the LIS Cross- definitions of factor income, the additional National Data Center in Luxembourg and the components included in prefiscal income World Inequality Database (Paris School of differs. This is important because the pre- Economics). In addition, there are two regional fiscal income is what each database uses to databases: EUROMOD (Institute for Social rank individuals prior to adding transfers and Economic Research, University of Essex), and subtracting taxes and will thus affect a tax-benefit microsimulation model for the the ensuing redistribution results (see point European Union, and the OECD­Eurostat on the treatment of pensions below). For Expert Group on Disparities in a National example, EUROMOD does not include the Accounts Framework (EGDNA).1 value of consumption of own production as part of prefiscal income, while the rest of One feature these databases have in common the databases do. EUROMOD, the Income is that they rely on fiscal incidence analysis, the Distribution and LIS do not include the method used to allocate taxes and public spend- (imputed) value of owner-occupied housing, ing to households so that incomes before taxes while the other three do. There is also a fun- and transfers can be compared with incomes damental difference in the treatment of con- after them. Standard fiscal incidence analysis tributory pensions (see the next paragraph). just looks at what is paid and what is received Finally, the World Inequality Database also without assessing the behavioural responses includes undistributed profits in its defini- that taxes and public spending may trigger for tion of prefiscal income. individuals or households. This is often referred · Second, EGDNA, EUROMOD, the Income to as the “accounting approach.”2 Distribution Database and the LIS treat old-age pensions from social security as The building block of fiscal incidence analysis pure transfers, while the World Inequality is the construction of income concepts. That is, Database treats them (together with un- starting from a prefiscal income concept, each employment benefits) as pure deferred new income concept is constructed by subtract- income. The CEQ Data Center on Fiscal ing taxes and adding the relevant components Redistribution presents results for both of public spending to the previous income scenarios. This assumption can make a sig- concept. While this approach is broadly the nificant difference in countries with a high same across all five databases mentioned, the proportion of retirees whose main or sole definition of the specific income concepts, the income stems from old-age pensions. For income concepts included in the analysis and example, in the European Union the redis- the methods to allocate taxes and public spend- tributive effect with contributory pensions ing differ. This spotlight focuses on comparing as pure transfers is 19.0 Gini points while the definition of income concepts—that is, on it is 7.7 Gini points when old-age pensions the types of incomes, taxes and public spending are treated as pure deferred income.3 In the included in the construction of the prefiscal and United States the values are 11.2 for pure transfers and 7.2 for pure deferred income.4

Chapter 3 Measuring inequality in income and wealth | 139

• Third, EUROMOD, the Income Notes

Distribution Database and the LIS present information on fiscal redistribution for di- The author is very grateful to Carlotta Balestra (EGDNA), Maynor rect taxes and direct transfers while the CEQ Cabrera (CEQ), Lucas Chancel (World Inequality Database, Paris Data Center on Fiscal Redistribution also School of Economics), Michael Forster and Maxime Ladaique (OECD includes the impact of indirect taxes and sub- Income Distribution Database), Teresa Munzi (Luxembourg Income sidies and transfers in kind, and the World Study), Daria Popova (EUROMOD, University of Essex) and Jorrit Inequality Database includes all government Zwijnenburg (EGDNA) for their inputs to the table on the compari- revenues and spending. EGDNA does not son of income concepts. include indirect taxes and subsidies but in- 1 Details on the methodologies applied by each database can be cludes transfers in kind (education, health and housing). found in the following: CEQ Data Center on Fiscal Redistribution: Lustig 2018a, chapters 1, 6 and 8; EGDNA: Zwijnenburg,

• Fourth, in the published information on Bournot and Giovannelli 2017; EUROMOD: Sutherland and Figari

preconstructed variables, the CEQ Data 2013; OECD Income Distribution Database: OECD 2017b; LIS: Center on Fiscal Redistribution reports forthcoming DART methodology document; World Inequality indicators based on income per capita, Database: Alvaredo and others 2016. EGDNA, EUROMOD, the Income 2 For an in-depth discussion of the fiscal incidence methodology, Distribution Database and LIS report them see, for example, Lustig (2018a). based on equivalized income5 and the World 3 The data for EU 28 are from EUROMOD statistics on distri- Inequality Database reports them based on bution and decomposition of disposable income, accessed at income per adult.6 www.iser.essex.ac.uk/euromod/statistics/ using EUROMOD version G3.0. The difference is probably an overestimation

• Fifth, all but EGDNA and the World because in many cases one cannot distinguish between

Inequality Database report incomes as they contributory and social pensions. appear in the microdata, while EGDNA and 4 See chapter 10 in Lustig (2018a). the World Inequality Database adjusts all 5 Equivalized income is equal to household income divided variables to match administrative totals in by square root of household members excluding domestic tax records and national accounts. servants. 6 An adult is defined by the World Inequality Database as an individual older than 20 years of age.

Source: Lustig forthcoming.

140 | HUMAN DEVELOPMENT REPORT 2019 TABLE S3.3.1

Income concept CEQ EGDNA EUROMOD IDD LIS WID.World Prefiscal Market income Market income Primary income Market income Market income Market income Pretax income plus pensions

Factor income Factor income Factor income Factor income Factor income Factor income Factor income

Undistributed profits

PLUS PLUS Old-age pensions Old-age from social pensions and security schemes unemployment benefits from social security

PLUS PLUS PLUS PLUS PLUS PLUS PLUS Transfers received Transfers received Imputed value of Transfers received Transfers received Transfers received Transfers received from nonprofit from nonprofit owner-occupied from nonprofit from nonprofit from nonprofit from nonprofit institutions and institutions and housing services institutions and institutions and institutions and institutions and other households, other households, and consumption other households other households other households other households, payments from imputed value of of own production and consumption and consumption payments from employment- owner-occupied of own production of own production employment- related pension housing services related pension schemes, imputed and consumption schemes, imputed value of owner- of own production value of owner- occupied housing occupied housing services and services and consumption of consumption of own production own production

MINUS MINUS Contributions to Contributions old-age pensions to old-age in social security pensions and schemes unemployment in social security

Chapter 3 Measuring inequality in income and wealth | 141

Income concept CEQ EGDNA EUROMOD IDD LIS WID.World Postfiscal: disposable Disposable Disposable Disposable Disposable Disposable Disposable Post-tax Postfiscal: consumable income income income income income income disposable income

Market income Market income Primary income Market income Market income Market income Market income

PLUS PLUS PLUS PLUS PLUS PLUS PLUS Old-age pensions Old-age pensions Old-age pensions Old-age pensions Other cash Other cash Old-age pensions and other cash and other and other and other benefits benefits received cash benefits cash benefits cash benefits (excluding old-age benefits and other from social received from received from received from pensions and security systems, social security social security social security unemployment (excluding old-age cash benefits social assistance systems and systems and systems and benefits) from benefits and social assistance social assistance social assistance public social pensions) from received from transfers received benefits benefits benefits insurance and from (paid social assistance social security and social security to) nonprofit benefits institutions and social assistance systems and other households

benefits social assistance

benefits

MINUS MINUS MINUS MINUS MINUS MINUS MINUS Contributions to Contributions to Contributions to Contributions to Contributions to Contributions to Contributions to other (excluding old-age pensions, old-age pensions, old-age pensions, old-age pensions, old-age pensions, other (excluding old-age pensions) unemployment unemployment unemployment unemployment unemployment old-age social security and other benefits and other benefits and other benefits and other benefits and other benefits pensions and schemes in social security in social security in social security in social security in social security unemployment) schemes schemes schemes schemes schemes in social security schemes schemes

MINUS MINUS MINUS MINUS MINUS MINUS MINUS Direct personal Direct personal Direct personal Direct personal Direct personal Direct personal Direct personal income and income taxes income taxes income taxes income taxes income taxes income and property taxes property taxes

Consumable Consumable na na na na na

Disposable Disposable

PLUS PLUS consumption consumption subsidies subsidies

MINUS MINUS

consumption taxes consumption taxes

(value added, (value added,

excise, sales and excise, sales and

the like) the like)

142 | HUMAN DEVELOPMENT REPORT 2019

Income concept CEQ EGDNA EUROMOD IDD LIS WID.World Postfiscal: including Final income na na transfers in kind Final income Adjusted na Post-tax national Memo items disposable income Contributory pensions Post-tax Welfare indicatora Consumable Consumable Disposable disposable income Total values income income income Unit PLUS PLUS PLUS PLUS Indirect Public spending Public spending Public spending consumption on education and on education, on education, subsidies public spending health and health and on health housing housing MINUS Indirect consumption taxes (value added, excise, sales and the like) and other taxes.

Public spending on education, health, defense, infrastructure and other public spending

Deferred income Government Government Government Government Government Deferred income transfer transfer transfer transfer transfer Income Income Income Income Income Income Income As implied by Match national microdata As implied by Match national As implied by As implied by As implied by accounts Per capita microdata accounts microdata microdata microdata Per adultc Per capita Equivalizedb Equivalizedb Equivalizedb Equivalizedb

na is not applicable. CEQ is the Commitment to Equity Institute Data Center on Fiscal Redistribution. EGDNA is the Organisation for Economic Co-operation and Development (OECD)­Eurostat Expert Group on Disparities in a National Accounts Framework. IDD is the OECD Income Distribution Database. LIS is the LIS Cross-National Data Center. WID.world is the World Inequality Database. a. When household surveys include only consumption expenditures (no information on income), CEQ Data Center on Fiscal Redistribution assumes that consumption expenditures equal disposable income and constructs the other income concepts as specified above, while the World Inequality Database transforms consumption distributions into income distributions using stylized savings profiles in countries where income data are not available. b. Equivalized income equals household income divided by the square root of household members (excluding domestic help). c. An individual is classified as an adult if he or she is older than age 20. Source: CEQ Data Center on Fiscal Redistribution: Lustig 2018a, chapter 6 (http://commitmentoequity.org/publications-ceq-handbook); OECD­Eurostat Expert Group on Disparities in a National Accounts Framework: www. oecd.org/officialdocuments/publicdisplaydocumentpdf/?cote=STD/DOC(2016)10&docLanguage=En; EUROMOD: www.euromod.ac.uk/publications/euromod-modelling-conventions; https://www.euromod.ac.uk/using-euromod/ statistics; LIS: forthcoming DART methodological document; OECD Income Distribution Database: www.oecd.org/els/soc/IDD-ToR.pdf; World Inequality Database: https://wid.world/document/dinaguidelines-v1/.

Chapter 3 Measuring inequality in income and wealth | 143 Chapter 4

Gender inequalities beyond averages: Between social norms and power imbalances 4.

Gender inequalities beyond averages: Between social norms and power imbalances

Gender disparities remain among the most persistent forms of inequality across all countries.1 Given that these disad- vantages affect half the world’s people, gender inequality is arguably one of the greatest barriers to human development. All too often, women and girls are discriminated against in health, in education, at home and in the labour market—with negative repercussions for their freedoms.

Progress in reducing gender inequality over Development Report’s Gender Inequality The world is not on the 20th century was remarkable in basic Index—a measure of women’s empowerment track to achieve achievements in health and education and par- in health, education and economic status— gender equality by 2030 ticipation in markets and politics (figure 4.1).2 shows that overall progress in gender inequali- Much of this progress was celebrated with the ty has been slowing in recent years.5 Beijing Platform for Action during the 1995 Fourth World Conference on Women.3 But Consider two developments. First, gender as the event’s 25th anniversary approaches gaps are deeper than originally thought. Time in 2020, many challenges to equality remain, magazine’s 2017 Person of the Year was “the particularly for enhanced capabilities that alter silence breakers,” women who denounced power relations and enhance agency. abuse. Accomplished women were unprotect- ed against persistent sexual abuse. The silence The world is not on track to achieve gender breakers were also given voice by the #MeToo equality by 2030. Based on current trends, movement, which uncovered abuse and vul- it would take 202 years to close the gender nerability for women, well beyond what is gap in economic opportunity.4 The Human covered in official statistics. In Latin America,

FIGURE 4.1

Remarkable progress in basic capabilities, much less in enhanced capabilities

Enhanced Agency capabilities and change

Social Tradeoffs/ norms power imbalances

Subsistence and Basic participation capabilities

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 147 Gender inequality is too, the #NiUnaMenos movement has shed structural barriers to equality. The tradeoffs correlated with a loss light on femicides and violence against women are influenced strongly by social norms and in human development from Argentina to Mexico.6 by a structure of mutually reinforcing gender gaps. These norms and gaps are not directly due to inequality Second, there are troubling signs of diffi- observable, so they are often overlooked and culties and reversals on the path towards gen- not systematically studied. der equality—for female heads of state and government and for women’s participation Gender inequality in in the labour market, even where there is a the 21st century buoyant economy and gender parity in access to education.7 And there are signs of a back- Gender inequality is intrinsically linked to hu- lash. In several countries the gender equality man development, and it exhibits the same dy- agenda is being portrayed as part of “gender namics of convergence in basic capabilities and ideolog y.”8 divergence in enhanced capabilities. Overall, it is still the case—as Martha Nussbaum has In other words, precisely when awareness is pointed out—that “women in much of the increasing more needs to be done to achieve world lack support for fundamental functions gender equality, the path becomes steeper. This of a human life.”9 This is evident in the Gender chapter explores why progress is slowing, identi- Inequality Index and its components—reflect- fying today’s active barriers that pose challenges ing gaps in reproductive health, empowerment for future prospects for equality, which include and the labour market. No place in the world personal and public beliefs as well as practices has gender equality. In Sub-Saharan Africa 1 in that generate biases against gender equality. It every 180 women giving birth dies (more than stresses that gender inequality reflects intrinsic 20 times the rate in developed countries), and imbalances in power—something well known adult women are less educated, have less access to women’s movements and feminist experts— to labour markets than men in most regions and documents two trends: and lack access to political power (table 4.1).

• Gender inequalities are intense, widespread

Gender inequality as a human and behind the unequal distribution of hu- development shortfall man development progress across levels of socioeconomic development. Gender inequality is correlated with a loss in

• Gender inequality tends to be more intense human development due to inequality (fig-

in areas of greater individual empowerment ure 4.2). No country has reached low inequal- and social power. This implies that progress ity in human development without restricting is easier for more basic capabilities and harder the loss coming from gender inequality. for more enhanced capabilities (chapter 1). Investing in women’s equality and lifting both The first trend indicates the urgency in ad- their living standards and their empowerment dressing gender inequality to promote basic are central to the human development agenda. human rights and development. The second “Human development, if not engendered, raises a red flag about future progress. Progress is endangered,” concluded the pioneer 1995 at the basics is necessary for gender equality, Human Development Report, based on similar but it is not enough. evidence.10 Social norms and gender-specific tradeoffs are key barriers to gender equality. Social and Today looks different from 1995. The 1995 cultural norms often foster behaviours that Human Development Report noted sizeable perpetuate inequalities, while power concen- gender disparities, larger than today’s, but doc- trations create imbalances and lead to capture umented substantial progress over the preced- by powerful groups such as dominant, patri- ing two decades, particularly in education and archal elites. Both affect all forms of gender health, where the prospect of equality was inequality, from violence against women to visible. The conclusion: “These impressions are the glass ceiling in business and politics. In cause for hope, not pessimism, for the future.”11 addition, gender-specific tradeoffs burden the complex choices women encounter in work, family and social life—resulting in cumulative

148 | HUMAN DEVELOPMENT REPORT 2019 TABLE 4.1 Gender Inequality Index: Regional dashboard

Gender Maternal Adolescent Share of Population with at least Labour force Inequality mortality birth rate seats in some secondary education participation rate (births per parliament (% ages 15 and older) Index ratio 1,000 women (% held by (% ages 25 and older) (deaths per ages 15­19) women) 100,000 live

births)

Female Male Female Male

Region 2018 2015 2015­2020 2018 2010­2018 2010­2018 2018 2018 Arab States 0.531 East Asia and the Pacific 0.310 148.2 46.6 18.3 45.9 54.9 20.4 73.8 Europe and Central Asia 0.276 Latin America and the Caribbean 0.383 61.7 22.0 20.3 68.8 76.2 59.7 77.0 South Asia 0.510 Sub-Saharan Africa 0.573 24.8 27.8 21.2 78.1 85.8 45.2 70.1

67.6 63.2 31.0 59.7 59.3 51.8 77.2

175.7 26.1 17.1 39.9 60.8 25.9 78.8

550.2 104.7 23.5 28.8 39.8 63.5 72.9

Source: Human Development Report Office (see Statistical table 5)

FIGURE 4.2 intense, and progress towards gender equality is slowing (figure 4.3). The space for gains based Gender inequality is correlated with a loss in on current strategies may be eroding, and un- human development due to inequality less the active barriers posed by biased beliefs and practices that sustain persistent gender Loss in human development due inequalities are addressed, progress towards to gender inequality (percent) equality will be far harder in the foreseeable 90 future.

60 Gender inequality and empowerment: Catching up in the basics, widening 30 gaps in enhanced capabilities

0 0 10 20 30 40 50 Accumulating capabilities requires achieve- ments of different natures. As chapter 1 Inequality in Human Development Index discussed, progress in human development is On the positive side distribution (percent) linked to expanding substantive freedoms, ca- women are catching pabilities and functionings from basic to more up in basic areas of Note: Countries mapped by their Gender Inequality Index performance relative enhanced. Progress towards equality tends to development. But to their performance on the Inequality-adjusted Human Development Index. The be faster for basic capabilities and harder for progress has been higher the loss due to gender inequality, the greater the inequality in human enhanced capabilities. Gender equality­related uneven as women development. capabilities follow a similar pattern. pull away from basic Source: Human Development Report Office. areas into enhanced On the positive side women are catching up ones, where gaps Today, the prospects are different. The past in basic areas of development. Legal barriers tend to be wider two decades have seen remarkable progress in to gender equality have been removed in most education, almost reaching parity in average countries: Women can vote and be elected, primary enrolment, and in health, reducing the they have access to education, and they can par- global maternal mortality ratio by 45 percent ticipate in the economy without formal restric- since 2000.12 But gains in other dimensions tions. But progress has been uneven as women of women’s empowerment have not been as pull away from basic areas into enhanced ones, where gaps tend to be wider.

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 149 FIGURE 4.3 Progress towards gender equality is slowing

Gender Inequality Index (mean value) 0.500

0.300 0.000

1995 2000 2005 2010 2011 2012 2013 2014 2015 2016 2017 2018

Source: Human Development Report Office (see Statistical table 5).

Women make greater These patterns can be interpreted as reflect- portfolios in affairs such as transport, econom- and faster progress ing the distribution of individual empower- ics or finance. Certain disciplines are typically ment and social power: Women make greater associated with feminine or masculine charac- where their individual and faster progress where their individual em- teristics, as also happens in education and the empowerment or powerment or social power is lower (basic capa- labour market. bilities). But they face a glass ceiling where they social power is lower have greater responsibility, political leadership Economic participation also shows a gra- (basic capabilities). and social payoffs in markets, social life and dient (see figure 4.4, right panel). When But they face a politics (enhanced capabilities) (figure 4.4). empowerment is basic and precarious, women glass ceiling where This view of gradients in empowerment is are over-represented, as for contributing fam- they have greater closely linked to the seminal literature on basic ily workers (typically not receiving monetary and strategic needs coming from gender plan- payment). Then, as economic power increases responsibility, political ning (box 4.1). from employee to employer, and from employ- leadership and social er to top entertainer and billionaire, the gender Take access to political participation (see gap widens. payoffs in markets, figure 4.4, left panel). Women and men vote in social life and politics elections at similar rates. So there is parity in Empowerment gradients appear even for (enhanced capabilities) entry-level political participation, where power a uniform set of companies, as with the gen- is very diffused. But when more concentrated der leadership gap in S&P 500 companies. political power is at stake, women appear Although women’s overall employment by severely under-represented. The higher the these companies might be close to parity, power and responsibility, the wider the gender women are under-represented in more senior gap—and for heads of state and government it positions. is almost 90 percent. In developing countries most women who Similar gradients occur even for women who receive pay for work are in the informal sec- reach higher power. Only 24 percent of na- tor. Countries with high female informal tional parliamentarians were women in 2019,13 work rates include Uganda, Paraguay, Mexico and their portfolios were unevenly distributed. and Colombia (figure 4.5), where more than Women most commonly held portfolios in 50 percent of women are protected by minimal environment, natural resources and energy, regulations; have few or no benefits; lack voice, followed by social sectors, such as social af- social security and decent work conditions; and fairs, education and family. Fewer women had are vulnerable to low salaries and possible job loss.

150 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 4.4 The greater the empowerment, the wider the gender gap

Global gender gap in politics Global gender gap per type of employment (gap with respect to parity, percent) (gap with respect to parity, percent)

100 100

80 80

60 40

40 20

0

-20

0 -40 Contributing Own Votea Lower house Upper house Speakers of Head of Employees Employers Top 100 Top 500 family account entertainers billionaires or single parliament government workers house

Basic Enhanced Basic Enhanced

a. Assumes an equal proportion of men and women in the voting population. Source: Human Development Report Office calculations based on data from the World Values Survey, the Inter-Parliamentary Union, ILO (2019b) and Forbes (2019).

BOX 4.1 and relations, such as a law condemning gender-based violence, equal access to credit, equal inheritance and Practical and strategic gender interests and needs others. Addressing these should alter gender power relations. Sometimes practical and strategic needs The notion of practical and strategic gender interests coincide—for example, the practical need for child and needs (pioneered by Caroline Moser),1 which in- care coincides with the strategic need to get a job out- forms much of the gender policy analysis framework, side the home.3 The difference is comparable to that is connected to the conception of basic and enhanced between basic and enhanced capabilities discussed capabilities and achievements in this Report. As ar- in this Report. Transformative changes that can bring ticulated in gender social policy analyses,2 practical about normative and structural shifts are the strongest gender needs refer to the needs of women and men predictors of practical and strategic interventions ex- to make everyday life easier, such as access to water, panding women’s agency and empowerment for gen- better transportation, child care facilities and so on. der equality. Addressing these will not directly challenge gender power relations but may remove important obstacles to women’s economic empowerment. Strategic gender needs refer to needs for society to shift in gender roles

1. Molyneux 1985; Moser 1989. 2. Moser 1989. 3. SIDA 2015.

Today, women are the most qualified in women’s reproductive roles (see Dashboard 2 history, and newer generations of women in the statistical annex), revealing one of the have reached parity in enrolment in primary moving targets discussed in chapter 1. Some education.14 But it now seems that this is not represent a natural part of the process of de- enough for achieving parity in adulthood. velopment—the constant need to push new The transition from the education system to boundaries to achieve more. Others represent the world of paid work is marked by a gen- the response of deeply rooted social norms to der equality discontinuity, associated with preserve the underlying structure of power.

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 151 FIGURE 4.5

The percentage of informal employment in nonagricultural employment in developing countries is generally higher for women than for men

Informal employment in nonagricultural Male Female employment, 2017 (percent)

0 Uganda Paraguay Mexico Colombia Costa Thailand Chile South Turkey Russian Serbia Ukraine Moldova Greece

Rica Africa Federation (Rep. of)

Source: ILO 2019b.

Gender inequality has Are social norms and power in many countries still cannot reach their full long been associated imbalances shifting? potential.19

with persistent Gender inequality has long been associated with Beliefs about what others do and what others discriminatory social persistent discriminatory social norms prescrib- think a person in some reference group should ing social roles and power relations between do, maintained by social approval and disap- norms prescribing men and women in society.15 Social norms held proval, often guide actions in social settings.20 social roles and power by individuals and their reference groups are So it is useful to measure the beliefs and atti- relations between men values, beliefs, attitudes and practices that assert tudes that create biases and prejudices towards and women in society preferred power dynamics for interactions be- women’s empowerment in society. tween individuals and institutions.16 As broader constructs, norms are operationalized through Social norms cover several aspects of an in- beliefs, attitudes and practices.17 dividual’s identity—gender, age, ethnicity, reli- gion, ability and so on—that are heterogeneous People’s expectations of individuals’ roles and multidimensional. Discriminatory social in households, communities, workplaces and norms and stereotypes reinforce gendered iden- societies can determine a group’s functioning. tities and determine power relations that con- Women often face strong conventional societal strain women’s and men’s behaviour in ways expectations to be caregivers and homem akers; that lead to inequality. Norms influence expec- men similarly are expected to be breadwin- tations for masculine and feminine behaviour ners.18 Embedded in these social norms are considered socially acceptable or looked down longstanding patterns of exclusion from house- on. So, they directly affect individuals’ choices, hold and community decisionmaking that freedoms and capabilities. limit women’s opportunities and choices. So, despite convergence on some outcome indica- Social norms also reflect regularities among tors—such as access to education at all levels groups of individuals. Rules of behaviour are and access to health care—women and girls set according to standards of behaviour or ideals attached to a group’s sense of identity.21 Individuals have multiple social identities and

152 | HUMAN DEVELOPMENT REPORT 2019 behave according to identity-related ideals, freedom, and beliefs about social censure and they also expect others sharing a common and reproach create barriers for individuals identity to behave according to these ideals. who transgress. For gender roles these beliefs Norms of behaviour related to these ideals can be particularly important in determining affect people’s perception of themselves and the freedoms and power relations with other others, thus engendering a sense of belong- identities—compounded when overlapping ing to particular identity groups. The beliefs and intersecting with those of age, race and people hold about appropriate behaviour class hierarchies (box 4.2). often determine the range of choices and preferences that they exercise—in that con- How prevalent are biases from social norms? text norms can determine autonomy and How are they evolving? How do they affect gender equality? These are difficult questions,

BOX 4.2 Overlapping and intersecting identities

When gender identities overlap with other identities, norms and stereotypes of exclusion can be associated Overlapping identities they combine and intersect to generate distinct prejudic- with different identities. For instance, regarding medi- must be considered in es and discriminatory practices that violate individuals’ an years of education completed in Angola and United research and policy equal rights in society. Intersectionality is the complex, Republic of Tanzania, an important gap distinguishes analysis because cumulative way the effects of different forms of discrim- women in the highest wealth quintile from those in the different social norms ination combine, overlap or intersect—and are amplified second or lowest quintile (see figure). If the differences and stereotypes of when put together.1 A sociological term, intersectionality are not explicitly considered, public programmes may exclusion can be refers to the interconnected nature of social categories leave women in the lowest quintiles behind. associated with such as race, class, gender, age, ethnicity, ability and different identities residence status, regarded as creating overlapping and Moreover, individuals’ different social identities interdependent systems of discrimination or disadvan- can profoundly influence their beliefs and experiences tage. It emerges from the literature on civil legal rights. about gender. People who identify with multiple minori- It recognizes that policies can exclude people who face ty groups, such as racial minority women, can easily be overlapping discrimination unique to them. excluded and overlooked by policies. But the invisibil- ity produced by interacting identities can also protect Overlapping identities must be considered in re- vulnerable individuals by making them less prototypical search and policy analysis because different social targets of common forms of bias and exclusion.2

How gaps in median years of education distinguish rich from poor in Angola and United Republic of Tanzania, 2015

6.5 6.6 7.0

6.5 6.0 6.2 6.4

4.4

1.4 0

Total Lowest Second Middle Fourth Highest Total Lowest Second Middle Fourth Highest

15­49 quintile quintile 15­49 quintile quintile

Angola United Republic of Tanzania

Note: Lowest quintile refers to the poorest 20 percent; highest quintile refers to the wealthiest 20 percent. Source: Demographic and Health Surveys.

1. IWDA 2018. 2. Biernat and Sesko 2013; Miller 2016; Purdie-Vaughns and Eibach 2008.

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 153 mainly because social norms and attitudes are (figure 4.9). At the other extreme, indicating a hard to observe, interpret and measure. But backlash, the share of men with no bias fell in using data from the World Values Survey wave 5 Sweden, Germany, India and Mexico. (2005­2009) and wave 6 (2010­2014), a social norms index can be constructed to capture how The share of women with no gender so- social beliefs can obstruct gender equality along cial norms bias increased the most in the multiple dimensions (figure 4.6 and box 4.3). Netherlands, Chile and Australia. But most countries in the sample showed a backlash, led Widespread biases and backlash by Sweden, India, South Africa and Romania (see figure 4.9).

The multidimensional The multidimensional gender social norms Gender inequality and social norms gender social norms count index and high-intensity index (see indices appear linked box 4.3) show widespread biases in gender The multidimensional gender social norms to gender inequality. social norms. According to the count index, indices appear linked to gender inequality, as only 14 percent of women and 10 percent of might be expected. In countries with higher bi- In countries with men worldwide have no gender social norm ases (measured through the multidimensional higher biases, overall bias (figure 4.7). Women are skewed towards gender social norms indices), overall inequality less bias against gender equality and women’s (measured by the Gender Inequality Index) is inequality is higher empowerment. Men are concentrated in the higher (figure 4.10). Similarly, the indices are middle of the distribution, with 52 percent positively related to the Gender Inequality having two to four gender social norms biases. Index in time spent on unpaid domestic chores The high-intensity index shows that more than and care work. half the world’s people have a high-intensity bias against gender equality and women’s Biases in social norms also show a gradient. empowerment. The political and economic dimensions of the multidimensional gender social norms index Both indices provide evidence of a stagnation indicate biases for basic women’s achievement or a backlash from 2005­2009 to 2010­2014. and against more enhanced women’s achieve- The share of both women and men worldwide ment (figure 4.11). Overall, the biases appear with no gender social norms bias fell (figure 4.8). more intense for more enhanced forms of women’s participation. The proportion of peo- Progress in the share of men with no gender ple favouring men over women for high-level social norms bias was largest in Chile, Australia, political and economic leadership positions is the United States and the Netherlands

FIGURE 4.6 How social beliefs can obstruct gender and women’s empowerment

Dimensions Political Educational Economic Physical integrity

Indicators Men make better Women have University is more Men should have Men make better Proxy for Proxy for political leaders the same rights important for a man more right to a business executives intimate reproductive than women do than for a woman job than women than women do partner violence as men rights

Dimension index Political empowerment Educational empowerment Economic empowerment Physical integrity index index index index

Multidimensional gender social norms index

154 | HUMAN DEVELOPMENT REPORT 2019 BOX 4.3 The multidimensional gender social norms index—measuring biases, prejudices and beliefs

Research prepared for this Report proposed the multidimensional gender so- includes 77 countries and territories accounting for 81 percent of the world cial norms index to capture how social beliefs can obstruct gender equality population. The second set consists of only countries with data for both along multiple dimensions. The index comprises four dimensions—political, wave 5 and wave 6. This set includes 32 countries and territories accounting educational, economic and physical integrity—and is constructed based for 59 percent of the world population. on responses to seven questions from the World Values Survey, which are used to create seven indicators (see figure 4.5 in the main text). The an- Definition of bias for the indicators of the multidimensional gender swer choices vary by indicator. For indicators for which the answer choices social norms index are strongly agree, agree, disagree and strongly disagree, the index defines individuals with a bias as those who answer strongly agree and agree. For Dimension Indicator Choices Bias definition the political indicator on women’s rights, for which the answer is given on a numerical scale from 1 to 10, the index defines individuals with a bias as Political Men make better Strongly agree, Strongly agree those who choose a rating of 7 or lower. For the physical integrity indicators, political leaders agree, disagree, and agree for which the answer also ranges from 1 to 10, the index defines individuals than women do strongly disagree with a bias using a proxy variable for intimate partner violence and one for Intermediate reproductive rights. Women have the 1, not essential, form: 1­7 same rights as to 10, essential Aggregation men For each indicator a variable takes the value of 1 when an individual has a bias and 0 when the individual does not. Two methods of aggregation are Educational University is Strongly agree, Strongly agree then used in reporting results on the index. more important agree, disagree, and agree for a man than strongly disagree The first consists of a simple count (equivalent to the union approach), for a woman where the indicators are simply summed and therefore have the same weight. This result has a minimum of 0 and a maximum of 7: Economic Men should have Agree, neither, Strongly agree more right to a disagree and agree The calculation is a simple addition of dichotomic variables, but it com- job than women plicates the disaggregation and analysis by dimension and indicator. Strongly agree, Agree Men make agree, disagree, To address this, the second method follows the Alkire­Foster methodol- better business strongly disagree ogy,1 which counts the different gender social norm biases that an individual executives than faces at the same time (following the intersection approach). These dimen- women do sions are analysed to determine who has a bias on each indicator. This result counts only people with high-intensity bias. Physical integrity Proxy for 1, never, to 10, Strongest form: intimate partner always 2­10 The methods are applied to two sets of countries. The first set consists violence of countries with data for either wave 5 (2005­2009) or wave 6 (2010­2014) 1, never, to 10, Weakest form: 1 of the World Values Survey and uses the latest data available. This set Proxy for always reproductive rights

1. Alkire and Foster 2011.

higher than the proportion of people favouring potential and are much more frequent among men over women in access to basic political men than women.22 rights or paid employment. Gradients in biases are likely to affect elec- Several theories linked to social norms could tions and economic and family decisions, mak- account for these differences. One suggests an ing gender equality more difficult to reach when inability to discern between confidence and higher levels of empowerment are at stake. competence. If people misinterpret confidence as a sign of competence, they can mistakenly What causes change—and what believe that men are better leaders than women determines the nature of change? when men are simply more confident. In other words, for leadership the only advantage that How can practices and behaviours either men have over women is that manifestations change or sustain traditional gender roles? of overconfidence, often masked as charisma or Norms can change as economies develop, by charm, are commonly mistaken for leadership changes in communications technology, by

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 155 FIGURE 4.7 would miss a deeper understanding of social change.24 Only 14 percent of women and 10 percent of men worldwide have no gender social norms biases Consider the subtle differences between descriptive and injunctive norms.25 Descriptive Percent of surveyed population responding norms are beliefs about what is considered a with biases towards gender equality normal practice in a social group or an area. and women’s empowerment Injunctive norms state what people in a com- munity should do. This distinction is important Female Male for practice, as it can lead to an understanding of why some aspects of gender norms and rela- 50 52 tions shift faster than others.26

24 18 21 The family sets norms, and experiences 12 from childhood create an unconscious gen- 14 10 der bias.27 Parents’ attitudes towards gender influence children through mid-adolescence, No gender 1 bias 2­4 biases More than and children at school perceive gender roles.28 biases 5 biases Parenting practices and behaviours are thus among the predictors of an individual’s Norms can change Note: Balanced panel of 77 countries and territories with data from wave 6 gendered behaviours and expectations. For as economies (2010­2014) of the World Values Survey, accounting for 81 percent of the world instance, children tend to mimic (in attitudes population. and actions) how their parents share paid and develop, by changes Source: Mukhopadhyay, Rivera and Tapia (2019), based on data from the World unpaid work.29 in communications Values Survey. technology, by new Parenting experiences may, however, in- laws, policies or new laws, policies or programmes, by social and fluence and change adults’ social norms and political activism and by exposure to new ideas established gender roles. In the “mighty girl programmes, by social and practices through formal and informal effect,” fathers raise their awareness of gender and political activism channels (education, role models and media).23 disadvantages when they are rearing daugh- and by exposure ters.30 Parenting a school-age girl makes it easi- to new ideas and Policymakers often focus on the tangible— er for men to put themselves in their daughter’s practices through on laws, policies, spending commitments, shoes, empathize with girls facing traditional formal and informal public statements and so forth. This is driven gender norms and embrace nontraditional channels (education, partly by the desire to measure impact (and ones that would not place their daughters at a thus prove effectiveness), by frustration disadvantage to men in the labour market.31 role models and media) with the vagueness of “talking shops” ar- guing about rights and norms and by sheer Adolescence is another key stage for gender impatience with the slow pace of change. socialization, particularly for boys.32 Young Yet neglecting the invisible power of norms

FIGURE 4.8

The share of both women and men worldwide with no gender social norms bias fell between 2005­2009 and 2010­2014

Percent of surveyed population responding 2005­2009 with biases towards gender equality 2010­2014

Indicated bias in one or Female 40.1 43.3 fewer questions from the Male

Indicated bias in two or Female 56.7 59.9 World Values Survey Male 69.7 70.4

Note: Balanced panel of 32 countries and territories with data from both wave 5 (2005­2009) and wave 6 (2010­2014) of the World Values Survey, accounting for 59 percent of the world population. Source: Mukhopadhyay, Rivera and Tapia (2019), based on data from the World Values Survey.

156 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 4.9

Progress in the share of men with no gender social norm bias from 2005­2009 to 2010­2014 was largest in Chile, Australia, the United States and the Netherlands, while most countries showed a backlash in the share of women with no gender social norms bias

Men Women

Chile Netherlands Australia Chile Netherlands Australia Argentina China

Poland Slovenia Thailand Japan

Japan Ukraine Trinidad and Tobago Trinidad and Tobago

Spain Thailand Korea (Republic of) United States

China Jordan Georgia Germany Romania Russian Federation Morocco Jordan Mexico Ghana Malaysia Malaysia Cyprus Rwanda Spain Russian Federation Uruguay Uruguay Brazil South Africa Ghana Rwanda Brazil Argentina Ukraine Poland Turkey Korea (Republic of) Slovenia Morocco Mexico Turkey Georgia India Romania Germany South Africa Sweden India

­0.1 ­0.05 0 0.05 0.1 0.15 ­0.1 0 0.1 0.2

Mean change (value) Mean change (value)

Note: Balanced panel of 32 countries and territories with data from both wave 5 (2005­2009) and wave 6 (2010­2014) of the World Values Survey, accounting for 59 percent of the world population. Source: Mukhopadhyay, Rivera and Tapia (2019), based on data from the World Values Survey.

FIGURE 4.10 Countries with higher social norms biases tend to have higher gender inequality

Gender Time spent on unpaid domestic chores Inequality Index (value) (ratio between women and men)

0.8 12

8

6

0.2 4

2

0

0 0.2 0.4 0.6 0 0.2 0.4 0.6

Social Norms Index (value) Social Norms Index (value)

Source: Mukhopadhyay, Rivera and Tapia (2019), based on data from the World Values Survey and Dashboard 2 in the statistical annex.

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 157 FIGURE 4.11

Biases in social norms show a gradient

Biases against women’s Men make better Men should have Men make better capabilities (percent) political leaders more rights to a job business executives than women do Women should have than women than women do the same political 49.6 42.1 rights as men 29.1 32.2

Basic Enhanced Basic Enhanced Politics The economy

Note: Balanced panel of 77 countries and territories with data from wave 6 (2011­2014) of the World Values Survey, accounting for 81 percent of the world population. Source: Mukhopadhyay, Rivera and Tapia (2019), based on data from the World Values Survey.

Powerlessness adolescents in different cultural settings com- or think differently.36 Because of intertwined manifests itself monly endorse norms that perpetuate gender social dynamics,37 challenging discriminatory as an inability to inequalities, and parents and peers are central norms that impede gender equality and wom- participate in or in shaping such attitudes. Some of the endorsed en’s empowerment requires acting on more influence decisions masculinity norms relate to physical toughness than one factor at a time. that profoundly affect (showing higher tolerance for pain, engaging in one’s own life, while fights, competing in sports), autonomy (being Restricted choices and power more powerful actors financially independent, protecting and pro- imbalances over the lifecycle make decisions despite viding for families), emotional stoicism (not neither understanding “acting like girls” or showing vulnerabilities, Gender inequality within households and the situation of the dealing with problems on their own) and het- communities is characterized by inequality vulnerable nor having erosexual prowess (having sex with many girls, across multiple dimensions, with a vicious their interests at heart exercising control over girls in relationships) cycle of powerlessness, stigmatization, discrim- (box 4.4).33 ination, exclusion and material deprivation all reinforcing each other. Powerlessness manifests Social convention refers to how compliance itself in many ways, but at its core is an inabil- with gender social norms is internalized in ity to participate in or influence decisions that individual values reinforced by rewards or profoundly affect one’s own life, while more sanctions. Rewards use social or psychological powerful actors make decisions despite neither approvals, while sanctions can range from ex- understanding the situation of the vulnerable clusion from the community to violence or le- nor having their interests at heart. Human gal action. Stigma can limit what is considered development is about expanding substantive normal or acceptable and be used to enforce freedoms and choices. This section presents stereotypes and social norms about appropri- evidence of restricted or even tragic choices ate behaviours. A social norm will be stickiest women face.38 when individuals have the most to gain from complying with it and the most to lose from Examples of restricted choices can be iden- challenging it.34 Social norms have enough tified in a lifecycle approach. Some represent power to keep women from claiming their legal blatant limits to basic freedoms and human rights due to pressure to conform to societal rights; others represent subtle manifestations of expectations.35 gender biases. The disparities of childhood and adolescence are amplified when women reach Social norms can also prevail when individ- uals lack the information or knowledge to act

158 | HUMAN DEVELOPMENT REPORT 2019 BOX 4.4 behaviours of the gender roles restrict men to act in a cer- Challenging rigid tain way that preserves existing power structures. In 2019 gender norms and The man box Promundo along with Unilever estimated the economic power dynamics impacts of the man box in Mexico, the United Kingdom in households and Engaging men and boys is a critical piece of advancing and the United States, considering bullying, violence, de- communities and the gender equality agenda. Gender equality implies pression, suicide, binge drinking and traffic accidents as involving men and changing and transforming the way individuals express costs of restricting men to masculine behaviours.2 Two of boys in making these and experience power in their lives, relationships and the most damaging consequences for men are related to changes are important communities. Reaching equality, women and men will their mental health: Men are less likely to seek mental have the same agency to make choices and participate health services than women are, and men are more likely in society. While women and girls bear the brunt of to die by suicide than women are. Besides the ethical and gender inequalities, men and boys are also affected by social gains of gender equality, men as individuals can traditional conceptions of gender. benefit from expressing freely, from having more options in their own experiences and behaviours and from having Gender is a social construct of attributes or roles better and healthier relationships with women and girls. associated with being male or female. What it means to be a man or a woman is learned and internalized So challenging rigid gender norms and power dy- based on experiences and messages over the course of namics in households and communities and involving a lifetime, normalized through social structures, culture men and boys in making these changes are important. and interactions. Though men usually have more agency Engaging men in preventing gender-based violence, than the women in their lives, men’s decisions and be- supporting women´s economic empowerment, pursuing haviours are also profoundly shaped by rigid social and change for reproductive health and acting as fathers or cultural expectations related to masculinity. caregivers are examples of how men can challenge their notions of masculinity and of their own selves. Masculinity is the pattern of social behaviours or practices associated with ideals about how men should behave.1 Some characteristics of masculinity relate to dominance, toughness and risk-taking, recently referred to as toxic masculinity or the man box, in that traditional

1. Ricardo and MenEngage 2014. 2. Heilman and others 2019.

adulthood, as exemplified in the differences in had an imbalanced sex ratio at birth, today 21 labour force participation and the representa- countries have a skewed ratio. The preference tion of women in decisionmaking positions for a son can lead to sex-selective abortions in business and in politics (see figure 4.4). and to a large number of “missing” women, For unpaid care work, women bear a bigger particularly in some South Asian countries.40 burden, providing more than three times as Discrimination continues through how much as men.39 And older women’s challenges households share resources. Girls and women accumulate through the life course: They are sometimes eat last and least in the household.41 less likely than men to have access to pensions, The gender politics of food—nurtured by as- even though they are expected to live three sumptions, norms and practices about women years longer. Along the way, social norms and needing fewer calories—can push women into path dependence—how outcomes today affect perpetual malnutrition and protein deficiency. outcomes tomorrow—interact to form a highly complex system of structural gender gaps. Education opportunities, including access and quality, are affected by both household Birth, early childhood and school age and community social norms. Gender differ- ences manifest first in girls’ families over edu- In some cultures traditional social norms can cation as a human right and later over respect affect girls even before they are born, since for women’s agency to decide to study and to some countries deeply prefer bearing sons choose her preferred field. Social norms can over daughters. While in the 1990s only some define the level of education a girl can attain countries had the technology available to de- or her choice of study. The restriction, control termine a baby’s gender and only 6 countries and monitoring of a girl’s or woman’s behav- iour and decisionmaking about her education

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 159 Social norms and or job, or her access to financial resources presents disproportionate risks to women’s and traditional behaviour or their distribution, constitute economic girls’ health, reflecting both biological differ- generally pose a threat violence against her (see spotlight 4.1 at the ences and social norms (see box 4.3). And early end of the chapter). And even when girls are marriage limits girls’ choices. to women’s health educated as well as boys, other effects of ine- quality—driven especially by gendered social The adolescent birth rate among women norms—reduce the likelihood that women ages 15­19 is 104.7 per 1,000 in Sub-Saharan will later attain positions of power and partici- Africa and 63.2 in Latin America and the pate in decisionmaking. Caribbean. When a teenage girl becomes preg- nant, her health is endangered, her education Worldwide, one in eight age-eligible girls and job prospects can abruptly end and her vul- does not attend primary or secondary school. nerability to poverty and exclusion multiplies.49 Only 62 of 145 countries have gender parity Adolescent pregnancy, often a result of a girl’s in primary and secondary education.42 Despite lack of opportunities and freedom, can reflect the progress in enrolment ratios for some a failure among those around her to protect her countries, large differences persist in learning rights. outcomes and education quality. Contraception is important in maintain- Even among children attending school, ing good sexual and reproductive health.50 determinants of occupational choices appear Contraceptive use is higher among unmarried very early. Girls are less likely to study subjects and sexually active adolescents, but so is the such as science, technology, engineering and unmet need for family planning, especially in mathematics, while boys are a minority of those Asia and the Pacific and Sub-Saharan Africa studying health and education.43 (figure 4.12). There is still a stigma in many countries around unmarried women needing Adolescence and early adulthood family planning services. And in some coun- tries regulations prevent access to these servic- Adolescence is when girls’ and boys’ futures es. Moreover, many women cannot afford to start to diverge; while boys’ worlds expand, pay for health care. girls’ worlds contract.44 Every year 12 mil- lion girls are victims of forced marriage.45 Girls Social norms and traditional behaviour gen- forced to get married as a child are victims of erally pose a threat to women’s reproductive a human rights violation and are condemned health. Women are more vulnerable to a loss to live a life with heavily restricted choices and of agency to have a satisfying and safe sex life, low human development. the capability to reproduce and the freedom to decide if, when and how often to do so.51 When Child marriage not only alienates girls men use their power to decide on women’s from their families and social networks but behalf, that limits women’s access to resources also increases their risk of becoming victims of domestic violence.46 It exacerbates overall FIGURE 4.12 gender inequality in education and employ- ment by greatly reducing a girl’s chances of Contraceptive use is higher among unmarried completing formal schooling and developing and sexually active adolescent girls, but so is the skills for employment outside the home.47 It unmet need for family planning, 2002­2014 also leads to early and multiple pregnancies, increasing health risks for both the married Currently married/in a union girls and their children, since the risks of Unmarried and sexually active newborn death and infant mortality and morbidity are higher in children born to 51 women under age 20.48 41

The health effects of early marriage are 20 23 among the many health risks that are higher for women and girls than for men and boys. One Contraceptive prevalence, Unmet need for of the most globally widespread cross-cutting any method (percent) family planning (percent) forms of horizontal inequality, early marriage Source: UNFPA 2016.

160 | HUMAN DEVELOPMENT REPORT 2019 and dictates women’s behaviour. More broadly, income regions have a narrower gap in unpaid Gender differences in if women are seen as objects rather than agents care work. The regions with the widest gaps paid and unpaid work in households and communities, this form of are the Arab States, South Asia, Sub-Saharan and the gradients horizontal inequality can lead to violence and Africa and Latin America and the Caribbean— in empowerment harassment (see spotlight 4.1 at the end of the the same regions that have the widest gaps combine multiple chapter), affecting women’s mental health.52 for women´s labour force participation (fig- elements that restrict ure 4.13). The struggle to reconcile care work women’s choices Adulthood and older age responsibilities with paid work can lead women to occupational downgrading, where they Globally, women do more unpaid work than choose employment below their skill level and men do.53 However, the global gender income accept poorer working conditions.57 gap is 44 percent (see Statistical table 4). Gender differences in paid and unpaid work Some constraints faced by women are invis- and the gradients in empowerment combine ible when gaps are seen in isolation. Statistics multiple elements that restrict women’s choices. typically record achievements (the func- The gaps illustrate the multidimensional effects tionings) but not the full set of choices (the of gender inequality on occupation choices, capabilities). This partial view tends to hide income and women’s financial independence the multidimensional biases in choices wom- and resilience to external shocks. en face. Take, for instance, a qualified woman who has children and must decide between A key constraint on women’s decisionmaking taking a job and staying home. Workplace is their disadvantages in the amount of unpaid inequalities (including pay gaps58 and the risk work they do, bearing disproportionate re- of harassment), social norms (pressure to fulfil sponsibility for housework, caring for family the role of mother) and imbalances at home (a members and performing voluntary communi- greater load of domestic unpaid work), among ty work.54 On average, women spend about 2.5 other factors, may deter her from participating times as much time on unpaid care and domes- in paid work. The woman’s choice may bring tic work as men do.55 This affects women’s la- feelings of guilt or regret. A large proportion bour force participation, lowers economywide of female homemakers feel that by staying productivity and limits their opportunities to home they are giving up a career or economic spend time in other ways.56 This sort of gender independence. A large proportion of mothers inequality is linked to levels of income: Higher employed in paid occupations face the stress of

FIGURE 4.13

The gap in unpaid care work persists in developing economies

Proportion of time spent on unpaid Female Male domestic chores and care work (percent)

21.6 21.0

19.4 19.2

16.8 15.2

5.8 5.5 5.9 5.2 5.6

Arab States East Asia and the Europe and Latin America South Asia Sub-Saharan Pacific Central Asia and the Africa Caribbean

Note: Aggregation rule has been relaxed; estimates not published in dashboard.

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 161 Women’s financial feeling that their choice implies suffering for stability61 or to legal discrimination and gender independence can their children (figure 4.14). norms.62 Women face restricted resources in areas besides finance, with climate change, in be dependent on Moreover, home-based inequalities exacer- particular, exacerbating existing inequalities in socioeconomic factors bate market-based gender inequality through women’s livelihoods and reducing their resil- the motherhood pay gap—a term that can refer ience (box 4.5). such as profession, to the difference in pay between mothers and earnings and income childless women, or to that between mothers As noted, girls and women of reproductive and fathers, rather than between all working age (15­49 years old) are more likely than stability or to legal men and women. The motherhood pay gap is boys and men of the same age to live in poor discrimination and usually bigger in developing countries, and in households (figure 4.16). This challenges the all countries it increases with the number of “headship definitions” approach to household gender norms children a woman has. The combination of composition for examining poverty profiles, in low earnings and dependants makes women which households with a male earner, a non- over-represented among poor people during income earner spouse and children are more their reproductive age: Women are 22 percent likely to have poor women. Children and other more likely than men to live in a poor house- dependants can be an important vulnerability hold between the ages of 25 and 34.59 factor for women in their reproductive health. For both genders, pooling resources and having According to the World Bank’s 2017 Global more adults working for pay in a household can Findex, of the 1.7 billion unbanked adults protect them from falling into poverty, as can in the world, 56 percent are women, while in education, especially for women.63 developing countries women are 9 percentage points more likely to be unbanked than men.60 For most people lifetime working conditions The Arab States and Sub-Saharan Africa have have a great impact on economic conditions the lowest percentage of women with an ac- and autonomy in older age. For women— count at a financial institution or with a mobile over-represented among older people—earlier money-service provider, but the percentage gender gaps in health, wages, productivity, is below 80 percent in all developing country labour participation, formal versus informal regions (figure 4.15). Women’s financial inde- work, remunerated versus nonremunerated pendence can be dependent on socioeconomic work, continuity in the labour market and the factors such as profession, earnings and income ability to own property and save are likely to

FIGURE 4.14

A large proportion of employed women believe that choosing work implies suffering for their children, while a large proportion of female homemakers feel that by staying home they are giving up a career or economic independence, 2010­2014

Employed women who agree that Homemakers who agree that a job is children suffer if women have a job the best way for a woman to be independent

Arab States Arab States

East Asia and the Pacific East Asia and the Pacific

Europe and Central Asia Europe and Central Asia Latin America and the Latin America and the Caribbean Caribbean South Asia South Asia

Sub-Saharan Africa Sub-Saharan Africa

0 20 40 60 80 0 20 40 60 80

Source: Human Development Report Office calculations based on data from wave 6 (2010­2014) of the World Values Survey.

162 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 4.15

The percentage of women with an account at a financial institution or with a mobile money-service provider is below 80 percent in all developing country regions in 2018

Women with access to an account with financial institution or a money service provider, ages 15 and older, simple average (percent)

0 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 South Asia Sub-Saharan Africa Developed East Asia and Europe and Latin America and Arab States the Pacific Central Asia the Caribbean

Source: Human Development Report Office based on data from the Global Findex database.

BOX 4.5 Climate change and gender inequality

Women tend to be responsible for procuring and pro- reduction and the mitigation of climate change effects Greater female viding food in households and are the primary workers and environmental degradation. participation in natural engaged in subsistence agriculture. They make up an resource management, average of 43 percent of the agricultural workforce in Greater female participation in natural resource productive agricultural developing countries.1 management, productive agricultural activities and nat- activities and natural ural disaster responses can enhance the effectiveness disaster responses Even so, they experience inequitable access to and sustainability of policies and projects. Closing the can enhance the land and agricultural inputs,2 which can affect their pro- gender gap in agricultural productivity would increase effectiveness and ductivity in the sector, generating a gap in comparison crop production by 7­19 percent in Ethiopia, Malawi, sustainability of with men’s productivity. In Ethiopia, Malawi, Rwanda, Rwanda, Tanzania and Uganda.4 policies and projects Tanzania and Uganda the gender gap in agricultural productivity ranges from 11 percent to 28 percent.3 The Climate change can affect women’s income, edu- difference is due to access to credit, ownership of land, cation, access to resources, access to technologies and use of fertilizers and seeds, and availability of labour. access to information.5 It is entangled with economic As in many other dimensions, gendered norms and and social consequences for women. Women in devel- traditions at the household level are behind the ineq- oping countries are highly vulnerable when they depend uitable allocations of production factors, thus limiting heavily on local natural resources for their livelihood. women’s agency, decisionmaking power and partici- Yet women are powerful agents of change. As key play- pation in the labour market. Furthermore, the gender ers in core productive sectors, they are well placed to agricultural gap hinders poverty reduction, inequality identify and adopt appropriate strategies to address climate change at the household and community levels.

1. FAO 2011. 2. UN Women, UNDP and UNEP 2018. 3. UN Women, UNDP and UNEP 2018. 4. UN Women, UNDP and UNEP 2018. 5. Brody, Demetriades and Esplen 2008.

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 163 FIGURE 4.16 expectations may lead to a perceived clash—a conflict, for example, of women’s rights with Girls and women of reproductive age are more traditional values—or reveal subconscious likely to live in poor households than boys and men biases. Still, even norms can be shifted towards gender equality. Poverty rate (percent) 25 The shift can be supported with a proactive stance, generating new regulations and policy 20 interventions that mainstream gender equality and women’s empowerment. This has been 15 Female happening but has not been enough to create long-term changes in stereotypes and tradition- 10 al gender roles. Entrenched inequalities persist Male due to discriminatory social norms and harmful behaviours and practices that undermine im- 5 plementation. Well intentioned interventions might fail or might have unintended conse- 0 quences if policymakers do not consider deeply 0­14 15­24 25­34 35­39 40­49 50­54 55­59 60+ rooted norms and practices. For instance, affirmative action or positive discrimination Age group (years) has sometimes overlooked or underplayed the effects of social norms on overall outcomes.67 Source: Munoz Boudet and others 2018. Efforts to promote women’s representation become later gender gaps in well-being.64 The in positions of leadership have yet to succeed, gap widens when pension systems are based and major prejudices persist about women’s on contributory schemes, and even more when ability to participate politically and function in they take the form of individual accounts.65 In high office. Representation quotas for women most developed countries women have equal sometimes do not deliver the envisaged trans- access to pensions. But in most developing formation and risk promoting tokenism by countries with data, there is a women’s pension introducing women’s presence while power gap (see Dashboard 2 in the statistical annex). remains entrenched in traditional hierarchies and privileges based on other identities such as The backlash against Empowering girls and class, race and ethnicity. changing gender women towards gender equality: A template to reduce Varied alternatives should be priorities in roles in households, horizontal inequalities light of multiple and complementary identities workplaces and rather than competing, conflicting ones—the Expanding opportunities for women and girls; multiple identities of an individual as a wom- politics affects entire promoting their economic, social and political an, a mother, a worker and a citizen should be societies influenced by participation; and improving their access to mutually supportive, not counterposed. So, shifting power relations social protection, employment and natural re- choices that enhance multiple freedoms are to sources make economies more productive. Such be prioritized over choices based on a singular investments reduce poverty and inequality and identity that diminish other freedoms. Any make societies more peaceful and resilient.66 All approach addressing gender inequality should that is well known. Social norms are shifting consider the multidimensional character of towards changed gendered roles in society. But gender and be sensitive to local social norms. while some conventional gender norms evolve Norm-aware interventions for women focus on in the private and public domains, their effects supporting them by providing solutions that are also facing a backlash from the conventional work around existing social norm constraints. power relations between men and women in today’s social hierarchies. Options to reduce gender inequalities—and many other horizontal ones—need to consider The backlash against changing gender roles how to directly target changes in unequal pow- in households, workplaces and politics affects er relationships among individuals within a entire societies influenced by shifting power community or to challenge deeply rooted roles. relations. The resistance to changes in gender

164 | HUMAN DEVELOPMENT REPORT 2019 This may include a combination of efforts in change—targeting both women and men is Analysis that goes education, raising awareness by providing new crucial. The importance of adequately engaging beyond averages information and changing incentives. men and boys in overcoming gender inequality requires more and or addressing their own gender-related vulnera- better data to keep An additional and important considera- bilities is acknowledged, but actions have a long pushing for gender tion to influence change in social norms and way to go. equality and to make traditional gender roles is for options to be other horizontal inclusive of both women and men, which may Finally, analysis that goes beyond averages re- inequalities visible hold also for other horizontal inequalities. quires more and better data to keep pushing for When choosing among alternatives—whether gender equality and to make other horizontal norm-aware or those pursuing social norm inequalities visible (box 4.6).

BOX 4.6

Better data are needed on gender inequalities

Gender data face challenges of quantity and quality. The disaggregating by sex and age, using outdated measure- first refers to not having enough data to depict women’s ments of time use and collecting data only on households current situation. For instance, among the Sustainable instead of individuals. Changes in these measurements Development Goals over 70 percent of data for 58 indi- can affect indicators such as the Multidimensional cators linked to gender equality and women’s empow- Poverty Index, calculated for households rather than indi- erment is missing.1 The second refers to current data viduals, so that complementary research may be needed that might not accurately reflect reality and that might to clarify the relationship between gender and poverty.2 underestimate women’s roles and contributions. More information is needed to get a better picture Some organizations perceive collecting and produc- of gender biases specific to a region, country or com- ing gender data as expensive in time and cost. Some munity, as with information on the impact of media and data collection methods are outdated and biased against social networks in reinforcing traditional norms and women because they follow gender social norms, such stereotypes.3 as interviewing only the male head of household, not

1. Human Development Report Office calculations based on data from UN Women (2017). 2. UNDP 2016. 3. Broockman and Kalla 2016; Paluck and others 2010.

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 165 SPOTLIGHT 4.1

Women’s unequal access to physical security—and thus to social and political empowerment

Violence against Violence against women is one of the cru- Decomposing these factors reveals inequality in women is one ellest forms of women’s disempowerment. the experience of violence, an insight that can of the cruellest Magnifying inequality, it happens throughout help in designing more focused interventions. the lifecycle, in different spaces—households, For instance, although violence can occur at all forms of women’s institutions, public spaces, politics and on- education levels, greater education attainment disempowerment line—in all societies, among all socioeconomic can protect women from partner violence. can be perpetuated groups and at all levels of education. And it Educated women have better access to informa- through social norms reflects the same social norms that legitimize tion and resources that help them identify an harassment and discrimination. abusive relationship and end it.3 Women’s eco- nomic empowerment through participation in More than a third of women—and more the workforce had mixed associations with the than two-thirds in some countries—have ex- risk of intimate partner violence,4 challenging the perienced physical or sexual violence inflicted notion that economic empowerment protects by an intimate partner or sexual violence in- women from gender-based violence. This finding flicted by a nonpartner (figure S4.1.1).1 Some highlights the heavy influence of social norms in 20 percent of women have experienced sexual women’s perceptions of their status in society in violence as children. Nearly a quarter of girls some cultures. In developing countries women ages 15­19 worldwide report having been vic- make up a large proportion of the informal sector tims of violence after turning 15.2 And violence workforce with low-paying jobs, a structure that is typically underestimated because of stigma, might perpetuate the myth of male superiority.5 denial, mistrust of authority and other barriers to women reporting an incident. Violence against women can be perpetuated through social norms. For example, female Intimate partner violence has been recurrent- genital mutilation and cutting remain wide- ly associated with such factors as age, wealth, spread. An estimated 200 million women and marital status, number of children, education girls living today have undergone female genital attainment and economic empowerment. mutilation, even though most men and women oppose the practice in many countries where FIGURE S4.1.1 it is performed.6 Violence against women and girls is sustained by individual behaviours and About a third of women ages 15 and older have beliefs as well as by social norms from the com- experienced physical or sexual violence inflicted munities and networks that can slow change. by an intimate partner, 2010 Violent actions, attitudes and behaviours are triggered by unequal power relations dictat- Percent Arab States East Asia ing gender roles at the household level. Some 35 and the examples are beliefs that a man has a right to Sub- 25 Pacific physically discipline a woman for an incorrect Saharan 15 behaviour, divorce is shameful or sex is a man’s 5 right in marriage. When women assert autonomy or aspire to South Asia Europe and exert power at any level—from the household Central Asia to the national government—they often face a backlash that can include violence (psycho- Latin America and the Caribbean logical, emotional, physical, sexual or econom- ic), whether as discrimination, harassment, Source: WHO 2013. assault or femicide. More than 85 percent of female members of European parliaments

166 | HUMAN DEVELOPMENT REPORT 2019 have experienced psychological violence, and or unwanted comments or purposely touching Women are 47 percent have received threats of death, or brushing up against someone. Women are harassed mostly rape, beating or kidnapping (figure S4.1.2).7 harassed mostly in public spaces, their workplac- in public spaces, Moreover, the only country in the world that es, their residences or their schools.10 their workplaces, has legally made political violence a separately their residences defined crime is Bolivia.8 Elsewhere, lacking Through social media and other online plat- or their schools laws, regulations and sanctions, women are left forms and applications, women are vulnerable unprotected from this type of violence. In 2016 to harassment and bullying in a new space—the the #NotTheCost campaign was launched to digital public space. Ensuring that this space raise awareness and stop violence against wom- is safe and empowering for women and girls is en in politics. The name alludes to how women a new challenge. Some 73 percent of women are told that harassment, threats, psychological online have been exposed to some type of cyber abuse and other forms of violence are “the cost” violence, and women are 27 times more likely of participating in politics.9 Traditional gender than men to be the victims of cyberviolence.11 norms play a role in such political violence. Besides the impact of violence against women and girls in other spaces, cyberviolence impedes Globally, there are some efforts to fight the their digital inclusion and keeps them from backlash. Political violence and sexual harass- enjoying digital dividends. Even though tech- ment and assault received attention in 2017 nology can connect and empower, it can also when American actress Alyssa Milano called for reinforce traditional gender roles and normalize women to come forward with their experienc- stereotypes reflecting a culture of misogyny and es. Some 1.7 million tweets using the hashtag marginalization. Security and harassment are #MeToo responded, and 85 countries had at among the top five barriers to women’s mobile least 1,000 #MeToo tweets. The movement gave phone ownership and use.12 Online harassment, visibility to the issue and propelled initiatives sexist attitudes and misogynistic remarks can to conduct more research on sexual harassment undermine women’s sense of legitimacy, compe- and assault, especially in the United States. Some tence and safety, making them mistrust technol- 81 percent of women and 43 percent of men in ogy and even opt out of its use. Besides hindering the United States reported experiencing some technological inclusion, violence against women form of sexual harassment or assault in their life- and girls in this space has a cumulative emotional time. The most common forms of sexual harass- and physical cost on them. ment are whistling, honking, saying disrespectful For each demographically “missing” woman, FIGURE S4.1.2 many more fail to get an education, job or po- litical responsibility they would have obtained Female members of European parliaments if they were men.13 Gender is a global factor in experience high rates of acts of political violence unequal human autonomy, physical security against women, 2018 and social, economic and political empower- ment. Women’s human development depends Reported incident on socioeconomic enabling factors such as Experienced this form of violence the ability to pursue a profession, to attain income stability and to achieve earnings com- Percent parable to men’s. Women’s empowerment in health, education, earning opportunities, and 39 23 political rights and participation can change social decisionmaking and development (fig- 25 21 22 ure S4.1.3). Women’s human development also requires positive gender norms and an absence 20 8 of gender discrimination, with laws preventing unequal treatment, harassment and violence Online 8 57 against women. Education, reproductive rights violence and political participation are key assets in all Threats Psychological Sexual Physical these areas, while the right to human security is fundamental. of physical harassment harassment violence

violence

Source: IPU 2019.

Chapter 4 Gender inequalities beyond averages—between social norms and power imbalances | 167 FIGURE S4.1.3 Traditional social norms encourage different forms of violence against women

Women’s Mental ealth empowerment in health health, education, Education earning opportunities, H and political rights and participation Psychological/ can change social emotional decisionmaking and development Labour Economic Forms of Sexual Physical violence against integrity

women

Financial Physical inclusion

Security

Source: Human Development Report Office based on UN General Assembly (2006).

Notes 8 Government of Bolivia 2012. 9 NDI 2019. 1 WHO 2013. 10 Kearl 2018. 2 UNICEF 2014a. 11 Broadband Commission for Digital Development Working 3 Flake 2005; Waites 1993. 4 Sardinha and Catalán 2018. Group on Broadband and Gender 2015; Messenger 2017. 5 Uthman, Lawoko and Moradi 2011. 12 GSMA Connected Women 2015. 6 UNICEF 2018a. 13 Duflo 2012. 7 IPU 2019. 14 Caprioli 2005. 15 Ouedraogo and Ouedraogo 2019. 16 Stone (2015) as cited in O’Reilly, Ó Súilleabháin and Paffenholz

(2015).

168 | HUMAN DEVELOPMENT REPORT 2019 Part III PART III.

This Report has taken us on a journey. It identifies the evolution of different inequalities in human development and examines the dynamic ways in which they limit human freedoms. It goes beyond averages to discover trends in the com- plete distribution of income and wealth. It also looks at gender inequality and delves into the factors holding back half of humanity. We are now almost at the end of the journey: What is to be done?

No single policy will suffice, nor will the same In this context, chapters 5 and 6 discuss two policies be appropriate for all countries. There key trends that could blunt the fight against are large and significant differences across inequalities in all countries. Understanding countries in history, institutions, incomes and these trends is essential because left on their administrative capabilities. Culture and social own, they will tend to increase inequalities in norms also matter, as the discussion on gender human development. inequality highlights (chapter 4). Moreover, inequalities in human development are linked. The first trend relates to climate change It is unlikely that households deprived of en- (chapter 5). Much has been written about this hanced capabilities, let alone basic capabilities, topic—the focus here is on its interactions will be at the top of the income scale. It is also with inequality. In a nutshell, increased vola- unlikely that women who suffer discrimina- tility in the world’s climate and rising average tion in access to education and jobs will be temperatures are likely to translate into more among the very rich. As parts I and II of the floods, droughts, hurricanes and related phe- Report highlight, inequalities along the vari- nomena. The chapter also documents that the ous dimensions interact and generate feedback impacts will not be distributed evenly within loops. This makes fighting inequality a daunt- or across countries. Some countries will suffer ing task. How can countries tackle the myriad more than others, and within countries some policies and institutions that stand behind all regions more than others. In parallel some the dimensions of inequality? Where should households will suffer more. they begin? Should they focus on capabilities, on income or on gender? What policies are All this will tend to increase inequalities—and more effective when and where? may even reduce the effectiveness of policies. For instance, countries might make progress against Part III of the Report, dealing with poli- income inequality through more progressive cies, addresses these questions. It proposes a taxation, but that progress could be undone by framework to support countries in tailoring households’ greater exposure to climate risks. a response to inequalities in human develop- Climate change may thus require strengthening ment to their specific circumstances, taking old tools and introducing new ones—from into account their political constraints and drought-resistant crops to new insurance ap- administrative capabilities. The aim is to help proaches. The chapter also considers interactions them craft their own responses—rather than in the other direction—how inequalities can offer a single recipe applicable to all. complicate responses to climate change. Indeed, it is far harder to rally around common respons- In beginning to think about what can be es in societies that are more polarized. done, it is essential to consider time and place. Addressing inequalities in human develop- Chapter 6 focuses on technological change. ment in the 21st century is not the same as it It has always been with us, but since the was before. Policymakers interested in fighting Industrial Revolution it has affected the distri- inequalities will take into account today’s com- bution of income and capabilities in far more plexities and challenges. Certainly, much is to profound and long-lasting ways, in part because be learned from what policies have worked and economic prosperity—and increasingly the na- what policies have failed in the past, but those ture of sustainability—is tied to the direction lessons have to be relevant to here and now. of technological change. Recent trends asso- ciated with robotics and artificial intelligence

PART III Beyond today | 171 pose new challenges but also create opportu- in human development. It does not provide a nities. The relative demands for skills and tasks recipe for all countries, since policies are coun- will change, as will the locations of economic try specific. Instead it presents a framework activity, given the dramatic increases in econ- to think about policies to address pernicious omies of scale and the dramatic reductions in inequalities in human development. It shows transportation costs. This will induce offshoring that the range of available policies is large and of some tasks and the disappearance of others. that it is feasible to address some of the under- Enhanced capabilities will be critical for people lying drivers of the inequalities in capabilities. to navigate the upheavals that technology can The central message is unequivocal. The trends bring. Technology itself can help in this regard, documented in parts I and II are not inevitable if policies are chosen such that technology helps —they result from policies and institutions, to reinstate demand for labour. and much can be done both nationally and in- ternationally to reform them. We have a choice. With these two chapters as background, chap- And we must act now. ter 7 deals with policies to combat inequalities

172 | HUMAN DEVELOPMENT REPORT 2019 Chapter 5

Climate change and inequalities in the Anthropocene 5.

Climate change and inequalities in the Anthropocene

The climate is in crisis. The effects are already unspooling in the form of melting ice sheets and, as is likely, record heatwaves and superstorms. Without bold collective action, these will only worsen over time, joined by a suite of other ca- lamities, from depressed crop yields to rising sea levels to potential conflict. As recognized in the Sustainable Development Goals and the Paris Climate Agreement, climate change is a global challenge.

But it will not affect everyone equally — not But climate change and inequality, and the in the same way, not at the same time, not at interaction of the two, are choices, not inevita- the same magnitude. Poorer countries and bilities. Even though the window for decisive poorer people will be hit earliest and hardest. and bold action on climate is shrinking, there Some countries could quite literally disappear. is still time to make different choices. Of all climate change’s disequalizing effects, perhaps none is greater than that on future This chapter suggests that by redressing ine- generations, which will shoulder the burden qualities, action on climate could also be made of previous generations’ fossil fuel-dependent easier and faster. To see why, consider two of development pathways. the multiple possible channels at play.1 The first relates to how individual consumption Inequality runs the gamut of climate change, decisions add up to total emissions (box 5.1).2 from emissions and impacts to resilience and The second, which is the focus of this chapter policy. Climate change is a recipe for more and likely more consequential, relates to how inequality in a world that already has plenty. inequality interacts with technological change

BOX 5.1 Household income, inequality and greenhouse gas emissions

Higher household incomes are associated with higher does,3 the increase in emissions by poor people would emissions, but the impact of inequality on aggregate be higher than the corresponding decrease in con- emissions depends on how quickly emissions increase sumption by rich people, leading to a net increase in as income rises.1 There is a wide range of empirical es- emissions. And one would expect to see the opposite timates for this relationship, showing that, on balance, in developing countries, with reductions in inequality emissions increase more slowly than income in most lowering emissions.4 However, the scale of the impact developed and middle-income countries but at the same of inequality through this channel tends to be small, rate (or even a little faster) in lower income countries.2 certainly when compared with other determinants of changes in emissions, such as technological change Taking this channel alone into account would imply and policies.5 that income inequality should be associated with low- er emissions in developed countries. To see how, con- Perhaps more important, the interplay of these sider the impact of transferring income from the rich consumption patterns within and across countries—al- to the poor in a developed country. Even though rich though trending towards lower emissions overall—ap- people emit more, given that the rate at which emis- pears unlikely to substantially reduce global aggregate sions increase is slower than the rate at which income emissions.6

1. It also depends on how inequality interacts with rising income. For a comprehensive description of the different possibilities, see Ravallion, Heil and Jalan (2000). 2. See, for instance, Liddle (2015). For a detailed estimate for the Philippines, see Seriño and Klasen (2015). 3. When this relationship is measured in terms of how much a percentage change in income is reflected in a corresponding percentage change in emissions—in technical terms, an elasticity—this implies an elasticity of less than 1. 4. More precisely, this would happen if the elasticity were greater than 1. For some empirical support of the hypothesis of this differential impact of inequality in emissions in developed and developing countries, see Grunewald and others (2017). 5. To illustrate, Sager (2017) calculated consumption-based carbon emissions Engel curves (showing the relationship between household income and average carbon dioxide emissions) for the United States for several years between 1996 and 2009. In a scenario where income is redistributed to perfect equality (a dramatic and extreme case), average carbon dioxide emissions in 2009 would have increased 2.3 percent, from the actual 33.9 tonnes per household to 34.7 tonnes. In contrast, had there been no technological change and assuming the same consumption composition between 1996 and 2009, average emissions would have increased 70 percent, to 57.9 tonnes. 6. Caron and Fally 2018.

Chapter 5 Climate change and inequalities in the Anthropocene | 175 Higher inequality tends and policy formation. There is some evidence Where emissions are being decoupled from to make collective that high inequality hinders the diffusion of economic growth—a hopeful sign that is action—key both new environmentally friendly technology.3 directionally right but not yet at scale, despite within and across Inequality can influence the relative power accelerating over the past two decades—this is countries to curb of interests arguing for and against curbing related to countries having “underlying policy climate change— emissions. Emissions would be expected to frameworks more supportive of renewable more difficult be higher when income is concentrated at the energy and climate change mitigation efforts,“9 top and when the resulting concentration of which shows the feasibility of a break from economic power coincides with the interests unsustainable development models that have of groups that oppose action on climate.4 endured for centuries.10 Still, countries with More generally, higher inequality tends to higher human development generally emit make collective action—key both within and more carbon per person and have higher per across countries to curb climate change—more capita ecological footprints (figure 5.1).11 Richer difficult.5 Information is critical for collective countries and communities may put a premium action, but the ability of different interest on local concerns, such as water and air quality, groups to communicate tends to be lower when but they tend not to experience locally the full inequality is high,6 with the concentration of extent of their impacts on the environment, income potentially leading to the suppression which are driven more by their income than or propagation of information in order to by “green” self-identities and associated behav- serve a particular interest.7 Other interacting iours.12 Instead, they often shift a significant mechanisms relate to how inequality shapes portion of the environmental impacts of their perceptions of fairness (with implications for consumption preferences to less-visible coun- compliance and enforcement).8 tries and communities elsewhere, including to

FIGURE 5.1

Per capita ecological footprints increase with human development

Ecological footprint, 2016 Medium human High human Very high human (global hectares per person) development development development

8

6

4

2

Biocapacity per person,

0 world average (1.7 global hectares)

0.4 0.5 0.6 0.7 0.8 0.9 1

Note: Covers 175 countries in the Global Ecological Footprint Network database (www.footprintnetwork.org/resources/data/; accessed 17 July 2018). As used here, the ecological footprint is a per capita measure of how much area of biologically productive land and water a country requires, domestically and abroad, to produce all the resources it consumes and to absorb the waste it generates. Each bubble represents a country, and the size of the bubble is proportional to the country’s population. Source: Cumming and von Cramon-Taubadel 2018.

176 | HUMAN DEVELOPMENT REPORT 2019 those along global supply chains.13 In the case was an existential threat to future generations, Some evidence of climate change, they also shift the impacts to exacerbating intergenerational economic suggests that future generations, which are even less visible. inequality, but also that it would increase in- development on its come inequality across and within countries.15 own is unlikely to Environmental burden shifting happens Recent research has confirmed, and made more offer protection from not just for greenhouse gas emissions but also precise, how disequalizing climate change can the negative impacts across many environmental domains.14 Thus, be: Income inequality across countries may al- of climate change this chapter goes beyond climate to examine ready be about 25 percent higher than it could inequalities and burden shifting in other im- have been without climate change.16 portant areas, such as waste generation, meat consumption and water use. Environmental This chapter takes that analysis further, show- burden shifting is linked to gradients in eco- ing how climate change exacerbates inequalities nomic and political power. Attempts to redress in other dimensions of human development and these power differences and how they manifest how inequality is also relevant to building climate environmentally are likely to be ever more rel- and disaster resilience. Some evidence suggests evant as humanity enters what has been called that “development on its own” may not offer the Anthropocene (box 5.2). protection from the negative impacts of climate change.17 New, broadly shared approaches to The 2007/2008 Human Development resilience may needed. Echoing a central theme Report showed not only how climate change

BOX 5.2 From Holocene to Anthropocene: Power—and who wields it—at the brink of a new era

The environment has a profound impact on people’s capabilities and on their human activities ranges from introducing invasive species to the plastics ep- ability to convert capabilities into achievements—and thus on human de- idemics in the oceans to fisheries stress and collapse to fossil fuel emissions velopment.1 Conversely, human activity affects the natural world, shaping and climate change.6 These and other activities have not just destabilized environmental processes and patterns at a global scale. Arguably, human- ecosystems but have also transformed planetary biogeochemical process- kind today is not just witnessing but also causing the sixth mass species es.7 Humanity is thought to have already breached at least four of nine plan- extinction in the Earth’s history.2 While the stratigraphy community has etary boundaries, the safe operating limits for different components of the yet to formally declare a new epoch (meaning that humanity is still in the Earth system seen as critical to maintaining a stable Holocene-like state.8 Holocene), the unfolding changes to the environment are so dramatic, and Two of these—climate change and biosphere integrity—are considered so heavily influenced by humans, that the expression Anthropocene has en- core boundaries, meaning they have the potential on their own to push the tered current use.3 Earth into a new state.9 Humans have exceeded the safe operating space for both; the risk of crossing a critical threshold, destabilizing the Earth system The Anthropocene portends a worrying mix of power, fragility and uncer- and exiting the Holocene is no longer assuredly low.10 tainty. The end of the last glacial period and the beginning of the Holocene more than 10,000 years ago ushered in a stable climate regime—a climatic This is the Anthropocene: human power at scale, without illusions of cradle for humans—with conditions favourable for permanent agriculture control and without fully grasping or heeding the consequences. Through and the dawn of civilizations. Rising populations, wealth and technological unmitigated greenhouse gas emissions and other actions, humans are pull- know-how have translated into greater, seemingly unbridled power, includ- ing themselves out of the relative stability of the current geological epoch ing over the environment. Yet fragilities have always been evident. Crops into the uncertainty of a new one. The Anthropocene is essentially a leap are susceptible to pests and bad weather. Infectious diseases have sprung into the unknown. Making a choice for sustainable human development, from (and through) domesticated animals and elsewhere.4 The interplay based on a country’s unique set of circumstances, is necessary. But it is not among humans, geography and the environment has been central to the easy—and it is made all the more difficult when persistently high inequality, way civilizations have come and gone.5 in its many forms, with its corrosive effects, implies that both people and planet lose. Choices rooted in inclusion and sustainability can turn the dam- Fast forward to today, and the intertwining of power, fragility and un- aging historical relationship between development and ecological footprints certainty has not changed. The differences are in the scale and the stakes. on its head—breaking humanity free from old development approaches that Humans have far more power to affect the environment, including at the simply will not work as it enters the brave new world of the Anthropocene. planetary level, but no greater control. The list of negative feedback from

1. Robeyns 2005. 2. Barnosky and others 2011; Ceballos, Ehrlich and Dirzo 2017; Ceballos, García and Ehrlich 2010; Ceballos and others 2015; Dirzo and others 2014; McCallum 2015; Pimm and others 2014; Wake and Vredenburg 2008. 3. Scott (2017) attributes to Paul Crutzen the introduction of the term and the proposal to date the start of this era to the late 18th century, coinciding with the invention of the steam engine, which unleashed the Industrial Revolution (even though Scott himself proposes the concept of a “thin Anthropocene,” which could be dated as far back as the hominid use of fire). In May 2019 the 34-member Anthropocene Working Group voted to designate the Anthropocene as a new geological epoch. The panel plans to submit a formal proposal to the International Commission on Stratigraphy, which oversees the official geological time chart. 4. Dobson and Carper 1996; McNeill 1976; Morand, McIntyre and Baylis 2014; Wolfe, Dunavan and Diamond 2007. 5. Crosby 1986; Diamond 1997, 2005. 6. Choy and others 2019; Early 2016; Millennium Ecosystem Assessment 2005; Seebens and others 2015; US NOAA 2018. 7. Campbell and others 2017; Steffen and others 2015. 8. Steffen and others 2015. 9. Steffen and others 2015. 10. Steffen and others 2015.

Chapter 5 Climate change and inequalities in the Anthropocene | 177 The challenge is to of this Report, this chapter finds convergence in rights. Climate policy can create virtuous feed- ensure that climate basic capabilities to cope with climate change and back loops in which emissions decline from resilience does not divergence in enhanced ones. Countries are con- direct effects (such as a carbon price) and from become the reserve verging—even though large disparities persist— indirect effects (such as lower inequality, which of only a select group in their preparedness to “normal” shocks, ones may facilitate even bolder climate policies). expected at a certain frequency and magnitude This chapter, as well as chapter 7, tees up some of countries and based on historical trends—a basic resilience of these key issues. communities that capability. Climate change impacts, however, do can most afford it not always conform to historical trends, with How climate change and more “surprises” than in the past.18 Shocks take inequalities in human on a new, unanticipated character. Building pre- development are intertwined paredness—which relies less on the experienced past and more on how science and technology, This section starts by expanding beyond ine- including advanced weather prediction systems, qualities in carbon emissions between coun- can help prepare for an uncertain future—is be- tries to inequalities within them, adding to the coming an enhanced capability in which gaps are more familiar story on how climate change will emerging. The challenge is to ensure that climate harm—and has already impacted—different resilience does not become the reserve of only a dimensions of human development. Finally, it select group of countries and communities that takes an illustrative look at climate resilience, can most afford it, thereby further exacerbating framing it as an enhanced capability that risks the inequality impacts of the climate crisis. divergence.

The urgency for action to combat climate From inequality in emissions change, including by fully implementing the to inequality in impact: Two Paris Agreement under the United Nations dimensions of climate injustice Framework Convention on Climate Change, cannot be overemphasized. So why isn’t more Carbon dioxide is not the most potent anthro- being done? True, there is renewed interest in pogenic greenhouse gas, but it is the most wide- many countries around the world in carbon spread, driven overwhelmingly by fossil fuel pricing, but to take just a simple illustration, combustion (87 percent of total carbon diox- only 5 percent of emissions are covered by a ide emissions over 2008­2017) for electricity, carbon price high enough to achieve the goals transportation and other uses.21 It is widespread of the Paris Agreement.19 Some even argue that because carbon emissions are deeply embedded carbon pricing will not be enough and that in current patterns of production and con- instead of relying on market signals, more fun- sumption, and powerful fossil fuel interests damental transformations of economies and have generally tried to keep it that way.22 societies will be needed.20 The various mech- anisms through which inequality influences The richest countries account for the lion’s technology diffusion and policies, reviewed share of cumulative carbon dioxide emissions briefly above, speak to the complex interplay (figure 5.2); they are still among the top pol- between climate change and inequality and luters on a per capita basis and in terms of even how action on climate can be hamstrung, aggregate country emissions today.23 These in- as in the case of the Mouvement des gilets jaunes equalities in cumulative emissions are central to (yellow vests movement), perhaps an instance the global conversation on climate, particularly when people felt as though they were being left for climate justice, burden sharing and differen- behind. tiated responsibilities.24

Addressing inequality and the climate crisis The same pattern of inequality plays out together can move countries towards inclusive within countries, with households at the top and sustainable human development. For in- of the income distribution responsible for stance, when carbon pricing is part of a broader more carbon emissions per person than those set of social policy packages, it is possible to at the bottom. While there is no direct way of address inequality and climate together while allocating emissions to individuals, estimates facilitating the realization of people’s human

178 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 5.2 on all continents, a third of them in emerging economies (figure 5.3).25 Today’s developed countries are responsible for the vast majority of cumulative carbon dioxide Inequality in global carbon dioxide equivalent emissions emissions between individuals has decreased, but within-country inequality is steadily rising Share of cumulative carbon dioxide and approaching the share of between-country emissions, 1750­2014 (percent) inequality in the global dispersion of carbon di- oxide equivalent emissions (figure 5.4). In 1998 66 a third of inequality in global carbon dioxide equivalent emissions was due to within-country 26 inequality; by 2013 half was.

7 Turning from emissions to impact, unmitigat- Part of the reason 1 ed climate change drives inequalities in human climate change development through two main mechanisms: and disasters are Low Medium High Very high differential exposure and vulnerability.26 Debate disequalizing is that continues on the relative importance of each. inequality exists Human development group This chapter takes the view that both matter. in the first place; Differential exposure is real: Climate change will they run along, Source: Human Development Report Office calculations based on Ritchie and hit the tropics harder first, and many developing exploit and deepen Roser (2018). countries are in the tropics.27 At the same time, existing social and developing countries and poor and vulnerable economic fault lines based on plausible approximations suggest that communities have fewer capacities to adapt to global carbon dioxide equivalent emissions are climate change and severe weather events than highly concentrated: The top 10 percent of do their richer counterparts. Part of the reason emitters account for 45 percent of global emis- climate change and disasters are disequalizing sions, while the bottom 50 percent account for is that inequality exists in the first place; they 13 percent. The top 10 percent of emitters live run along, exploit and deepen existing social and economic fault lines. These fault lines were dramatically laid bare when Hurricane Katrina

FIGURE 5.3

Of the top 10 percent of global emitters of carbon dioxide equivalent emissions, 40 percent are in North America, and 19 percent are in the

Top 10 percent of emitters: Middle 40 percent of emitters: Bottom 50 percent of emitters: 45 percent of world emissions 42 percent of world emissions 45 percent of world emissions

North Other Asia Other rich Russian Fed./C. Asia South Africa China Latin America Middle East & N. Africa America 5% 6% 7% 3% 35% 9% 4% 40% South Africa Russian Fed./C. Asia European Union Other Asia 2% Other rich 7% 18% 23%

4% India Russian Fed./C. Asia

China Asia 36% 1% 10% 8% South Africa North China 10% America 16% Middle East & 19% 7% N. Africa 5% India Middle East Latin 1% & N. Africa America 5% 7%

6% India 5%

Source: Chancel and Piketty 2015.

Chapter 5 Climate change and inequalities in the Anthropocene | 179 FIGURE 5.4 catastrophic events, whose impacts are gener- ally not systematically captured in many mod- Within-country inequality in carbon dioxide els.32 As Martin Weitzman once claimed, “All equivalent emissions is now as important as damage functions are made up—especially between-country inequality in driving the global for extreme situations,“33 yet many of the dispersion of carbon dioxide equivalent emissions most widely used economic models of climate change rely on “smooth” damage functions that Level of may not fully account for the possibility of cat- inequality astrophic events.34 Over the past few years research has at- Between tempted to incorporate tipping points into integrated assessment models. The findings of 0.4 such work have generally strengthened the case for a greater precautionary approach to the 0.3 climate.35 The bottom line is that estimates of Within economic effects of future climate change give some broad directional agreement, and while 0.2 uncertainties abound, the costs of potential catastrophic events coupled with the pace at 1998 2003 2008 2013 which the scientific evidence is accumulating on the scale of damages reinforce arguments for Note: In 2008 the within-country component of the Theil index, which measures early and forceful action.36 For example, there is dispersion in the distribution of a variable that can be perfectly decomposed into strong evidence that the economic damages of between-group and within-group components, was 0.35 and the between- extreme natural hazards have increased globally country component was 0.40—that is, between-country inequality accounted for over the past several decades (figure 5.5). Some 53 percent of total inequality new modelling approaches that attempt to Source: Chancel and Piketty 2015. incorporate risk and uncertainty point to large costs associated with delays in taking forceful The complexities of struck New Orleans in 2005. A more recent action on mitigation, with these costs com- the climate system example is the tragic loss of life and devastation pounding over time (a five-year delay implies a wrought by Hurricane Dorian in the Bahamas cost of $24 trillion, and a 10-year delay implies make significant in 2019. Dorian was the strongest hurricane to a cost of $100 trillion).37 tipping points and strike the country since recordkeeping began in thresholds possible 1851.28 The communities hardest hit included The negative impacts of climate change extend shantytowns populated mostly by poor Haitian to health and education. Between 2030 and immigrants, some of whom had fled the devas- 2050 climate change is expected to cause some tating 2010 earthquake in their home country.29 250,000 additional deaths a year from mal- nutrition, malaria, diarrhoea and heat stress.38 The global economic impacts of climate Hundreds of millions more people could be ex- change have been modelled many times, posed to deadly heat by 2050, and the geograph- producing a range of estimates, each with its ic range for disease vectors—such as mosquito own range of possible outcomes. From these species that transmit malaria or dengue—will estimates, two key points emerge: First, climate likely shift and could expand.39 Lower agricultur- change will reduce global GDP, especially in al yields due to temperature changes can affect the long run, and second, negative economic food security, and food insecurity can worsen impacts are generally worse at higher tempera- nutrition. Good nutrition is essential for healthy ture thresholds.30 Moving beyond these general pregnancies and for early childhood survival trends to more precise estimates is challenging. and development, which can reduce inequalities The exact magnitude of the economic effects in human development (chapter 2). It is also of climate change is highly uncertain, and it important for school attendance, performance varies by geography and many other variables. and achievement.40 Malnutrition, by contrast, Nonlinearities complicate matters: Each addi- complicates the course of other illnesses, such as tional unit of change in the climate is unlikely tuberculosis and AIDS. to yield the same incremental impact over time.31 The complexities of the climate system make significant tipping points and thresh- olds possible—for example, the possibility for

180 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 5.5

Economic damages from extreme natural hazards have been increasing

Economic damages (US$ billions)

5

0

1960 1970 1980 1990 2000 2010

Note: Data are the yearly distribution of economic damages associated with 10,901 disasters that occurred worldwide between 1960 and 2015. Partial boxplots are coloured by decade. The lower hinge is the median, the middle line is the 75th percentile, the upper hinge is the 90th percentile and the upper whisker is the 99th percentiles. The red dashed line tracks the time progression of the 99th percentile. Source: Coronese and others 2019.

By the end of the 21st century, unmitigat- human development. An analysis of the last 40 Climate change is likely ed climate change could cause an additional years further substantiates the general pattern: to have already been 1.4 billion drought exposure events a year and Temperature-related shocks hit poorer coun- a force for increasing 2 billion more extreme rainfall exposure events tries harder than richer countries.44 In fact, even income inequality a year, inevitably increasing flood risk.41 The though some richer countries may have enjoyed between and within impact of these shocks on livelihoods can im- small benefits on average from temperature in- countries. It is likewise pede human development, influencing factors creases, the evidence suggests that all countries driving inequality in ranging from the availability of food to the will eventually be negatively affected by climate other dimensions of ability to pay for health care and schooling. change.45 human development Out-of-pocket health spending pushes almost 100 million people into extreme poverty each For health, the evidence from large-scale em- year.42 Even where schooling is free, livelihood pirical studies on climate impacts shows:46 shocks can siphon children from school into in- · In all regions the proportion of people vul- come-generating activities. These interrelated, overlapping shocks, when combined, will also nerable to heat exposure is rising. The elderly have consequences for mental health, which account for a significant portion of that vul- now appears in some countries’ national health nerability (see spotlight 5.2 at the end of the strategies for adapting to climate change.43 chapter). Heat stress, cardiovascular disease and renal disease are among the many causes Climate change is likely to have already been of heat-related illness and death.47 In 2017, a force for increasing income inequality be- 153 billion labour hours were lost because tween and within countries (see spotlight 5.1 of heat, an increase of more than 62 billion at the end of the chapter), as noted in the hours since 2000. opening of this chapter. Climate change is like- · Global vectorial capacity48 for the transmis- wise driving inequality in other dimensions of sion of dengue fever virus continues to rise, reaching a record high in 2016. In other

Chapter 5 Climate change and inequalities in the Anthropocene | 181 Climate change’s words, conditions are becoming more fa- misconstruing ability to pay for willingness to biophysical and vourable for transmission of dengue. pay, thereby systematically undervaluing those

• In the highlands of Sub-Saharan Africa, communities’ needs and desires.57

social dimensions are malaria vectorial capacity has increased working in the same 27.6 percent since the 1950 baseline. Consider the impact of climate change on direction: towards · In the Baltic region, changes in sea surface crop yields. Without improved crop varieties, worsening inequality temperatures have steadily increased suitabil- climate change will cause significant declines in ity for cholera outbreaks. average crop yields over the course of the 21st Since poor countries—and poor and vulner- century in many regions. The largest declines able people within countries—are dispropor- will occur where food insecurity is already a tionately burdened by these health conditions, threat.58 Climate change­related inequality is climate change has already put pressure towards partly a biophysical phenomenon of differen- greater health inequalities, both within and be- tial exposure. In regions where natural climate tween countries.49 variability is lower—such as the tropics, where In many developing countries exposure to many developing countries are found—climate floods, droughts and hurricanes in utero and signals will emerge from the “noise” more during early life impair later education and quickly and easily in the tropics.59 Recent mod- cognitive outcomes. In Southeast Asia higher elling shows that poorer countries will gener- than average temperatures during the prenatal ally experience weather-related changes before period and early life are associated with fewer richer countries. Regional heat extremes, for years of schooling, perhaps because heat has example, are expected to change noticeably in a negative impact on education attainment Africa, large parts of India and most of South where local climates are historically warm America after 1.5°C of warming, but mid-lat- and wet.50 In some developed countries there itude regions will not see such changes until is also evidence that prenatal heat exposure global temperatures increase by about 3°C.60 increases the risk of maternal hospitalization and of hospital readmission in the first year of Climate-induced inequality is also a social life for newborns, with differentiated impacts phenomenon. Vulnerable people will suffer across segments of the population that tend more because, for instance, with less irrigation, to increase maternal health gaps.51 These and yields are more weather dependent. With few- other potential impacts of climate change on er and less robust cereal market stabilization education outcomes have clear inequality im- mechanisms, livelihoods can be volatile. With plications, both within and across generations. less income and wealth, poor people are less As noted above, climate impacts are often able to absorb spikes in food prices. With dis- framed as the interaction of exposure and vul- criminatory laws, marginalized groups are bur- nerability.52 Exposure can be driven by vulner- dened with compounding insecurities. Climate ability, as vulnerable groups are driven to less change is expected to exacerbate these and secure, more disaster-prone locations, especially other vulnerabilities, its biophysical and social in urban areas.53 Such vulnerability-driven ex- dimensions working in the same direction: posure is widespread. The location or operation towards worsening inequality.61 of polluting factories and expressways, waste management54 and landfills, gazetted parks and Recent modelling has started to capture the conservation areas55, and even airports56 and interaction between biophysical and social other transportation hubs (and their expan- aspects through the spatial correlation of coun- sion) in or near vulnerable communities rests tries’ cereal productivity and gains from trade. on decisions that can take advantage of those Climate change, instead of affecting countries’ communities’ relative lack of power—either cereal yields uniquely or independently, will explicitly or implicitly. For example, cost-ben- cause regional changes that affect countries’ efit analyses for policy decisions—analyses that yields more similarly the closer countries are to purport to be objective, impartial or efficient— one another. So, developing countries will take can, among other potential pitfalls, implicitly a direct hit from climate change as cereal yields take advantage of vulnerable communities by decline and an additional hit when neighbour- ing countries also experience a decline. The decline in productivity across neighbouring trade networks reduces the gains from trade,

182 | HUMAN DEVELOPMENT REPORT 2019 which could worsen income inequality among change overwhelm response capacities, as typi- Countries have already countries by an additional 20 percent over the cally conceived, across many—perhaps all—lev- started adopting course of the 21st century.62 els of human development? For countries where tools, implementing climate change is an existential threat, the answer policies and making Feedback mechanisms have long been im- is a resounding yes. For others, if exposure ulti- investments that build portant in climate science, especially in terms of mately matters much more than vulnerability, resilience to climate biophysical systems. Economic feedback mecha- climate change may not be something that coun- change and other kinds nisms, such as knock-on trade effects, are coming tries can necessarily grow or “develop” out of. of shocks, precisely increasingly into view. Another is the impact of because old ways climate-induced GDP declines on carbon emis- Countries have already started adopting of doing things are sions. Climate-driven decreases in GDP may in tools, implementing policies and making invest- insufficient to the task turn decrease energy use and carbon emissions ments that build resilience to climate change over the course of the 21st century. In some and other kinds of shocks, precisely because scenarios fossil fuel emissions drop 13 percent, old ways of doing things are insufficient to the enough to offset positive carbon emission feed- task.67 They are charting different development back mechanisms from natural systems.63 paths that try to respond to the sobering, unfolding reality of climate change. Data and Here again recent empirical analysis com- technology, ranging from satellite imagery to plements income inequality projections. One drought-tolerant seeds, are seen as important study using longitudinal data from more than parts of forward-looking climate adaptation.68 11,000 districts in 37 countries suggests that So are fiscal rules that help protect economies since 2000, warming has made tropical coun- from unexpected climate shocks.69 Plus, build- tries at least 5 percent poorer than they other- ing resilience is a good economic investment. wise would be.64 The study also sheds light on The Global Commission on Adaptation found the importance of exposure and vulnerability that every $1 invested in adaptation could re- as mechanisms for climate-related inequalities: sult in benefits worth $2­$10.70 Disparities in the economic impacts of warm- ing are driven more by differences in exposure So, empirical analyses that emphasize ex- than differences in underlying vulnerability. In posure-driven pathways need not undermine other words the negative impacts of warming the rationale for resilience. On the contrary, cut similarly across communities of all levels such studies provide important historical of development. Richer ones are not insulated lessons for why conscious efforts to build from warming because they are rich, and poorer resilience matter—and matter urgently. From ones are not uniquely vulnerable because they a forward-looking inequality perspective the are poor. Part of the challenge is that exposure challenge is to ensure that climate resilience to damaging temperatures is much more com- is a broadly shared capability and a collective mon in poor regions. investment in human development rather than a capability that is the reserve of only a select That study’s findings, which imply a primacy group of countries and communities that can of exposure, correspond to those of another most afford it, thereby opening a new area of recent study on climate’s impacts on education divergence in the face of a global climate crisis. across 29 countries, mostly in the tropics. It found that the level of education of the head of As some analysts have noted, some impacts household did not buffer households from the of climate change may be smaller than the long-term impacts of adverse climate events.65 impacts of demographic change and economic In fact, children from more educated house- growth.71 Poverty projections at certain levels holds suffered greater education penalties, of warming similarly depend at least as much with hot temperatures having a levelling effect on development scenarios as on warming it- on education attainment. On the other hand, self.72 The 2011 Human Development Report a recent study using global data spanning four probed the ways various environmental and decades found the opposite: that richer coun- inequality scenarios might affect human de- tries are more insulated than poorer countries velopment across low, medium, high and very from the effects of temperature increases.66 high human development countries.73

Thus, the debate continues around an unset- A world of greater inequality is one possible tled, and unsettling, question: Might climate future, depending on the choices societies

Chapter 5 Climate change and inequalities in the Anthropocene | 183 ultimately make. Although unmitigated climate FIGURE 5.6 change will continue to narrow those choices over time—and indeed some climate change is Human development crises are more frequent and already baked in, owing to legacy emissions— deeper in developing countries much can still be changed. Carbon dioxide and other greenhouse gas emissions are the Average reduction in Human Developed product of human choices mediated largely by Development Index (HDI) value biophysical processes as well as by economic (percent of HDI in previous year) and social systems.74 Development paths that prioritize resilience and inclusion can be cho- Developing sen, too. The disproportionate impacts on poor countries—and poor and vulnerable people 0.5 within countries—largely reflect and are likely driven at least in part by structural inequalities. 1.2 If such inequalities—across income, wealth, health, education and other elements of human 13.5 8.2 development—are in no small part the result of social choices, as this Report argues, the course Frequency of reduction in HDI (percent) of climate change and the way it ultimately affects inequality have a lot of choice built in. Source: Human Development Report Office calculations for countries with There still is time to choose differently. annual data for 1980­2017.

The effects of shocks Differentiated paths in the ability ability to move and more resources with which do not appear to be to adapt to climate change: to recover. People in low human development Convergence in basic, divergence in countries are 10 times more likely than people randomly distributed; enhanced capabilities yet again? in very high human development countries to instead, they seem die due to natural hazards leading to disasters. to do more harm to This section considers asymmetries in capabili- And the relative cost (as a percentage of GDP) ties relevant to withstanding disasters linked to of disasters is about four times lower in very the more vulnerable natural hazards. The effects of shocks (linked high human development countries than in not only to disasters but also to other causes other countries (figure 5.7). These results are ranging from conflict to terms-of-trade crises) merely suggestive and should be seen in the do not appear to be randomly distributed context of broader trends in the global reduc- across different groups; instead, they seem to tion in causalities linked to natural hazards and do more harm to the more vulnerable. Over accelerating increases in the economic damag- 1980­2017 developing countries recorded es—with asymmetric impacts across climate re- a higher frequency of crises in human devel- gions depending on the nature of the hazard.76 opment, measured as a yearly reduction in Human Development Index (HDI) value, than Developing countries tend to have fewer developed countries did, and the impact of resources to prevent and respond to disasters these reductions was more severe. The average linked to natural hazards.77 The support and reduction in HDI value when facing a crisis was enforcement of building codes, the construc- 0.5 percent for developed countries but 1.2 per- tion and maintenance of basic infrastructure, cent for developing countries (figure 5.6). and the development of contingency plans, among other investments, demand resources. Low human development countries are more And with poverty and deprivation much more exposed to the human and economic losses prevalent in developing countries, people are from shocks from all sources. While some more vulnerable.78 extreme negative shocks can have an equaliz- ing effect within countries,75 people in very Within countries the effects of disasters vary high human development countries are better with income. Poorer people are more likely shielded from the costs because they have to be affected by natural hazards. In 12 of 13 more options for responding to shocks, greater country studies from developing countries, the

184 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 5.7 have declined. In the 1960s and 1970s there were twice as many deaths, despite a fraction of The lower the level of human development, the the number of recorded events, as over the past more deadly the disasters 20 years (figure 5.9). This reflects good work on disaster prevention, preparation and response. Deaths by natural Damage of disasters, 1997­2016 natural disasters, 1997­2016 International instruments—including the (per million inhabitants) (percent of GDP of one year) Yokohama Strategy (1994) and the Hyogo Framework for Action (2005), leading to the 1,000 16 2015 Sendai Framework for Disaster and Risk Reduction—have mobilized stakeholders across 800 12 the globe to invest in disaster risk reduction.80 As a result, developing and developed countries 600 are converging to lower vulnerability.81

8 But progress in reducing the absolute number 400 of deaths appears to have plateaued since the 1990s—likely the result of two forces. One 200 4 is further progress in adaptation, leading to convergence towards greater preparation to 0 0 recurrent events. Second is the greater frequen- Low cy and severity of shocks, possibly related to Medium High Very high climate change—increasing the human cost in poorer areas, creating inequalities. The IPCC’s Human development group 2014 Synthesis Report warned that “continued emission of greenhouse gases will cause further Note: Data are simple averages across human development groups. Country warming […] increasing the likelihood of severe, values are the sum of population or GDP over 20 years divided by the population pervasive and irreversible impacts for people and or GDP in one representative year. ecosystems.”82 Climate change “risks are uneven- Source: Human Development Report Office calculations based on data from ly distributed and are generally greater for disad- the Centre for Research on the Epidemiology of Disasters’ Emergency Events vantaged people and communities in countries Database (www.emdat.be/database; accessed 28 October 2019). at all levels of development.”83 If disasters tend to hit disadvantaged people harder, climate change percentage of poor people affected by natural could make vicious cycles of low outcomes and If disasters tend to hit hazards was larger than that of nonpoor peo- low opportunities more persistent.84 disadvantaged people ple.79 In El Salvador and Honduras people in harder, climate change the lower quintiles of the income distribution could make vicious were more likely to be affected by floods and cycles of low outcomes landslides (figure 5.8). and low opportunities more persistent There has been progress curbing the effects of recurring shocks behind disasters. Even though too many preventable casualties remain from events such as flooding, drought and earth- quakes, total causalities per recorded event

FIGURE 5.8

In El Salvador and Honduras people in the lower quintiles of the income distribution were more likely to be affected by floods and landslides

Physical damage from floods Physical damage from landslides (percent of population, per quintile of income) (percent of population, per quintile of income)

22 21

17 17 18 16

10 11 11 13 12 12

9 78 10 9

5 6

Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5

San Salvador (El Salvador) Tegucigalpa (Honduras)

Source: Hallegatte and others (2017), based on Fay (2005).

Chapter 5 Climate change and inequalities in the Anthropocene | 185 FIGURE 5.9 Fewer deaths in the 2000s than in the 1960s and 1970s despite more occurrences of natural disasters

Average number of deaths Average number of occurrences (thousands) 450

450

300 300

150 150

0 0

1950 1956 1962 1968 1974 1980 1986 1992 1998 2004 2010 2016

Note: Data are rolling 20-year averages. Source: Centre for Research on the Epidemiology of Disasters’ Emergency Events Database (www.emdat.be/database).

Environmental Shocks, including those related to climate them to prepare for and respond better to sur- inequalities become a change, can push people into poverty. In prise shocks, including climate-related ones.89 lens to understand and Senegal, households affected by a natural dis- aster were 25 percent more likely than others Environmental inequalities and address other forms to fall into poverty during 2006­2011.85 The injustices are pervasive—a of inequality, and the impacts of natural disasters go beyond income. global snapshot of waste, meat distribution of power In Ethiopia, Kenya and Niger children born consumption and water use and decisionmaking during droughts are more likely to suffer from malnutrition.86 In Cameroon climate shocks Environmental inequalities and environmental more broadly reduce girls’ chances of finishing primary school injustices have much deeper roots than the by 8.7 percentage points. In Mongolia, wildfires current climate crisis.90 The environmental jus- reduced the probability of completing second- tice movement has had strong links with other ary school by 14.4 percentage points.87 social justice movements.91

Climate change may also increase forced Ultimately, environmental inequalities—and population displacements. In 2017 there were environmental justice—are not just about the en- 18.8 million new internal displacements asso- vironment. They give expression to stigmatizing ciated with disasters across 135 countries and social norms and discriminatory laws and prac- territories, most caused by floods (8.6 million) tices, which are manifestations of inequality in and storms, including cyclones, hurricanes and different dimensions, many taking shape as hori- typhoons (7.5 million). While countries at zontal inequalities.92 Environmental inequalities different incomes were affected, most displace- thus become a lens to understand and address ments took place in developing countries,88 other forms of inequality, and the distribution of where the risk of becoming homeless due to power and decisionmaking more broadly. disasters is more than three times higher than in developed countries. Many environmental inequalities and injus- tices persist around the world. They are many, In sum, climate change impacts mediated pervasive and persistent because differences by disasters differ across the globe, with shifts in power (and how it is wielded) are as well. in both the nature of the events and their Environmental inequalities operate at many probability. This affects the ability to measure scales, reproducing and reinforcing familiar the effects and to formulate policies (box 5.3). gradients, as seen in the preceding climate dis- Developed countries appear to have a broader cussion and elsewhere in this Report. The rest set of resources and institutions that enable

186 | HUMAN DEVELOPMENT REPORT 2019 BOX 5.3 When history is no longer a good guide

When an event recurs, societies are likely to adapt through aspects outlined above. And with climate change, it ap- learning about four aspects: pears that communities around the world will confront

• The nature of the shock. more and more “surprises” (shocks outside of the his-
• The probability of occurrence. torical experience).2
• The effects of the event on well-being.
• The actions to reduce damage. With climate change the basic structure of shocks

does not disappear but evolves into a different process. Common knowledge accumulates over time, informed Current policy frameworks may become incomplete. by historical conditions, with lessons learned about what Some effects of climate change might take the form of works to reduce the negative effects of shocks. So when “black swans,” low-probability but high-impact events the events are uncertain but their effects are “known” to which both public and private institutions are ill-pre- from historical experience, coping mechanisms are eas- pared to respond. In other cases the effects are com- ier to develop. The upshot: a substantial reduction in the pletely unknown and unpredictable: when events never negative effects of shocks.1 This sort of adaptation occurs experienced before are observed (such as new record in all societies in different ways. temperatures). The ability to successfully adapt to cli- mate change depends on resources for an enhanced However, when events fall outside of the histori- system of preparation and response.3 cal norm, there is significant unpredictability in the four

1. See, for instance, Clarke and Dercon (2016). 2. For an example based on the climate impact on ocean temperature, see Pershing and others (2019); for the implications in terms of the need to develop a more prospective, as opposed to retrospective, ability to respond to surprise shocks, see Ottersen and Melbourne-Thomas (2019). 3. See, for instance, Farid and others (2016).

of this chapter takes a look at a few of them, in concentrate it in enormous garbage patches. More than 270,000 the forms of waste, meat and water use. Three have been identified so far: one in the tonnes of plastic North Pacific (the Great Pacific Garbage Patch), waste are in the Waste one in the South Pacific and one in the North world’s oceans, where Atlantic.96 The Great Pacific Garbage Patch gyres concentrate Waste93 comes from the flow of materials, often measures 1.6 million square kilometres (three it in enormous in the form of products, through society. More times the size of France), and parts of it have garbage patches waste generally means more upstream extraction upwards of 100 kilograms of plastic per square of raw materials, from mining to deforestation, kilometre.97 Plastics can circulate in oceans for with negative impacts on natural habitats. It also years, degrading in sunlight into microplastics, means more conversion of raw materials into forming a sort of peppery soup that birds and products, which usually entails the intensive use fish consume.98 Marine microplastics are not of industrial energy (especially from fossil fuels), confined to the sea surface; they have also been the consumption of water and the emission of documented in the water column and animal pollutants across interconnected networks. communities of the deep sea.99 The largest living space on earth, the deep sea, may also prove to Waste management requires transportation be one of the largest reservoirs of microplastics, and energy. It is a notable contributor to climate which have also been found in the atmosphere change. Nearly 5 percent of global greenhouse and remote mountains.100 gas emissions are due to waste management (ex- cluding transportation), driven mainly by food In 2016 the world generated just over 2 billion waste and improper management.94 When metric tonnes of solid waste, or 0.74 kilogram per burned openly, waste contributes to air pol- person per day, an average that varies widely by lution and health hazards; when deposited in country (0.11­4.54 kilograms).101 Under a busi- landfills, it takes up space and can leach toxins ness-as-usual scenario total waste is expected to into soil and groundwater. grow to 3.4 billion metric tonnes by 2050—and to grow fastest in low-income countries, tripling Waste also finds its way into waterways and by 2050. Richer countries produce more waste oceans. More than 270,000 tonnes of plastic per capita and poorer countries less (figure 5.10). waste are in the world’s oceans,95 where gyres

Chapter 5 Climate change and inequalities in the Anthropocene | 187 FIGURE 5.10

Richer countries generate more waste per capita

Waste generation per capita (kilograms per capita per day)

1.5

High income, 683

1.0 Lower middle

income,

586 Upper middle income, 655

Low income, 93

0

0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 40,000 45,000 50,000

GDP per capita in purchasing power parity terms (constant 2011 international $)

Source: Kaza and others 2018.

Livestock is the world’s Rates of waste collection vary consider- Meat consumption largest agricultural ably between and within countries. Waste collection is nearly universal in high-income Livestock production is important for live- user of land resources, countries, with little disparity between urban lihoods and economies. It employs at least with pasture and and rural areas. At lower income levels waste 1.3 billion people worldwide and supports the collection rates decline steadily, and stark livelihoods of some 600 million poor house- cropland dedicated disparities between urban and rural areas holds, mostly in developing countries,104 where to the production of open up. About 40 percent of global waste is it accounts for 20 percent of total agricultural feed accounting for disposed of in landfills, and one-third is open- output. Animal-source foods are important almost 80 percent of ly dumped. The vast majority of waste in low- components of healthy, nutritious diets, all agricultural land income countries is openly dumped, and open contributing especially to children’s balanced dumping steadily declines in favour of land- growth and cognitive development. Among fills, as country income increases. Incineration many other benefits, livestock can also help is used primarily among upper-middle and cushion households from negative impacts of high-income countries. Industrial waste shocks, such as droughts.105 typically far exceeds municipal solid waste and shows a steep gradient by country in- Livestock is the world’s largest agricultural come. Generally, recycling is a significant user of land resources, with pasture and cropland waste disposal method only in high-income dedicated to the production of feed accounting countries.102 for almost 80 percent of all agricultural land (while providing only 37 percent of the world’s In addition to urban-rural divides, inequal- protein and 18 percent of its calories—after in- ities in waste are evident within countries.103 cluding aquaculture).106 About a fifth of available Waste sites, polluting factories, and noisy air- freshwater is directed to livestock production.107 ports and expressways are eyesores and health The intensity of resource use by livestock is hazards that no community wants to be near. closely tied, directly and indirectly, to energy in- Their location in poorer communities thus re- efficiencies in animal food production systems. flects other forms of inequality. Most plant matter that animals ingest, including

188 | HUMAN DEVELOPMENT REPORT 2019 feed, is used up by the animals themselves rather as do the bottom 10 percent of emitters. The Up to 80 percent than stored as muscle or fat for consumption by problem is concentrated at the top: The major- of greenhouse gas people. The loss ratio varies but has been estimat- ity of emissions from beef herders come from emissions generated ed to be as high as 90 percent,108 making animals the highest impact 25 percent of producers. by the global a highly inefficient source of calories for people. One-size-fits-all approaches are unlikely to agricultural sector For each calorie, the production of animal foods work, but significant opportunities exist to re- are from livestock requires much more land and resources than the duce variability among farms and mitigate the production, which adds production of an equivalent amount of plant- environmental impacts of beef, livestock and up to 7.1 gigatonnes based foods.109 agricultural production generally. Reducing of carbon dioxide losses across the supply chain is another option, equivalent per year— Up to 80 percent of greenhouse gas emissions as is reducing demand for meat where possible or 14.5 percent of generated by the global agricultural sector are and appropriate. For instance, on a per unit of global anthropogenic from livestock production, which adds up to protein basis, greenhouse gas emissions from greenhouse gas 7.1 gigatonnes of carbon dioxide equivalent per the bottom 10 percent of beef producers still emissions year—or 14.5 percent of global anthropogenic exceed those from peas by a factor of 36.113 greenhouse gas emissions.110 Emissions emanate from across the supply chain, with feed produc- The environmental benefits of dietary change tion, enteric fermentation, animal waste and exceed what producers can achieve on their land use changes among the most important own (box 5.4).114 But the trend is in the op- sources at the farm level).111 Cattle are respon- posite direction, owing mostly to population sible for about two-thirds of livestock-related growth but also to other variables, such as ur- carbon dioxide equivalent emissions, largely in banization and rising per capita incomes, that the form of methane emissions, a greenhouse tend to increase demand for animal foods.115 gas roughly 30 times more potent than carbon Between 2000 and 2014 the global production dioxide in trapping heat.112 of meat rose 39 percent and milk 38 percent. The Food and Agriculture Organization of the Improving farm management is one way to United Nations estimates that by 2030 meat reduce these and other environmental impacts. production will increase another 19 percent For many major agricultural products, green- from that in 2015­2017, with developing house gas emissions vary widely across farms. countries accounting for almost all the increase Livestock is no exception. For beef the top (figure 5.11). Milk production is projected to 10 percent of emitters produce up to 12 times grow 33 percent in the same period.116 Even as much greenhouse gases per unit of protein

BOX 5.4 The impacts of a global dietary shift on sustainable human development

A global dietary shift favouring more plant-based of recent systematic reviews have, with some contro- foods and following guidelines for good nutrition versy, called into question the degree to which reduc- would impact several dimensions of sustainable hu- ing red and processed meat consumption improves key man development, both in aggregate and in distribu- health indicators).4 Numerous studies have estimated tion. The climate would also benefit. One estimate is the impacts of nutritious, plant-based diets, including that dietary changes could reduce growth in food-re- on overall mortality reduction.5 The benefits, however, lated greenhouse gas emissions by 29­70 percent by are not evenly shared. On a per capita basis, high- and 2050.1 On a per capita basis, food-related emissions middle-income countries might benefit more, owing to could fall twice as much in richer countries as in poor- reduced red meat consumption and lower energy in- er ones, narrowing the inequality in carbon dioxide takes.6 A global shift to sustainable, nutritious, plant- equivalent emissions between them.2 This would be based diets, therefore, could improve health overall driven primarily by reductions in red meat consump- globally while potentially worsening some kinds of tion, which also has health benefits3 (though a series health inequalities among countries.

1. Springmann and others 2016. 2. Springmann and others 2016. 3. Springmann and others 2016. 4. Han and others forthcoming; Vernooij and others forthcoming; Zeraatkar, Han and others forthcoming; Zeraatkar, Johnston and others forthcoming. See also Carroll and Doherty (2019) and Johnston and others (forthcoming). 5. Key and others 2009; Le and Sabaté and 2014; Orlich and others 2013; Springmann and others 2016; Tilman and Clark 2014. 6. Springmann and others 2016.

Chapter 5 Climate change and inequalities in the Anthropocene | 189 FIGURE 5.11 Developing countries will drive most of the rise in meat production to 2030

Metric tonnes (carcass weight equivalent/ready to cook)

60 6.9 0.7

3.1 2.9 23%

50 3.2

19.1

30 13.5

77%

10 10.1

0

Beef Pork Poultry Sheep Beef Pork Poultry Sheep

Total

Developing Developed increase

Source: FAO 2018.

Global water though developing countries will drive future Water use withdrawal has nearly growth in meat production, the world’s richer septupled over the last countries eat meat most intensively, and this is Water and sanitation are essential for human expected to continue well into the future.117 development. They have also been recognized century, outpacing as human rights.122 Despite the expansion of population growth by a As incomes rise, food expenditures favour safely managed drinking water and sanitation factor of 1.7. Most of it more nutrient-rich foods, such as animal foods services over the past two decades, significant (Bennett’s Law).118 This is explained partly gaps remain. As of 2017, 29 percent of peo- is for agricultural use by the nutritional benefits of meat and other ple worldwide lacked access to safe drinking animal products, especially for children in water. The gap is even higher for sanitation, at poorer households. There are clear inequalities 55 percent.123 in spending on meat across income quintiles, but as incomes increase, inequalities in meat How much water humans use and in what consumption decline.119 ways have consequences for the environment and societies. Global water withdrawal has Projections of meat consumption—and ine- nearly septupled over the last century, outpac- qualities—do not account for wild cards such as ing population growth by a factor of 1.7.124 technological breakthroughs that could greatly Most of it is for agricultural use (69 percent), alter current trajectories and reduce environ- followed by industry (19 percent) and munic- mental damages. An estimated 31 start-ups are ipalities (12 percent).125 Attempts have been working to become the first company to market made to establish a meaningful safe operating synthetic animal protein.120 Competition will space for water use at the global level.126 The also come from elsewhere, particularly novel conceptual underpinnings are also being revis- vegan meat replacements121 New areas of diver- ited to consider subnational boundaries and to gence could open up, since products are likely expand beyond consumptive use of blue water to be rolled out initially in rich countries. And (freshwater in the form of rivers, lakes, ground- if these foods offer additional benefits in re- water and so on) to include green water (soil ducing noncommunicable diseases, they could moisture that evaporates or transpires) and exacerbate health inequalities.

190 | HUMAN DEVELOPMENT REPORT 2019 other elements of the dynamic, global hydro- to higher overall consumption of those prod- In many countries, logical cycle. Much analytical, management ucts per se,135 though the latter can be relevant basic water and and policy work remains at the national level as well.136 This points to the enormous potential sanitation coverage for and at smaller spatial scales, such as the basin.127 that remains for efficiency improvements. the wealthiest quintile is at least twice that It is at these spatial scales where water stress, Water access and consumption also vary for the poorest quintile scarcity and crises are manifest. By some es- greatly within countries. Consider access to timates, as many as 4 billion people, about safe drinking water and sanitation, where two-thirds of the global population, live under significant inequalities persist between and conditions of severe water scarcity for at least within countries. Gaps in coverage between one month of the year.128 Half a billion people rural and urban areas have long been impor- face water scarcity year-round.129 One-third of tant. Globally, over the past two decades the the world’s 37 largest aquifer systems are con- gaps have narrowed, falling from 47 percentage sidered stressed.130 Globally, enough freshwater points to 32 for safely managed water services is available to meet annual demand, but spatial and from 14 percentage points to 5 for safely and temporal mismatches between water and managed sanitation services. In many countries supply drive water scarcity. The 2006 Human inequalities by wealth are significant. In some, Development Report argues forcefully that basic water and sanitation coverage for the limits on physical supply are not the central wealthiest quintile is at least twice that for the problem but rather that “the roots of the crisis poorest quintile (figure 5.12). For water, wealth in water can be traced to poverty, inequality inequalities generally exceed urban-rural ones and unequal power relationships, as well as within the same country. While water and sani- flawed water management policies that exacer- tation coverage has generally improved over the bate scarcity.”131 past two decades across most, but not all, coun- tries, inequalities by wealth have shown no such Water footprints are one way to understand general trend. In some countries inequalities and measure human use of water. Every country have declined; in others they have increased.137 has a national water footprint, the amount of water produced or consumed per capita. The As with urban-rural divides, national averages footprint includes virtual water, which is the can mask differences and deprivations at lower water used in the production of such goods as levels. In South Africa the national Gini index food or industrial products. Across countries, for piped water is .36, but this varies consid- agriculture constitutes the single greatest com- erably across the country’s provinces, from ponent (92 percent) of the water consumption .06 (least unequal) to .57 (most unequal).138 footprint, with cereals the largest subcompo- Reducing inequality in water access and use can- nent (27 percent), followed by meat (22 per- not mean denying people their right to water, a cent) and milk products (7 percent).132 Because right embedded in South Africa’s constitution the national water footprint of consumption and affirmed by legislation that includes sanita- includes imported virtual water, some coun- tion.139 The human right to water and sanitation tries have water footprints much larger than is also affirmed in the Sustainable Development might be expected based on national water re- Goals. The very realization of this right should source endowments alone. The transboundary go a long way in reducing inequalities. movement of virtual water is significant. Over 1996­2005 about one-fifth of the global water Increasingly severe water-related crises footprint was bound up in exported goods, around the world are driving what some have with trade in crops the lion’s share.133 argued is a fundamental transition in freshwater resources and their management. Approaches Water footprints vary considerably across that focus singularly on meeting water demand countries. The widest variation is for developing are giving way to more multifaceted ones that countries. Indeed, some of them have national recognize various limits on supply, broader eco- water footprints of consumption on par, or logical and social values of water, and the costs exceeding, those in developed countries.134 The and efficiency of human use. Nexus approaches high water footprints in some developing coun- are emerging that identify and respond to the tries have been attributed more to lower effi- way in which water is linked to other resources, ciencies of water use in consumed products than such as energy, food and forests.140

Chapter 5 Climate change and inequalities in the Anthropocene | 191 FIGURE 5.12

In some countries basic water and sanitation coverage for the wealthiest quintile is at least twice that for the poorest quintile

Rural/Urban Wealth quintile Rural Urban Poorest quintile Richest quintile

Democratic Republic of the Congo Somalia Ethiopia

Cameroon

United Republic of Tanzania Honduras Pakistan

Dominican Republic India

Kazakhstan Guyana

Montenegro Turkey

Thailand

0 20 40 60 80 100 0 20 40 60 80 100

Source: UNICEF and WHO 2019.

Environmental Economic production systems, demographic inextricably linked with inequalities in human inequalities are largely trends and climate change are all playing big development. They reflect the way economic parts in this shift. So is technology. Over the and political power—and the intersection of a choice, made by past two decades, for example, the spread of so- the two—is distributed and wielded, both those with the power phisticated precision irrigation technology has across countries and within them. Often, these to choose. Remedying improved efficiency of water use in agriculture. environmental inequalities and injustices are them is also a choice Modern technologies are also transforming the legacy of entrenched gradients in power go- wastewater treatment and reuse, as well as the ing back decades; for climate change, centuries. economic viability of seawater desalination. Countries and communities with greater power Remote sensing provides real-time data. Smart have, consciously or not, shifted some of the water meters and improved water pricing environmental consequences of their consump- policies can both improve efficiency.141 The tion onto poor and vulnerable people, onto response to and shaping of these new tools and marginalized groups, onto future generations. trends—the extent to which inclusion is made Environmental inequalities are largely a choice. a bedrock principle of a shift to freshwater sus- Remedying them is also a choice, but doing so tainability—will play a big role in determining cannot come at the expense of achieving the whether the human rights to water and sani- full suite of people’s human rights. tation are progressively realized, inequalities in access to both are reduced and a path of Technology has been central to the climate sustainable water use is embarked on. story. It has underpinned development trajec- tories that are directly linked to the climate A break from the past: crisis. Technology, in the form of renewables Making new choices for and energy efficiency, offers a glimpse that the people and planet future may break from the past—if the oppor- tunity can be seized quickly enough and broad- This chapter has shown that environmen- ly shared.142 If so, both people and planet win. tal inequalities are many and that they are The way people grapple with these and other technologies so that they encourage, rather than threaten, sustainable and inclusive human

192 | HUMAN DEVELOPMENT REPORT 2019 development is the subject of the following at building climate resilience. But much more Historical development chapter on technology. on the policy front needs to be done urgently, paths have exacted with developed and developing countries work- environmental and The uptake and broad diffusion of cli- ing together, to avoid dangerous climate tipping social tolls that are mate-protecting technologies old and new will points and to ensure that poor and vulnerable too great. They must be critical in charting new development paths people are not left behind. Chapter 7, which change, and there for all countries. Historical development paths takes a panoramic look at policy options across are encouraging have exacted environmental and social tolls that the Report, discusses some potential policies signs that they are are too great. They must change, and there are that help address climate change and inequality encouraging signs that they are. The SDGs, the together in the hope that they help countries Paris Agreement and renewed interest in and chart their paths for more sustainable, more expansion of progressive carbon pricing offer inclusive human development. promising paths forward. So do efforts thus far

Chapter 5 Climate change and inequalities in the Anthropocene | 193 Spotlight 5.1

Measuring climate change impacts: Beyond national averages

Climate change will A recent study that moved beyond national av- mechanism for poor people, thereby worsening worsen inequality erages to a more granular look at climate change inequality. Mobility in the United States has in the United impacts in 3,143 counties across the continental fallen in recent decades.3 States because United States1 could signal the future for cli- the worst impacts mate change economic impact assessments— While in middle-income countries warming are concentrated partly because some of the model’s parameters has increased emigration to cities and other in regions that are were linked to real-world, observed data. countries, in poorer countries warming has re- already poorer duced the likelihood of emigration.4 Although on average The study found significant spatial heter- this does not mean that poorer people in rich ogeneity in agricultural yields and all-cause countries are less likely to migrate in response mortality. Projected economic impacts varied to climate change, it does indicate that other widely across counties, from median losses ex- variables—perhaps poverty-related ones at vari- ceeding 20 percent of gross county product to ous levels—can interact with climate change to median gains exceeding 10 percent. Negative shape migration likelihood and overall coping economic impacts were concentrated in the capacity. It also suggests that migration as a South and Midwest, while the North and coping mechanism for climate change is less West showed smaller negative impacts—or common in poorer countries than in richer even net gains. ones.

The study concluded that climate change will Granular analyses, adapted for differences worsen inequality in the United States because in data availability and quality, could be useful the worst impacts are concentrated in regions in other contexts. They could also be linked that are already poorer on average. By the latter to deprivation and vulnerability data so that part of the 21st century, the poorest third of climate exposure, impacts and vulnerabilities counties are projected to experience damages could be brought together, superimposed and of 2­20 percent of county income. Effects in integrated for policy-relevant analysis and the richest third are projected to be less severe, visualization, perhaps using geographic infor- ranging from damages of 6.7 percent of county mation systems. Vulnerability hotspots could income to benefits of 1.2 percent. Nationally, be identified—spatially and by population—for each 1°C increase in global mean surface tem- policy action, including through impact mitiga- perature will cost 1.2 percent of GDP. tion and resilience building. Granular analyses would also be key in developing place-specific The study does not address one of the main adaptation pathways, which could advance coping mechanisms for climate change: migra- climate change adaptation, structural inequality tion. Migration would affect national impact reduction and broader Sustainable Development estimates as well as the absolute costs and Goal achievement by “identifying local, socially benefits for individual counties. In theory, mi- salient tipping points before they are crossed, gration could also dampen the impact on ine- based on what people value and tradeoffs that are quality, as those experiencing the most negative acceptable to them.”5 impacts move to areas less affected and with more opportunities. The United States has a Notes long history of migration for economic op- portunity, including in times of environmental 1 Hsiang and others 2017. and economic crisis (such as the Dust Bowl).2 2 Hornbeck 2012. In practice today, however, some evidence sug- 3 Carr and Wiemers 2016. gests migration may not be a significant coping 4 Cattaneo and Peri 2016. 5 Roy and others 2019, p. 458.

194 | HUMAN DEVELOPMENT REPORT 2019 Spotlight 5.2

Climate vulnerability

Much like economic feedback mechanisms, assistance (such as for regional public insurance Some worsening attention to structural inequalities and de- mechanisms); and global governance.4 of inequality due to velopment deficits in the context of climate climate change is change is a fairly recent advance. In a literature The IPCC’s Fifth Assessment Report con- already “baked in.” review in four climate-change journals through cluded with very high confidence that climate The idea of “soft” and 2012, 70 percent of published studies articulat- change would worsen existing poverty and ex- “hard” adaptation ed climate change itself as the main source of acerbate inequalities.5 The IPCC’s 2018 special limits is a recognition vulnerability, while less than 5 percent engaged report summarized subsequent literature show- of the variability of with the social roots of vulnerability.1 The ing that “the poor will continue to experience communities and Intergovernmental Panel on Climate Change’s climate change severely, and climate change will human institutions (IPCC) Fifth Assessment Report in 2014 exacerbate poverty (very high confidence).“6 The to respond to and helped redress this imbalance.2 special report cites evidence of poorer subsist- cope with climate ence communities already affected by climate change impacts How the variables of social (or structural) change through declines in crop production vulnerability aggregate at different levels— and quality, increases in crop pests and diseases, from individuals and households to towns and and disruption to culture. A series of studies cities to districts and provinces to countries referenced in the special report indicates that and regions—will shape the patterns of cli- children and the elderly are disproportionally mate-related impacts across space and across affected by climate change and that it can in- populations in those spaces. Different patterns crease gender inequality. The special report also of inequality may emerge at different scales cites a 2017 report that claims that by 2030, and depending on the kind of inequality being 122 million additional people could become measured. The impact on inequalities at those extremely poor, due mainly to higher food pric- different levels depends critically on whether es and worse health. The poorest 20 percent more negative impacts are disproportionally across 92 countries would suffer substantial borne by those on the lower ends of existing income losses. Lower-income countries are inequality distributions—that is, those already projected to experience disproportional socio- experiencing various forms of greater depriva- economic losses from climate change, placing tion or development deficits. Given that struc- pressure towards greater inequality between tural inequalities exist in various forms and are countries and countering prevailing trends of inextricably linked to communities’ and coun- recent decades towards less inequality between tries’ capacities to cope with climate change, countries.7 Furthermore, the special report then absent mitigating factors, some worsening identifies critical research gaps, stating that “im- inequality due to climate change is already pacts are likely to occur simultaneously across “baked in.” Furthermore, the idea of “soft” and livelihood, food, human, water and ecosystem “hard” adaptation limits, as well as “loss and security…but the literature on interacting and damage” and “residual climate-related risks,” in cascading effects remains scarce.”8 the climate change literature is a recognition of the variability of communities and human in- A 2016 United Nations Department of stitutions to respond to and cope with climate Economic and Social Affairs (UNDESA) change impacts.3 The IPCC’s 2018 special report summarizes the literature on structural report on global warming of 1.5°C briefly sum- inequalities and their relationship to climate-re- marizes the latest literature on approaches and lated exposure and vulnerability.9 Within policy options to address residual risk and loss countries, the UNDESA report notes that and damage, looking at adaptation and disaster many poor people live in floodplains, along risk reduction strategies; compensatory, dis- riverbanks or on precarious hillsides for lack tributive and procedural equity considerations; of alternatives, putting them at greater risk of litigation and litigation risks; international flooding, mudslides and other weather-related disasters. A climate change axiom is that wetter

Chapter 5 Climate change and inequalities in the Anthropocene | 195 areas will become wetter and dry areas drier. Notes Flood frequencies are expected to double for 450 million more people in flood-prone are- 1 Tschakert (2016), based on data from Bassett and Fogelman as.10 Climate change will also place additional (2013). drought-related stress on those in arid and semi-arid areas, where large concentrations of 2 IPCC 2014. poor and marginalized people live. Poor people 3 Klein and others (2014), as cited in Roy and others (2019). are expected to be more exposed to droughts 4 Roy and others 2019. for warming scenarios above 1.5°C in several 5 IPCC 2014. countries in Asia and in Southern and West 6 Roy and others 2019, p. 451. Africa.11 The rural poor in poor countries will 7 Pretis and others (2018), as cited in Roy and others (2019). suffer a double whammy from climate change: a 8 Roy and others 2019, p. 452. negative shock to their livelihoods and spikes in 9 UNDESA 2016. food prices resulting from drops in global yields. 10 Arnell and Gosling (2016), as cited in Roy and others (2019). 11 Winsemius and others (2018), as cited in Roy and others (2019).

196 | HUMAN DEVELOPMENT REPORT 2019 Chapter 6

Technology’s potential for divergence and convergence: Facing a century of structural transformation 6.

Technology’s potential for divergence and convergence: Facing a century of structural transformation

Will the technological transformations unfolding before our eyes increase inequality? Many think so, but the choice is ours. There certainly is historical precedent for technological revolutions to carve deep and persistent inequalities. The Industrial Revolution may have set humanity on a path towards unprecedented improvements in well-being. But it also opened the Great Divergence,1 separating societies that industrialized,2 producing and exporting manufacturing goods, from many that depended on primary commodities well into the middle of the 20th century.3 And by shifting the sources of energy towards the intensive use of fossil fuels (starting with coal), the Industrial Revolution launched production pathways culminating in the climate crisis (chapter 5).4

Whether the ongoing changes in technology powered by artificial intelligence techniques can be characterized as a revolution is for fu- known as machine learning—particularly ture historians to determine. The digitalization deep learning—which enables machines to of information and the ability to share infor- match, or even surpass, what humans can do mation and communicate instantaneously and on tasks ranging from translating languages to globally have been building over several dec- recognizing images and speech.7 As artificial ades, as with computers, mobile phones and intelligence continues to improve the bench- the internet. The 2001 Human Development mark performance in a wider range of tasks,8 Report considered how to make these and it is likely to reshape the world of work in fun- other new technologies work for human devel- damental ways—for workers performing those opment, focusing on their potential to benefit tasks and across the entire labour market.9 developing countries and poor people.5 While the report did not address technology’s impact Artificial intelligence is not the only relevant on jobs and earnings in detail, it highlighted technology. Nor does it work in isolation. It the growing demand for technology skills and interacts with digital technologies in ways that the potential for job creation in both devel- are reshaping knowledge-based labour mar- oped and developing economies, suggesting kets, economies and societies.10 Perhaps for the the possibility for reducing inequality within first time in human history, these technologies and across countries. But recent advances in are known almost everywhere. East Asian technologies such as automation and artificial countries are investing heavily in artificial in- intelligence, as well as developments in labour telligence and in advances in its use (discussed markets over the course of the 21st century, later in the chapter). And African countries show that these technologies are replacing have seized the potential of mobile phones to tasks performed by humans—raising with foster financial inclusion.11 heightened urgency the question of whether technology will give rise to a New Great These technologies also change politics, cul- Divergence. ture and lifestyles. Basic artificial intelligence algorithms meant to increase the number of Advances in artificial intelligence grabbed clicks in social media have led millions towards headlines when a computer programme be- hardened extreme views.12 In some countries came, in just a few hours, the world’s best chess family and friends are being displaced by the in- player. The programme had no prior informa- ternet as the main vehicle for couples to meet, tion on how to play the game. Given only the partly because of better artificial intelligence rules, it taught itself how to win—not only at algorithms for matching people.13 The world chess but also at Go and Shogi.6 This was the of finance is being fundamentally reshaped, latest of several technological breakthroughs with nonfinancial technology firms provid- ing payment services. China leads the way in

Chapter 6 Technology’s potential for divergence and convergence: Facing a century of structural transformation | 199 Technology is not mobile payments, which represent 16 percent Can artificial intelligence enhance human something outside of GDP, followed by the United States, India development? The direction of technological and Brazil—but at a distance, at still less than change involves many decisions by govern- economies and 1 percent of GDP.14 These firms are also extend- ments, firms and consumers.27 But making tech- societies that ing credit and other financial services. In China nology work for people and nature is already artificial intelligence enables online lenders part of the conversation in some countries.28 determines outcomes to make decisions on loans in seconds, with Public policy and public investment will drive on its own new credit granted to more than 100 million technological change, as they have historical- people.15 And central banks from China16 to ly.29 But so will the distribution of capabilities. Rwanda17 are considering digital currencies. The cleavages that may open are not necessarily between developed and developing countries Now take a step back. Technology has always or between people at the top and people at progressed in every society, creating disruptions the bottom of the income distribution. North and opportunities (from gunpowder to the America and East Asia, for instance, are far printing press). But the advances were typically ahead in expanding access to broadband inter- one-off and did not translate into the sustained net, accumulating data and developing artificial and rapid progress18 that Simon Kuznets intelligence.30 described as “modern economic growth.”19 Sustained improvements in productivity and This chapter shows that while access to basic living standards depend on constantly intro- technologies is converging, there is a growing ducing new ideas and using them productive- divergence in the use of advanced ones, echoing ly.20 But having these gains in productivity and the findings in part I of the Report. The chap- well-being reach everyone is not a given, and ter describes how some aspects of technology people who lack access can face new and deeper are associated with the rise of some forms of deprivations when access is simply assumed.21 inequality—for instance, by shifting income towards capital and away from labour and the Technology is not something outside econ- increasing market concentration and power omies and societies that determines outcomes of firms. It then examines the potential for on its own.22 It co-evolves with social, political artificial intelligence and frontier technologies and economic systems. This implies that it to narrow inequalities in health, education takes time for the productive use of technology and governance—pointing to technology’s to settle, because it requires complementary potential in redressing inequalities in human changes in economic and social systems.23 But development. It concludes that technology how technology will shape the evolution and can either replace or reinstate labour—it is distribution of human development in the 21st ultimately a matter of choice, a choice not de- century does not need to be left to chance. At termined by technology alone. a minimum another Great Divergence should be avoided while simultaneously addressing the Inequality dynamics in access climate crisis. to technology: Convergence in basic, divergence in enhanced The impact of technical change can be an ex- plicit concern for policymakers.24 With a clear A refrain throughout this Report is that de- emphasis on enhancing human development, it spite convergence in basic capabilities, gaps can increase the employability of workers and remain large in enhanced capabilities—and improve the reach and quality of social services. are often widening. This is also the case for Investments in artificial intelligence need not technology, especially for access, the focus simply automate tasks performed by humans; here. To be sure, this is only a partial perspec- they can also generate demand for labour. For tive, given the inequalities in leveraging new example, artificial intelligence can define more technologies, having a seat at the table in the detailed and individualized teaching needs and development of these technologies and being thus generate more demand for teachers to pro- trained or reskilled for working with them. vide a wider range of education services.25 More There are also gender disparities, with women generally, technological change can be directed to both reduce inequality and promote envi- ronmental sustainability.26

200 | HUMAN DEVELOPMENT REPORT 2019 and girls under-represented in education and (box 6.1). But digital gaps can also become bar- In more empowering careers in science, technology, engineering and riers not only in accessing services or enabling areas of technology, mathematics.31 Still, the evidence on access in economic transactions but also in being part involving access to this chapter shows that despite convergence in of a “learning society.”37 It is thus important to more information and access to basic technologies (which is still far complement this static picture of gaps with an a potential transition from equal), there is divergence in the access to analysis of how they are evolving. from consuming and use of advanced ones. content to producing Catching up in the basics, widening it, the gaps are larger In fact, the ability to access and use digital gaps in advanced technologies and widening technologies has a defining role both in the pattern of production and consumption and Inequalities in access to basic entry-level in how societies, communities and even house- technologies are shrinking. Mobile phones, holds are organized. More and more depends— including basic service, have spread rapidly in to a great extent—on the ability to connect to most parts of the world (figure 6.2, left panel). digital networks. This section shows that: In 2007 there were 102 mobile subscriptions

• Groups with lower human development have per 100 inhabitants in developed countries

compared with 39 in developing countries. By systematically less access to a wide range of 2017 the gap had narrowed, with 127 mobile technologies, as is widely established. subscriptions per 100 inhabitants in developed

• Gaps in basic entry-level technologies, countries and 99 in developing countries. This

though still evident, are closing—reflecting convergence reflects both rapid expansion at convergence in basic capabilities. the bottom and a binding constraint at the top,

• Gaps in advanced technologies32 (even when with little room for further growth.

considered commonplace by the standards of many) are widening—mirroring the pattern In more empowering areas of technology, in enhanced capabilities identified earlier in involving access to more information and a the Report. potential transition from consuming content to producing it, the gaps are larger and widening Inequalities in access to (figure 6.2, right panel). Low human develop- technology are widespread ment countries have made the least progress in these technologies—a trend consistent with the The higher the level of human development, widening gaps in installed broadband capacity, the greater the access to technology (figure 6.1, especially in absolute differences, to which the top panel). The digital revolution has moved chapter turns next in some detail.38 fast and had enormous impact, but it is far from universal. In 2017 almost 2 billion people The distinction between the number of still did not use a mobile phone.33 And of the telecommunication subscriptions and the 5 billion mobile subscribers in the world, nearly availability of bandwidth mattered little when 2 billion—most of them in low- and middle- there was only fixed-line telephony, since all income countries—do not have access to the the connections had essentially the same band- internet.34 In 2017 the number of fixed broad- width. But as artificial intelligence and related band subscriptions per 100 inhabitants was technologies continue to evolve, bandwidth only 13.3 globally and 9.7 in developing coun- will be increasingly important (as will be cloud tries, and the number of mobile broadband computing, which depends on the ability to subscriptions per 100 inhabitants was 103.6 connect computers with each other). Access to in developed countries compared with only bandwidth, comparable in quantity and quality 53.6 in developing countries.35 Inequalities are to that in developed countries, is essential for much greater for advanced technologies, such developing countries to cultivate their own as access to a computer, internet or broadband artificial intelligence and related applications. (figure 6.1, bottom panel). Also essential are transferring and adopting technologies developed by leaders in the digital The convergence in basic technologies, such world. Taking these two groups of countries in as mobile phones,36 has empowered tradition- the aggregate, there has been convergence. In ally marginalized and excluded people—with 2007 high-income countries had 22.4 times greater financial inclusion a good illustration

Chapter 6 Technology’s potential for divergence and convergence: Facing a century of structural transformation | 201 FIGURE 6.1 Digital divides: Groups with higher development have greater access, and inequalities are greater for advanced technologies, 2017

The higher the level of human development, the greater the access to technology.

Mobile-cellular subscriptions Households with internet Households with a computer Fixed broadband subscriptions

(per 100 inhabitants) (percent) (percent) (per 100 inhabitants)

Very high human development 131.6 84.1 80.7 28.3

High human development 116.7 51.7 47.0 11.3

Medium human development 90.6 26.8 20.0 2.3

Low human development 67.0 15.0 9.7 0.8

Concentration curves

Inequalities are much greater for advanced technologies.

Mobile-cellular subscriptions Households with internet Households with a computer Fixed broadband subscriptions (per 100 inhabitants) (percent) (percent) (per 100 inhabitants)

Cumulative Cumulative Cumulative Cumulative outcome proportion outcome proportion outcome proportion outcome proportion 1.0 1.0 1.0 0.8 Perfect equality 0.8 1.0 0.8 0.6 0.8 0.6 0.6 0.4 0.6 0.4 0.2 0.4 0.2 0.4 0.0 0.2 0.0 0.2 0 20 40 60 80 100 0 20 40 60 80 100 Percent of population 0 20 40 60 80 100 Percent of population 0.0 Percent of population

More equal Less equal

Note: Data are simple averages across human development groups. Shaded areas are 95 percent confidence intervals. Source: Human Development Report Office calculations based on country-level data from the International Telecommunications Union.

the bandwidth per capita of other countries; by reflects the population distribution (meaning 2017 the ratio had fallen to 3.4 (figure 6.3). that the distribution of both is roughly equiv- alent), and in East Asia and the Pacific mobile While the convergence in broadband among subscriptions have already caught up with the re- developing countries as a whole is positive, the gion’s share in the global population (figure 6.4). pattern of convergence in technologies has dif- In Africa there is still a difference, though con- fered across regions. Take mobile subscriptions vergence is not far off. But the distribution of and installed broadband potential. The regional installed bandwidth potential follows neither distribution of mobile subscriptions already

202 | HUMAN DEVELOPMENT REPORT 2019 BOX 6.1 Mobile technology promotes financial inclusion

Financial inclusion is the ability to access and use a Increases in e-commerce have also been dramatic, in- range of appropriate and responsibly provided finan- cluding individuals and small businesses selling products cial services in a well regulated environment.1 Mobile and services on online platforms. In particular, inclusive money, digital identification and e-commerce have given e-commerce, which promotes the participation of small many more people the ability to save money and trans- firms in the digital economy, is important because it can act business securely without needing cash, to insure create new opportunities for traditionally excluded groups. against risks and to borrow to grow their businesses In China, for example, an estimated 10 million small and and reach new markets. medium enterprises sell on the Taobao platform; nearly half the entrepreneurs on the platform are women, and In 2017, 69 percent of adults had an account with a more than 160,000 are people with disabilities.4 financial institution, up 7 percentage points from 2014.2 That means more than half a billion adults gained ac- From artificial intelligence to cryptography, innova- cess to financial tools in three years. tion in financial technology is transforming the financial sector globally.5 While financial technology offers many Well known examples of mobile money—plat- potential benefits, there are also considerable concerns forms that allow users to send, receive and store mon- about these new technologies’ vulnerabilities. Blockchain ey using a mobile phone—include Kenya’s M-Pesa and technology, for one, provides applications that include a China’s Alipay. Mobile money has brought financial secure digital infrastructure to verify identity, facilitate services to people long ignored by traditional banks. It faster and cheaper cross-border payments and protect reaches remote regions without physical bank branch- property rights. But these technologies bring new risks es. It can also help women access financial servic- that are not fully considered by existing regulations.6 es—an important aspect of equality, since women in Policymakers will need to address several tradeoffs to many countries are less likely than men to have a bank reap financial technology’s potential benefits. account.3

1. UNCDF 2019. 2. Demirgüç-Kunt and others 2018. 3. McKinsey 2018; World Bank 2016. 4. Luohan Academy 2019. 5. He and others 2017. 6. Sy and others 2019.

the distribution of the population nor the distri- suggesting that by 2030 about 70 percent of East Asia and the bution of gross national income. East Asia and the global economic benefits tied to artificial Pacific has already the Pacific has already taken the lead in installed intelligence will accrue to North America and taken the lead in bandwidth potential, with 52 percent in 2017. East Asia.39 installed bandwidth potential, with So, the emerging technology cleavages do New technologies tend to have higher prices 52 percent in 2017 not follow a simple developed­developing when initially introduced, with prices falling country dichotomy, and the emerging dis- and quality increasing as the technologies dif- parities are fairly recent. From 1987 to 2007 fuse.40 Thus, every innovation has the potential little changed in the global ranking of installed to initially carve a divide, at the beginning of bandwidth potential (figure 6.5). In 1987 a the diffusion process—a point also made in group of developed countries were in the top chapter 2, in the discussion of how gradients global ranks: The United States, Japan, France in health emerged when health technologies and Germany hosted more than half the global became available. The contribution here is to bandwidth, mainly through fixed-line teleph- show that the gaps for advanced technologies ony. At the turn of the millennium things are widening, not closing—in a new geography started to change, notably with the expansion of divergence that goes beyond developed and of bandwidth in East and North Asia: By developing countries. Avoiding a New Great 2007 Japan, the Republic of Korea and China Divergence implies paying attention to the occupied ranks 1, 3 and 5. And in 2011 China evolution of technology distribution, because took the lead in installed bandwidth. Beyond benevolent technology diffusion is neither broadband, projections on the distribution of automatic nor instantaneous.41 Instead, tech- future economic benefits linked to artificial nology may well catalyse divergence in human intelligence confirm this shifting geography development outcomes. By what processes? of technology divergence, with estimates That is the topic of the next section.

Chapter 6 Technology’s potential for divergence and convergence: Facing a century of structural transformation | 203 FIGURE 6.2 Dynamics of access to technology

Declining inequality Increasing inequality

Mobile-cellular subscriptions Fixed broadband subscriptions Change between 2007 and 2017 (per 100 inhabitants) Change between 2007 and 2017 (per 100 inhabitants)

59.5 12.3

49.3 49.3 8.9

26.1

Low Medium High Very high Low Medium High Very high

Human development group Human development group

Basic Enhanced

Convergence Divergence Access to mobile Access to broadband

Change between 2007 and 2017 Change between 2007 and 2017 De ve lo ped (per 100 inhabitants) (per 100 inhabitants) South Asia Europe and Central Asia 60 East Asia and the Pacific

50 Sub-Saharan Africa Europe and Central Asia 10

40 Arab States Latin America 8 and the Caribbean

and the Caribbean 6 Arab States

De ve lo ped 4 East Asia and the Pacific

10 South Asia 2

0 0

0 20 40 60 80 100 120 0 5 10 15 20 25

Subscriptions per 100 inhabitants, 2007 Subscriptions per 100 inhabitants, 2007

Note: Convergence and divergence are tested for in two ways: by using the slope of an equation that regresses the change over 2007­2017 with respect to the initial value in 2007 (with ordinary least squares, robust and median quantile regressions) and by comparing the gains of very high human development countries and the gains of low and medium human development countries. For mobile subscriptions there is convergence according to both metrics (p-values below 1 percent). For fixed broadband subscriptions there is divergence according to both metrics (p-values below 1 percent). Source: Human Development Report Office calculations based on data from the International Telecommunication Union.

204 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 6.3 The bandwidth gap between high-income and other countries fell from 22-fold to 3-fold

Bandwidth potential per capita Ratio of bandwidth potential per capita (megabits per second) for high-income countries to that for the rest of the world

80 Ratio: high-income countries 25 to the rest of the world 63.39 20 60 22.4

20 High Rest of 18.50 5 income the world

1.69 0.08 3.4

0 0

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

Technology is reshaping the growth.44 And it may have been the result of For most of the 20th world: How will it shape creating and strengthening such institutions as century the shares inequality in human development? trade unions and social insurance.45 However, of national income with a decline in labour’s share of income since going to labour and to Technology is reshaping lives—not only econ- the 1980s across both developed and develop- capital held remarkably omies but also societies and even politics. What ing economies, this empirical regularity has constant across specific changes will bear on inequality in hu- been unravelling.46 For developed economies, many economies man development? This question is difficult to technology has been a key driver of the decline, address, in part because it may never be possible in part by replacing routine tasks, as described to assign to technology alone any of the major in chapter 2.47 For developing countries the evi- changes that will reshape inequality in human dence is ambiguous, with both technology and development, especially with globalization and globalization playing important roles.48 its interaction with technological change also playing a major role. Still, this section high- A related trend is the steep decline in the lights some emblematic ways in which tech- price of machinery and equipment, such as nology is upending previously stable patterns computers (generally designated as capital or in the distribution of income and economic investment goods), relative to the price of con- power. The aim is not so much to attribute cau- sumer goods.49 Since 1970 the relative prices sality as to give a sense of technology’s potential of investment goods in developing countries to reshape inequalities in human development fell by almost 60 percent, with 75 percent of over the next few years. the decline occurring from 1990 onwards.50 Among investment goods the price decline has Unravelling stable trends42 been dramatic for computing and communi- cations equipment, pointing to a link between For most of the 20th century the shares of technology and the incentives for firms to national income going to labour and to capital replace labour with capital, a process that in held remarkably constant across many econo- developing countries was also associated with mies.43 This was far from a foregone conclusion greater integration in global value chains.51 to those witnessing the evolution of economic Another recent development—linked to the two trends just noted as well as to the increase in

Chapter 6 Technology’s potential for divergence and convergence: Facing a century of structural transformation | 205 FIGURE 6.4

The distribution of mobile subscriptions is converging to the distribution of population by region, but installed bandwidth potential is not

Share of total (percent) Mobile subscriptions Population

0 2016 2017 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016

East Asia and the Pacific North America Europe and Central Asia South Asia Latin America and the Caribbean Sub-Saharan Africa Middle East and North Africa Share of total (percent) Gross national 100 Installed bandwidth potential income

0

1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2016 2017

FIGURE 6.5

From 1987 to 2007 little changed in the global ranking of installed bandwidth potential, but at the turn of the millennium things started to change, with the expansion of bandwidth in East and North Asia

1987 2007 2017

Republic Rest of United Netherlands Rest of Japan Rest of China of Korea the world States 2% the world 19% the world 37% 28% 2% 28% Italy 21% 29% 2% United Indonesia Russian States 2% Federation 17% Spain Japan Germany 2% 11% 3% Republic United China of Korea 2% United Kingdom 8% Canada 3% 10% Brazil States France Germany 2% India 12% 3% Italy 6% 9% France 2% Russian 4% 7% Federation Germany France Republic Japan 4% 5% United 6% 2% of Korea Russian 4% Kingdom Federation

5% 3%

206 | HUMAN DEVELOPMENT REPORT 2019 corporate profits (discussed below) and changes There is evidence that both manifestations of There has been in corporate income tax rates (discussed in chap- market power are increasing, and even though a sharp increase ter 7)—is the shift in the balance of savings held technology is not the only element driving this in markups (the by households and by firms. National savings shift, it is playing an important role. difference between (comprising household, corporate and govern- what a firm charges ment savings) are needed to fund investments. There has been a sharp increase in markups and the marginal Until the late 1980s most savings were held by (the difference between what a firm charges and cost of production), households, but today as much as two-thirds the marginal cost of production), and this has and this has been are accounted for by the corporate sector.52 And been linked directly to labour’s declining share linked directly to given that corporate investment has been stable, in income.55 While the trend of increased mar- labour’s declining this means that corporations have been holding ket power is widely shared across several sectors share in income on to these savings, in some countries using and industries, firms in sectors that intensively them to repurchase their own stock. use information and communications technolo- gies have witnessed faster, and greater, increases Perhaps more consequential for the distribu- in markups (figure 6.6), suggesting that tech- tion of income is a breakdown in many coun- nology’s relevance pervades across a wide range tries in the association between improvements of firms.56 Look now at big digital companies, in labour productivity and the typical worker’s commonly known as Big Tech, and explore how earnings, well documented for developed coun- they have been acquiring market power. tries. This Report has already shown the trend towards the accumulation of income at the Many Big Tech firms are platforms. Uber, top in several countries (chapter 3). Here the the ride sharing company, is a platform where emphasis is specifically on labour income. This drivers offer their services and customers come breakdown between productivity and earnings looking for those services. Gojek and Grab work not only goes against what used to be stable in the same way in Asia. Amazon is a platform trends but is also inconsistent with simple linking sellers of products with potential buyers. models of the labour market. All platforms benefit from network effects— that is, the value of the platform increases when As workers become more productive (in part there are more participants on both sides of the as a result of technological change), one would market. For Amazon the more sellers and the expect their earnings to increase. That is, after all, more buyers, the better for each group—and, the assumed process for technological change to of course, for Amazon as well.57 Getting big deliver improvements in living standards—per- supports staying big, since buyers are reluctant haps not to everyone immediately but for the ma- to leave a platform where they find sellers, and jority over time. And indeed, until the 1980s real sellers, buyers. Social media companies such as average earnings for the bottom 90 percent of the Facebook and Instagram also benefit directly population (a proxy for the income of a typical from network effects—people stay on the net- household) increased in step with productivity work where their friends and family are. growth for many countries.53 Since then, there has been a decoupling in the evolution of these two Big Tech intensively uses data and, in- indicators, with the earnings of a typical family creasingly, artificial intelligence, so another remaining flat or increasing less than productivity network spillover common to all platforms is growth. The International Labour Organization economies of scale in data use, making these has documented a similar decoupling for 52 de- firms prone to acquiring market power.58 Even veloped economies, showing that from 1999 to though these platforms lower prices for con- 2017, labour productivity increased 17 percent sumers (and so, from that perspective, a more while real wages rose 13 percent.54 traditional measure of market power such as markups may not seem to apply), they can Shifting economic power exercise market power by potentially limiting competition and choice.59 The big players The market power of firms can be manifested spend vast amounts on lobbying to influence in their ability to charge prices above the cost policies that keep them in place and potential of production or by paying lower wages than new entrants out.60 And they can use their would be needed in an efficient labour market. vast reserves of cash to simply buy up new platforms starting to make a mark. Google

Chapter 6 Technology’s potential for divergence and convergence: Facing a century of structural transformation | 207 FIGURE 6.6 Market power is on the rise, particularly for firms intensive in information and communication technology

Average markup Information and communication technology­intensive firms (index, 1.00 = 2000) All firms 1.15

1.10

1.05

1.00

0.95

0.90

2000 2003 2006 2009 2012 2015

Note: Values are average markups for firms from 20 countries, advanced and emerging, both publicly listed and privately held. Source: Diez, Fan and Villegas-Sánchez 2019.

In parallel to the rise bought its competitors DoubleClick and industrialized economies, and in developing of monopoly power YouTube. Facebook first acquired Instagram, economies it adds to the casual labour force.65 then WhatsApp. Both companies, like others, in product markets is are the products of hundreds of mergers.61 The discussion here illustrates how technology the growing market is already shaping the distribution of income66 power in labour In parallel to the rise of monopoly power and of economic power through rising markups, in product markets is the growing market with firms exercising power at the expense of markets—monopsony power in labour markets—monopsony power workers and consumers, as reflected in the declin- power (exercised by (exercised by employers), which, once again, ing share of labour income and the decoupling of employers), which, is linked to the decline in labour’s share of median wages from labour productivity.67 Further income.62 And when employers have power in advances in technology, linked to advances in au- once again, is linked to labour markets, the impact of technological tomation and artificial intelligence, could accel- the decline in labour’s change on inequality can be magnified.63 erate these dynamics,68 while pushing to the limit existing frameworks to curb market power. The share of income Technology is enabling monopsony power merit of antitrust action is still assessed primarily in online platforms that are carving out tasks by how much consumer prices have risen.69 But to assign to humans based on who charges the technology platforms are based on an exchange lowest price. This includes work in digital la- of user data for “free services.” So there are calls bour markets such as TaskRabbit and Amazon to revisit current antitrust approaches and how to Mechanical Turk, commonly referred to as extend them to curb monopsony power.70 crowdwork. The availability of online work may lower search costs, which would make Harnessing technology markets competitive. But market power is high for a Great Convergence even in this large and diverse spot market. For in human development Amazon Mechanical Turk, employers capture much of the surplus created by the platform. This chapter started by asserting that avoiding This has implications for distributing the gains another Great Divergence was a matter of from digital labour markets, which will likely choice—though that does not imply that the become greater over time.64 While crowdwork task will be easy. It ends with indications of is a product of technological advances, it also represents a return to the past casual labour in

208 | HUMAN DEVELOPMENT REPORT 2019 how to exercise that choice and unleash a Great increase in productivity, boosting the demand Automation can Convergence in human development. The fo- for all factors of production, including labour be leveraged to cus will remain on digital and related technol- (figure 6.7). After elaborating on the potential create new tasks—a ogies, guided by a broad set of principles linked of this framework to identify opportunities reinstatement effect, to the implementation of the 2030 Agenda for to use artificial intelligence to increase labour which would counter Sustainable Development (box 6.2). It first pro- demand, the discussion moves towards some of the displacement effect vides a framework to analyse the impact of arti- broader risks associated with it. ficial intelligence and automation that suggests opportunities to generate demand for labour. Artificial intelligence’s potential for The discussion also considers the challenges of reinstating work artificial intelligence, including the potential to exacerbate horizontal inequalities, as well In addition to the amount, it is important to as the ethics of it. It then provides concrete consider the quality of work. Do the kinds of illustrations of how technology can, in practice, new tasks created through technology differ reduce inequality, particularly addressing the fundamentally from past ones? For example, divergence in enhanced capabilities identified the rise of platforms may push down the in part I of the Report. number of workers in brick-and-mortar retail stores while increasing the number employed Automation, artificial intelligence in fulfilment centres preparing online orders and inequality: Will it be possible to for shipping.73 Work available on platforms increase the demand for labour? has introduced flexibility and extended work opportunities in some sectors but created chal- Automation and artificial intelligence do not lenges such as how to handle the large amount have to shrink the net demand for labour.71 of data on workers, which poses risks for work- Automation can be leveraged to create new er privacy and could have other consequences, tasks—a reinstatement effect, which would depending on how the data are used.74 counter the displacement effect.72 The impact on inequality will depend on how technology In addition to offering new work opportuni- changes the task content of production— ties, platforms can enhance financial inclusion. whether it displaces or reinstates labour This is happening in South-East Asia (where through the creation of new kinds of tasks. For more than three-quarters of the population is example, jobs such as fulfilment centre work- unbanked) thanks to ride-hailing services such er, social media adviser and YouTube media as Gojek and Grab.75 Once drivers become part personality did not exist a few decades ago. of these platforms, they get support to open Technological advance also results in an overall bank accounts, and the apps have become vehi- cles to handle financial transactions, including

BOX 6.2

Digital technologies for the Sustainable Development Goals: Creating the right conditions

Digital technologies have transformational potential. several recommendations under broad themes, such as Different actors at different levels have to participate build an inclusive digital economy and society; protect for these applications to be taken to scale. Many appli- human rights and human agency while promoting digital cations are yet to be developed. Policies are needed— trust, security and stability; and fashion a new global at the national and global levels—to provide the right digital cooperation architecture.1 incentives to developers and adopters of technology in the fields most beneficial for human development. As a follow up to that report, the Global Charter for a Sustainable Digital Age provides a set of principles The UN Secretary-General established the High- and standards for the international community, aiming level Panel on Digital Cooperation in July 2018 to identi- to link the digital age with the global sustainability per- fy examples of and propose ways for cooperating across spective. It sets out concrete guidelines for action for sectors, disciplines and borders. Its final report made dealing with the challenges of the digital age.2

1. UN 2019a. 2. German Advisory Council on Global Change website (www.wbgu.de/en/publications/charter).

Chapter 6 Technology’s potential for divergence and convergence: Facing a century of structural transformation | 209 FIGURE 6.7

Technology can displace some tasks but also reinstate new ones

Displacement Productivity Reinstatement effect effect effect

(tasks related to accounting + (cybersecurity and bookkeeping, experts, digital travel agents) Net change in transformation specialists, demand for data scientists) - labour +

Basing the impact of cash. Incentives to adopt more formalized pay- Some occupations have several tasks that artificial intelligence ment methods extend to retailers, such as food could be easily replaced by artificial intelli- merchants using the platform to deliver to their gence bundled with other tasks that are diffi- and automation customers.76 cult or impossible for machines to replace. A on the assumption radiologist’s task of checking medical images Basing the impact of artificial intelligence to identify anomalies can be performed by that technology and automation on the assumption that tech- artificial intelligence, but a machine cannot could replace nology could replace entire occupations can set priorities, consult with the medical team, lead to high estimates of how many jobs are make treatment plans or communicate with entire occupations at risk.77 An approach based on tasks (with patients and family—all tasks the radiolo- can lead to high occupations defined by a bundle of different gist performs. This suggests that when tasks estimates of how tasks) provides a more balanced and more within a job can be separated and rebundled, actionable framework to understand the there is potential for job-redesign or job-craft- many jobs are at risk impact—and potential—of artificial intelli- ing.79 With the prevalence of highly accurate gence and automation. There is evidence that, medical image recognition, radiologists can within occupations, the possibility of tasks spend less time looking at images and more being replaced with artificial intelligence time interacting with other medical teams varies greatly, and different occupations have and with patients and family. Job-redesign different resulting levels of susceptibility and job-crafting thus offer opportunities to (table 6.1).78

210 | HUMAN DEVELOPMENT REPORT 2019 leverage artificial intelligence to increase la- TABLE 6.1 bour demand. Different tasks have different potential for being replaced by artificial intelligence

The ability of artificial intelligence to iden- Occupations with Suitability Occupations with Suitability tify patterns, relationships and trends and to low suitability for for machine high suitability for for machine automatically display them through interactive machine learning learning score machine learning learning score dashboards or create automated reports is constantly improving. This implies updated Massage therapists 2.78 Concierges 3.90 task structures for many jobs, including stock 3.90 market traders, copywriters and even journal- Animal scientists 3.09 Mechanical drafters ists and editors. While a lot of tasks will be au- 3.89 tomated, high-level management and oversight Morticians, undertakers and of automated systems tasks are less susceptible. 3.78 However, an occupation’s aggregate suitability Archaeologists 3.11 funeral directors 3.78 for machine learning score is not correlated with wages.80 So it is not inevitable that artifi- Public address system and other cial intelligence will replace or depress wages in certain occupations, as some argue about announcers 3.13 Credit authorizers previous waves of automation.81 Plasterers and stucco masons 3.14 Brokerage clerks A human-centred agenda thus requires atten- tion to technology’s broader role in advancing Source: Brynjolfsson, Mitchell and Rock 2018. decent work. Technology can free workers from drudgery and arduous labour. There is even decision task is broader, requiring the collection Realizing technology’s potential for collaborative robots, or cobots, to and organization of data, the ability to take an potential in the future reduce work-related stress and injury. Realizing action based on a decision and the judgement of work depends on technology’s potential in the future of work to evaluate the payoffs associated with different fundamental choices depends on fundamental choices about work outcomes. For individual workers, advances in about work design, design, including detailed job-crafting discus- artificial intelligence will matter to the degree including detailed job- sions between workers and management.82 that prediction is a core skill in the tasks that crafting discussions make up their occupation. The diagnosis that a between workers Intelligence augmentation (using computers radiologist provides can also be partially made and management to extend people’s ability to process informa- by artificial intelligence, but that is very differ- tion and reason about complex problems) ent from a decision on the course of treatment means that artificial intelligence, instead of or its implementation by a surgeon. Automated aiming at automation, can integrate human prediction thus enhances rather than replaces agency and automation in a way that enhanc- the value of these occupations. es both. The augmentation can take place in everyday human tasks. This happens already Exercising choices to seize on in spelling and grammar checking in word technology’s potential: Balancing risks processors, which highlight text to correct and opportunity errors, and in autocompletions of text input in internet search engines. Automatic suggestions, After establishing artificial intelligence’s poten- easily dismissible, can accelerate the search and tial to reinstate work, this section elaborates on refine ambiguous queries. These provide value, elements to consider in seizing the opportuni- promoting efficiency, accuracy and the consid- ties that artificial intelligence, and technology eration of alternate possibilities. They augment, more broadly, are presenting. Doing so implies but do not replace, user interaction.83 also having a clear-headed perspective on risks. For instance, artificial intelligence can accentu- Finally, the recent advances in artificial ate biases and horizontal inequalities (box 6.3), intelligence do not increase artificial general including exacerbating gender disparities in the intelligence that could substitute machines for workforce, leading to even more women being all aspects of human cognition. Artificial in- in low-quality service jobs.85 Women, on aver- telligence has been very effective at one aspect age, perform more routine or codifiable tasks of intelligence: prediction.84 But prediction than men and fewer tasks requiring analytical is only an input into decisionmaking. The input or abstract thinking.86 These differences are also present in gender gaps in education and employment linked to technology.87 LinkedIn

Chapter 6 Technology’s potential for divergence and convergence: Facing a century of structural transformation | 211 BOX 6.3 Artificial intelligence and the risk of bias: Making horizontal inequalities worse?

Artificial intelligence applications have the potential to and sensors on a large scale. How does this differ from support positive social change—indeed, in some do- mass surveillance? mains their impact could be revolutionary. But as with any new technology, actually achieving these positive Machine learning algorithms are not biased inher- results is challenging and risky. ently; they learn to be biased. Algorithmic bias occurs when the learning algorithm is trained on biased da- Many groups of people across the globe may be on tasets and subsequently “accurately” learns the pat- the receiving end of artificial intelligence’s downside. terns of bias in the data.2 In some cases the learned They may lose their jobs as more tasks are performed representations within machine learning algorithms can by machine learning—even if net job loss is contained, even exaggerate these biases.3 For example, women are inequalities in income and wealth could rise, and the less likely to receive targeted ads for high-paying jobs quality of jobs fall. Workers may see strong biases potentially because the algorithm that targets the ads against their skin colour or gender embedded in ma- trained on data in which women had lower paying jobs.4 chine learning, and they may be subjects of surveillance. And a computer programme used in the United States to Algorithms for job matching may reproduce historical bi- assess the risk of reoffending by individuals in the crim- ases and prejudices. Companies need policies on trans- inal justice system incorrectly flagged black defendants parency and data protection so that workers know what as high risk nearly twice as often as white defendants.5 is being tracked. Regulation may be needed to govern data use and algorithm accountability in the world of Facial recognition services can be much less accu- work. rate in identifying women or people with darker skin.6

As uses of artificial intelligence become pervasive, The well recognized lack of diversity among the questions arise about the rise of propaganda and manip- people designing and developing artificial intelligence ulation, undermining democracy, and about surveillance is another problem. Few women work in artificial intel- and the loss of privacy. For example, artificial intelli- ligence, as in the tech sector in general, and among the gence applications are linked with the development of men, racial diversity is limited.7 Diverse teams, bringing smart cities.1 This involves collecting data from cameras diverse perspectives, representative of the general pop- ulation, could check biases.

1. Glaeser and others 2018. 2. Caliskan, Bryson and Narayanan 2017; Danks and London 2017. 3. Zhao, Wang and others 2017. 4. Spice 2015. 5. IDRC 2018. 6. Boulamwini and Gebru 2018. 7. IDRC 2018.

LinkedIn and the World and the World Economic Forum found a will oppose autonomous weapons, which Economic Forum significant gap between female and male search and engage targets without human representation among artificial intelligence intervention.89 Many companies—from Big found a significant professionals—only 22 percent worldwide are Tech to startups—are formulating corporate gap between female.88 Racial and ethnic differences among ethical principles overseen by ethics officers women in access to training and employment or review boards. Still, it is unclear how ac- female and male opportunities can exacerbate these disparities. countable they will hold themselves to the representation among Artificial intelligence and technology more principles—which points to the need for broadly developed by teams that reflect a coun- regulation.90 Governments increasingly use artificial intelligence try’s population can counter such risks. When artificial intelligence themselves, and some are professionals—only teams are not diverse, artificial intelligence developing data ethics principles (box 6.4). 22 percent worldwide will tend to be trained on data that may have When artificial intelligence systems inform built-in biases that a more representative envi- decisionmaking that affects humans (such as are female ronment could avoid. medical diagnosis or providing a judge with an assessment of potential recidivism), avoid- Researchers, firms and governments are ing bias and errors across different contexts responding to manage the risks of artificial and communities is especially important. And intelligence—which include accentuation given the global application and reach of many of biases as well as development of deceptive artificial intelligence innovations, collective and malicious applications. For instance, action may be needed at some point on some thousands of artificial intelligence researchers regulatory aspects. have signed an open letter stating that they

212 | HUMAN DEVELOPMENT REPORT 2019 BOX 6.4 The United Kingdom’s Data Ethics Framework principles

1. Start with clear user need and public benefit. Using 5. Ensure robust practices and work within your data in more innovative ways has the potential to skillset. Insights from new technology are only transform how public services are delivered. We as good as the data and practices used to create must always be clear about what we are trying to them. You must work within your skillset recog- achieve for users—both citizens and public servants. nising where you do not have the skills or experi- ence to use a particular approach or tool to a high 2. Be aware of relevant legislation and codes of prac- standard. tice. You must have an understanding of the rele- vant laws and codes of practice that relate to the 6. Make your work transparent and be accountable. use of data. When in doubt, you must consult rel- You should be transparent about the tools, data and evant experts. algorithms you used to conduct your work, working in the open where possible. This allows other re- 3. Use data that is proportionate to the user need. searchers to scrutinise your findings and citizens to The use of data must be proportionate to the user understand the new types of work we are doing. need. You must use the minimum data necessary to achieve the desired outcome. 7. Embed data use responsibly. It is essential that there is a plan to make sure insights from data are 4. Understand the limitations of the data. Data used used responsibly. This means that both develop- to inform policy and service design in government ment and implementation teams understand how must be well understood. It is essential to consider findings and data models should be used and moni- the limitations of data when assessing if it is ap- tored with a robust evaluation plan. propriate to use it for a user need.

Source: UK Department for Digital, Culture, Media and Sport 2018.

A broader set of disruptions to the world of technology disruptions on specific income Compensation of work, powered in part by artificial intelli- groups and the resistance to those changes.94 for crowdwork is gence, is linked to digital labour platforms— During adjustments, vulnerable workers often below the alluded to earlier. These applications allow typically face periods of unemployment or minimum wage the outsourcing of work to geographically see their earnings eroded. But if technology dispersed people, generating crowdwork. changes rapidly, it might be more challenging While they provide new sources of income to to find decent jobs in a new techno-economic many workers in different parts of the world, paradigm95 than after a more “standard” eco- the work is sometimes poorly paid, and no nomic recession. Social insurance programmes official mechanisms are in place to address un- can provide affected workers with sustenance fair treatment. Compensation for crowdwork during transition periods, but the nature of the is often below the minimum wage.91 True, transition matters as well: Sectors and locations much policy innovation is already under way, where the displacement effect is stronger may with subnational regulators stepping up.92 need targeted social protection schemes.96 But the dispersed nature of the work across international jurisdictions makes it difficult Active labour market policies—including to monitor compliance with applicable labour wage subsidies, job placement services and laws. That is why the International Labour special labour market programmes—can fa- Organization suggests developing an inter- cilitate adaptation to a new techno-economic national governance system for digital labour paradigm. The ideal would be a social protec- platforms that sets minimum rights and tion floor that affords a basic level of protection protections and requires platforms (and their to all in need, complemented by contributory clients) to respect them.93 social insurance schemes that provide increased protection.97 The design of these systems pre- Providing social protection sents policymakers with choices ranging from ensuring coverage at the bottom while curbing A related challenge is providing social protec- leakage to the better off98 to balancing the tion to help address both the adverse impact generosity of transfers and the losses in efficien- cy99 and ultimately to assessing the fiscal cost

Chapter 6 Technology’s potential for divergence and convergence: Facing a century of structural transformation | 213 against alternative uses.100 Narrowly targeted FIGURE 6.8 policies could include measures to facilitate geographic mobility, supporting housing and Workers in medium and high wage jobs are more moving costs,101 particularly if technology likely to participate in adult learning creates jobs in one region while contributing to their elimination in others. Participation in adult learning by wage level (percent) Ultimately, social protection will be only part of the response, because workers whose jobs are 0 10 20 30 40 50 60 70 80 partially or fully automatable will need to adjust Italy to substantially changed or entirely new occupa- Turkey tions. Since automation affects some tasks and Lithuania creates others, the nature and content of jobs Poland change constantly. And this requires workers Slovakia to learn throughout their lives. Artificial intelli- Japan gence and automation tend to make high-skilled Greece workers more valuable and in demand. There is France evidence that those are the workers who avail Belgium themselves of lifelong learning opportunities, Spain while participation among low-skill, low-wage Israel workers is much lower (figure 6.8). Thus, there Korea, Rep. of is a risk of patterns of divergence emerging in Austria workplace and lifelong learning that are simi- Slovenia lar to those in enhanced capabilities. Lifelong Germany learning risks creating a wedge by enabling the Chile highly skilled to race further ahead.102 Singapore Taxation and data regulations Estonia Beyond the impact of artificial intelligence on Canada labour markets, two systemic challenges and Finland risks merit particular attention: taxation and United Kingdom data regulation. As the potential for machines United States to replace tasks performed by humans grows, Australia some have argued that there is an efficiency ra- Netherlands tionale for taxing robots103 and for channelling Sweden technology to reinstate, rather than replace, New Zealand labour.104 In addition, digitally intensive eco- Norway nomic activities, where the value of companies Denmark is linked less to their physical presence in a country and more to the number of members Average of networks around the world, are challenging longstanding assumptions underlying princi- Low-wage Medium/high wage ples of taxation. Some proposed actions and ideas serve the interest of particular tax juris- Source: OECD 2019c. dictions,105 but given that digital activities are global and many companies operate across bor- or deciding the placement of drivers waiting for ders, there is a clear need for an international rides, the revenues of more and more firms are consensus on how to tax digital activities, and tied to collecting and analysing huge amounts efforts to broker such an international agree- of data. The free flow and use of data are im- ment are under way.106 portant for businesses and governments. But there is also a need to protect personal data, Data are at the centre of the digital economy. data embodying intellectual property and Whether targeting ads, managing supply chains data related to national security. For now the ownership and use of data are governed mostly by default norms and rules. But many jurisdic- tions at different levels are working out data policies to ensure that advances in innovation also protect users.107 European governments,

214 | HUMAN DEVELOPMENT REPORT 2019 through the European Union’s General Data with the availability of real-time objective data Technology can Protection Regulation, have instituted data pri- on mood—from smartphone keystrokes, for help, for instance, in vacy rules.108 Beyond regulation are proposals example—artificial intelligence can help men- enabling individually to pay users for their data, to spread the wealth tal health diagnosis. Elderly care providers are customized content generated by artificial intelligence. Firms could starting to offload some parts of care to artifi- to “teach at the generate better data by paying. Data-providing cial intelligence, from early diagnosis of disease right level” labour could come to be seen as useful work, to at-home health monitoring and fall detec- conferring the same sort of dignity as paid tion.115 Artificial intelligence has also been used employment.109 to pore through genetic data to discover that a shortage of the element selenium could be asso- Deploying technology as a force for ciated with premature births in Africa.116 convergence in human development Applications of artificial intelligence extend For education to drive convergence implies beyond education and health to other public preparing young people today for the world of services, leading not only to greater efficiency work of tomorrow. Technology can help, for and enhanced transparency, but also to broad- instance, in enabling individually customized er participation in various aspects of public content to “teach at the right level.” This is es- life. For example, linguistic diversity, a given pecially important because the rapid expansion in most countries, can make e-governance ser- of access to primary and secondary education vices inaccessible for entire groups. In South in developing countries has led to the enrol- Africa, with 11 official languages, the Centre ment of millions of first-generation learners. If for Artificial Intelligence Research is work- they fall behind and lack instructional support ing on machine translation approaches to at home, they may learn very little in school.110 broaden access to government services.117 In One example of how technology can help Uganda the AI Research Group at Makerere in middle school grades is a technology-led University is developing source datasets instructional programme called Mindspark for some of the dozens of languages spoken used in urban India. It benchmarks the initial there.118 learning level of every student and dynamically personalizes material to match the individual’s The potential returns are huge in service level and rate of progress. In just 4.5 months delivery during and after disasters. Artificial those with access to the programme scored Intelligence for Disaster Response is an open- higher in math and in Hindi.111 In partnership source project that applies artificial intelli- with the programme, India’s government is gence to mine, classify and tag Twitter feeds providing a personal learning platform called during humanitarian crises, turning the raw Diksha. Pointing a cell phone at a printed QR tweets into an organized source of information code opens a universe of interactive content— that can improve response times. Soon after lesson plans for teachers and study guides for Ecuador experienced a major earthquake in students and parents.112 2016, Zooniverse, a web-based platform for crowd-sourced research, launched a website Digital health solutions can also drive con- that combined inputs from volunteers and vergence. Still in their early days, they show an artificial intelligence system to review the potential for expanding service coverage. 1,300 satellite images and, two hours after Services include digitizing supply chains and the website’s launch, produced a heat map of patient data, with integrated digital platforms damages.119 for information, bookings, payments and com- plementary services. They are important in areas For social protection, technology is helping that are remote and that have inadequate access in targeting payments and other benefits, pro- to health care providers. Artificial intelligence viding timely delivery and reducing opportu- is already taking hold, for example, in machine nities for fraud. Public platforms that support pattern recognition for medical scans and skin interoperability and data exchange can reduce lesions.113 There is also potential for machine the administrative burden and the time to learning to aid personalized nutrition.114 And deliver services to poor, vulnerable and margin- alized groups, promoting social and economic inclusion.120

Chapter 6 Technology’s potential for divergence and convergence: Facing a century of structural transformation | 215 The direction of Technology can also improve the availability Recall that the public sector has supported technological of data and information for policymakers and fundamental research for technology that was businesses—and inform public debate. For subsequently commercialized by the private change can be an instance, as digital imagery becomes ubiquitous sector.132 Technological innovation will be explicit concern and machine vision techniques improve, auto- crucial to meet the Sustainable Development for policymakers mated systems lend themselves to measuring Goals.133 Harnessing technology for that demographics with fine spatial resolution in purpose calls for all countries to shape global close to real time.121 The same applies to meas- and national institutions and policies that will uring poverty and other social and economic determine the impact of technological change indicators, often combining mobile phone data on sustainability and inclusion in a way that is and satellite imagery, with the use of multiple nationally relevant.134 It is in this context that lenses obtained from diverse datasets helping international intellectual property rights mat- capture information on living standards more ter. An overly stringent intellectual property accurately.122 For instance, in Senegal the rights regime can make technology diffusion Multidimensional Poverty Index can be accu- harder (box 6.5). rately predicted for 552 communes using call data records and environmental data (related to The successful generation, diffusion and food security, economic activity and accessibility adoption of technology for development take to facilities). This approach can generate poverty place in a network of multiple actors—in- maps more frequently, and its diagnostic capa- cluding the private sector, government and bility is likely to assist policymakers in designing academia, often referred to as a national inno- better interventions to eradicate poverty.123 vation system.135 Public policies to influence the direction of technology are nested in such In the same way that artificial intelligence systems. Across countries, there are enormous can chart individualized learning paths for stu- asymmetries in the size and organization of dents, artificial intelligence’s potential to collect innovation efforts. Research and development detailed and frequent data can be leveraged to are still more intensive in developed countries obtain very specific localized information.124 For (figure 6.9), and on average the gap with other instance, using an artificial intelligence algorithm countries is widening, but at the same time new to analyse weather and local rice crop data in regions are emerging as scientific and techno- Colombia125 led to distinct recommendations logical powerhouses, as in East Asia. for different towns, helping 170 farmers in Córdoba avoid direct economic losses estimated Important in the ability to invest nationally at $3.6 million and potentially improving rice in science and technology, the diffusion of production. Other applications include using innovation will remain a powerful driver to cutting-edge artificial intelligence to tackle urban increase productivity. Enhancing the pro- challenges related to traffic, safety and sustaina- ductivity and employability of every work- bility. These applications range from artificial er—reaching those currently in informal and intelligence traffic management126 to artificial in- precarious forms of employment and excluded telligence systems that locate pipes at risk of fail- from more modern productive systems—will ure.127 Global telecommunication networks and tend to reduce income inequality while increas- cloud services can enable artificial intelligence ing incomes.136 insights to be transferred and adapted in dif- ferent contexts.128 Sharing artificial intelligence For this mechanism to work, workers must results among machines enables transfer learn- be able to use technology and benefit from the ing,129 through which knowledge moves and is rise in productivity. Between 2007 and 2017 customized into new contexts,130 supplementing the median income in many countries grew less resources in previously underserved areas. than productivity per worker, even though in- come and productivity are strongly correlated *** (figure 6.10, left panel). Moreover, the higher the productivity, the greater the share of pro- The direction of technological change can ductivity that the median worker receives as be an explicit concern for policymakers.131 compensation (see figure 6.10, right panel). Decoupling median labour income from pro- ductivity implies that increasing productivity

216 | HUMAN DEVELOPMENT REPORT 2019 BOX 6.5 Intellectual property rights, innovation and technology diffusion

In principle, intellectual property rights can be a pow- Under the World Trade Organization Trade Related erful driver to incentivize innovation and creativity, Aspects of Intellectual Property System, developing even if they impose temporary restrictions on free countries are encouraged to increase the level and access to new knowledge. But in some cases they stringency of their intellectual property provisions in have generated patent thickets, patent trolls and ev- order to enhance international transfers of technology ergreening1—potentially curbing not only diffusion, and spur innovative domestic firms.5 The point is that but also innovation itself. Patent thickets imply long intellectual property protection will give them the right and costly negotiations to obtain multiple permissions. to the profits from research and development break- Patent trolling—where innovators face suits from oth- throughs. But country case studies show mixed evi- ers who own intellectual property simply to profit by dence of intellectual property rights being important licensing patents rather than undertaking production for foreign investment inflows, domestic technological themselves—is costly.2 And evergreening—where development or technology transfers.6 companies extend their patent protection by inventing new follow-on patents that are closely linked but allow Assigning patents to a shell company in a low tax for a longer period of monopoly than would otherwise country, paying royalties on their own patents to the be permitted—curbs competition. shell companies and parking the income offshore illus- trates how intellectual property rights can be used for On balance, while weak patent systems may in- tax avoidance.7 These mechanisms further concentrate crease innovation only mildly, strong patent systems income, wealth and market power. Here, as in other can slow innovation.3 In the last few decades a higher areas, economic institutions and laws created in the concentration of patent ownership, echoing the broader 20th century to manage industrialization in developed pattern of market concentration, has contributed to de- economies may need to be reconsidered in the 21st clines in knowledge diffusion and business dynamism.4 century.

1. Baker, Jayadev and Stiglitz 2017. 2. Bessen and Meurer 2014. 3. Boldrin and Levine 2013. 4. Akcigit and Ates 2019. 5. Baker, Jayadev and Stiglitz 2017. 6. Maskus 2004. 7. Dharmapala, Foley, and Forbes 2011; Lazonick and Mazzucato 2013.

FIGURE 6.9

There are enormous asymmetries in research and development across human development groups

Researchers, 2017 Change in researchers, 2007­2017 Expenditure, 2015 Change in expenditure, 2005­2015 (per million people) (per million people) (percent of GDP) (percentage points)

4,000 800 1.80 0.30

3,000 600 0.20 1.20 2,000 400 0.10 1,000 200 0.60

0.00

0 High Very high 0 0 Low and High Very high ­0.1 0 Low and medium medium

Human development group Human development group

Source: Human Development Report Office calculations based on data from the World Bank’s World Development Indicators database.

Chapter 6 Technology’s potential for divergence and convergence: Facing a century of structural transformation | 217 FIGURE 6.10

Income and productivity are strongly correlated, and the higher the productivity, the greater the share of productivity that the median worker receives as compensation

Log of workers’ median income Ratio of workers’ median income (2011 PPP dollars) to output per worker

12 0.6

Technology diffusion 10 matters not only for 0.4

incomes, but also 8 for addressing other challenges, including 0.2 6 those related to climate change

4 0

7 8 9 10 11 12 7 8 9 10 11 12

Productivity (log of output per worker, 2011 PPP dollars) Productivity (log of output per worker, 2011 PPP dollars)

Source: Human Development Report Office calculations based on data for 94 countries from the International Labour Organization.

is not enough to increase wages, as discussed countries—predominately Organisation for earlier.137 But higher productivity can push the Economic Co-operation and Development envelope for greater absolute compensation members with very high human develop- and for a more balanced distribution between ment—have been reducing their carbon workers and capital owners—and much of this dioxide emissions per capita, reflecting more push towards higher productivity depends on efficient forms of production (figure 6.11).139 technology diffusion. Technology diffusion will be key to extending that decoupling to countries at all levels of Technology diffusion matters not only development. for incomes, but also for addressing other challenges, including those related to climate This chapter has examined the distribution change (chapter 5). Technological inequality of enhanced capabilities related to technology. between developing and developed countries There is potential for harnessing technology harms developing countries’ potential to move for convergence in human development. At beyond traditional patterns of production the same time, there is a possibility that these and consumption.138 A significant decoupling technologies end up causing more divergence. of emissions from economic development is Making the right choices and policies, in this taking place, and over the last decade several area and more broadly, are the topic of chapter 7.

218 | HUMAN DEVELOPMENT REPORT 2019 FIGURE 6.11

A significant decoupling of emissions from development has allowed some countries to reduce their carbon dioxide emissions, reflecting more efficient forms of production

Carbon dioxide emissions Change in carbon dioxide emissions per capita, 2014 (tonnes) per capita, 1995­2014 (tonnes)

50 Human development group

40 20 Low Medium High Very high

30 10

0

10 ­10

0 ­20 0.400 0.600 0.800 1.000 0.600 0.800 1.000

Human Development Index value Human Development Index value

Note: Each bubble represents a country, and the size of the bubble is proportional to the country’s population. Source: Human Development Report Office based on data from the World Bank’s World Development Indicators database.

Chapter 6 Technology’s potential for divergence and convergence: Facing a century of structural transformation | 219 Chapter 7

Policies for reducing inequalities in human development in the 21st century: 7.

Policies for reducing inequalities in human development in the 21st century:

Three trends in inequalities in human development are revealed by looking beyond income and beyond averages. They frame the context for policies as we look beyond today to a world of mounting impacts of climate change and revolutionary advances in technology:

• Inequalities in basic capabilities are falling (some quite rapidly) but remain high, with many people still left behind.

Moreover, the pace of convergence is not fast enough to eradicate extreme deprivations, as called for in the Sustainable Development Goals (SDGs).

• Inequalities in human development are growing in areas likely to be central to people over the next decades. Inequality

in enhanced capabilities—those fast-becoming essential as we move to the 2020s—are increasing, both between and within countries.

• Inequalities in the distribution of opportunities between men and women have improved, but further progress may get

harder, as the challenge of gender equality moves from basic to enhanced capabilities. There is even evidence of back- lash in some countries.

This is both a hopeful and sobering picture. So we must act—but how? Hopeful because progress in reducing gaps This chapter proposes a framework for pol- icies that links the expansion and distribution in basic capabilities shows that with appropri- of both capabilities and income. With the ate policies, results follow. Policies have been overarching objective of redressing inequalities insufficient to completely close gaps in basic in both basic and enhanced capabilities, the capabilities, yet it may still be possible to get framework includes two blocks (figure 7.1). on track and eliminate extreme deprivations, The first block (the one on the left in figure 7.1) as pledged in the 2030 Agenda for Sustainable encompasses policies towards the convergence Development. But aspirations are moving. So and expansion of capabilities, looking beyond considering just how to catch up in basic capa- income.1 The policy goals are to accelerate con- bilities is not enough: Reversing the divergence vergence in basic capabilities while reversing in enhanced capabilities is becoming increas- divergences in enhanced capabilities and elim- ingly important. Turning attention rapidly to inating gender and other horizontal inequali- this task could possibly avoid an entrenchment ties. The timing of many of these policies along of divergences in enhanced capabilities. the lifecycle matters, in relation to when they have an impact over the course of people’s lives. Sobering because the compound effect of The earlier in life some policies are pursued, the emerging inequalities, technological change and less interventions may be needed through other the climate crisis could make remedial actions policies (which may be both more expensive down the road more challenging. We know this and less effective) later in life. from the lifecycle approach that has informed The second block (the one on the right in so much of the analysis in this Report—that figure 7.1) considers policies for the inclusive capabilities accumulate over time, as can dis- expansion of income. The policy objective advantages (chapters 1 and 2). The 2020s will is to jointly advance equity and efficiency in welcome children who are expected to live into markets, increasing productivity that translates the 22nd century, so gaps that would seem small into widely shared growing incomes—redress- in the next few years can be amplified over dec- ing income inequality. The framework is based ades, compounding already large inequalities in income and political power.

Chapter 7 Policies for reducing inequalities in human development in the 21st century: We have a choice | 223 FIGURE 7.1 A framework for designing policies to redress inequalities in human development

Redressing inequalities in basic and enhanced capabilities

Policies to: Premarket Premarket In-market Policies for

• Accelerate convergence Postmarket inclusive expansion
• Reverse divergence in in incomes

enhanced capabilities (productivity and equity)

Source: Human Development Report Office calculations.

The framework is on an integrated approach, because the two (which connect the right block back to the multidimensional, blocks of policies are interdependent. Policies left). Wages, profits and labour participation emphasizing the to advance capabilities beyond income often rates are typically determined in markets, intrinsic importance require resources to fund government pro- which are conditioned by prevailing regula- of indivisible human grammes, which are financed by taxes. And the tions, institutions and policies (in-market). freedoms: Redressing overall resources available are, in turn, linked to But those outcomes also depend on policies inequalities in basic and productivity, which is linked in part to people’s that affect people before they become active in enhanced capabilities capabilities. The two blocks can thus work to- the economy (premarket). Premarket policies is the overarching gether in a virtuous policy cycle. can reduce disparities in capabilities, help- intended outcome ing everyone enter the labour market better The framework is multidimensional, empha- equipped—even though it is important to sizing the intrinsic importance of indivisible emphasize that this is far from the only reason human freedoms: Redressing inequalities in why capabilities matter and that by enhancing basic and enhanced capabilities is the over- capabilities the contributions to expanding arching intended outcome. Thus, it is not incomes go beyond participating in the labour consistent with the reduction of inequalities in market (they can, for instance, enhance polit- some capabilities at the expense of the drastic ical participation). In-market policies affect deterioration of others. Or with approaches the distribution of income and opportunities that either reduce living standards­compro- when individuals are working, shaping out- mising sustainable growth through ill-designed comes that can be either more or less inclusive. distributive policies­or that simply pursue the Postmarket policies affect inequalities once the creation of wealth while violating human rights market, along with in-market policies, have and our planet’s sustainability. determined the distribution of income and op- portunities. These sets of policies interact. The Multidimensionality also allows a better provision of public services premarket may integration of the instrumental analysis of depend in part on the effectiveness of postmar- income and nonincome mechanisms behind ket policies (taxes on market income to fund the formation and progressive equalization of health and education, for instance), which capabilities. The policy cycle can be described matter in mobilizing government revenue to as one composed of premarket policies (pri- pay for those services. And taxes, in turn, are marily within the block on the left of figure 7.1 informed by how much society is willing to on nonincome capabilities and feeding into redistribute income from those with more to the block on the right), in-market policies those with less.2 (predominantly in the right block on inclusive income expansion) and postmarket policies

224 | HUMAN DEVELOPMENT REPORT 2019 A corollary is that considering policies in Towards convergence in Universal policies isolation has limited effectiveness. Take, for capabilities beyond income: From built on extensive instance, recommendations linked to the redis- basic to enhanced universalism coverage only— tribution of income, which tend to dominate without adequate the policy debate. Tony Atkinson simulated the Policies with universal reach speak to the fulfil- resources or designed effect of an ambitious redistributive package ment of the pledge to “leave no one behind” to ensure both quality on income inequality in the United Kingdom, of the 2030 Agenda and to the Universal and equity—are not showing that it would only halve the gap with Declaration of Human Rights.5 Progress genuinely universal Sweden in the Gini coefficient for disposable towards universal achievements has been income and would be insufficient to reverse its remarkable: 91 percent of children attend increase between the late 1970s and 2013.3 This primary education,6 more than 8 in 10 births should not be read as indicating that redistribu- are attended by a skilled professional7 and more tion does not matter—the chapter argues quite than 90 percent of people have access to an the opposite—but decisive change depends on improved water supply.8 These averages may a wider and more systemic approach to policies. hide the prevalence of deprivations (chapter 1) but are massive achievements.9 They did not Using this framework, the chapter has two happen by chance: They were the result of sections, each corresponding roughly to poli- policy choices. This section is about recalibrat- cies associated with the two blocks. The aim of ing ambitions and actions for the 21st century the chapter is to illustrate with specific exam- and for new generations that will see the 22nd ples of policies how the framework proposed century. It starts by arguing that convergence can be used to redress inequalities in human de- in capabilities beyond income should build on velopment—it is not meant to provide a com- these achievements, but be further enhanced. prehensive analysis of all relevant policies. And Such enhancement would call for both politi- given the large heterogeneity across countries cal support (which would require overcoming and the uncertainties associated with future constraints in social choice, as elaborated in pathways (due not only to climate change and spotlight 7.1 at the end of the chapter), as well technology but also to other factors not con- as financial resources (to be addressed in the sidered in the Report4), each country will have second half of the chapter). Beyond enhanced to determine the most suitable set of policies universalism, this section considers policies on for its unique circumstances. eliminating horizontal inequalities (with a fo- cus on gender inequality) and the enhancement The first section discusses how to expand of capabilities for climate shocks and to harness capabilities beyond income, addressing both technology. vertical and horizontal inequalities in human development. It considers both the structure Towards enhanced universal systems and the design of education and health sys- tems, as well as policies related to the emerging Universal policies built on extensive coverage challenges of technology and climate change. only—without adequate resources or designed Among horizontal inequalities, the focus is on to ensure both quality and equity—are not gender equality, responding to the challenges genuinely universal.10 They are useful: They outlined in chapter 4. boost floors, providing access to essential services, and can be credited for some of the The second section addresses policies that can convergence in basic capabilities. But they are jointly lift productivity in ways that are trans- unable to address on their own the persistence lated into widely shared incomes—redressing of inequalities in human development, as man- income inequality. Those policies have a bearing ifested in gradients in achievements. on how markets for goods and services as well as for labour and capital function. The section also This section argues that enhanced universal discusses the effect of redistributive policies at systems (illustrated with services linked to the national level. Because national policies can education and health) could be more effective be constrained or facilitated by globalization, in reducing human development inequalities if the section considers how international col- based on two pillars: lective action—or the lack thereof—can shape inequalities in the 21st century.

Chapter 7 Policies for reducing inequalities in human development in the 21st century: We have a choice | 225 Relying on private, · Comprehensive social services ensuring there are two gaps to address: the gap between feebased schools equal access to quality services in line with poor and rich within countries and the gap for basic education the new demands and aspirations of the 21st between the top achievers in each country and can leave the poorest century.11 As chapter 2 noted, inequality in the SDG target.15 even further behind, human development is multidimensional— due in part to unequal transmitted through different channels, in- Children from lower socioeconomic groups access and lower cluding markets, family networks and social have a double disadvantage—fewer years of accountability for networks—and can be compounded by fac- school and less learning each year. Policies that quality, which tends tors such as violence. Health outcomes, for focus on outcomes rather than inputs require to harm poor students instance, depend on access to services but are data on learning rather than just on enrolment, disproportionally, also socially determined. Enhanced universal investing in children’s mastery of basic concepts systems would incorporate these dimensions. from an early stage and combining overall especially girls improvements with targeted interventions

• Complementary special policies for excluded for groups that are especially disadvantaged.16

groups. Even though poor and marginalized Relying on private, fee-based schools for basic people may benefit from universal policies, education can leave the poorest even further these alone might not be enough to reach behind,17 due in part to unequal access and those furthest behind, including due to lower accountability for quality, which tends to group-based discrimination. For instance, harm poor students disproportionally, especial- children from households facing overlapping ly girls. Free quality public education, improv- deprivations. Leaving no one behind thus ing teachers’ training and enhancing inclusivity, also requires targeted policies addressing especially for girls and disabled students, can horizontal and group inequalities.12 mitigate these risks.18

Ensuring universal access to knowledge Early childhood interventions that can help and lifelong learning flatten gradients are showing results in devel- oping countries (box 7.1). Several countries Policies to ensure equitable access to quality have been expanding coverage in preprimary early childhood education have long-term con- education, with Ethiopia having pushed for sequences for health, cognitive development a significant jump in coverage since 2010 and employment prospects—and they even (box 7.2). This not only is likely to contribute benefit a person’s siblings and children (chap- to equalization of capabilities in the long run ter 2).13 Focusing primarily on providing access but can also affect the distribution of unpaid to education towards a minimum national work, favouring the inclusion of women in the standard has not always closed achievement labour market (as elaborated in the discussion gaps, even in developed countries.14 Given that about gender inequality later in this chapter). SDG target 4.6 calls for all young people to achieve numeracy and literacy skills, even equal Furthermore, technology demands updat- grade attainment between rich and poor house- ing skills throughout life (chapter 6). Lifelong holds in the same country would not necessari- learning would enhance both economic ly ensure that this target is met. In fact, learning and social outcomes and help achieve more achievements in many developing countries are equitable opportunities at every age.19 The below the SDG target even for students from International Labour Organization has made richer families—and children from poorer a concrete proposal on how to implement a households have even worse school attainment. system of entitlements to training, through This implies that simple equalization—lifting reconfigured employment insurance or social up the children from the lowest socioeconomic funds that would allow workers to take paid status to the grade attainment achieved by chil- time off to engage in training.20 Workers dren from the highest socioeconomic status in would be entitled to a number of hours of each country—would not achieve the SDG tar- training, regardless of the type of work they get of quality learning for all. Thus, enhancing do. In countries where most workers work learning outcomes to achieve the SDG target informally, national or sectoral education of universal numeracy and literacy implies that and training funds to provide informal work- ers access to education and training could be established. Policies to reduce informal

226 | HUMAN DEVELOPMENT REPORT 2019 BOX 7.1 Enhancing capabilities in China: Tackling inequality at its roots

In addition to cognitive skills, social and emotional skills from 1.34 (on a scale of 1 to 5) in 2010 to 2.67 in 2014. have been found to mark a productive adult.1 But these For younger children from the richest quintile the aver- skills are often left to the family. While weak social and age score increased from 2.37 to 3.17—less than for emotional skills may be an emerging source of inequal- the other wealth quintiles. Average scores for older chil- ity, they can also be a consequence because the root dren showed a similar pattern, rising from 3.41 in 2010 can lie in inequalities in parents’ education that may to 3.61 in 2014 for children in the lowest quintile and be transmitted to the next generation. But investing in from 3.49 to 3.65 for children in the richest quintile. So, these skills also provides an opportunity to break the inequality in parenting test scores between richer and vicious cycle of inequalities by creating a level start for poorer quintiles almost disappeared.2 all children. China’s improvements are linked to its national China’s scores in positive parenting and socioemo- campaign to promote early childhood development, tional development improved substantially between launched with the United Nations Children’s Fund in 2010 and 2014, especially for children from poorer fam- 2010. The campaign has the ambitious goal of universal ilies. Positive parenting was measured by survey ques- early childhood education. It emphasizes brain develop- tions that ask caregivers how often they intervene to ment in early childhood and provides parenting support enhance their children’s age-specific skills (for instance, through internet portals, websites and mobile phone read to them or play outside with them). Socioemotional applications. It also includes substantial investments development was measured by an assessment of chil- in kindergarten and teacher training, especially in rural dren’s attitudes, behaviour and relation to others. areas and for poor and migrant children in urban areas, and government support for early learning development For younger children from the lowest income quin- guidelines, tools and national standards.3 tile the average positive parenting test score increased

1. Heckman, Stixrud and Urzua 2016.; Kautz and others 2014. 2. Li and others 2018. 3. Greubel and van der Gaag 2012; UNICEF 2019c.

BOX 7.2

Unlocking the potential of preprimary education for advancing human development in Ethiopia

An estimated 50 percent of children in the world are not enrolled in any form Since its introduction the 0-Class has achieved high enrolment rates of early childhood education.1 In developing countries children face even and is now by far the most widely available preschool, especially in rural higher barriers—with only 20 percent enrolment—and often receive lower areas.4 In its first year, the programme enrolled almost three times more quality preprimary education. Sustainable Development Goal target 4.2 calls children than had access to kindergarten in the previous year. Fuelled by for all girls and boys to have access to quality early childhood development, these early successes, further solutions to increase rural enrolment have care and preprimary education by 2030, but the poorest households have the been explored in Ethiopia. The United Nations Children’s Fund and Save the least access to these learning opportunities. Children piloted the Accelerated School Readiness model to reach children who did not attend 0-Class, including children in emergency situations.5 The Ethiopia shows how preprimary education can enable developing model consists of a two-month summer programme before grade 1. Run by countries to improve education outcomes. Starting from one of the lowest primary school teachers and supported by low-cost learning kits, it provides preprimary enrolment rates in the world, just 1.6 percent in 2000, Ethiopia young children with a basic curriculum in preliteracy and prenumeracy. saw the rate rise to 45.9 percent in 2017—representing more than 3 million children.2 Most of the growth was between 2007 and 2017, initiated by the The impacts of preprimary education have been evaluated in multiple National Policy Framework for Early Childhood Care and Education in 2010. case studies in Ethiopia. A Save the Children project on advancing literacy and math skills found that children from lower socioeconomic backgrounds Acknowledging the key role of equitable access to preprimary education achieved significantly higher education gains—practically closing the learn- for human development, a core pillar of the policy framework is the expan- ing gap with their peers from higher socioeconomic backgrounds.6 Young sion of preschool and school readiness programmes.3 Led by the Ministry Lives, an international study of childhood poverty led by researchers at the of Education, the main catalyst for the growth in preprimary education has University of Oxford, followed the education achievements of two cohorts been the “0-Class,” a year of preschool intended for vulnerable households of children between 2002 and 2016 across Ethiopia.7 Urban children who that aims to prepare young children for entry into grade 1, the first year of attended preschool programmes had a 25.7 percent higher likelihood of primary school. Although the ministry had initially considered two years of completing secondary education than their non-preschool counterparts. preprimary education, the plans were changed to broaden access.

1. UNICEF 2019c. 2. UNICEF 2019c. 3. Rossiter and others 2018. 4. Woodhead and others 2017. 5. UNICEF 2019c. 6. Dowd and others 2016. 7. Woldehanna and Araya 2017.

Chapter 7 Policies for reducing inequalities in human development in the 21st century: We have a choice | 227 Political commitment employment could be powerful, since formal insurance to civil servants and formal sector needs to go hand in jobs are associated with larger firms that in- workers. Next was expanding coverage to poor hand with financial vest more in worker training and with longer and vulnerable people, which required a strong resources dedicated employment spells, where more on-the-job political commitment. In Brazil and Thailand to universal health learning can occur. social movements played an important role (see coverage, and box S7.1.1 at the end of the chapter for the role Enabling everyone to lead a long and of social movements more broadly in redressing different countries take healthy life inequalities). different approaches While inequalities in health outcomes are often Political commitment needs to go hand in unrelated to the availability of health services hand with financial resources dedicated to (chapter 2 and box 7.3), universal health cov- universal health coverage, and different coun- erage, a priority in SDG target 3.8, has the tries take different approaches. France used potential to increase equality in health-related earmarked taxes: first a payroll tax and later capabilities.21 Thailand and Rwanda have earmarked income and capital taxes. Brazil and rolled out universal health coverage schemes. Ghana earmark part of their social security In Thailand the policy, implemented in 2001, contributions and value added tax. By contrast, spread to all provinces the following year Japan, Thailand, Turkey and Viet Nam do not and reached 98 percent of the population in have specific amounts earmarked but give it 2011.22 Rwanda has the highest enrolment in budget priority. In addition to financing, a health insurance in Sub-Saharan Africa, with major implementation challenge is the shortage community-based health insurance covering of health care personnel. In many cases private more than 75 percent of the population.23 In and unregulated public health care of variable Bangladesh, Brazil, Ethiopia, France, Ghana, quality may increase sharply. In response, Indonesia, Japan, Peru, Thailand, Turkey Indonesia reformed its accreditation of health and Viet Nam—with a wide range of health professionals and standardized the processes systems and incomes—governments used an for certifying them. Brazil and Ethiopia broad- incremental approach to create and expand ened their health professional recruitment their universal health coverage programmes.24 pools for health extension and offered more The process typically began by providing health flexible career opportunities to community health workers.25

BOX 7.3 The persistence of health gradients even with universal health coverage

Even countries with low income inequality and universal against human papillomavirus, compared with health coverage have not eliminated gradients in health. 40 percent of women from top-income households. Sweden has an outstanding health care system, with · Risky births are more common in poorer families in broad coverage, minimal out-of-pocket costs and spe- Sweden, since more than 30 percent of mothers in cial help for vulnerable groups. But this equal access to the bottom 1 percent smoke before or during preg- health care does not produce equal health outcomes. nancy compared with only 5 percent of mothers in For example: the top group.

• Mortality rates in Sweden are strongly correlated Such persistent inequality in health outcomes can be

accounted for in part by unequal access to health exper- with socioeconomic status. At the bottom more tise outside the formal health system. Some policies that than 40 percent of people die by age 80, compared could mimic family access to health professionals include with fewer than 25 percent at the top. People of long-term visiting-nurse programmes, making more gen- lower socioeconomic status are twice as likely as eral practitioners available and ensuring that more pro- those at the top to suffer from heart attacks, lung viders are culturally compatible with their communities, cancer, type 2 diabetes and heart failure. since this increases trust. Such policies would be even

• Only 10 percent of women from bottom-income more effective if targeted at the poorest.

households in Sweden receive the vaccination

Source: Human Development Report Office, based on Chen, Persson and Polyakova (2019).

228 | HUMAN DEVELOPMENT REPORT 2019 Addressing horizontal inequalities: training or media campaigns against gender Several historical Focus on gender inequality stereotyping. To change incentives, protective examples show that mechanisms can confront possible harm due a combination of Universal policies can provide basic floors but to traditional gender norms or a backlash, such universal and targeted may not be enough to eliminate horizontal as school bullying or workplace harassment. policies can reduce inequalities. The latter are often rooted in Changing incentives can also be introduced horizontal inequalities. long-standing social norms and social exclu- to delay early marriage and reduce teenage But there is also a risk sion. Social exclusion happens when people are pregnancies. The three dimensions (education, that targeted policies unable to fully participate in economic, social awareness, incentives) often reinforce each oth- further reinforce and political life because they are excluded on er, as the examples of policies included in this group differences the basis of cultural, religious, racial or other section suggest. or grievances, since reasons.26 This may mean a lack of voice, lack members receive of recognition or lack of capacity for active For example, Québec’s 2006 nontransfer- benefits precisely participation. It may also mean exclusion from able parental leave for fathers shifted incen- because of their decent work, assets, land, opportunities, access tives so that fathers became more involved in group identity to social services or political representation.27 home caregiving. With new benefits fathers increased their participation in parental leave When there are large horizontal inequalities, by 250 percent,31 contributing to reverse the targeted or affirmative action policies that social norm that expected mothers to take sole directly support disadvantaged groups—for ex- responsibility for care work. And in households ample, the provision of access to credit, schol- where men had the opportunity to use the ben- arships or certain group quotas in employment efit, fathers’ daily time in household work was and education—can complement universal 23 percent higher than in households where policies. Several historical examples show that new fathers did not participate, long after the a combination of universal and targeted pol- leave period ended.32 This example also shows icies can reduce horizontal inequalities.28 But the importance of including men in gender there is also a risk that targeted policies further equality policies. In fact, according to a survey reinforce group differences or grievances, since of Organisation for Economic Co-operation members receive benefits precisely because and Development (OECD) countries on of their group identity. Targeted policies are implementing gender strategies or policies, particularly relevant when a group has clearly almost everyone considers changing men’s and been disadvantaged historically,29 with policies boys’ attitudes towards care activities to be the having a defined timeframe so that they are ap- first priority.33 Yet, even though the importance plied only as long as the targeted group is truly of adequately engaging men and boys in over- disadvantaged. Clear communication about coming gender inequality or addressing their the policies is crucial to prevent grievances and own gender-related vulnerabilities is widely feelings of disadvantage. acknowledged, public policies have yet to fully consider that dimension.34 Given that gender remains one of the most prevalent bases of discrimination, policies Thus, laws and regulations can balance the addressing deep-seated discriminatory norms distribution of care work in households—say, and harmful gender stereotypes, prejudices by increasing the duration of paid parental and practices are key for the full realization leave, as in the Québec example. But only of women’s human rights.30 Policies can target about half of the countries in the world offer social norms directly. Interventions to change paternity leave in addition to maternity leave, unequal power relationships among individuals and half of those offer fewer than 3 weeks for within a community or to challenge deeply fathers and 80 percent offer fewer than 14 rooted gender roles can be achieved through weeks for mothers.35 Moreover, it is not enough education, by raising awareness or by changing for the policy to be gender-neutral; it must incentives. Education and raising awareness are explicitly target men (as in Québec), precisely both based on providing individuals with new because otherwise social norms may prevail, information and knowledge that can foster impeding people from taking leave. In 2007 the different values and behaviours. Such initiatives Republic of Korea started to reserve a year of might include formal education, workplace paternal leave, and by 2014 the number of male

Chapter 7 Policies for reducing inequalities in human development in the 21st century: We have a choice | 229 Balancing the workers who took advantage of it had tripled.36 from violence and discrimination to access to distribution of care, And some countries offer economic incentives public services. But the way in which policies particularly for children, for workers to use leave, as in Sweden, where are designed and implemented is determined, is crucial precisely parents receive a small gender-equality finan- in part, by participation in politics. Thus, af- because much of the cial bonus for every day they use parental leave firmative action quotas that increase minority difference in earnings equally. This way, fathers’ share of childcare participation in politics can result in stronger throughout the lifecycle during the early months or years of a child’s life institutional commitment to equality and is generated before can be increased, which may allow for changes nondiscrimination. Even though Tunisia is a age 40, leading women in social norms around childcare that can be young democracy (its first constitution was to miss many labour reflected throughout a child’s life. ratified in 2014), today it has one of the world’s market opportunities most progressive gender parity laws. It has leg- during the early stages Balancing the distribution of care, particular- islated candidate, constitutional and electoral ly for children, is crucial precisely because much law quotas. The regulations guarantee equal of their careers of the difference in earnings throughout the li- opportunities for women and men at all levels fecycle is generated before age 40, leading wom- of responsibility in all fields and ask candidates en to miss many labour market opportunities to file candidacy applications on the basis of during the early stages of their careers.37 These parity between men and women alternating. missed opportunities coincide with childbirth, By 2018 women occupied 47 percent of local which can encourage women to withdraw from council positions.40 Almost all countries with the labour market. Offering access to affordable high female political representation have such childcare can provide mothers opportunities to enabling measures as positive discrimination make their own work-life decisions, allowing and affirmative action. them to engage in paid work. Mothers tend to adjust their choices around paid work to the Policies can also increase the representation demands of childcare.38 Hence, accessible and of girls in science, technology, engineering affordable childcare is relevant for mothers’ and mathematics (STEM; box 7.4). The Costa freedom to engage in paid work.39 Rican Technological Institute set up a special- ized training centre to build women’s capacity The impact of regulations and laws goes in STEM and entrepreneurship. It celebrated beyond changing the balance of care. Policies the first all-female hackathon in Central are important in areas ranging from protection

BOX 7.4

Girls’ coding choices and opportunities

In Latin America 30 million young people are not in edu- The chosen women face different barriers such as cation, employment or training, and 76 percent of them living on the outskirts of cities and having to spend are women. As an additional challenge, studying is not 2­3 hours to commute to class or growing up be- a guarantee for a bright future for women and girls: Less lieving tech sector jobs requiring mathematics skills than 20 percent of women in the region transition from were beyond their reach. In the courses the women studying to formal jobs.1 learn coding essentials to build websites, apps and games. Classes follow the agile classroom model, Laboratoria is a nonprofit organization established learning as if they were working. When students in 2014 that targets girls from low-income families who near the completion of the training and begin their face major barriers to accessing higher education. It job search, Laboratoria pairs them with mentors from combines applied coding education (including six-month the technology field. Tech companies such as IBM, coding bootcamps), socioemotional training, deep em- Google, LinkedIn and Microsoft have partnered with ployer engagement and job placement services to cre- Laboratoria to increase the supply of female devel- ate opportunities for students. Operating in Brazil, Chile, opers. The companies participating in and sponsoring Mexico and Peru, it has graduated more than 820 girls Talent Fest have first access to Laboratoria’s pool of and aims to reach 5,000 young women by 2021. More talent, but other businesses can pay to browse student than 80 percent of students get jobs as developers, profiles as well. which often triples their incomes.

1. OECD 2017. Source: Human Development Report Office based on Guaqueta (2017), Laboratoria (2019) and World Bank (2013).

230 | HUMAN DEVELOPMENT REPORT 2019 America in 2018, using technology and STEM women and men. For example, SASA!, a pro- For climate change, expertise to bolster sustainable development.41 gramme designed by Raising Voices and first enhanced capabilities Cenfotec University and the institute estab- implemented in Kampala, Uganda, targets tra- encompass those lished a follow-up strategy to create technol- ditional social norms that perpetuate violence that enable people to ogy training spaces and support all women against women. Addressing both women and prepare and respond interested in a STEM career. NiñaSTEM men in households, it approaches the power not only to shocks (GirlSTEM), launched in early 2017 by the imbalance at the individual and structural lev- that have historical Mexican government in partnership with the els by making communities rethink household precedence but OECD, invites women with prominent science relationship dynamics. Today the programme’s also to the more and mathematics careers to act as mentors, results have been widely tested and standard- unprecedented visiting schools and encouraging girls to choose ized, as in Haiti and Tanzania, and it has been disruptions that STEM subjects and be ambitious.42 scaled up to 25 countries.46 climate change is likely to bring about For girls to choose STEM they must be in Towards enhanced capabilities for school. Some interventions can change incen- climate shocks and technology tives for girls to stay in school by either delaying marriage or reducing adolescent pregnancy. Climate change and technology are likely to Cash transfers have been proven to increase shape inequalities in human development over school attendance. The Zomba Cash Transfer the course of the 21st century, as explored in Programme in Malawi, where pregnancy is the chapters 5 and 6. Enhanced capabilities related main reason girls drop out, gave conditional to these two factors are ultimately about how and unconditional transfers to girls in school empowered people are to navigate the challeng- and girls who had recently dropped out. It es and opportunities associated with them in significantly reduced HIV prevalence and preg- the coming decades. nancy and early marriage rates and improved language test scores.43 For climate change, enhanced capabilities encompass those that enable people to prepare As with education, it is important to consid- and respond not only to shocks that have er how women may be uniquely vulnerable to historical precedence but also to the more un- health inequalities because of their sexual and precedented disruptions that climate change is reproductive health care needs. Reproductive likely to bring about. Insurance can help in this health, which gives women agency and con- regard. Article 8 of the 2015 Paris Agreement trol over their own body and fertility, still has of the United Nations Framework Convention much room for progress. In Tigray, Ethiopia, a on Climate Change calls for risk insurance fa- service delivery model that combines commu- cilities, climate risk pools and other insurance nity-based distribution of contraception with solutions.47 That same year the Group of 7 social marketing benefits women and their launched an initiative on climate risk insurance, communities.44 In Bujumbura, the capital of pledging to reach 400 million uninsured peo- Burundi, the government started a national ple in poor countries.48 Insurance, however, has module for comprehensive sexuality education well recognized challenges (such as moral haz- in all schools to empower girls and women ard and adverse selection) that imply the need through awareness of and access to sexual and for appropriate regulation. This also applies to reproductive health assistance and family plan- the design of climate-related insurance systems. ning services—and to provide the community Index-based microinsurance linking payouts to a platform for dialogue on sexual education independently observed weather parameters, and sexual and reproductive rights. The govern- such as rainfall, can address some of these chal- ment has received support from international lenges, while sovereign insurance pools have organizations, including the United Nations also been proposed and implemented.49 Population Fund, to develop the school club model and two manuals for teachers and Still, climate change poses unique challenges students.45 to, and perhaps limits on, the viability and function of insurance if it is difficult to share Finally, social norms mould individuals’ risks. Climate change is expected to affect large behaviours and beliefs about violence against geographies in similar ways. As risks become women. Preventive policies can target both

Chapter 7 Policies for reducing inequalities in human development in the 21st century: We have a choice | 231 The most salient more correlated, the benefits of risk sharing of magnitude,53 including by developing narra- self-reported barrier that insurance affords can become smaller. tives that would facilitate more SDG-focused For instance, the probability that the top four analyses, with climate as one objective among for mobile internet maize-producing countries will experience other SDGs.54 use is limited digital a simultaneous production loss greater than 10 percent is now virtually zero. But as temper- When it comes to technology, chapter 6 literacy and skills: atures increase by 2°C, mean yields drop and highlights the importance of harnessing 34 percent in Africa, absolute variability increases, the probability technological change towards inclusion and 35 percent in East Asia, increases to 7 percent. At an increase of 4°C, it sustainability and the crucial role that “being 37 percent in South reaches 86 percent.50 connected” plays in enabling countries and Asia and 28 percent people to leverage the potential of digital Policies—local, national and international— and artificial intelligence technologies. Even in Latin America thus have a major role in the design and im- though the impact of technology on human plementation of climate-related insurance that development goes beyond access, the discus- includes poor and vulnerable people. Policies sion here illustrates steps that can enhance can support the application of new technolo- capabilities (without suggesting that this is the gies. Drones, for example, have shown promise most important policy response). Chapter 6 in gathering accurate data on weather-related documents divergence in access to advanced damage to crops and property.51 Or insurance communication technologies, which can be premiums could be directly subsidized, and accounted for in part by gaps in relative costs. even means-tested. Reinsurance will also be The Broadband Commission has set a target important for affordable premiums, especially for 2025: entry-level broadband services where insurance is local and climate-related risk (1 gigabyte) at a cost of less than 2 percent profiles are fairly homogeneous. of monthly gross national income per capita. Currently most developed countries, almost The special report of the 2018 Inter half of developing countries that are not governmental Panel on Climate Change least developed countries and a small portion discusses place-specific adaptation pathways of least developed countries have met the as opportunities for addressing structural in- target.55 equalities, power imbalances and governance mechanisms that give rise to and exacerbate Still, the most salient self-reported barrier for inequalities in climate risks and impacts.52 mobile internet use is limited digital literacy But the report warns that such pathways can and skills: 34 percent in Africa, 35 percent in also reinforce inequalities and imbalances. East Asia, 37 percent in South Asia and 28 per- Adaptation narratives built around self-reli- cent in Latin America.56 Indeed, more than half ance, for example, may intensify climate bur- the world’s people lack basic information and dens on poor people and marginalized groups. communication technology skills. There are significant differences across income groups. The special report also lists recent research For instance, in lower-middle-income countries that has linked long-term climate change mit- only 6 percent of adults have sent an email with igation and adaptation pathways to individual an attachment compared with 70 percent in SDGs, to varying degrees. It calls for more developed countries.57 Thus, education for both nexus approaches, which investigate a subset of young and older people will be key to increas- sustainable development dimensions together. ing digital literacy. Examples include a water­energy­climate nexus, leveraging the widely used shared Connectivity can also be enhanced through socioeconomic pathways. Using new methods public Wi-Fi services offered in public facili- for poverty and inequality projections, shared ties such as libraries and community centres. socioeconomic pathway­based assessments Singapore and North Macedonia are two have been undertaken for the local sustainable pioneers. In 2005 Singapore implemented the development implications of avoided impacts Wireless@SG programme to connect citi- and related adaptation needs. zens through a network of hotspots in public and commercial facilities. In 2006 North A focus on sustainable development can re- Macedonia developed a plan to connect 460 duce the climate risk exposure of populations primary and secondary schools and provide vulnerable to poverty by more than an order

232 | HUMAN DEVELOPMENT REPORT 2019 680 Wi-Fi kiosks with free access to internet FIGURE 7.2 services. Indonesia recently launched an ambi- tious plan to have public access across many of Higher labour productivity is associated with a its 17,000 islands by 2022. In the Philippines lower concentration of labour income at the top the Free Public Access Program is expanding connectivity through the country: In 2019, Inequality 2,677 access points were operational, and 6,000 (share of top 10 percent) are expected to be added in a second phase. In Thailand the government is extending con- Human development group nectivity to 4,000 villages. In the Dominican Republic the government is installing 5,000 100 Low Medium High Very high hotspots. In Madagascar the government has started a plan to connect schools and 80 hospitals.58 In fact, access to the internet is so important that it is making its way to being 60 acknowledged as a right. In 2016 the United Nations General Assembly passed a resolution 40 stressing the importance of “applying a com- prehensive human rights-based approach in 20 providing and in expanding access to Internet,” requesting “all States to make efforts to bridge 7 8 9 10 11 12 the many forms of digital divides.” This expan- sion must be consistent with general human Log of productivity (output per worker) Higher labour rights principles, “the same rights that people productivity is have offline must also be protected online, in Note: Includes 94 countries with microdata. associated with a particular freedom of expression.”59 Source: Human Development Report Office based on data from ILO (2019a). lower concentration of labour income at Towards inclusive income FIGURE 7.3 the top. Improving expansion: Raising productivity capabilities across and enhancing equity The relationship between labour productivity and the population concentration of labour income appears to hold also unleashes the Episodes of rapid economic growth and over time at most levels of human development productive potential structural transformation can go along with of a country increases in economic inequality (chapter 2),60 Inequality but higher labour productivity is associated (share of top 10 percent) with a lower concentration of labour income at the top (figure 7.2).61 While the evolution of 70 Low these two variables cannot be inferred simply by looking at a cross-section that represents a 60 snapshot in time, the relationship appears to hold over time at all levels of human develop- 50 Medium ment (except for the Group of 7 economies; figure 7.3). This suggests that pathways that 40 High deliver both improvements in economic per- 30 Very high-G7 formance and labour incomes that are not con- G7 centrated at the top are not only feasible but 8 also common—even if not inevitable, because 9 10 11 12 this evidence does not indicate the direction of causality.62 The challenge, therefore, is to Log of productivity (output per worker) identify those policies that are consistent with a framework of inclusive income expansion.63 Note: Includes 94 countries with microdata. Source: Human Development Report Office based on data from ILO (2019a).

Importantly, environmental sustainability also needs to be considered, especially the climate crisis, which spotlight 7.2 at the end of the chapter addresses.

Improving capabilities across the population also unleashes the productive potential of a country. Discussed here are policies primarily in-market and postmarket that have a bearing on the rate of expansion and distribution of income. The market distribution of income depends on how much people can use their as- sets and capabilities, the return on those assets

Chapter 7 Policies for reducing inequalities in human development in the 21st century: We have a choice | 233 A minimum wage and capabilities, and their ability to respond that enhance women’s participation in the la- can be an instrument to shocks.64 Policies that improve the func- bour market, in a context in which mothers and tioning of markets are thus crucial to increase caregivers are empowered with the conditions of efficiency when productivity, also determining the distribution discussed earlier in the chapter to exercise their there is a monopsony of income. Postmarket policies reflect primary free choice, would clearly achieve both objec- choices associated with government taxes, tives (box 7.5). The remainder of this section (companies with transfers and public spending. This second covers other relevant labour market institutions excessive power in half of the chapter considers policies in these and policies. the labour market) or dimensions. when the economy Monopsonies, minimum increases productivity Balancing power: Equitable and wage and efficiency in response to higher efficient labour markets Another important labour market policy is a labour costs Most people receive income from work (a few minimum wage, which exists in 92 percent of also from capital gains), which is determined to countries.73 As collective bargaining in firms a great extent by how markets are organized and becomes more challenging, broader subnation- regulated. Thus, labour markets and the world al or national negotiations appear to be gaining of work are important determinants of income relevance as a way to protect worker interests.74 inequality. For instance, increases in labour A minimum wage is an instrument to transmit income towards the bottom of the distribution productivity gains to the incomes of workers were central in Latin American countries that with limited bargaining power. But a minimum reduced income inequality in the 2000s.65 wage that is too high can reduce employment or provide incentives for informal employment. Markets are not a baseline on which govern- ments intervene;66 rather, they are embedded Across countries, minimum wages show a in society (to use Karl Polanyi’s expression).67 negative relationship with inequality in labour And market outcomes are shaped by a number income (figure 7.4).75 This association does not of policies and institutions, some of which are prove any causality, but it is consistent with considered in this section. For instance, unions literature documenting that a minimum wage endow workers with the capacity to collectively can, when well calibrated, increase salaries of bargain for their share of income, exercising low-income groups with limited effects on agency and contributing to the outcome of ne- employment.76 The distributive role is linked, gotiations, shaping the distribution of market in turn, to productivity. income.68 Due in part to the fragmentation of production associated with globalization un- A minimum wage can be an instrument of ionization has become more difficult, with the efficiency when there is a monopsony (compa- influence of unions declining in many coun- nies with excessive power in the labour market, tries,69 although with variations by country and as alluded to in chapter 6) or when the econo- over time.70 While the relationship between my increases productivity in response to higher changing inequality in human development labour costs.77 Indeed, monopsony is likely to and changing union density varies across coun- increase inequality, reducing the labour share.78 tries, in practice, promoting equity through The higher the concentration, the greater the stronger unions is consistent with sustained firms’ labour market power to determine wages, gains in productivity.71 given workers’ lack of alternative employment opportunities. In some cases firms can cooper- Policies and institutions underpinned by ate to reduce wages even further.79 Monopsony the respect for human rights determine what is more prevalent when the geographical mo- constitutes illicit labour markets, outlawing bility of labour is low, due either to laws such as practices like slavery, human trafficking, child residency requirements or to low skills of work- labour, human degradation, harassment and ers, which makes them easily substitutable. discrimination.72 But beyond eradicating those practices, how can in-market policies contrib- Public policy can play a key role in such ute to a fairer distribution of incomes without cases. Although opinions are split on whether hurting incentives for productivity? Policies minimum wages reduce employment in com- petitive markets, when labour market power

234 | HUMAN DEVELOPMENT REPORT 2019 BOX 7.5 gender pay gap calculations.5 Currently, equal pay for equal work is constitutionally guaranteed in only 21 per- Gender equality in the labour market cent of countries.6

Women’s contribution to measured economic activity Other examples to improve the quality of working does not correspond to their share of the population: conditions include defining identical criteria to promote It is far below their full potential. This has important men and women, having flexible working arrangements macroeconomic implications. The loss in GDP per capita and increasing the supply of care options to broaden that is attributable to gender gaps in the labour market choices. In Belgium, France, Germany and New Zealand is estimated to be as high as 27 percent in some re- all employees in companies of a certain size are enti- gions.1 Women’s economic empowerment boosts pos- tled to request flexible working arrangements. Japan itive development outcomes, such as productivity, and and the Republic of Korea provide mothers and fathers increases economic diversification and income equality.2 one year of nontransferable paid parental leave each. And Nordic countries often reserve parts of the parental Policies that aim to mitigate gender biases and leave period for the exclusive use of each parent for a guarantee equal pay can promote economic growth few months.7 and could be magnified through a stronger presence of skilled women in the labour market.3 Barriers to wom- It is not enough to adopt these policies if they are en’s participation act as brakes on the national econo- not accompanied by training or awareness campaigns to my, stifling its ability to grow. So implementing policies change gender social norms. For the workplace it is very that remove labour market distortions and create a level important to change attitudes towards caregiving and playing field for all would boost the demand for wom- taking leave from work to care for dependents by men en’s labour—with action also on the supply side to al- so that fathers who take leave are not stigmatized. This low women to exercise their free choice to participate.4 can help balance workloads at home and change atti- These measures range from changes in discriminatory tudes towards gender roles in households. As in other regulations and practices to ensuring gender equality in dimensions, it is critical to engage men. One way is by pay and fairer working conditions for women. establishing male role models to drive changes in gen- der stereotypes. An alternative is to raise awareness Modifying regulations could require employers to through sensitivity training to recognize male privilege, review their pay practices or to report gender gap cal- discern signs of sexism and understand exclusion and culations. Since 2001 both France and Sweden have “micromachismos.”8 asked employers to review their practices and develop an annual plan for gender equality. Australia, Germany, Japan, Sweden and the United Kingdom require organ- izations with 250 or more employees to publish their

1. Cuberes and Teignier 2012. 2. IMF 2018. 3. Agenor, Ozdemir and Moreira 2018. 4. Elborgh-Woytek and others 2013. 5. Australian Government 2019; OECD 2017a. 6. Human Development Report Office calculations using data from the WORLD Policy Analysis Center’s Gender Database 2019. 7. OECD 2016. 8. A series of strategies, gestures, comments and actions of daily life that are subtle, almost imperceptible, but perpetuate and transmit gender-based violence from one generation to another (Gómez 2014).

is concentrated by firms, minimum wages can workers’ wages are marked down by 6 percent or actually increase employment, when the min- more from their marginal product.81 imum wage acts as a price floor, preventing a profit-maximizing firm with monopsony power Minimum wages can also be effective in the from reducing wages through lower hiring.80 context of high informality. A common mis- The positive effect on employment and wages at conception is that the informal sector, since it the bottom is expected to reduce inequalities. has no formal barriers to entry, is more compet- itive than the formal sector. But the difficulty Further efforts to reduce inequalities by check- of enforcing contracts in the informal economy ing the labour market power of firms are ham- can create a holdup problem, where workers pered by the dearth of research and data on the cannot be certain they will be paid once the topic of monopsony, especially compared with work is done. If this happens, employers in research and data on monopoly. An internation- informal markets have considerable power over ally comparable indicator and dataset on labour their workers.82 This would turn on its head the market power would enable monitoring across concern that labour market regulations, such countries and prompt action to reduce it. There as a minimum wage, could increase informal- is ample scope for policy, since in some cases ity. When this mechanism holds, enforcing

Chapter 7 Policies for reducing inequalities in human development in the 21st century: We have a choice | 235 FIGURE 7.4 Productivity Minimum wage: a tool to share the fruit of progress?

Inequality of labour income

Share of top 10 percent Log of output per worker (2011 PPP$) R-squared = 0.653 R-squared = 0.798 11

9

40 8

20 6 7 8 9 10 7 6 7 8 9 10 5 5

Log of minimum wage (2011 PPP$) Log of minimum wage (2011 PPP$)

Note: Includes 60 countries with microdata and observed minimum wage. Data are for the most recent year available. Source: Human Development Report Office based on data from the International Labour Organization’s ILOSTAT database and ILO (2019a).

Platforms generate minimum wages can alleviate the holdup prob- countries, with no major reduction of employ- automatic digital lem by providing a commitment device, which ment.84 But minimum wages apply only to records, so there could increase both efficiency and equity. workers earning wages—often only in the is an opening for formal sector in developing countries, thus minimum wages In India, minimum wage laws had been large- covering a small share of all workers. ly ineffective because the overwhelming ma- under new forms of jority of the workforce has informal contracts To sum up, minimum wages can be a vehi- e-formalization and there is little monitoring or culpability for cle of equity and efficiency if well calibrated employers. But since the mid-2000s the laws to local conditions, including productivity have played an important role alongside right- growth and its distribution in the economy, to-work legislation. The Mahatma Gandhi the presence of monopsony and the level of National Rural Employment Guarantee Act informality. Technological change is affecting promised 100 days of employment per rural those parameters, often raising productivity household, at the official minimum wage, in in combination with monopsony power (see public works generated by local administra- chapter 6). Platforms generate automatic digi- tions. Poor people self-select for the programme tal records, so there is an opening for minimum because it involves arduous physical work at the wages under new forms of e-formalization.85 As minimum wage. It has helped move market noted, whether work happens in the formal or wages closer to the legal minimum, reduce ex- informal sector can matter. ploitative working conditions and protect the rights of routinely discriminated groups such Informality’s challenges as women and workers from Scheduled Castes and Tribes.83 Around the world 61 percent of employed workers (2 million people) are in informal In Sub-Saharan Africa moderately higher employment. The rate of informality is higher minimum wages were correlated with high- in developing countries and emerging countries er economic growth, especially in poorer

236 | HUMAN DEVELOPMENT REPORT 2019 (70 percent) than in developed countries and poverty around the world are urban street The challenge is to (18 percent).86 On average, informal work- vendors and people who work from home pro- open a path to formality ers are poorer, are less educated, have lower ducing for global supply chains. by tackling some of the productivity and lower salaries, and are more structural causes—low vulnerable to shocks.87 They also contribute The challenge is to open a path to formality education and health less to social protection schemes, which is an by tackling some of the structural causes—low and low-productivity obstacle—both from the financial point of education and health and low-productivity sectors—while also view and from the access point of view—to sectors—while also providing options for so- providing options for consolidating high-quality universal systems.88 cial protection, with a flexible mix that might social protection, with combine contributory and noncontributory a flexible mix that might While most informal workers in the world are systems to ensure financial sustainability.91 combine contributory men,89 informal female workers are particularly and noncontributory vulnerable.90 Unpaid family workers, industrial There are different complementary strategies, systems to ensure outworkers, home workers and casual workers given the heterogeneity of conditions facing financial sustainability are predominantly women with low earnings and informal workers. Some countries have a top- a high risk of poverty, while employees and regu- down approach, extending the protections and lar informal workers with higher wages and less benefits enjoyed by formal workers to home risk of poverty are more often men (figure 7.5). workers and other subcontractors. Bottom-up This hierarchy intersects with other horizontal strategies to protect informal workers are also inequalities, such as the marginalization of ethnic possible. Organizing workers, especially poor groups. Groups with high rates of insecure work women, into collectives enables them to pool assets and skills to produce larger quantities of

FIGURE 7.5

Unpaid family workers, industrial outworkers, home workers and casual workers are predominantly women with low earnings and a high risk of poverty, while employees and regular informal workers with higher wages and less risk of poverty are more often men

Poverty risk Average earnings Low High

Employers

Informal wage workers: “Regular”

Own-account operators

Informal wage workers: casual

Industrial outworkers / homeworkers

High Unpaid family workers Low

Source: Chen 2019. Segmentation by sex: Predominantly men Men and women Predominantly women

Chapter 7 Policies for reducing inequalities in human development in the 21st century: We have a choice | 237 Rising market power higher quality goods, acquire new technology financial fees deliver capital gains mainly to the of firms (measured and skills and enhance voice and agency, in- wealthy. In some cases the key increase in finan- by markups) in creasing their bargaining power and increasing cial gains has favoured the top 20 percent of the recent decades has political clout. income distribution—the professional-mana- gone along with the gerial class—rather than the top 1 percent.101 In Technology can help in the move from infor- the euro area, wealth inequality is closely linked reduction in labour’s mality towards better protection for workers. to capital gains on equities (stocks), which ben- share of income Many modern business models rely on the efit the top of the distribution.102 In contrast, collection and use of large amounts of data on credit for productive activities leads to broader and, in many cases, the actions of consumers and workers. Such gains in income for most of the labour force.103 increases in inequality data could improve conditions for informal workers. Apps and sensors can make it easier Productive credit had a positive effect on for companies and social partners to monitor economic growth in 46 countries (both devel- working conditions and labour law compliance oped and developing, including some least de- in supply chains. Governments can invest in veloped countries).104 Combined with the link incubating and testing digital technologies, between credit use and inequality, this evidence including blockchain, that could support social strengthens the case for policies that encourage security payments for those working on labour financing for productive purposes.105 An effec- platforms.92 tive banking and financial sector regulatory framework is also important to the extent that Making finance inclusive it can prevent banking or financial crises—both of which can be very regressive, depending on Financial development can enhance econom- the way the crises are resolved. ic development by reducing asymmetries of information, resolving problems of scale and Antitrust policies for greater equity reallocating capital efficiently.93 Still, ques- tions remain about whether too much finance Rising market power of firms (measured by increases inequality and, perhaps more impor- markups) in recent decades has gone along tant, what type of finance is most inclusive.94 with the reduction in labour’s share of income and, in many cases, increases in inequal- Empirical evidence is mixed. Some studies ity (chapter 6).106 The increase has been led find that financial development reduces ine- by firms at the top 10 percent of the markup quality, especially in developing countries.95 But distribution (figure 7.6), with information and others find that financial deepening increases inequality in both developing and developed FIGURE 7.6 countries.96 Possible channels of increasing ine- quality, beyond the creation of rent by financial The rising market power of firms in recent decades institutions, are the rising compensation of has been led by firms at the top 10 percent of the executives at the top of the distribution and the markup distribution increased indebtedness at the bottom.97 The Bank for International Settlements has revisit- Percent High-markup (top decile) ed the question, focusing on financial structure 1.5 and its relationship to inequality.98 Looking at 97 countries (both developed and emerging 1.2 economies), it found a nonlinear relationship, with financial development reducing inequality 1.3 up to a point and increasing it afterwards.99 Analysing the composition of financial flows provides a more granular notion of finance 1.1 Middle (percentiles 50­90) than simply considering the amount. It also sheds light on mechanisms connecting finan- 1.0 cial growth and inequality besides those assum- ing that all credit goes to productive uses.100 Laggards (bottom half) Dividends, rental income, and interest and 0.9

2000 2003 2006 2009 2012 2015

Source: Diez, Fan and Villegas-Sánchez 2019.

238 | HUMAN DEVELOPMENT REPORT 2019 communication technology­intensive firms in- certain occupations and the legal restrictions Where concentration creasing their markups significantly more than that protect the position of incumbent firms is inefficient, several the rest (chapter 6).107 and regulating monopolies through prices or, policies are available for technology firms, through rules on data to reduce it and its Greater market power for firms can increase ownership, privacy and open interfaces.113 negative impacts on inequality, when shareholders and executives inclusive growth. The accumulate more wealth than workers.108 With the legal principles behind antitrust most basic antitrust Some evidence suggests that antitrust policies law varying by country, global firms face het- policy is the detection could redistribute wealth without the indirect erogeneous regulations. Over the last few years and sanctioning costs of taxation and have a positive effect on European regulators have been particularly of collusion the economy as a whole.109 Market concentra- active in scrutinizing potential anticompetitive tion can affect poor households significantly practices of big tech companies—for exam- (box 7.6). For those with fewer options to ple, the European Commission fined Google diversify expenditure, lower purchasing power 8.25 billion in 2017­2019.114 as a result of anticompetitive practices, such as collusion and monopoly, translates into Fiscal progressivity for reduced capabilities.110 But caution is needed sustainable development when assessing concentration in various mar- kets. An increasing concentration of revenues Redistribution through taxation and public nationally does not necessarily imply more spending is a key determinant of inequality, market power. In many cases geographic mar- not just of income inequality but also of capa- kets for products are local, but concentration is bilities affected by education, health care and measured nationally, so it reflects a shift from other publicly provided services. Several of the local to national firms rather than market pow- policies discussed in the first half of this chapter er. This requires looking at individual markets would likely be making larger claims on public in more detail. Markups are also difficult to resources in many countries. Direct income observe objectively, as different assumptions tax and transfer schemes thus matter not only and measurement methods lead to different because they tend to reduce disposable income results for markup levels and trends.111 There inequality. Spending on in-kind transfers is also a difference between efficient concen- such as education and health can also reduce tration—due to intense price competition, inequalities in capabilities, in turn reducing investment in intangibles and rising produc- income inequality. Importantly, reductions in tivity of leading firms—and inefficient concen- inequalities in income and opportunity can tration—when leading firms are entrenched also reinforce each other. with less competition, higher barriers to entry, lower investment and productivity growth, The effect of redistribution on income in- and higher prices.112 equality can be seen by comparing inequality before and after taxes and transfers (direct and Where concentration is inefficient, several in kind). While the analysis of the impact of policies are available to reduce it and its neg- redistribution can be affected by differences ative impacts on inclusive growth. The most in income concepts and definitions relating to basic antitrust policy is the detection and “before” and “after” taxes and transfers (see spot- sanctioning of collusion. In many countries light 3.3 at the end of chapter 3), the effects can cartels are already illegal, but more resources be sizable. There generally is evidence of larger could be devoted to enforcement. Mergers are effects of redistribution in developed countries another route to market concentration, and than in developing countries (box 7.7). stricter merger enforcement could help tackle rising market power by posing legal challenges Nora Lustig’s fiscal incidence analysis has to mergers that may stifle competition. Policy illuminated several features of the impact of can also prevent dominant firms from using fiscal redistribution in low-income and emerg- their position and network effect to exclude ing economies.115 Her analysis goes beyond their competitors from markets, by investigat- direct taxes and transfers (and pensions), ing such cases more rigorously. Other policies which dominate the literature, to add both include reducing the licencing requirements in indirect taxes and estimates of the monetized benefits accruing from the public provision of

Chapter 7 Policies for reducing inequalities in human development in the 21st century: We have a choice | 239 BOX 7.6 How market concentration can disproportionately affect poor people

A grasp of the distributive effects of competition is central to policymaking. the poverty headcount by 0.8 percentage point and the Gini coefficient by Poorer households are typically the most affected by market concentration 0.32 point (box figure 1).4 because they consume a more homogeneous set of goods, have less op- portunity to substitute consumption and have limited access to markets.1 In mobile telecommunications relative gains are fairly evenly distribut- Inducing more competition in concentrated markets could reduce poverty, ed across income groups. For corn products a decline in market concentra- increase household welfare2 and boost growth and productivity. tion would benefit households at the bottom of the income distribution more (in relative terms), since they allocate a larger share of their consumption to Mexico is well known for its history of monopolies, including Telmex these products. Corn is especially relevant in the diet for low-income groups for fixed-line communications (privatized in 1990) and an oligopoly in corn in Mexico, therefore, for households in the four lowest deciles, moving from products, an important household staple. Plagued by low productivity and a concentrated market to perfect competition would increase their average limited innovation that have resulted in high prices for consumers, these income by 1.6­2.9 percent (box figure 2). By contrast, the increase among monopolies have become an integral part of Mexico’s paradoxical growth, households in the three highest deciles would be only about 0.4 percent leading to an average 98 percent markup in goods across households, ac- (though the absolute impacts increase in higher income deciles). cording to recent estimates.3 Competition-enhancing policies that reduce concentration in key markets One study using the Welfare and Competition tool to simulate the dis- can benefit households. The hypothetical case shows that market concentra- tributional effects of a rise in competition in mobile telecommunications and tion in key sectors of the Mexican economy reduces welfare, especially among corn products in Mexico found that increasing competition from 4 to 12 firms poor and vulnerable households. Moving towards competitive markets, among in the mobile telecommunications industry and reducing the market share of the main objectives of the Mexican government, requires removing market the oligopoly in corn products from 31.2 percent to 7.8 percent would reduce imperfections and economic distortions to enhance economic performance.

Box figure 1 Mexico: Expenditure share in mobile communications and corn, by income decile

Expenditure share by decile (percent)

10 9.3

8 7.6

6.6 6.0

6 5.2 5.0

4.5 4.1 3.6 Decile 8 2.5 4 Decile 9 Decile 10 2

0 Decile 4 Decile 5 Decile 6 Decile 7 Decile 1 Decile 2 Decile 3

Mobile communications Corn

Note: The simulation relies on the assumption that the mobile telecommunication market behaves as an oligopoly and that corn markets mimic a partial collusive oligopoly. The price elasticity of demand is estimated to be -0.476 for mobile communications and -0.876 for corn products. Source: Rodríguez-Castelán and others 2019.

Box figure 2 Mexico: Relative impact on household budgets after moving from a concentrated market to perfect competition by income decile

Relative impact on household budgets by income decile (percent of average income)

4

3.1

2.4 2.1

2 1.7

1.4 1.3 1.1

0.9 0.6 0.3 Decile 8 0 Decile 9 Decile 10 Decile 1 Decile 2 Decile 3 Decile 4 Decile 5 Decile 6 Decile 7

Mobile communications Corn

Source: Rodríguez-Castelán and others 2019.

1. Creedy and Dixon 1998; Urzúa 2013. 2. Atkin, Faber and Gonzalez-Navarro 2018; Busso and Galiani 2019. 3. Aradillas 2018. 4. The reduction in Gini 0.32 point is on a 0­100 scale. See details in Rodríguez-Castelán and others (2019). Source: Based on Rodríguez-Castelán and others (2019).

240 | HUMAN DEVELOPMENT REPORT 2019 BOX 7.7 David Coady, Fiscal Affairs Department, International Monetary Fund The power of fiscal redistribution

Fiscal policy can do much to address inequality in in- dominant factor in recent decreases in income inequal- come and opportunity. A comparison of income inequal- ity.1 From an inclusive growth perspective, expanding ity across advanced and emerging economies shows access to human capital is a win­win prospect. the redistributive role of direct tax and transfer systems (box figure 1). While direct taxes and transfers in ad- Box figure 1 Redistributive direct taxes and vanced economies reduce the Gini coefficient by 0.17 transfers explain nearly all the difference in point (from 0.48 to 0.31), they reduce it much less, by disposable income inequality between advanced 0.04 (from 0.49 to 0.45), in emerging and developing and emerging economies economies, which include Latin American countries with some of the highest income inequality in the world. So, Income inequality Advanced economies on average, the redistributive impact of direct income (absolute reduction in Emerging markets taxes and transfers explains nearly all the difference in Gini coefficient) and developing countries disposable income inequality between advanced and emerging economies. 0.48 0.49

The redistributive reach of fiscal policy is greater 0.45 when the analysis includes the impact of in-kind public spending on education and health. For instance, rising 0.31 spending on education has been instrumental in in- creasing access to education and reducing inequality of Before After education outcomes. As more educated cohorts enter the labour market, income inequality decreases as the Note: Emerging markets and developing economies are Argentina, Armenia, inequality of education outcomes falls and the higher Plurinational State of Bolivia, Brazil, Chile, Colombia, Costa Rica, Dominican human capital stock leads to a reduction in returns to Republic, Ecuador, El Salvador, Ethiopia, Georgia, Ghana, Guatemala, high skills. The decline in education outcome inequality Honduras, Indonesia, Islamic Republic of Iran, Jordan, Mexico, Nicaragua, reduced disposable income inequality in emerging and Peru, Russian Federation, South Africa, Sri Lanka, United Republic of developing economies over 1990­2005 by an estimat- Tanzania, Tunisia, Uganda, Uruguay and Bolivarian Republic of Venezuela. ed 2­5 Gini points on average (box figure 2). In Latin Source: Based on IMF (2017a). America improved education outcomes have been the

Box figure 2 Absolute decrease in Gini for disposable income due to reduced inequality in education outcomes

(absolute decline in Gini for disposable income, 1990­2005) 6

5

4

3

2

1

0 Latin America Middle East Sub-Saharan Advanced Emerging Asia and and the and North Africa Africa economies Europe the Pacific Caribbean

Source: Coady and Dizioli 2018.

1. Azevedo, Inchauste and Sanfelice 2013.

Chapter 7 Policies for reducing inequalities in human development in the 21st century: We have a choice | 241 From the perspective health and education services (which consume Hence, all the countries included in the study of fiscal effort, many much more government resources than either had room for more redistribution.120 direct transfers or pensions). It confirms that countries have the fiscal redistribution is a powerful tool to redress But tax rates have been declining. For exam- scope to increase income inequality.116 Net direct taxes and gov- ple, the top marginal personal income tax rate ernment spending on health and education are has tended to decline in both developed and redistribution by always equalizing forces (measured as the mar- developing countries over the past few decades increasing tax revenues ginal contribution to reduce inequality). Even (figure 7.7). Corporate income taxes have also indirect taxes equalize more often than not. fallen since 1990, in both developed and devel- The equalizing effect of health and education oping countries.121 spending (including tertiary education in some countries) is particularly relevant: Not only are Several domestic factors might explain to- they a more powerful equalizing force, but they day’s low tax rates (chapter 2).122 And tax com- also bolster human development capabilities.117 petition among countries may also have been a factor, especially for corporate income taxes, as The impact of fiscal policies varies considerably discussed below. across countries. This variation can be explained by differences in the size of the taxes and transfers Recent policy debates have returned to taxes budget—that is, fiscal effort—and differences in on wealth, intended to both raise public revenue the progressivity of taxes and transfers—that is, and lower inequality (by flattening the wealth fiscal progressivity (see also spotlight 7.3 at the gradient and by using the funds raised for pub- end of the chapter). lic social services expenditure or infrastructure investment). The advantage of taxing wealth, From the perspective of fiscal effort, many especially real estate, is that it is harder to hide countries have the scope to increase redistribu- than income, to a point. Wealth taxation is also tion by increasing tax revenues. A recent study on very progressive due to the very high concentra- whether (personal income) tax rates are optimal tion of wealth at the top. However, the reporting for maximizing revenues, which depends on how of wealth could fall by as much as an estimated responsive revenues are to taxes, found that tax 15 percent in response to such a tax. And of 12 rates were significantly below optimal levels in countries with a wealth tax in the 1990s, only 3 all the countries examined, implying that they (in Europe) still have the measure in place.123 This could raise tax rates and still increase revenue.118 is due partly to concerns about efficiency and Some studies have also found that the decreasing potential distortive effects on the economy.124 progressivity of taxation in many countries was The OECD recommends a low tax rate targeted not associated with higher economic growth.119 at the very wealthy, with few exemptions and the possibility of paying in instalments.125

FIGURE 7.7

Top personal income tax rates have declined around the world

Percent Low-income developing countries Emerging markets Advanced economies

1974 1978 1982 1986 1990 1994 1998 2002 2006 2010 2014 2018

Source: International Monetary Fund Fiscal Affairs Department’s Tax Policy Reform Database.

242 | HUMAN DEVELOPMENT REPORT 2019 However, analysis of progressivity must go mix of policies to pursue to redress inequality. A more integrated beyond the progressivity of individual taxes— What is clear is that the social value of redistri- global economy also or even aggregate taxes. It is not enough to look bution increases where inequality is higher (see requires international solely at the progressivity of individual tax rates spotlight 7.3 at the end of the chapter). cooperation and rules because fiscal systems are designed with both to ensure fair play revenues and expenditures in mind. The pro- New principles for and to avoid a race to gressivity of net transfers is more informative international taxation the bottom in taxes than the progressivity of the individual taxes and transfers. For example, even an efficient but Globalization and the increased integration of regressive tax—such as a typical value added countries have meant more than just increased tax—can be equalizing if it is complemented by flows of goods, services, finance and people. transfers that target poor people.126 Decisions by corporations on how they struc- ture their supply chains can shape investment, Assessments of fiscal redistribution should production, trade, migration and taxation thus consider both taxation and spending to- around the world. Global value chains define gether.127 Public policy can also maximize the modern manufacturing production especially impact of redistribution through deliberate and in recent decades have been accompanied design of how resources are allocated to different by the distribution of research and develop- groups in society and to different areas of spend- ment129 and other segments of the value chain. ing. Fiscal policy should tilt towards greater Multinational corporations distribute activities spending on the lower deciles, through more in cities and countries to take advantage of transfers (both direct and in kind) to the lower differences in costs, availability of skills, inno- deciles or through greater spending on pro- vation capabilities and logistical advantages. grammes to support disadvantaged groups and communities. Investments in public goods— Evidence suggests that the domestic spillo- including the education system, infrastructure, ver of global value chains have contributed to sanitation and security—could also dispropor- significant gains in productivity and incomes tionately benefit people in lower deciles who in many economies.130 There can also be an would otherwise not have access to such services. association with increasing inequality in some developing countries, through the skill premi- Regardless of the type of tax, support for um, and in developed economies, if jobs are redistribution has strengthened since 1980—at displaced.131 So a more integrated global econ- least in OECD countries. The OECD’s new omy also requires international cooperation Risks that Matter survey asked more than and rules to ensure fair play and to avoid a race 22,000 people in 21 countries about their per- to the bottom in taxes (particularly corporate ceptions of social and economic risks, how well income taxes), disclosure and regulations.132 they think their government addresses those risks and their desired policies and preferenc- Thus, international tax cooperation must en- es for social protection. In almost all OECD sure that transparency is maintained in order to countries more than half the respondents— detect and deter tax evasion; that multinational especially older and low-income ones—think corporations are prevented from shifting profits their government should do more for their eco- to no- or low-tax jurisdictions; that countries nomic and social security, though this does not can get their fair share of taxes, especially with necessarily imply support for higher tax rates.128 the advent of new digitally intensive business models; and that countries, particularly devel- In sum, redistribution can be a powerful in- oping countries, can develop capacities to deal strument to redress inequalities in both income with these challenges.133 and capabilities. Fiscal effort is one part of this tool. The other side of redistribution is fiscal Wealthy people can use offshore centres to progressivity, how net transfers are allocated— hide their money and reduce their tax burdens. to whom they are transferred and how and on The wealth of individuals in offshore centres what public services they are spent on and for in 2014 was an estimated $7.6 trillion, more whose benefit. Decomposing these two aspects than the capitalization of the world’s 20 larg- shows great variation—and thus suggests mul- est companies or the accumulated assets of the tiple options for countries to consider—in the wealthiest 1,645 people (figure 7.8). In April

Chapter 7 Policies for reducing inequalities in human development in the 21st century: We have a choice | 243 FIGURE 7.8 of Information for Tax Purposes (Global Forum). Besides exchanges of information on Offshore wealth is bigger than the value of top request, a significant step towards tax trans- corporations or of billionaires parency has been achieved through automatic exchange of information frameworks such as $ trillions Comparable to the Common Reporting Standard under the 7.6 Global Forum and the US Foreign Account Tax Compliance Act. The first wave of auto- 5.9 6.4 matic exchange of information reporting in 2017, and the bulk following in 2018, allowed Offshore Market capitalization Assets of the information on 47 million offshore accounts— wealth with a total value of around 4.9 trillion—to be of the 20 largest wealthiest exchanged for the first time.

global companies 1,645 billionaires Also stepping up is global coordination to combat base erosion and profit shifting by Source: Based on Zucman (2015), Forbes and the FT 500. corporates, most notably through the Group of 20­OECD BEPS Project.The project address- International tax 2016 the Panama Papers offered a glimpse into es tax avoidance by establishing internationally rules also need the extent of the problem. The fiscal cost to na- agreed standards backed by peer review process- to be modified to tional governments has been estimated at more es to root out harmful tax practices and ensure capture new forms than $190 billion a year.134 that profits are taxed where the economic activ- of value creation ities giving rise to them are conducted.139 It in- in the economy And because capital is mobile, large multi- cludes the review of preferential tax regimes by national corporations often have an advantage the Forum on Harmful Tax Practices.Where a over national governments in determining how regime is assessed by the forum as harmful, the much and where they pay their taxes. In August jurisdiction is required to amend or abolish the 2016 the European Commission determined regime or face being put on blacklists, which that the effective corporate tax rate Apple paid could come with punitive consequences. Many was 0.005 percent in fiscal year 2014, thanks jurisdictions have since amended their tax laws to a special tax regime in Ireland, where profits in line with the internationally agreed stand- from sales across Europe could be recorded.135 ards under the project.

In 2015 an estimated 40 percent of the prof- International collaboration and collective ac- its of multinational firms globally were attrib- tion have thus addressed harmful tax practices uted to no- or low-tax jurisdictions.136 In some and enhanced tax transparency. But more needs low-tax jurisdictions, too, government revenues to be done. Corporates and wealthy individuals have increased as tax rates have fallen.137 Where bent on evading or avoiding taxes will continue the profits thus attributed are not generated to exploit loopholes in the current international by underlying economic activities, the practice tax framework. For example, individuals could is harmful. In such cases governments in the use residence and citizenship by investment countries where the underlying economic activ- schemes, often referred to as “golden passports,” ities are conducted lose tax revenue. Moreover, to avoid disclosure of their offshore assets.140 the firms are not shifting productive capital— Potential tax evaders could also hide wealth which could raise wages and reduce inequality in cryptocurrencies and physical assets, which in the receiving countries—but shifting profits the automatic exchange of information frame- on paper. The benefits to such countries are work does not currently cover.141 Information typically narrowly concentrated. exchanges can also be asymmetrical, with ju- risdictions collecting more information from Significant efforts have been made in the last overseas on its own taxpayers but sharing little decade to combat tax evasion138 by wealthy the other way.142 individuals, most notably through the par- ticipation of more than 100 jurisdictions in the International tax rules also need to be mod- Global Forum on Transparency and Exchange ified to capture new forms of value creation in the economy. With digitalization, firms today

244 | HUMAN DEVELOPMENT REPORT 2019 no longer need to maintain a physical operat- countries in how long and how healthily people This Report intends ing presence to sell their goods and services. can expect to live, how much they can learn and to help policymakers Business models based on digital networks can how high their overall standard of living can and stakeholders also generate value through active and mean- be. Some of the gaps are shrinking, especially everywhere ingful interactions with a vast consumer or in basic capabilities such as life expectancy at understand the user base. Some take the view that jurisdictions birth, access to primary education and basic challenges they where users are located should be allowed to connectivity through technologies such as mo- confront with long- tax a proportion of those businesses’ profits.143 bile phones. But not fast enough: The world in standing and new Discussions at the Group of 20 and OECD not on track to eradicate basic deprivations by inequalities in human have also expanded beyond digitalized business 2030. And in the meantime, gaps in enhanced development—and to include broad-based changes to the entire capabilities are growing—life expectancy at the options available economy to reallocate profits and taxing rights older ages, access to higher education, advanced to address them to market jurisdictions.144 skills and the use of frontier technologies.

Any major revisions to international rules on It is possible to reduce inequalities in human corporate taxation should be shaped by clear development in a sustainable way. Because principles. A fair playing field is needed to tack- each country has its own specifics, there is no le tax avoidance without reducing the incentive universal route. While the impacts of climate for countries to invest in their competitiveness change and technology are universal, they also and capabilities for value creation and without vary in how they affect countries. Thus, various losing the substantial efficiency gains brought elements are needed to design a country-spe- by global value chains. cific path based on a diagnosis of the drivers of inequality along each of the dimensions Beyond tax rules aimed at new business mod- considered in this Report (and others). Among els, a further option being debated is anacross- the array of policies available in each dimen- the-board minimum tax rate.145 Differential sion, countries need to choose ones that are tax rates might also be used to stimulate invest- most appropriate and politically feasible. Their ments to fight climate change.146 Developing choices should be driven by a pragmatic view of countries should have an active presence in what could work given their context and insti- these definitions. The Inclusive Framework tutions. Those at the bottom of the distribution on BEPS is an effort in that direction, but the of income or capabilities care about narrowing United Nations remains a far more inclusive the difference with those at the top, not about forum for these deliberations. The principles of the policy used. So countries need to measure, efficiency and equity, from a global perspective evaluate and, when needed, adjust. this time, must be central in this debate. Much can be done to reduce inequalities in Postscript: We have a choice human development. This Report intends to help policymakers and stakeholders everywhere Big strides have been made in advancing human understand the challenges they confront with development and in enhancing capabilities over long-standing and new inequalities in human the past three decades. But progress has been development—and the options available to ad- uneven. Large gaps exist between and within dress them. There is nothing inevitable in how these inequalities will evolve in the 21st century.

Chapter 7 Policies for reducing inequalities in human development in the 21st century: We have a choice | 245 Spotlight 7.1

Addressing constraints in social choice

A full-fledged universal system is demanding. Figure S7.1.1 identifies three schematic tra- Even if resources are available, reducing inequal- jectories for extending both the coverage and ities in human development is a social choice. the quality of social services, describing some of Politics and context matter. They have interests the political challenges potentially associated and identities. Elements conditioning choices in- with each: clude history and social norms, the prevalence of · Top-down extensions of benefits associated inequality, and the overall resources available and competing claims on their use.147 Social norms, with a small formal workforce may be dif- in particular, are hard to change.148 Even with ficult to implement because those already legislation setting equal rights, society might benefiting (at the top) have little incentive to close and open doors selectively. This Report’s extend services to those below them if they analysis of gender inequality shows that reac- fear that it will reduce quality. Instead, they tions often become more intense in areas where may press to expand the benefits they already more power is involved, potentially culminating have, even if this requires higher payroll con- with a backlash towards the very principles of tributions. They often have the resources to gender equality (chapter 4). Explicit policies for opt out. destigmatization and recognition of low-status · Starting from the bottom of the income lad- groups are relevant to reduce inequalities.149 der can also be challenging if the middle class avoids using services perceived as tailored One challenge in several developing countries for poor people, preferring to use market is how to enhance the existing coverage and options instead. The upper middle class can quality of services already provided to those also oppose financing services that benefit at the bottom. In many cases this challenge other groups. emerges after targeted programmes, such as · Starting with a unified system that initially conditional cash transfers, have already pushed covers nonpoor but vulnerable individuals forward advances in basic capabilities. 150 Those such as formal workers with low wages, higher up the income ladder may have expand- policies can then be expanded upward and ed their access to enhanced capabilities in the downward, as long as there is an emphasis meantime. The middle class may be caught in on quality (thus providing incentives for between. What could be the next steps? high-income individuals to participate, while

FIGURE S7.1.1

Strategies for practical universalism in (unequal) developing countries

Top-down Bottom-up Lower middle-up and trajectory trajectory

Wealthy and high Low High

Hard to expand, as it would Effective to address urgent needs. Relatively high quality can attract compromise quality. But hard to expand because of high-income groups to join middle resource constraints and because class. This might be used to finance low quality does not attract expansion to the poor (interclass participation of the middle class. alliance).

Source: Human Development Report Office, based on the discussion in Martínez and Sánchez-Ancochea (2016).

246 | HUMAN DEVELOPMENT REPORT 2019 allowing expansions to poor people). This In developing countries the challenge is to so- approach, successful in Costa Rica, reduces lidify social policies for a still vulnerable middle the risk of creating different programmes for class. In Latin America there is evidence that poor and nonpoor people. the middle class pays more than it receives in In the end the road to universalism may social services. That, coupled with perceptions of depend on a combination of the three trajec- low-quality education and health services, feeds tories, specific to each context. For instance, resistance to further expanding social policies.152 countries where social insurance reaches less One consequence is the preference for private than 20 percent of the population require a providers: The share of students going to private very different policy trajectory from those school for primary education in Latin America where social insurance reaches more than rose from 12 percent in 1990 to 19 percent in 60 percent. Building broad support requires 2014.153 The larger the share of the private sector, that revenues be generated from a diversity the larger the segmentation in social services for of sources, including copayments for those different groups.154 A natural response would be who can afford them, payroll contributions to add resources from those at the top. But those (depending on the proportion of formal work- groups, while a minority, have often been an ob- ers) and general taxes. In countries with deep stacle to expanding universal services, using their horizontal inequalities, it is also important to economic and political power through structural create stakeholders in different communities and instrumental mechanisms (figure S7.1.2).155 and to avoid the identification of services with specific groups. What to do about all this? Overcoming a In developed countries the challenge may be narrative of tradeoffs between efficiency and to keep social policies that provide enhanced redistribution would be a first important step capabilities to the broadest base. Those systems because gains in equality in human develop- are sustainable to the extent that they work for ment and productivity can march together most of the population, and particularly for the under some policies. Strengthening the capac- middle classes. That has been eroded recently in ity and autonomy of the state to reduce the some OECD countries, where the middle class ability to turn economic power into political perceives itself as progressively left behind in power could also help—through transparency, real income, affordable access to quality educa- promotion of a free independent press and tion and health, and security.151 opening of space for a range of actors to act and engage in productive social dialogue.156

FIGURE S7.1.2 Power of the economic elite and action mechanisms

Structural - Threat of withholding investment as a power response to state decisions

Instrumental - Lobbying power - Control of the press

• Funding of electoral campaigns and/or political parties
• Creation of pro-business political parties
• Promotion of “revolving doors” for politicians
• Promotion of pro-business think tanks

Note : “Structural power” comes from the elite’s control of business decisions and its influence on investment—and economic growth. “Instrumental power” refers to the private sector’s active engagement in the political process through lobbying, publicity, and other tools that many other members of society may not have. Source: Adapted from Martinez and Sánchez-Ancochea (2019), based on Fairfield (2015) and Schiappacase (2019).

Chapter 7 Policies for reducing inequalities in human development in the 21st century: We have a choice | 247 BOX S7.1.1 Ben Phillips, Being right is not enough: Reducing inequality needs a movement from below author of the forthcoming book How to Fight Inequality (Polity Press, October 2020)

It is a remarkable achievement. Just a few years ago there was no consensus Naidoo, who led the trade union movement that helped bring down apartheid that inequality needed to be tackled. Now inequality is recognized as harm- in South Africa, emphasized that “power is built at the grassroots, village by ful and dangerous by mainstream economists, the International Monetary village, street by street.” Organizing is not just about marches. It is about the Fund, the Organisation for Economic Co-operation and Development and whole process, about what happens between the most visible moments. It is the World Bank. And all governments have, in adopting the Sustainable about people forming groups so they can be strong enough to act and be harder Development Goals, pledged to reduce inequality. to ignore, suppress or exploit because they have collective power. In Nepal the Mahila Adhikar Manch, a grassroots women’s movement, started as communi- But winning on words has not meant winning on action. Inequalities ty- and district-level women’s forums, organizing local campaigns on violence continue to worsen, and the broad thrust of government action is at best against women. After six years of grassroots actions, community leaders came insufficient to address them. The mainstream consensus has shifted to rec- together for two days’ deliberations and formed a national secretariat. Since ognize the inequality crisis without a sufficient shift in action. The problem then Mahila Adhikar Manch has grown to be a membership-based organization in beating inequality is not being unsure of what needs to be done; it is not that has spread to more than 30 districts with 50,000 members. gathering the collective power to overcome those stopping it. Old divides across groups need to be broken down to form a winning coali- Some leaders made commitments to tackle inequality without a determined tion. The Usawa (“equality”) Festival in Nairobi deliberately brings together rural intention to implement them, but even when leaders are more inclined to effect and urban, young and old of all communities in a common celebration and plan- change, they cannot act without the wind at their back that ordinary people, ning process, because only by breaking down barriers and building community when organized, can give them. Remember the story of US President Lyndon can it build the unity needed for change. So too the dividing line between unions Johnson telling Martin Luther King, Jr., “I know what I have to do, but you have and social movements has never been wide when they have been at their most to make me do it.” Politicians are under so much pressure from the ever more effective. The movement in El Salvador to protect water as a public good has powerful 1 percent that even the best-intentioned ones need pressure. been effective, its leaders note, only because it brought together such a broad range of the church, social movements, academics, resident groups and non- Inequality is so hard to break because it is a vicious cycle. The power im- governmental organizations—a narrower coalition would not have been strong balance that comes with the concentration of wealth—and its interaction with enough to win. William Barber II calls these movements “fusion coalitions” politics, economics, society and narrative—enables the further concentration because their power comes from bringing so many different groups together. of wealth and a worsened power imbalance. The imbalance of power is what matters for fixing the injustice. As history shows—in the birth of the European Build a new story welfare state, the US New Deal and Great Society, free education in Kenya, the The third lesson is to build a new story of society. Previous victories against in- National Rural Employment Guarantee Act in India, free HIV medicines in South equality built one, and a new one is needed again. Such a new story will not be Africa and the declines in inequality in Latin America in the early 21st century— built in policy papers. The Mexican social movement secured the passing of a the momentum for action to tackle inequality requires pressure from below. labour law reform, ensuring domestic workers access to social security and the right to paid holidays, due in part to the popularity of the movie Roma, which How can inequality be beaten again? Three key lessons stand out from has no explicit policy message but moved millions to understand with greater research and observation. empathy the likes of domestic workers. Similarly, a new narrative is needed to shift from the old Millennium Development Goals to the new Sustainable Overcome deference Development Goals, which embody a new vision of mutuality. But it requires The first lesson is to overcome deference. John Lewis, who helped lead a new narrative to bring it alive. Possible parts of the story might assert that a the US civil rights movement, describes how, as a child, he was urged by good society is about the values we want to live by and the relationships we his mother, “Don’t get in the way; don’t get in trouble.” But as a teenager, want to have, that we need a ceiling as well as a floor and that our society and inspired by activists fighting inequality, he realized that making change re- economy are something we build together. In Laudato Si Pope Francis set out a quired him to “get in trouble, good trouble, necessary trouble.” So too with vision of community over competition, dignity over materialism. South Africa’s Treatment Action Campaign for access to antiretroviral med- icines, the Gambia’s Has Decided movement to ensure that the loser of the The shift in recognizing the problem of inequality and the formal commit- election there stood down as promised and Bolivia’s landless workers, who ment to tackle it have been necessary but insufficient conditions for tackling demanded access to land. All were treated as troublemakers before they inequality. Likewise, analysis of the trends and impacts of inequality and pol- were recognized for prompting needed change. So too were the suffragettes, icy advice on how to tackle it are vitally important but not enough. The one who struggled for women’s right to vote. Resistance does not always work, generalizable lesson of social change seems to be that no one saves others; but acceptance never works. And no one gets to initiate major social shifts people liberate themselves by standing together. Change can be slow, and it without being criticized—that is part of the journey to greater equality. is always complicated and sometimes fails—but it is the only way it works. Change is not given; it is won. By overcoming deference, building collective Build collective power power and building a new story, inequality can beat inequality. The second lesson is to build collective power by organizing. As the saying goes, “There is no justice, just us.” But “just us”—organized—is powerful. Jay

248 | HUMAN DEVELOPMENT REPORT 2019 Spotlight 7.2

Productivity and equity while ensuring environmental sustainability

The analysis in this chapter assumes room for dioxide equivalent to $127.8 Only 5 percent of economic growth along pathways that combine greenhouse gas emissions are covered by a car- equity and increases in productivity. But over bon price considered high enough to achieve the next decades countries will face demands the goals of the Paris Agreement.9 About half for different patterns of development to keep of emissions covered by carbon pricing are global warming below 2°C.1 priced at less than $10 per tonne of carbon di- oxide equivalent, well below what is considered So countries may need to recalibrate the tools necessary to fight climate change.10 used to promote both equity and productivity in a more sustainable way, and new opportu- Raising the price of carbon, seen in isola- nities may lay therein.2 The question is how to tion, may be considered regressive since poor make room for the expansion of productivity people generally spend a greater share of their in a way that does not destroy the planet. The income on energy-intensive goods and services consensus expressed by the Intergovernmental than rich people do.11 Some research paints Panel on Climate Change is that the world a more nuanced picture: an inverse U-shape needs to decarbonize the economy, reaching relationship between energy expenditure share net zero emissions by mid-century.3 This and income, leading to suggestions that carbon requires a shift in patterns of consumption, pricing can, on average, be regressive for coun- employment and production and in the struc- tries with an income per capita above roughly ture of government taxes and transfers, with $15,000 but progressive for poorer countries.12 significant implications for the distribution of However, the inequality impact of fiscal re- income and human development. distribution measures should not be seen as piecemeal and isolated from how the collected Take, for instance, carbon prices—either funds are to be used and how the incidence of through a carbon tax or a market-based emis- taxes is implemented, as discussed in chapter 7. sions trading scheme. By raising the relative Nothing mechanical determines that pricing price of carbon-emitting activities to better carbon must be regressive. reflect the social damages of carbon, incen- tives to produce less carbon would be in place. Carbon pricing can, for instance, reduce The United States pioneered successful mar- inequality if the revenues from a carbon tax are ket-based trading schemes for some pollutants, returned to taxpayers according to a budget-neu- notably sulphur dioxide, nitrogen oxides and tral concept called revenue recycling. One study leaded gasoline.4 The largest emissions trading in the United States showed that if just 11 per- scheme for carbon is the European Union cent of revenues were returned to the bottom Emission Trading Scheme, but other jurisdic- income quintile, those households would not be tions are either planning or considering carbon worse off on average.13 The fiscal transfer could pricing as a way to meet their commitments un- be increased, either through cash transfers or tax der the Paris Agreement of the United Nations credits, to reduce inequality as carbon emissions Framework Convention on Climate Change, fell. Reductions in energy subsidies function which represents 55 percent of greenhouse similarly to the introduction of a carbon tax gas emissions.5 Still, only about 20 percent of because both increase the price of fossil fuels. A global greenhouse gas emissions are covered by study in India showed that phasing out energy one of the 57 carbon pricing initiatives either subsidies and returning the government savings in operation or scheduled for implementa- to people in the form of a universal basic income tion.6 Administered across 46 national and would be progressive, significantly benefiting the 28 subnational jurisdictions, these initiatives poorest, who typically spend far less on energy generated approximately $44 billion in 2018, than the richest do.14 up $11 billion from 2017.7 Carbon prices vary widely, from less than $1 per tonne of carbon Where ambitious emission reduction targets are set, carbon pricing can generate sustained

Chapter 7 Policies for reducing inequalities in human development in the 21st century: We have a choice | 249 revenues over decades that could also be spent Organization study projected scenarios of de- on other important areas, such as health and carbonization consistent with limiting global education.15 And to the extent that those warming to 2°C (over preindustrial levels). It investments disproportionately benefit poor found that the net effect on employment by and vulnerable people, inequality in human 2030 would be positive, with 24 million jobs development could also decline. Some revenue created and 6 million jobs lost. Going beyond recycling options reduce inequalities more than the averages also applies to policies: Even if the others.16 So carbon pricing using equity-promot- world is better off in employment, the gains ing revenue recycling options could be a triple and losses are not equally distributed, and some win: a way to reduce carbon emissions, reduce communities will be more affected than others. or avoid climate-related inequalities and reduce The management of that dynamic can be very other inequalities in human development. consequential for human development and for the political sustainability of the process.18 Where opportunities for equity-promoting revenue recycling face real-world constraints, Notes some have argued for alternatives, such as establishing sector-specific carbon prices 1 Some even argue that economic growth objectives may not supplemented by regulation and public invest- be consistent with keeping global warming below 2°C (Hickel ments.17 If higher carbon prices can be assigned 2019). to different sectors or to different products and uses where the rich tend to spend, lower carbon 2 As proposed, for instance, by advocates of strategies such as prices can be set in areas where poor people “green new deals.” See UNCTAD (2019) as well as the work of spend differentially. For a given emissions the New Economy Commission. See also Rodrik (2007). reductions target a portfolio of differentiated carbon prices, direct regulation and investment 3 IPCC 2018. means those with higher incomes will ex ante 4 Newell and Rogers 2003. bear more of the costs of compliance. Such ap- 5 World Bank 2019d. proaches can alleviate some of the undesirable 6 World Bank 2019d. distributional impacts of a single carbon price, 7 World Bank 2019d. especially where the ability to address distribu- 8 World Bank 2019d. tional concerns ex post are limited. 9 World Bank 2019d. 10 World Bank 2019d. The other side of the adjustment is in produc- 11 Grainger and Kolstad 2010. tion and employment. A drastic reduction in 12 Dorband and others 2019. fossil fuels implies the progressive reduction of 13 Mathur and Morris 2012. jobs in those sectors. An International Labour 14 Coady and Prady 2018. 15 Jakob and others 2019. 16 Klenert and others 2018. 17 Stern and Stiglitz 2017; Stiglitz 2019a. 18 See discussion on the management of phasing out jobs in

chapter 5 of UNDP (2015).

250 | HUMAN DEVELOPMENT REPORT 2019 Spotlight 7.3

Variation in the redistributive impact of direct taxes and transfers in Europe

David Coady, Fiscal Affairs Department, International Monetary Fund

While the redistributive impact of direct Ireland, Denmark, Estonia and Latvia have rel- income taxes and transfers in European coun- atively low fiscal effort, this is offset by relatively tries is large, so is the variation in the extent of high fiscal progressivity, resulting in relatively fiscal redistribution across countries. Euromod high overall fiscal redistribution. The relatively data for 28 EU countries in 2016 shows that low fiscal redistribution in Cyprus and Slovakia the social welfare1 impact of redistributive reflects the combination of low fiscal effort and fiscal policy (the extent of fiscal redistribu- low fiscal progressivity. The relatively high fiscal tion) is highest (above 35 percent) in Ireland, redistribution in Finland reflects the combina- Denmark, Belgium, Estonia and Finland and tion of high progressivity and fiscal effort. lowest (below 13 percent) in Greece, Hungary, Slovakia and Cyprus (figure 7.3.1). High progressivity can reflect either of two factors, or a combination. First, high progres- This variation can be explained by differences sivity may reflect a high share of net transfers in the size of the tax and transfer budget—fiscal going to lower income deciles—high targeting effort—and difference in the progressivity of performance. Second, high progressivity can taxes and transfers—fiscal progressivity. On aver- reflect high market (pre­taxes and transfers) in- age, countries with higher fiscal effort have lower come inequality2—high targeting returns, that fiscal progressivity (figure 7.3.2). For instance, is, redistribution has a high social return where while Greece, Italy and Hungary have relatively market income inequality is high. So even when high fiscal effort, this is offset by their relatively countries have the exact same tax and transfer low fiscal progressivity, resulting in relatively low policies in terms of fiscal effort and target- overall fiscal redistribution. By contrast, while ing performance—for example, where every

FIGURE S7.3.1 Median = 0.22 Fiscal redistribution in European countries, 2016

Proportional change in welfare

0

Note: The proportional change in social welfare is the product of fiscal progressivity and fiscal effort.

Chapter 7 Policies for reducing inequalities in human development in the 21st century: We have a choice | 251 FIGURE S7.3.2

Fiscal progressivity and fiscal effort in European countries, 2016

Fiscal progressivity

2.0 Ireland Median = 0.341

1.8

1.4 Latvia Estonia R2 = 0.449

1.2 Belgium 1.0 Finland

0.8 Malta Bulgaria Germany Croatia Median = 0.672

0.6 Czechia Sweden Slovenia Austria

0.4 Romania Spain Italy Slovakia Cyprus France

0.2 Poland

Luxembourg Hungary

0

0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5

Fiscal effort

FIGURE S7.3.3 Fiscal policy Market income inequality and variation in fiscal redistribution Initial market inequality

Absolute difference in welfare impact from median

0

Note: Countries are ordered by extent of fiscal redistribution from figure S7.3.1. Fiscal policy is the combined impact of fiscal effort and targeting performance. Initial market inequality captures the impact of differences in targeting returns due to differences in pre­tax and transfer income inequality. Differences are relative to a reference country with median values for fiscal policy and targeting returns.

252 | HUMAN DEVELOPMENT REPORT 2019 country has the same transfer budget used to fi- Notes nance a uniform transfer—there can still be sub- stantial differences in fiscal redistribution across 1 Derived using constant elasticity social welfare functions countries, reflecting solely differences in market in which an indicator of inequality can be interpreted as the income inequality. On average, 37 percent of social welfare cost of disparities in income distribution. the differences in fiscal redistribution across countries in figure S7.3.1 is due to differences 2 Since there is very little social benefit from redistributing in the inequality of market income. Overall, income in countries where incomes before taxes and transfers high fiscal redistribution—countries to the left (that is, market incomes) vary little across households, it is in figure S7.3.3—is driven predominantly by possible that a country with relatively high fiscal effort and high targeting returns, reflecting high market targeting performance can still have low fiscal redistribution income inequality, rather than by differences in because it has low market income inequality. Conversely, is underlying fiscal policies. This is particularly so it also possible that a country with low fiscal effort and tar- for Denmark, Estonia, Latvia and Lithuania. geting performance can have high fiscal redistribution simply because it has high market income inequality.

Chapter 7 Policies for reducing inequalities in human development in the 21st century: We have a choice | 253 Notes and references

Overview caution. Still, the dynamic of gaps in with data. Ravallion (2018a, 2018b) gross enrolment rates (mostly from life expectancy opening up at older has clarified how these differing administrative data), the figure for “in 1 Sources for most data and factual ages is robust to changes in age (it views emerge, often using exactly the tertiary education” would be 66 per- statements in this overview are remains valid at age 60), and even same data. It depends partially on the cent for very high human development included in the Report but are included though there is some heterogeneity measures of income and consumption countries and 7 percent for low human here where precision or qualifications across countries and over time, the inequality that are used (for instance, development countries. are important. same pattern is broadly confirmed absolute versus relative), as well as 2 Chetty and others 2016. within countries, as described in more the social welfare weight that is given 3 Acemoglu, Johnson and Robinson 2 Estimates for the United States, based detail in chapter 1. to different segments of the population 2001. on Chetty and others (2016). Kreiner, 14 Brown, Ravallion and Van de Walle (the consumption of those who are 4 UNDP 2016. Nielsen and Serena (2018) argue that 2017. living below the extreme poverty line, 5 The discussion in these paragraphs these results overestimate life expec- 15 Stiglitz, Sen and Fitoussi 2009a. for example, has barely budged, even draws from Basu and Lopez-Calva tancy gaps across different income though many have been able to move (2011) and from Sen (1993, 1999). groups because they ignore income Part I above the line). 6 Basu and Lopez-Calva 2011, p. 153. mobility (by their method, the overesti- 7 For example, gaps in life expectancy 7 Rejecting, at the same time, a “grand mation could be as high as 50 percent), 1 Sen (1980), rephrasing the original appear marked in the United States mausoleum [of] one fixed and final list but they also find that these gaps have question: “Equality of what?” across socioeconomic groups, with of capabilities,” (Sen 2005, p. 160), been increasing over time and that the those at the top of the income distribu- especially if the list was derived pri- overestimation is attenuated at higher 2 This despite the fact that formal tion pulling away from everyone else, marily from theoretical considerations ages (disappearing completely at age decompositions of the contributions while those at the bottom have differ- that did not take into consideration the 80). Mackenbach and others (2018) of income inequality to differences in ent experiences, with lower achieve- real concerns and aspirations of the note that health inequalities generally social welfare aggregating utility using ments in less prosperous places, with time. This is the approach also taken increased in Europe from the 1980s different social welfare functions— the degree of prosperity assessed in in this Report. though the late 2000s, with some over time and across countries—show terms of overall level of education, 8 Article 19 of the Universal Declaration narrowing in several countries since that while inequality matters, income income and government expenditures. of Human Rights reads: “Everyone has then. levels and income growth matter See Chetty and others (2016). See also the right to freedom of opinion and ex- much more, even when the degree Case and Deaton (2017). pression; this right includes freedom to 3 This is discussed in more detail in of inequality aversion is high (Dollar, 8 Williams, Neighbors and Jackson hold opinions without interference and chapter 2. Kleineberg and Kraay 2015; Gaspar, 2003. to seek, receive and impart information Mauro and Poghosyan 2017). See 9 Kearl 2018. and ideas through any media and re- 4 As suggested in UN (2019b), which also the discussion in chapter 2 on 10 The historical analysis should be con- gardless of frontiers” (www.un.org/en/ identified reducing inequalities and inequality and economic growth. sidered along with the argument that universal-declaration-human-rights/). promoting capabilities as “entry in preindustrial societies the limited 9 See, for instance, the discussion in points” to the transformations 3 Based on Google’s Ngram count of amount of resources may have deter- Basu and Lopez-Calva (2011). needed to achieve the Sustainable the expressions “global growth” and mined a maximum level of inequality 10 The survival of a child during the first Development Goals. See also Lusseau “global inequality” from 1950 to 2008; consistent with subsistence of those five years of life (historically the main and Mancini (2019), who found that in- “global inequality” overtook “global at the very bottom. See Milanovic, variable determining the cross-section equalities are a key hurdle in achieving growth” around 2002. Lindert and Williamson (2010). variation of life expectancy at birth) the Sustainable Development Goals 11 See, for instance, evidence of the is an entry point to the prospect of across all countries and that reducing 4 Including inequality reduction as a effects of democracy on human having a long and healthy life. It is an them would have compound positive development priority was contentious development in Gerring, Thacker and achievement that does not depend effects on the entire set of Sustainable during the negotiations for the SDGs, Alfaro (2012). Evidence of the effect on the agency of the child, but on Development Goals. in part because of disagreements of democracy on economic growth is social and family conditions. Instead, on what kind of inequality should be found to be positive and significant in sequential survival—one year after 5 Also a premise of the Deaton Review, reflected in the SDGs. As Fukuda-Parr Acemoglu and others (2019). the other—to become an old healthy a multiyear project examining inequal- (2019) argues, the political compro- 12 As suggested in UN (2019b), which adult represents the realization of ities in the United Kingdom (Joyce and mises required to have aspirations to identified reducing inequalities and that ideal. It is the result of social and Xu 2019). reduce inequalities reflected in the promoting capabilities as entry points family conditions, as before, but also 2030 Agenda led to a dilution of the to the transformations needed to of personal agency and empowerment. 6 Atkinson 2015. ambition of some, especially those implement the SDGs. See also Lusseau 11 Sen 1992, p. 45. 7 Deaton (2017) has argued that in developing countries, who had and Mancini (2019), who found that in- 12 Moser 1989. advocated for stronger commitments, equalities are key hurdles in achieving 13 These two drivers of change are governments often do more to increase especially on inequality across the SDGs across all countries and that already a source of public concern. inequality than to reduce it. countries. For a comprehensive review reducing them would have compound See, for instance, Saad (2019) on fear 8 See, for instance, Saad (2019) on fear of the emergence of research and positive effects on the entire set of of climate change and Reinhart (2018) of climate change and Reinhart (2018) policy interest on global inequality, see SDGs. on artificial intelligence and jobs. on artificial intelligence and jobs. Christiansen and Jensen (2019). 14 Crocker 2008, p. 16. 9 Sen 1980. Chapter 1 15 Crocker 2008, based on an analysis of 10 Expression used by Angus Deaton to 5 Deaton 2013a. Sen’s work. place in perspective the evolution of 6 The optimistic view of development 1 These are estimates for people in 16 For instance, inequality in mean years inequalities (Belluz 2015). higher education based on household of schooling is based on a simple 11 To borrow the expression from Deaton progress is not universally shared. For surveys. Since questionnaires are sum that assumes that one year in (2013a). instance, Hickel (2017a, 2017b) argues different for different groups of primary education counts the same 12 UNDP and OPHI 2019. that we are facing a “development countries, there might be heterogene- as one year in secondary or tertiary 13 Many developing countries lack delusion,” given that global inequality ity and biases. Using fully harmonized education, even if these achievements complete vital registration systems, increased and that those left behind are qualitatively different. In particular, so the country-level estimates of life are further apart from the better off. expectancy at older ages used in this On the other hand, World Bank (2018a) Report, drawn from United Nations shows that within-country inequality Population Division official statistics, has fallen in most developing countries are subject to significant measurement errors and should be interpreted with

Notes | 257 it leads to a potential underestimation human development countries during (2018) found no significant changes 75 See, for instance, Anand (2017), of the role of inequalities in tertiary the second half of the 20th century. over time and little gradient. Anand, Roope and Peichl (2016) and education, typically amounting to few- Between 1955 and 1995 the gain was 41 Szwarcwald and others (2016) and Richardson and others (2019). er years than primary and secondary 63 percent higher in very high human Saikia, Bora and Luy (2019) are education. development countries than in low among the first attempt to examine 76 World Bank (2018a) provides an 17 Permanyer and Smits 2019. human development countries. In the the growing disparities in health alternative interpretation. 18 Deaton (2007) warns about how 21st century there has been marked and life expectancy for Brazil and conclusions about inequality can increase: Between 1995 and 2015 the India. Large-scale data, going beyond 77 Deaton 2013a, 2013b. change depending on the definition increase was 223 percent higher in surveys and covering populationwide 78 OECD 2019f. of the indicator. In this chapter, very high human development coun- socioeconomic and health status 79 This view seems plausible in many unless explicitly stated—as in tries than in low human development. information, are urgently needed to the Inequality-adjusted Human The contrast is even starker in relative provide more convincing evidence on cases. Deaton (2013a) discusses how Development Index—comparisons terms. the socioeconomic health gradients some forms of progress are likely to of inequality in human development 29 The discussion is limited to people and fill these knowledge gaps. spread gradually. depart from summary measures. They under age 80 because people rarely 42 See, for instance, Auerbach and others 80 Based on Kuznets’s (1955) seminal compare the achievements across survive beyond age 100. (2017). paper. See a broader discussion in groups (countries, castes, quintiles 30 Bragg and others 2017; Di Cesare and 43 Education is often a variable used for chapter 2. based on living standards and so on). others 2013; Gonzaga and others 2014; direct measurement of social mobility. 81 Milanovic (2016) describes Kuznets’s Comparisons are made with respect to Oyebode and others 2015; Sommer See for instance Narayan and others waves for income inequality but based the original base (typically, percentage and others 2015. (2018) and OECD (2018a). on a wider set of mechanisms of of the population). This serves three 31 UNDESA 2019. 44 While there is endogeneity (enrolment malign forces and benign forces. purposes. First, it expresses progress 32 Estimates for the United States, based ratios are linked to expected years of 82 OECD 2019f. with respect to an invariant base with on Chetty and others (2016). These schooling, one of the four variables 83 Models of the complex relationship be- intrinsic value—the base is linked to results might overestimate life expec- used to calculate the HDI), these tween aspirations and inequality can people across indicators. In the case tancy gaps across different income relationships hold when using other be found in Besley (2017) and Genicot of indicators based on ratios the base groups because they ignore income development groupings in the analysis, and Ray (2017). represents people with access. In mobility. Kreiner, Nielsen and Serena including income. the case of life expectancy the base (2018) argue that overestimation could 45 Heckman 2011b. Chapter 2 represents years of human life. Life be as high as 50 percent. Using data 46 Montenegro and Patrinos 2014. claims should be universal (Anand for Denmark, they also find that gaps 47 See Goldin and Katz (2009) and 1 Deaton 2018. 2018). Second, in the context of across socioeconomic groups have Agarwal and Gaule (2018). 2 Sen 1999. bounded indicators this comparison been increasing over time and that 48 Akmal and Pritchett 2019; UNESCO 3 An analysis based on this year’s satisfies the mirror axiom (Erreygers the overestimation is attenuated at 2019b. 2009), ensuring that conclusions are higher ages (disappearing completely 49 Banerjee and Duflo 2011; Pritchett and Multidimensional Poverty Index (MPI) robust to changes in convention in the at age 80). Mackenbach and others Beatty 2015. shows no correlation between MPI construction of the indicator going (2018) note that health inequalities 50 Bruns and Luque 2015; Filmer and value and income inequality (measured from achievement to shortfall and vice generally increased in Europe between Pritchett 1999. by the Gini coefficient) but a strong versa. Third, in practical terms, they the 1980s and the late 2000s, with 51 Rözer and Van De Werfhorst 2017. correlation between MPI value and avoid extreme sensitivity from variable some narrowing in several countries 52 UN Inter-agency Group for Child percentage loss in HDI value due to bases of comparison. since then. Mortality Estimation 2018. inequality in both health and education 19 World Bank 2018a. 33 Chetty and others 2016. Also, 53 World Bank 2018a. (Kovacevic 2019; UNDP and OPHI 20 The convergence in primary education Finkelstein, Gentzkow and Williams 54 UNESCO 2019b. 2019). is based on between-country and (2019) estimate that moving from a 55 UN Inter-agency Group for Child 4 Recent research has not only concep- within-country comparisons over the 10th percentile location to a 90th Mortality Estimation 2018. tually clarified causal mechanisms but last decade. UNESCO (2019b) presents percentile location increases life 56 UNESCO 2019b. also marshalled supporting empirical similar results over that period but expectancy at age 65 by 1.1 years in 57 World Bank 2019c. evidence. While much of the evidence highlights that over the last few the United States. 58 UNESCO 2018b. is specific to countries with enough years there have been no evidence of 34 Baker, Currie and Schwandt 2017. 59 UNESCO 2019b. data, that the empirical work is tied convergence between countries. 35 Brønnum-Hansen 2017; Kreiner, 60 UNDP and OPHI 2019. to general hypotheses lends universal 21 Deaton 2013a. Nielsen and Serena 2018. 61 Dercon 2001. relevance to the analysis. 22 This analysis is based on simple aver- 36 van Raalte, Sasson and Martikainen 62 Nussbaum 2011. 5 Deaton 2013b. ages. In Statistical table 1 the analysis 2018. 63 Sen 1999. 6 Persistently low mobility along with in- is based on population-weighted 37 Suzuki and others 2012. 64 See the discussion of recognition and creasing income inequality compounds averages and reveals a gap of 18.2 38 Buchan and others 2017. challenges for destigmatization in the disadvantages among people who years. 39 Currie and Schwandt 2016. Lamont (2018). are unable to move up. As Chetty and 23 UNDESA 2019. 40 Majer and others 2011. Murtin and 65 UNDP Chile 2017. others (2014, p. 1) put it, ”[…] the 24 UNDESA 2019. others (2017) assess inequality in 66 See Hojman and Miranda (2018). consequences of the “birth lottery”— 25 UN 2015a. longevity across education and gender 67 Stewart 2005, 2016a. the parents to whom a child is born— 26 Permanyer and Smits 2019. groups in 23 Organisation for Economic 68 UN 2015c. are larger today than in the past. A 27 Consistent results with this divergence Co-operation and Development coun- 69 ECLAC 2018a. useful visual analogy is to envision in life expectancy at older ages are tries. Their estimates of expected lon- 70 Pew Research Center 2014. the income distributions as a ladder, documented by Engelman, Canudas- gevity at ages 25 and 65 by education 71 Eurobarometer 2018. with each percentile representing a Romo and Agree (2010) and Permanyer and gender show that the gap in life 72 Latinobarometro 2018. different rung. The rungs of the ladder and Scholl (2019). Seligman, expectancy between highly educated 73 Hauser and Norton 2017. Alesina, have grown further apart (inequality Greenberg and Tuljapurkar (2016) also and poorly educated people is 8 years Stantcheva and Teso (2018) find has increased) but children’s changes find a disassociation between equity for men and 5 years for women at age that lower perceptions about social of climbing from lower to higher rungs and length of lifespan. 25 and 3.5 years for men and 2.5 years mobility tend to increase preferences have not changed.” 28 Based on data from UNDESA (2019), for women at age 65. This implies for redistribution. 7 Corak 2013. The curve was introduced the absolute gain in life expectancy at that relative inequalities in longevity 74 Cruces, Pérez-Truglia and Tetaz 2013. in a 2012 speech by Alan Krueger age 70 was higher in very high human by education increase with age. For (chairman of the Council of Economic development countries than in low France, Currie, Schwandt and Thuilliez Advisers; Krueger 2012) and in the President’s Economic Report to Congress (US Government 2012) based on Corak’s data.

258 | HUMAN DEVELOPMENT REPORT 2019 8 See, for instance, the seminal dis- the dynamics of social structures, as Duncan, Brooks-Gunn and Klebanov social status). Health measures cussion in Solon (1999) and the more argued in Xie, Cheng and Zhou (2015). (1994), Heckman and Carneiro (2003) include general self-rated health and comprehensive review in Black and 24 The contribution of assortative mating and Phillips and Shonkoff (2000). Cantril’s self-anchoring measure of life Devereux (2011). to levels and changes in income 29 Black and others 2017. satisfaction. inequality varies in the literature. 30 Wilkinson and Pickett 2018. 52 Babones 2008; Curran and Mahutga 9 Corak 2013, p. 85. Blundell, Joyce, Keiller and Ziliak 31 Garcia and others 2016; Heckman 2018; Kim and Saada 2013; Torre and 10 Corak 2013, p. 98. (2018) estimate that, for the United 2017. Myrskylä 2014; Wilkinson and Pickett 11 Brunori, Ferreira and Peragine 2013. Kingdom and the United States, assor- 32 UNESCO 2018a. 2011. Multivariate regressions of tative mating contributed slightly more 33 Similar results have been found for income inequality and life expectancy This conclusion was drawn from two than half of the increase in household Australia, Canada, the United Kingdom as well as income inequality and infant different measures of mobility: inter- earnings for the group between the and the United States (Bradbury and mortality with recent data from coun- generational persistence in income 5th and 95th percentiles in the period others 2015; Heckman 2011a). Genes tries at all levels of human develop- and intergenerational persistence in 1994­2015 (table 2, p. 58). Greenwood can usually explain only part of such ment show that other variables—such education. and others (2014) report a very large divergences. See, for example, Rowe as GDP per capita, education level, 12 For an earlier analysis on inequality of impact of assortative mating on ine- (1994). Environmental influences government health expenditure, ethnic opportunity, see World Bank (2006). quality by simulating what would have affect gene expression, as shown in diversity and, in the case of high and The report found that a quarter of happened to income inequality in the an identical twin study. Raised apart, very high human development coun- all differences in earnings between United States in 2005 if mating had twins already differed by age 3 due tries, democratization—better explain workers can be attributed to similar been random; but they later corrected to different exposure to stimuli in variations in these health indicators circumstances as the ones mentioned these findings as an overestimation their living and learning environments than income inequality (Bernardi and above. (Greenwood and others 2015). The (Fraga and others 2005; Lee and others Plavgo forthcoming). 13 Narayan and others 2018. The corrected estimates are in line with 2018). 53 McEniry and others 2018. The measure for mobility is intergenera- those of Eika, Mogstad and Zafar 34 See, for example, Jensen and Nielsen article examines the relation between tional persistence in education and the (forthcoming) for the United States and (1997) and Khanam (2008). socioeconomic status and health measure for inequality of opportunity other developed countries, which show 35 Akmal and Pritchett 2019. For the defi- conditions for people age 60 or older. is the inequality of economic oppor- that assortative mating accounts for nition of learning profiles, see Pritchett Socioeconomic status is measured by tunity index developed in Brunori, a non-negligible amount of income ine- and Sandefur (2017). education attainment. Ferreira and Peragine (2013). quality, but with other factors playing 36 Bernardi 2014; Bernardi and 54 Chen, Persson and Polyakova 2019. 14 Brunori, Ferreira and Peragine 2013. a greater role. (Hryshko, Juhn and Boado 2013; Bernardi and Plavgo 55 Kuznets 1955. Lewis’s dual model is For a critical literature review on McCue 2017 also find a small effect forthcoming; Blossfeld and others similar in spirit to Kuznets’s, but Lewis equality and inequality of opportunity for the United States). Hakak and 2016; Hartlaub and Schneider 2012; assumes that holders of capital in the focusing on the principles of compen- Firpo (2017) find similar evidence for Heckman and Krueger 2005; Yanowitch modern sector can accumulate wealth sation and reward, see Ferreira and Brazil, showing that the counterfactual 1977. while paying a constant wage to a Peragine (2016). income Gini with assortative mating 37 Bernardi and Plavgo forthcoming. See “reserve army” of workers that are 15 Even in rather equal societies there is would have been slightly slower also Yastrebov, Kosyakova and Kurakin available in the agricultural sector, evidence that children of wealthy par- than it actually was over a period (2018). thus having very different implications ents are well off themselves. Recent of 20 years (see also Torche 2010, 38 Heckman 2010. on income distribution (Lewis 1954). evidence drawn from the wealth who finds an isomorphism between 39 OECD 2010. 56 Kuznets 1955, p. 17. He also con- of adoptees in Norway (Fagereng, assortative mating and inequality not 40 Bernardi and Ballarino 2016; Bernardi sidered the implications of a higher Mogstad and Ronning 2019) and only for Brazil but also for Chile and and Plavgo forthcoming. savings rate, and thus accumulation Sweden (Black and others 2019) sug- Mexico). Further, these studies show 41 Bussolo, Checchi and Peragine 2019; of capital and assets, at the top of the gests that the wealth of the adoptive that the patterns of assortative mating Kramarz and Skans 2014. income distribution, emphasizing the parents was a strong determinant of across income groups and over time 42 Bussolo, Checchi and Peragine 2019. impact of policies and taxes in limiting children’s accumulation of wealth. An vary, and with several other factors 43 Shanmugaratnam 2019. the accumulation of wealth at the important caveat is that these findings driving inequality, it is difficult to 44 Deaton 2013b. very top. The “unravelling” of these pertain to the intergenerational attribute unambiguously the impact of 45 Deaton 2003, 2013b; Galama and Van policies and tax structures in many transmission of wealth, which may be assortative mating to inequality. Still, Kippersluis 2018; Lindahl and others market-based economies is document- different from that of income, which is the evidence strongly suggests that 2016. ed by Piketty (2014), who argues that the focus of this section. assortative mating takes place across 46 See for instance, Almond and Currie the mid-20th century, when inequality 16 Roemer 1998. countries and makes a non-negligible (2011), both on the impact of health was low, was an exceptional period 17 There is consensus among many contribution to income inequality. before age 5 on adult health and on during which institutions curbed the economic thinkers that final welfare 25 For an argument and evidence showing the potential to redress some of the tendency of returns to capital running is inappropriate for assessing distrib- how assortative mating is important negative impacts in an early age later ahead income growth and that the utive justice. Compare, for example, for intergenerational mobility see in life. more normal course of capitalism is to Dworkin (1981), Rawls (1971), Roemer Chadwick and Solon (2002). 47 For an example on how foetuses have a high concentration of income (1998) and Sen (1985). 26 Most of the analysis in this section are affected by pollution, see Currie and wealth at the top—which is the 18 Narayan and others 2018. considers what happens from one gen- (2011). trend that has dominated in several 19 Deaton 2013b, p. 265. eration to the next, but even though 48 Currie 2009. advanced economies since the 1980s. 20 For a historical perspective on health the evidence is contentious in the 49 Case and Paxson 2010; Currie 2009, This would thus be a rejection of gradients in the United Kingdom and literature, the persistence can carry 2011. Kuznets-like arguments grounded on the evolution of political and academic even across multiple generations, with 50 Skelton and others 2011. structural change. debates, see Macintyre (1997). the effects dissipating over time (see 51 Elgar and others 2016. The sample 57 Kanbur 2017. 21 See Case and Paxson (2008). Solon 2018 for a recent review). for this study consists of 1,371 ado- 58 Milanovic 2016. Thus, the recent in- 22 Some evidence suggests that it 27 Regression of respondents’ years of lescents in seven European countries. crease in inequality in many advanced is not only levels of income that schooling on their parents’ highest Measures used to assess socio- economies can be interpreted as the matter; variation in incomes during years of schooling (EqualChances economic status included parent-re- transition to societies adjusting to childhood has adverse health effects 2019). Data are for the 1980 cohort ported material assets and household the joint effects of globalization and (especially on mental health) later in and for the most recent year available. income as well as youth-reported technological change (Conceição and life (Bjorkenstam and others 2017). 28 See, among others, Blossfeld and oth- material assets and subjective social Galbraith 2001). 23 This behaviour does not need to reflect ers (2017), Chevalier and Lanot (2001), status (MacArthur scale of subjective 59 Tinbergen 1974, 1975. rational choices or individual pref- erences but may even be shaped by

Notes | 259 60 In particular for the United States, see 79 Chenery and others 1974; López-Calva discussion on the data’s limitation, 126 UN and World Bank 2018. Goldin and Katz (2009). and Rodríguez-Castelán 2016. see Kennedy and Prat (2019). Around 127 Kelley and others 2015. 80 percent of individuals in the sample 128 Schleussner and others 2016. 61 OECD 2019f. 80 Bourguignon 2003. watch news on television, 40 percent 129 Von Uexkull and others 2016. 62 Acemoglu and Autor 2011; Autor, Levy 81 Lakner and others 2019. read newspapers and only 30 percent 130 Hillesund 2019. 82 Aiyar and Ebeke 2019. Some empirical use pure internet sources. Internet 131 Langer and Stewart 2015; Miodownik and Murnane 2003; Goos, Manning sources are consumed more widely and Salomons 2014. evidence suggests that high income when they are associated with a tradi- and Nir 2016. 63 One of the reasons for contesting inequality can reduce public school tional platform, especially newspaper 132 Scheidel 2017 this theory is the large dispersion in attendance because parents opt to websites. 133 Bircan, Brück and Vothknecht 2017. earnings within, as opposed to across, send children either to work (low 108 Prat 2015. occupations. See Mishel, Schmitt and socioeconomic status families) or to 109 Kennedy and Prat 2019. The authors use cross-country panel Shierholz (2013). private school (high socioeconomic 110 Fake news is defined as “intentionally data (annual observations from 161 64 Jaumotte, Lall and Papageorgiou status families), diminishing support false or misleading stories” (Clayton countries) for 1960­2014. (2013) show that technology accounts for public education and expenditure and others forthcoming, p. 1). 134 Gates and others 2012. For infant for the increase in inequality in devel- per student, which could equalize op- 111 Rodrik 2018. mortality, see Dahlum and others oping countries and that exposure to portunity. Gutiérrez and Tanaka 2009. 112 For a case study of Latin America, see (forthcoming). globalization does not reduce inequal- 83 While Marrero, Gustavo and Juan Piñeiro, Rhodes-Purdy and Rosenblatt 135 Bircan, Brück and Vothknecht 2017. ity, as one might expect if, through Rodríguez (2013) and Aiyar and Ebeke 2016. 136 UN and World Bank 2018. trade, production were to move from (2019) find supporting evidence, 113 Rodrik 2018. 137 Stewart 2016b. developed to developing countries. The Ferreira and others (2018) do not. 114 This paragraph draws on the analysis reason is that countries are also ex- 84 ECLAC 2018a. in World Bank (2017b). Part II posed to financial globalization, which 85 Birdsall, Ross and Sabot 1995. 115 Bernardi and Plavgo forthcoming. Due counters the equalizing effect of trade 86 ECLAC 2018a. to a few cases of extreme outliers, 1 This also limits understanding of globalization in developing countries. 87 Bowles and others 2012. for the multivariate analysis in the whether people at the bottom are get- 65 Bhorat and others 2019. 88 Alvaredo and others 2018. background papers for this Report, ting closer to moving out of poverty. In 66 See Hunt and Nunn (2019) for the 89 Berger-Schmitt 2000. homicide rates were transferred into fact, some evidence suggests that peo- United States. For more evidence, 90 Uslaner 2002. their natural logarithmic form. See ple who remain below the poverty line including Organisation for Economic 91 Uslaner and Brown 2005. also Kawachi, Kennedy and Wilkinson have seen little movement towards Co-operation and Development 92 Wilkinson and Pickett 2011 (data (1999), Pickett, Mookherjee and the line (Ravallion 2016), while many countries, see Autor (2014, 2019). For on trust are from the World Values Wilkinson (2005) and Wilkinson and who have made it over the line remain an extensive review, see Salverda and Survey). When including countries Pickett (2011). poor when using other metrics (Brown, Checchi (2015). with lower human development using 116 This was determined through an Ravallion and Van de Walle 2017), 67 For a sense of the evolution of the data from the Gallup World Poll from interaction effect between the Gini co- vulnerable to falling back (Lopez-Calva debate over time, see Aghion, Caroli 2010 (the year with most coverage), efficient and mean years of schooling. and Ortíz-Juarez 2014). and Garcia-Penalosa (1999), Baymul there is no significant correlation There is no such moderating effect for and Sen (2018), Eicher and Turnovsky (Human Development Report Office low and medium human development 2 Rose (2016) describes many of the pit- (2003), Galbraith (2012), Milanovic calculation). countries. falls of relying on averages to design (2005), Ostry, Loungani and Berg 93 Paskov and Dewilde 2012. 117 Enamorado and others 2016. and implement policy, going as far as (2019), Piketty (2006), Stiglitz (2012) 94 Dinesen and Sønderskov 2015; Leigh 118 Gilligan (1996), as cited by Pickett, suggesting that policies that promote and World Bank (2006). 2006. Mookherjee and Wilkinson (2005). equal access, if guided by the ideal of 68 See, for instance, Banerjee and Duflo 95 Buttrick and Oishi 2017. 119 Kawachi, Kennedy and Wilkinson what would be needed on average, are (2003). Kuznets (1955) starts with a 96 Van Zomeren 2019. 1999. bound to not fully create opportunity lengthy discussion of the ideal data 97 Connolly, Corak and Haeck 2019, p. 35. 120 Alesina and Perotti 1996. for everyone. needed to investigate the relation 98 Connolly, Corak and Haeck 2019. 121 Collier and Hoeffler 1998; Fearon and between growth and inequality, recog- 99 European Commission, Directorate- Laitin 2003. 3 Borrowing from Ravallion (2001). nizing that his requirements sounded General for Research and Innovation 122 Stewart 2005, 2009, 2016a, 2016b. 4 Ferreira (2012) made a similar point, like a statistician’s pipe dream. 2014. 123 Cederman, Gleditsch and Buhaug 69 See Piketty (2006, 2014). Kuznets’s 100 See Ramos and others (2019) for a 2013. See also Stewart (2005). arguing for the importance of using arguments are not inconsistent with study on religious diversity. One of the mechanisms behind this growth incidence curves. Piketty’s assertion, given that Kuznets 101 OECD 2010. was explained a long time ago by 5 Criado-Perez 2019. himself recognized several limitations 102 Lancee and Van de Werfhorst 2012; Horowitz’s (2001) comprehensive study 6 Atkinson 1970, p. 261­262. of his article (for example, that it Solt 2008. Ethnic Groups in Conflict. Ethnicity 7 Ravallion 2018a. excludes government transfers). 103 On the influence of the upper middle is equivalent to the family concept, 8 Anand 2018. 70 Scheidel 2017. class in political processes in the generating solidarity and a strong 9 Coyle 2015. 71 Okun 1975. United States, see Reeves (2018). sense of belonging that can transform 10 Rockoff 2019, p. 147. 72 Lucas 2004, p. 20. See also Gilens and Page (2014), Igan into intense emotional outbursts and 11 See Deaton (2005) as well as Ferreira 73 Cingano 2014; Ostry and Berg 2011; and Mishra (2011) and Karabarbounis sometimes even hatred (Cederman, and Lustig (2015). Ostry, Loungani and Berg 2019. See (2011). Clientelism can be defined Gleditsch and Buhaug 2013). Another 12 Smith and others 2019. also Alesina and Rodrik (1994), Assa as “a political strategy characterized explanation is that groups protest 13 See, for instance, Galbraith (2018). The (2012), Barro (2008) and Stiglitz (2016). by an exchange of material goods in when they perceive inequalities as objections include the observation that 74 Neves, Afonso and Silva 2016. return for electoral support” (World unfair and try to cope with them income tax data are sparse and patchy. 75 See, for instance, Kraay (2015) and Bank 2017b, p. 10, based on Stokes collectively instead of individually When there are large gaps in the data, Bourguignon (2015b). 2009). (Van Zomeren 2019). Sen (2008b, p. 5) assumptions have to be made that are 76 Furman 2019. In discussing Furman’s 104 For a more comprehensive discussion suggests that the “coupling between very significant and open to scrutiny arguments, Rodrik (2019) and on this, see UNDP (2016). cultural identities and poverty” makes (Galbraith and others 2016). Shanmugaratnam (2019) come to 105 World Bank 2017b. inequality more important and may 14 Criado-Perez 2019. support the same basic argument. 106 Chadwick 2017, p. 4. thus contribute to violence. 77 López-Calva and Rodríguez-Castelán 107 Kennedy and Prat 2019. Data are 124 Langer 2005. Chapter 3 2016. from the Reuters Digital News Report 125 Stewart 2009. 78 Mendez Ramos 2019. survey, which covers more than 72,000 1 See, for instance, Zucman (2013, individuals in 36 countries. For a 2015). Also discussed in Alvaredo and others (2018).

2 See also chapter 5 and Chancel (2017).

260 | HUMAN DEVELOPMENT REPORT 2019 3 Zucman 2014. size of Canada as compared with the 46 Theil values are obtained from http:// 66 Alstadsæter, Johannesen and Zucman 4 See UN (2009). United States (implying that different WID.world/gpinter, using the global 2018; Zucman 2014. 5 See Alvaredo and others (2016). assumptions on the distribution of inequality dataset constructed for 6 See Zucman (2019). national income in Canada only have Alvaredo and others (2018); see http:// 67 “Capital” and “wealth” are used 7 See Zucman (2014). a marginal impact on the distribution wir2018.wid.world). interchangeably in this chapter. 8 In India the government stopped of growth in the United States and Canada combined). The two countries 47 Gathered by the World Bank and 68 More details are in Alvaredo and publishing data between 2000 and are merged into a single aggregate. available on PovcalNet. others (2018). 2010 (see Chancel and Piketty 2017). This allows a simple estimate of 9 This section is based in part in inequality in a region that is broadly 48 The values for Africa are based on 69 As for income inequality data, the set Ferreira, Lustig and Teles (2015). comparable in size to Western Europe, interpolation of PovcalNet data (see of countries with available wealth­ 10 Formerly known as the Luxembourg while taking into consideration the Chancel and others 2019, which income ratios is constantly growing. Income Study (www.lisdatacenter.org). main differences in national income includes technical details for this See Ravallion (2015). levels and growth trajectories between section). The values presented for the 70 Alvaredo and others 2018; Garbinti, 11 Gasparini and Tornarolli2015. the United States and Canada. See Americas, Asia and Europe are based Goupille-Lebret and Piketty 2016; Saez 12 Galbraith 2016. Chancel, Clarke and Gethin (2017). on distributional national accounts. and Zucman 2016. 13 See Lustig (2018a). 34 Sub-Saharan Africa is the merged 14 See www.wider.unu.edu/project/wiid- distribution of Sub-Saharan African 49 For full details on methods and sources 71 See Alvaredo and others (2018, section world-income-inequality-database. countries for which survey data are in this section on Africa, see Chancel 4) for a longer discussion of the under- 15 See Bourguignon (2015a). available from the World Bank’s and others (2019). lying data. 16 See, for instance, ECLAC (2018b). PovcalNet. Survey data are corrected 17 European Union Statistics on Income with available tax data estimates 50 The incomes of the bottom 10 percent 72 See Piketty, Yang and Zucman. 2019 and Living Conditions website (https:// (which are available at this stage for of the distribution are reduced ec.europa.eu/eurostat/web/microdata/ the recent period only for Côte d’Ivoire 25­50 percent, while the incomes of Chapter 4 european-union-statistics-on-income- and South Africa; the gap between the top 1 percent are increased by the and-living-conditions, accessed 10 surveys and tax data in those countries same proportion, when moving from 1 UNDP 2018a; UN Women 2019; WEF October 2019. is used to correct survey estimates in consumption to income inequality 2018; World Bank 2012b. 18 See, for example, Galbraith and others other African countries). See Chancel, (Chancel and others 2019). 2015; Ravallion 2018b. Clarke and Gethin (2017) and Chancel 2 UNDP 2018a; UN Women 2019; WEF 19 Alvaredo and others 2018; Morgan and Czajka (2017). 51 See Morgan (2017) on Brazil, Assouad 2018; World Bank 2012b. 2017. 35 Here pensions and unemployment ben- (2017) on Lebanon and Czajka (2017) 20 Kuznets 1953; Atkinson and Harrison efits are considered deferred income on Côte d’Ivoire, among others. 3 UN Women 1995. (1978). and are therefore counted as part 4 WEF 2018. 21 See Piketty (2001, 2003). of pretax and government transfers 52 Chancel and others 2019, p. 11. See 5 UNDP 2018a. 22 See Piketty and Saez (2003). income; see spotlight 3.1 at the end of also Alvaredo and Atkinson (2010) for 6 Giraldo-Luque and others 2018. 23 See Alvaredo and others (2013). the chapter. an analysis of South Africa in historical 7 Fletcher, Pande and Moore 2017. 24 See Alvaredo and others (2016, 2018). 36 Blundell, Joyce, Norris and Ziliak 2018. perspective. 8 Butler 2019; McDonald and White 25 See Alvaredo and others (2016). 37 Although it should be emphasized 26 See Alvaredo and others (2018). that the top 1 percent and bottom 53 See also Odusola and others (2017). 2018; UN News 2019. 27 See Piketty (2014). 50 percent of the population are not 54 The extreme poverty rate went from 9 Nussbaum 2001, p. 1. 28 See https://wid.world/. necessarily composed of the same 10 UNDP 1995, p. 1. Methodological details can be found in individuals in 1980 and 2016. 36.6 percent in 1996 to 16.9 percent 11 UNDP 1995, p. 29. Blanchet and Chancel (2016). 38 The “elephant curve” was popularized in 2008 and to 18.9 percent in 2014. 12 UN 2015a. 29 UN 2009. by Lakner and Milanovic (2016). See http://povertydata.worldbank. 13 UN Women and IPU 2019. 30 See results in Stiglitz, Sen and Fitoussi 39 This is discussed in Alvaredo and org/poverty/country/ZAF (accessed 6 14 UNICEF 2018b. (2009b). others (2018). November 2019). 15 Keleher and Franklin 2008; Marcus 31 The research here on the levels and 40 Ravallion 2018a. 55 See Alvaredo and others (2018). evolution of global income inequality 41 The share of the population living on 56 Parts of this section draw on Alvaredo 2018; Marcus and Harper 2014; Munoz draws heavily on Alvaredo and others less than $1.90 a day decreased from and others (2018) and Blanchet, Boudet and others 2012; Sen, Ostlin (2018), which provides full details on 46 percent in 1993 to 21.2 percent in Chancel and Gethin (2019). and George 2007. methods and sources. 2011 (World Bank 2012a). 57 Blanchet, Chancel and Gethin 2019; 16 Marcus and Harper 2014. 32 Here, Europe corresponds to Western 42 In particular, Lakner and Milanovic Piketty, Saez and Zucman 2018. 17 Bicchieri 2006; Fehr, Fischbacher and Europe. Western Europe is built by (2016) and Anand and Segal (2014). 58 See details in Blanchet, Chancel and Gächter 2002; Ostrom 2000. merging the income distributions See also other attempts to measure Gethin (2019). 18 Galvan and Garcia-Peñalosa 2018. of France, Germany and the United global income inequality: Bourguignon 59 It is important to stress that the focus 19 OECD 2017a; UNDP and UN Women Kingdom and an aggregate merging and Morrisson (2002), Niño-Zarazúa, here is solely on monetary income 2019; UN Women 2015b; WEF 2017. other Western European countries (28 Roope and Tarp (2017) and Ortiz and inequality, which was unusually low 20 Mackie and others 2015. countries in total) to cover 420 million Cummings (2011). Indeed, when meas- in the Russian Federation and Eastern 21 Charles 2012. people. See Chancel, Clarke and ured in absolute terms the elephant Europe under communism. Other forms 22 Chamorro-Premuzic 2013. Gethin (2017). The Middle East is de- curve looks more like a hockey stick of inequality prevalent at the time, in 23 Marcus and Harper 2015. fined as the region from Egypt to Iran (Ravallion 2018a). This fact is illus- terms of access to public services or 24 Green 2016. and from Turkey to the Gulf countries trated by focusing on shares of total consumption of other forms of in-kind 25 Gintis 2007. and covers 410 million people. See growth captured and not only on the benefits, may have enabled local elites 26 Cislaghi, Manji and Heise 2018; Alvaredo, Assouad and Piketty (2018). growth rates of each income group. to enjoy much higher standards of Cooper and Fletcher 2013; Marcus and 33 United States­Canada is built as fol- 43 See Alvaredo and others (2018) for a living than their income levels suggest. Harper 2014. lows. Canadian growth is distributed more detailed discussion of national 60 The percentage of the population at 27 Bandura 2003; Mackie and others to the Canadian population assuming trajectories. risk of poverty is defined as the share 2015; Munoz Boudet and others 2012; the same distribution as the one 44 See also Milanovic (2005). of adults living on less than 60 percent Sood, Menard and Witte 2009. observed in the United States. The 45 Indeed, the two scenarios are not addi- of the national median income. 28 Bian, Leslie and Cimpian 2017; simplification seems acceptable given tive, in the sense that global inequality 61 See https://data.oecd.org/socialexp/ Cunningham 2001. the similar trajectories of top income is not the sum of the two curves. social-spending.htm. 29 OECD 2017a. shares observed in the two countries 62 For comparisons of the United 30 Borrell-Porta, Costa-Font and Philipp and is justified by the relatively small States and Europe, see OECDStats 2018. (https://stats.oecd.org/Index. 31 Borrell-Porta, Costa-Font and Philipp aspx?DataSetCode=MIN2AVE#). 2018. 63 Parts of this section draw on Alvaredo 32 Amin and others 2018. and others (2018). 33 Kågesten and others 2016. 64 Piketty and Zucman 2014. 34 Mackie and Le Jeune 2009; Mackie 65 Atkinson and Harrison 1978. and others 2015; UNICEF Innocenti Research Centre 2010. 35 UN Women 2015b.

Notes | 261 36 Mackie and Le Jeune 2009; Marcus make a chair available for someone to 11 Cumming and von Cramon-Taubadel (trade liberalization via preferential and Harper 2014; UNICEF 2013. sit on). 2018. trade agreements) than what this 3 Because fewer consumers have the chapter asserts has to do with trade 37 Cialdini, Kallgren and Reno 1991; purchasing power to afford “green” 12 See Moser and Kleinhückelkotten generally, regardless of the extent of Etzioni 2000; Jacobs and Campbell goods and services, keeping prices (2017) for an exploration of the liberalization, as well as to do with 1961. up and generating less demand for complexities of environmental identity, non-trade-related burden shifting. See further technological change (Vona and intentions and impacts, combining in- also Roca (2003). 38 Nussbaum 2003. Patriarca 2011). tent- and impact-oriented perspectives 15 Given that the negative effects of cli- 39 Addati and others 2018. 4 There is evidence supporting this that have been motivating questions mate change fall disproportionally on 40 The term was coined by Amartya Sen hypothesis for the case of states in the environmental psychology those with lower incomes and fewer in the United States, with analysis literature. capabilities (UNDP 2007). to capture the fact that the proportion showing that there is no impact 16 Based on simulations of the evolution of women is lower than what would be of the Gini index on emissions by 13 For instance, the trend (that is, in income inequality across countries expected if girls and women through- state (thus, confirming that the first removing cyclical changes in income) from 1961 to 2010, measured by the out the developing world were born mechanism is either absent or weak) elasticity between income and emis- 90:10 ratio of population-weighted and died at the same rate, relative to but that there is a positive relationship sions for a typical developed country GDP per capita (Diffenbaugh and boys and men (Sen 1990). between state-level emissions and the has been estimated to be essentially Burke 2019a). These results have 41 UNDP 2016. concentration of income among the top zero for production-based emissions been contested as an overestimation 42 UNDP 2016. 10 percent, consistent with “approach- (effectively meaning that emissions (Rosen 2019), but the authors stand by 43 OECD 2017a; UNESCO 2019a. es that focus on political economy are decoupled from growth), while their findings (Diffenbaugh and Burke 44 Bill & Melinda Gates Foundation 2019. dynamics […] which highlight the jumping to 0.5 for consumption-based 2019b). 45 UNICEF 2019a. potential political and economic power emissions (implying still quite strong 17 Burke and Tanutama 2019; Randell and 46 Kishor and Johnson 2004. […] of the wealthy” (Jorgenson, coupling); the elasticities estimated for Gray 2019. Although a lot depends on 47 Loaiza and Wong 2012. Schor and Huang 2017, p. 40). Market developing countries are about 0.7 for how impact is assessed (for example, 48 Chandra-Mouli, Camacho and Michaud concentration was key in the history both production-based emissions and economic damages or causalities) 2013. of the development of the Montreal consumption-based emissions (Cohen and on the nature of hazards linked 49 Blum and Gates 2015. Protocol in 1987 to combat ozone-de- and others 2018). to climate change. For instance, there 50 Statistics for contraceptive prevalence stroying chlorofluorocarbons. For years, is evidence that extreme temperature focus on married women because in dominant firms opposed regulatory 14 A related concept—and narrower one, events have increased mortality in both developing countries most sexually action until they saw how they could in certain formulations—discussed at developed and developing countries, active adolescent girls are married, benefit economically from regulation length in the literature is the pollution deaths linked to extreme drought have while in some it is claimed that sexual that would create a profitable market haven hypothesis, first postulated gone down in both groups of countries activity happens only within marriage. for chemical substitutes (Hamann and by Copeland and Taylor (1994) in the and there is increasing polarization As a result, household surveys do not others 2018; Maxwell and Briscoe context of the North American Free between developed and developing collect data on unmarried women. Still, 1998). Trade Agreement. In its most general countries in deaths associated with ex- unmarried women need to be con- 5 Much of the evidence applies to formulation, the pollution haven hy- treme storms increasing in developing sidered when designing policies and common pool resources, as opposed to pothesis posits that trade liberalization countries (Coronese and others 2019). interventions for reproductive health. a global public good such as climate encourages more polluting firms and 18 Pershing and others (2019) provide a 51 UNFPA 2019. stability, but the broader mechanism industries to move some of their oper- conceptual overview of the impact of 52 Kumar and Rahman 2018. showing how inequality makes collec- ations to countries with laxer environ- climate change on “surprise” events 53 UNDP 2016. tive action more challenging remains mental standards, thereby increasing and empirical analysis in 65 large 54 UNIFEM 2000. valid. See, for instance, Alesina and La pollution in receiving countries. marine ecosystems. 55 ILO 2017a. Ferrara (2000); Anderson, Mellor and Evidence for the hypothesis is mixed; 19 World Bank 2019d. 56 Alonso and others 2019. Milyo (2008); Bardhan (2000); Costa see Gill, Viswanathan and Abdul Karim 20 Klein 2019. 57 Hegewisch and Gornick 2011. and Kahn (2003); and Varughese and (2018) for a comprehensive review. A 21 Le Quéré and others 2018. 58 The global average gap for the same Ostrom (2001). conceptual sticking point is causal- 22 Brulle 2018; Dunlap and McCright job is 77 percent (UN Women 2017). 6 For both the impact of inequality in ity—that is, whether firms relocate 2011; Van den Hove, Le Menestrel and 59 Munoz Boudet and others 2018. reducing cooperation and the value because of the laxer environmental De Bettignies 2002. 60 World Bank 2017a. of communication in enabling it, see standards or for some other reason 23 Ritchie and Roser 2018. 61 Schmidt and Sevak 2006; Sierminska, Tavoni and others (2011). that also happens to correlate with 24 Some have argued that the uncon- Frick and Grabska 2010. 7 Berger and others 2011. weaker standards. Some compelling ditional pledges under the Paris 62 Demirgüç-Kunt, Klapper and Singer 8 And how inequality exacerbates evidence exists for the hypothesis for Agreement worsen existing inequal- 2013. social status competition and could carbon dioxide emissions, including ities in carbon emissions and that an 63 Munoz Boudet and others 2018. encourage growth policies at the ex- recently in Itzhak, Kleimeier and Viehs emissions trading scheme across a 64 See UN Women (2019). pense of environmental ones (Baland, (2018), which used innovative micro subset of major signatories shows 65 See UN Women (2015a). Bardhan and Bowles 2007; Berthe data. Environmental burden shifting that although emissions trading re- 66 ILO 2017a; UNDP 2016; UN Women and Elie 2015; Chaigneau and Brown as presented in this chapter is broader duces the costs of meeting emissions 2015b; World Bank 2012b. 2016; Franzen and Vogl 2013; Magnani than under the pollution haven hypoth- reductions targets, most of the gains 67 Deschamps 2018. 2000). esis and is not preconditioned on dif- would go to richer countries, implying 9 Cohen and others 2018 p. 1. ferences in environmental regulation. additional inequality. See Rose, Wei Chapter 5 10 Some evidence suggests that decoup- It can occur between countries—in the and Bento (2019). ling is associated with reductions in form of net flows of virtual pollution or 25 Chancel and Piketty 2015. 1 This framing is adapted from Berthe income inequality—more specifically, resource use (such as freshwater use) 26 Cardona and others 2012. and Elie (2015). that it is negatively associated with bundled in traded goods—or within 27 Even though growth in the economic increasing the income share of the top them—for example, in the siting of damages associated with extreme 2 This may seem to exclude production, 20 percent and positively associated waste disposal facilities. Kolcava, hazards in temperate regions has ac- but it is possible to consider not only with increasing the income share of Nguyen and Bernauer (2019) show only celerated (Coronese and others 2019). direct emissions from consumption the bottom 20 percent (McGee and partial support for a link between trade 28 The only stronger hurricane ever (such as driving a car) but also the in- Greiner 2018). liberalization via preferential trade recorded in the Atlantic was Hurricane direct emissions related to production agreements and environmental burden Allen in 1980, but it weakened before of a good or service (for instance, the shifting, as measured by aggregate steel processing, manufacturing and ecological footprints. Still, that study transportation activities required to has tested a narrower hypothesis

262 | HUMAN DEVELOPMENT REPORT 2019 making landfall (Le Page 2019). See 52 Other, related frameworks have been 85 Dang, Lanjouw and Swinkels 2014. 111 Processing, distribution and retail mat- also Rice (2019). proposed for the channels by which 86 Fuentes-Nieva and Seck 2010. ter, too, with losses often accounting 29 Semple 2019. climate change affects inequality. 87 Kim 2010. for the greatest share of emissions 30 Burke, Davis and Diffenbaugh 2018; See Islam and Winkel (2017), which 88 IDMC 2018. at these stages. Poore and Nemecek Kahn and others 2019; Kompas, Pham proposes three channels: exposure, 89 For instance, when an ocean heat 2018. and Che 2018; Pretis and others 2018; susceptibility and ability to cope Tol 2018. and recover. With the discussion on wave in the North Atlantic in 2012 led 112 FAO 2017; Science Daily 2014. 31 Burke and Tanutama 2019; Carleton resilience, this chapter broadly encom- the lobster capture to peak one month 113 Beef refers to beef obtained from beef and Hsiang 2016. passes this framework. earlier than normal, this led to a glut 32 Cooper 2019. in supply and drop in prices. After herders and dairy herders. Poore and 33 Weitzman 2012, p. 234. 53 See Winsemius and others 2018. The this “surprise” shock, investments in Nemecek 2018. 34 Some of the most widely used models authors also highlight a potential marketing and processing capacity en- 114 Poore and Nemecek 2018. rely on smooth damage functions as pathway going in other direction: abled the industry to respond to sharp 115 OECD and FAO 2018. a “best fit” for the underlying data the impact of hazard-prone areas on increases in temperature—such as the 116 FAO 2018. rather than on damage functions with poverty. one that occurred in 2016, in which the 117 OECD and FAO 2017, 2018. nonlinearities (that is, thresholds, industry hit record value (Pershing and 118 Bennett 1941; Block and others 2004; tipping points), which may be more 54 Demaria 2010. others 2019). Bouis, Eozenou and Rahman 2011. characteristic of potential catastrophic 55 Boillat and others 2018; Hart 2014; 90 Examples of environmental justice ac- 119 Because income elasticities for meat events under climate change. Smooth tivism include the mobilization against consumption are higher for lower functions are a “best fit” precisely be- Jones 2009. the siting of toxic dumping sites in income groups. Humphries and others cause the underlying data themselves 56 Martinez-Alier and others 2016; the 1980s (Bullard 1983, 1990; Margai 2014. have made minimizing assumptions 2001; Taylor 2000). 120 Burton 2019. about catastrophic events. To help Sobotta, Campbell and Owens 2007. 91 Milman 2018; US EPA 2015. 121 A.T. Kearney 2019. remedy this and other shortcomings in 57 Wenz 2007. 92 Thus, some environmental justice liter- 122 Giupponi and Paz 2015; Government of the underlying data from other models, 58 Asseng and others 2015; Battisti and ature is focused on procedural justice Ecuador 2008; State of California 2012; a “fudge factor” of 25 percent is added questions rather than on distributive Takacs 2016; UN General Assembly to the DICE damage function (Cooper Naylor 2009; Challinor and others outcomes (Curran 2018). 2010; United Nations Human Rights 2019; Nordhaus and Moffat 2017). 2016; Porter and others 2014; Zhao, Liu 93 In this chapter “waste” refers to solid Council 2010. 35 Cai, Judd and Lontzek 2013; Cai and and others 2017. waste. 123 Data in this paragraph are from others 2015; Lemoine and Traeger 59 King and Harrington 2018; King and 94 Kaza and others 2018. UNICEF and WHO (2019). 2014. others 2015; Mora and others 2013. 95 Eriksen and others 2014 124 FAO 2016. 36 Burke, Davis and Diffenbaugh 2018; 60 Schiermeier 2018. 96 US NOAA 2018. 125 FAO 2016. Kahn and others 2019; Kompas, Pham 61 For the general mechanisms through 97 Lebreton and others 2018. 126 Gerten and others 2015; Jaramillo and and Che 2018; Pretis and others 2018; which a weather shock can lead to 98 US NOAA 2018. Destouni 2015; Rockström and others Tol 2018. devastating food insecurity, see, for 99 Choy and others 2019; Woodall and 2009; Steffen and others 2015. 37 Daniel, Litterman and 2019. Forceful instance, Devereux (2009). others 2014. 127 Gleeson and others forthcoming. action in this context means pricing 62 Dingel, Meng and Hsiang 2019. 100 Allen and others 2019; Gasperi and 128 Mekonnen and Hoekstra 2016. carbon based on probabilistic assump- 63 Woodard, Davis and Randerson 2019. others 2018. 129 Mekonnen and Hoekstra 2016. tions of climate damages. The price 64 Burke and Tanutama 2019. 101 Kaza and others 2018. 130 Richey and others 2015. of carbon in this model would be high, 65 Randell and Gray 2019. 102 This paragraph is based on Kaza and 131 UNDP 2006, p. v. and increasing, over a few years, but 66 Mejia and others 2019. others (2018). 132 Hoekstra and Mekonnen 2012. would come progressively down as the 67 European Environment Agency 2018; 103 Bullard 1983, 1990; Margai 2001; 133 Mekonnen and Hoekstra 2011. insurance value decreases and tech- Parry and Terton 2016. Taylor 2000. For a literature review, 134 Hoekstra and Mekonnen 2012. nology makes emissions cuts cheaper. 68 Devex n.d.; Parry and Terton 2016; UK see Martuzzi, Mitis and Forastiere 135 In other words, more water is used to 38 WHO 2018. Space Agency 2018. (2010). See also Elliott and others produce the meat and cereals that are 39 Hoegh-Gulberg and others 2018. 69 Nakatani 2019. (2001); Harper, Steger and Filcak consumed rather than consuming more 40 Global Panel on Agriculture and Food 70 Global Commission on Adaptation (2009); Johnson, Lora-Wainwright meats and cereals overall. Systems for Nutrition 2016; US CDC 2019. and Lu 2018; Laurian (2008); McLaren, 136 Hoekstra and Mekonnen 2012. 2014. 71 Vörösmarty and others 2000. Cottray and Taylor (1999); Steger and 137 UNICEF and WHO 2019. 41 An individual can be exposed once or 72 Hallegatte and Rozenberg 2017; others 2007; Varga, Kiss and Ember 138 Cole and others 2018. multiple times during a year. Each time Rozenberg and Hallegatte 2015. (2002); Varró, Gombköto and Szeremi 139 Republic of South Africa 1996; South a person is exposed counts as an expo- 73 UNDP 2011. (2001); Walker and others (2003). Africa Department of Water and sure event. Watts and others 2015. 74 Liu and others 2007. 104 Thornton and others (2006), as cited in Sanitation 2016. 42 WHO and World Bank 2017. 75 As documented by Scheidel (2017). FAO (2018). 140 Gleick 2018. 43 Watts, Amann, Arnell and others 2018. And the response to a shock can be 105 Data in this paragraph are from FAO 141 Gleick 2018. 44 Mejia and others 2019. equalizing, even when the impact is (2018). 142 The cost of transitioning to a 45 Kahn and others 2019. not. For instance, Hurricane Mitch hit 106 FAO 2014; Poore and Nemecek 2018. carbon-free electricity system in the 46 Watts, Amann, Arnell and others 2018. the poorest households the hardest in 107 “Agriculture uses approximately United States has dropped markedly, 47 Watts, Amann, Ayeb-Karlsson and Honduras, but the response generated 70 percent of the available freshwater driven by declines in the cost of re- others 2018. an opportunity to address longstanding supply, and roughly 30 percent of newable energy technologies, such as 48 “Vectorial capacity is a measure of inequalities (McSweeney and Coomes global agricultural water goes on wind and solar, as well as energy stor- the capacity for vectors to transmit a 2011). livestock production” (FAO 2018, age systems (Heal 2019). See Haegel pathogen to a host and is influenced by 76 Coronese and others 2019. p. 51). Calculation based on 30 percent and others (2019) and Veers and others vector, pathogen, and environmental 77 Clarke and Dercon 2016. of 70 percent = 21 percent. (2019) for reviews of global cost and factors” (Watts, Amann, Ayeb-Karlsson 78 See, for instance, Hallegatte and 108 Godfray and others 2010; Rask and capacity trends for photovoltaic and and others 2018, p. 2487). others (2017). Rask 2011. wind technologies, respectively, as 49 Watts, Amann, Arnell and others 2018. 79 Hallegatte and others 2017. 109 Gerbens-Leenes and Nonhebel well as a discussion of challenges 50 Randell and Gray 2019. 80 UNDRR 2019. 2002; Pimentel and Pimentel 2003; and opportunities for further scaling 51 Kim, Lee and Rossin-Slater 2019. 81 As an example, consider the reduction Wirsenius, Azar and Berndes 2010. up. Davis and others (2018) explore in vulnerability to floods (Jongman and 110 FAO 2006, 2017; Gerber and others challenges and opportunities in decar- others 2015). 2013; Tubiello and others 2014. bonizing energy services and industrial 82 IPCC 2014, p. 8. processes, such as long-distance 83 IPCC 2014, p. 13. For food security, see FAO (2018). 84 IPCC 2014.

Notes | 263 freight transport and air travel, that of these technological innovations and to the availability of technology per se 30 Lee 2018. are difficult to provide without emitting their significance, see McNeill (2001). (Scott 2017). In fact, humans’ ability 31 UNESCO n.d. carbon dioxide. Despite increasingly 5 The report also considered the to domesticate plants and animals 32 UNDESA 2018. favourable conditions for renewable importance of balancing incentives for already existed almost 4,000 years 33 GSMA 2017. energy and associated technologies, investments in new technology with before. But the technology became 34 GSMA 2018. global energy growth is still outpacing its diffusion and discussed the many consequential only when institutional 35 ITU 2019. decarbonization (Jackson and others barriers that developing countries face innovations such as the creation of 36 OECD 2019b. 2018). in benefiting from that diffusion (UNDP the state enabled small settlements to 37 See, for instance, Gonzales (2016) and 2001). grow into the early civilizations of the Chapter 6 6 Silver and others 2018. Fertile Crescent and the Nile Delta. Rosenberg (2019). 7 LeCun, Bengio and Hinton 2015. 19 As articulated, for instance, in 38 Hilbert 2019. 1 The expression became widely used Moreover, it is conceivable that Kuznets’s Nobel Prize lecture (Kuznets 39 For projections on the gains from by economic historians after Kenneth machines will learn not only by 1971). As Mokyr (2016, p. 339) put it: Pomeranz’s (2000) book The Great themselves, but also from other “In only one case did such an accumu- artificial intelligence, see PwC (2017). Divergence, even though the book machines, so that what is learned by lation of knowledge become sustained For the growth of artificial intelligence presented what was at the time an one can be shared with all the others. and self-propelling to the point of in North America and East Asia, and original thesis on how and why the This can be done much faster than the becoming explosive and changing the particularly in China, see Lee (2018). Industrial Revolution happened (claim- sharing of information by humans, who material basis of human existence 40 See, for instance, Utterback and ing that it was happenstance that it communicate at rates of 10 bits per more thoroughly and more rapidly Abernathy (1975). originated in Europe, given that East second, 100 million times slower than than anything before in the history 41 Hilbert 2011. Asia had very similar conditions in machines (Pratt 2015). of human on this planet. That one 42 This section draws in part from the late 17th century and, further, that 8 And beyond the impact on labour instance occurred […] during and after Conceição (2019a). Europe’s advantage was due largely to markets, artificial intelligence is also the Industrial Revolution.” Mokyr’s 43 So much so that a constant labour the large natural resources extracted starting to raise very deep philosoph- central thesis is that the European share of income has been one of from the “New World” colonies). ical questions. As of now, and in the Enlightenment, itself a construction the features that economic growth This view is contested, with a recent near future, artificial intelligence sim- that took several hundred years to models were expected to account for, review of the debate on the causes ply implements tasks that are defined build and that was far from preor- since Kaldor (1961) identified this as of the Great Divergence and multiple by humans, but it is conceivable that dained, provided the fertile ground an empirical regularity characterizing hypothesis being put forward, included machines might eventually be able for a “market for ideas” to flourish, economic growth. On the constant la- in Vries (2016). For a recent economic to set their own objectives—raising as well as the conviction that humans bour share of income, see Giovannoni perspective, see O’Rourke, Rahman profound questions about the human could understand “natural regularities (2014). and Taylor (2019). species and how people interact with and exploit them to their advantage” 44 Autor and Salomons (2017) quoted technology. See Russell (2018). (p. 7). Keynes as stating that this regularity 2 The aspiration to “develop” in 9 As shown in the extensive discussion 20 See Vries (2016). As Nobel laureate was a bit of a miracle. much of the second half of the 20th of possible channels and impacts of ar- Paul Romer (1990) argued, since we 45 As argued, for instance, by Rodrik century was almost synonymous with tificial intelligence on employment and live on a planet where our resources (2015). Avent (2016) goes a step “industrialization.” And, indeed, over earnings by Frank and others (2019). and abilities to produce things are further and argues that in the digital the second half of the 20th century, 10 The literature on this transformation bounded, it is ideas and abilities to age a new kind of institution will be manufacturing moved to several de- is vast, but recent works touching on combine things in ever more efficient needed. Since the promise of the dig- veloped countries—though not to all it include Goldin and Katarna (2016), ways that have driven economic ital revolution is an end to work, what and not at the same time—leading to Iversen and Soskice (2019), Unger growth. Perhaps the best way to fully will also be needed is institutions that some convergence in incomes across (2019) and the more speculative leverage technology is through sus- provide for people who do not work countries. A testimony of the enduring analysis in Harari (2016). taining what Stiglitz and Greenwald because their work is not necessary to appeal of industrialization is reflected 11 World Bank 2019a. (2014) called a “learning society.” generate economic growth. in the fact that industrializing remains 12 Russell 2018. 21 Basu, Caspi and Hockett (2019) show 46 Karabarbounis and Neiman 2013. For one of the goals of the 2030 Agenda 13 In the United States in 1995, only that the new technology underlying the the global dimension of the decline in for Sustainable Development. 2 percent of heterosexual couples platform economy, while expanding the labour share, see Dao and others had met online, while about half met the world’s production possibility (2017). 3 Until the mid-19th century the through family and friends. By 2017, frontier, can leave vast segments of 47 The erosion of demand for routine largest ratio of real income per capita close to 40 percent had met online, the population marooned and without tasks linked to technological change between the richest and the poorest compared with less than a third bargaining power. can account for about half of the de- society was 5 to 1 (Vries 2016). Human who met through family and friends 22 Mokyr 2002. cline in the labour share in developed Development Report Office calcula- (Rosenfeld, Thomas and Hausen 2019). 23 Vickers and Zierbarth 2019. countries (IMF 2017b). For evidence tions based on country estimates in 14 Frost and others 2019. 24 Atkinson 2015. on Europe, see Dimova (2019). The the Maddison Project Database of in- 15 Fintech News Hong Kong 2019. 25 Acemoglu and Restrepo 2019. decrease in intensity in trade unions come per person (Bolt and others 2018) 16 People’s Bank of China 2019. 26 See, for instance, Acemoglu and has also been an important factor in show that the ratio had grown to 50 17 Butera 2019. others (2012) for a treatment of some countries, including the United to 1 by the middle of the 20th century. 18 The Neolithic, or agricultural, directed technical change to address States (see Farber and others 2018). These estimates are contested but still Revolution that happened more than environmental challenges. 48 In developed countries falling provide a useful reference point. On 10,000 years ago is often invoked as 27 For instance, the full impact of elec- labour shares reflect a significant inequality within countries, Milanovic, another instance of a technological tricity on manufacturing productivity substitution of capital for labour, Lindert and Williamson (2010) show transformation on the scale of the did not fully materialize until factories but the explanation for the trend in that the Gini coefficient for income Industrial Revolution. But while the evolved to become single story and to developing countries is different. Firms was, on average, as high in preindus- historical consequences of the shift have multiple electrical motors tied to in advanced countries automate the trial as in industrial economies, with from hunter-gatherers living in small different pieces of equipment (David routine tasks. Therefore, tasks with similar variation across economies. nomadic groups to more sedentary 1990). lower factor substitutability are more and larger segments cultivating plants 28 For instance, in the context of Japan’s likely to be offshored. In developing 4 The intensive use of coal as a source and tending after livestock is beyond Society 5.0 (Government of Japan countries the decline in labour share is of energy continued throughout the dispute, recent scholarship shows that 2017). explained mainly by global integration, 20th century and was exacerbated by these transformations were not related 29 Mazzucato 2013. particularly the expansion of global the widespread use of the internal value chains, which contributed to combustion engine. For a historical ac- raising the overall capital intensity count of the environmental dimensions

264 | HUMAN DEVELOPMENT REPORT 2019 of production in developing countries Pitterle (2017); Goos, Manning and of programmes is another key consid- 119 IDRC 2018. (Dao and others 2017). Salomons (2014); and World Bank eration (Coady 2018). 120 World Bank 2019b. 49 For a description of how the decline (2016). 101 Berger and Frey 2016. 121 Consider the American Community in the relative price of investment 82 ILO 2019c. 102 OECD 2019c. goods interacts with technology and 83 Consider Voyager, an interactive 103 In fact, one reason why businesses Survey. Automated systems for moni- globalization to decrease the labour system for exploratory analysis that deploy so many robots, despite their toring demographics could become an share, see Lian (2019). For the decline combines manual and automated chart sometimes questionable contribution increasingly practical supplement to in the relative prices of investment specification. Given a dataset, Voyager to the bottom line, is that automation the survey. Some of the character- goods, see Lian and others (2019). spots potential quality or coverage is often subsidized. Subsidies induce istics relevant to the survey—such 50 Measured as the decline in the median issues. As users interact, Voyager firms to substitute capital for labour as income, race, education and developing country (Lian and others recommends views. Users report that even when the substitution does not voting patterns by postal code and 2019). Voyager helped promote data quality socially save costs, though it privately precinct—can be accurately estimated 51 Karabarbounis and Neiman 2013; Lian assessment and combat confirmation benefits the firm (Acemoglu and by applying artificial intelligence to and others 2019. bias (Heer 2019). Restrepo 2018; Guerreiro, Rebelo and images gathered by Google Street 52 Chen, Karabarbounis and Neiman 84 Agarwal, Gans and Goldfarb 2019. Teles 2018). View (Gebru and others 2017). 2017. Corporate savings are profits 85 Cheng, Chauhan and Chintala 2019; 104 The Republic of Korea, the most 122 Pokhriyal and Jacques 2017. that are not paid to taxes, labour, or IWPR 2019. robotized country in the world, 123 Rains, Krishna and Wibbels 2019. debt or equity holders. 86 Brussevich, Dabla-Norris and Khalid reduced the tax deduction on business 124 Tödtling and Trippl 2005. 53 Furman 2014. 2019. investments in automation, which is, 125 Cariboni 2014. 54 ILO 2018b. 87 It has been documented that boys in effect, a robot tax (Porter 2019). 126 Pla-Castells and others 2015. 55 Autor and others 2017; De Loecker outnumber girls in computer science By contrast, the European Parliament 127 Employing inspection of pipes that and Eeckjout 2017; Furman and Orszag Advanced Placement exams, and in rejected a motion to emphasize that have already been replaced by evaluat- 2015. 2013 only 26 percent of computer “consideration should be given to the ing soil dynamics and electromag- 56 Diez, Fan and Villegas-Sánchez 2019. professionals were women (AAUW possible need to introduce corporate netic forces coming from power lines 57 The importance of these network 2015; IDRC 2018). reporting requirements on the extent (Terdiman 2017). externalities has long been recognized 88 WEF 2018. and proportion of the contribution of 128 Mann and Hilbert 2018. as a key feature of all platforms, not 89 Metz 2019. robotics and AI to the economic results 129 Goodfellow, Bengio and Courville only technological ones. See Rochet 90 Metz 2019. of a company for the purpose of taxa- 2016. and Tirole (2003). 91 ILO 2018a. tion and social security contributions” 130 Mann and Hilbert 2018. 58 Moazed and Johnson 2016. 92 The US state of California recently (European Parliament 2016, p. 10). 131 Atkinson 2014; Conceição 2019b. 59 Khan 2017. declared all drivers on ride sharing 105 One proposal is a tax on revenues from 132 Mazzucato 2011. 60 Dellinger 2019. platforms to be employees of the sales of targeted digital ads, the key 133 Many efforts are ongoing, under the 61 Wu and Thompson 2019. companies (Szymkowski 2019). This to the operation of platforms such as umbrella of the United Nations and 62 Naidu, Posner and Weyl 2018. ensures that labour laws apply to Facebook and Google (Romer 2019). others, to accelerate technology trans- 63 Chau and Kanbur 2018. these jobs. The New York City Taxi and 106 Tankersley and Rappeport 2019. fer in order to achieve the Sustainable 64 Dube and others 2018. Limousine Commission has approved 107 The Group of 20, under the Development Goals. For example, the 65 ILO 2018a. new rules designed to provide a Presidentship of Japan in 2019, has Technology Bank for Least Developed 66 See, for example, Atkinson (2014) and minimum hourly wage of $17.22 (after proposed expanding the World Trade Countries, established in 2018, Kanbur (2018). expenses) for drivers who work with Organization rules to include trade in following the call in the Istanbul 67 See, for example, Basu (2019b) and app-based services such as Uber, Lyft, data (Bradsher and Bennhold 2019). Programme of Action for the Least Stiglitz (2019b). Via and Juno (Ha 2018). 108 The General Data Protection Developed Countries and the 2030 68 Furman and Seamans 2019. 93 ILO (2019c) indicates that the Maritime Regulation requires companies to, Agenda for Sustainable Development, 69 Wu 2018. Labour Convention of 2006—in effect among other things, obtain a person’s is working to make science, technology 70 Basu 2019b; Stiglitz 2019b; Sunstein a global labour code for seafar- freely given consent before collecting and innovation resources available 2018. ers—was a source of inspiration in personal information, sharing it to institutions and individuals in 71 More broadly, it is possible to consider addressing the challenges of workers, among applications and using it in least developed countries and to how to steer artificial intelligence in a employers, platforms and clients any way (Wolford n.d.). The European strengthen the science, technology way that integrates ethical values and operating in different jurisdictions. Commission is also bringing legislation and innovation ecosystem in least economic value (see Korinek 2019). 94 Korinek and Stiglitz 2017. that will give EU citizens explicit rights developed countries. See www.un.org/ 72 Acemoglu and Restrepo 2018. 95 Freeman and Perez 1990. over the use of their facial recognition technologybank/. 73 This can have varied geographic im- 96 Moreover, technology in itself could data (Khan 2019). 134 Conceição and Heitor 2007. pacts. For instance, there is evidence also provide opportunities for devel- 109 Arrieta-Ibarra and others 2018. 135 Freeman 1987; Nelson 1993; UNDESA that in the United States smaller cities oping countries to reimage prevailing 110 Banerjee and Duflo 2011; Pritchett and 2018. faced a greater negative impact from industrial era policies in updated Beatty 2015. 136 López-Calva and Rodríguez-Castelán automation, while larger cities faced a social protection systems, offering 111 Muralidharan, Singh and Ganimian 2016. much smaller impact, given the abun- more effective risk sharing (Rutkowski 2018. 137 Schwellnus, Kappeler and Pionnier dance of occupations and professions 2018). 112 Digital technology can also help with 2017. with tasks not easily automatable 97 Individual savings can be a voluntary ageing workers, opening opportunities 138 ECLAC 2018a. (Frank and others 2018). option to supplement stable, equitable in training, including by overcoming 139 See, for instance, the case of China 74 ILO 2019c. and adequate mandatory social insur- time and resource constraints through (Zhao, Zhang and Shao 2016). 75 The Economist 2019. ance benefits (ILO 2019c). flexible and shorter learning options. 76 The Economist 2019; Maulia 2018. 98 For universal basic income, see, for ex- 113 O’Connor 2019; PwC n.d. Chapter 7 77 Bruckner, LaFleur and Pitterle 2017. ample, Francese and Prady (2018). See 114 O’Connor 2019. 78 Brynjolfsson, Mitchell and Rock 2018. also Hanna, Khan and Olken (2018). 115 Sanyal 2018. 1 Expansion and convergence because if 79 Wrzesniewski and Dutton 2001. 99 For example, overly generous un- 116 A 15 percent reduction in prematu- the objective were to be conver- 80 Brynjolfsson, Mitchell and Rock 2018. employment benefits can disincentivize rity is expected, which would save gence alone, that could conceivably 81 As discussed in chapter 2. See also labour market participation. See about 80,000 lives a year in Africa be achieved by diminishing the Acemoglu and Autor (2011); Autor, Farber and Valletta (2015). (Shankland 2019). capabilities of those that already have Katz and Kearney (2006); Bhorat and 100 Such as for health, education and oth- 117 World Wide Web Foundation 2017. them—whereas the goal, clearly, has others (2019); Bruckner, LaFleur and er spending areas. Fiscal sustainability 118 IDRC 2018. to be to move those that are lagging behind to the higher achievements of

Notes | 265 others. For brevity, the chapter will experienced a 20 percentage point participation of young women in the London School of Hygiene & Tropical refer to convergence only, but it should increase in their time in employment labour force. This incentive to join Medicine and Center for Domestic be understood to mean by expanding and more than $180,000 in additional the labour force or to work more also Violence Prevention 2015). the capabilities of those at the bottom. cumulative income. The benefits of yielded substantial lifecycle labour 47 UN 2015b. 2 Which, in turn, are shaped both by early childhood education extended supply effects (Lefebvre, Merrigan and 48 Surminsky, Bouwer and Linnerooth- history and by political economy to health at later ages. Boys from the Verstraete 2009). And when Québec Bayer 2016; UNFCCC 2015. considerations—each of which is treatment group were less likely to introduced universal access to low-fee 49 Surminsky, Bouwer and Linnerooth- also not independent from the level have excessive cholesterol and arterial child care in 2008, nearly 70,000 more Bayer 2016. of inequality in society (Piketty 1995, inflammation. Girls who received the mothers took jobs than if no such pro- 50 Tigchelaar and others 2018. Future 2014). education support had less long-term gramme had existed, for an increase warming increases the probability of 3 The policies included higher and more stress and a lower chance of having of 3.8 percent in women’s employment globally synchronized maize production progressive income taxes, earned diabetes or experiencing substance and 1.7 percent increase in Quebec’s shocks. income discounts at low income abuse. The early intervention boosted GDP (Fortin, Godbout and St-Cerny 51 Betkowski 2018. levels, taxable benefits paid for each the well-being and capabilities of not 2012; Herrera 2019.). 52 Roy and others 2019. child and a minimum income for all only the children as they grew up but 40 UN Women 2018. 53 Roy and others 2019. individuals. See Scheidel (2018), based also of their children and siblings. 41 Shackelford 2018. 54 Riahi and others 2017. Other ways on Atkinson (2015). Participants’ children had higher levels 42 OECD 2017a. include providing scenarios with high 4 For instance, the Report does not deal of employment and education than 43 Baird, McIntosh and B. Özler 2013; regional resolution (Fujimori and others with the trends linked to migration, nonparticipants’ children. They were Baird and others 2013; Hagen-Zanker 2017), representing institutional and ageing, urbanization, trade and others. suspended less often from school and and others 2017. governance changes associated with 5 “Everyone, as a member of society, had less criminal activity, especially 44 The programme recruited volunteer shared socioeconomic pathways more has the right to social security and children of fathers who had early community health workers who explicitly (Zimm, Spurling and Busch is entitled to realization, through childhood education (Heckman and administered injectable contracep- 2018) and calling up local and spatially national effort and international Karapakula 2019b). tives, charging a small fee, or provided explicit estimates of vulnerability, co-operation and in accordance with 14 For instance, in the United States such counselling and referrals for other poverty and inequality, which have the organization and resources of each policies since 1960 included school methods. The option to have communi- emerged in recent publications based State, of the economic, social and desegregation, equalizing funding ty meetings and provide contraceptives on the shared socioeconomic path- cultural rights indispensable for his across school districts, compensatory door to door took the cultural and ways and are essential to investigate dignity and the free development of resources for schools with many social conditions into account in in- equity dimensions (Byers and others his personality” (Universal Declaration low-income students and additional creasing the awareness, acceptability 2018; Klinsky and Winkler 2018. of Human Rights, article 22). early childhood education support for and use of modern contraceptives 55 Broadband Commission for 6 UNESCO 2019b. poor families. But achievement gaps (Bixby Center for Population Health Sustainable Development 2019. 7 The figure for the world in 2014 was between the bottom and the top of the and Sustainability 2014). 56 Broadband Commission for 80.1 percent. Based on data from socioeconomic distribution have been 45 As well as family planning services, to Sustainable Development 2019. the World Development Indicators large and persistent for nearly half a provide the community with a platform 57 Broadband Commission for database (http://datatopics.worldbank. century (Hanushek and others 2019). for dialogue on sexual education Sustainable Development 2019. org/world-development-indicators/) 15 Akmal and Pritchett 2019. and sexual and reproductive rights. 58 Broadband Commission for accessed 10 October 2019. 16 Akmal and Pritchett 2019. Information about sexual and repro- Sustainable Development 2019. 8 See UN (2019b). 17 Akmal and Pritchett 2019. ductive health is disseminated through 59 UN General Assembly 2016. 9 See, for instance, Ritchie (2019). 18 Malouf Bous and Farr 2019. youth peer networks, many of which 60 The relationship between income 10 UNDP 2016. 19 Shanmugaratnam 2019. are affiliated with school, community, inequality and economic growth has 11 This is consistent with the categories 20 ILO 2019c. religious and youth associations. often been presented as a tradeoff of coverage, generosity and equity 21 See also Braveman and Gottlieb The government has received United (chapter 2). This framing has led to po- discussed in Martínez and Sánchez- (2014). Nations Population Fund support to lar policy approaches. At one extreme Ancochea (2018, 2019a, 2019b). 22 George 2016. develop the school club model and two overemphasizing pro-equality policies 12 For important frameworks and guides 23 Chemouni 2018. manuals for teachers and students can neglect economic incentives to on operationalizing the pledge to leave 24 Reich and others 2016. (UNFPA 2019). innovate and produce. At the other ex- no one behind, see UNSDG (2019) and 25 Reich and others 2016. 46 The word “sasa”, which means “now” treme pro-growth policies can neglect UNDP (2018b). For a more conceptual 26 Stewart 2006. in Kiswahili, serves as an acronym inclusion and sustainability. Choosing analysis, see Klasen and Fleurbaey 27 UNDESA 2009. for the four phases of the approach: one or the other side of this tradeoff (2018). 28 Stewart 2016a. Start, Awareness, Support, Action. The often ends up in poor performance on 13 For instance, in a cohort of socioec- 29 Langer and Stewart 2015; Stewart programme begins by partnering with both growth and equality. To fix ideas, onomically disadvantaged minority 2016a. a local organization, which selects the experience of Latin America— children in Michigan followed from 30 UN CEB 2017. an equal number of female and male maybe the most unequal region in the age 3 to age 55, children in the 31 Silcoff 2018. community activists—regular people world, with waves of policy experi- treatment group received 2.5 hours of 32 Patnaik 2019. interested in issues of violence, power mentation—offers some illustrative education a day and a weekly home 33 OECD 2017a. and rights, as well as institutional examples of these two approaches: visit to help parents engage with them. 34 Barker and others 2016. activists working for the police and populist experiences in the 1970s The effects of combined education 35 Human Development Report Office in health care, local government and and 1980s, followed by conservative and parent engagement at an early calculations based on data from the faith-based groups. The activists policies in the 1990s consistent with age were significant. When the boys WORLD Policy Analysis Center’s receive training in new concepts and the so-called Washington Consensus. grew up, they spent on average Gender Database 2019. ways to approach power imbalances. Some populist experiences in Latin 8 percent fewer days in jail between 36 Park 2015. They then take the lead in organ- America are analysed in Dornbusch ages 20 and 50 than those who did 37 OECD 2017a. izing informal activities with their and Edwards (1991). The reforms in not participate in the programme. 38 OECD 2017a. community networks to encourage Latin America during the 1990s are Only 7 percent of the boys in the 39 Del Boca 2015; Jaumotte 2003; Olivetti open discussions and critical thinking. discussed in Ffrench-Davis (2000). An treatment group were convicted of a and Petrongolo 2017; Thévenon 2013. Combined, the strategies ensure that analysis of long-term inequality in violent felony at least once, compared Québec introduced a low-fee universal different community members are Latin America is found in Gasparini with 30 percent in the control group. child care programme in 1997 for exposed and receive information from and Lustig (2011). Between the ages of 26 and 40, they children up to age 4, increasing the people they trust (Raising Voices,

266 | HUMAN DEVELOPMENT REPORT 2019 61 See similar analysis in ECLAC (2018a), monopsony power in China and India, 108 Ennis, Gonzaga and Pike 2017; Gans Nations and the World Bank Group. figure I.1, using overall inequality even though the degree of monopsony and others 2018. The objectives of the platform, measured with the Gini coefficient. fell over time in both countries (Brooks launched in 2016, are domestic re- and others 2019). If firms in China 109 Gans and others 2018. source mobilization and the state; the 62 This negative relationship is statisti- had no labour market power, labour’s 110 See Atkinson (1995). role of taxes in supporting sustainable cally significant. There is a decoupling share of income would have been 111 Basu 2019a. economic growth, investment and between the two variables for the 10 percentage points higher in 1999 112 Covarrubias, Gutiérrez and Philippon trade; the social dimensions of taxes Group of 7 economies. The weakening and 5 percentage points higher in (poverty, inequality and human devel- of policies that jointly support growth 2007. If firms in India had no labour 2019. opment); tax capacity development; and equity has been suggested as a market power, labour’s share would 113 Shapiro 2018. and tax cooperation (see PCT 2019). leading factor in the case of the United have been 13 percentage points higher 114 European Commission 2019. 134 Zucman 2015. States (Furman and Orszag 2018). in 1999 and 6 percentage points higher 115 See Lustig (2018a). 135 European Commission 2016. in 2011. 116 For the cash portion—redistribution 136 Tørsløv, Wier and Zucman 2018. 63 López-Calva and Rodríguez-Castelán 79 Brooks and others 2019. 137 Tørsløv, Wier and Zucman 2018. 2016. 80 Bhaskar, Manning and To 2002. aside from the in-kind benefits of 138 The OECD defines tax evasion as 81 Falch 2010; Ridder and van den Berg publicly provided health care and generally referring to illegal arrange- 64 López-Calva and Rodríguez-Castelán 2003; Staiger, Spetz and Phibbs 2010. education—taxes can sometimes ments where liability to tax is hidden 2016. 82 Basu, Chau and Kanbur 2015. increase the number of people living or ignored—that is, the taxpayer pays 83 See Ghosh (2016, 2019). in poverty or reduce their income. less tax than legally obligated by 65 Lustig, Lopez-Calva and Ortiz-Juarez 84 Bhorat, Kanbur and Stanwix 2017. In Armenia, Bolivia, Brazil, Ethiopia, hiding income or information from the 2013. 85 See Chacaltana, Leung and Lee (2018). Ghana, Guatemala, Honduras, Sri tax authorities. 86 ILO 2018. Lanka and Tanzania income redistribu- 139 The BEPS Project provides 15 action 66 The authors are grateful to Angus 87 ILO 2018. tion increased the number of people plans that equip governments with the Deaton for emphasizing this point to 88 OECD and ILO 2019. below the $2.50 a day poverty line. domestic and international instruments them. 89 ILO 2018. In Indonesia, Mexico, the Russian needed to tackle tax avoidance. 90 OECD and ILO 2019. Federation and Tunisia redistribution The OECD defines tax avoidance as 67 Polanyi 1944. 91 OECD and ILO 2019. also reduced the income of about half generally describing the arrangement 68 Kus 2012. 92 This paragraph is based on ILO (2019c). the poor population (Lustig 2018b, of a taxpayer’s affairs that is intended 69 One estimate suggests that, for the 93 Levine 2005. 2018c). Some countries may simply to reduce tax liability; although the 94 Arcand, Berkes and Panizza 2015. have too few people above the poverty arrangement could be strictly legal, United States, the decline in unions There is no conceptual consensus. line with incomes high enough to tax it is usually in contradiction with the may have accounted for as much as Some suggest that financial develop- (Bolch, Ceriani and Lopez-Calva 2017). intent of the law it purports to follow. half of the increase in the income ment can reduce income inequality 117 This paragraph is based mainly on 140 OECD 2018b. share of the top 10 percent from 1980 (by, say, increasing access to credit Lustig (2018b). 141 Noked 2018. to 2010 (Jaumotte and Osorio 2015; or other financial services, such as 118 Klemm and others 2018. 142 Shaxton 2019. Marx, Soares and Van Acker 2015). insurance; Banerjee and Newman 119 Ostry, Berg and Tsangarides 2014. 143 OECD 2018d. 70 National policies towards unions range 1998; Galor and Zeira 1993). Others 120 Grigoli and Robles 2017. 144 OECD 2019a. widely, from outright resistance, to predict a nonlinear relationship where 121 Average statutory corporate income 145 OECD 2019d. Several small or tripartite cooperation with workers inequality first increases as access to tax rates fell from 1990 to 2015, from developing countries have used lower and employers, to actively promoting finance is restricted to a minority and about 45 percent to 25 percent in ad- corporate tax rates or preferential tax collective bargaining as part of a then decreases as access to credit vanced economies and from just under rates for specified activities as part broader wage policy (Hayter 2015). spreads across society Greenwood 40 percent in emerging economies and of a package of measures to attract 71 A recent meta-analysis of 42 studies and Jovanovic 1990. See also Bolton, about 35 percent in low-income coun- investment and stimulate growth, and 269 estimates concluded that Santos, and Scheinkman (2016); tries to just over 20 percent in both rather than compete by holding wages there is no significant effect of unions Gennaioli, Shleifer, and Vishny (2012); cases (IMF 2017a). There is evidence low indefinitely. on productivity growth, even though Korinek and Kreamer (2014); and that effective corporate income tax 146 FitzGerald and Ocampo 2019. there are differences across sectors Thakor (2012). rates have also declined significantly 147 Piketty 2014. (Doucouliagos, Freeman and Laroche 95 Beck, Demirgüç-Kunt and Levine 2007; since the 1980s (FitzGerald and 148 UNDP 2016. 2017). Clarke, Xu and Zou 2006; Demirgüç- Ocampo 2019). 149 Lamont (2018) calls for a new research 72 See UNDP (2015). Kunt and Levine 2009. 122 See also Ardanaz and Scartascini agenda on policies in the area, defin- 73 ILO 2016a. 96 De Haan and Sturm 2017; Jauch and (2011) and Martínez and Sánchez- ing some policy principles. 74 ILO 2016b. Watzka 2016; Jaumotte, Lall and Ancochea (2019a). 150 Well implemented conditional cash 75 Data on minimum wages are hetero- Papageorgiou 2013. 123 OECD 2018c. transfers programmes appear to be geneous but have been harmonized by 97 Rajan 2011. 124 Saez and Zucman 2019. effective and to have positive long- the International Labour Organization 98 Brei, Ferri and Gambacorta 2018. 125 OECD 2018c. term effects. See Bouguen and others to express monthly minimum wages in 99 In contrast to the prediction of some 126 This paragraph is based on Lustig (2019). 2011 purchasing power parity dollars, theoretical models. (2018b). 151 OECD 2019f. subject to problems in price conver- 100 Favara and Imbs 2015. 127 Aaberge and others (2018) do this for 152 Daude and others (2017) examine nine sions. As data validation, cases where 101 Mitnik, Cumberworth and Grusky 2016. Nordic countries, showing that, on Latin American countries. the minimum wage resulted in larger 102 Adam and Tzamourani 2016. the whole, the impact has been less 153 Martínez and Sánchez-Ancochea output per capita and cases where the 103 Bezemer and Samarina 2016. progressive than in the past. 2019a; Verget and others 2017. monthly minimum wage was lower 104 Bezemer, Grydaki and Zhang 2016. 128 OECD 2019e. 154 Murillo and Martínez-Garrido 2017. than $10 were excluded. 105 Bezemer and others 2018; Mazzucato 129 See, for example, Branstetter, Glennon 155 Martínez and Sánchez-Ancochea 76 See the literature review in section I.5 and Semieniuk 2017. and Jensen 2019 2019a, based on Fairfield (2015) and of ILO (2016b). 106 Barkai 2016; De Loecker and Eeckhout 130 World Bank 2020. Schiappacasse (2019). 77 Riley and Bondibene 2017. 2017; Eggertsson, Robbins and Wold 131 Timmer and others 2014 . 156 Martínez and Sánchez-Ancochea 78 See discussion of recent evidence in 2018; Gutiérrez and Philippon 2019. 132 FitzGerald and Ocampo 2019. 2019a. Nolan, Richiardi and Valenzuela (2018). 107 Diez, Fan and Villegas-Sánchez 2019. 133 A response in this direction is Felix and Portugal (2017) provide the Platform for Collaboration on evidence of the link monopsony-wage Tax launched by the International dispersion in Portugal. Webber (2015) Monetary Fund, the OECD, the United uses US data to document monopsony power on wages, which is strongest in the lower half of the earnings distribution. There is also evidence of

Notes | 267 References

Aaberge R., C. André, A. Boschini, L. Calmfors, K. Agénor, P.R., K. Ozdemir and E. P. Moreira. 2018. “Gender Alonso, C., M. Brussevich, M.E. Dabla-Norris, Y. Gunnarsson, M. Hermansen, A. Langørgen, P. Gaps in the Labor Market and Economic Growth.” Policy Kinoshita and M.K. Kochhar. 2019. “Reducing Lindgren, C. Orsetta, J. Pareliussen, P-O Robling, Research Working Paper 8661. World Bank, Washington, and Redistributing Unpaid Work: Stronger Policies to J. Roine and J. E. Søgaard. 2018. Increasing Income DC. Support Gender Equality.” IMF Working Paper 19/225. Inequality in the Nordics: Nordic Economic Policy Review International Monetary Fund, Washington, DC. 2018. Copenhagen: Nordic Council of Ministers. https:// Aghion, P., E. Caroli and C. Garcia-Penalosa. 1999. norden.diva-portal.org/smash/get/diva2:1198429/ “Inequality and Economic Growth: The Perspective of the Alstadsæter, A., N. Johannesen and G. Zucman. 2018. FULLTEXT01.pdf. Accessed 10 October 2019. New Growth Theories.” Journal of Economic Literature “Who Owns the Wealth in Tax Havens?” Macro Evidence 37(4): 1615­1660. and Implications for Global Inequality.” Journal of Public AAUW (American Association of University Women). Economics 162: 89­100. 2015. Solving the Equation: The Variables for Women’s Aguilar, G.R., and A. Sumner. 2019. “Who Are the World’s Success in Engineering and Computing. Washington, DC. Poor? A New Profile of Global Multidimensional Poverty.” ------. 2019. “Tax Evasion and Inequality.” American www.aauw.org/research/solving-the-equation/. Accessed Working Paper 499. Center for Global Development, Economic Review 109(6): 2073­2103. 24 October 2019. Washington, DC. Alvaredo, F., L. Assouad and T. Piketty. 2018. “Measuring Abud, M.J., G.G. Molina and E. Ortiz-Juárez. 2016. Aiyar, S.S., and C. Ebeke. 2019. “Inequality of Opportunity, lnequality in the Middle East 1990­2016: The World’s “Out-of-Poverty and Back-to-Poverty Transitions using Inequality of Income and Economic Growth.” IMF Working Most Unequal Region?” Review of Income and Wealth Panel Data.” Supporting document to the 2016 Regional Paper 19/34. International Monetary Fund, Washington, Human Development Report for Latin America and the DC. www.theboxisthereforareason.com/wp-content/ Alvaredo, F., A.B. Atkinson. 2010. “Colonial Rule, Caribbean. United Nations Development Programme, uploads/2019/02/WPIEA2019034.pdf. Accessed 1 August Apartheid and Natural Resources: Top Incomes in South Human Development Report Office, New York. 2019. Africa 1903-2007.” CEPR Discussion Paper 8155. Center for Economic and Policy Research, Washington, DC. Acemoglu, D., P. Aghion, L. Bursztyn and D. Hemous. Akcigit, U., and S.T. Ates. 2019. “What Happened to 2012. “The Environment and Directed Technical Change.” U.S. Business Dynamism?” NBER Working Paper 25756. Alvaredo, F, A.B. Atkinson, L. Chancel, T. Piketty, E. American Economic Review 102(1): 131­166. National Bureau of Economic Research, Cambridge, MA. Saez and G. Zucman. 2016. “Distributional National Accounts (DINA) Guidelines: Concepts and Methods Used Acemoglu, D., and D. Autor. 2011. “Skills, Tasks and Akmal, M., and L. Pritchett. 2019. “Learning Equity in the World Wealth and Income Database.” WID.world Technologies: Implications for Employment and Earnings.” Requires More than Equality: Learning Goals and Working Paper 2016/2. World Inequality Database. http:// In O. Ashenfelter and D. Card, eds., Handbook of Labor Achievement Gaps between the Rich and the Poor in Five wid.world/document/dinaguidelines-v1/. Accessed 10 Economics, Vol. 4B. New York: Elsevier. Developing Countries.” Working Paper 504. Center for October 2019. Global Development, Washington, DC. Acemoglu, D., S. Johnson and J.A. Robinson. 2001. Alvaredo, F., A.B. Atkinson, T. Piketty and E. Saez. “The Colonial Origins of Comparative Development: An Alesina, A., and E. La Ferrara. 2000. “Participation in 2013. “The Top 1 Percent in International and Historical Empirical Investigation.” American Economic Review Heterogeneous Communities.” Quarterly Journal of Perspective.” Journal of Economic Perspectives 27(3): 91(5): 1369­1401. Economics 115(3): 847­904. 3­20.

Acemoglu, D., S. Naidu, P. Restrepo and J.A. Robinson. Alesina, A., and R. Perotti. 1996. Income Distribution, Alvaredo, F., L. Chancel, T. Piketty, E. Saez and G. 2019. “Democracy Does Cause Growth.” Journal of Political Instability, and Investment. European Economic Zucman, eds. 2018. World Inequality Report 2018. Political Economy 127(1): 47­100. Review 40(6): 1203­1228. Cambridge, MA: Belknap Press.

Acemoglu, D., and P. Restrepo. 2018. “Artificial Alesina, A., and D. Rodrik. 1994. “Distributive Politics and Amin, A., A. Kågesten, E. Adebayo and V. Chandra- Intelligence, Automation and Work.” NBER Working Economic Growth.” The Quarterly Journal of Economics Mouli. 2018. “Addressing Gender Socialization and Paper 24196. National Bureau of Economic Research, 109(2): 465­490. Masculinity Norms among Adolescent Boys: Policy and Cambridge, MA. Programmatic Implications.” Journal of Adolescent Health Alesina, A., S. Stantcheva and E. Teso. 2018. 62(3): S3­S5. ------. 2019. “The Wrong Kind of AI? Artificial Intelligence “Intergenerational Mobility and Preferences for and the Future of Labor Demand.” NBER Working Redistribution.” American Economic Review 108(2): Anand, P. 2017. “Happiness, Well-Being and Human Paper 25682. National Bureau of Economic Research, 521­554. Development: The Case for Subjective Measures.” Cambridge, MA. Background paper for Human Development Report Alkire, S., and J. Foster. 2011. “Counting and 2016. United Nations Development Programme, Human Acemoglu, D., and J.A. Robinson. 2012. Why Nations Fail: Multidimensional Poverty Measurement.” Journal of Development Report Office, New York. The Origins of Power, Prosperity, and Poverty. New York: Public Economics 95(7): 476­487. Crown Publishers. ------. 2018. “Recasting Human Development Measures.” Allen, S., D. Allen, V.R. Phoenix, G. Le Roux, P.D. Occasional Paper. United Nations Development Adam, K., and P. Tzamourani. 2016. “Distributional Jiménez, A. Simonneau, S. Binet and D. Galop. 2019. Programme, Human Development Report Office, New York. Consequences of Asset Price Inflation in the Euro Area.” “Atmospheric Transport and Deposition of Microplastics European Economic Review 89: 172­192. in a Remote Mountain Catchment.” Nature Geoscience Anand, P., L. Roope and A. Peichl. 2016. “Wellbeing 12(5): 339­344. Evidence for the Assessment of Progress.” IZA Discussion Addati, L., U. Cattaneo, V. Esquivel and I. Valarino. 2018. Paper 9840. Institute of Labor Economics, Bonn, Germany. Care Work and Care Jobs for the Future of Decent Work. Almeida, J., R.M. Johnson, H.L. Corliss, B.E. Molnar Geneva: International Labour Office. and D. Azrael. 2009. “Emotional Distress Among LGBT Anand, S., and P. Segal. 2014. “The Global Distribution Youth: The Influence of Perceived Discrimination Based on of Income.” In A.B. Atkinson and F. Bourguignon, eds., Agarwal, A., J.S. Gans and A. Goldfarb. 2019. “Artificial Sexual Orientation.” Journal of Youth and Adolescence Handbook of Income Distribution, Vol. 2. New York: Intelligence: The Ambiguous Labor Market Impact of 38(7): 1001­1014. Elsevier. Automating Prediction.” Journal of Economic Perspectives 33(2): 31­50. Almond, D., and J. Currie. 2011. “Human Capital Anderson, E. 1999. “What is the Point of Equality?” Ethics Development Before Age Five.” In O. Ashenfelter and D. 109(2): 287­337. Agarwal, R., and P. Gaule. 2018. Invisible Geniuses: Could Card, eds., Handbook of Labor Economics. Amsterdam: the Knowledge Frontier Advance Faster? Washington, DC: Elsevier. Anderson, L.R., J.M. Mellor and J. Milyo. 2008. International Monetary Fund. “Inequality and Public Good Provision: An Experimental Analysis.” Journal of Socio-economics 37: 1010­1028.

268 | HUMAN DEVELOPMENT REPORT 2019 Aradillas, A. 2018. “Estudio Sobre el Impacto que Tiene data/wgea-research/international-gender-equality-report- Baland, J.M., P. Bardhan and S. Bowles. 2007. el Poder de Mercado en el Bienestar de los Hogares ing-schemes. Accessed 9 October 2019. Inequality, cooperation and environmental sustainability. Mexicanos.” In COFECE, Poder de Mercado y Bienestar Princeton, NJ: Princeton University Press, and New York: Social, México. Autor, D. 2014. “Polanyi’s Paradox and the Shape of Russell Sage Foundation. Employment Growth.” NBER Working Paper 20485. Arcand, J.L., E. Berkes and U. Panizza. 2015. “Too Much National Bureau of Economic Research, Cambridge, Bandura, A. 2003. “Social Cognitive Theory for Personal and Finance?” Journal of Economic Growth 20(2): 105­148. MA. www.nber.org/papers/w20485.pdf. Accessed 16 Social Change by Enabling Media.” In A. Singhal, M. Cody, September 2019. E. Rogers and M. Sabido, eds., Entertainment-Education Ardanaz, M., and C. Scartascini. 2011. “Why Don’t We and Social Change: History, Research, and Practice. Tax the Rich? Inequality, Legislative Malapportionment, ------. 2019. “Work of the Past, Work of the Future.” London: Routledge. and Personal Income Taxation around the World.” IDB NBER Working Paper 25588. National Bureau of Economic Working Paper 282. Inter-American Development Bank, Research, Cambridge, MA. www.nber.org/papers/w25588. Banerjee, A.V., and E. Duflo. 2003. “Inequality and Growth: Washington, DC. pdf?sy=588. Accessed 16 September 2019. What Can the Data Say?” Journal of Economic Growth 8(3): 267­299. Arnell, N.W., and S.N. Gosling. 2016. “The Impacts of Autor, D., D. Dorn, L.F. Katz, C. Patterson and J. Van Climate Change on River Flood Risk at the Global Scale.” Reenen. 2017. “The Fall of the Labor Share and the Rise ------. 2011. Poor Economics: A Radical Rethinking of the Climatic Change 134(3): 387­401. of Superstar Firms.” NBER Working Paper 23396. National Way to Fight Global Poverty. New York: Public Affairs. Bureau of Economic Research, Cambridge, MA. Arrieta-Ibarra, I., L. Goff, D. Jiménez-Hernández, J. Banerjee, A.V., and A.F. Newman. 1998. “Information, the Lanier and E.G. Weyl. 2018. “Should We Treat Data Autor, D., L. Katz and M. Kearney. 2006. “The Polarization Dual Economy, and Development.” Review of Economic as Labor? Moving beyond `Free’.” AEA Papers and of the U.S. Labor Market.” American Economic Review Studies 65(4): 631­653. Proceedings 108: 38­42. 96(2): 189­194. Bardhan, P. 2000. “Irrigation and Cooperation: an Empirical Assa, J. 2012. “Financialization and Its Consequences: The Autor, D.H., F. Levy and R.J. Murnane. 2003. “The Skill Analysis of 48 Irrigation Communities in South India.” OECD Experience.” Finance Research 1(1): 35­39. Content of Recent Technological Change: An Empirical Economic Development and Cultural Change 48(4): Exploration.” Quarterly Journal of Economics 118(4): 847­865. Asseng, S., F. Ewert, P. Martre, R.P. Rötter, D.B. Lobell, 1279­1333. D. Cammarano, B.A. Kimball, M.J. Ottman, G.W. Barkai, S. 2016. “Declining Labor and Capital Shares.” Wall, J.W. White and M.P. Reynolds. 2015. “Rising Autor, D., and A. Salomons. 2017. “Does Productivity New Working Paper 2. Stigler Center for the Study of the Temperatures Reduce Global Wheat Production.” Nature Growth Threaten Employment?” Paper presented at Economy and the State, Chicago, IL. Climate Change 5(2): 143­147. the ECB Forum on Central Banking, 26­28 June, Sintra, Portugal. www.ecbforum.eu/uploads/originals/2017/ Barker, G., M. Greene, E.G. Siegel, M. Nascimento, Assouad, L. 2017. “Rethinking the Lebanese Economic speakers/papers/D_Autor_A_Salomons_Does_productivi- M. Segundo, C. Ricardo, J.G. Figueroa, J. Franzoni, Miracle: The Extreme Concentration of Income and Wealth ty_growth_threaten_employment_Final_Draft_20170619. J. Redpath, R. Morrell, R. Jewkes, D. Peacock, F. in Lebanon, 2005-2014.” WID.world Working Paper pdf. Accessed 24 October 2019. Aguayo, M. Sadler, A. Das, S.K. Singh, A. Pawar and 2017/13. World Inequality Database. P. Pawlak. 2016. What Men Have to Do with It: Public Avent, R. 2016. The Wealth of Humans: Work, Power, and Policies to Promote Gender Equality. Rio de Janeiro, Brazil: A.T. Kearney. 2019. “How Will Cultured Meat and Meat Status in the Twenty-First Century. New York: St. Martin’s Instituto Promundo, and Washington, DC: International Alternatives Disrupt the Agricultural and Food Industry.” Press. Center for Research on Women. Chicago, IL. www.atkearney.com/retail/article/?/a/how- will-cultured-meat-and-meat-alternatives-disrupt-the- Azevedo, J P., G. Inchauste and V. Sanfelice. 2013. Barnosky, A.D., N. Matzke, S. Tomiya, G.O. Wogan, B. agricultural-and-food-industry. Accessed 15 August 2019. “Decomposing the Recent Inequality Decline in Latin Swartz, T.B. Quental, C. Marshall, J.L. McGuire, E.L. America.” Policy Research Working Paper 6715. World Lindsey, K.C. Maguire and B. Mersey. 2011. “Has the Atkin, D., B. Faber and M. Gonzalez-Navarro. 2018. Bank, Washington, DC. Earth’s Sixth Mass Extinction Already Arrived?” Nature “Retail Globalization and Household Welfare: Evidence 471(7336): 51­57. from Mexico.” Journal of Political Economy 126(1): 1­73. Azevedo, J.P., G. Inchauste, S. Olivieri, J. Saavedra and H. Winkler. 2013. “Is Labor Income Responsible for Barro, R.J. 2008. “Inequality and Growth Revisited.” ADB Atkinson, A.B. 1970. “On the Measurement of Inequality.” Poverty Reduction? A Decomposition Approach.” Policy Working Paper Series on Regional Economic Integration Journal of Economic Theory 2: 244­263. Research Working Paper 6414. World Bank, Washington, 11. Asian Development Bank, Manila. www.econstor. DC. eu/bitstream/10419/109529/1/wp-011.pdf. Accessed 1 ------. 1995. “Capabilities, Exclusion, and the Supply of August 2019. Goods.” Welfare State Programme Discussion Paper 097. Babones, S.J. 2008. “Income Inequality and Population Centre for Analysis of Social Exclusion, The London School Health: Correlation and Causality.” Social Science & Bassett, T.J., and C. Fogelman. 2013. “Déjà Vu or of Economics and Political Science, London. Medicine 66(7): 1614­1626. Something New? The Adaptation Concept in the Climate Change Literature.” Geoforum 48: 42­53. ------. 2014. “After Piketty?” British Journal of Sociology Baird, S.J., E. Chirwa, J. De Hoop and B. Özler. 2013. 65(4): 619­638. “Girl Power: Cash Transfers and Adolescent Welfare: Basu, S. 2019a. “Are Price-Cost Markups Rising in the Evidence from a Cluster-Randomized Experiment in United States? A Discussion of the Evidence.” NBER ------. 2015. Inequality What Can Be Done? Cambridge, Malawi.” NBER Working Paper 19479. National Bureau of Working Paper 26057. National Bureau of Economic MA: Harvard University Press. Economic Research, Cambridge, MA. Research, Cambridge, MA.

Atkinson, A.B., and A.J. Harrison. 1978. Distribution of Baird, S.J., C. McIntosh and B. Özler. 2016. “When the Basu, K. 2019b. “New Technology and Increasing Returns: Personal Wealth in Britain. Cambridge, UK: Cambridge Money Runs Out: Do Cash Transfers Have Sustained The End of the Antitrust Century?” IZA Policy Paper 146. University Press. Effects on Human Capital Accumulation?” Policy Research Institute of Labor Economics, Bonn, Germany. Working Paper 7901. World Bank, Washington, DC. Atkinson, A.B., T. Piketty and E. Saez. 2011. “Top Incomes Basu, K., A. Caspi and R. Hockett. 2019. “The Law in the Long Run of History.” Journal of Economic Literature Baker, D., A. Jayadev and J. Stiglitz. 2017. “Innovation, and Economics of Markets with Digital Platforms.” 49(1): 3­71. Intellectual Property, and Development: A Better Set of Unpublished manuscript. Approaches for the 21st Century.” AccessIBSA: Innovation Auerbach, A.J., K.K. Charles, C.C. Coile, W. Gale, D. & Access to Medicines in India, Brazil & South Africa. Basu, K., and L.-F. Lopez-Calva. 2011. “Functionings and Goldman, R. Lee, C.M. Lucas, P.R. Orszag, L.M. Capabilities.” In K.J. Arrow, A. Sen and K. Suzumura, eds., Sheiner and B. Tysinger. 2017. “How the Growing Gap Baker, M., J. Currie and H. Schwandt. 2017. “Mortality Handbook of Social Choice and Welfare, Vol. 2. New York: in Life Expectancy May Affect Retirement Benefits and Inequality in Canada and the US: Divergent or Convergent North Holland. Reforms.” The Geneva Papers on Risk and Insurance- Trends?” National Bureau of Economic Research, Issues and Practice 42(3): 475­499. Cambridge, MA. Basu, A., N. Chau and R. Kanbur. 2015. “Contractual Dualism, Market Power and Informality.” Economic Australian Workplace Gender Equality Agency. 2019. Journal 125(589): 1534­1573. “International Gender Equality Reporting Schemes.” Australian Government, Canberra. www.wgea.gov.au/

Notes | 269 Battisti, D.S., and R.L. Naylor. 2009. “Historical Warnings Bessen, J., and M. Meurer. 2014. “The Direct Costs from Disorder: A Swedish Population-Based Study.” Journal of of Future Food Insecurity with Unprecedented Seasonal NPE Disputes.” Cornell Law Review 99(2): 387­424. Epidemiology and Community Health 71(7): 648­654. Heat.” Science 323(5911): 240­244. Betkowski, B. 2018. “How Drones Could Improve Crop Black, M.M., S.P. Walker, L.C. Fernald, C.T. Andersen, Baymul, C., and K. Sen. 2018. “Was Kuznets Right? Damage Estimates.” Phys.org, 27 April. https://phys.org/ A.M. DiGirolamo, C. Lu, C., D.C. McCoy, G. Fink, Y.R. New Evidence on the Relationship between Structural news/2018-04-drones-crop.html. Accessed 13 April 2019. Shawar, J. Shiffman and A.E. Devercelli. 2017. “Early Transformation and Inequality.” Working Paper 2018-027. Childhood Development Coming of Age: Science Through University of Manchester Global Development Institute, Bezemer, D., M. Grydaki and L. Zhang. 2016. “More the Life Course.” The Lancet 389(10064): 77­90. Manchester, UK. http://hummedia.manchester.ac.uk/insti- Mortgages, Lower Growth?” Economic Inquiry 54(1): tutes/gdi/publications/workingpapers/GDI/gdi-working- 652­674. Black, S.E., and P.J. Devereux. 2011. “Recent paper-2018027-baymul-sen.pdf. Accessed 6 August 2019. Developments in Intergenerational Mobility.” In O. Bezemer, D., J. Ryan-Collins, F. van Lerven and L. Ashenfelter and D. Card, eds., Handbook of Labor Bayoumi, M.T., and J. Barkema. 2019. “Stranded! How Zhang. 2018. “Credit Where it’s Due: A Historical, Economics. Amsterdam: Elsevier. Rising Inequality Suppressed US Migration and Hurt Those Theoretical and Empirical Review of Credit Guidance Left Behind.” IMF Working Paper 19/122. International Policies in the 20th Century.” Working Paper 2018-11. Black, S.E., P.J. Devereux, P. Lundborg and K. Majlesi. Monetary Fund, Washington, DC. www.imf.org/en/ University College of London, Institute for Innovation and 2019. “Poor Little Rich Kids? The Role of Nature versus Publications/WP/Issues/2019/06/03/Stranded-How- Public Purpose, London. Nurture in Wealth and Other Economic Outcomes and Rising-Inequality-Suppressed-US-Migration-and-Hurt- Behaviors.” NBER Working Paper 21409. National Bureau Those-Left-Behind-46824. Accessed 5 August 2019. Bezemer, D., and A. Samarina. 2016. “Debt Shift, Financial of Economic Research, Cambridge, MA. Development and Income Inequality in Europe.” SOM Beck, T., A. Demirgüç-Kunt and R. Levine. 2007. Research Reports 16020-GEM. University of Groningen, Blanchet, T., and L. Chancel. 2016. “National Accounts “Finance, Inequality and the Poor.” Journal of Economic SOM Research School, The Netherlands. Series Methodology.” WID.world Working Paper 2016/1. Growth 12(1): 27­49. World Inequality Database. http://wid.world/docu- Bhaskar, V., A. Manning and T. To. 2002. “Oligopsony and ment/1676/. Accessed 10 October 2019. Belluz, J. 2015. “Nobel Winner Angus Deaton Talks about Monopsonistic Competition in Labor Markets.” Journal of the Surprising Study on White Mortality He Just Co- Economic Perspectives 16(2): 155­174. Blanchet, T., L. Chancel and A. Gethin. 2019. “How Authored.” Vox, 7 November. Unequal Is Europe? Evidence from Distributional National Bhorat, H., R. Kanbur and B. Stanwix. 2017. “Minimum Accounts, 1980-2017.” WID.world Working Paper Bennett, M.K. 1941. “Wheat in National Diets.” Wheat Wages in Sub-Saharan Africa: A Primer.” World Bank 2019/06. World Inequality Database. Studies 18(1388-2016-116736): 37­76. Research Observer 32(1): 21­74. Block, S.A., L. Kiess, P. Webb, S. Kosen, R. Moench- Berger, A., C. Brown, C. Kousky and R. Zeckhauser. Bhorat, H., K. Lilenstein, M. Oosthuizen and A. Pfanner, M.W. Bloem and C.P. Timmer. 2004. “Macro 2011. “The Challenge of Degraded Environments: How Thornton. 2019. “The Rise of the Missing Middle' in an Shocks and Micro Outcomes: Child Nutrition during Common Biases Impair Effective Policy." Risk Analysis Emerging Economy: The Case of South Africa." Paper pre- Indonesia's Crisis." Economics & Human Biology 2(1): 31(9): 1423­33. sented at the Centre for the Study of African Economies 21­44. Conference, 17­19 March, Oxford, UK. Berger, T., and C. Frey. 2016. "Structural Transformation Blossfeld, H.P., S. Buchholz, J. Skopek and M. Triventi, in the OECD: Digitalisation, Deindustrialisation and the Bhorat, H., M. Oosthuizen, K. Lilenstein and A. eds. 2016. Models of Secondary Education and Social Future of Work." OECD Social, Employment and Migration Thornton. 2019. "The Rise of the Missing Middle’ in an Inequality: An International Comparison. Northampton, Working Paper 193. Organisation for Economic Co- Emerging Economy: The Case of South Africa.” University MA: Edward Elgar Publishing. operation and Development, Paris. of Cape Town, Development Policy Research Unit, Cape Town, South Africa. Blossfeld, H.P., N. Kulic, J. Skopek and M. Triventi, eds. Berger-Schmitt, R. 2000. “Social Cohesion as an Aspect 2017. Childcare, Early Education and Social Inequality: of the Quality of Societies: Concept and Measurement.” Bian, L., S.J. Leslie and A. Cimpian. 2017. “Gender An International Perspective. Northampton, MA: Edward EuReporting Working Paper 14. Zentrum für Umfrage, Stereotypes about Intellectual Ability Emerge Early Elgar Publishing. Methoden und Analysen, Mannheim, Germany. www. and Influence Children’s Interests.” Science 355(6323): gesis.org/fileadmin/upload/dienstleistung/daten/soz_indi- 389­391. Blum, R.W., and W.H. Gates, Sr. 2015. Girlhood Not katoren/eusi/paper14.pdf. Accessed 1 August 2019. Motherhood: Preventing Adolescent Pregnancy. New York: Bicchieri, C. 2006. The Grammar of Society: The Nature and United Nations Population Fund. Bernardi, F. 2014. “Compensatory Advantage as a Dynamics of Social Norms. Cambridge, UK: Cambridge Mechanism of Educational Inequality: A Regression University Press. Blundell, R., R. Joyce, A.N. Keiller and J.P. Ziliak. 2018. Discontinuity Based on Month of Birth.” Sociology of “Income Inequality and the Labour Market in Britain and Education 87(2): 74­88. Biernat, M., and A.K. Sesko. 2013. “Evaluating the the US.” Journal of Public Economics 162: 48­62. Contributions of Members of Mixed-Sex Work Teams: Bernardi, F., and G. Ballarino, eds. 2016. Education, Race and Gender Matter.” Journal of Experimental Social Blundell, R., R. Joyce, A. Norris and J. Ziliak. 2018. Occupation and Social Origin: A Comparative Analysis Psychology 49(3): 471­476. “Income Inequality and the Labour Market in Britain and of the Transmission of Socio-Economic Inequalities. the US.” Journal of Public Economics 162: 48­62. Northampton, MA: Edward Elgar Publishing. Bill & Melinda Gates Foundation. 2019. Examining Inequality: How Geography and Gender Stack the Deck for Boillat, S., J.D. Gerber, C. Oberlack, J.G. Zaehringer, Bernardi, F., and H.C. Boado. 2013. “Previous School (or against) You. Seattle, WA. C.I. Speranza and S. Rist. 2018. “Distant Interactions, Results and Social Background: Compensation and Power, and Environmental Justice in Protected Area Imperfect Information in Educational Transitions.” Bircan, Ç., T. Brück and M. Vothknecht. 2017. “Violent Governance: A Telecoupling Perspective.” Sustainability European Sociological Review 30(2): 207­217. Conflict and Inequality.” Oxford Development Studies 10(11): 3954. 45(2): 125­144. Bernardi, F., and I. Plavgo. Forthcoming. “Education as Boldrin, M., and D. Levine. 2013. “The Case against an Equalizer for Human Development?” Background Paper Birdsall, N., D. Ross and R. Sabot. 1995. “Inequality and Patents.” Journal of Economic Perspectives 27(1): 3­22. for Human Development Report 2019. United Nations Growth Reconsidered: Lessons from East Asia.” World Development Programme, Human Development Report Bank Economic Review 9(3): 477­508. Bolch, K B., L. Ceriani and L.F. Lopez-Calva. 2017. Office, New York. “Arithmetics and Politics of Domestic Resource Bixby Center for Population Health and Sustainability. Mobilization.” Policy Research Working Paper 8029. World Berthe, A., and L. Elie. 2015. “Mechanisms Explaining 2014. “Ethiopia: CBD of Injectable Contraceptives Bank, Washington, DC. the Impact of Economic Inequality on Environmental Scaling Up Community-Based Distribution of Injectable Deterioration.” Ecological Economics 116(C): 191­200. Contraceptives in Tigray, Ethiopia.” Berkeley, CA. Bolt, J., R. Inklaar, H. de Jong and J.L. van Zanden. 2018. Maddison Project Database, version 2018. Besley, T. 2017. “Aspirations and the Political Economy of Björkenstam, E., S. Cheng, B. Burström, A.R. Pebley, University of Groningen, Maddison Project, Groningen, the Inequality.” Oxford Economic Papers 69(1): 1­35. C. Björkenstam and K. Kosidou. 2017. “Association Netherlands. between Income Trajectories in Childhood and Psychiatric

270 | HUMAN DEVELOPMENT REPORT 2019 Bolton, P., T. Santos and J.A. Scheinkman. 2016. Brei, M., G. Ferri and L. Gambacorta. 2018. “Financial for Occupations and the Economy?” AEA Papers and “Cream-Skimming in Financial Markets.” Journal of Structure and Income Inequality.” BIS Working Paper 756. Proceedings 108: 43­47. Finance 71(2): 709­736. Bank of International Settlements, Basel, Switzerland. Bublitz, E. 2016. “Perceptions of Inequality Survey Borrell-Porta, M., J. Costa-Font and J. Philipp. 2018. Broadband Commission for Digital Development 2015/2016.” Bertelsmann Stiftung, Gütersloh, Germany. “The Mighty Girl' Effect: Does Parenting Daughters Alter Working Group on Broadband and Gender. 2015. Attitudes towards Gender Norms?" Oxford Economic Cyber Violence against Women and Girls: A Worldwide Buchan, I.E., E. Kontopantelis, M. Sperrin, T. Chandola Papers 71(1): 25­46. Wake Up Call. Geneva: International Telecommunications and T. Doran. 2017. "North-South Disparities in English Union, and Paris: United Nations Educational, Scientific Mortality1965­2015: Longitudinal Population Study." Bouguen, A., Y. Huang, M. Kremer and E. Miguel. and Cultural Organization. Journal of Epidemiology and Community Health 71(9): 2018. "Using RCTs to Estimate Long-Run Impacts in 928­936. Development Economics." NBER Working Paper 25356. ------. 2019. "The State of Broadband: Broadband as National Bureau of Economic Research, Cambridge, MA. a Foundation for Sustainable Development." Geneva: Bullard, R.D. 1983. "Solid Waste Sites and the Black International Telecommunications Union, and Paris: Houston Community." Sociological Inquiry 53(2­3): Bouis, H.E., P. Eozenou and A. Rahman. 2011. "Food United Nations Educational, Scientific and Cultural 273­288. Prices, Household Income, and Resource Allocation: Organization. www.itu.int/dms_pub/itu-s/opb/pol/S-POL- Socioeconomic Perspectives on their Effects on Dietary BROADBAND.20-2019-PDF-E.pdf. Accessed14 October ------. 1990. Dumping in Dixie: Race, Class, and Quality and Nutritional Status." Food and Nutrition 2019. Environmental Quality. New York, NY: Routledge. Bulletin 32(1_suppl1): S14-S23. Brody, A., J. Demetriades and E. Esplen. 2008. "Gender Burke, M., W.M. Davis and N.S. Diffenbaugh. 2018. Boulamwini, J., and T. Gebru. 2018. "Gender Shades: and Climate Change: Mapping the Linkages: A Scoping "Large Potential Reduction in Economic Damages under Intersectional Accuracy Disparities in Commercial Study on Knowledge and Gaps." University of Sussex, UN Mitigation Targets." Nature 557(7706): 549­553. Gender Classification." Proceedings of Machine Learning Institute of Development Studies, Brighton, UK. Research 81: 1­15. Burke, M., and V. Tanutama. 2019. "Climatic Constraints Brønnum-Hansen, H. 2017. "Socially Disparate Trends in on Aggregate Economic Output." NBER Working Bourguignon, F. 2003. "The Growth Elasticity of Poverty Lifespan Variation: A Trend Study on Income and Mortality Paper 25779. National Bureau of Economic Research, Reduction: Explaining Heterogeneity across Countries and Based on Nationwide Danish Register Data." BMJ Open Cambridge, MA. www.nber.org/papers/w25779.pdf. Time Periods." Working Paper. World Bank, Washington, 7(5): e014489. Accessed 8 August 2019. DC. http://documents.worldbank.org/curated/ en/503161468780002293/pdf/28104.pdf. Accessed 1 Broockman, D., and J. Kalla. 2016. "Durably Reducing Burton, R.J. 2019. "The Potential Impact of Synthetic Animal August 2019. Transphobia: A Field Experiment on Door-to-door Protein on Livestock Production: The New "War Against Canvassing." Science 352(6282): 220­224. Agriculture"?" Journal of Rural Studies 68: 33­45. ------. 2015a. "Appraising Income Inequality Databases in Latin America." Journal of Economic Inequality 13: Brooks, W. J., J.P. Kaboski, Y.A. Li and W. Qian. 2019. Busso, M., and S. Galiani. 2019. "The Causal Effect of 557­578. "Exploitation of Labor? Classical Monopsony Power and Competition on Prices and Quality: Evidence from a Labor's Share." NBER Working Paper 25660. National Field Experiment." American Economic Journal: Applied ------. 2015b. "Revisiting the Debate on Inequality and Bureau of Economic Research, Cambridge, MA. Economics, 11 (1): 33­56. Economic Development." Revue d'économie politique 125(5): 633­663. Brown, C., M. Ravallion and D. Van de Walle. 2017. Bussolo, M., D. Checchi and V. Peragine. 2019. "Long- "Are Poor Individuals Mainly Found in Poor Households? Term Evolution of Inequality of Opportunity." Policy Bourguignon F., and C. Morrisson. 1998. "Inequality Evidence Using Nutrition Data for Africa." NBER Working Research Working Paper 8700. World Bank, Washington, and Development: The Role of Dualism." Journal of Paper 24047. National Bureau of Economic Research, DC. Development Economics 57: 233­257. Cambridge, MA. Butera, S. 2019. "Rwandan Central Bank Studying Ways ------. 2002. "Inequality among World Citizens: 1820- Bruckner, M., M. LaFleur and I. Pitterle. 2017. "Frontier of Issuing Digital Currency." BNN Bloomberg, 22 August. 1992." American Economic Review 92(4): 727­744. Issues: The Impact of the Technological Revolution on www.bnnbloomberg.ca/rwandan-central-bank-studying- Labour Markets and Income Distribution." United Nations ways-of-issuing-digital-currency-1.1305321. Accessed 6 Bowles, S., C.M. Fong, H. Gintis and U. Pagano. 2012. Department of Economic and Social Affairs, New York. October 2019. The New Economics of Inequality and Redistribution. Cambridge, UK: Cambridge University Press. Brulle, R.J. 2018. "The Climate Lobby: a Sectoral Analysis of Butler, J. 2019. "Judith Butler: The Backlash against Gender Lobbying Spending on Climate Change in the USA, 2000 to Ideology’ Must Stop.” New Statesman America, 21 Bradbury, B., M. Corak, J. Waldfogel and E. 2016.” Climatic Change 149: 289­303. January. www.newstatesman.com/2019/01/judith-butler- Washbrook. 2015. Too Many Children Left Behind: The backlash-against-gender-ideology-must-stop. Accessed 19 US Achievement Gap in Comparative Perspective. New Brunori, P., F.H.G. Ferreira and V. Peragine. 2013. August 2019. York: Russell Sage Foundation. “Inequality of Opportunity, Income Inequality and Economic Mobility: Some International Comparisons.” Buttrick, N.R., and S. Oishi. 2017. “The Psychological Bradsher, K., and K. Bennhold. 2019. “World Leaders at IZA Working Paper. Institute for the Study of Labor, Bonn, Consequences of Income Inequality.” Social and Davos Call for Global Rules on Tech.” New York Times, Germany. http://anon-ftp.iza.org/dp7155.pdf. Accessed 12 Personality Psychology Compass 11(3): 12304. 23 January. www.nytimes.com/2019/01/23/technology/ September 2019. world-economic-forum-data-controls.html. Accessed 19 Byers, E., M. Gidden, D. Leclere, P. Burek, K. Ebi, P. August 2019. Bruns, B., and J. Luque. 2015. Great Teachers: How Greve, D. Grey and P. Havlik. 2018. “Global Exposure to Raise Student Learning in Latin America and the and Vulnerability to Multi-Sector Development and Bragg, F., M.V. Holmes, A. Iona, Y. Guo, H. Du, Y. Chen, Caribbean. Washington, DC: World Bank. Climate Change Hotspots.” Environmental Research Z. Bian, L. Yang, W. Herrington and D. Bennett. Letters 13(5). 2017. “Association between Diabetes and Cause-Specific Brunwasser, M. 2019. “Comforting the Afflicted and Mortality in Rural and Urban Areas of China.” Journal of Afflicting the Comfortable: The Impact of Investigative Cai, Y., K.L. Judd, T.M. Lenton, T.S. Lontzek and D. the American Medical Association 317(3): 280­289. Journalism on Inequality.” Background paper for Human Narita. 2015. “Environmental Tipping Points Significantly Development Report 2019. United Nations Development Affect the Cost-Benefit Assessment of Climate Policies.” Branstetter, L.G., B. Glennon and J.B. Jensen. 2019. Programme, Human Development Report Office, New York. Proceedings of the National Academy of Sciences 112(15): “The Rise of Global Innovation by US Multinationals 4606­4611. Poses Risks and Opportunities.” PIIE Policy Brief. Peterson Brussevich, M., E. Dabla-Norris and S. Khalid. 2019. “Is Institute for International Economics, Washington DC. Technology Widening the Gender Gap? Automation and Cai, Y., K.L. Judd and T.S. Lontzek. 2013. “The Social Cost the Future of Female Employment.” IMF Working Paper of Stochastic and Irreversible Climate Change.” NBER Braveman P., and L. Gottlieb. 2014. “The Social WP/19/91. International Monetary Fund, Washington, DC. Working Paper 18704. National Bureau of Economic Determinants of Health: It’s Time to Consider the Causes Research, Cambridge, MA. www.nber.org/papers/w18704. of the Causes.” Public Health Reports 129(Suppl 2): 19­31. Brynjolfsson, E., T. Mitchell and D. Rock. 2018. pdf. Accessed 24 October 2019. “What Can Machines Learn and What Does It Mean

References | 271 Cakal, H., M. Hewstone, G. Schwär and A. Heath. Signaled by Vertebrate Population Losses and Declines.” Equitable Adaptation Fund.” Working Paper 2015/7. 2011. “An Investigation of the Social Identity Model of Proceedings of the National Academy of Sciences 114(30): World Inequality Lab, Paris. https://wid.world/document/ Collective Action and the `Sedative’ Effect of Intergroup E6089­E6096. chancel-l-piketty-t-carbon-and-inequality-from-kyoto- Contact among Black and White Students in South to-paris-wid-world-working-paper-2015-7/. Accessed 9 Africa.” British Journal of Social Psychology 50: 606­627. Ceballos, G., A. García and P.R. Ehrlich. 2010. “The Sixth August 2019. Extinction Crisis: Loss of Animal Populations and Species.” Caliskan, A., J.J. Bryson and A. Narayanan. 2017. Journal of Cosmology 8: 1821­1831. ------. 2017. “Indian Income Inequality, 1922-2015: From “Semantics Derived Automatically from Language Corpora British Raj to Billionaire Raj?” WID.world Working Paper Contain Human-Like Biases.” Science 356(6334): 183­186. Cederman, L.E., K.S. Gleditsch and H. Buhaug. 2013. 2017/11. World Inequality Database. Inequality, Grievances, and Civil War. Cambridge, UK: Campbell, B.M., D.J. Beare, E.M. Bennett, J.M. Cambridge University Press. Chandra-Mouli, V., A.V. Camacho and P.A. Michaud. Hall-Spencer, J.S. I. Ingram, F. Jaramillo, R. Ortiz, 2013. “WHO Guidelines on Preventing Early Pregnancy N. Ramankutty, J.A. Sayer and D. Shindell. 2017. Cefai, C., P.A. Bartolo, V. Cavioni and P. Downes. 2018. and Poor Reproductive Outcomes among Adolescents “Agriculture Production as a Major Driver of the Earth “Strengthening Social and Emotional Education as a in Developing Countries.” Journal of Adolescent Health System Exceeding Planetary Boundaries.” Ecology and Core Curricular Area across the EU. A Review of the 52(5): 517­522. Society 22(4): 8. International Evidence.” NESET II Report. Publications Office of the European Union, Luxembourg. https:// Charles, A. 2012. “Identity, Norms, and Ideals.” In Exchange Caprioli, M. 2005. “Primed for Violence: The Role of Gender boa.unimib.it/retrieve/handle/10281/188490/268947/ Entitlement Mapping: Theories and Evidence. Perspectives Inequality in Predicting Internal Conflict.” International Strengthening-Social-and-Emotional-Education.pdf. from Social Economics. New York: Palgrave Macmillan. Studies Quarterly 49(2): 161­178. Accessed 30 July 2019. Chau, N.H., and R. Kanbur. 2018. “Employer Power, Cardona, O.D., M.K. van Aalst, J. Birkmann, M. Cerra, V., and S.C. Saxena. 2008. “Growth Dynamics: The Labor Saving Technical Change, and Inequality.” CEPR Fordham, G. McGregor, R. Perez, R.S. Pulwarty, E.L.F. Myth of Economic Recovery.” American Economic Review Discussion Paper DP12925. Centre for Economic Policy Schipper and B.T. Sinh. 2012. “Determinants of Risk: 98(1): 439­457. Research, London. Exposure and Vulnerability.” In C.B. Field, V. Barros, T.F. Stocker, D. Qin, D.J. Dokken, K.L. Ebi, M.D. Mastrandrea, Chacaltana, J., V. Leung and M. Lee. 2018. “New Chemouni, B. 2018. “The Political Path to Universal Health K.J. Mach, G.-K. Plattner, S.K. Allen, M. Tignor and P.M. Technologies and the Transition to Formality: The Trend Coverage: Power, Ideas and Community-Based Health Midgley, eds., A Special Report of Working Groups I and towards E­formality.” ILO Employment Working Paper Insurance in Rwanda.” World Development 106: 87­98. II of the Intergovernmental Panel on Climate Change. 247. International Labour Office, Geneva. Cambridge, UK: Cambridge University Press. Chen, M. 2019. “Informality and Inequality: In a Globalized Chadwick, A. 2017. The Hybrid Media System: Politics and and Urbanized World.” Background paper for Human Cariboni, D. 2014. “Colombia Rice Growers Saved from Ruin Power, 2nd ed. New York: Oxford University Press. Development Report 2019. United Nations Development after Being Told Not to Plant Their Crop.” The Guardian, Programme, Human Development Report Office, New York. 30 September. www.theguardian.com/global-devel- Chadwick, L., and G. Solon. 2002. “Intergenerational opment/2014/sep/30/colombia-rice-growers-climate- Income Mobility among Daughters.” American Economic Chen, P., L. Karabarbounis and B. Neiman. 2017. “The change. Accessed 15 August 2019. Review 92(1): 335­344. Global Rise of Corporate Saving.” Journal of Monetary Economics 89: 1­19. Carleton, T.A., and S.M. Hsiang. 2016. “Social and Chaigneau, T., and K. Brown. 2016. “Challenging the Economic Impacts of Climate.” Science 353(6304). Win-Win Discourse on Conservation and Development: Chen, Y., P. Persson and M. Polyakova. 2019. “The Analyzing Support for Marine Protected Areas.” Ecology Roots of Health Inequality and the Value of Intra-Family Caron, J., and T. Fally. 2018. “Per Capita Income, and Society 21(1): 36. Expertise.” Working Paper 25618. National Bureau of Consumption Patterns, and CO2 Emissions.” NBER Economic Research, Cambridge, MA. www.nber.org/ Working Paper 24923. National Bureau of Economic Challinor, A.J., A.K. Koehler, J. Ramirez-Villegas, S. papers/w25618.pdf. Accessed 1 August 2019. Research, Cambridge, MA Whitfield and B. Das. 2016. “Current Warming Will Reduce Yields Unless Maize Breeding and Seed Systems Chenery, H., M.S. Ahluwalia, J.H. Duloy, C.L.G. Bell Carr, M., and E.E. Wiemers. 2016. “The Decline in Lifetime Adapt Immediately.” Nature Climate Change 6(10): and R. Jolly. 1974. Redistribution with Growth; Policies Earnings Mobility in the US: Evidence from Survey-Linked 954­958. to Improve Income Distribution in Developing Countries Administrative Data.” Washington Center for Equitable in the Context of Economic Growth. Oxford, UK: Oxford Growth, Washington, DC. www.sole-jole.org/16399.pdf. Chamorro-Premuzic, T. 2013. “Why Do So Many University Press. Accessed 16 August 2019. Incompetent Men Become Leaders?” Harvard Business Review, 22 August. Cheng, S., B. Chauhan and S. Chintala. 2019. “The Rise Carroll, A.E., and T.S. Doherty. 2019. “Meat Consumption of Programming and the Stalled Gender Revolution.” and Health: Food For Thought.” Annals of Internal Chancel, L. 2017. Insoutenables inégalités : Pour une justice Sociological Science 6: 321­351. Medicine, 1 October. sociale et environnementale. Paris: Les Petits Matins. Chetty, R., N. Hendren and L.F. Katz. 2016. “The Effects Case, A., and A. Deaton. 2017. “Mortality and Morbidity in Chancel, L., R. Clarke and A. Gethin. 2017. “World of Exposure to Better Neighborhoods on Children: New the 21st Century.” Brookings Papers on Economic Activity: Inequality Report 2018 Technical Notes for Figures Evidence from the Moving to Opportunity Experiment.” 397­476. and Tables.” WID.world Technical Note 2017/8. World American Economic Review 106(4): 855­902. Inequality Database. http://wid.world/static/technical- Case, A., and C. Paxson. 2008. “Height, Health, and notes-for-figures-and-tables.pdf. Accessed 10 October Chetty, R., N. Hendren, P. Kline, E. Saez and N. Turner. Cognitive Function at Older Ages.” American Economic 2019. 2014. “Is the United States Still a Land of Opportunity? Review 98(2): 463­467. Recent Trends in Intergenerational Mobility.” American Chancel, L., D. Cogneau, A. Gethin and A. Myczkowski. Economic Review 104(5): 141­147. ------. 2010. “Causes and Consequences of Early-Life 2019. “Income Inequality in Africa, 1990-2017.” WID. Health.” Demography 47(1): S65­S85. world Issue Brief 2019/6. World Inequality Database. Chetty, R., M. Stepner, S. Abraham, S. Lin, B. Scuderi, N. Turner, A. Bergeron and D. Cutler. 2016. “The Cattaneo, C., and G. Peri. 2016. “The Migration Response Chancel, L., and L. Czajka. 2017. “Estimating the Regional Association between Income and Life Expectancy in the to Increasing Temperatures.” Journal of Development Distribution of Income in Sub-Saharan Africa.” WID.world United States, 2001-2014.” Journal of the American Economics 122: 127­146. Technical Note 2017/6. World Inequality Database. Medical Association 315(16): 1750­1766.

Ceballos, G., P.R. Ehrlich, A.D. Barnosky, A. García, Chancel, L., A. Hough and T. Voituriez. 2018. “Reducing Chevalier, A., and G. Lanot. 2001. “The Relative Effect R.M. Pringle and T.M. Palmer. 2015. “Accelerated Inequalities within Countries: Assessing the Potential of of Family and Financial Characteristics on Educational Modern Human­Induced Species Losses: Entering the the Sustainable Development Goals.” Global Policy 9(1): Achievement.” Centre for the Economics of Education, Sixth Mass Extinction.” Science Advances 1(5). 5­16. London School of Economics and Political Science, London. Ceballos, G., P.R. Ehrlich and R. Dirzo. 2017. “Biological Chancel, L., and T. Piketty. 2015. “Carbon and Inequality: Annihilation via the Ongoing Sixth Mass Extinction from Kyoto to Paris: Trends in the Global Inequality of Carbon Emissions (1998-2013) & Prospects for an

272 | HUMAN DEVELOPMENT REPORT 2019 Chiam, Z., S. Duffy and M.G. Gil. 2017. Trans Legal Cole, M.J., R.M. Bailey, J.D. Cullis and M.G. New. 2018. Creedy, J., and R. Dixon. 1998. “The Relative Burden Mapping Report 2017: Recognition before the Law. “Spatial Inequality in Water Access and Water Use in of Monopoly on Households with Different Incomes.” Geneva: International Lesbian, Gay, Bisexual, Trans and South Africa.” Water Policy 20(1): 37­52. Economic New Series 65(258): 285­293. Intersex Association. Collier, P., and A. Hoeffler. 1998. “On Economic Causes of Criado-Pérez, C. 2019. Invisible Women: Data Bias in a Choy, C.A., B.H. Robison, T.O. Gagne, B. Erwin, E. Civil War.” Oxford Economic Papers 50(4): 563­573. World Designed for Men. New York: Abrams Press. Firl, R.U. Halden, J.A. Hamilton, K. Katija, S.E. Lisin, C. Rolsky and K.S. Van Houtan. 2019. “The Conceição, P. 2019a. “Fear and Loathing of Technological Crocker, D. 2008. “Sen’s Concept of Agency.” University of Vertical Distribution and Biological Transport of Marine Progress? Leveraging Science and Innovation for the Maryland, Silver Spring, MD. Microplastics across the Epipelagic and Mesopelagic Implementation of the 2030 Agenda for Sustainable Water Column.” Scientific Reports 9(1). Development.” In A. Baimenov and P. Liverakos, eds., Crosby, A. 1986. Ecological Imperialism. Cambridge, UK: Public Service Excellence in the 21st Century. Singapore: Cambridge University Press. Christiansen, C.O., and S.L.B. Jensen, eds. 2019. Palgrave Macmillan. Histories of Global Inequality: New Perspectives. New Cruces, G., R. Pérez-Truglia and M. Tetaz. 2013. “Biased York: Palgrave Macmillan. ------. 2019b. “How Science and Technology Policy Shape Perceptions of Income Distribution and Preferences for Inequality.” In Financing the UN Development System: Redistribution: Evidence from a Survey Experiment.” Cialdini, R.B., C.A. Kallgren and R.R. Reno. 1991. “A Time for Hard Choices. Dag Hammarskjöld Foundation and Journal of Public Economics 98: 100­112. Focus Theory of Normative Conduct: A Theoretical United Nations Multi-Partner Trust Fund Office. Uppsala, Refinement and Reevaluation of the Role of Norms in Sweden, and New York. Cuaresma, J.C., W. Fengler, H. Kharas, K. Bekhtiar, M./ Human Behavior.” Advances in Experimental Social Brottrager and M. Hofer. 2018. “Will the Sustainable Psychology 24: 201­234). Conceição, P., and J.K. Galbraith. 2001. “A New Development Goals Be Fulfilled? Assessing Present and Kuznets Hypothesis: Theory and Evidence on Growth Future Global Poverty.” Palgrave Communications 4(29). Cingano, F. 2014. “Trends in Income Inequality and Its and Inequality.” In J.K. Galbraith and M. Berner, eds., Impact on Economic Growth.” OECD Social, Employment Inequality and Industrial Change: A Global View. New Cuberes, D., and M. Teignier. 2012. “Gender Gaps in the and Migration Working Paper 163. Organisation for York: Cambridge University Press. Labor Market and Aggregate Productivity.” Working Paper. Economic Co-operation and Development, Paris. www. University of Sheffield, Department of Economics, UK. oecd.org/els/soc/trends-in-income-inequality-and-its- Conceição, P., and M.V. Heitor. 2007. “Diversity and http://eprints.whiterose.ac.uk/74398/1/serps_2012017. impact-on-economic-growth-SEM-WP163.pdf. Accessed Integration of Science and Technology Policies.” pdf. Accessed 14 October 2019. 23 August 2019. Technological Forecasting and Social Change 74(1): 1­17. Cumming, G.S., and S. von Cramon-Taubadel. 2018. Cislaghi, B., K. Manji and L. Heise. 2018. Social Norms Connolly, M., M. Corak and C. Haeck. 2019. “Linking Economic Growth Pathways and Environmental and Gender-Related Harmful Practices: What Assistance “Intergenerational Mobility between and within Canada Sustainability by Understanding Development as Alternate from the Theory to the Practice? Learning Report 02. and the United States.” Journal of Labor Economics Social­Ecological Regimes.” Proceedings of the National London: London School of Hygiene & Tropical Medicine. 37(S2): S595­S641. Academy of Sciences 115(38): 9533­9538.

Clarke, D.J., and S. Dercon. 2016. Dull Disasters? How Cooper, L.B., and E. Fletcher. 2013. “Reducing Societal Cunningham, M. 2001. “The Influence of Parental Attitudes Planning Ahead Will Make a Difference. New Delhi: SAGE Discrimination against Adolescent Girls Using Social and Behaviours on Children’s Attitudes towards Gender Publications. Norms to Promote Behavior Change.” Girl Hub, London. and Household Labor in Early Adulthood.” Journal of Marriage and Family 63(1): 111­122. Clarke, G, L. Xu and H-F. Zou. 2006. “Finance and Income Cooper, R. 2019. “The Deadly Hidden Risks within the Most Inequality: What Do the Data Tell Us?” Southern Economic Prominent Economic Model of Climate Change.” The Curran, D. 2018. “Environmental Justice Meets Risk-Class: Journal 72(3): 578­596. Week, 4 September. The Relational Distribution of Environmental Bads.” Antipode 50(2): 298­318. Clayton, K., S. Blair, J.A. Busam, S. Forstner, J. Glance, Copeland, B.R., and M.S. Taylor. 1994. “North-South Trade G. Green, A. Kawata and others. Forthcoming. “Real and the Environment.” Quarterly Journal of Economics Curran, M., and M.C. Mahutga. 2018. “Income Inequality Solutions for Fake News? Measuring the Effectiveness of 109(3): 755­787. and Population Health: A Global Gradient?” Journal of General Warnings and Fact-Check Tags in Reducing Belief Health and Social Behavior 59(4): 536­553. in False Stories on Social Media.” Political Behavior. Corak, M. 2013. “Income Inequality, Equality of Opportunity, and Intergenerational Mobility.” Journal of Economic Currie, J. 2009. “Healthy, Wealthy, and Wise: Coady, D. 2018. “Creating Fiscal Space.” Finance and Perspectives 27(3): 79­102. Socioeconomic Status, Poor Health in Childhood, and Development 55(4): 24­27. Human Capital Development.” Journal of Economic Corcoran, K.E., D. Pettinicchio and J.T. Young. 2011. Literature 47(1): 87­122. Coady, D., D. D’Angelo and B. Evans. 2019. “Fiscal “The Context of Control: A Cross-National Investigation Redistribution and Social Welfare.” IMF Working Paper of the Link between Political Institutions, Efficacy, and ------. 2011. “Inequality at Birth: Some Causes and WP/19/51. International Monetary Fund, Washington, DC. Collective Action.” British Journal of Social Psychology Consequences.” American Economic Review 101(3): 1­22. 50: 575­605. Coady, D., and A. Dizioli. 2018. “Income Inequality and Currie, J., and H. Schwandt. 2016. “Inequality in Mortality Education Revisited: Persistence, Endogeneity and Coronese, M., F. Lamperti, K. Keller, F. Chiaromonte Decreased among the Young While Increasing for Older Heterogeneity.” Applied Economics 50(2): 1­15. and A. Roventini. 2019. “Evidence for Sharp Increases Adults, 1990­2010.” Science 352(6286): 708­712. in the Economic Damages of Extreme Natural Disasters.” Coady, D., and D. Prady. 2018. “Universal Basic Income in Proceedings of the National Academy of Sciences 116(43): Currie, J., H. Schwandt and J. Thuilliez. 2018. “Pauvreté, Developing Countries: Issues, Options and Illustrations 21450­21455. Egalité, Mortalité: Mortality (In) Equality in France and the for India.” IMF Working Paper WP/18/174. International United States.” Journal of Population Economics 1­35. Monetary Fund, Washington, DC. Costa, D., and M.E. Kahn. 2003. “Civic Engagement and Community Heterogeneity: An Economist’s Perspective.” Cutler, D.M., and A. Lleras-Muney. 2010. “Understanding Cohen, J., R. Desai and H. Kharas. 2019. “Spatial Perspectives on Politics 1: 103­111. Differences in Health Behaviors by Education.” Journal of Targeting of Poverty Hotspots.” In H. Kharas, J. McArthur Health Economics 29(1): 1­28. and I. Ohno, eds., Leave No One Behind: Time for Specifics Covarrubias, M., G. Gutiérrez and T. Philippon. 2019. on the Sustainable Development Goals. Washington, DC: “From Good to Bad Concentration? US Industries over Czajka, L. 2017. “Income Inequality in Côte d’Ivoire: The Brookings Institution. the Past 30 years.” NBER Working Paper 25983. National 1985­2014.” WID.world Working Paper 2017/8. World Bureau of Economic Research, Cambridge, MA. Inequality Database. Cohen, G., J.T. Jalles, P. Loungani and R. Marto. 2018. “The Long-Run Decoupling of Emissions and Output: Coyle, D. 2015. GDP: A Brief but Affectionate History: Dahlum, S., H.M. Nygard, S.A. Rustad and G. Ostby. Evidence from the Largest Emitters.” IMF Working Paper. Revised and Expanded Edition. Princeton, NJ: Princeton Forthcoming. “The Conflict­Inequality Trap? Linking International Monetary Fund, Washington, DC. University Press. Internal Armed Conflict to Horizontal Inequality.” Background Paper for Human Development Report 2019. United Nations Development Programme, Human Development Report Office, New York.

References | 273 Dang, H.A., P.F. Lanjouw and R. Swinkels. 2014. “Who ------. 2017. “Without Governments, Would Countries Diez, M.F.J., J. Fan and C. Villegas-Sánchez. 2019. Remained in Poverty, Who Moved Up, and Who Fell Have More Inequality, or Less?” The Economist, 13 July. “Global Declining Competition” IMF Working Paper Down? An Investigation of Poverty Dynamics in Senegal www.economist.com/the-world-if/2017/07/13/without- WP/19/82. International Monetary Fund, Washington, DC. in the Late 2000s.” Policy Research Working Paper 7141. governments-would-countries-have-more-inequality-or- World Bank, Washington, DC. https://papers.ssrn.com/ less. Accessed 10 October 2019. Diffenbaugh, N.S., and M. Burke. 2019a. “Global sol3/papers.cfm?abstract_id=2540771. Accessed 16 Warming Has Increased Global Economic Inequality.” August 2019. ------. 2018. “How Inequality Works”. Mint, 1 January. Proceedings of the National Academy of Sciences 116(20): www.livemint.com/Opinion/sMRTHlLePT4cfXTkjM7JOM/ 9808­9813. Daniel, K., R.B. Litterman and G. Wagner. 2019. Angus-Deaton—How-inequality-works.html. Accessed 10 “Declining CO2 Price Paths.” Proceedings of the National September 2019. ------. 2019b. “Reply to Rosen: Temperature­Growth Academy of Sciences 116(42): 20886­20891. Relationship Is Robust.” Proceedings of National De Haan, J., J.E. Sturm. 2017. “Finance and Income Academies of Sciences 116(33): 16171­16172. Danks, D., and A.J. London. 2017. “Algorithmic Bias in Inequality: A Review and New Evidence.” European Autonomous Systems.” In C. Sierra, ed., Proceedings Journal of Political Economy 50: 171­195. Dimova, D. 2019. “The Structural Determinants of the Labor of the Twenty-Sixth International Joint Conference on Share in Europe.” IMF Working Paper 19/67. International Artificial Intelligence. Freiburg, Germany: International Del Boca, D. 2015. “Child Care Arrangements and Labour Monetary Fund, Washington, DC. Joint Conferences on Artificial Intelligence. Supply.” IDB Working Paper 569. Inter-American Development Bank, Washington, DC. Dinesen, P.T., and K.M. Sønderskov. 2015. “Ethnic Dao, M.C., M. Das, Z. Koczan and W. Lian 2017. “Why Is Diversity and Social Trust: Evidence from the Micro- Labor Receiving a Smaller Share of Global Income? Theory Dellinger, A.J. 2019. “How the Biggest Tech Companies Context.” American Sociological Review 80(3): 550­573. and Empirical Evidence. IMF Working Paper WP/17/169. Spent Half A Billion Dollars Lobbying Congress.” Forbes, International Monetary Fund, Washington, DC. 30 April. Dingel, J.I., K.C. Meng and S.M. Hsiang. 2019. “Spatial Correlation, Trade, and Inequality: Evidence from the Daude, C., N. Lustig, A. Melguizo and J. R. Perea. De Loecker, J., and J. Eeckhout. 2017. “The Rise of Global Climate.” NBER Working Paper 25447. National 2017. “On the Middle 70%: The Impact of Fiscal Policy Market Power and the Macroeconomic Implications.” Bureau of Economic Research, Cambridge, MA. www.nber. on the Emerging Middle Class in Latin America Using NBER Working Paper 23687. National Bureau of Economic org/papers/w25447.pdf. Accessed 13 August 2019. Commitment to Equity.” Working Papers 1716. Tulane Research, Cambridge, MA. University, Department of Economics, New Orleans, LA. Dirzo, R., H.S. Young, M. Galetti, M., Ceballos, N.J. Demaria, F. 2010. “Shipbreaking at Alang-Sosiya (India): An Isaac and B. Collen. 2014. “Defaunation in the David, P. 1990. “The Dynamo and the Computer: An Ecological Distribution Conflict.” Ecological Economics Anthropocene.” Science 345(6195): 401­406. Historical Perspective on the Modern Productivity 70(2): 250­260. Paradox.” American Economic Review 80(2): 255­361. Dixon, J., K. Durrheim and C. Tredoux. 2005. “Beyond the Demirgüç-Kunt, A., L. Klapper and D. Singer. 2013. Optimal Contact Strategy: A Reality Check for the Contact David, A., N. Guilbert, H. Hino, M. Leibbrandt, E. “Financial Inclusion and Legal Discrimination against Hypothesis.” American Psychologist 60: 697­711. Potgieter and M. Shiva. 2018. “Social Cohesion and Women: Evidence from Developing Countries.” Policy Inequality in South Africa.” Research Paper 2018-63. Research Working Paper 6416. World Bank, Washington, Dobson, A.P., and E.R. Carper. 1996. “Infectious Diseases Agence Française de Développement, Paris. www.afd.fr/ DC. and Human Population History.” Bioscience 46(2): en/social-cohesion-and-inequality-south-africa. Accessed 115­126. 26 August 2019. Demirgüç-Kunt, A., L. Klapper, D. Singer, S. Ansar and J. Hess. 2018. The Global Findex Database 2017: Dollar, D., T. Kleineberg and A. Kraay. 2015. “Growth, Davis, S.J., N.S. Lewis, M. Shaner, S. Aggarwal, D. Measuring Financial Inclusion and the Fintech Revolution. Inequality and Social Welfare: Cross-Country Evidence.” Arent, I.L. Azevedo, S.M. Benson, T. Bradley, J. Washington, DC: World Bank. Economic Policy 30(82): 335­377. Brouwer, Y-M. Chiang, C.T.M. Clack, A. Cohen, S. Doig, J. Edmonds, P. Fennell, C.B. Field, B. Demirgüç-Kunt, A., and R. Levine. 2009. “Finance and Dorband I., M. Jakob, M. Kalkhul and J. Steckel. Hannegan, B-M. Hodge, M.I. Hoffert, E. Ingersoll, P. Inequality: Theory and Evidence.” Annual Review of 2019. “Poverty and Distributional Effects of Carbon Jaramillo, K.S. Lackner, K.J. Mach, M. Mastrandrea, Financial Economics 1(1): 287­318. Pricing in Low- and Middle-Income Countries—A Global J. Ogden, P.F. Peterson, D.L. Sanchez, D. Sperling, Comparative Analysis.” World Development. 115: J. Stagner, J.E. Trancik, C-J. Yang and K. Caldeira. Dercon, S. 2010. “Risk, Poverty and Human Development: 246­257. 2018. “Net-Zero Emissions Energy Systems.” Science What Do We Know, What Do We Need to Know?” In 360(6396): 1419. R. Fuentes-Nieva and P. Seck, eds., Risks, Shocks, and Dornbusch, R., and S. Edwards. 1991. “The Human Development on the Brink. New York: Palgrave Macroeconomics of Populism.” In R. Dornbusch and S. De Loecker, J., and J. Eeckhout. 2017. “The Rise of Macmillan. Edwards, eds., The Macroeconomics of Populism in Latin Market Power and the Macroeconomic Implications.” America. Chicago, IL: University of Chicago Press. NBER Working Paper 23687. National Bureau of Economic Deschamps, P. 2018. “Gender Quotas in Hiring Committees: Research, Cambridge, MA. A Boon or a Bane for Women?” LIEPP Working Paper 82. Doucouliagos, H., R. Freeman and P. Laroche 2017. The Sciences Po, Paris. Economics of Trade Unions: A Study of a Research Field Deaton, A. 2003. “Health, Inequality, and Economic and its Findings. New York: Routledge. Development.” Journal of Economic Literature 41(1): Devereux, S. 2009. “Why Does Famine Persist in Africa?” 113­158. Food Security 1:25. Dowd, A.J., I. Borisova, A. Amente and A. Yenew. 2016. “Realizing Capabilities in Ethiopia: Maximizing Early ------. 2005. “Measuring Poverty in a Growing World Devex. n.d. “Turning the Tide.” https://pages.devex.com/ Childhood Investment for Impact and Equity.” Journal of (or Measuring Growth in a Poor World).” Review of turningthetide.html. Accessed 13 August 2019. Human Development and Capabilities 17(4): 477­493. Economics and Statistics 87(1): 1­19. Dharmapala, D., C.F. Foley and K.J. Forbes. 2011. Dube, A., J. Jacobs, S. Naidu and S. Suri. 2018. ------. 2007. Global Patterns of Income and Health: Facts, “Watch What I Do, Not What I Say: The Unintended “Monopsony in Online Labor Markets.” NBER Working Interpretations, and Policies. WIDER Annual Lecture 10. Consequences of the Homeland Investment Act.” Journal Paper 24416. National Bureau of Economic Research, Helsinki: United Nations University World Institute for of Finance 66(3): 753­787. Cambridge, MA. Development Economics Research. Di Cesare, M., Y.-H. Khang, P. Asaria, T. Blakely, Duflo, E. 2012. “Women Empowerment and Economic ------. 2013a. The Great Escape: Health, Wealth, and the M.J. Cowan, F. Farzadfar, R. Guerrero, N. Ikeda, C. Development.” Journal of Economic Literature 50(4): Origins of Inequality. Princeton University Press. Kyobutungi and K.P. Msyamboza. 2013. “Inequalities 1051­1079. in Non-Communicable Diseases and Effective Responses.” ------. 2013b. “What Does the Empirical Evidence Tell Us The Lancet 381(9866): 585­597. Duncan, G.J., J. Brooks-Gunn and P.K. Klebanov. 1994. about the Injustice of Health Inequalities.” Inequalities in “Economic Deprivation and Early Childhood Development.” Health: Concepts Measures, and Ethics 263: 281. Diamond J. 1997. Guns, Germs and Steel. New York: W.W. Child Development 65(2): 296­318. Norton & Company. Dunlap, R.E., and A.M. McCright. 2011. “Organized ------. 2005. Collapse: How Societies Choose to Fail or Climate Change Denial.” In J.S. Dryzek, R.B. Norgaard Succeed. New York: Penguin Books.

274 | HUMAN DEVELOPMENT REPORT 2019 and D. Schlosberg, eds., The Oxford Handbook of Climate Eriksen, M., L.C. Lebreton, H.S. Carson, M. Thiel, C.J. ------. 2018. World Livestock: Transforming the Livestock Change and Society. Oxford, UK: Oxford University Press. Moore, J.C. Borerro, F. Galgani, P.G. Ryan and J. Sector through the Sustainable Development Goals. Rome. Reisser. 2014. “Plastic Pollution in the World’s Oceans: www.fao.org/3/CA1201EN/ca1201en.pdf. Accessed 15 Dworkin, R. 1981. “What is Equality? Part 2: Equality of More than 5 Trillion Plastic Pieces Weighing over 250,000 August 2019. Resources.” Philosophy and Public Affairs 10(3): 283­345. Tons Afloat at Sea.” PLOS ONE 9(12): e111913. Farber, H.S., D. Herbst, I. Kuziemko and S. Naidu. 2018. Early, R., B.A. Bradley, J.S. Dukes, J.J. Lawler, J.D. Erreygers, G. 2009. “Correcting the Concentration Index.” “Unions and Inequality over the Twentieth Century: New Olden, D.M. Blumenthal, P. Gonzalez, E.D. Grosholz, Journal of Health Economics 28(2): 504­515. Evidence from Survey Data.” NBER Working Paper 24587. I. Ibañez, L.P. Miller and C.J. Sorte. 2016. “Global National Bureau of Economic Research, Cambridge, MA. Threats from Invasive Alien Species in the Twenty-First Etzioni, A. 2000. “Social Norms: Internalization, Persuasion, Century and National Response Capacities.” Nature and History.” Law & Society Review 34(1): 157­178. Farber, H.S., and R.G. Valletta. 2015. “Do Extended Communications 7. Unemployment Benefits Lengthen Unemployment Spells? Eurobarometer. 2018. “Fairness, Inequality and Evidence from Recent Cycles in the US Labor Market.” ECLAC (Economic Commission for Latin America and Inter-Generational Mobility.” Special Eurobarometer Journal of Human Resources 50(4): 873­909. the Caribbean). 2018a. The Inefficiency of Inequality. 471. European Commission, Directorate-General for Thirty-Seventh Session of ECLAC, 7­11 May, Havana. Communication, Brussels. Farid, M., M. Keen, M.G. Papaioannou, I.W. Parry, Santiago. C.A. Pattillo and A. Ter-Martirosyan. 2016. “After European Commission. 2016. “State Aid: Ireland Gave Paris; Fiscal, Macroeconomic and Financial Implications ------. 2018b. Social Panorama of Latin America 2018. Illegal Tax Benefits to Apple Worth up to Euro 13 Billion.” of Global Climate Change.” IMF Staff Discussion Note Santiago. Press Release, 30 August. https://europa.eu/rapid/press- 16/01. International Monetary Fund, Washington, DC. release_IP-16-2923_en.htm. Accessed 7 November 2019. www.imf.org/en/Publications/Staff-Discussion-Notes/ The Economist. 2019. “In South-East Asia, Grab and Gojek Issues/2016/12/31/After-Paris-Fiscal-Macroeconomic- Bring Banking to the Masses.” 2 May. www.economist. ------. 2019. “Antitrust: Commission Fines Google Euro and-Financial-Implications-of-Global-Climate- com/special-report/2019/05/02/in-south-east-asia-grab- 1.49 Billion for Abusive Practices in Online Advertising.” Change-43484. Accessed 5 November 2019. and-gojek-bring-banking-to-the-masses. Accessed 24 Press Release, 20 March. https://europa.eu/rapid/press- October 2019. release_IP-19-1770_en.htm. Accessed 7 November 2019. Favara, G., and J. Imbs. 2015. “Credit Supply and the Price of Housing.” American Economic Review 105(3): 958­992. Eggertsson, G., J. Robbins and E. Wold. 2018. “Kaldor European Commission, Directorate-General for and Piketty’s Facts: The Rise of Monopoly Power in the Research and Innovation. 2014. “Trust at Risk: Fay, M. 2005. The Urban Poor in Latin America. Directions in United States.” NBER Working Paper 24287. National Implications for EU Policies and Institutions.” Report of the Development. Washington, DC: World Bank. Bureau of Economic Research, Cambridge, MA. Expert Group. Luxembourg. https://publications.europa.eu/ en/publication-detail/-/publication/e512c11b-e922-11e6- Fearon, J.D., and D.D. Laitin. 2003. “Ethnicity, Insurgency, Eicher, T.S., and S.J. Turnovsky. 2003. Inequality and ad7c-01aa75ed71a1. Accessed 1 August 2019. and Civil War.” American Political Science Review 97(1): Growth: Theory and Policy Implications. Cambridge, MA: 75­90. MIT Press. European Environment Agency. 2018. “Environmental Indicator Report 2018: Number of Countries That Have Fehr, E., U. Fischbacher and S. Gächter. 2002. “Strong Eika, L., M. Mogstad and B. Zafar. Forthcoming. Adopted a Climate Change Adaptation Strategy/Plan.” Reciprocity, Human Cooperation, and the Enforcement of “Educational Assortative Mating and Household Income Copenhagen. www.eea.europa.eu/airs/2018/environment- Social Norms.” Human Nature 13(1): 1­25. Inequality.” Journal of Political Economy. and-health/climate-change-adaptation-strategies. Accessed 13 August 2019. Felix, S., and P. Portugal. 2017. “Labor Market Elborgh-Woytek, K., M. Newiak, K. Kochhar, S. Imperfections and the Firm’s Wage Setting Policy.” Fabrizio, K. Kpodar, Ph. Wingender, B. Clemments European Parliament. 2016. “Draft Report with recom- Working Paper. Banco de Portugal, Economics and and G. Schwartz. 2013. “Women, Work, and the mendations to the Commission on Civil Law Rules on Research Department, Lisbon. Economy: Macroeconomic Gains from Gender Equity.” Robotics.” 2015/2103(INL). Brussels. www.europarl.euro- IMF Staff Discussion Note, International Monetary pa.eu/doceo/document/JURI-PR-582443_EN.pdf?redirect. Ferreira, F.H.G. 2012. “Distributions in Motion: Economic Fund, Washington, DC. www.imf.org/external/pubs/ft/ Accessed 24 October 2019. Growth, Inequality, and Poverty Dynamics.” In P.N. sdn/2013/sdn1310.pdf. Accessed14 October 2019. Jefferson, ed., The Oxford Handbook of the Economics of Fagereng, A., M. Mogstad and M. Ronning. 2019. “Why Poverty. New York: Oxford University Press. Elgar, F.J., B. McKinnon, T. Torsheim, C.W. Schnohr, Do Wealthy Parents Have Wealthy Children?” Working J. Mazur, F. Cavallo and C. Currie. 2016. “Patterns of Paper 2019-22. University of Chicago, Becker Friedman Ferreira, F.H.G., C. Lakner, M.A. Lugo and B. Özler. 2018. Socioeconomic Inequality in Adolescent Health Differ Institute for Economics, Chicago, IL. “Inequality of Opportunity and Economic Growth: How According to the Measure of Socioeconomic Position.” Much Can Cross-Country Regressions Really Tell Us?” Social Indicators Research 127(3): 1169­1180. Fairfield, T. 2015. “Structural Power in Comparative Political Review of Income and Wealth 64(4): 800­827. Economy: Perspectives from Policy Formulation in Latin Eliason, M.J., S. Dibble and P.A. Robertson. 2011. America.” Business and Politics 17(3): 411­441. Ferreira, F.H.G., and N. Lustig. 2015. “Special Issue on “Lesbian, Gay, Bisexual, and Transgender (LGBT) “Appraising Cross-National Income Inequality Databases.” Physicians’ Experiences in the Workplace.” Journal of Falch, T. 2010. “The Elasticity of Labor Supply at the Journal of Economic Inequality 13(4). Homosexuality 58(10): 1355­1371. Establishment Level.” Journal of Labor Economics 28(2): 237­266. Ferreira, F., N. Lustig and D. Teles. 2015. “Appraising Elliott, P., D. Briggs, S. Morris, C. de Hoogh, C. Hurt, Cross-National Income Inequality Databases: An T.K. Jensen, I. Maitland, S. Richardson, J. Wakefield FAO (Food and Agriculture Organization of the United Introduction.” Journal of Economic Inequality 13(4): and L. Jarup. 2001. “Risk of Adverse Birth Outcomes in Nations). 2006. Livestock’s Long Shadow: Environmental 497­526. Populations Living near Landfill Sites.” BMJ 323: 363. Issues and Options. www.fao.org/3/a0701e/a0701e00. htm. Accessed 29 October 2019 Ferreira, F., and V. Peragine. 2016 “Individual Enamorado, T., L.F. López-Calva, C. Rodríguez-Castelán Responsibility and Equality of Opportunity.” In M.D. and H. Winkler. 2016. “Income Inequality and Violent ------. 2011. The State of Food and Agriculture: Closing Adler and M. Fleurbaey, eds., The Oxford Handbook of Crime: Evidence from Mexico’s Drug War.” Journal of the Gender Gap for Development. Rome. Well-Being and Public Policy. New York: Oxford University Development Economics 120: 128­143. Press. ------. 2014. “Animal Production.” Rome. www.fao.org/ Engelman, M., V. Canudas-Romo and E.M. Agree. animal-production/en/. Accessed 15 August 2019. Festinger, L. 1954. “A Theory of Social Comparison 2010. “The Implications of Increased Survivorship for Processes.” Human Relations 7: 117­140. Mortality Variation in Aging Populations.” Population and ------. 2016. AQUASTAT database. Rome. www.fao.org/ Development Review 36(3): 511­539. nr/water/aquastat/water_use/index.stm. Accessed 10 Ffrench-Davis, R. 2000. Reforming the Reforms in Latin October 2019. America: Macroeconomics, Trade, Finance. London/New Ennis, S.F., P. Gonzaga and C. Pike. 2019. “Inequality: York: Macmillan/Palgrave. A Hidden Cost of Market Power.” Oxford Review of ------. 2017. “Livestock Solutions for Climate Change.” Economic Policy 35(3): 518­549 Rome. www.fao.org/3/a-i8098e.pdf. Accessed 15 August Fiala, O., and R. Watkins. 2019. “The Power of 2019. Convergence: Applying the Capabilities Approach to

References | 275 the 2030 Goals and `Leaving No One Behind.” Save the Frank, M.R., L. Sun, M. Cebrian, H. Youn and I. ------. 2018. “Sparse, Inconsistent and Unreliable: Tax Children manuscript prepared for Human Development Rahwan. 2018. “Small Cities Face Greater Impact from Records and the World Inequality Report.” Development Report Office. Automation.” Journal of the Royal Society Interface and Change 50(2): 329­346. 15(139). Filmer, D., and L. Pritchett. 1999. “The Effect of Household Galbraith, J.K., J. Choi, B. Halbach, A. Malinowska Wealth on Educational Attainment: Evidence from 35 Franzen, A., and D. Vogl. 2013. “Acquiescence and the and W. Zhang. 2015. “A Comparison of Major World Countries.” Population and Development Review 25(1): Willingness to Pay for Environmental Protection: A Inequality Data Sets: LIS, OECD, SILC, WDI and EHII.” UTIP 85­120. Comparison of the ISSP, WVS, and EVS.” Social Science Working Paper 69. The University of Texas at Austin. Quarterly 94(3): 637­659. Finkelstein, A., M. Gentzkow and H. L. Williams. ------. 2016. “A Comparison of Major World Inequality 2019. “Place-Based Drivers of Mortality: Evidence from Freedom House. 2019. Freedom in the World 2019: Data Sets: LIS, OECD, EU-SILC, WDI, and EHII.” In L. Migration.” National Bureau of Economic Research, Democracy in Retreat. Washington, DC. https://freedom- Cappellari, S.W. Polachek and K. Tatsiramos, eds., Cambridge, MA. house.org/report/freedom-world/freedom-world-2019/ Income Inequality Around the World. Research in Labor democracy-in-retreat. Economics, Vol. 44. Bingley, UK: Emerald Group Publishing Fintech News Hong Kong. 2019. “How China’s Virtual Limited. Banks Are Offering Loans to Micro-Businesses within Freeman, C. 1987. Technology Policy and Economic Minutes.” 22 August. https://fintechnews.hk/9925/virtual- Performance: Lessons from Japan. London: Pinter. Galor, O., and J. Zeira. 1993. “Income Distribution and banking/virtual-banking-china-ai-big-data-microbusiness/. Macroeconomics.” Review of Economic Studies 60(1): Accessed 6 October 2019. Freeman, C., and C. Perez. 1990. “The Diffusion of 35­52. Technological Innovations and Changes of Techno- FitzGerald, V., and J. A. Ocampo. 2019. “Towards economic Paradigm.” In F. Arcangeli, P. David and G. Gans, J., A. Leigh, M. Schmalz and A. Triggs. 2018. Fairer Global Taxation.” Background paper for Human Dos, eds., The Diffusion of New Technologies. New York: “Inequality and Market Concentration, When Shareholding Development Report 2019. United Nations Development Oxford University Press. Is More Skewed than Consumption.” Oxford Review of Programme, Human Development Report Office, New York. Economic Policy 35(3): 550­563. Frost, J., L. Gambacorta, Y. Huang, H.S. Shin and P. Flake, D.F. 2005. “Individual, Family, and Community Risk Zbinden. 2019. “BigTech and the Changing Structure of Galvan, E., and C. Garcia-Peñalosa. 2018. “Gender Norms Markers for Domestic Violence in Peru.” Violence Against Financial Intermediation.” BIS Working Paper 779. BIS, and Labour Supply: Identifying Heterogeneous Patterns Women 11(3): 353­73. Basel, Switzerland. https://papers.ssrn.com/sol3/papers. across Groups of Women.” Preliminary version of paper cfm?abstract_id=3369011. Accessed 6 October 2019. presented at the Eighth Meeting of the Society for the Fletcher, E., R. Pande and C. T. Moore. 2017. “Women Study of Economic Inequality, 3­5 July, Paris. and Work in India: Descriptive Evidence and a Review of Fuentes-Nieva, R., and P. Seck. 2010. “The Short- Potential Policies.” HKS Faculty Research Working Paper and Medium-Term Human Development Effects of Garbinti, G., J. Goupille-Lebret and T. Piketty. 2016. RWP18-004. Harvard University, John F. Kenendy School Climate-Related Shocks: Some Empirical Evidence.” “Accounting for Wealth Inequality Dynamics: Methods, of Government, Cambridge, MA. In Risks, Shocks and Human Development. New York: Estimates and Simulations for France (1800-2014).” WID. Palgrave-Macmillan. world Working Paper 2016/5. World Inequality Database. Forbes. 2019. “The World´s Billionaires.” 5 March. www. forbes.com/billionaires/#1360ce67251c. Accessed 9 Fujimori, S., T. Hasegawa, T. Masui, K. Takahashi, D.S. García, J.L., J.J. Heckman, D.E. Leaf and M.J. Prados. October 2019. Herran, H. Dai, Y. Hijioka and M. Kainuma. 2017. 2016. “The Life-cycle Benefits of an Influential Early “SSP3: AIM Implementation of Shared Socioeconomic Childhood Program.” NBER Working Paper 22993. Fortin, P., L. Godbout and S. St-Cerny. 2012. “Impact Pathways.” Global Environmental Change 42: 268­283. National Bureau of Economic Research, Cambridge, MA. of Quebec´s Universal Low-Fee Childcare Program on Female Labour Force Participation, Domestic Income, and Fukuda-Parr, S. 2019. “Keeping Out Extreme Inequality Gaspar, V., P. Mauro and T. Poghosyan. 2017. “Lessons Government Budgets.” University of Quebec at Montreal from the SDG Agenda ­ The Politics of Indicators.” Global from the Old Masters on Assessing Equity and Efficiency: and University of Sherbooke. www.oise.utoronto.ca/atkin- Policy 10(S1): 61­69. A Primer for Fiscal Policymakers.” IMF Working Paper No. son/UserFiles/File/News/Fortin-Godbout-St_Cerny_eng. 17/214. International Monetary Fund, Washington, DC. pdf. Accessed 9 October 2019. Furman, J. 2014. “Global Lessons for Inclusive Growth.” Presentation to the Institute of International and European Gasparini, L., and N. Lustig, 2011. “The Rise and Fall Foster, J., S. Seth, M. Lokshin and Z. Sajaia. 2013. A Affairs, 7 May, Dublin. of Income Inequality in Latin America.” Working Paper Unified Approach to Measuring Poverty and Inequality. 1110. Tulane University, Department of Economics, New Washington, DC: World Bank. ------. 2019. “Should Policymakers Care Whether Orleans, LA. Inequality Is Helpful or Harmful for Growth?” In Foster, J.E., and E. Ok. 1999. “Lorenz Dominance and the O. Blanchard and L.H. Summers, eds., Evolution or Gasparini, L., and L. Tornarolli. 2015. “A Review of the Variance of Logarithms.” Econometrica 67(4): 901­907. Revolution? Rethinking Macroeconomic Policy after the OECD Income Distribution Database.” Journal of Economic Great Recession. Cambridge, MA: MIT Press. Inequality 13: 579­602. Fourie, C., F. Schuppert and I. Wallimann-Helmer. 2015. Social Equality: On What It Means to Be Equals. Oxford, Furman, J., and P. Orszag. 2015. “A Firm-Level Perspective Gasperi, J., S.L. Wright, R. Dris, F. Collard, C. Mandin, UK: Oxford University Press. on the Role of Rents in the Rise in Inequality.” M. Guerrouache, V. Langlois, F.J. Kelly and B. Tassin. Presentation at “A Just Society” Centennial Event in 2018. “Microplastics in Air: Are We Breathing It In?” Fraga, M.F., E. Ballestar, M.F. Paz, S. Ropero, F. Setien, Honor of Joseph Stiglitz, 16 October, New York. Current Opinion in Environmental Science & Health 1: 1­5. M.L. Ballestar, D. Heine-Suñer, J.C. Cigudosa, M. Urioste, J. Benitez and M. Boix-Chornet. 2005. ------. 2018. “Slower Productivity and Higher Inequality: Gates, S., H. Hegre, H.M. Nygård and H. Strand. 2012. “Epigenetic Differences Arise during the Lifetime of Are They Related?” Working Paper 2018-4. Peterson “Development Consequences of Armed Conflict.” World Monozygotic Twins.” Proceedings of the National Institute for International Economics, Washington, DC. Development 40(9): 1713­1722. Academy of Sciences 102(30): 10604­10609. Furman, J., and R. Seamans. 2019. “AI and the Economy.” GDIM. 2018. Global Database on Intergenerational Mobility. Francese, M., and D. Prady. 2018. “Universal Basic Innovation Policy and the Economy 19: 161­191. World Bank, Development Research Group, Washington, Income: Debate and Impact Assessment.” IMF Working DC. Paper WP/18/273. International Monetary Fund, Galama, T.J., and H. Van Kippersluis. 2018. “A Theory of Washington, DC. Socio-Economic Disparities in Health over the Life Cycle.” Gebru, T., J. Krause, Y. Wang, D. Chen, J. Deng, E. The Economic Journal 129(617): 338­374. Lieberman Aiden and L. Fei-Fei. 2017. “Using Deep Frank, M.R., D. Autor, J.E. Bessen, E. Brynjolfsson, M. Learning and Google Street View to Estimate the Cebrian, D.J. Deming, M. Feldman, M. Groh, J. Lobo, Galbraith, J. 2012. Inequality and Instability: A Study of the Demographic Makeup of Neighborhoods across the United E. Moro and D. Wang. 2019. “Toward Understanding the World Economy Just before the Great Crisis. Oxford, UK: States.” Proceedings of the National Academy of Sciences Impact of Artificial Intelligence on Labor.” Proceedings of Oxford University Press. 114(50): 13108­13113. the National Academy of Sciences 116(14): 6531­6539. ------. 2016. Inequality: What Everyone Needs to Know. Gemici, K. 2007. “Karl Polanyi and the Antinomies of Oxford, UK: Oxford University Press. Embeddedness.” Socio-Economic Review 6: 5­33.

276 | HUMAN DEVELOPMENT REPORT 2019 Genicot, G., and D. Ray. 2017. “Aspirations and Inequality.” Giupponi, M.B.O., and M.C. Paz. 2015. “The 1, Article 12. http://pdba.georgetown.edu/Constitutions/ Econometrica 85(2): 489­519. Implementation of the Human Right to Water in Ecuador/english08.html. Accessed 15 August 2019. Argentina and Colombia.” Anuario Mexicano de Derecho Gennaioli, N., A. Shleifer and R. Vishny. 2012. Internacional 15(1): 323­352. Government of Japan. 2017. Realizing Society 5.0. Tokyo. “Neglected Risks, Financial Innovation, and Financial https://www.japan.go.jp/abenomics/_userdata/abenom- Fragility.” Journal of Financial Economics 104(3): 452­468. Glaeser, E.L., S.D. Kominers, M. Luca and N. Naik. 2018. ics/pdf/society_5.0.pdf. Accessed 6 October 2019. “Big Data and Big Cities: The Promises and Limitations of George, S. 2016. “What Thailand Can Teach the World Improved Measures of Urban Life.” Economic Inquiry 56: Graham, C. 2012. Happiness around the World: The Paradox about Universal Healthcare.” The Guardian, 24 May. 114­137. of Happy Peasants and Miserable Millionaires. Oxford, www.theguardian.com/health-revolution/2016/may/24/ UK: Oxford University Press. thailand-universal-healthcare-ucs-patients-government- Gleeson, T., S.C. Zipper, L.W. Erlandsson, M. Porkka, political. Accessed 9 October 2019. M., F. Jaramillo, D. Gerten, I. Fetzer, S. Cornell, Grainger, C., and C. Kolstad. 2010. “Who Pays a Price on L. Piemontese, L. Gordon and J. Rockström. Carbon?” Environmental & Resource Economics 46(3): Gerbens-Leenes, P.W., and S. Nonhebel. 2002. Forthcoming. “The Water Planetary Boundary: A 359­376. “Consumption Patterns and their Effects on Land Required Roadmap to Illuminate Water Cycle Modifications in the for Food.” Ecological Economics 42(1­2): 185­199. Anthropocene.” Water Resources Research. https://ear- Greef, K.D. 2019. “Botswana’s High Court Decriminalizes tharxiv.org/vfg6n/. Accessed 15 August 2019. Gay Sex.” New York Times, June 11. www.nytimes. Gerber, P.J., H. Steinfeld, B. Henderson, A. Mottet, C. com/2019/06/11/world/africa/botswana-ruling-criminal- Opio, J. Dijkman, A. Falcucci and G. Tempio. 2013. Gleick, P.H. 2018. “Transitions to Freshwater Sustainability.” ize-gay-sex.html. Accessed 18 October 2019. “Tackling Climate Change through Livestock—A Global Proceedings of the National Academy of Sciences 115(36): Assessment of Emissions and Mitigation Opportunities.” 8863­8871. Green, D. 2016. How Change Happens. Oxford, UK: Oxford Food and Agriculture Organization of the United Nations, University Press. Rome. Global Commission on Adaptation. 2019. Adapt Now: A Global Call for Leadership on Climate Resilience. Green, D.P., J. Glaser and A. Rich. 1998. “From Lynching Gerring, J., S.C. Thacker and R. Alfaro. 2012. “Democracy Rotterdam, The Netherlands: Global Center on Adaptation, to Gay Bashing: The Elusive Connection between and Human Development.” The Journal of Politics 74(1): and Washington, DC: World Resources Institute. Economic Conditions and Hate Crime.” Journal of 1­17. Personality and Social Psychology 75: 82­92. Global Panel on Agriculture and Food Systems for Gert, G., and H. Kharas. 2018. “Leave No Country Behind: Nutrition. 2016. “The Cost of Malnutrition: Why Policy Greenwood, J., N. Guner, G. Kocharkov and C. Santos. Ending Poverty in the Toughest Places.” Brookings Action is Urgent.” Technical Brief 3. London, UK. www.glo- 2014. “Marry Your Like: Assortative Mating and Income Institution blog, 6 February. www.brookings.edu/research/ pan.org/sites/default/files/pictures/CostOfMalnutrition. Inequality.” American Economic Review 104(5): 348­353. leave-no-country-behind/. Accessed 10 October 2019. pdf. Accessed 9 August 2019. ------. 2015. “Corrigendum to Marry Your Like: Assortative Gerten, D., J. Rockström, J. Heinke, W. Steffen, K. Godfray, H.C.J., J.R. Beddington, I.R. Crute, L. Haddad, Mating and Income Inequality.” www.jeremygreenwood. Richardson and S. Cornell. 2015. “Response to D. Lawrence, J.F. Muir, J. Pretty, S. Robinson, S.M. net/papers/ggksPandPcorrigendum.pdf. Accessed 10 Comment on “Planetary boundaries: Guiding Human Thomas and C. Toulmin. 2010. “Food Security: The October 2019.. Development on a Changing Planet.” Science 348(6240): Challenge of Feeding 9 Billion People.” Science 327(5967): 1217­1217. 812­818. Greenwood, J., and B. Jovanovic. 1990. “Financial Development, Growth, and the Distribution of Income.” Ghosh, J. 2016. “The Role of Labour Market and Sectoral Goldin, C.D., and L.F. Katz. 2009. The Race between Journal of Political Economy 98(5): 1076­1107. Policies in Promoting More and Better Jobs in Low Middle Education and Technology. Cambridge, MA: Harvard Income Countries: Issues, Evidence and Policy Options: University Press. Greubel, L., and J. van der Gaag. 2012. “Early Childhood The Case of India.” Employment Working Paper 206. Development: A Chinese National Priority and Global International Labour Organization, Geneva Goldin, I., and C. Kutarna. 2016. Age of Discovery: Concern for 2015.” The Brookings Institution, Washington, Navigating the Risks and Rewards of our New DC. www.brookings.edu/blog/up-front/2012/06/29/early- ------. 2019. “Asian Approaches to Tackle Inequalities.” Renaissance. Oxford, UK: Bloomsbury Publishing. childhood-development-a-chinese-national-priority-and- Background paper for Human Development Report global-concern-for-2015/. Accessed 8 November 2019. 2019. United Nations Development Programme, Human Gómez, L. 2014. “Micromachismos, un Machismo Silencioso Development Report Office, New York. y Sutil.” Tinta Libre, December. Grigoli, F., and A. Robles. 2017. “Inequality Overhang.” IMF Working Paper WP/17/76. International Monetary Gilens, M., and B.I. Page. 2014. “Testing Theories of Gonzaga, C.M., R. Freitas-Junior, M. R. Souza, M.P. Fund, Washington, DC. American Politics: Elites, Interest Groups, and Average Curado and N.M. Freitas. 2014. “Disparities in Female Citizens.” Perspectives on Politics 12: 564­581. Breast Cancer Mortality Rates between Urban Centers and Grunewald, N., S. Klasen, I. Martínez-Zarzoso and C. Rural Areas of Brazil: Ecological Time-Series Study.” The Muris. 2017. “The Trade-off between Income Inequality Gill, F.L., K. Viswanathan and M.Z. Abdul Karim. 2018. Breast 23(2): 180­187. and Carbon Dioxide Emissions.” Ecological Economics “The Critical Review of the Pollution Haven Hypothesis.” 142(C): 249­256. International Journal of Energy Economics and Policy 8(1): Gonzales, A. 2016. “The Contemporary US Digital Divide: 167­174. From Initial Access to Technology Maintenance.” GSMA. 2017. “Number of Mobile Subscribers Worldwide Information, Communication & Society 19(2): 234­248. Hits 5 Billion.” www.gsma.com/newsroom/press-release/ Gilligan, J. 1996. Violence: Our Deadly Epidemic and Its number-mobile-subscribers-worldwide-hits-5-billion/. Causes. New York: GP Putnam. Goodfellow, I., Y. Bengio and A. Courville. 2016. Deep Accessed 6 October 2019. Learning. Cambridge, MA: MIT Press. Gintis, H. 2007. “A Framework for the Unification of the ------. 2018. State of Mobile Internet Connectivity 2018. Behavioral Sciences.” Behavioral and Brain Sciences Goos, M., A. Manning and A. Salomons. 2014. London. www.gsma.com/mobilefordevelopment/resourc- 30(1): 1­16. “Explaining Job Polarization: Routine-Biased Technological es/state-of-mobile-internet-connectivity-2018/. Accessed Change and Offshoring.” American Economic Review 4 November 19. Giovannoni, O.G. 2014. “What Do We Know about the 104(8): 2509­2526. Labor Share and the Profit Share? Part III: Measures and GSMA Connected Women. 2015. Bridging the Gender Gap: Structural Factors.” Working Paper 805. Levy Economics Government of Bolivia. 2012. Ley contra el acoso y Mobile Access and Usage in Low- and Middle-Income Institute, Annandale-On-Hudson, NY. https://pdfs.seman- violencia política hacia las mujeres. Ley número 243. Countries. London. ticscholar.org/b3bc/2e58434772393b96caaea7ff7b96c00c Gaceta Oficial del Estado Plurinacional de Bolivia. 28 de a40a.pdf. Accessed 24 October 2019. mayo de 2012. Guaqueta, J. 2017. “Bootcamps: Raising Expectations for Girls in Math, Science and Technology.” World Bank Blog, Giraldo-Luque, S., N. Fernández-García and J.C. Pérez- Government of Ecuador, National Assembly, 24 February. https://blogs.worldbank.org/education/ Arce. 2018. “La centralidad temática de la movilización Legislative and Oversight Committee. 2008. bootcamps-raising-expectations-girls-math-science-and- #Niunamenos en Twitter.” El profesional de la información Constitution of the Republic of Ecuador, Chapter 2, Section technology. Accessed 9 October 2019. 27(1).

References | 277 Guerreiro, J., S. Rebelo and P. Teles. 2018. “Should C. Valli, M. Rabassa, N. Rehman, M.K. Parvizian, M. ------. 2011b. “The Economics of Inequality: The Value of Robots Be Taxed?” NBER Working Paper 23806. National Zworth, J.J. Bartoszko, LC. Lopes, D. Sit, M.M. Bala, Early Childhood Education.” American Educator 35(1): 31. Bureau of Economic Research, Cambridge, MA. P. Alonso-Coello and B.C. Johnston. Forthcoming. “Reduction of Red and Processed Meat Intake and Cancer ------. 2017. “There’s More to Gain by Taking Gutiérrez, G., and T. Philippon. 2019. “The Failure of Free Mortality and Incidence: A Systematic Review and Meta- a Comprehensive Approach to Early Childhood Entry.” NBER Working Paper 26001. National Bureau of analysis of Cohort Studies.” Annals of Internal Medicine. Development.” The Heckman Equation. https://heck- Economic Research, Cambridge, MA. manequation.org/www/assets/2017/01/F_Heckman_ Hanna, R., A. Khan and B. Olken. 2018. “Targeting the CBAOnePager_120516.pdf. Accessed 20 August 2019. Gutiérrez, C., and R. Tanaka. 2009. “Inequality and Poor.” Finance and Development 55(4): 28­31. Education Decisions in Developing Countries.” Journal of Heckman, J.J., and P. Carneiro. 2003. “Human Capital Economic Inequality 7(1): 55­81. Hanushek, E., P. Peterson, L. M. Talpey and L. Policy.” Working Paper 9495. National Bureau of Economic Woessmann. 2019. “The Unwavering SES Achievement Research, Cambridge, MA. www.nber.org/papers/w9495. Ha, A. 2018. “New York’s Taxi and Limousine Commission Gap: Trends in US Student Performance.” NBER Working pdf. Accessed 30 July 2019. Approves Minimum Wage Rules for App-Based Drivers.” Paper 25648. National Bureau of Economic Research, Tech Crunch, 4 December. Cambridge, MA. Heckman, J. J., and G. Karapakula. 2019a. “Intergenerational and Intragenerational Externalities of Haegel, N.M., H. Atwater Jr., T. Barnes, C. Breyer, A. Harari, Y.N. 2016. Homo Deus: A Brief History of Tomorrow. the Perry Preschool Project.” NBER Working Paper 25889. Burrell, Y-M. Chiang, S. De Wolf, B. Dimmler, D. London: Random House. National Bureau of Economic Research, Cambridge, MA. Feldman, S. Glunz, J.C. Goldschmidt, D. Hochschild, R. Inzunza, I. Kaizuka, B. Kroposki, S. Kurtz, S. Leu, Harcourt, B.E. 2011. The Illusion of Free Markets: Heckman, J. J., and G. Karapakula. 2019b. “The Perry R. Margolis, K. Matsubara, A. Metz, W.K. Metzger, Punishment and the Myth of Natural Order. Cambridge, Preschoolers at Late Midlife: A Study in Design-Specific M. Morjaria, S. Niki, S. Nowak, I.M. Peters, S. MA: Harvard University Press. Inference.” NBER Working Paper 25888. National Bureau Philipps, T. Reindl, A. Richter, D. Rose, K. Sakurai, of Economic Research, Cambridge, MA. R. Schlatmann, M. Shikano, W. Sinke, R. Sinton, Harper, G.W., and M. Schneider. 2003. “Oppression B.J. Stanbery, M. Topic, W. Tumas, Y. Ueda, J. van and Discrimination among Lesbian, Gay, Bisexual, and Heckman, J.J., and A.B. Krueger. 2005. Inequality de Lagemaat, P. Verlinden, M. Vetter, E. Warren, Transgendered People and Communities: A Challenge for in America: What Role for Human Capital Policies? M. Werner, M. Yamaguchi and A.W. Bett. 2019. Community Psychology.” American Journal of Community Cambridge, MA: MIT Press. “Terawatt-Scale Photovoltaics: Transform Global Energy.” Psychology 31(3­4): 243­252. Science 364(6443): 836­838. Heckman, J. J., J. Stixrud and S. Urzua. 2006. “The Harper, K., T. Steger and R. Filcák. 2009. “Environmental Effects of Cognitive and Noncognitive Abilities on Labor Hagen-Zanker, J., L. Pellerano, F. Bastagli, L. Harman, Justice and Roma Communities in Central and Eastern Market Outcomes and Social Behavior.” Journal of Labor V. Barca, G. Sturge, T., Schmidt and C. Laing. 2017. Europe.” Environmental Policy and Governance 19(4): Economics 24(3): 411­482. “The Impact of Cash Transfers on Women and Girls.” 251­268. Briefing. Overseas Development Institute, London. Heer, J. 2019. “Agency Plus Automation: Designing Artificial Hart, C. 2014. “The Role of Environmental Justice in Intelligence into Interactive Systems.” Proceedings of the Hakak, L., and S. Firpo. 2017. “Household Income Biodiversity Conservation: Investigating Experiences of National Academy of Sciences 116(6): 1844­1850. Inequality and Education in Marriage Market in Brazil: Communities near Kruger National Park, South Africa.” An Empirical Study.” University of São Paulo, Faculty of Dalhousie Journal of Interdisciplinary Management 10(1): Hegewisch, A., and J. Gornick. 2011. “The Impact of Economics, Brazil. www.fea.usp.br/sites/default/files/ 1­16. Work-Family Policies on Women’s Employment: A Review anexo-evento/chapter2_article_new_version_3_15.pdf. of Research from OECD Countries.” Community, Work and Accessed 9 September 2019. Hartlaub, V., and T. Schneider. 2012. “Educational Family 14(2): 119­138. Choice and Risk Aversion: How Important Is Structural Hall, J. 2013. “From Capabilities to Contentment: Testing vs. Individual Risk Aversion?” SOEPpapers on Heilman, B., C.M. Guerrero-López, C. Ragonese, M. the Links Between Human Development and Life Multidisciplinary Panel Data Research 433. German Kelberg and G. Barker. 2019. The Cost of the Man Box: Satisfaction.” In J. Helliwell, R. Layard and J. Sachs, eds., Institute for Economic Research (DIW Berlin), Berlin. www. A Study on the Economic Impacts of Harmful Masculine World Happiness Report 2013. New York: UN Sustainable diw.de/documents/publikationen/73/diw_01.c.394455.de/ Stereotypes in the United States. Washington, DC, and Development Solutions Network. diw_sp0433.pdf. Accessed 31 July 2019. London: Promundo-US and Unilever.

Hallegatte, S., and J. Rozenberg. 2017. “Climate Change Hauser, O.P., and M.I. Norton. 2017. “(Mis) Perceptions of Helliwell, J. 2019. “Inequality in Subjective Well-Being.” through a Poverty Lens.” Nature Climate Change 7(4): Inequality.” Current Opinion in Psychology 18: 21­25. Background paper for Human Development Report 250­256. 2019. United Nations Development Programme, Human Hayter, S. 2015. “Unions and Collective Bargaining.” In J. Development Report Office, New York. Hallegatte, S., A. Vogt-Schilb, M. Bangalore and J. Berg, ed., Labour Markets, Institutions and Inequality: Rozenberg. 2017. Unbreakable: Building the Resilience Building Just Societies in the 21st Century. Cheltenham, Herrera, A. 2019. “What We Can Learn from Canada’s of the Poor in the Face of Natural Disasters. Climate UK: Edward Elgar Publishing. Universal Child Care Model.” The World [Radio program], Change and Development Series. Washington, DC: 5 February. www.pri.org/stories/2019-02-05/what-we- World Bank. http://documents.worldbank.org/curated/ He, D., R. Leckow, V. Haksar, T. Mancini-Griffoli, N. can-learn-canada-s-universal-child-care-model. Accessed en/512241480487839624/pdf/110618-PUB-Box396333B- Jenkinson, M. Kashima, T. Khiaonarong, C. Rochon 14 October 2019. PUBLIC-PUBDATE-11-24-16-UNIT-ITSKI.pdf. Accessed 16 and H. Tourpe. 2017. “Fintech and Financial Services: August 2019. Initial Considerations.” IMF Staff Discussion Note Hickel, J. 2017a. The Divide: A Brief Guide to Global SDN/17/05. International Monetary Fund, Washington, Inequality and Its Solutions. New York: Random House. Hamann, M., K. Berry, T. Chaigneau, T. Curry, R. DC. Heilmayr, P.J.G. Henriksson, J. Hentati-Sundberg, A. ------. 2017b. “The Development Delusion: Foreign Aid Jina, E. Lindkvist, Y. Lopez-Maldonado, E. Nieminen, Heal, G. 2019. “The Cost of a Carbon-free Electricity System and Inequality.” American Affairs 1(3): 160­173. M. Piaggio, J. Qiu, J.C. Rocha, C. Schill, A. Shepon, in the U.S.” NBER Working Paper 26084. National Bureau A.R. Tilman, I. van den Bijgaart and T. Wuet. 2018. of Economic Research, Cambridge, MA. www.nber.org/ ------. 2019. “The Contradiction of the Sustainable “Inequality and the Biosphere.” Annual Review of papers/w26084. Accessed 4 November 2019. Development Goals: Growth versus Ecology on a Finite Environment and Resources 43: 61­83. Planet.” Sustainable Development (2019): 1­12. https:// Heckman, J.J. 2010. “Cognitive Skills Are Not Enough.” onlinelibrary.wiley.com/doi/abs/10.1002/sd.1947. Hamilton, J. 2016. Democracy’s Detectives: The Economics The Heckman Equation. https://heckmanequation.org/ Accessed 14 October 2019. of Investigative Journalism. Cambridge, MA: Harvard resource/cognitive-skills-are-not-enough/. Accessed 30 University Press. July 2019. Hilbert, M. 2011. “The End Justifies the Definition: The Manifold Outlooks on the Digital Divide and Their Practical Han, M.A., D. Zeraatkar, G.H. Guyatt, R.W.M. Vernooij, ------. 2011a. “The American Family in Black & White: Usefulness for Policy-Making.” Telecommunications Policy R. El Dib, Y. Zhang, A. Algarni, G. Leung, D. Storman, A Post-Racial Strategy for Improving Skills to Promote 35(8): 715­736. Equality.” Journal of Daedalus 140(2): 70­89. ------. 2019. “Making New Technologies Work for Equality.” Background paper for Human Development

278 | HUMAN DEVELOPMENT REPORT 2019 Report 2019. United Nations Development Programme, Really Polarizing?” NBER Working Paper 26064. National Contribution of Working Groups I, II and III to the Fifth Human Development Report Office, New York. Bureau of Economic Research, Cambridge, MA. Assessment Report of the Intergovernmental Panel on Climate Change. Geneva. Hillesund, S. 2019. “Choosing Whom to Target: Horizontal IDMC (Internal Displacement Monitoring Centre). 2018. Inequality and the Risk of Civil and Communal Violence.” Global Report on Internal Displacement 2018. Geneva. ------. 2018. “Summary for Policymakers.” Global Journal of Conflict Resolution 63(2): 528­554. Warming of 1.5°C. An IPCC Special Report on the Impacts IDRC (International Development Research Centre). of Global Warming of 1. °C above Preindustrial Levels. Hoegh-Guldberg, O., D. Jacob, M. Taylor, M. Bindi, S. 2018. Artificial Intelligence and Human Development: World Meteorological Organization, Geneva. Brown, I. Camilloni, A. Diedhiou, R. Djalante, K. Ebi, Toward a Research Agenda. Ottawa. F. Engelbrecht, J. Guiot and others. 2018. “Impacts of IPU (Inter-Parliamentary Union). 2019. Women in national 1.5 °C Global Warming on Natural and Human Systems.” Igan, D., and P. Mishra. 2011. “Three’s Company: Wall parliaments, as of 1 February 2019. http://archive.ipu.org/ In V. Masson-Delmotte, P. Zhai, H.-O. Pörtner, D. Roberts, Street, Capitol Hill, and K Street: Political Influence and wmn-e/classif.htm. Accessed 9 October 2019. J. Skea, P.R. Shukla, A. Pirani, W. Moufouma-Okia, C. Financial Regulation.” Journal of Law & Economics 57(4): Péan, R. Pidcock, S. Connors, J.B.R. Matthews, Y. Chen, 1063­1084. Islam, S.N., and J. Winkel. 2017. “Climate Change and X. Zhou, M.I. Gomis, E. Lonnoy, T. Maycock, M. Tignor Social Inequality.” DESA Working Paper 152. United and T. Waterfield, eds., Global Warming of 1.5°C. An IIPS (International Institute for Population Sciences) Nations Department of Economic and Social Affairs, New IPCC Special Report on the Impacts of Global Warming and ICF International. 2017. “India National Family York. of 1.5°C Above Pre-Industrial Levels and Related Global Health Survey NFHS-4 2015-16.” Mumbai, India. Greenhouse Gas Emission Pathways, in the Context ITU (International Telecommunications Union). 2019. of Strengthening the Global Response to the Threat of IIPS (International Institute for Population Sciences) Statistics. www.itu.int/en/ITU-D/Statistics/Pages/stat/ Climate Change, Sustainable Development, and Efforts to and Macro International. 2007. “India National Family default.aspx. Accessed 8 October 2019. Eradicate Poverty. Cambridge, UK: Cambridge University Health Survey NFHS-3 2005-06.” Mumbai, India. Press. www.ipcc.ch/site/assets/uploads/sites/2/2019/05/ Itzhak, B.-D., S. Kleimeier and M. Viehs. 2018. “Exporting SR15_Chapter3_Low_Res.pdf. Accessed 9 August 2019. ILGA (International Lesbian, Gay, Bisexual, Trans Pollution.” NBER Working Paper 25063. National Bureau and Intersex Association). 2019. State-Sponsored of Economic Research, Cambridge, MA. Hoekstra, A.Y., and M.M. Mekonnen. 2012. “The Water Homophobia 2019. Geneva. Footprint of Humanity.” Proceedings of the National Iversen, T., and D. Soskice. 2019. Democracy and Academy of Sciences 109(9): 3232­3237. ILO (International Labour Organization). 2016a. Prosperity: Reinventing Capitalism through a Turbulent Minimum Wage Policy Guide. Geneva. www.ilo.org/ Century. Princeton: Princeton University Press. Hojman, D.A., and Á. Miranda. 2018. “Agency, Human wcmsp5/groups/public/---ed_protect/---protrav/—travail/ Dignity, and Subjective Well-Being.” World Development documents/publication/wcms_508566.pdf. Accessed 14 Iversen, V., A. Krishna and K. Sen. 2019. “Beyond Poverty 101: 1­15. October 2019. Escapes—Social Mobility in Developing Countries: A Review Article.” World Bank Research Observer 34(2): Hornbeck, R. 2012. “The Enduring Impact of the American ILO (International Labour Organization). 2016b. 239­273. Dust Bowl: Short and Long-Run Adjustments to Global Wage Report 2016/17: Wage Inequality in the Environmental Catastrophe.” American Economic Review Workplace, Geneva. www.ilo.org/wcmsp5/groups/public/- IWDA (International Women’s Development Agency). 102(4): 1477­1507. —dgreports/---dcomm/---publ/documents/publication/ 2018. “What Does Intersectional Feminism Actually wcms_537846.pdf. Accessed 14 October 2019. Mean?” 11 May. https://iwda.org.au/what-does-intersec- Horowitz, D.L. 2001. Ethnic Groups in Conflict, 2nd ed. tional-feminism-actually-mean/. Accessed 9 October 2019. London, UK: University of California Press. ILO (International Labour Organization). 2017a. World Employment Social Outlook: Trends for Women 2017. IWPR (Institute for Women’s Policy Research). Hryshko, D., C. Juhn and K. McCue. 2017. “Trends in Geneva: International Labour Office. 2019. “Women, Automation, and the Future of Work.” Earnings Inequality and Earnings Instability among U.S. Washington, DC. http://iwpr.org/wp-content/up- Couples: How Important Is Assortative Matching?” Labour ------. 2017b. World Social Protection Report: Universal loads/2019/03/C476_Automation-and-Future-of-Work.pdf. Economics 48: 168-182. Social Protection to Achieve the Sustainable Development Accessed 16 August 2019. Goals 2017-2019. Geneva: International Labour Office. Hsiang, S., R. Kopp, A. Jina, J. Rising, M. Delgado, S. Jackson, R.B., C. Le Quéré, R.M. Andrew, J.G. Canadell, Mohan, D.J. Rasmussen, R. Muir-Wood, P. Wilson, ------. 2018a. Digital Labour Platforms and the Future of J.I. Korsbakken, Z. Liu, G.P. Peters and B. Zheng. 2018. M. Oppenheimer and K. Larsen. 2017. “Estimating Work: Towards Decent Work in the Online World. Geneva: “Global Energy Growth is Outpacing Decarbonization.” Economic Damage from Climate Change in the United International Labour Office. Environmental Research Letters 13: 120401. States. Science 356(6345): 1362­1369. ------. 2018b. Global Wage Report: What Lies Behind Jacobs, R.C., and D.T. Campbell. 1961. “The Perpetuation Hsu, Y.-C., and Tapia, H. 2019. “Older People Facing New Gender Pay Gaps. Geneva: International Labour Office. of an Arbitrary Tradition through Several Generations of a Inequalities: Life Expectancy in Chile.” Background paper Laboratory Microculture.” Journal of Abnormal and Social for Human Development Report 2019. United Nations ------. 2019a. “The Global Labour Income Share and Psychology 62(3): 649­658. Development Programme, Human Development Report Distribution.” July. www.ilo.org/global/statistics-and- Office, New York. databases/publications/WCMS_712232/lang—en/index. Jagers, R.J., D. Rivas-Drake and T. Borowski. 2018. htm. Accessed 10 October 2019. “Equity & Social and Emotional Learning: A Cultural Human Rights Watch. 2017. “`I Want to Be Like Nature Analysis.” Framework Briefs, Special Issues Series. Made Me’: Medically Unnecessary Surgeries on ------. 2019b. Labour statistics. https://ilostat.ilo.org/ The Collaborative for Academic, Social, and Emotional Intersex Children in the US.” 25 July. www.hrw.org/ data/. Accessed 9 October 2019. Learning, Chicago, IL. https://measuringsel.casel.org/ report/2017/07/25/i-want-be-nature-made-me/medically- wp-content/uploads/2018/11/Frameworks-Equity.pdf. unnecessary-surgeries-intersex-children-us. Accessed 26 ------. 2019c. Work for a Brighter Future: Global Accessed 30 July 2019. July 2019. Commission on the Future of Work. Geneva: International Labour Office. Jakob, M., O. Edenhofer, U. Kornek, D. Lenzi and J. Humphries, D.L., J.R. Behrman, B.T. Crookston, Minx. 2019. “Governing the Commons to Promote Global K.A. Dearden, W. Schott and M.E. Penny. 2014. IMF (International Monetary Fund). 2017a. “Tackling Justice: Climate Change Mitigation and Rent Taxation.” In “Households Across All Income Quintiles, Especially the Inequality.” Fiscal Monitor, October. Washington, DC. R. Kanbur and H. Shue, eds., Climate Justice: Integrating Poorest, Increased Animal Source Food Expenditures Economics and Philosophy. Oxford, UK: Oxford University Substantially During Recent Peruvian Economic Growth.” ------. 2017b. World Economic Outlook, April. Washington, Press. PLOS ONE 9(11). DC. Jaramillo, F., and G. Destouni. 2015. “Comment on Hunt, J., and R. Nunn. 2019. “Is Employment Polarization ------. 2018. “Pursuing Women’s Economic “Planetary Boundaries: Guiding Human Development on a Informative About Wage Inequality and Is Employment Empowerment.” Policy Paper. Washington, DC. www.imf. Changing Planet.” Science 348(6240): 1217. org/en/Publications/Policy-Papers/Issues/2018/05/31/ pp053118pursuing-womens-economic-empowerment. Accessed 14 October 2019.

IPCC (Intergovernmental Panel on Climate Change). 2014. Climate Change 2014: Synthesis Report.

References | 279 Jauch, S., and S. Watzka. 2016. “Financial Development Kahn, M.E., K. Mohaddes, R.N.C. Ng, M.H. Pesaran, M. Khanam, R. 2008. “Child Labour and School Attendance: and Income Inequality: A Panel Data Approach.” Empirical Raissi and J.C. Yang. 2019. “Long-Term Macroeconomic Evidence from Bangladesh.” International Journal of Economics 51(1): 291­314. Effects of Climate Change: A Cross-Country Analysis.” Social Economics 35(1/2): 77­98. Globalization Institute Working Paper 365. Federal Jaumotte, M.F., and M.C. Osorio. 2015. “Inequality and Reserve Bank of Dallas, Dallas, TX. Kiatpongsan S., and M. Norton. 2014. “How Much (More) Labor Market Institutions.” IMF Staff Discussion Note Should CEOs Make? A Universal Desire for More Equal SDN/15/14. International Monetary Fund, Washington, Kaldor, N. 1961. “Capital Accumulation and Economic Pay.” Perspectives on Psychological Science 9: 587­593. DC. Growth.” In F.A. Lutz and D.C. Hague, eds., The Theory of Capital. New York: St. Martin’s Press. Kidd, S., and D. Athias. 2019. Hit and Miss: An Assessment Jaumotte, F., S. Lall and C. Papageorgiou. 2013. “Rising of Targeting Effectiveness in Social Protection. Working Income Inequality: Technology, or Trade and Financial Kanbur, R. 2017. “Structural Transformation and Income Paper. Development Pathways, Orpington, UK. Globalization.” IMF Economic Review 61: 271­309. Distribution: Kuznets and Beyond.” IZA Discussion Paper 10636. Institute for the Study of Labour, Bonn, Germany. Kim, D., and A. Saada. 2013. “The Social Determinants Jaumotte, M.F., and M.C. Osorio. 2015. “Inequality and www.econstor.eu/bitstream/10419/161259/1/dp10636. of Infant Mortality and Birth Outcomes in Western Labor Market Institutions.” IMF Staff Discussion Note pdf. Accessed 1 August 2019. Developed Nations: A Cross-Country Systematic Review.” SDN/15/14. International Monetary Fund, Washington, International Journal of Environmental Research and DC. ------. 2018. “On Three Canonical Responses to Labour Public Health 10(6): 2296­2335. Saving Technical Change.” VoxEU, 8 January. https:// Jensen, P., and H.S. Nielsen. 1997. “Child Labour or voxeu.org/article/three-canonical-responses-labour- Kim, J., A. Lee and M. Rossin-Slater. 2019. “What to School Attendance? Evidence from Zambia.” Journal of saving-technical-change. Accessed 8 August 2019. Expect When It Gets Hotter: The Impacts of Prenatal Population Economics 10(4): 407­424. Exposure to Extreme Heat on Maternal and Infant Health.” Karabarbounis, L. 2011. “One Dollar, One Vote.” The NBER Working Paper 26384. National Bureau of Economic Jetten, J., C. Haslam, S.A. Haslam and N.R. Economic Journal 121(553): 621­651. Research, Cambridge, MA. www.nber.org/papers/w26384. Branscombe. 2009. “The Social Cure.” Scientific Accessed 28 October 2019. American Mind 20(5): 26­33. Karabarbounis, L., and B. Neiman. 2013. “The Global Decline of the Labor Share.” Quarterly Journal of Kim, N. 2010. “Impact of Extreme Climate Events on Johnson, T., A. Lora-Wainwright and J. Lu. 2018. Economics 129(1): 61­103. Educational Attainment: Evidence from Cross-Section Data “The Quest for Environmental Justice in China: Citizen and Welfare Projection.” In R. Fuentes-Nieva and P.A. Participation and the Rural­Urban Network against Kautz, T., J.J. Heckman, R. Diris, B. Ter Weel and Seck, eds. Risk, Shocks, and Human Development. London: Panguanying’s Waste Incinerator.” Sustainability Science L. Borghans. 2014. “Fostering and Measuring Skills: Palgrave Macmillan. 13(3): 733­746. Improving Cognitive and Non-Cognitive Skills to Promote Lifetime Success.” NBER Working Paper 20749. National King, A.D., M.G. Donat, E.M. Fischer, E. Hawkins, Johnston, B.C., D. Zeraatkar, M.A. Han, R.W.M. Bureau of Economic Research, Cambridge, MA. L.V. Alexander, D.J. Karoly, A.J. Dittus, S.C. Lewis Vernooij, C. Valli, R. El Dib, C. Marshall, P.J. Stover, and S.E. Perkins. 2015. “The Timing of Anthropogenic S. Fairweather-Taitt, G. Wójcik, F. Bhatia, R. de Kawachi, I., B.P. Kennedy and R.G. Wilkinson. 1999. Emergence in Simulated Climate Extremes.” Souza, C. Brotons, J.J. Meerpohl, C.J. Patel, B. “Crime: Social Disorganization and Relative Deprivation.” Environmental Research Letters 10(9): 094015. Djulbegovic, P. Alonso-Coello, M.M. Bala and Social Science & Medicine 48(6): 719­731. G.H. Guyatt. Forthcoming. “Unprocessed Red Meat King, A.D., and L.J. Harrington. 2018. “The Inequality and Processed Meat Consumption: Dietary Guideline Kaza, S., L. Yao, P. Bhada-Tata and F. Van Woerden. of Climate Change from 1.5 to 2 C of Global Warming.” Recommendations from the Nutritional Recommendations 2018. What a Waste 2.0: A Global Snapshot of Solid Geophysical Research Letters 45(10): 5030­5033. (NutriRECS) Consortium.” Annals of Internal Medicine. Waste Management to 2050. Urban Development Series. Washington, DC: World Bank. https://openknowledge. Kishor, S., and K. Johnson. 2004. Profiling Domestic Jones, P.J.S. 2009. “Equity, Justice and Power Issues worldbank.org/handle/10986/30317. Accessed 14 August Violence: A Multi-Country Study. Calverton, MD: ORC Raised by No-take Marine Protected Area Proposals.” 2019. Macro. Marine Policy 33(5): 759­765. Kearl, H. 2018. “The Facts Behind the# metoo Movement: Klasen, S. 2018. “Human Development Indices and Jongman, B., H.C. Winsemius, J.C. Aerts, E.C. de A National Study on Sexual Harassment and Assault: Indicators: A Critical Evaluation.” Occasional Paper. United Perez, M.K. van Aalst, W. Kron and P.J. Ward. 2015. Executive Summary.” Stop Street Harassment, Reston, VA. Nations Development Programme, Human Development “Declining Vulnerability to River Floods and the Global Report Office, New York. Benefits of Adaptation.” Proceedings of the National Keleher, H., and L. Franklin. 2008. “Changing Gendered Academy of Sciences 112(18): E2271­E2280. Norms about Women and Girls at the Level of Household Klasen, S., and M. Fleurbaey. 2018, “Leaving No One and Community: A Review of the Evidence.” Global Public Behind: Some Conceptual and Empirical Issues.” Jorgenson, A., J. Schor and X. Huang. 2017. “Income Health 3(Sup1): 42­57. CDP Background Paper 44 ST/ESA/2018/CDP/44. United Inequality and Carbon Emissions in the United States: A Nations Department of Economic and Social Affairs, New State-Level Analysis, 1997­2012.” Ecological Economics Kelley, C.P., S. Mohtadi, M.A. Cane, R. Seager and Y. York. 134(C): 40­48. Kushnir. 2015. “Climate Change in the Fertile Crescent and Implications of the Recent Syrian Drought.” Klein, N. 2019. On Fire: The (Burning) Case for a Green New Jost, J.T. 2019. “A Quarter Century of System Justification Proceedings of the National Academy of Sciences 112(11): Deal. New York: Simon & Schuster. Theory: Questions, Answers, Criticisms, and Societal 3241­3246. Applications.” British Journal of Social Psychology 58(2): Klein, R.J.T., G.F. Midgley, B.L. Preston, M. Alam, F.G.H. 263­314. Kennedy, P.J., and A. Prat. 2019. “Where do People Get Berkhout, K. Dow and M.R. Shaw. 2014. “Adaptation their News?” Economic Policy 34(97): 5­47. Opportunities, Constraints, and Limits.” In C.B. Field, V.R. Jost, J.T., A. Ledgerwood and C.D. Hardin. 2008. “Shared Barros, D.J. Dokken, K.J. Mach, M.D. Mastrandrea, T.E. Reality, System Justification, and the Relational Basis of Key, T.J., P.N. Appleby, E.A. Spencer, R.C. Travis, Bilir, M. Chatterjee, K.L. Ebi, Y.O. Estrada, R.C. Genova, Ideological Beliefs.” Social and Personality Psychology A.W. Roddam and N.E. Allen. 2009. “Mortality B. Girma, E.S. Kissel, A.N. Levy, S. MacCracken, P.R. Compass 2: 171­186. in British Vegetarians: Results from the European Mastrandrea and L.L. White, eds. Climate Change 2014: Prospective Investigation into Cancer and Nutrition (EPIC- Impacts, Adaptation, and Vulnerability. Part A: Global and Joyce, R., and X. Xu. 2019. “Inequalities in the Twentieth- Oxford).” American Journal of Clinical Nutrition 89(5): Sectoral Aspects. Contribution of Working Group II to the First Century.” Introducing the IFS Deaton Review. Institute 1613S­1619S. Fifth Assessment Report of the Intergovernmental Panel for Fiscal Studies, London. on Climate Change. Cambridge, UK: Cambridge University Khan, L.M. 2017. “The Ideological Roots of America’s Press. www.ipcc.ch/site/assets/uploads/2018/02/ Kågesten A, S. Gibbs, R.W. Blum, C. Moreau, V. Market Power Problem.” Yale Law Journal Forum 127: WGIIAR5-Chap16_FINAL.pdf. Accessed 16 August 2019. Chandra-Mouli, A. Herbert and A. Amin. 2016. 960. “Understanding Factors that Shape Gender Attitudes in Klemm, A., L. Liu, V. Mylonas and P. Wingender. Early Adolescence Globally: A Mixed-Methods Systematic Khan, M. 2019. “EU Plans Sweeping Regulation of Facial 2018. “Are Elasticities of Taxable Income Rising?” IMF Review.” PLoS ONE 11(6): e0157805. Recognition.” Financial Times, 22 August.

280 | HUMAN DEVELOPMENT REPORT 2019 Working Paper WP/18/132. International Monetary Fund, Kuznets, S. 1953. Shares of Upper Income Groups in Income Le Page, M. 2019. “Hurricane Dorian Is Joint Strongest Washington, DC. and Savings. New York: National Bureau of Economic Atlantic Storm Ever to Hit Land.” New Scientist, 2 Research. September. Klenert, D., L. Mattauch, E. Combet, O. Edenhofer, C. Hepburn, R. Rafaty and N. Stern. 2018. “Making Carbon ------. 1955. “Economic Growth and Income Inequality.” Le Quéré, C., R.M. Andrew, P. Friedlingstein, S. Sitch, Pricing Work for Citizens.” Nature Climate Change 8(8): American Economic Review 45(1): 1­28. J. Hauck, J. Pongratz, P.A. Pickers, J.I. Korsbakken, 669­677. G.P. Peters, J.G. Canadell and A. Arneth. 2018. ------. 1971. “Nobel Prize Lecture: Modern Economic “Global Carbon Budget 2018.” Earth System Science Data Klinsky, S., and H. Winkler. 2018. “Building Equity in: Growth: Findings and Reflections.” Nobelprize.org. www. (Online) 10(4). Strategies for Integrating Equity into Modelling for a 1.5 nobelprize.org/prizes/economic-sciences/1971/kuznets/ C World.” Philosophical Transactions of the Royal Society lecture. Accessed 6 October 2019. Lebreton, L., B. Slat, F. Ferrari, B. Sainte-Rose, J. A: Mathematical, Physical and Engineering Sciences Aitken, R. Marthouse, S. Hajbane, S. Cunsolo, A. 376(2119). Laboratoria. 2019. “Building the Digital Skills Young Schwarz, A. Levivier and K. Noble. 2018. “Evidence Women in Latin America Need to Thrive in Tech.” www. that the Great Pacific Garbage Patch is Rapidly Kolcava, D., Q. Nguyen and T. Bernauer. 2019. “Does laboratoria.la/en/impact. Accessed 9 October 2019. Accumulating Plastic.” Scientific Reports 8(1). Trade Liberalization Lead to Environmental Burden Shifting in the Global Economy?” Ecological Economics 163: Lakner, C., D.G. Mahler, M. Negre and E.B. Prydz. 2019. LeCun, Y., Y. Bengio and G. Hinton. 2015. “Deep 98­112. “How Much Does Reducing Inequality Matter for Global Learning.” Nature 521(7553): 436. Poverty?” Policy Research Working Paper 8869. World Kompas, T., V.H. Pham and T.N. Che. 2018. “The Effects Bank, Washington, DC. https://openknowledge.worldbank. Lee, J.J., R. Wedow, A. Okbay, E. Kong, O. Maghzian, of Climate Change on GDP by Country and the Global org/bitstream/handle/10986/31796/WPS8869.pdf. M. Zacher, T.A. Nguyen-Viet, P. Bowers, J. Economic Gains from Complying with the Paris Climate Accessed 1 August 2019. Sidorenko, R.K. Linnér and M.A. Fontana. 2018. “Gene Accord.” Earth’s Future 6(8): 1153­1173. Discovery and Polygenic Prediction from a Genome-Wide Lakner, C., and B. Milanovic. 2015. “Global Income Association Study of Educational Attainment in 1.1 Million Korinek, A. 2019. “Integrating Ethical Values and Economic Distribution: From the Fall of the Berlin Wall to the Great Individuals.” Nature Genetics 50(8): 1112­1121. Value to Steer Progress in Artificial Intelligence.” NBER Recession.” World Bank Economic Review 30(2): 203­232. Working Paper 26130. National Bureau of Economic Lee, K.-F. 2018. AI Superpowers: China, Silicon Valley, and Research, Cambridge, MA. Lambert, S., and P. De Vreyer. 2017. “By Ignoring Intra- the New World Order. Boston, MA: Houghton Mifflin Household Inequality Do We Underestimate the Extent of Harcourt. Korinek, A., and J. Kreamer. 2014. “The Redistributive Poverty?” Working Paper DT/2017/05. Développement, Effects of Financial Deregulation.” Journal of Monetary Institutions et Mondialisation, Paris. Lefebvre, P., P. Merrigan and M. Verstraete. 2009. Economics 68: S55­S67. “Dynamic Labour Supply Effects of Childcare Subsidies: Lamont, M. 2018. “Addressing Recognition Gaps: Evidence from a Canadian Natural Experiment on Low-Fee Korinek, A., and J. Stiglitz, J. 2017. “Artificial Intelligence Destigmatization and the Reduction of Inequality.” Universal Child Care.” Labour Economics 16(5): 490­502. and Its Implications for Income Distribution and American Sociological Review 83(3): 419­44. Unemployment” NBER Working Paper 24174. National Leigh, A. 2006. “Trust, Inequality and Ethnic Heterogeneity.” Bureau of Economic Research, Cambridge, MA. Lancee, B., and H.G. Van de Werfhorst. 2012. “Income Economic Record 82(258): 268­280. Inequality and Participation: A Comparison of 24 European Kousky, C., and R. Cooke. 2012. “Explaining the Failure to Countries.” Social Science Research 41(5): 1166­1178. Lemoine, D., and C. Traeger. 2014. “Watch Your Step: Insure Catastrophic Risks.” Geneva Papers on Risk and Optimal Policy in a Tipping Climate.” Economic Policy 6(1): Insurance-Issues and Practice 37(2): 206­227. Langer, A. 2005. “Horizontal Inequalities and Violent 137­66. Conflict.” Côte d’Ivoire Country Paper. Occasional Paper Kovacevic, M. 2019. “Poverty and Inequality.” Unpublished 2005/32. United Nations Development Programme, Human Levine, R. 2005. “Finance and Growth: Theory and manuscript. Development Report Office, New York. http://hdr.undp. Evidence.” Handbook of Economic Growth 1: 865­934. org/sites/default/files/hdr2005_langer_arnim_32.pdf. Kraay, A. 2015. “Weak Instruments in Growth Regressions: Accessed 6 August 2019. Lewis, A.W. 1954. “Economic Development with Unlimited Implications for Recent Cross-Country Evidence on Supplies of Labor.” Manchester School of Economic and Inequality and Growth.” Policy Research Working Paper Langer, A., and F. Stewart. 2015. “Regional Imbalances, Social Studies 22: 139­191. 7494. World Bank, Washington, DC. Horizontal Inequalities, and Violent Conflicts: Insights from Four West African Countries.” World Bank, Washington, Li, Z., Y. Jiang, M. Li and C. Lu. 2018. “Inequalities in Kramarz, F., and O.N. Skans. 2014. “When Strong Ties Are DC. http://documents.worldbank.org/curated/ Socio-Emotional Development and Positive Parenting Strong: Networks and Youth Labour Market Entry.” Review en/768071468191326719/Regional-imbalances-horizontal- During Childhood: Evidence from China 2010­2014.” SSM- of Economic Studies 81(3): 1164­1200. inequalities-and-violent-conflicts-insights-from-four-West- Population Health 5: 8­16. African-countries. Accessed 6 August 2019. Kreiner, C.T., T.H. Nielsen and B.L. Serena. 2018. “Role Lian, W. 2019. “Technological Changes, Offshoring, and the of Income Mobility for the Measurement of Inequality in Larson, J.L. 2010. The Market Revolution in America: Labor Share.” IMF Working Paper 19.142. International life Expectancy.” Proceedings of the National Academy of Liberty, Ambition, and the Eclipse of the Common Good. Monetary Fund, Washington, DC. Sciences 115(46): 11754­11759. Cambridge, UK: Cambridge University Press. Lian, W., N. Novta, E. Pugacheva, Y. Timmer and P. Krishna, A. 2010. One Illness Away: Why People Become Latinobarometro. 2018. Informe Latinobarómetro Topalova. 2019. “The Price of Capital Goods: A Driver of Poor and How They Escape Poverty. Oxford, UK: Oxford 2018. Santiago. www.latinobarometro.org/latdocs/ Investment under Threat.” IMF Working Paper WP/19/134. University Press. INFORME_2018_LATINOBAROMETRO.pdf. Accessed 18 International Monetary Fund, Washington, DC. October 2019. Krueger, A.B. 2012. “The Rise and Consequences of Liddle, B. 2015. “What Are the Carbon Emission Elasticities Inequality in the United States.” Speech at the Center for Laurian, L. 2008. “Environmental Injustice in France.” for Income and Population? Bridging STIRPAT and ECK American Progress, 12 January, Washington, DC. www. Journal of Environmental Planning and Management 51(1): via Robust Heterogeneous Panel Estimates.” Global govinfo.gov/content/pkg/ERP-2012/pdf/ERP-2012.pdf. 55­79. Environmental Change 31: 62­73. Accessed 7 August 2019. Lazonick, W.H., and M. Mazzucato. 2013. “The Risk- Lindahl, M., E. Lundberg, M. Palme and Emilia Kumar, A., and T. Rahman. 2018. “Can a Women’s Rural Reward Nexus in the Innovation-Inequality Relationship: Simeonova. 2016. “Parental Influences on Health and Livelihood Program Improve Mental Health? Experimental Who Takes the Risks? Who Gets the Rewards?” Industrial Longevity: Lessons from a Large Sample of Adoptees.” Evidence from India.” Paper presented at the Annual & Corporate Change 22(4): 1093­1128. NBER Working Paper 21946. National Bureau of Economic Meeting of the Agricultural and Applied Economics Research, Cambridge, MA. Association, 5­7 August, Washington, DC. Le, L.T., and J. Sabaté. 2014. “Beyond Meatless, the Health Effects of Vegan Diets: Findings from the Adventist Liu, J., T. Dietz, S.R. Carpenter, M. Alberti, C. Folke, Kus, B. 2012. “Financialisation and Income Inequality in Cohorts.” Nutrients 6(6): 2131­2147. E. Moran, A.N. Pell, P. Deadman, T. Kratz, J. OECD Nations: 1995-2007.” Economic and Social Review Lubchenco, E. Ostrom, Z. Ouyang, W. Provencher, 43(4): 477­495. C.L. Redman, S.H. Schneider and W.W. Taylor. 2007.

References | 281 “Complexity of Coupled Human and Natural Systems.” P. Vineis, C. White, B. Wojtyniak, Y. Hu and W.J. ------. 2018. “Undoing Segmentation? Latin American Science 317(5844): 1513­1516. Nusselder. 2018. “Trends in Health Inequalities in Healthcare Policy During the Economic Boom.” Social 27 European Countries.” Proceedings of the National Policy & Administration 52(6): 1181­1200. Loaiza Sr., E., and S. Wong. 2012. Marrying Too Young: Academy of Sciences 115(25): 6440­6445. End Child Marriage. New York: United Nations Population ------. 2019a. “The Relationship between Universal Social Fund. Mackie, G., and J. Le Jeune. 2009. “Social Dynamics of Policy and Inequality: A Comparative Political Economy Abandonment of Harmful Practices: A New Look at the Approach.” Background paper for Human Development López-Calva, L.F., and E. Ortiz-Juarez. 2014. “A Theory.” Special Series on Social Norms and Harmful Report 2019. United Nations Development Programme, Vulnerability Approach to the Definition of the Middle Practices. Innocenti Working Paper 2009-06. Florence, Human Development Report Office, New York. Class.” Journal of Economic Inequality 12(1): 23­47. Italy: United Nations Children’s Fund, Innocenti Research Centre. ------. 2019b. “Undoing Segmentation? Latin American López-Calva, L.F., and C. Rodríguez-Castelán. 2016. Health Care and Pensions after the Economic Boom.” “Pro-Growth Equity: A Policy Framework for the Twin Mackie, G., F. Moneti, H. Shakya and E. Denny. 2015. Goals.” Policy Research Working Paper 7897. World “What Are Social Norms? How Are They Measured?” Martinez-Alier, J., L. Temper, D. Del Bene and A. Bank, Washington, DC. http://documents.worldbank.org/ Working Paper 1. United Nations Children’s Fund, New Scheidel. 2016. “Is There a Global Environmental Justice curated/en/911711479931074058/pdf/WPS7897.pdf. York, and University of California­San Diego Centre on Movement?” Journal of Peasant Studies 43(3): 731­755. Accessed 23 August 2019. Global Justice, San Diego, CA. Martuzzi, M., F. Mitis and F. Forastiere. 2010. Lucas, R.E. 2004. “The Industrial Revolution: Past and Magnani, E. 2000. “The Environmental Kuznet’s Curve, “Inequalities, Inequities, Environmental Justice in Waste Future.” 2003 Annual Report Essay. Federal Reserve Environmental Protection Policy and Income Distribution.” Management and Health.” European Journal of Public Bank of Minneapolis, Minneapolis, MN. https://ideas. Ecological Economics 32: 431­443. Health 20(1): 21­26. repec.org/a/fip/fedmar/y2004imayp5-20nv18no.1.html. Accessed 26 September 2019. Majer, I.M., W.J. Nusselder, J.P. Mackenbach and A.E. Marx, A., J. Soares and W. Van Acker. 2015. “The Kunst. 2011. “Socioeconomic Inequalities in Life and Protection of the Rights of Freedom of Association and Luohan Academy. 2019. “Digital Technology and Inclusive Health Expectancies around Official Retirement Age in 10 Collective Bargaining. A Longitudinal Analysis over 30 Growth.” Hangzhou, China. https://gw.alipayobjects. Western-European Countries.” Journal of Epidemiology Years in 73 Countries.” In A. Marx, J. Wouters, G. Rayp com/os/antfincdn/DbLN6yXw6H/Luohan_Academy- and Community Health 65(11): 972­979. and L. Beke, eds., Global Governance of Labour Rights: Report_2019_Executive_Summary.pdf. Accessed 14 Assessing the Effectiveness of Transnational Public and August 2019. Major, B. 1994. “From Social Inequality to Personal Private Policy Initiatives. Northampton, MA: Edward Elgar Entitlement: The Role of Social Comparisons, Legitimacy Publishing. Lusseau, D., and F. Mancini. 2019. “Income-Based Appraisals, and Group Membership.” Advances in Variation in Sustainable Development Goal Interaction Experimental Social Psychology 26: 293­355. Maskus, K.E. 2004. “Encouraging International Technology Networks.” Nature Sustainability 2: 242­247. Transfer.” UNCTAD-ICTSD Issue Paper 7. United Nations Malouf Bous, K., and J. Farr. 2019. “False Promises: How Conference on Trade and Development and International Lustig, N. 2000. “Crises and the Poor: Socially Responsible Delivering Education Through Public-Private Partnerships Centre for Trade and Sustainable Development, Geneva. Macroeconomics.” Economía Journal 1: 1­30. Risks Fueling Inequality Instead of Achieving Quality Education for All.” Oxfam Briefing Paper. Oxfam GB, Mathur, A., and A. Morris. 2012. “Distributional Effects ------, ed. 2018a. Commitment to Equity Handbook: Oxford, UK. https://policy-practice.oxfam.org.uk/publica- of a Carbon Tax in the Context of Broader Fiscal Reform.” Estimating the Impact of Fiscal Policy on Inequality and tions/false-promises-how-delivering-education-through- Climate and Energy Economics Discussion Paper. The Poverty. Washington, DC: Brookings Institution Press. public-private-partnerships-ris-620720. Accessed 10 Brookings Institution, Washington, DC. www.brookings. October 2019. edu/research/distributional-effects-of-a-carbon-tax-in- ------. 2018b. “Fiscal Policy, Income Redistribution, and the-context-of-broader-fiscal-reform/. Accessed 13 August Poverty Reduction in Low- and Middle- Income Countries.” Mann, S., and M. Hilbert. 2018. “AI4D: Artificial 2019. In N. Lustig, ed., Commitment to Equity Handbook: Intelligence for Development.” https://ssrn.com/ab- Estimating the Impact of Fiscal Policy on Inequality and stract=3197383. Accessed 15 August 2019. Maulia, E. 2018. “Go-Jek Sparks an Indonesian Banking Poverty. Washington, DC: The Brookings Institution. Revolution.” Nikkei Asian Review, 29 August. https:// Marcus, R. 2018. The Norms Factor: Recent Research asia.nikkei.com/Spotlight/Cover-Story/Go-Jek-sparks-an- ------. 2018c. “The Sustainable Development Goals, on Gender, Social Norms, and Women’s Economic Indonesian-banking-revolution. Accessed 24 October 2019. Domestic Resource Mobilization and the Poor.” In J.A. Empowerment. Ottawa: International Development Ocampo and J. Stiglitz, eds., The Welfare State Revisited. Research Centre. Maxwell, J., and F. Briscoe. 1998. “There’s Money in New York: Columbia University Press. the Air: The CFC Ban and DuPont’s Regulatory Strategy.” Marcus, R., and C. Harper. 2014. “Gender Justice and Business Strategy and the Environment 6(5): 276­286. ------, ed. Forthcoming. Commitment to Equity Social Norms—Processes of Change for Adolescent.” Handbook: Estimating the Impact of Fiscal Policy on Girls: Towards a Conceptual Framework 2. Overseas Mazzucato, M. 2011. The Entrepreneurial State. London: Inequality and Poverty, 2nd edition. Washington, DC: Development Institute, London. Demos. Brookings Institution Press, and New Orleans, LA: Tulane University. ------. 2015. “How Do Gender Norms Change?” Overseas Mazzucato, M. 2013. The Entrepreneurial State: Debunking Development Institute, London. Public vs. Private Sector Myths, Vol. 1. London: Anthem Lustig, N., L.F. Lopez-Calva and E. Ortiz-Juarez. 2013. Press. “Deconstructing the Decline in Inequality in Latin Margai, F.L. 2001. “Health Risks and Environmental Inequity: America.” Policy Research Working Paper 6552. World A Geographical Analysis of Accidental Releases of Mazzucato, M., and G. Semieniuk. 2017. “Public Bank, Washington, DC. Hazardous Materials.” Professional Geographer 53(3): Financing of Innovation: New Questions.” Oxford Review 422­434. of Economic Policy 33(1): 24­48. MacInnis, C.C., and G. Hodson. 2019. “Extending the Benefits of Intergroup Contact beyond Attitudes: When Marmot, M. 2005. “Social Determinants of Health McCallum, M.L. 2015. “Vertebrate Biodiversity Losses Point Does Intergroup Contact Predict Greater Collective Action Inequalities.” Lancet 365(9464): 1099­1104. to a Sixth Mass Extinction.” Biodiversity and Conservation Support?” Journal of Theoretical Social Psychology 3: 24(10): 2497­2519. 11­22. Marrero, G., and J. Rodríguez. 2013. “Inequality of Opportunity and Growth.” Journal of Development McDonald, P., and D. White. 2018. “The Backlash against Macintyre, S. 1997. “The Black Report and Beyond: What Economics 104: 107­122. Gender Equality Is Arising in New Forms. Opponents of Are the Issues?” Social Science & Medicine 44(6): Gender Equality Now Question the Research Methods in 723­745. Martínez, J., and D. Sánchez-Ancochea. 2016. Studies of Sexual Harassment and Assault. London School “Achieving Universalism in Developing Countries.” of Economics Blogs.” London School of Economic and Mackenbach, J.P., J.R. Valverde, B. Artnik, M. Bopp, Background paper for Human Development Report Political Science, London. H. Brønnum-Hansen, P. Deboosere, R. Kalediene, K. 2016. United Nations Development Programme, Human Kovács, M. Leinsalu, P. Martikainen, G. Menvielle, Development Report Office, New York. McEniry, M., R. Samper-Ternent, C.E. Flórez, R. Pardo E. Regidor, J. Rychtaríková, M. Rodriguez-Sanz, and C. Cano-Gutierrez. 2018. “Patterns of SES Health

282 | HUMAN DEVELOPMENT REPORT 2019 Disparities among Older Adults in Three Upper Middle-and ------. 2016. Global Inequality: A New Approach for the 2001-2015.” WID.world Working Paper 2017/12. World Two High-Income Countries.” Journals of Gerontology: Age of Globalization. Cambridge, MA: Harvard University Inequality Database. Series B 74(6): e25­e37. Press. Moser, C.O. 1989. “Gender Planning in the Third World: McGee, J.M., and P.T. Greiner. 2018. “Can Reducing Milanovic, B., P.H. Lindert and J.G. Williamson. 2010. Meeting Practical and Strategic Gender Needs.” World Income Inequality Decouple Economic Growth from CO2 “Pre-Industrial inequality.” Economic Journal 121(551): Development 17(11): 1799­1825. Emissions?” Socius: Sociological Research for a Dynamic 255­72. World 4: 1­11. Moser, S., and S. Kleinhückelkotten. 2017. “Good Millennium Ecosystem Assessment. 2015. “Ecosystems Intents, but Low Impacts: Diverging Importance of McKinsey. 2018. “Mobile Money in Emerging Markets: The and Human Well-being: Synthesis.” Washington, DC: Motivational and Socioeconomic Determinants Explaining Business Case for Financial Inclusion.” www.mckinsey. Island Press. www.millenniumassessment.org/documents/ Pro-environmental Behavior, Energy Use and Carbon com/~/media/McKinsey/Industries/Financial%20 document.356.aspx.pdf. Accessed 8 August 2019. Footprint.” Environment and Behavior 50(6): 1­31. Services/Our%20Insights/Mobile%20money%20in%20 emerging%20markets%20The%20business%20case%20 Miller, D. 2016. “Intersectionality: How Gender Interacts Mosse, D. 2018. “Caste and Development: Contemporary for%20financial%20inclusion/Mobile-money-in-emerging- with Other Social Identities to Shape Bias.” The Perspectives on a Structure of Discrimination and markets.ashx. Accessed 14 August 2019. Conversation, 4 February. Advantage.” World Development 110: 422­436.

McLaren, D., O. Cottray and M. Taylor. 1999. Pollution Milman, O. 2018. “Robert Bullard: `Environmental Justice Moyer, J., D. Bohl, T. Hanna, B. Mapes and M. Rafa. Injustice: The Geographic Relation between Household Isn’t Just Slang, It’s Real.’” The Guardian, 20 December. 2019. Assessing the Impact of War on Development in Income and Polluting Factories. London: Friends of the Yemen. Sana’a: United Nations Development Programme. Earth Trust. Miodownik, D., and L. Nir. 2016. “Receptivity to Violence www.arabstates.undp.org/content/rbas/en/home/library/ in Ethnically Divided Societies: A Micro-Level Mechanism crisis-response0/assessing-the-impact-of-war-on-develop- McNeill, J.R. 2001. Something New under the Sun: An of Perceived Horizontal Inequalities.” Studies in Conflict & ment-in-yemen-.html. Accessed 24 October 2019. Environmental History of the Twentieth-Century World. Terrorism 39(1): 22­45. Global Century Series. New York: WW Norton & Company. Mukhopadhyay, T., C. Rivera and H. Tapia. 2019. “Gender Mishel, L., J. Schmitt and H. Shierholz. 2013. “Assessing Inequality and Multidimensional Social Norms.” Working McNeill, W.H. 1976. Plagues and Peoples. Garden City, NY: the Job Polarization Explanation of Growing Wage Paper. United Nations Development Programme, Human Anchor Press. Inequality.” Working Paper. Economic Policy Institute, Development Report Office, New York. Washington, DC. www.epi.org/files/2012/wp295-assess- McSweeney, K., and O.T. Coomes. 2011. “Climate-Related ing-job-polarization-explanation-wage-inequality.pdf. Munoz Boudet, A.M., P. Buitrago, B. Leroy De La Briere, Disaster Opens a Window of Opportunity for Rural Poor Accessed 18 September 2019. D.L. Newhouse, E.C. Rubiano Matulevich, K. Scott in Northeastern Honduras.” Proceedings of the National and P. Suarez Becerra. 2018. “Gender Differences in Academy of Sciences 108(13): 5203­5208. Mishra, S., and R.N. Carleton. 2015. “Subjective Relative Poverty and Household Composition through the Life- Deprivation Is Associated with Poorer Physical and Mental Cycle: A Global Perspective.” Policy Research Working Mejia, S.A., C. Baccianti, M. Mrkaic, N. Novta, E. Health.” Social Science & Medicine 147: 144­149. Paper 8360. World Bank, Washington, DC. Pugacheva and P. Topalova. 2019. “Weather Shocks and Output in Low-income Countries: Adaptation and Mitnik, P.A., E. Cumberworth and D.B. Grusky. 2016. Munoz Boudet, A.M., P. Petesch and C. Turk, with A. the Role of Policies.” IMF Working Paper. International “Social Mobility in a High-Inequality Regime.” The Annals Thumala. 2012. On Norms and Agency: Conversations Monetary Fund, Washington, DC. of the American Academy of Political and Social Science about Gender Equality with Women and Men in 20 663(1): 140­184. Countries. Washington, DC: World Bank. Mekonnen, M.M., and A.Y. Hoekstra. 2011. “National Water Footprint Accounts: The Green, Blue and Grey Moazed, A., and N.L. Johnson. 2016. Modern Monopolies: Muralidharan, K., A. Singh and A.J. Ganimian. Water Footprint of Production and Consumption.” Value What It Takes to Dominate the 21st Century Economy. New 2019. “Disrupting Education? Experimental Evidence of Water Research Report 50. UNESCO-IHE Institute York, NY: St. Martin’s Press. on Technology-Aided Instruction in India.” American for Water Education, Delft, The Netherlands. https:// Economic Review 109(4): 1426­1460. waterfootprint.org/media/downloads/Report50- Mokyr, J. 2002. The Gifts of Athena: Historical Origins of the NationalWaterFootprints-Vol1.pdf. Accessed 15 August Knowledge Economy. Princeton, NJ: Princeton University Murillo, J., and C. Martínez Garrido. 2017. “Segregación 2019. Press. Social en las Escuelas Públicas y Privadas en América Latina.” Educação & Sociedade 38(140): 727­750. ------. 2016. “Four Billion People Facing Severe Water ------. 2016. A Culture of Growth: The Origins of the Scarcity.” Science Advances 2(2). Modern Economy. Princeton, NJ: Princeton University Murtin, F., J. Mackenbach, D. Jasilionis and M.M. Press. d’Ercole. 2017. “Inequalities in Longevity by Education Meltzer, A.H., and S.F. Richards. 1981. “A Rational Theory in OECD Countries: Insights from New OECD Estimates.” of the Size of Government.” Journal of Political Economy Molyneux, M. 1985. “Mobilization without Emancipation? OECD Statistics Working Papers 2017/02. Organisation for 89(5): 914­927. Women’s Interests, the State, and Revolution in Economic Co-operation and Development, Paris. Nicaragua.” Feminist Studies 11(2): 227­254. Mendez Ramos, F. 2019. “Uncertainty in Ex-Ante Poverty Myers, M., and L. Juma. 2018. Defending Independent and Income Distribution: Insights from Output Growth and Montenegro, C.E., and H.A. Patrinos. 2014. Comparable Media: A Comprehensive Analysis of Aid Flows. Natural Resource Country Typologies.” Policy Research Estimates of Returns to Schooling around the World. Washington, DC: National Endowment for Democracy, Working Paper 8841. World Bank, Washington, DC. Washington, DC: World Bank. Center for International Media Assistance. www.cima. https://openknowledge.worldbank.org/bitstream/han- ned.org/publication/comprehensive-analysis-media-aid- dle/10986/31666/WPS8841.pdf?sequence=4. Accessed 1 Mora, C., A.G. Frazier, R.J. Longman, R.S. Dacks, flows/. Accessed 1 October 2019. August 2019. M.M. Walton, E.J. Tong, J.J. Sanchez, L.R. Kaiser, Y.O. Stender, J.M. Anderson, C.M. Ambrosino, Naidu, S., E.A. Posner and G. Weyl. 2018. “Antitrust Messenger, C. 2017. “Cyber Violence against Women and I. Fernandez-Silva, L.M. Giuseffi and T.W. Remedies for Labor Market Power.” Harvard Law Review Girls Exacerbates Digital Exclusion.” Digital @ DAI, 20 Giambelluca. 2013. “The Projected Timing of Climate 132: 536. November. Departure from Recent Variability.” Nature 502(7470): 183­187. Nakatani, R. 2019. “A Possible Approach to Fiscal Rules Metz, C. 2019. “Is Ethical AI Even Possible?” New York in Small Island—Incorporating Natural Disasters and Times, 1 March. www.nytimes.com/2019/03/01/business/ Morand, S., K.M. McIntyre and M. Baylis. 2014. Climate Change.” IMF Working Paper. International ethics-artificial-intelligence.html. Accessed 19 August “Domesticated Animals and Human Infectious Diseases of Monetary Fund, Washington, DC. 2019. Zoonotic Origins: Domestication Time Matters.” Infection, Genetics and Evolution 24: 76­81. Narayan, A., R. Van der Weide, A. Cojocaru, C. Lakner, Milanovic, B. 2005. “Can We Discern the Effect of S. Redaelli, D.G. Mahler, R.G.N. Ramasubbaiah and Globalization on Income Distribution? Evidence from Morgan, M. 2017. “Falling Inequality beneath Extreme S. Thewissen. 2018. Fair Progress?: Economic Mobility Household Surveys.” World Bank Economic Review 19(1): and Persistent Concentration: New Evidence for Brazil across Generations Around the World. Washington, DC: 21­44. Combining National Accounts, Surveys and Fiscal Data, World Bank.

References | 283 NDI (National Democratic Institute). 2019. “#NotTheCost ------. 2017a. The Pursuit of Gender Equality: An Uphill Orlich, M.J., P.N. Singh, J. Sabaté, K. Jaceldo-Siegl, J. Stopping Violence against Women in Politics.” Battle. Paris. Fan, S. Knutsen, W.L. Beeson and G.E. Fraser. 2013. Washington, DC. “Vegetarian Dietary Patterns and Mortality in Adventist ------. 2017b. “Terms of Reference: OECD Project on the Health Study 2.” JAMA Internal Medicine 173(13): Nelson, R. 1993. National Innovation Systems. A Distribution of Household Incomes.” Income Distribution 1230­1238. Comparative Analysis. New York/Oxford: Oxford University Database. Paris. Press. O’Rourke, K.H., A. Rahman and A.M. Taylor. 2019. “Trade, ------. 2018a. A Broken Social Elevator? How to Promote Technology, and the Great Divergence.” NBER Working Neves, P.C., Ó. Afonso and S.T. Silva. 2016. “A Meta- Social Mobility. Paris. Paper 25741. National Bureau of Economic Research, Analytic Reassessment of the Effects of Inequality on Cambridge, MA. Growth.” World Development 78: 386­400. ------. 2018b. “OECD Clamps Down on CRS Avoidance through Residence and Citizenship by Investment Ortiz, I., and M. Cummins. 2011. “Global Inequality: Beyond Newell, R., and K. Rogers. 2010. “Leaded Gasoline in the Schemes.” 16 October. https://oecd.org/tax/oecd-clamps- the Bottom Billion­A Rapid Review of Income Distribution United States: The Breakthrough of Permit Trading.” In W. down-on-crs-avoidance-through-residence-and-citizen- in 141 Countries.” Social and Economic Policy Working Harrington, R. Morgenstern and T. Sterner, eds. Choosing ship-by-investment-schemes.htm. Accessed 7 November Paper. United Nations Children’s Fund, New York. Environmental Policy. New York: Routledge. 2019. Ostrom, E. 2000. “Collective Action and the Evolution of Niño-Zarazúa, M., L. Roope and F. Tarp. 2017. “Global ------. 2018c. “The Role and Design of Net Wealth Taxes Social Norms.” Journal of Economic Perspectives 14(3): Inequality: Relatively Lower, Absolutely Higher. Review of in the OECD.” OECD Tax Policy Studies 26. Paris. 137­158. Income and Wealth 63(4): 661­684. ------. 2018d. Tax Challenges Arising from Digitalisation— Ostry, J.D., and Berg, A. 2011. “Inequality and Noked, N. 2018. “Tax Evasion and Incomplete Tax Interim Report 2018. Paris. Unsustainable Growth: Two Sides of the Same Coin?” Transparency.” Laws 7(3): 31. IMF Staff Discussion Note 11/08). International Monetary ------. 2019a. “Addressing the Tax Challenges of the Fund, Washington, DC. Nolan, B., R. Richiardi and L. Valenzuela. 2018. “The Digitalisation of the Economy—Public Consultation Drivers of Inequality in Rich Countries.” INET Oxford Document.” OECD Public Consultation Document. Paris. Ostry, J.D., A. Berg and C.G. Tsangarides. 2014. Working Paper No. 2018-15. University of Oxford, Institute www.oecd.org/tax/beps/public-consultation-document- “Redistribution, Inequality, and Growth.” IMF Staff for New Economic Thinking, Oxford, UK. addressing-the-tax-challenges-of-the-digitalisation-of-the- Discussion Note SDN/14/02. International Monetary Fund, economy.pdf. Accessed 7 November 2019. Washington, DC. Nordhaus, W., and A. Moffat. 2017. “A Survey of Global Impacts of Climate Change: Replication, Survey Methods, ------. 2019b. Getting Skills Right: Future-Ready Adult Ostry, J.D., P. Loungani and A. Berg. 2019. “Confronting and a Statistical Analysis.” Discussion Paper 2096. Learning Systems. Paris. Inequality: How Societies Can Choose Inclusive Growth.” Cowles Foundation for Research in Economics, New Chichester, NY: Columbia University Press. Haven, CT. https://cowles.yale.edu/sites/default/files/ ------. 2019c. Getting Skills Right: Engaging Low-Skilled files/pub/d20/d2096.pdf. Accessed 24 October 2019. Adults in Learning. Paris. Ottersen, G., and J. Melbourne-Thomas. 2019. “Time to Look Forward to Adapt to Ocean Warming.” Proceedings Norton, M.I., and D. Ariely. 2011. “Building a Better ------. 2019d. “Programme of Work to Develop a of the National Academy of Sciences 116 (37): America—One Wealth Quintile at a Time.” Perspectives Consensus Solution to the Tax Challenges Arising from the 18157­18158. on Psychological Science 6: 9­12. Digitalisation of the Economy.” Paris www.oecd.org/tax/ beps/programme-of-work-to-develop-a-consensus-solu- Ouedraogo, R., and I. Ouedraogo. 2019. “Gender ------. 2013. “America’s Desire for Less Wealth Inequality tion-to-the-tax-challenges-arising-from-the-digitalisation- Equality and Electoral Violence in Africa: Unlocking the Does Not Depend on How You Ask Them.” Judgement and of-the-economy.pdf. Accessed 7 November 2019. Peacemaking Potential of Women.” IMF Working Paper Decision Making 8: 393­394. WP/19/174. International Monetary Fund, Washington, ------. 2019e. Risks That Matter: Main Findings from the DC. Nussbaum, M.C. 2001. Women and Human Development: 2018 Risks That Matter Survey. Paris. www.oecd.org/ The Capabilities Approach, Vol. 3. Cambridge, UK: social/risks-that-matter.htm. Accessed 10 October 2019. OutRight Action International. 2019. Agenda 2030 for Cambridge University Press. LGBTI Health and Well-Being. New York. ------. 2019f. Under Pressure: The Squeezed Middle Class. ------. 2003. “Tragedy and Human Capabilities: A Paris. Oxfam. 2019. “Public Good or Private Wealth?” Oxfam Response to Vivian Walsh.” Review of Political Economy Briefing Paper. Oxford, UK. www.oxfam.org/en/research/ 15(3): 413­418. OECD (Organisation for Economic Co-operation and public-good-or-private-wealth. Accessed 26 September Development) and FAO (Food and Agriculture 2019. ------. 2011. Creating Capabilities: The Human Organization of the United Nations). 2017. OECD-FAO Development Approach. Cambridge, MA: Harvard Agricultural Outlook 2017-2026. Paris and Rome. www. Oyebode, O., U.J. Pape, A.A. Laverty, J.T. Lee, N. University Press. fao.org/3/a-i7465e.pdf. Accessed 29 October 2019. Bhan and C. Millett. 2015. “Rural, Urban and Migrant Differences in Non-Communicable Disease Risk-Factors O’Connor, A. 2019. “How Artificial Intelligence Could ------. 2018. OECD-FAO Agricultural Outlook 2018-2027. in Middle Income Countries: A Cross-Sectional Study of Transform Medicine.” New York Times, 11 March. www. Paris and Rome. https://read.oecd-ilibrary.org/agriculture- WHO-SAGE Data.” PLOS ONE 10(4): e0122747. nytimes.com/2019/03/11/well/live/how-artificial- and-food/oecd-fao-agricultural-outlook-2018-2027_agr_ intelligence-could-transform-medicine.html. Accessed 24 outlook-2018-en#page1. Accessed 15 August 2019. Paluck, E.L., and E. Ball, with C. Poynton and S. October 2019. Siedloff. 2010. “Social Norms Marketing Aimed at OECD (Organisation for Economic Co-operation Gender-Based Violence: A Literature Review and Critical O’Reilly, M., A.Ó. Súilleabháin and T. Paffenholz. 2015. and Development) and ILO (International Labour Assessment.” International Rescue Committee, New York. Reimagining Peacemaking: Women’s Roles in Peace Organization). 2019. Tackling Vulnerability in the Processes. New York: International Peace Institute. Informal Economy. Development Centre Studies. Paris. Park, J. 2015. “South Korean `Superdads’ on Paternity Leave https://doi.org/10.1787/939b7bcd-en. Accessed 10 Break with Tradition.” Reuters, 24 September. www. Odusola, A., G.A. Cornia, H. Bhorat and P. Conceição. October 2019. reuters.com/article/us-southkorea-superdads/south- 2017 Income Inequality Trends in Sub-Saharan Africa: korean-superdads-on-paternity-leave-break-with-tradition- Divergence, Determinants and Consequences. New York: Okun, A.M. 1975. Equality and Efficiency: The Big Tradeoff. idUSKBN0U626220151224. Accessed 9 October 2019. United Nations Development Programme. Washington, DC: Brookings Institution Press. Parry, J.-E., and A. Terton. 2016. “How Are Vulnerable OECD (Organisation for Economic Co-operation and Oleske, J.M. 2015. “State Inaction: Equal Protection, and Countries Adapting to Climate Change?” International Development). 2010. Improving Health and Social Religious Resistance to LGBT Rights.” University of Institute for Sustainable Development, Winnipeg, Canada. Cohesion through Education. Paris. Colorado Law Review 87(1): 63. www.iisd.org/faq/adapting-to-climate-change/. Accessed 13 August 2019. ------. 2016. “Be Flexible!” Background Brief on How Olivetti, C., and B. Petrongolo. 2017. “The Economic Workplace Flexibility Can Help European Employees to Consequences of Family Policies: Lessons from a Century Balance Work and Family. Paris. of Legislation in High-Income Countries.” Journal of Economic Perspectives 31(1): 205­230.

284 | HUMAN DEVELOPMENT REPORT 2019 Paskov, M., and C. Dewilde. 2012. “Income Inequality and Piketty, T., and G. Zucman. 2014. “Capital Is Back: Wealth- North-Holland. www.columbia.edu/~ap3116/papers/ Solidarity in Europe.” Research in Social Stratification and Income Ratios in Rich Countries 1700-2010.” Quarterly MediaHandbookPrat_v6.pdf. Accessed 26 August 2019. Mobility 30(4): 415­432. Journal of Economics 129(3): 1155­1210. Pratt, G.A. 2015. “Is a Cambrian Explosion Coming for Patnaik, A. 2019. “Reserving Time for Daddy: The Pimentel, D., and M. Pimentel. 2003. “Sustainability of Robotics?” Journal of Economic Perspectives 29(3): 51­60. Consequences of Fathers’ Quotas.” Journal of Labor Meat-Based and Plant-Based Diets and the Environment.” Economics 37(4): 1009­1059. American Journal of Clinical Nutrition 78(3): 660S­663S. Pretis, F., M. Schwarz, K. Tang, K. Haustein and M.R. Allen. 2018. “Uncertain Impacts on Economic Growth PCT (Platform for Collaboration on Tax). 2019. “PCT Pimm, S.L., C.N. Jenkins, R. Abell, T.M. Brooks, J.L. When Stabilizing Global Temperatures at 1.5 C or 2 C Progress Report 2018-2019.” Paris. www.oecd.org/ctp/ Gittleman, L.N. Joppa, P.H. Raven, C.M. Roberts and Warming.” Philosophical Transactions of the Royal Society tax-global/platform-for-collaboration-on-tax-progress- J.O. Sexton. 2014. “The Biodiversity of Species and their A: Mathematical, Physical and Engineering Sciences report-2018-2019.pdf. Accessed 10 October 2019. Rates of Extinction, Distribution, and Protection.” Science 376(2119). 344(6187): 1246752. People’s Bank of China. 2019. “PBC Holds Video Pritchett, L., and A. Beatty. 2015. “Slow Down, You’re Conference on Its Work in H2 2019.” www.pbc.gov.cn/ Piñeiro, R., M. Rhodes-Purdy and F. Rosenblatt. Going Too Fast: Matching Curricula to Student Skill english/130721/3872760/index.html. Accessed 6 October 2016. “The Engagement Curve: Populism and Political Levels.” International Journal of Educational Development 2019. Engagement in Latin America.” Latin American Research 40: 276­288. Review 51(4): 3­23. Permanyer, I., and N. Scholl. 2019. “Global Trends Pritchett, L., and J. Sandefur. 2017. “Girls’ Schooling and in Lifespan Inequality: 1950-2015.” PLOS ONE 14(5): Pistor, K. 2019. The Code of Capital: How the Law Creates Women’s Literacy: Schooling Targets Alone Won’t Reach e0215742. Wealth and Inequality. Princeton, NJ: Princeton University Learning Goals.” CGD Policy Paper 104. Center for Global Press. Development, Washington, DC. Permanyer, I., and J. Smits. 2019. “Uncovering Subnational Variation in Human Development around Pizer, J.C., B. Sears, C. Mallory and N.D. Hunter. Purdie-Vaughns, V., and R.P. Eibach. 2008. “Intersectional the World: 1990-2017.” Background paper for Human 2012. “Evidence of Persistent and Pervasive Workplace Invisibility: The Distinctive Advantages and Disadvantages Development Report 2019. United Nations Development Discrimination Against LGBT People: The Need for Federal of Multiple Subordinate-Group Identities.” Sex Roles Programme, Human Development Report Office, New York. Legislation Prohibiting Discrimination and Providing for 59(5­6): 377­391. Equal Employment Benefits.” Loyola of Los Angeles Law Pershing, A., N.R. Record, B.S. Franklin, B.T. Kennedy, Review 45(3): 715­780. PwC. 2017. “Sizing the Prize: What’s the Real Value of AI L. McClenachan, K.E. Mills, J.D. Scott, A.C. Thomas for Your Business and How Can You Capitalise?” www. and N.H. Wolff. 2019. “Challenges to Natural and Human Pla-Castells, M., J.J. Martinez-Durá, J.J. Samper- pwc.com/gx/en/issues/analytics/assets/pwc-ai-analysis- Communities from Surprising Ocean Temperatures.” Zapater and R.V. Cirilo-Gimeno. 2015. “Use of ICT sizing-the-prize-report.pdf. Accessed 13 August 2019. Proceedings of the National Academy of Sciences 116(37): in Smart Cities: A Practical Case Applied to Traffic 18378­18383. Management in the City of Valencia.” 2015 Smart Cities ------. n.d. “No Longer Science Fiction, AI and Robotics Symposium, Prague. Are Transforming Healthcare.” www.pwc.com/gx/en/in- Pew Research Center. 2014. “Emerging and Developing dustries/healthcare/publications/ai-robotics-new-health/ Economies Much More Optimistic than Rich Countries Pokhriyal, N., and Jacques, D.C. 2017. “Combining transforming-healthcare.html. Accessed 24 October 2019. about the Future.” Washington, DC. Disparate Data Sourced for Improved Poverty Prediction and Mapping.” Proceedings of the National Academy of Rains, E., A. Krishna and E. Wibbels. 2019. “Combining Phillips, D.A., and J.P. Shonkoff, eds. 2000. From Neurons Sciences 114(46): E9783­E9792. Satellite and Survey Data to Study Indian Slums: Evidence to Neighborhoods: The Science of Early Childhood on the Range of Conditions and Implications for Urban Development. Washington, DC: National Academies Polanyi, K. 1944. The Great Transformation. Boston, MA: Policy.” Environment and Urbanization 31(1): 267­292. Press. Beacon Press. Raising Voices, London School of Hygiene & Tropical Pickett, K.E., J. Mookherjee and R.G. Wilkinson. 2005. Pomeranz, K. 2000. The Great Divergence: China, Europe, Medicine and Center for Domestic Violence “Adolescent Birth Rates, Total Homicides, and Income and the Making of the Modern World Economy. Princeton, Prevention. 2015. Is Violence Against Women Inequality in Rich Countries.” American Journal of Public NJ: Princeton University Press. Preventable? Findings from the SASA! Study Summarized Health 95(7): 1181­1183. for General Audiences. Kampala: Raising Voices. Poore, J., and T. Nemecek. 2018. “Reducing Food’s Piketty, T. 1995. “Social Mobility and Redistributive Environmental Impacts through Producers and Rajan, R. G. 2011. Fault Lines: How Hidden Fractures Still Politics.” The Quarterly Journal of Economics 110(3): Consumers.” Science 360(6392): 987­992. Threaten the World Economy. Princeton, NJ: Princeton 551­584. University Press. Porter, E. 2019. “Don’t Fight the Robots. Tax Them.” New ------. 2001. Les hauts revenus en France au XXème siècle. York Times, 23 February. www.nytimes.com/2019/02/23/ Ramos, M.R., M.R Bennett, D.S Massey and M. Paris : Grasset. sunday-review/tax-artificial-intelligence.html. Accessed Hewstone. 2019. “Humans Adapt to Social Diversity over 19 August 2019. Time.” Proceedings of the National Academy of Sciences ------. 2003. “Income Inequality in France, 1901­1998.” 116(25) 12244­12249. Journal of Political Economy 111(5): 1004­1042. Porter, J.R., L. Xie, A.J. Challinor, K. Cochrane, S.M. Howden, M.M. Iqbal, D.B. Lobell and M.I. Travasso. Randell, H., and C. Gray. 2019. “Climate Change ------. 2006. “The Kuznets Curve: Yesterday and 2014. “Food Security and Food Production Systems.” and Educational Attainment in the Global Tropics.” Tomorrow.” In A.V. Banerjee, R. Benamou and D. In Field, C.B., V.R. Barros, D.J. Dokken, K.J. Mach, Proceedings of the National Academy of Sciences 116(18): Mookherjee, eds., Understanding Poverty. New York: M.D. Mastrandrea, T.E. Bilir, M. Chatterjee, K.L. Ebi, 8840-8845. Oxford University Press. Y.O. Estrada, R.C. Genova, B. Girma, E.S. Kissel, A.N. Levy, S. MacCracken, P.R. Mastrandrea and L.L. White, Rask, K.J., and N. Rask. 2011. “Economic Development and ------. 2014. Capital in the Twenty-First Century. eds. 2014. Climate Change 2014: Impacts, Adaptation, Food Production­Consumption Balance: A Growing Global Cambridge, MA: Harvard University Press. and Vulnerability. Part A: Global and Sectoral Aspects. Challenge.” Food Policy 36(2): 186­196. Contribution of Working Group II to the Fifth Assessment Piketty, T., and E. Saez. 2003. “Income Inequality in Report of the Intergovernmental Panel on Climate Change. Ravallion, M. 2001. “Growth, Inequality and Poverty: the United States, 1913­1998.” Quarterly Journal of Cambridge, UK: Cambridge University Press. www.ipcc.ch/ Looking Beyond Averages.” World Development 29(11): Economics 118(1): 1­41. site/assets/uploads/2018/02/WGIIAR5-PartA_FINAL.pdf. 1803­1815. Accessed 16 August 2019. Piketty, T., E. Saez and G. Zucman. 2018. “Distributional ------. 2015. “The Luxembourg Income Study.” Journal of National Accounts: Methods and Estimates for the United Prat, A. 2015. “Media Capture and Media Power.” In Economic Inequality 13: 527­547. States.” Quarterly Journal of Economics 133(2): 553­609. S.P. Anderson, J. Waldfogel and D. Stromberg, eds., Handbook of Media Economics, Vol. 1A. Amsterdam: ------. 2016. “Are the World’s Poorest Being Left Behind?” Piketty, T. L. Yang and G. Zucman. 2019. “Capital Journal of Economic Growth 21(2): 139­164. Accumulation, Private Property and Rising Inequality in China, 1978-2015.” American Economic Review 100(7): ------. 2018a. “Inequality and Globalization: A Review 2469­2496. Essay.” Journal of Economic Literature 56(2): 620­642.

References | 285 ------. 2018b. “What Might Explain Today’s Conflicting Robeyns, I. 2005. “The Capability Approach: A Theoretical Social Inequality Hypotheses.” New Media & Society Narratives on Global Inequality?” WIDER Working Paper Survey.” Journal of Human Development 6(1): 93­114. 21(2): 464­482. 2018/141. United Nations University­World Institute for Development Economics Research, Helsinki. ------. 2019. “What, If Anything, Is Wrong with Extreme Rosenfeld, M.J., R.J. Thomas and S. Hausen. 2019. Wealth?” Journal of Human Development and Capabilities “Disintermediating Your Friends: How Online Dating in Ravallion, M., M. Heil and J. Jalan. 2000. “Carbon 20(3): 251­266. https://doi.org/10.1080/19452829.2019.1 the United States Displaces Other Ways of Meeting.” Emissions and Income Inequality.” Oxford Economic 633734. Accessed 10 October 2019. Proceedings of the National Academy of Sciences 116(36): Papers 52(4): 651­669. 17753­17758. Roca, J. 2003. “Do Individual Preferences Explain the Rawls, J. 1971. A Theory of Justice. Cambridge, MA: Environmental Kuznet’s Curve?” Ecological Economics Rossiter, J., B. Hagos, P. Rose, T. Teferra and T. Harvard University Press. 45(1): 3­10. Woldehanna. 2018. Early Learning in Ethiopia: Equitable Access and Learning. System Diagnostic Report for World Reich, M.R., J. Harris, N. Ikegami, A. Maeda, C. Cashin, Rochet, J.-C., and J. Tirole. 2003. “Platform Competition Bank Early Learning Program. Cambridge, UK: University E.C. Araujo, K. Takemi and T.G. Evans. 2016. “Moving in Two-Sided Markets.” Journal of the European Economic of Cambridge, Research for Equitable Access and Learning towards Universal Health Coverage: Lessons from 11 Association 1(4): 990­1029. Centre. Country Studies.” Lancet 387(10020): 811­816. Rockoff, H. 2019. “On the Controversies Behind the Origins Rowe, D.C. 1994. The Limits of Family Influence: Genes, Reeves, R.V. 2018. Dream Hoarders: How the American of the Federal Economic Statistics.” Journal of Economic Experience, and Behavior. New York: Guilford Press. Upper Middle Class Is Leaving Everyone Else in the Perspectives 33(1): 147­164. Dust, Why That Is a Problem, and What To Do about It. Roy, J., P. Tschakert, H. Waisman, S. Abdul Halim, P. Washington, DC: Brookings Institution Press. Rockström, J., W. Steffen, K. Noone, A. Persso, F.S. Antwi-Agyei, P. Dasgupta, B. Hayward, M. Kanninen, Chapin III, E.F. Lambin, T.M. Lenton, M. Scheffer, C. D. Liverman, C. Okereke, P.F. Pinho, K. Riahi and A.G. Reinhart, C., and K. Rogoff. 2009. “The Aftermath of Folke, H.J. Schellnhuber and B. Nykvist. 2009. “A Suarez Rodriguez. 2019. “Sustainable Development, Financial Crises.” American Economic Review 99(2): Safe Operating Space for Humanity.” Nature 461(7263): Poverty Eradication and Reducing Inequalities.” In V. 466­472. 472­475. Masson-Delmotte, P. Zhai, H.-O. Pörtner, D. Roberts, J. Skea, P.R. Shukla, A. Pirani, W. Moufouma-Okia, C. Reinhart, R.J. 2018. “AI Seen as Greater Job Threat Than Rodriguez-Castelan, C., A. Araar, E. A. Malasquez, S. Péan, R. Pidcock, S. Connors, J.B.R. Matthews, Y. Chen, Immigration, Offshoring.” Gallup, 9 March. https://news. D. Olivieri and T. Vishwanath. 2019. “Distributional X. Zhou, M.I. Gomis, E. Lonnoy, T. Maycock, M. Tignor gallup.com/poll/228923/seen-greater-job-threat-immigra- Effects of Competition: A Simulation Approach.” Policy and T. Waterfield, eds., Global Warming of 1.5°C. An tion-offshoring.aspx. Accessed 18 October 2019. Research Working Paper 8838. World Bank, Washington, IPCC Special Report on the Impacts of Global Warming DC. of 1.5°C Above Pre-Industrial Levels and Related Global Republic of South Africa. 1996. The Constitution of the Greenhouse Gas Emission Pathways, in the Context Republic of South Africa. Chapter 2, 27. www.justice.gov. Rodrik, D. 2007. One Economics, Many Recipes: of Strengthening the Global Response to the Threat of za/legislation/constitution/SAConstitution-web-eng.pdf. Globalization, Institutions, and Economic Growth. Climate Change, Sustainable Development, and Efforts to Accessed 15 August 2019. Princeton, NJ: Princeton University Press. Eradicate Poverty. Cambridge, UK: Cambridge University Press. www.ipcc.ch/site/assets/uploads/sites/2/2019/02/ Riahi, K., D.P. Van Vuuren, E. Kriegler, J. Edmonds, B.C. ------. 2015. “Premature Deindustrialization.” NBER SR15_Chapter5_Low_Res.pdf. Accessed 16 August 2019. O’Neill, S. Fujimori, N. Bauer, K. Calvin, R. Dellink, O. Working Paper 20935. National Bureau of Economic Fricko and W. Lutz. 2017. “The Shared Socioeconomic Research, Cambridge, MA. Rozenberg, J., and S. Hallegatte. 2015. “The Impacts Pathways and their Energy, Land Use, and Greenhouse of Climate Change on Poverty in 2030 and the Gas Emissions Implications: An Overview.” Global ------. 2018. “Populism and the Economics of Potential from Rapid, Inclusive, and Climate-Informed Environmental Change 42: 153­168. Globalization.” Journal of International Business Policy Development.” Policy Research Working Paper 7483. 1­2: 12­33. World Bank, Washington, DC. https://openknowledge. Ricardo, C., and MenEngage. 2014. Men, Masculinities, worldbank.org/bitstream/handle/10986/23447/ and Changing Power: A Discussion Paper on Engaging ------. 2019. “Policy, Inequality and Growth.” In O. The0impacts0of0informed0development.pdf?sequence=1. Men in Gender Equality from Beijing 1995 to 2015. Blanchard and L.H. Summers, eds., Evolution or Accessed 13 August 2019. Washington, DC: MenEngage. Revolution? Rethinking Macroeconomic Policy after the Great Recession. Cambridge, MA: MIT Press. Rözer, J., and H. Van De Werfhorst. 2017. “Inequalities Rice, D. 2019. “Dorian’s Legacy: The Slowest, Strongest in Educational Opportunities by Socioeconomic and Hurricane to Ever Hit the Bahamas.” USA Today, 6 Roemer, J.E. 1998. Equality of Opportunity. Cambridge, MA: Migration Background: A Comparative Assessment September. Harvard University Press. across European Societies.” ISOTIS Report. University of Amsterdam. Richardson, R., N. Schmitz, S. Harper and A. Nandi. Rohland, E. 2018. “Adapting to Hurricanes: A Historical 2019. “Development of a Tool to Measure Women’s Perspective on New Orleans from Its Foundation to Russell, S. 2018. “How to Make AI That Works, for Us.” Agency in India.” Journal of Human Development and Hurricane Katrina, 1718­2005.” WIREs Climate Change Science Focus, 16 November. www.sciencefocus.com/ Capabilities 20(1): 26­53. 9(1): e488. future-technology/how-to-make-ai-that-works-for-us/. Accessed 6 October 2019. Richey, A.S., B.F. Thomas, M.H. Lo, J.T. Reager, J.S. Romer, P. 1990. “Endogenous Technological Change.” Famiglietti, K. Voss, S. Swenson and M. Rodell. Journal of Political Economy 98(5): S71­S102. Rutgers. 2014. “Burundi Embraces Comprehensive Sexuality 2015. “Quantifying Renewable Groundwater Stress with Education.” 17 December. www.rutgers.international/ GRACE.” Water Resources Research 51(7): 5217­5238. ------. 2019. “A Tax That Could Fix Big Tech.” New York news-opinion/news-archive/burundi-embraces-compre- Times, 6 May. www.nytimes.com/2019/05/06/opinion/ hensive-sexuality-education. Accessed 9 October 2019. Ridder, G., and G. van den Berg. 2003. “Measuring Labor tax-facebook-google.html. Accessed 19 August 2019. Market Frictions: A Cross-Country Comparison.” Journal of Rutkowski, M. 2018. “Reimagining Social Protection.” the European Economic Association 1(1): 224­244. Rose, A., D. Wei and A. Bento. 2019. “Equity Implications Finance and Development 55(4). of the COP21 Intended Nationally Determined Riley, R., and C.R. Bondibene. 2017. “Raising the Contributions to Reduce Greenhouse Gas Emissions.” In Saad, L. 2019. “Americans as Concerned as Ever About Standard: Minimum Wages and Firm Productivity.” Labour R. Kanbur and H. Shue, eds., Climate Justice: Integrating Global Warming.” Gallup, 25 March. https://news.gallup. Economics 44: 27­50. Economics and Philosophy. Oxford, UK: Oxford University com/poll/248027/americans-concerned-ever-global- Press. warming.aspx. Accessed 18 October 2019. Ritchie, H. 2019. “12 Key Metrics to Understand the State of the World.” https://ourworldindata.org/12-key-metrics. Rose, T. 2016. The End of Average. New York: Harper Collins. Saez, E., and G. Zucman 2016. “Wealth Inequality in the Accessed 10 October 2019. United States since 1913: Evidence from Capitalized Rosen, R.A. 2019. “Temperature Impact on GDP Growth Is Income Tax Data.” Quarterly Journal of Economics 131(2): Ritchie, H., and M. Roser. 2018. “CO2 and Greenhouse Gas Overestimated.” Proceedings of National Academies of 519­578. Emissions.” https://ourworldindata.org/co2-and-other- Sciences 116(33): 16170. greenhouse-gas-emissions. Accessed 9 August 2019. Rosenberg, D. 2019. “Use of E-Government Services in a Deeply Divided Society: A Test and an Extension of the

286 | HUMAN DEVELOPMENT REPORT 2019 ------. 2019. “How Would a Progressive Wealth Tax Work? Sears, B., and C. Mallory. 2011. “Documented Evidence Shanmugaratnam, T. 2019. “Absolute Mobility Matters, Evidence from the Economics Literature.” Unpublished of Employment Discrimination & Its Effects on LGBT Too: Regenerating People and Cities.” In O. Blanchard and manuscript. People.” The Williams Institute, Los Angeles, CA. https:// L.H. Summers, eds., Evolution or Revolution? Rethinking williamsinstitute.law.ucla.edu/wp-content/uploads/ Macroeconomic Policy after the Great Recession. Sager, L. 2017. “Income Inequality and Carbon Consumption: Sears-Mallory-Discrimination-July-20111.pdf. Accessed Cambridge, MA: MIT Press. Evidence from Environmental Engel Curves.” GRI Working 26 July 2019. Paper 285. Grantham Research Institute on Climate Shapiro, C. 2018. “Antitrust in a Time of Populism.” Change and the Environment, London. Seebens, H., F. Essl, W. Dawson, N. Fuentes, D. Moser, International Journal of Industrial Organization J. Pergl, P. Pysek, M. van Kleunen, E. Weber, M. 61:714­748. Saguy, T. 2018. “Downside of Intergroup Harmony? When Winter and B. Blasius. 2015. “Global Trade Will Reconciliation Might Backfire and What to Do.” Policy Accelerate Plant Invasions in Emerging Economies Shaxton, N. 2019. “Tackling Tax Havens.” IMF Finance and Insights from the Behavioral and Brain Sciences 5: 75­81. under Climate Change.” Global Change Biology 21(11): Development Magazine, September 2019. www.imf. 4128­4140. org/external/pubs/ft/fandd/2019/09/tackling-global-tax- Saikia, N., J.K. Bora and M. Luy. 2019. “Socioeconomic havens-shaxon.htm. Accessed 7 November 2019. Disparity in Adult Mortality in India: Estimations using the Seligman, B., G. Greenberg and S. Tuljapurkar. 2016. Orphanhood Method.” Genus 75(1): 7. “Equity and Length of Lifespan Are Not the Same.” Shorrocks, A.F., and J.E. Foster. 1987. “Transfer Sensitive Proceedings of the National Academy of Sciences 113(30): Inequality Measures.” Review of Economic Studies 54(3): Salverda, W., and D. Checchi. 2015. “Labor Market 8420­8423. 485­497. Institutions and the Dispersion of Wage Earnings.” In A.B. Atkinson and F. Bourguignon, eds., Handbook of Income Semple, K. 2019. “Corpses Strewn, People Missing a SIDA (Swedish International Development Cooperation Distribution, Vol. 2. New York: Elsevier. Week After Dorian Hit the Bahamas.” New York Times, 8 Agency). 2015. “Gender Analysis Principles and September. Elements.” Gender Tool Box. Stockholm. Sanyal, S. 2018. “How Is AI Revolutionizing Elderly Care.” Forbes, 31 October. Sen, A. 1980. “Equality of What?” In S. McMurrin, ed., Sierminska, E.M., J.R. Frick and M.M. Grabka. 2010. Tanner Lectures on Human Values. Vol. I. Cambridge, UK: “Examining the Gender Wealth Gap.” Oxford Economic Sardinha, L., and H.E. Catalan. 2018. “Attitudes towards Cambridge University Press. Papers 62(4): 669­690. Domestic Violence in 49 Low- and Middle-Income Countries: A Gendered Analysis of Prevalence and ------. 1982. “Food Battles: Conflict in the Access to Food.” Silcoff, M. 2018. “`The Daddy Quota’: How Quebec Got Men Country-Level Correlates.” PLoS One 13(10): e0206101. Coromandel Lecture, 13 December. to Take Parental Leave.” The Guardian, 15 June. www. theguardian.com/world/2018/jun/15/the-daddy-quota- Scheidel, W. 2017. The Great Leveler: Violence and the ------. 1985. Commodities and Capabilities. Amsterdam: how-quebec-got-men-to-take-parental-leave. Accessed 9 Global History of Inequality from the Stone Age to the North-Holland. October 2019. Present. Princeton, NJ: Princeton University Press. ------. 1990. “More than 100 Million Women Are Silver, D., T. Hubert, J. Schrittwieser, I. Antonoglou, Schelling, T. 1978. Micromotives and Macrobehavior. New Missing.” The New York Review of Books 37(20): 61­66. M. Lai, A. Guez, M. Lanctot, L. Sifre, D. Kumaran, York: Norton. T. Graepel and T. Lillicrap. 2018. “A General ------. 1992. Inequality Reexamined. Cambridge, MA: Reinforcement Learning Algorithm that Masters Chess, Schiappacasse, I. 2019. “Business Elites and the Harvard University Press. Shogi, and Go through Self-Play.” Science 362(6419): Development of Social Policy in Latin America: The 1140­1144. Case of Pensions in Chile”. Unpublished PhD thesis at ------. 1993. “Capability and Well-Being.” In M. Nussbaum the University of Oxford, Department of International and A. Sen, The Quality of Life. Oxford, UK: Oxford Sim, A., E. Lim, C. Forde and B. Cheon. 2018. “Personal Development, Oxford, UK. University Press. Relative Deprivation Increases Self-Selected Portion Sizes and Food Intake.” Appetite 121: 268­274. Schiermeier, Q. 2018. “Telltale Warming Likely to Hit Poorer ------. 1999. Development as Freedom. Oxford, UK: Oxford Countries First.” Nature 556: 415­416. University Press. Skelton, J.A., M.B. Irby, J.G. Grzywacz and G. Miller. 2011. “Etiologies of Obesity in Children: Nature and Schiffrin, A. 2019. “The Contribution of Free Media to the ------. 2005. “Human Rights and Capabilities.” Journal of Nurture.” Pediatric Clinics 58(6): 1333­1354. Fight against Corruption and to Human Development.” Human Development 6(2): 151­166. Background paper for Human Development Report Skopek, J., and G. Passaretta. 2018. “The Social 2019. United Nations Development Programme, Human ------. 2008a. “The Economics of Happiness and Stratification of Skills from Infancy to Adolescence­ Development Report Office, New York. Capability.” In L. Bruni, F. Comim and M. Pugno, eds., Evidence from an Accelerated Longitudinal Design.” Trinity Capabilities and Happiness. Oxford, UK: Oxford University College Dublin, Department of Sociology, Dublin. https:// Schleussner, C.-F., J.F. Donges, R.V. Donner and H.J. Press. osf.io/preprints/socarxiv/xkctv/. Accessed 8 August 2019. Schellnhuber. 2016. “Armed-Conflict Risks Enhanced by Climate-Related Disasters in Ethnically Fractionalized ------. 2008b. “Violence, Identity and Poverty.” Journal of Smith, H. J., T.F. Pettigrew, G.M. Pippin and S. Countries.” Proceedings of the National Academy of Peace Research 45(1): 5­15. Bialosiewicz. 2012. “Relative Deprivation: A Theoretical Sciences 113(33): 9216­9221. and Meta-Analytic Review.” Personality and Social Sen, G., P. Ostlin and A. George. 2007. “Unequal, Unfair, Psychology Review 16(3): 203­232. Schlozman, K.L. 2012. The Unheavenly Chorus: Unequal Ineffective and Inefficient Gender Inequity in Health: Why Political Voice and the Broken Promise of American It Exists and How We Can Change It.” Final Report to Smith, M., D. Yagan, O. Zidar and E. Zwick. 2019. Democracy. Princeton, NJ: Princeton University Press. the WHO Commission on Social Determinants of Health. “Capitalists in the Twenty-First Century.” NBER Working Women and Gender Equity Knowledge Network, Geneva. Paper 25442. National Bureau of Economic Research, Schmidt, L., and P. Sevak. 2006. “Gender, Marriage, and Cambridge, MA. Asset Accumulation in the United States.” Feminist Seriño, M.N.V., and S. Klasen. 2015. “Estimation and Economics 12(1­2): 139­166. Determinants of the Philippines Household Carbon Sobotta, R.R., H.E. Campbell and B.J. Owens. 2007. Footprint.” Developing Economies 53(1): 44­62. “Aviation Noise and Environmental Justice: The Barrio Schwellnus, C., A. Kappeler and P.-A. Pionnier. 2017. Barrier.” Journal of Regional Science 47(1): 125­154. “The Decoupling of Median Wages from Productivity in Shackelford, A.H. 2018. “The Why and the How of Central OECD Countries.” International Productivity Monitor 32: America’s First All Female Hackathon.” 28 March. https:// Soens, T. 2011. “Floods and Money: Funding Drainage and 44­60. undg.org/silofighters_blog/the-why-and-the-how-of- Flood Control in Coastal Flanders from the Thirteenth to central-americas-first-all-female-hackathon/. Accessed 9 the Sixteenth Centuries.” Continuity and Change 26(3): Science Daily. 2014. “A More Potent Greenhouse Gas October 2019. 333­365. than Carbon Dioxide, Methane Emissions Will Leap as Earth Warms.” 27 March. www.sciencedaily.com/re- Shankland, S. 2019. “Bill Gates Says AI Should Improve ------. 2018. “Resilient Societies, Vulnerable People: leases/2014/03/140327111724.htm. Accessed 15 August Education and Medicine.” CNET, 18 March. www.cnet. Coping with North Sea Floods before 1800.” Past & 2019. com/news/bill-gates-says-ai-should-improve-education- Present 241: 143­177. and-medicine/. Accessed 24 October 2019. Scott, J.C. 2017. Against the Grain: A Deep History of the Earliest States. New Haven, CT: Yale University Press.

References | 287 Solon, G. 1999. “Intergenerational Mobility in the Labor Stern, N., and J.E. Stiglitz. 2017. Report of the High-Level com/2016/04/11/want-more-panama-papers-heres-how/. Market.” In O. Ashenfelter and D. Card, eds., Handbook of Commission on Carbon Prices. Washington, DC: World Accessed 1 October 2019. Labor Economics, Vol. III. Amsterdam: Elsevier. Bank. Sunstein, C. 2018. “A New View of Antitrust Law That ------. 2018. “What Do We Know So Far about Stewart, F. 2005. “Horizontal Inequalities: A Neglected Favors Workers.” Bloomberg, 14 May. www.bloomberg. Multigenerational Mobility?” Economic Journal 128(612): Dimension of Development.” In UNU-WIDER, A.B. com/opinion/articles/2018-05-14/antitrust-law-gets-a- F340­F352. Atkinson, B. Kaushik, J.N. Bhagwati, D.C. North, D. chicago-school-makeover. Accessed 8 October 2018. Rodrik, F. Stewart, J.E. Stiglitz, J.G. Williamson, eds., Solt, F. 2008. “Economic Inequality and Democratic Political Wider Perspectives on Global Development. London, UK: Surminski, S., L. M. Bouwer and J. Linnerooth-Bayer. Engagement.” American Journal of Political Science 52(1): Palgrave Macmillan. 2016. “How Insurance can Support Climate Resilience.” 48­60. Nature Climate Change 6(4): 333­334. ------. 2006. “Social Exclusion and Conflict: Analysis Sommer, I., U. Griebler, P. Mahlknecht, K. Thaler, and Policy Implications.” Oxford Centre for Research on Sutherland, H., and F. Figari. 2013. “EUROMOD: The K. Bouskill, G. Gartlehner and S. Mendis. 2015. Inequality, Human Security and Ethnicity, Oxford, UK. European Union Tax-Benefit Microsimulation Model.” “Socioeconomic Inequalities in Non-Communicable www.qeh.ox.ac.uk/pdf/pdf-research/crise-pp1. Accessed International Journal of Microsimulation 6(1): 4­26. Diseases and Their Risk Factors: An Overview of 9 October 2019. Systematic Reviews.” BMC Public Health 15(1): 914. Suzuki, E., S. Kashima, I. Kawachi and S. Subramanian. ------. 2009. “Horizontal Inequalities as a Cause of 2012. “Social and Geographic Inequalities in Premature Sood, S., T. Menard and K. Witte. 2009. “The Theory Conflict.” Bradford Development Lecture, 21 October. Adult Mortality in Japan: A Multilevel Observational Study behind Entertainment-Education.” In A. Singhal, M. Cody, University of Bradford, West Yorkshire, UK. from 1970 to 2005.” BMJ Open 2(2): e000425. E. Rogers and M. Sabido, eds., Entertainment-Education and Social Change: History, Research, and Practice. ------. 2016a. “The Dynamics of Horizontal Inequalities.” Sy, A.N., R. Maino, A. Massara, H. Perez-Saiz and London: Routledge. Think Piece for Human Development Report 2016. United P. Sharma. 2019. “FinTech in Sub-Saharan African Nations Development Programme, Human Development Countries. A Game Changer?” IMF Departmental Paper South Africa Department of Water and Sanitation. 2016. Report Office, New York. http://hdr.undp.org/sites/default/ 19/04. International Monetary Fund, Washington, DC. “National Sanitation Policy.” Pretoria. www.dwa.gov.za/ files/stewart_layout.pdf. Access 6 August 2019. Documents/sanitation/17005SC_POLICY_National%20 Szwarcwald, C.L., P.R.B. de Souza Júnior, A.P. Sanitation%20Policy%202016%20FINAL310117.pdf. ------, ed. 2016b. Horizontal Inequalities and Conflict: Marques, W.d.S. de Almeida and D.E.R. Montilla. Accessed 15 August 2019. Understanding Group Violence in Multiethnic Societies. 2016. “Inequalities in Healthy Life Expectancy by Brazilian New York: Springer. Geographic Regions: Findings from the National Health Spice, B. 2015. “Questioning the Fairness of Targeting Ads Survey, 2013.” International Journal for Equity in Health Online.” Carnegie Mellon University News, 7 July. www. Stewart, F., G. Ranis and E. Samman. 2018. Advancing 15(1): 141. cmu.edu/news/stories/archives/2015/july/online-ads- Human Development: Theory and Practice. Oxford, UK: research.html. Accessed 24 October 2019. Oxford University Press. Szymkowski, S. 2019. “California Bill Passes to Classify Lyft, Uber Drivers as Employees.” Roadshow, 11 Springmann, M., H.C.J. Godfray, M. Rayner and P. Stiglitz, J.E. 2012. The Price of Inequality: How Today’s September. www.Cnet.Com/Roadshow/News/California- Scarborough. 2016. “Analysis and Valuation of the Divided Society Endangers our Future. New York: WW Ab5-Uber-Lyft-Employees/. Accessed 24 October 2019. Health and Climate Change Cobenefits of Dietary Norton & Company. Change.” Proceedings of the National Academy of Takacs, D. 2016. “South Africa and the Human Right to Sciences 113(15): 4146­4151. ------. 2016. “Inequality and Economic Growth.” In M. Water: Equity, Ecology and the Public Trust Doctrine.” Mazzucato and M. Jacobs, eds., Rethinking Capitalism. Berkeley Journal of International Law 34(2): 55­108. Spruyt, B., and T. Kuppens. 2015. “Education-Based Hoboken, NJ: Wiley-Blackwell. Thinking and Acting? Towards an Identity Perspective for Takasu, Y. 2019. “SDGs and Japan: Human Security Studying Education Differentials in Public Opinion and ------. 2019a. “Addressing Climate Change through Price Indicators (HSI) of Japan.” Unpublished manuscript. Political Participation.” European Journal of Cultural and and Non-Price Interventions.” NBER Working Paper 25939. Political Sociology 2: 291­312. National Bureau of Economic Research, Cambridge, MA. Tankersley, J., and A. Rappeport. 2019. “As Nations Look to Tax Tech Firms, U.S. Scrambles to Broker a Deal.” New Spruyt, B., T. Kuppens, R. Spears and J. van Noord. ------. 2019b. “Market Concentration Is Threatening York Times, 12 July. www.nytimes.com/2019/07/12/busi- Forthcoming. “Talking Politics? Educational Category the US Economy.” Project Syndicate, 11 March. www. ness/economy/tech-company-taxes.html. Accessed 24 Salience Reinforces Differences in People’s Willingness project-syndicate.org/commentary/united-states- October 2019. to Participate in Deliberative Initiatives.” Political economy-rising-market-power-by-joseph-e-stiglitz- Psychology. 2019-03?barrier=accesspaylog. Accessed 8 October 2019. Tavoni, A., A. Dannenberg, G. Kallis and A. Loschel. 2011. “Inequality, Communication, and the Avoidance Staiger, D.O., J. Spetz and C.S. Phibbs. 2010. “Is There Stiglitz, J.E., and B.C. Greenwald. 2014. Creating of Disastrous Climate Change in a Public Goods Game.” Monopsony in the Labor Market? Evidence from a Natural a Learning Society: A New Approach to Growth, Proceedings of the National Academy of Sciences 108(29): Experiment.” Journal of Labor Economics 28(2): 211­236. Development, and Social Progress. New York: Columbia 11825­11829. University Press. State of California. 2012. Assembly Bill No. 685 ­ Chapter Taylor, D.E. 2000. “The Rise of the Environmental Justice 524. https://leginfo.legislature.ca.gov/faces/billNavClient. Stiglitz, J., A. Sen and J.-P. Fitoussi. 2009a. “The Paradigm: Injustice Framing and the Social Construction of xhtml?bill_id=201120120AB685 Accessed 10 October Measurement of Economic Performance and Social Environmental Discourses.” American Behavioral Scientist 2019. Progress Revisited: Reflections and Overview.” 43: 508­580. Commission on the Measurement of Economic Steffen, W., K. Richardson, J. Rockström, S.E. Cornell, Performance and Social Progress, Paris. Terdiman, D. 2017. “How Robots And AI Could Save I. Fetzer, E.M. Bennett, R. Biggs, S.R. Carpenter, W. American Water Utilities Half A Trillion Dollars.” Fast De Vries, C.A. De Wit and C. Folke. 2015. “Planetary ------. 2009b. “Report by the Commission on the Company, 24 February. www.fastcompany.com/3068423/ Boundaries: Guiding Human Development on a Changing Measurement of Economic Performance and Social how-robots-and-ai-could-save-american-water-utilities- Planet.” Science 347(6223). Progress.” Paris. half-a-trillion-dollar. Accessed 15 August 2019.

Steger, T., A. Antypas, L. Atkins, F. Borthwick and Stokes, S. 2009. “Political Clientelism.” In C. Boix and Thakor, A.V. 2012. “Incentives to Innovate and Financial C. Cahn. 2007. “Making the Case for Environmental S. Stokes, eds., The Oxford Handbook of Comparative Crises.” Journal of Financial Economics 103(1): 130­148. Justice in Central and Eastern Europe.” CEU Center for Politics. Oxford, UK: Oxford University Press. Environmental Law and Policy, Budapest. www.env-health. Thévenon, O. 2013. “Drivers of Female Labour Force org/IMG/pdf/Making_the_case_for_environmental_jus- Stone, L. 2015. “Women Transforming Conflict: A Participation in the OECD.” OECD Social, Employment and tice.pdf. Accessed 15 August 2019. Quantitative Analysis of Female Peacemaking.” Working Migration Working Paper 145. Organisation for Economic Paper. Seton Hall University, South Orange, NJ. Co-operation and Development, Paris.

Sullivan, D. 2016. “Want More `Panama Papers’? Here’s Thornton, P.K., P.G. Jones, T. Owiyo, R.L. Kruska, M.T. How.” Foreign Policy, 11 April 2016. https://foreignpolicy. Herrero, P.M. Kristjanson, A.M.O. Notenbaert,

288 | HUMAN DEVELOPMENT REPORT 2019 N. Bekele and A. Omolo. 2006. “Mapping Climate government/case-studies/commonsensing. Accessed 13 hdr_20072008_en_complete.pdf. Accessed 8 August Vulnerability and Poverty in Africa.” Report to the August 2019. 2019. Department for International Development. Nairobi. https://cgspace.cgiar.org/bitstream/handle/10568/2307/ UN (United Nations). 2009. System of National Accounts ------. 2011. Human Development Report 2011: Mapping_Vuln_Africa.pdf?sequence=1%26isAllowed=y. 2008. New York. Sustainability and Equity: A Better Future for All. New Accessed 15 August 2019. York. http://hdr.undp.org/sites/default/files/reports/271/ ------. 2015a. The Millennium Development Goals Report. hdr_2011_en_complete.pdf. Accessed 13 August 2019. Tigchelaar, M., D.S. Battisti, R.L. Naylor and D.K. Ray. New York. 2018. “Future Warming Increases Probability of Globally ------. 2015. Human Development Report 2015: Work for Synchronized Maize Production Shocks.” Proceedings of ------. 2015b. Paris Agreement. https://unfccc.int/sites/ Human Development. New York. the National Academy of Sciences 115(26): 6644­6649. default/files/english_paris_agreement.pdf. Accessed 14 October 2019. ------. 2016. Human Development Report 2016: Human Tilman, D., and M. Clark. 2014. “Global Diets Link Development for Everyone. New York. http://hdr.undp.org/ Environmental Sustainability and Human Health.” Nature ------. 2015c. Transforming Our World: The 2030 Agenda sites/default/files/2016_human_development_report.pdf. 515(7528): 518­522. for Sustainable Development. New York. Accessed 5 August 2019.

Timmer, M.P., A. Erumban, B. Los, R. Stehrer and G. De ------. 2019a. The Age of Interdependence. Report of the ------. 2018a. Human Development Indices and Indicators: Vries. 2014. “Slicing Up Global Value Chains.” Journal of High-Level Panel on Digital Cooperation. New York. Statistical Update 2018. New York. Economic Perspectives 28(2): 99­118. ------. 2019b. Global Sustainable Development Report: ------. 2018b. “What Does It Mean to Leave No One Tinbergen, J. 1974. “Substitution of Graduate by Other The Future is Now: Science for Achieving Sustainable Behind? A UNDP Discussion Paper and Framework for Labour.” Kyklos 27(2): 217­226. Development. New York: United Nations. Implementation.” New York. www.undp.org/content/ undp/en/home/librarypage/poverty-reduction/what ------. 1975. “Substitution of Academically Trained by UN (United Nations) and World Bank. 2018. Pathways for -does-it-mean-to-leave-no-one-behind-.html. Accessed 14 Other Manpower.” Review of World Economics 111(3): Peace: Inclusive Approaches to Preventing Violent Conflict. October 2019. 466­476. Washington, DC. https://openknowledge.worldbank.org/ handle/10986/28337. Accessed 6 August 2019. UNDP (United Nations Development Programme) Chile. Tödtling, F., and M. Trippl. 2005. “One Size Fits All? 2017. Desiguales: Orígenes, cambios y desafíos de la Towards a Differentiated Regional Innovation Policy UNCDF (United Nations Capital Development Fund). brecha social en Chile. Santiago. Approach.” Research Policy 34(8): 1203­1219. 2019. “Financial Inclusion.” www.uncdf.org/financial- inclusion. Accessed 13 August 2019. UNDP (United Nations Development Programme) and Tol, R.S.J. 2018. “The Economic Impacts of Climate OPHI (Oxford Poverty and Human Development Change.” Review of Environmental Economics and Policy UN CEB (United Nations System Chief Executives Initiative). 2019. Global Multidimensional Poverty Index 12(1): 4­25. Board for Coordination). 2017. Leaving No One 2019: Illuminating Inequalities. New York and Oxford, UK. Behind: Equality and Non-Discrimination at the Heart of http://hdr.undp.org/sites/default/files/mpi_2019_publica- Torche, F. 2010. “Educational Assortative Mating and Sustainable Development. New York. tion.pdf. Accessed 9 September 2019. Economic Inequality: A Comparative Analysis of Three Latin American Countries.” Demography 47(2): 481­502. UNCTAD (United Nations Conference on Trade and UNDP (United Nations Development Programme) Development). 2019. Trade and Development Report and UN Women (United Nations Entity for Gender Torre, R., and M. Myrskylä. 2014. “Income Inequality 2019: Financing A Global Green New Deal. Geneva. Equality and the Empowerment of Women). 2019. and Population Health: An Analysis of Panel Data for 21 “Gender Equality as an Accelerator for Achieving the Developed Countries, 1975­2006.” Population Studies UNDESA (United Nations Department of Economic SDGs.” New York. 68(1): 1­13. and Social Affairs). 2009. Creating an Inclusive Society: Practical Strategies to Promote Social Integration. New UNDRR (United Nations Office for Disaster Risk Tørsløv, T.R., L.S. Wier and G. Zucman. 2018. “The York. www.un.org/esa/socdev/egms/docs/2009/Ghana/ Reduction). 2019. Global Assessment Report on Disaster Missing Profits of Nations.” NBER Working Paper 24701. inclusive-society.pdf. Accessed 10 October 2019. Risk Reduction 2019. Geneva. National Bureau of Economic Research, Cambridge, MA. ------. 2016. World Economic and Social Survey 2016: UNESCO (United Nations Educational, Scientific and Tschakert, P. 2016. “The Role of Inequality in Climate- Climate Change Resilience—An Opportunity for Reducing Cultural Organization). 2018a. Data Centre. http://uis. Poverty Debates.” Policy Research Working Paper 7677. Inequalities. New York. www.un.org/development/ unesco.org/. Accessed 2 October 2019. World Bank, Washington, DC. https://openknowledge. desa/dpad/wp-content/uploads/sites/45/publication/ worldbank.org/bitstream/handle/10986/24515/ WESS_2016_Report.pdf. Accessed 16 August 2019. ------. 2018b. “One in Five Children, Adolescents and The0role0of0in0mate0poverty0debates.pdf;sequence=1. Youth Is Out of School.” Paris. http://uis.unesco.org/ Accessed 16 August 2019. ------. 2018. World Economic and Social Survey 2018: sites/default/files/documents/fs48-one-five-children- Frontier Technologies for Sustainable Development. New adolescents-youth-out-school-2018-en.pdf. Accessed 10 Tubiello, F.N., M. Salvatore, R.D. Cóndor Golec and A. York. October 2019. Ferrara. 2014. “Agriculture, Forestry and Other Land Use Emissions by Sources and Removals by Sinks: 1990-2011 ------. 2019. World Population Prospects 2019. New York. ------. 2019a. Global Education Monitoring Report 2019: Analysis.” Food and Agriculture Organization of the United Gender Report: Building Bridges for Gender Equality. Paris. Nations, Statistics Division, Rome. UNDP (United Nations Development Programme). 1995. Human Development Report 1995: Gender and Human ------. 2019b. Meeting Commitments: Are Countries on Turchin, P., and S.A. Nefedov. 2009. Secular Cycles. Development. New York. Track To Achieve SDG4? Montreal. Princeton, NJ: Princeton University Press. ------. 2001. Human Development Report 2001: Making ------. n.d. “Girls’ and Women’s Education in Science, Uchino, B.N. 2006. “Social Support and Health: A Review New Technologies Work for Human Development. New Technology, Engineering and Mathematics (STEM).” Paris. of Physiological Processes Potentially Underlying Links to York. http://hdr.undp.org/sites/default/files/reports/262/ https://en.unesco.org/themes/education-and-gender- Disease Outcomes.” Journal of Behavioral Medicine 29(4): hdr_2001_en.pdf. Accessed 6 October 2019. equality/stem. Accessed 24 October 2019. 377­387. ------. 2006. Human Development Report 2006: Beyond UNFCCC (United Nations Framework Convention on UK Department for Digital, Culture, Media and Sport. Scarcity: Power, Poverty and the Global Water Crisis. New Climate Change). 2015. “G7 Climate Risk Insurance 2018. “Guidance. Data Ethics Framework.” www.gov.uk/ York: Palgrave Macmillan. www.undp.org/content/dam/ Initiative—Stepping Up Protection for the Most government/publications/data-ethics-framework/data- undp/library/corporate/HDR/2006%20Global%20HDR/ Vulnerable.” Paris. https://unfccc.int/news/g7-climate- ethics-framework. Accessed 19 August 2019. HDR-2006-Beyond%20scarcity-Power-poverty-and-the- risk-insurance-initiative-stepping-up-protection-for-the- global-water-crisis.pdf. Accessed 15 August 2019. most-vulnerable. Accessed 13 August 2019. UK Space Agency. 2018. “Case Study: CommonSensing. Fostering Climate Resilience for Small Island Nations ------. 2007. Human Development Report 2007/2008: UNFPA (United Nations Population Fund). 2016. Using Remote Sensing.” London. www.gov.uk/ Fighting Climate Change: Human Solidarity in a Universal Access to Reproductive Health: Progress and Divided World. New York: Palgrave Macmillan. Challenges. New York. http://hdr.undp.org/sites/default/files/reports/268/

References | 289 ------. 2019. “In Burundi, Sexual Health Education Helps United Nations Human Rights Council. 2010. “The Uslaner, E.M., and M. Brown. 2005. “Inequality, Trust, Youth Protect Themselves, Their Futures.” 18 January. Human Right to Safe Drinking Water and Sanitation.” and Civic Engagement.” American Politics Research 33(6): Resolution A/HRC/RES/18/1. New York. www.right-docs. 868­894. UN General Assembly. 2006. “In-depth Study on All Forms org/doc/a-hrc-res-18-1/. Accessed 15 August 2019. of Violence against Women: Report of the Secretary Uthman, O., S. Lawoko and T. Moradi. 2011. “Factors General.” A/61/122/Add.1. New York. UN News. 2019. “Despite Progress, Companies Face Gender Associated with Attitudes towards Intimate Partner Equality `Backlash’: UN Business Body.” 18 March. Violence against Women.” In S.D. Palmer, ed., Social ------. 2010. “The Human Right to Water and Sanitation.” Work and Community Practice. New York: Taylor & Francis. Resolution 64/292. New York. www.un.org/en/ga/search/ UNSDG (United Nations Sustainable Development view_doc.asp?symbol=A/RES/64/292. Accessed 15 Group). 2019. “Leaving No One Behind: A UNSDG Utterback, J.M., and W.J. Abernathy. 1975. “A Dynamic August 2019. Operational Guide for UN Country Teams.” Interim Draft, Model of Process and Product Innovation.” Omega 3(6): March 2019. New York. 639­656. ------. 2016. “Promotion and Protection of All Human Rights, Civil, Political, Economic, Social and Cultural UN Women (United Nations Entity for Gender Equality van Bavel, B. 2016. The Invisible Hand? How Market Rights, Including the Right to Development.” New York and the Empowerment of Women). 1995. Beijing Economies have Emerged and Declined since AD 500. www.article19.org/data/files/Internet_Statement_ Declaration and Platform for Action: Beijing+ 5 Political Oxford, UK: Oxford University Press. Adopted.pdf. Accessed 14 October 2019. Declaration and Outcome. Adopted at the Fourth World Conference on Women, 4­15 September, Beijing. ------. 2019. “Open Societies before Market Economies: Unger, R.M. 2019. The Knowledge Economy. London: Verso Historical Analysis.” Socio-Economic Review, mwz007. Books. ------. 2015a. “Making National Social Protection Floors Work for Women.” Policy Brief 1. New York. van Bavel, B., D.R. Curtis and T. Soens. 2018. “Economic UNHRC (United Nations Human Rights Council). 2018. Inequality and Institutional Adaptation in Response to “The Safety of Journalists.” Resolution adopted by the ------. 2015b. Progress of the World’s Women 2015-2016: Flood Hazards: A Historical Analysis.” Ecology and Society Human Rights Council, 27 September. A/HRC/RES/39/6. Transforming Economies, Realizing Rights. New York. 23(4): 30. ------. 2017. “Equal Pay for Work of Equal Value.” www. Van den Hove, S., M. Le Menestrel and H.C. De UNICEF (United Nations Children’s Fund). 2013. Female unwomen.org/en/news/in-focus/csw61/equal-pay. Bettignies. 2002. “The Oil Industry and Climate Change: Genital Mutilation/Cutting: A Statistical Overview and Accessed 9 October 2019. Strategies and Ethical Dilemmas.” Climate Policy 2(1): Exploration of the Dynamics of Change. New York. 3­18. ------. 2018. “Historic Leap in Tunisia: Women Make Up ------. 2014a. A Statistical Snapshot of Violence against 47 Per Cent of Local Government.” 27 August. New York. van Raalte, A.A., I. Sasson and P. Martikainen. 2018. Adolescent Girls. New York. www.unwomen.org/en/news/stories/2018/8/feature-tuni- “The Case for Monitoring Life-Span Inequality.” Science sian-women-in-local-elections. Accessed 9 October 2019. 362(6418): 1002­1004. ------. 2014b. UNICEF Annual Report 2014: China. New York. www.unicef.org/about/annualreport/files/China_ ------. 2019. Progress of the World’s Women Report: van Zomeren, M. 2019. “Key Insights from the Social Annual_Report__2014.pdf. Accessed 8 November 2019. Families in a Changing World. New York. Psychology of Inequality and Human Development: From Social Embeddedness and Relative Deprivation to ------. 2018a. “Female Genital Mutilation.” February UN Women (United Nations Entity for Gender Equality Health and Participation.” Background paper for Human New York. https://data.unicef.org/topic/child-protection/ and the Empowerment of Women) and IPU (Inter- Development Report 2019. United Nations Development female-genital-mutilation/. Accessed 9 October 2019. Parliamentary Union). 2019. “Women in Politics 2019.” Programme, Human Development Report Office, New York. ------. 2018b. “Gender and Education.” June 2018. https:// Varga, C., I. Kiss and I. Ember. 2002. “The Lack of data.unicef.org/topic/gender/gender-disparities-in-educa- UN Women (United Nations Entity for Gender Equality Environmental Justice in Central and Eastern Europe.” tion/. Accessed 9 October 2019. and the Empowerment of Women), UNDP (United Environmental Health Perspectives 110(11): A662­A663. Nations Development Programme) and UNEP ------. 2019a. “Child Marriage Is a Violation of Human (United Nations Environment Programme). 2018. The Varró, M.J., G. Gombköto and M. Szeremi. 2001. “Risk Rights but Is All Too Common.” June. New York. https:// Cost of the Gender Gap in Agricultural Productivity: Five Factors of a Mass Lead Exposure, Heves, Hungary.” data.unicef.org/topic/child-protection/child-marriage/. African Countries. New York. Egészségtudomány 45: 167­180. Accessed 9 October 2019. Urzua, C. 2013. “Distributive and Regional Effects of Varughese, G., and E. Ostrom. 2001. “The Contested Role ------. 2019b. “Cyclone IDAI and Kenneth Post-Impact Monopoly Power.” Economía Mexicana 22(2): 279­295. of Heterogeneity in Collective Action: Some Situation.” May 2019. Paris. www.unicef.org/appeals/ Evidence from Community Forestry in Nepal.” World files/UNICEF-Idai-Situation-and-Response-12-May-2019. US CDC (Centers for Disease Control and Prevention). Development 29(5): 747­765. pdf. Accessed 10 October 2019. 2014. “Health and Academic Achievement.” Atlanta, GA. www.cdc.gov/healthyyouth/health_and_academics/pdf/ Veers, P., K. Dykes, E. Lantz, S. Barth, C.L. Bottasso, ------. 2019c. A World Ready to Learn: Prioritizing Quality health-academic-achievement.pdf. Accessed 9 August O. Carlson, A. Clifton, J. Green, P. Green, H. Early Childhood Education. New York. 2019. Holttinen, D. Laird, V. Lehtomäki, J.K. Lundquist, J. Manwell, M. Marquis, C. Meneveau, P. Moriarty, UNICEF (United Nations Children’s Fund) Innocenti US EPA (Environmental Protection Agency). 2015. X. Munduate, M. Muskulus, J. Naughton, L. Pao, Research Centre. 2010. The Dynamics of Social Change: “Environmental Justice Timeline.” Washington, DC. www. J. Paquette, J. Peinke, A. Robertson, J.S. Rodrigo, Towards the Abandonment of Female Genital Mutilation/ epa.gov/environmentaljustice/environmental-justice- A.M. Sempreviva, J.C. Smith, A. Tuohy and R. Wiser. Cutting in Five African Countries. Innocenti Insight. timeline. Accessed 14 August 2019. 2019. “Grand Challenges in the Science of Wind Energy.” Florence, Italy. Science 366(6464): 443. US Government. 2012. Economic Report of the President. UNICEF (United Nations Children’s Fund) and WHO Washington, DC. https://obamawhitehouse.archives.gov/ Verger, A., M. Moschetti and C. Fontdevila. 2017. La (World Health Organization). 2019. Progress on sites/default/files/microsites/ERP_2012_Complete.pdf. Privatización Educativa en Améríca Latina: Una Cartografía Household Drinking Water, Sanitation and Hygiene 2000- Accessed 9 September 2019. de Políticas, Tendencias y Trayectorias. Barcelona: 2017: Special Focus on Inequalities. New York. www.who. Educación Internacional. int/water_sanitation_health/publications/jmp-2019-full- US NOAA (National Oceanic and Atmospheric report.pdf?ua=1. Accessed 15 August 2019. Administration). 2018. “Garbage Patches: How Gyres Vernooij, R.W.M, D. Zeraatkar, M.A. Han, R. El Dib, M. Take Our Trash Out to Sea.” Ocean Podcast: Episode 14. Zworth, K. Milio, D. Sit, Y. Lee, H. Gomaa, C. Valli, UNIFEM (United Nations Development Fund for https://oceanservice.noaa.gov/podcast/mar18/nop14- M.J. Swierz, Y. Chang, S.E. Hanna, P.M. Brauer, J. Women). 2000. Progress of the World’s Women 2000: ocean-garbage-patches.html. Accessed 8 August 2019. Sievenpiper, R. de Souza, P. Alonso-Coello, M.M. UNIFEM Biennial Report. New York. Bala, G.H. Guyatt and B.C. Johnston. Forthcoming. Uslaner, E.M. 2002. The Moral Foundations of Trust. “Patterns of Red and Processed Meat Consumption UN Inter-agency Group for Child Mortality Estimation. Cambridge, UK: Cambridge University Press. and Risk for Cardiometabolic and Cancer Outcomes: a 2018. Levels & Trends in Child Mortality: Report 2019. United Nations Children’s Fund.

290 | HUMAN DEVELOPMENT REPORT 2019 Systematic Review and Meta-analysis of Cohort Studies.” Environmental Justice and Environmentalism: The Social Wolfe, N.D., C.P. Dunavan and J. Diamond. 2007. Annals of Internal Medicine. Justice Challenge to the Environmental Movement. “Origins of Major Human Infectious Diseases.” Nature Cambridge, MA: The MIT Press. 447(7142): 279­282. Vickers, C., and N. Zierbarth. 2019. “Lessons for Today from Past Periods of Rapid Technological Change.” WHO (World Health Organization). 2013. Global Wolford, B. n.d. “What Are the GDPR Consent DESA Working Paper 158. United Nations Department of and Regional Estimates of Violence against Women: Requirements?” https://gdpr.eu/gdpr-consent-require- Economic and Social Affairs, New York. Prevalence and Health Effects of Intimate Partner Violence ments/. Accessed 24 October 2019. and Non-Partner Sexual Violence. Geneva. Von Uexkull, N., M. Croicu, H. Fjelde and H. Buhaug. Woodall, L.C., A. Sanchez-Vidal, M. Canals, G.L.J. 2016. “Civil Conflict Sensitivity to Growing-Season ------. 2017. World Malaria Report. Geneva. Paterson, R. Coppock, V. Sleight, A. Calafat, A.D. Drought.” Proceedings of the National Academy of Rogers, B.E. Narayanaswamy and R.C. Thompson. Sciences 113(44): 12391­12396. ------. 2018. “Fact Sheet on Climate Change and Human 2014. “The Deep Sea is a Major Sink for Microplastic Health.” Geneva. www.who.int/en/news-room/fact- Debris.” Royal Society Open Science 1: 140317. Vona, F., and F. Patriarca. 2011. “Income Inequality and the sheets/detail/climate-change-and-health. Accessed 9 Development of Environmental Technologies.” Ecological August 2019. Woodard, D.L., S.J. Davis and J.T. Randerson. 2019. Economics 70(11): 2201­2213. “Economic Carbon Cycle Feedbacks May Offset Additional ------. 2019. “Ebola Virus Disease: Democratic Republic of Warming from Natural Feedbacks.” Proceedings of the Vörösmarty, C.J., P. Green, J. Salisbury and R.B. the Congo.” External Situation Report 40. Geneva. https:// National Academy of Sciences 116(3): 759­764. Lammers. 2000. “Global Water Resources: Vulnerability apps.who.int/iris/bitstream/handle/10665/312264/ from Climate Change and Population Growth.” Science SITREP_EVD_DRC_20190507-eng.pdf. Accessed 18 Woodhead, M., J. Rossiter, A. Dawes and A. Pankhurst. 289(5477): 284­288. October 2019. 2017. “Scaling-up Early Learning in Ethiopia: Exploring the Potential of O-Class.” Young Lives Working Paper 163. Vries, P. 2016. “What We Do and Do not Know about the WHO (World Health Organization) Study Group on University of Oxford, Oxford Department of International Great Divergence at the Beginning of 2016.” Historische Female Genital Mutilation and Obstetric Outcome. Development, Oxford, UK. Mitteilungen der Ranke-Gesellschaft 28: 249 ­297. www. 2006. “Female Genital Mutilation and Obstetric Outcome: researchgate.net/publication/290920219_What_we_do_ WHO Collaborative Prospective Study in Six African World Bank. 2006. World Development Report 2006: Equity and_do_not_know_about_the_Great_Divergence_at_ Countries.” Lancet 367(9525): 1835­1841. and Development. Washington, DC. the_beginning_of_2016. Accessed 6 October 2019. WHO (World Health Organization) and World Bank. ------. 2012a. World Development Indicators 2012. Waites, E.A. 1993. Trauma and Survival: Post-Traumatic and 2017. Tracking Universal Health Coverage: 2017 Global Washington, D.C.. http://datatopics.worldbank.org/world- Dissociative Disorders in Women. New York: Norton. Monitoring Report.” Geneva and Washington, DC. http:// development pubdocs.worldbank.org/en/193371513169798347/2017- -indicators/. Accessed 01 November 2019. Wake, D.B., and V.T. Vredenburg. 2008. “Are We in global-monitoring-report.pdf. Accessed 9 August 2019. the Midst of the Sixth Mass Extinction? A View from ------. 2012b. World Development Report 2012: Gender the World of Amphibians.” Proceedings of the National Wilensky, U. 1997 “Netlogo Segregation Model.” Center Equality and Development. Washington, DC. Academy of Sciences 105(Supplement 1): 11466­11473. for Connected Learning and Computer-Based Modeling, Northwestern University, Evanston, IL. http://ccl.north- ------. 2013. “Solutions for Youth Employment.” Walker, G., J. Fairburn, G. Smith and G. Mitchell. 2003. western.edu/netlogo/models/Segregation. Accessed 18 Washington, DC. www.s4ye.org/sites/default/files/ “Environmental Quality and Social Deprivation.” R&D September 2019. S4YE%20Digital%20Jobs%20Case%20Study%20-%20 Technical Report E2-067/1/TR. Bristol, UK: Environment 13.%20Laboratoria.pdf. Accessed 7 November 2019. Agency. www.researchgate.net/publication/237606377_ Wilkinson, R., and K. Pickett. 2011. The Spirit Level: Why Environmental_Quality_and_Social_Deprivation. Greater Equality Makes Societies Stronger. New York: ------. 2016. World Development Report 2016: Digital Accessed 14 August 2019. Bloomsbury Publishing USA. Dividends. Washington, DC.

Watts, N., W.N. Adger, P. Agnolucci, J. Blackstock, P. ------. 2018. The Inner Level: How More Equal Societies ------. 2017a. Global Findex database. https://globalfin- Byass, W. Cai, S. Chaytor, T. Colbourn, M. Collins, Reduce Stress, Restore Sanity and Improve Everyone’s dex.worldbank.org/. Accessed 9 October 2019. A. Cooper and P.M Cox. 2015. “Health and Climate Well-Being. New York: Penguin Press. Change: Policy Responses to Protect Public Health.” ------. 2017b. World Development Report 2017: Lancet 386(10006): 1861­1914. Williams, D.R., H.W. Neighbors and J.S. Jackson. 2003. Governance and the Law. Washington, DC. “Racial/Ethnic Discrimination and Health: Findings from Watts, N., M. Amann, N. Arnell, S. Ayeb-Karlsson, Community Studies.” American Journal of Public Health ------. 2018a. Poverty and Shared Prosperity: Piecing Belesova., H. Berry, T. Bouley, M. Boykoff, P. Byass, 93(2): 200­208. www.ncbi.nlm.nih.gov/pmc/articles/ Together the Poverty Puzzle. Washington, DC: World Bank. W. Cai and D. Campbell-Lendrum. 2018. “The 2018 PMC2518588/. Report of the Lancet Countdown on Health and Climate ------. 2018b. “The Human Capital Project.” World Bank, Change: Shaping the Health of Nations for Centuries to Wilson, G. 2012. “Intersex Genital Mutilation ­ IGM: The Washington, DC. Come.” Lancet 392(10163): 2479­2514. Fourteen Days of Intersex.” 25 Feburary. Organization Intersex International. http://oiiinternational.com/2574/ ------. 2019a. “Afronomics: M-Pesa and Rise of Digital Watts, N., M. Amann, S. Ayeb-Karlsson, K. Belesova, T. intersex-genital-mutilation-igm-fourteen-days-intersex/. Financial Services in Africa.” https://olc.worldbank.org/ Bouley, M. Boykoff, P. Byass, W. Cai, D. Campbell- Accessed 26 July 2019. content/afronomics-m-pesa-and-rise-digital-financial- Lendrum, J. Chambers and P.M. Cox. 2018. “The services-africa. Accessed 6 October 2019. Lancet Countdown on Health and Climate Change: from Winsemius, H.C., B. Jongman, T.I. Veldkamp, S. 25 Years of Inaction to a Global Transformation for Public Hallegatte, M. Bangalore and P.J. Ward. 2018. ------. 2019b. “Mainstreaming Disruptive Technologies at Health.” Lancet 391(10120): 581­630. “Disaster Risk, Climate Change, and Poverty: Assessing the World Bank Group.” Development Committee Paper. the Global Exposure of Poor People to Floods and Washington, DC. Webber, D. 2015. “Firm Market Power and the Earnings Droughts.” Environment and Development Economics Distribution.” Labour Economics 35(C): 123­134. 23(3): 328­348. ------. 2019c. “World Bank, UNESCO Institute for Statistics Join Forces to Help Countries Measure WEF (World Economic Forum). 2017. The Global Gender Wirsenius, S., C. Azar and G. Berndes. 2010. “How Student Learning.” Press Release, 4 July. World Bank, Gap Report 2017. Geneva. Much Land Is Needed for Global Food Production under Washington, DC. www.worldbank.org/en/news/press- Scenarios of Dietary Changes and Livestock Productivity release/2019/07/03/world-bank-unesco-institute-for- ------. 2018. The Global Gender Gap Report 2018. Geneva. Increases in 2030?” Agricultural Systems 103(9): 621­638. statistics-join-forces-to-help-countries-measure-student- learning. Accessed 11 November 2019. Weitzman, M. 2012. “GHG Targets as Insurance against Woldehanna, T., and M.W. Araya. 2017. “Early Investment Catastrophic Climate Damages.” Journal of Public in Preschool and Completion of Secondary Education in ------. 2019d. State and Trends of Carbon Pricing Economic Theory 14(2): 221­244. Ethiopia: Lessons from Young Lives.” Young Lives Working 2019. Washington, DC: World Bank. http://documents. Paper 168. University of Oxford, Oxford Department of worldbank.org/curated/en/191801559846379845/pdf/ Wenz, P. 2007. “Does Environmentalism Promote Injustice International Development, Oxford, UK. State-and-Trends-of-Carbon-Pricing-2019.pdf. Accessed 1 for the Poor?” In R. Sandler and P.C. Pezzullo, eds., November 2019.

References | 291 ------. 2020. World Development Report 2020: Trading Background in the Context of Inconsistent Selection Proceedings of the 2017 Conference on Empirical Methods for Development in the Age of Global Value Chains. Mechanisms in Higher Education.” Sociology of Education in Natural Language Processing. Washington, DC. 91(3): 224­241. Zhao, X., X. Zhang and S. Shao. 2016. “Decoupling CO2 World Policy Analysis Center. 2019. Gender Data Base. Zeraatkar, D., B.C. Johnston, J. Bartoszko, K. Cheung, Emissions and Industrial Growth in China over 1993­2013: www.worldpolicycenter.org/maps-data/data-download. M.M. Bala, C. Valli, M. Rabassa, D. Sit, K. Milio, The role of Investment.” Energy Economics 60: 275­292. Accessed 10 October 2019. B. Sadeghirad, A. Agarwal, A.M. Zea, Y. Lee, M.A. Han, R.W.M. Vernooij, P. Alonso-Coello, G.H. Guyatt Zheng, B. 2018. “Almost Lorenz Dominance.” Social Choice World Wide Web Foundation. 2017. “Artificial and R. El Dib. Forthcoming. “Effect of Lower Versus and Welfare 51(1): 51­63. Intelligence: The Road Ahead in Low and Middle-Income Higher Red Meat Intake on Cardiometabolic and Cancer Countries.” White Paper. Washington, DC. Outcomes: a Systematic Review of Randomized Trials.” Zimm, C., F. Sperling and S. Busch. 2018. “Identifying Annals of Internal Medicine. Sustainability and Knowledge Gaps in Socio-Economic Wrzesniewski, A., and J.E. Dutton. 2001. “Crafting a Job: Pathways vis-à-vis the Sustainable Development Goals.” Revisioning Employees as Active Crafters of Their Work.” Zeraatkar, D., M.A. Han, G.H. Guyatt, R.W.M. Vernooij, Economies 6(2): 20. Academy of Management Review 25: 179­201. R. El Dib, K. Cheung, K. Milio, M. Zworth, J.J. Bartoszko, C. Valli, M. Rabassa, Y. Lee, J. Zajac, A. Zucman, G. 2013. “The Missing Wealth of Nations: Are Wu, T. 2018. The Curse of Bigness: Antitrust in the New Prokop-Dorner, C. Lo, M.M. Bala, P. Alonso-Coello, Europe and the U.S. Net Debtors or Net Creditors?” Gilded Age. New York: Columbia Global Reports. SE. Hanna and B.C. Johnston. Forthcoming. “Red Quarterly Journal of Economics 128(3): 1321­1364. and Processed Meat Consumption and Risk for All-Cause Wu, T., and S.A. Thompson. 2019. “The Roots of Big Tech Mortality and Cardiometabolic Outcomes: a Systematic ------. 2014. “Taxing across Borders: Tracking Personal Run Disturbingly Deep.” New York Times, 7 June. www. Review of Meta-analysis of Cohort Studies.” Annals of Wealth and Corporate Profits.” Journal of Economic nytimes.com/interactive/2019/06/07/opinion/google- Internal Medicine. Perspectives 28(4): 121­148. facebook-mergers-acquisitions-antitrust.html. Accessed 9 August 2019. Zhao, C., B. Liu, S. Piao, X. Wang, D.B. Lobell, Y. Huang, ------. 2015. The Hidden Wealth of Nations: The Scourge M. Huang, Y. Yao, S. Bassu, P. Ciais and J.L. Durand. of Tax Havens. Chicago, IL: University of Chicago Press. Xie, Y., S. Cheng and X. Zhou. 2015. “Assortative Mating 2017. “Temperature Increase Reduces Global Yields of without Assortative Preference.” Proceedings of the Major Crops in Four Independent Estimates.” Proceedings ------. 2019. “Global Wealth Inequality.” Annual Review of National Academy of Sciences 112(19): 5974­5978. of the National Academy of Sciences 114(35): 9326­9331. Economics 11: 109­138.

Yanowitch, M. 1977. Social and Economic Inequality in the Zhao, J., T. Wang, M. Yatskar, V. Ordonex and K.-W. Zwijnenburg, J., S. Bournot and F. Giovannelli. 2017. Soviet Union: Six Studies. White Plains, NY: Sharpe. Chang. 2017. “Men Also Like Shopping: Reducing Gender “OECD Expert Group on Disparities in a National Accounts Bias Amplification Using Corpus-level Constraints.” Framework ­ Results from the 2015 Exercise.” Working Yastrebov, G., Y. Kosyakova and D. Kurakin. 2018. Paper 76. Organisation for Economic Co-operation and “Slipping Past the Test: Heterogeneous Effects of Social Development, Paris.

292 | HUMAN DEVELOPMENT REPORT 2019 Statistical annex HUMAN DEVELOPMENT REPORT 2019 Readers guide Inequalities in human development in the 21st century

Statistical tables 295

Human development composite indices 300 1 Human Development Index and its components 304 2 Human Development Index trends, 1990­2018 308 3 Inequality-adjusted Human Development Index 312 4 Gender Development Index 316 5 Gender Inequality Index 320 6 Multidimensional Poverty Index: developing countries Human development dashboards 323 7 Quality of human development 328 8 Life-course gender gap 333 9 Women’s empowerment 338 10 Environmental sustainability 343 11 Socioeconomic sustainability 348 Developing regions 349

Statistical annex | 293

Readers guide

The 20 statistical tables in this annex provide an overview of see Technical note 6 at http://hdr.undp.org/sites/default/files/ key aspects of human development. The first six tables contain hdr2019_technical_notes.pdf. the family of composite human development indices and their components estimated by the Human Development Report Comparisons over time and across editions Office (HDRO). The sixth table is produced in partnership with the Oxford Poverty and Human Development Initiative Because national and international agencies continually (OPHI). The remaining tables present a broader set of indi- improve their data series, the data—including the HDI values cators related to human development. The five dashboards and ranks—presented in this report are not comparable to use colour coding to visualize partial groupings of countries those published in earlier editions. For HDI comparability according to performance on each indicator. across years and countries, see table 2, which presents trends using consistent data, or http://hdr.undp.org/en/data, which Tables 1­6 and dashboards 1­5 are part of the printed version presents interpolated consistent data. of the 2019 Human Development Report. The full set of 20 sta- tistical tables is part of the digital version of the report and is post- Discrepancies between national and ed at http://hdr.undp.org/en/human-development-report-2019. international estimates

Unless otherwise noted, tables use data available to the HDRO National and international data can differ because international as of 15 July 2019. All indices and indicators, along with techni- agencies harmonize national data using a consistent methodol- cal notes on the calculation of composite indices and additional ogy and occasionally produce estimates of missing data to allow source information, are available at http://hdr.undp.org/en/data. comparability across countries. In other cases international agencies might not have access to the most recent national data. Countries and territories are ranked by 2018 Human Devel- When HDRO becomes aware of discrepancies, it brings them opment Index (HDI) value. Robustness and reliability analysis to the attention of national and international data authorities. has shown that for most countries differences in HDI are not statistically significant at the fourth decimal place. For this rea- son countries with the same HDI value at three decimal places are listed with tied ranks.

Sources and definitions Country groupings and aggregates

Unless otherwise noted, the HDRO uses data from interna- The tables present weighted aggregates for several country tional data agencies with the mandate, resources and expertise groupings. In general, an aggregate is shown only when data are to collect national data on specific indicators. available for at least half the countries and represent at least two- thirds of the population in that grouping. Aggregates for each Definitions of indicators and sources for original data com- grouping cover only the countries for which data are available. ponents are given at the end of each table, with full source details in Statistical references.

Methodology updates Human development classification

The 2019 Report retains all the composite indices from the HDI classifications are based on HDI fixed cutoff points, family of human development indices—the HDI, the Inequality- which are derived from the quartiles of distributions of the adjusted Human Development Index (IHDI), the Gender Devel- component indicators. The cutoff points are HDI of less than opment Index (GDI), the Gender Inequality Index (GII) and the 0.550 for low human development, 0.550­0.699 for medium Multidimensional Poverty Index (MPI). The methodology used human development, 0.700­0.799 for high human develop- to compute the indices is the same as the one used in the 2018 ment and 0.800 or greater for very high human development. Statistical Update. For details, see Technical notes 1­5 at http:// hdr.undp.org/sites/default/files/hdr2019_technical_notes.pdf. Regional groupings

The 2019 Report has five colour-c oded dashboards (quality of Regional groupings are based on United Nations Development human development, life-course gender gap, women’s empower- Programme regional classifications. Least Developed Countries ment, environmental sustainability and socioeconomic sustain- and Small Island Developing States are defined according to ability). For details on the methodology used to create them, UN classifications (see www.unohrlls.org).

Readers guide | 295 Developing countries Monetary Fund; International Telecommunication Union; Inter-Parliamentary Union; Luxembourg Income Study; The developing countries aggregates include all countries that Office of the United Nations High Commissioner for Human are included in a regional grouping. Rights; Office of the United Nations High Commissioner for Refugees; Organisation for Economic Co-operation and Devel- Organisation for Economic Co-operation and opment; Socio-Economic Database for Latin America and the Development Caribbean; Syrian Center for Policy Research; United Nations Children’s Fund; United Nations Conference on Trade and Of the 36 Organisation for Economic Co-operation and Development; United Nations Department of Economic and Development members, 33 are considered developed countries Social Affairs; United Nations Economic and Social Commis- and 3 (Chile, Mexico and Turkey) are considered developing sion for West Asia; United Nations Educational, Scientific and countries. Aggregates refer to all countries from the group for Cultural Organization Institute for Statistics; United Nations which data are available. Entity for Gender Equality and the Empowerment of Women; United Nations Office on Drugs and Crime; World Bank; Country notes and World Health Organization. The international education database maintained by Robert Barro (Harvard University) Data for China do not include Hong Kong Special Administra- and Jong-Wha Lee (Korea University) was another invaluable tive Region of China, Macao Special Administrative Region of source for the calculation of the Report’s indices. China or Taiwan Province of China. Statistical tables As of 2 May 2016, Czechia is the short name to be used for the Czech Republic. The first six tables relate to the five composite human devel- opment indices and their components. Since the 2010 Human As of 1 June 2018, the Kingdom of Eswatini is the name of Development Report, four composite human development the country formerly known as Swaziland. indices—the HDI, the IHDI, the GII and the MPI for devel- oping countries—have been calculated. The 2014 Report intro- As of 14 February 2019, the Republic of North Macedonia duced the GDI, which compares the HDI calculated separately (short form: North Macedonia) is the name of the country for- for women and men. merly known as the former Yugoslav Republic of Macedonia. The remaining tables present a broader set of human develop- Symbols ment indicators and provide a more comprehensive picture of a country’s human development. A dash between two years, as in 2012­2018, indicates that the For indicators that are global Sustainable Development Goals data are from the most recent year available during the period indicators or can be used in monitoring progress towards specif- ic goals, the table headers include the relevant goals and targets. specified. A slash between years, as in 2013/2018, indicates that Table 1, Human Development Index and its components, the data are the average for the years shown. Growth rates are ranks countries by 2018 HDI value and details the values of the three HDI components: longevity, education (with two usually average annual rates of growth between the first and last indicators) and income per capita. The table also presents the difference in rankings by HDI value and gross national income years of the period shown. per capita, as well as the rank on the 2017 HDI, calculated using the most recently revised historical data available in 2019. The following symbols are used in the tables: Table 2, Human Development Index trends, 1990­2018, .. Not available provides a time series of HDI values allowing 2018 HDI values to be compared with those for previous years. The table uses the 0 or 0.0 Nil or negligible most recently revised historical data available in 2019 and the same methodology applied to compute 2018 HDI values. The — Not applicable table also includes the change in HDI rank over the last five years and the average annual HDI growth rate across four time Statistical acknowledgements intervals: 1990­2000, 2000­2010, 2010­2018 and 1990­2018

The Report’s composite indices and other statistical resources Table 3, Inequality-adjusted Human Development Index, draw on a wide variety of the most respected international contains two related measures of inequality—the IHDI and data providers in their specialized fields. HDRO is particularly the loss in HDI due to inequality. The IHDI looks beyond the grateful to the Centre for Research on the Epidemiology of Disasters; Economic Commission for Latin America and the Caribbean; Eurostat; Food and Agriculture Organization; Gallup; ICF Macro; Internal Displacement Monitoring Centre; International Labour Organization; International

296 | HUMAN DEVELOPMENT REPORT 2019 average achievements of a country in longevity, education and HUMAN DEVELOPMENT REPORT 2019 income to show how these achievements are distributed among Beyond income, beyond averages, beyond today: its residents. The IHDI value can be interpreted as the level of Inequalities in human development in the 21st century human development when inequality is accounted for. The rel- ative difference between IHDI and HDI values is the loss due ratios and total fertility rates, which can help assess the burden to inequality in distribution of the HDI within the country. of support that falls on the labour force in a country. The table presents the coefficient of human inequality, which is the unweighted average of inequalities in the three dimensions. Table 8, Health outcomes, presents indicators of infant In addition, the table shows each country’s difference in rank health (percentage of infants who are exclusively breastfed in on the HDI and the IHDI. A negative value means that taking the 24 hours prior to the survey, percentage of infants who lack inequality into account lowers a country’s rank on the HDI. immunization for DPT and measles and infant mortality rate) The table also presents the income shares of the poorest 40 and of child health (percentage of children under age 5 who are percent, the richest 10 percent and the richest 1 percent of the stunted and under-five mortality rates). The table also contains population, as well as the Gini coefficient. indicators of adult health (adult mortality rates by gender, mortality rates due to noncommunicable diseases by gender Table 4, Gender Development Index, measures dispar- incidence of malaria and tuberculosis and HIV prevalence ities on the HDI by gender. The table contains HDI values rates). Finally, it includes healthy life expectancy at birth and estimated separately for women and men; the ratio of which is current health expenditure as a percentage of GDP. the GDI value. The closer the ratio is to 1, the smaller the gap between women and men. Values for the three HDI compo- Table 9, Education achievements, presents standard edu- nents—longevity, education (with two indicators) and income cation indicators. The table provides indicators of educational per capita—are also presented by gender. The table includes five attainment—adult and youth literacy rates and the share of the country groupings by absolute deviation from gender parity in adult population with at least some secondary education. Gross HDI values. enrolment ratios at each level of education are complemented by primary school dropout rate and survival rate to the last grade Table 5, Gender Inequality Index, presents a composite of lower secondary general education. The table also presents measure of gender inequality using three dimensions: repro- government expenditure on education as a percentage of GDP. ductive health, empowerment and the labour market. The reproductive health indicators are the maternal mortality ratio Table 10, National income and composition of resources, and the adolescent birth rate. The empowerment indicators are covers several macroeconomic indicators such as gross domestic the share of parliamentary seats held by women and the share product (GDP), labour share of GDP (which includes wages of population with at least some secondary education by gender. and social protection transfers), gross fixed capital formation, The labour market indicator is participation in the labour force and taxes on income, profit and capital gains as a percentage of by gender. A low GII value indicates low inequality between total tax revenue. Gross fixed capital formation is a rough indi- women and men, and vice-versa. cator of national income that is invested rather than consumed. In times of economic uncertainty or recession, gross fixed Table 6, Multidimensional Poverty Index, captures the capital formation typically declines. General government final multiple deprivations that people in developing countries face in consumption expenditure (presented as a share of GDP and as their health, education and standard of living. The MPI shows average annual growth) is an indicator of public spending. In both the incidence of nonincome multidimensional poverty (a addition, the table presents two indicators of debt—domestic headcount of those in multidimensional poverty) and its inten- credit provided by the financial sector and total debt service, sity (the average deprivation score experienced by poor people). both measured as a percentage of GDP or GNI. The consumer Based on deprivation score thresholds, people are classified as price index, a measure of inflation, is also presented. vulnerable to multidimensional poverty, multidimensionally poor or in severe multidimensional poverty. The table includes Table 11, Work and employment, contains indicators on the contribution of deprivation in each dimension to overall four topics: employment, unemployment, work that is a risk to multidimensional poverty. It also presents measures of income human development and employment-related social security. poverty—population living below the national poverty line and The employment indicators are the employment to popula- population living on less than $1.90 in purchasing power parity tion ratio, the labour force participation rate, employment in terms per day. MPI values are based on a revised methodology agriculture and employment in services. The unemployment developed in partnership with OPHI. For details, see Technical indicators are total unemployment, youth unemployment and note 5 at http://hdr.undp.org/sites/default/files/hdr2019_tech- youth not in school or employment. The indicators on work nical_notes.pdf and OPHI’s website (http://ophi.org.uk/ that is a risk to human development are child labour, the multidimensional-poverty-index/). working poor and the proportion of informal employment in nonagricultural employment. A new indicator on skill-level Table 7, Population trends, contains major population employment—high-skill to low-skill employment ratio—has indicators, including total population, median age, dependency been added. The indicator on employment-related social secu- rity is the percentage of the eligible population that receives an old-age pension.

Table 12, Human security, reflects the extent to which the population is secure. The table begins with the percentage of births

Readers guide | 297 that are registered, followed by the number of refugees by country countries globally. A country that is in the top third group on all of origin and the number of internally displaced people. It then indicators can be considered a country with the highest quality shows the size of the homeless population due to natural disasters, of human development. The dashboard shows that not all coun- the number of deaths and missing people attributed to disasters, tries in the very high human development group have the high- the population of orphaned children and the prison population. It est quality of human development across all quality indicators also provides homicide and suicide rates (by gender), an indicator and that many countries in the low human development group on justification of wife beating and an indicator on the depth of are in the bottom third of all quality indicators in the table. food deficit (average dietary energy supply adequacy). Dashboard 2, Life-course gender gap, contains a selection of Table 13, Human and capital mobility, provides indica- indicators that indicate gender gaps in choices and opportunities tors of several aspects of globalization. International trade is over the life course—childhood and youth, adulthood and older captured by measuring exports and imports as a share of GDP. age. The indicators refer to health, education, labour market and Financial flows are represented by net inflows of foreign direct work, seats in parliament, time use and social protection. Most investment and flows of private capital, net official develop- indicators are presented as a ratio of female to male values. Sex ment assistance and inflows of remittances. Human mobility ratio at birth is an exception to grouping by tercile—countries is captured by the net migration rate, the stock of immigrants, are divided into two groups: the natural group (countries with the net number of tertiary students from abroad (expressed as a value of 1.04­1.07, inclusive) and the gender-biased group (all a percentage of total tertiary enrolment in the country) and other countries). Deviations from the natural sex ratio at birth the number of international inbound tourists. International have implications for population replacement levels; they can communication is represented by the percentages of the total suggest possible future social and economic problems and may and female populations that use the Internet, the number of indicate gender bias. Countries with values of a parity index con- mobile phone subscriptions per 100 people and the percentage centrated around 1 form the group with the best achievements change in mobile phone subscriptions between 2010 and 2017. in that indicator. Deviations from parity are treated equally regardless of which gender is overachieving. Table 14, Supplementary indicators: perceptions of well-be- ing, includes indicators that reflect individuals’ perceptions of Dashboard 3, Women’s empowerment, contains a selec- relevant dimensions of human development—education quality, tion of woman-specific empowerment indicators that allows health care quality, standard of living, personal safety, freedom of empowerment to be compared across three dimensions: repro- choice and overall life satisfaction. The table also presents indi- ductive health and family planning, violence against girls and cators reflecting perceptions about community and government. women and socioeconomic empowerment. Most countries have at least one indicator in each tercile, which implies that wom- Table 15, Status of fundamental human rights treaties, en’s empowerment is unequal across indicators and countries. shows when countries ratified key human rights conventions. The 11 selected conventions cover basic human rights and free- Dashboard 4, Environmental sustainability, contains a doms related to elimination of all forms of racial and gender selection of indicators that cover environmental sustainability discrimination and violence, protection of children’s rights, and environmental threats. The environmental sustainability rights of migrant workers and persons with disabilities. They indicators present levels of or changes in energy consumption, also cover torture and other cruel, inhuman and degrading carbon dioxide emissions, forest area, fresh water withdrawals treatment as well as protection from enforced disappearance. and natural resource depletion. The environmental threats indicators are mortality rates attributed to household and Dashboard 1, Quality of human development, contains a ambient air pollution and to unsafe water, sanitation and selection of indicators associated with the quality of health, edu- hygiene services, proportion of land that is degraded mostly by cation and standard of living. The indicators on quality of health human activities and practices, and the International Union for are lost health expectancy, number of physicians and number Conservation of Nature Red List Index value, which measures of hospital beds. The indicators on quality of education are aggregate extinction risk across groups of species. pupil­teacher ratio in primary schools; primary school teachers trained to teach; proportion of primary and secondary schools Dashboard 5, Socioeconomic sustainability, contains a with access to the Internet; and Programme for International selection of indicators that cover economic and social sustain- Student Assessment (PISA) scores in mathematics, reading and ability. The economic sustainability indicators are adjusted science. The indicators on quality of standard of living are the net savings, total debt service, gross capital formation, skilled proportion of employment that is in vulnerable employment, labour force, diversity of exports and expenditure on research the proportion of rural population with access to electricity, and development. The social sustainability indicators are the the proportion of population using at least basic drinking-water old age dependency ratio projected to 2030, the ratio of edu- services and the proportion of population using at least basic cation and health expenditure to military expenditure, change sanitation facilities. A country in the top third of an indicator in overall loss in HDI value due to inequality and changes in distribution has performed better than at least two-thirds of gender and income inequality.

298 | HUMAN DEVELOPMENT REPORT 2019 composite indices TABLE 1 Human Development Index and its components

TABLE Human Development of schooling Gross national income GNI per capita rank 1 (years) (GNI) per capita minus HDI rank HDI rank HDI rank 2018a (2011 PPP $) 2017 VERY HIGH HUMAN DEVELOPMENT 12.6 2018 2018 1 1 Norway 0.954 82.3 18.1 b 13.4 2 2 Switzerland 0.946 83.6 16.2 12.5 c 68,059 5 3 3 Ireland 0.942 82.1 18.8 b 14.1 4 4 Germany 0.939 81.2 17.1 12.0 59,375 8 6 4 Hong Kong, China (SAR) 0.939 84.7 16.5 12.7 c 5 6 Australia 0.938 83.3 22.1 b 12.5 c 55,660 9 7 6 Iceland 0.938 82.9 19.2 b 12.4 7 8 Sweden 0.937 82.7 18.8 b 11.5 46,946 15 9 9 Singapore 0.935 83.5 16.3 12.2 10 10 Netherlands 0.933 82.1 18.0 b 12.6 60,221 5 11 11 Denmark 0.930 80.8 19.1 b 12.4 12 12 Finland 0.925 81.7 19.3 b 13.3 c 44,097 15 13 13 Canada 0.922 82.3 16.1 12.7 c 14 14 New Zealand 0.921 82.1 18.8 b 13.0 e 47,566 12 15 15 United Kingdom 0.920 81.2 17.4 13.4 15 15 United States 0.920 78.9 16.3 11.8 47,955 9 17 17 Belgium 0.919 81.5 19.7 b 12.5 g 18 18 Liechtenstein 0.917 80.5 f 14.7 12.8 i 83,793 d ­6 19 19 Japan 0.915 84.5 15.2 12.6 20 20 Austria 0.914 81.4 16.3 12.2 e 50,013 3 21 21 Luxembourg 0.909 82.1 14.2 13.0 22 22 Israel 0.906 82.8 16.0 12.2 48,836 4 22 22 Korea (Republic of) 0.906 82.8 16.4 12.3 24 24 Slovenia 0.902 81.2 17.4 9.8 41,779 12 25 25 Spain 0.893 83.4 17.9 12.7 27 26 Czechia 0.891 79.2 16.8 11.4 43,602 10 26 26 France 0.891 82.5 15.5 11.3 28 28 Malta 0.885 82.4 15.9 10.2 e 35,108 18 29 29 Italy 0.883 83.4 16.2 13.0 c 30 30 Estonia 0.882 78.6 16.1 12.1 39,507 13 31 31 Cyprus 0.873 80.8 14.7 10.5 31 32 Greece 0.872 82.1 17.3 12.3 56,140 ­4 33 32 Poland 0.872 78.5 16.4 13.0 34 34 Lithuania 0.869 75.7 16.5 11.0 43,821 5 35 35 United Arab Emirates 0.866 77.8 13.6 10.2 38 36 Andorra 0.857 81.8 f 13.3 j 9.7 e 99,732 d,h ­16 36 36 Saudi Arabia 0.857 75.0 17.0 e 12.6 c 37 36 Slovakia 0.857 77.4 14.5 12.8 c 40,799 6 39 39 Latvia 0.854 75.2 16.0 9.2 40 40 Portugal 0.850 81.9 16.3 9.7 46,231 0 40 41 Qatar 0.848 80.1 12.2 10.4 42 42 Chile 0.847 80.0 16.5 9.1 i 65,543 ­13 43 43 Brunei Darussalam 0.845 75.7 14.4 11.9 44 43 Hungary 0.845 76.7 15.1 9.4 e 33,650 13 45 45 Bahrain 0.838 77.2 15.3 11.4 e 46 46 Croatia 0.837 78.3 15.0 9.7 36,757 8 47 47 Oman 0.834 77.6 14.7 10.6 c 48 48 Argentina 0.830 76.5 17.6 12.0 e 32,143 13 49 49 Russian Federation 0.824 72.4 15.5 12.3 l 50 50 Belarus 0.817 74.6 15.4 11.8 i 35,041 8 51 50 Kazakhstan 0.817 73.2 15.3 11.8 51 52 Bulgaria 0.816 74.9 14.8 11.4 e 31,597 12 51 52 Montenegro 0.816 76.8 15.0 11.0 51 52 Romania 0.816 75.9 14.3 12.4 e 40,511 0 56 55 Palau 0.814 73.7 f 15.6 e 10.6 m 51 56 Barbados 0.813 79.1 15.2 e 7.3 34,795 6 57 57 Kuwait 0.808 75.4 13.8 8.7 58 57 Uruguay 0.808 77.8 16.3 7.7 36,141 2 59 59 Turkey 0.806 77.4 16.4 e 11.5 e 60 60 Bahamas 0.805 73.8 12.8 n 10.2 30,379 10 61 61 Malaysia 0.804 76.0 13.5 9.7 j 62 62 Seychelles 0.801 73.3 15.5 33,100 5

24,909 20

27,626 13

29,775 7

66,912 ­28

48,641 k ­20

49,338 ­22

30,672 3

26,301 10

27,935 4

110,489 d ­40

21,972 17

76,389 d ­39

27,144 4

40,399 ­18

23,061 9

37,039 ­18

17,611 18

25,036 2

17,039 18

22,168 8

19,646 9

17,511 15

23,906 2

16,720 14

15,912 18

71,164 ­52

19,435 5

24,905 ­6

28,395 ­17

27,227 ­15

25,077 ­12

300 | HUMAN DEVELOPMENT REPORT 2019

Human Development of schooling Gross national income GNI per capita rank TABLE (years) (GNI) per capita minus HDI rank HDI rank 1 Value (years) (years) 2017 2018a (2011 PPP $) HDI rank 2018 2018 14.8 HIGH HUMAN DEVELOPMENT 0.799 75.8 13.0 e 11.2 15,218 15 65 63 Serbia 14.7 63 Trinidad and Tobago 0.799 73.4 15.0 11.0 l 28,497 ­21 63 65 Iran (Islamic Republic of) 12.9 66 Mauritius 0.797 76.5 15.4 10.0 18,166 ­2 63 67 Panama 15.2 68 Costa Rica 0.796 74.9 15.4 9.4 i 22,724 ­10 66 69 Albania 14.0 70 Georgia 0.795 78.3 14.4 10.2 i 20,455 ­7 66 71 Sri Lanka 13.6 e 72 Cuba 0.794 80.1 12.5 e 8.7 14,790 12 68 73 Saint Kitts and Nevis 13.8 j 74 Antigua and Barbuda 0.791 78.5 14.3 10.1 m 12,300 20 69 75 Bosnia and Herzegovina 14.7 e 76 Mexico 0.786 73.6 16.6 12.8 9,570 34 70 77 Thailand 15.4 78 Grenada 0.780 76.8 14.6 11.1 e 11,611 24 72 79 Brazil 13.2 e 79 Colombia 0.778 78.7 14.7 e 11.8 e 7,811 o 43 71 81 Armenia 13.5 82 Algeria 0.777 74.6 f 13.8 8.5 n 26,770 ­25 73 82 North Macedonia 13.9 e 82 Peru 0.776 76.9 14.9 e 9.3 j 22,201 ­17 73 85 China 12.4 e 85 Ecuador 0.769 77.3 15.1 e 9.7 12,690 10 75 87 Azerbaijan 14.1 88 Ukraine 0.767 75.0 13.9 e 8.6 17,628 ­11 76 89 Dominican Republic 15.1 89 Saint Lucia 0.765 76.9 14.2 e 7.7 16,129 ­6 77 91 Tunisia 11.3 92 Mongolia 0.763 72.4 12.7 e 8.8 n 12,684 8 78 93 Lebanon 13.6 e 94 Botswana 0.761 75.7 13.1 e 7.8 e 14,068 2 78 94 Saint Vincent and the Grenadines 12.8 e 96 Jamaica 0.761 77.1 13.0 e 8.3 12,896 4 78 96 Venezuela (Bolivarian Republic of) 14.4 e 98 Dominica 0.760 74.9 12.7 e 11.8 9,277 26 81 98 Fiji 12.9 e 98 Paraguay 0.759 76.7 11.9 e 8.0 l 13,639 0 81 98 Suriname 13.1 102 Jordan 0.759 75.7 12.1 q 9.7 l 12,874 2 81 103 Belize 14.3 e 104 Maldives 0.759 76.5 12.7 e 9.2 12,323 6 85 105 Tonga 11.6 106 Philippines 0.758 76.7 10.9 e 7.9 m 16,127 ­13 86 107 Moldova (Republic of) 12.0 108 Turkmenistan 0.758 76.8 12.8 n 9.0 10,141 17 84 108 Uzbekistan 12.9 110 Libya 0.754 72.9 12.5 e 10.5 15,240 ­10 87 111 Indonesia 13.7 111 Samoa 0.750 72.0 14.0 s 11.3 m 7,994 25 88 113 South Africa 12.9 n 114 Bolivia (Plurinational State of) 0.745 73.9 13.1 7.9 15,074 ­10 90 116 Egypt 0.745 76.1 12.4 e 8.5 11,528 7 89 MEDIUM HUMAN DEVELOPMENT 12.7 l 117 Marshall Islands 0.739 76.5 12.8 7.2 e 10,677 10 91 118 Viet Nam 11.1 q 119 Palestine, State of 0.735 69.7 13.1 e 10.2 e 10,784 7 94 120 Iraq 13.4 121 Morocco 0.730 78.9 11.5 e 8.7 n 11,136 5 93 123 Guyana 0.728 69.3 9.3 m 15,951 ­21 97

0.728 72.4 8.6 n 11,746 ­2 95

0.726 74.4 9.8 e 7,932 18 96

0.726 72.1 10.3 9,070 p 14 92

0.724 78.1 f 7.8 j 9,245 10 98

0.724 67.3 10.9 i 9,110 11 102

0.724 74.1 8.5 11,720 ­5 99

0.724 71.6 9.1 11,933 ­8 99

0.723 74.4 10.5 i 8,268 10 99

0.720 74.5 9.8 l 7,136 17 103

0.719 78.6 6.8 q 12,549 ­17 105

0.717 70.8 11.2 i 5,783 26 104

0.712 71.1 9.4 e 9,540 ­1 106

0.711 71.8 11.6 6,833 16 106

0.710 68.1 9.8 q 16,407 ­38 108

0.710 71.6 11.5 6,462 18 109

0.708 72.7 7.6 m 11,685 r ­16 111

0.707 71.5 8.0 11,256 ­14 111

0.707 73.2 10.6 j 5,885 18 110

0.705 63.9 10.2 11,756 ­22 111

0.703 71.2 9.0 6,849 8 114

0.702 66.2 8.3 q 15,794 ­40 114

0.700 71.8 7.3 i 10,744 ­16 116

0.698 73.9 f 10.9 e 4,633 21 116

0.693 75.3 8.2 i 6,220 10 118

0.690 73.9 9.1 5,314 15 119

0.689 70.5 7.3 e 15,365 ­44 120

0.676 76.5 5.5 i 7,480 ­3 121

0.674 71.3 10.9 l 3,317 30 122

0.670 69.8 8.5 l 7,615 ­7 123

TABLE 1 Human Development Index and its components | 301 TABLE 1 HUMAN DEVELOPMENT INDEX AND ITS COMPONENTS

TABLE Human Development of schooling Gross national income GNI per capita rank 1 (years) (GNI) per capita minus HDI rank HDI rank HDI rank 2018a (2011 PPP $) 2017 124 El Salvador 6.9 2018 2018 124 125 Tajikistan 0.667 73.1 12.0 10.7 q 126 126 Cabo Verde 0.656 70.9 11.4 e 6.2 6,973 ­3 128 126 Guatemala 0.651 72.8 11.9 6.5 127 126 Nicaragua 0.651 74.1 10.6 6.8 i 3,482 26 125 129 India 0.651 74.3 12.2 s 6.5 e 129 130 Namibia 0.647 69.4 12.3 6.9 i 6,513 ­1 129 131 Timor-Leste 0.645 63.4 12.6 q 4.5 q 131 132 Honduras 0.626 69.3 12.4 e 6.6 7,378 ­7 133 132 Kiribati 0.623 75.1 10.2 7.9 j 132 134 Bhutan 0.623 68.1 11.8 e 3.1 e 4,790 11 134 135 Bangladesh 0.617 71.5 12.1 e 6.1 136 135 Micronesia (Federated States of) 0.614 72.3 11.2 7.7 j 6,829 ­5 135 137 Sao Tome and Principe 0.614 67.8 11.5 j 6.4 e 138 138 Congo 0.609 70.2 12.7 e 6.5 m 9,683 ­27 136 138 Eswatini (Kingdom of) 0.608 64.3 11.6 n 6.7 l 138 140 Lao People’s Democratic Republic 0.608 59.4 11.4 e 5.2 i 7,527 ­14 140 141 Vanuatu 0.604 67.6 11.1 6.8 l 141 142 Ghana 0.597 70.3 11.4 e 7.2 i 4,258 7 142 143 Zambia 0.596 63.8 11.5 7.1 q 144 144 Equatorial Guinea 0.591 63.5 12.1 q 5.6 j 3,917 11 143 145 Myanmar 0.588 58.4 9.2 n 5.0 q 146 146 Cambodia 0.584 66.9 10.3 4.8 i 8,609 ­23 145 147 Kenya 0.581 69.6 11.3 e 6.6 i 148 147 Nepal 0.579 66.3 11.1 e 4.9 i 4,057 6 148 149 Angola 0.579 70.5 12.2 5.1 q 147 150 Cameroon 0.574 60.8 11.8 q 6.3 l 3,700 10 150 150 Zimbabwe 0.563 58.9 12.7 8.3 e 153 152 Pakistan 0.563 61.2 10.5 5.2 3,024 20 151 153 Solomon Islands 0.560 67.1 8.5 5.5 q 152 LOW HUMAN DEVELOPMENT 0.557 72.8 10.2 e 5,804 ­8 154 Syrian Arab Republic 5.1 t 154 155 Papua New Guinea 0.549 71.8 8.9 e 4.6 i 9,359 ­32 155 156 Comoros 0.543 64.3 10.0 e 4.9 q 156 157 Rwanda 0.538 64.1 11.2 e 4.4 e 6,317 ­13 158 158 Nigeria 0.536 68.7 11.2 6.5 q 157 159 Tanzania (United Republic of) 0.534 54.3 9.7 l 6.0 i 2,808 17 160 159 Uganda 0.528 65.0 8.0 6.1 q 160 161 Mauritania 0.528 63.0 11.2 e 4.6 i 4,099 ­2 159 162 Madagascar 0.527 64.7 8.5 6.1 n 162 163 Benin 0.521 66.7 10.4 3.8 m 3,582 7 163 164 Lesotho 0.520 61.5 12.6 6.3 i 164 165 Côte d’Ivoire 0.518 53.7 10.7 5.2 i 17,796 ­80 165 166 Senegal 0.516 57.4 9.6 3.1 e 166 167 Togo 0.514 67.7 9.0 4.9 q 5,764 ­13 166 168 Sudan 0.513 60.8 12.6 3.7 i 168 169 Haiti 0.507 65.1 7.7 e 5.4 q 3,597 2 169 170 Afghanistan 0.503 63.7 9.5 n 3.9 i 170 171 Djibouti 0.496 64.5 10.1 4.0 j 3,052 9 171 172 Malawi 0.495 66.6 6.5 e 4.6 i 172 173 Ethiopia 0.485 63.8 11.0 q 2.8 q 2,748 13 173 174 Gambia 0.470 66.2 8.7 e 3.7 q 178 174 Guinea 0.466 61.7 9.5 e 2.7 q 5,555 ­16 175 176 Liberia 0.466 61.2 9.0 e 4.7 i 173 177 Yemen 0.465 63.7 9.6 e 3.2 m 3,291 3 175 178 Guinea-Bissau 0.463 66.1 8.7 e 3.3 l 177 179 Congo (Democratic Republic of the) 0.461 58.0 10.5 n 6.8 2,661 12 179 180 Mozambique 0.459 60.4 9.7 e 3.5 e 180 181 Sierra Leone 0.446 60.2 9.7 3.6 i 5,190 ­17 181 182 Burkina Faso 0.438 54.3 10.2 e 1.6 q 183 182 Eritrea 0.434 61.2 8.9 3.9 n 2,027 13 182 184 Mali 0.434 65.9 5.0 2.4 l 184 185 Burundi 0.427 58.9 7.6 3.1 q 2,725 r 7 185 0.423 61.2 11.3 3,686 ­9

2,426 7

1,959 11

5,086 ­22

2,805 0

1,752 11

3,746 ­17

1,404 19

2,135 2

3,244 ­9

3,589 ­16

3,256 ­12

1,593 10

3,962 ­26

1,665 6

1,746 1

3,601 u ­24

1,159 11

1,782 ­4

1,490 4

2,211 ­10

1,040 9

1,433 r 3

1,593 ­2

800 8

1,154 4

1,381 1

1,705 ­8

1,708 u ­9

1,965 ­17

660 4

302 | HUMAN DEVELOPMENT REPORT 2019

SDG 3 SDG 4.3 SDG 4.6 SDG 8.5 Life expectancy Expected years Mean years Human Development of schooling Gross national income GNI per capita rank TABLE (years) (GNI) per capita minus HDI rank HDI rank 1 2018a (2011 PPP $) 2017 2018 2018 2018a 4.8 186 HDI rank 0.413 57.6 5.0 e 2.4 q 2018 2018 187 0.401 54.0 7.5 e 4.3 i 188 186 South Sudan 0.381 52.8 7.6 e 2.0 e 1,455 u ­7 189 187 Chad 0.377 62.0 6.5 188 Central African Republic .. 1,716 ­15 .. 189 Niger .. 72.1 10.8 e .. .. OTHER COUNTRIES OR TERRITORIES … .. 777 0 .. .. Korea (Democratic People’s Rep. of) .. .. 11.3 e .. 912 ­3 .. .. Monaco .. 15.1 … Nauru .. 57.1 … San Marino .. .. 12.0 — .. Somalia 0.892 12.3 8.3 .. .. — .. Tuvalu 0.750 79.5 6.4 — Human development groups 0.634 75.1 16.4 4.8 17,313 .. — 0.507 69.3 13.8 7.4 — Very high human development 0.686 61.3 11.7 .. .. High human development 71.1 9.3 7.1 — Medium human development 0.703 12.2 7.9 .. .. — Low human development 0.741 71.9 10.2 — Developing countries 0.779 75.3 12.0 8.6 5,409 .. — Regions 0.759 74.2 13.4 6.5 — Arab States 0.642 75.4 14.6 5.7 40,112 — — East Asia and the Pacific 0.541 69.7 14.5 4.8 — Europe and Central Asia 0.528 61.2 11.8 8.6 14,403 — — Latin America and the Caribbean 0.723 65.0 10.0 South Asia 71.8 9.8 12.0 6,240 — — Sub-Saharan Africa 0.895 12.2 8.4 — Least developed countries 0.731 80.4 2,581 — Small island developing states 72.6 16.3 Organisation for Economic 12.7 10,476 — World 15,721 —

14,611 —

15,498 —

13,857 —

6,794 —

3,443 —

2,630 —

15,553 —

40,615 —

15,745 —

NOTES o Based on a cross-country regression and the Life expectancy at birth: Number of years a revised data available in 2019 that were used to projected growth rate from UNECLAC (2019). newborn infant could expect to live if prevailing calculate HDI values for 2018. a Data refer to 2018 or the most recent year patterns of age-specific mortality rates at the time of available. p HDRO estimate based on data from World Bank birth stay the same throughout the infant’s life. MAIN DATA SOURCES (2019a), United Nations Statistics Division b In calculating the HDI value, expected years of (2019b) and UNECLAC (2019). Expected years of schooling: Number of years Columns 1 and 7: HDRO calculations based on schooling is capped at 18 years. of schooling that a child of school entrance age data from UNDESA (2019b), UNESCO Institute for q Updated by HDRO based on data from ICF Macro can expect to receive if prevailing patterns of Statistics (2019), United Nations Statistics Division c Based on data from OECD (2018). Demographic and Health Surveys for 2006­2018. age-specific enrolment rates persist throughout the (2019b), World Bank (2019a), Barro and Lee (2018) child’s life. and IMF (2019). d In calculating the HDI value, GNI per capita is r HDRO estimate based on data from World Bank capped at $75,000. (2019a), United Nations Statistics Division Mean years of schooling: Average number of Column 2: UNDESA (2019b). (2019b) and projected growth rates from years of education received by people ages 25 and e Updated by HDRO based on data from UNESCO UNESCWA (2018). older, converted from education attainment levels Column 3: UNESCO Institute for Statistics (2019), Institute for Statistics (2019). using official durations of each level. ICF Macro Demographic and Health Surveys, UNICEF s Updated by HDRO based on data from CEDLAS Multiple Indicator Cluster Surveys and OECD (2018). f Value from UNDESA (2011). and World Bank (2018). Gross national income (GNI) per capita: Aggregate income of an economy generated by Column 4: UNESCO Institute for Statistics (2019), g Imputed mean years of schooling for Austria. t Updated by HDRO based on Syrian Center for its production and its ownership of factors of Barro and Lee (2018), ICF Macro Demographic and Policy Research (2017). production, less the incomes paid for the use of Health Surveys, UNICEF Multiple Indicator Cluster h Estimated using the purchasing power parity (PPP) factors of production owned by the rest of the world, Surveys and OECD (2018). rate and projected growth rate of Switzerland. u HDRO estimate based on data from World Bank converted to international dollars using PPP rates, (2019a), United Nations Statistics Division divided by midyear population. Column 5: World Bank (2019a), IMF (2019) and i Based on Barro and Lee (2018). (2019b) and IMF (2019). United Nations Statistics Division (2019b). GNI per capita rank minus HDI rank: Difference j Based on data from the national statistical office. DEFINITIONS in ranking by GNI per capita and by HDI value. A Column 6: Calculated based on data in columns negative value means that the country is better 1 and 5. k Estimated using the PPP rate and projected Human Development Index (HDI): A composite ranked by GNI than by HDI value. growth rate of Spain. index measuring average achievement in three basic dimensions of human development— a long and HDI rank for 2017: Ranking by HDI value for 2017, l Updated by HDRO based on data from United healthy life, knowledge and a decent standard of which was calculated using the same most recently Nations Children’s Fund (UNICEF) Multiple living. See Technical note 1 at http://hdr.undp.org/ Indicator Cluster Surveys for 2006­2018. sites/default/files/hdr2019_technical_notes.pdf for details on how the HDI is calculated. m Updated by HDRO using Barro and Lee (2018)

n Based on cross-country regression.

TABLE 1 Human Development Index and its components | 303 TABLE 2 Human Development Index trends, 1990­2018

Human Development Index (HDI) Change in Average annual HDI growth

Value (%)

HDI rank 1990 2000 2010 2013 2015 2016 2017 2018 2013­2018a 1990­2000 2000­2010 2010­2018 1990­2018

TABLE VERY HIGH HUMAN DEVELOPMENT 0.850 0.917 0.942 0.946 0.948 0.951 0.953 0.954 0 0.76 0.27 0.16 0.41 1 Norway 0.832 0.889 0.932 0.938 0.943 0.943 0.943 2 2 Switzerland 0.764 0.857 0.890 0.908 0.926 0.936 0.939 0.946 0 0.67 0.47 0.18 0.46 3 Ireland 0.801 0.869 0.920 0.927 0.933 0.936 0.938 4 Germany 0.781 0.827 0.901 0.916 0.927 0.931 0.936 0.942 13 1.16 0.38 0.71 0.75 4 Hong Kong, China (SAR) 0.866 0.898 0.926 0.926 0.933 0.935 0.937 6 Australia 0.804 0.861 0.892 0.920 0.927 0.932 0.935 0.939 0 0.82 0.57 0.25 0.57 6 Iceland 0.816 0.897 0.906 0.927 0.932 0.934 0.935 8 Sweden 0.718 0.818 0.909 0.923 0.929 0.933 0.934 0.939 6 0.58 0.86 0.51 0.66 9 Singapore 0.830 0.876 0.911 0.924 0.927 0.929 0.932 10 Netherlands 0.799 0.863 0.910 0.926 0.926 0.928 0.929 0.938 0 0.37 0.30 0.17 0.29 11 Denmark 0.784 0.858 0.903 0.916 0.919 0.922 0.924 12 Finland 0.850 0.868 0.895 0.910 0.917 0.920 0.921 0.938 3 0.69 0.35 0.64 0.55 0.820 0.870 0.899 0.907 0.914 0.917 0.920 13 Canada 0.775 0.867 0.905 0.914 0.916 0.918 0.919 0.937 ­4 0.96 0.09 0.42 0.49 14 New Zealand 0.860 0.881 0.911 0.914 0.917 0.919 0.919 15 United Kingdom 0.806 0.873 0.903 0.908 0.913 0.915 0.917 0.935 ­1 1.31 1.07 0.35 0.95 15 United States 0.862 0.904 0.912 0.912 0.915 0.916 17 Belgium .. 0.855 0.885 0.900 0.906 0.910 0.913 0.933 ­3 0.55 0.39 0.31 0.42 18 Liechtenstein 0.816 0.838 0.895 0.896 0.906 0.909 0.912 19 Japan 0.795 0.860 0.893 0.892 0.899 0.904 0.908 0.930 ­6 0.77 0.54 0.27 0.54 20 Austria 0.790 0.853 0.887 0.895 0.901 0.902 0.904 21 Luxembourg 0.792 0.817 0.882 0.893 0.899 0.901 0.904 0.925 ­2 0.90 0.52 0.30 0.59 22 Israel 0.728 0.824 0.881 0.884 0.886 0.892 0.899 22 Korea (Republic of) 0.829 0.825 0.865 0.875 0.885 0.888 0.891 0.922 2 0.21 0.31 0.38 0.29 24 Slovenia 0.754 0.796 0.862 0.874 0.882 0.885 0.888 25 Spain 0.730 0.842 0.872 0.882 0.888 0.887 0.890 0.921 4 0.59 0.34 0.30 0.42 26 Czechia 0.780 0.787 0.847 0.861 0.877 0.881 0.883 26 France 0.744 0.830 0.871 0.873 0.875 0.878 0.881 0.920 ­3 1.13 0.43 0.21 0.62 28 Malta 0.769 0.780 0.844 0.863 0.871 0.875 0.879 29 Italy 0.730 0.799 0.850 0.854 0.864 0.869 0.871 0.920 ­3 0.24 0.34 0.12 0.24 30 Estonia 0.731 0.796 0.857 0.858 0.868 0.866 0.871 31 Cyprus 0.753 0.785 0.835 0.851 0.858 0.864 0.868 0.919 ­1 0.80 0.33 0.22 0.47 32 Greece 0.712 0.755 0.824 0.840 0.855 0.860 0.866 32 Poland 0.732 0.782 0.821 0.839 0.860 0.863 0.864 0.917 ­4 .. 0.48 0.17 .. 34 Lithuania 0.723 0.759 0.828 0.846 0.850 0.854 0.852 35 United Arab Emirates 0.744 0.810 0.846 0.857 0.857 0.856 0.915 0 0.47 0.34 0.42 0.41 36 Andorra .. 0.763 0.829 0.844 0.849 0.851 0.854 36 Saudi Arabia 0.698 0.728 0.817 0.834 0.842 0.845 0.849 0.914 0 0.54 0.66 0.26 0.50 36 Slovakia 0.739 0.785 0.822 0.837 0.843 0.846 0.848 39 Latvia 0.698 0.816 0.834 0.857 0.851 0.847 0.848 0.909 2 0.85 0.37 0.22 0.50 40 Portugal 0.711 0.753 0.800 0.830 0.839 0.843 0.845 41 Qatar 0.757 0.805 0.832 0.844 0.843 0.844 0.843 0.906 ­1 0.74 0.39 0.27 0.48 42 Chile 0.703 0.769 0.826 0.835 0.835 0.838 0.841 43 Brunei Darussalam 0.768 0.792 0.796 0.807 0.834 0.839 0.839 0.906 0 1.17 0.77 0.33 0.78 43 Hungary 0.704 0.749 0.811 0.825 0.830 0.832 0.835 45 Bahrain 0.736 0.704 0.793 0.811 0.827 0.834 0.833 0.902 0 ­0.05 0.67 0.29 0.30 46 Croatia 0.670 0.770 0.818 0.824 0.828 0.828 0.832 47 Oman 0.721 0.780 0.803 0.813 0.817 0.822 0.893 1 0.90 0.47 0.40 0.60 48 Argentina .. 0.682 0.792 0.808 0.811 0.812 0.815 49 Russian Federation 0.707 0.685 0.764 0.791 0.806 0.808 0.813 0.891 1 0.86 0.80 0.41 0.71 50 Belarus 0.734 0.712 0.779 0.792 0.807 0.812 0.813 50 Kazakhstan 0.793 0.801 0.807 0.809 0.813 0.891 ­1 0.77 0.35 0.27 0.48 52 Bulgaria .. .. 0.797 0.800 0.806 0.808 0.813 52 Montenegro 0.690 0.709 0.776 0.811 0.803 0.808 0.811 0.885 2 0.56 0.74 0.55 0.62 52 Romania 0.694 0.736 0.799 0.812 0.812 0.814 0.813 55 Palau 0.771 0.794 0.798 0.807 0.809 0.809 0.883 ­1 0.77 0.48 0.17 0.49 56 Barbados .. 0.786 0.774 0.797 0.802 0.806 0.807 57 Kuwait 0.701 0.742 0.743 0.781 0.800 0.800 0.805 0.882 ­1 0.67 0.79 0.54 0.68 57 Uruguay 0.655 0.795 0.797 0.799 0.800 0.804 59 Turkey .. 0.787 0.773 0.787 0.797 0.801 0.802 0.873 2 0.90 0.62 0.34 0.64 60 Bahamas 0.732 0.724 0.762 0.782 0.801 0.801 0.800 61 Malaysia 0.712 0.712 0.872 ­1 0.56 0.74 0.22 0.53 62 Seychelles 0.692 0.579 0.872 2 0.98 0.62 0.54 0.72

.. 0.869 5 0.31 0.88 0.67 0.62 0.644 0.866 5 0.78 0.48 0.68 0.65 0.857 ­1 .. 0.88 0.43 ..

0.857 ­1 0.64 0.85 0.71 0.74

0.857 1 0.33 0.82 0.42 0.53

0.854 4 0.41 1.16 0.56 0.72

0.850 1 0.98 0.46 0.42 0.64

0.848 ­9 0.74 0.22 0.22 0.41

0.847 2 0.70 0.61 0.71 0.67

0.845 ­6 0.47 0.33 0.19 0.34

0.845 ­1 0.89 0.72 0.28 0.65

0.838 6 0.74 0.06 0.64 0.46

0.837 ­1 1.12 0.79 0.41 0.80

0.834 1 .. 1.19 0.63 ..

0.830 ­2 0.86 0.61 0.18 0.58

0.824 3 ­0.18 0.79 0.69 0.41

0.817 0 .. 1.50 0.39 ..

0.817 9 ­0.07 1.10 0.84 0.61

0.816 6 0.26 0.90 0.58 0.58

0.816 1 .. .. 0.36 ..

0.816 2 0.11 1.18 0.29 0.54

0.814 ­7 .. 0.53 0.60 ..

0.813 ­9 0.53 0.35 0.22 0.38

0.808 ­2 1.00 0.10 0.22 0.45

0.808 ­1 0.69 0.42 0.54 0.55

0.806 5 1.26 1.26 1.03 1.19

0.805 ­4 .. 0.10 0.16 ..

0.804 ­1 1.18 0.66 0.49 0.80

0.801 1 .. 0.68 0.63 ..

304 | HUMAN DEVELOPMENT REPORT 2019

Human Development Index (HDI) Change in Average annual HDI growth

Value (%)

HDI rank 1990 2000 2010 2013 2015 2016 2017 2018 2013­2018a 1990­2000 2000­2010 2010­2018 1990­2018

HIGH HUMAN DEVELOPMENT 0.706 0.710 0.762 0.775 0.785 0.791 0.794 0.799 4 0.06 0.71 0.60 0.45 TABLE 63 Serbia 0.667 0.721 0.788 0.787 0.796 0.796 0.799 0.799 63 Trinidad and Tobago 0.577 0.671 0.756 0.785 0.789 0.799 0.799 0.797 ­3 0.78 0.90 0.17 0.65 2 65 Iran (Islamic Republic of) 0.620 0.674 0.748 0.775 0.786 0.790 0.793 0.796 66 Mauritius 0.659 0.719 0.758 0.775 0.782 0.788 0.793 0.795 ­3 1.53 1.20 0.68 1.17 67 Panama 0.655 0.711 0.754 0.777 0.786 0.789 0.792 0.794 68 Costa Rica 0.644 0.667 0.740 0.781 0.788 0.788 0.789 0.791 1 0.84 1.04 0.79 0.90 69 Albania 0.669 0.732 0.756 0.771 0.776 0.783 0.786 70 Georgia .. 0.687 0.750 0.765 0.772 0.774 0.776 0.780 0 0.87 0.53 0.60 0.67 71 Sri Lanka 0.625 0.686 0.776 0.762 0.768 0.771 0.777 0.778 72 Cuba 0.676 0.747 0.767 0.769 0.772 0.774 0.777 ­2 0.82 0.59 0.64 0.69 73 Saint Kitts and Nevis .. 0.771 0.767 0.770 0.772 0.774 0.776 74 Antigua and Barbuda .. .. 0.714 0.748 0.755 0.765 0.767 0.769 ­5 0.35 1.05 0.84 0.74 75 Bosnia and Herzegovina .. 0.669 0.739 0.750 0.759 0.764 0.765 0.767 76 Mexico .. 0.705 0.721 0.731 0.746 0.753 0.762 0.765 5 .. 0.90 0.91 .. 77 Thailand 0.652 0.649 0.743 0.750 0.756 0.760 0.760 0.763 78 Grenada 0.574 .. 0.726 0.752 0.755 0.757 0.760 0.761 2 0.95 0.88 0.49 0.80 79 Brazil .. 0.684 0.729 0.746 0.753 0.759 0.760 0.761 79 Colombia 0.613 0.662 0.729 0.743 0.748 0.751 0.758 0.760 2 0.15 1.24 0.02 0.50 81 Armenia 0.600 0.649 0.730 0.746 0.751 0.755 0.758 0.759 82 Algeria 0.633 0.646 0.735 0.743 0.753 0.757 0.758 0.759 ­2 .. .. 0.48 .. 82 North Macedonia 0.578 0.669 0.721 0.742 0.750 0.755 0.756 0.759 82 Peru .. 0.679 0.702 0.727 0.742 0.749 0.753 0.758 ­3 .. .. 0.08 .. 85 China 0.613 0.591 0.716 0.751 0.758 0.756 0.757 0.758 85 Ecuador 0.501 0.669 0.732 0.741 0.749 0.749 0.752 0.754 5 .. 0.65 0.93 .. 87 Azerbaijan 0.642 0.641 0.732 0.744 0.742 0.746 0.747 0.750 88 Ukraine .. 0.671 0.701 0.712 0.733 0.738 0.741 0.745 2 0.79 0.48 0.47 0.59 89 Dominican Republic 0.705 0.653 0.730 0.726 0.736 0.744 0.744 0.745 89 Saint Lucia 0.593 0.694 0.717 0.725 0.731 0.736 0.738 0.739 12 1.24 1.05 0.74 1.03 91 Tunisia .. 0.653 0.697 0.728 0.736 0.730 0.729 0.735 92 Mongolia 0.569 0.589 0.751 0.741 0.728 0.725 0.732 0.730 0 .. .. 0.33 .. 93 Lebanon 0.583 .. 0.660 0.699 0.714 0.719 0.724 0.728 94 Botswana .. 0.578 0.711 0.714 0.721 0.725 0.726 0.728 ­3 1.11 0.59 0.59 0.78 94 Saint Vincent and the Grenadines 0.570 0.674 0.723 0.720 0.722 0.722 0.725 0.726 96 Jamaica .. 0.669 0.753 0.772 0.763 0.752 0.735 0.726 2 0.99 0.96 0.54 0.85 96 Venezuela (Bolivarian Republic of) 0.641 0.672 0.733 0.730 0.729 0.729 0.723 0.724 98 Dominica 0.638 0.694 0.694 0.707 0.718 0.718 0.721 0.724 3 0.24 1.17 0.52 0.65 98 Fiji .. 0.675 0.692 0.709 0.718 0.718 0.722 0.724 98 Paraguay 0.640 0.640 0.701 0.724 0.730 0.725 0.722 0.724 ­1 1.11 1.23 0.49 0.97 98 Suriname 0.588 .. 0.728 0.720 0.721 0.722 0.722 0.723 102 Jordan .. 0.702 0.693 0.707 0.715 0.722 0.719 0.720 2 .. 0.94 0.41 .. 103 Belize 0.616 0.643 0.669 0.693 0.709 0.713 0.716 0.719 104 Maldives 0.613 0.610 0.692 0.699 0.714 0.715 0.717 0.717 4 1.03 0.59 0.65 0.76 105 Tonga .. 0.666 0.672 0.692 0.702 0.704 0.709 0.712 106 Philippines 0.645 0.631 0.681 0.702 0.703 0.705 0.709 0.711 7 1.66 1.74 0.95 1.48 107 Moldova (Republic of) 0.590 0.609 0.673 0.691 0.701 0.706 0.708 0.710 108 Turkmenistan 0.653 .. 0.665 0.688 0.696 0.701 0.707 0.710 ­8 0.41 0.68 0.71 0.59 108 Uzbekistan .. 0.596 0.757 0.707 0.691 0.690 0.704 0.708 110 Libya .. 0.728 0.666 0.688 0.696 0.700 0.704 0.707 0 .. 1.34 0.36 .. 111 Indonesia 0.676 0.604 0.690 0.696 0.699 0.704 0.706 0.707 111 Samoa 0.525 0.638 0.662 0.683 0.699 0.702 0.704 0.705 ­5 ­0.49 0.87 0.29 0.22 113 South Africa 0.621 0.629 0.655 0.673 0.685 0.692 0.700 0.703 114 Bolivia (Plurinational State of) 0.625 0.616 0.658 0.679 0.692 0.696 0.700 0.702 10 0.97 0.71 0.76 0.82 115 Gabon 0.540 0.627 0.666 0.681 0.690 0.695 0.696 0.700 116 Egypt 0.619 0.611 4 .. 0.50 0.26 .. MEDIUM HUMAN DEVELOPMENT 0.546 0.696 117 Marshall Islands .. 0.690 3 1.40 0.93 0.39 0.94 118 Viet Nam .. 0.578 0.689 119 Palestine, State of 0.475 0.684 ­1 0.11 1.70 0.66 0.83 120 Iraq .. 0.675 121 Morocco .. 0.608 0.671 ­6 .. .. ­0.36 .. 122 Kyrgyzstan 0.574 0.531 0.668 123 Guyana 0.458 0.594 11 0.14 1.34 1.22 0.88 0.618 0.606 0.537 4 .. 0.54 0.29 ..

0 0.42 0.78 0.05 0.44

­26 0.51 1.14 ­0.45 0.46

­8 .. 0.54 ­0.15 ..

3 0.53 0.28 0.52 0.44

2 0.85 0.80 0.56 0.75

­3 .. .. 0.41 ..

­6 1.31 0.36 ­0.07 0.57

­2 0.49 0.74 0.49 0.58

4 .. 0.92 0.90 ..

0 0.31 0.39 0.45 0.38

3 0.67 0.62 0.73 0.67

­3 ­0.70 1.12 0.56 0.30

2 .. .. 0.67 ..

3 .. 1.10 0.83 ..

­9 0.74 0.39 ­0.84 0.16

0 1.40 0.99 0.74 1.07

­4 0.26 0.79 0.30 0.46

0 0.06 0.52 0.78 0.43

3 1.31 0.63 0.88 0.94

1 0.13 0.48 0.81 0.45

­2 1.13 0.86 0.62 0.89

… .. 0.698 … .. .. 0.653 0.673 0.680 0.685 0.671 0.681 0.685 0.687 0.693 ­1 1.99 1.23 0.74 1.36 0.652 0.662 0.665 0.672 0.618 0.646 0.660 0.669 0.690 ­5 .. .. 0.35 .. 0.636 0.658 0.666 0.669 0.639 0.656 0.663 0.666 0.689 ­1 0.58 0.71 0.68 0.65

0.676 2 1.48 1.53 1.14 1.40

0.674 ­1 ­0.39 0.69 0.73 0.31

0.670 ­1 1.21 0.53 0.61 0.79

TABLE 2 Human Development Index trends, 1990­2018 | 305 TABLE 2 HUMAN DEVELOPMENT INDEX TRENDS, 1990­2018

Human Development Index (HDI) Change in Average annual HDI growth

Value (%)

HDI rank 1990 2000 2010 2013 2015 2016 2017 2018 2013­2018a 1990­2000 2000­2010 2010­2018 1990­2018

TABLE 124 El Salvador 0.529 0.608 0.659 0.662 0.660 0.662 0.665 0.667 ­5 1.40 0.82 0.14 0.83 125 Tajikistan 0.603 0.538 0.630 0.643 0.642 0.647 0.651 0.656 ­1 ­1.13 1.60 0.50 0.30 2 126 Cabo Verde 0.564 0.626 0.641 0.643 0.645 0.647 0.651 ­1 1.06 0.48 126 Guatemala .. 0.546 0.602 0.616 0.647 0.648 0.649 0.651 2 .. 0.98 0.98 .. 126 Nicaragua 0.477 0.568 0.614 0.630 0.644 0.649 0.653 0.651 0 1.36 0.77 0.74 1.11 129 India 0.494 0.497 0.581 0.607 0.627 0.637 0.643 0.647 1 1.41 1.57 1.34 0.99 130 Namibia 0.431 0.543 0.588 0.622 0.637 0.639 0.643 0.645 ­3 1.43 0.78 1.17 1.46 131 Timor-Leste 0.579 0.505 0.620 0.613 0.628 0.628 0.624 0.626 ­2 ­0.64 2.06 0.13 0.38 132 Honduras 0.555 0.598 0.603 0.613 0.618 0.621 0.623 0 0.76 0.51 132 Kiribati .. 0.564 0.589 0.605 0.619 0.622 0.623 0.623 ­1 .. 0.43 0.71 .. 134 Bhutan 0.508 0.571 0.594 0.606 0.610 0.615 0.617 0 0.88 0.98 0.73 135 Bangladesh .. 0.549 0.572 0.588 0.599 0.609 0.614 5 .. 1.40 135 Micronesia (Federated States of) .. 0.470 0.595 0.599 0.606 0.608 0.612 0.614 ­2 .. 1.56 0.41 .. 137 Sao Tome and Principe .. 0.541 0.546 0.568 0.590 0.593 0.603 0.609 5 .. 0.95 1.36 .. 138 Congo 0.388 0.480 0.557 0.581 0.614 0.613 0.609 0.608 ­1 1.95 1.31 1.12 1.65 138 Eswatini (Kingdom of) .. 0.495 0.513 0.558 0.585 0.596 0.603 0.608 6 .. 1.19 2.15 .. 140 Lao People’s Democratic Republic 0.437 0.468 0.546 0.579 0.594 0.598 0.602 0.604 ­2 0.94 0.92 1.28 1.19 141 Vanuatu 0.531 0.466 0.585 0.588 0.592 0.592 0.595 0.597 ­6 ­0.71 1.60 0.26 0.49 142 Ghana 0.545 0.554 0.578 0.585 0.587 0.591 0.596 ­3 ­1.51 0.91 0.39 143 Zambia 0.399 .. 0.531 0.559 0.570 0.580 0.589 0.591 0 1.55 .. 1.35 1.49 144 Equatorial Guinea .. 0.483 0.580 0.588 0.593 0.592 0.590 0.588 ­9 .. 1.39 0.18 .. 145 Myanmar 0.454 0.428 0.523 0.551 0.565 0.571 0.577 0.584 2 0.61 2.17 1.39 0.97 146 Cambodia 0.424 0.520 0.535 0.555 0.566 0.572 0.578 0.581 ­1 0.11 1.09 1.05 1.20 147 Kenya .. 0.424 0.533 0.551 0.562 0.568 0.574 0.579 0 .. 2.13 1.04 .. 147 Nepal 0.349 0.419 0.527 0.555 0.568 0.572 0.574 0.579 ­2 1.94 2.46 1.18 1.85 149 Angola 0.384 0.446 0.510 0.547 0.565 0.570 0.576 0.574 0.89 1.79 1.50 1.49 150 Cameroon 0.467 0.446 0.471 0.531 0.548 0.556 0.560 0.563 1 ­0.46 1.70 2.26 0.77 150 Zimbabwe 0.380 0.394 0.472 0.527 0.543 0.549 0.553 0.563 3 1.61 2.63 2.22 1.52 152 Pakistan .. 0.438 0.524 0.537 0.550 0.556 0.558 0.560 4 .. 0.71 0.85 .. 153 Solomon Islands 0.445 0.452 0.524 0.550 0.555 0.553 0.555 0.557 ­1 ­0.15 0.43 0.78 0.84 LOW HUMAN DEVELOPMENT 0.498 0.449 ­4 ­0.95 1.55 0.44 154 Syrian Arab Republic 0.404 0.476 0.544 1.06 0.97 1.17 155 Papua New Guinea .. 0.543 .. .. 156 Comoros 0.590 0.539 157 Rwanda 0.558 0.436 0.644 0.572 0.540 0.539 0.529 0.549 ­14 0.57 0.88 ­1.98 ­0.06 158 Nigeria 0.377 0.457 0.510 0.521 0.539 0.541 0.533 1.45 159 Tanzania (United Republic of) 0.337 0.513 0.532 0.535 0.537 0.522 0.543 0 1.58 0.80 1.31 159 Uganda .. 0.488 0.506 0.515 0.525 0.522 .. 161 Mauritania 0.245 .. 0.484 0.520 0.527 0.528 0.524 0.538 ­4 3.24 1.15 0.60 .. 162 Madagascar 0.395 0.487 0.503 0.519 0.518 0.518 163 Benin .. 0.395 0.489 0.503 0.515 0.520 0.515 0.536 2 .. 3.77 1.19 2.84 164 Lesotho 0.373 0.446 0.490 0.511 0.521 0.519 0.514 0.59 165 Côte d’Ivoire 0.312 0.456 0.504 0.509 0.514 0.515 0.512 0.534 ­2 2.37 .. 1.25 .. 166 Senegal 0.378 0.398 0.473 0.500 0.510 0.512 0.510 1.67 167 Togo 0.444 0.461 0.486 0.499 0.507 0.510 0.528 2 2.10 1.03 1.25 168 Sudan .. 0.407 0.454 0.475 0.494 0.508 0.507 .. 169 Haiti 0.348 0.390 0.468 0.494 0.504 0.506 0.501 0.528 2 1.36 2.16 0.97 1.89 170 Afghanistan 0.488 0.426 0.468 0.490 0.502 0.506 0.493 ­0.93 171 Djibouti 0.391 0.403 0.471 0.477 0.501 0.505 0.492 0.527 ­4 0.40 0.94 0.91 1.19 172 Malawi 0.377 0.440 0.467 0.483 0.492 0.497 0.482 0.36 173 Ethiopia 0.405 0.345 0.464 0.485 0.490 0.491 0.466 0.521 ­4 0.50 1.01 0.42 .. 174 Gambia 0.332 0.361 0.446 0.467 0.482 0.489 0.459 1.97 174 Guinea 0.412 0.362 0.437 0.463 0.475 0.478 0.463 0.520 0 0.67 1.74 1.19 1.45 176 Liberia 0.298 0.283 0.412 0.439 0.453 0.460 0.466 1.47 177 Yemen 0.382 0.437 0.448 0.454 0.456 0.463 0.518 2 0.37 1.46 0.21 178 Guinea-Bissau .. 0.335 0.408 0.439 0.449 0.456 0.460 .. 179 Congo (Democratic Republic of the) 0.303 0.422 0.441 0.463 0.463 0.463 0.456 0.516 5 1.79 1.09 1.61 0.99 180 Mozambique 0.432 0.499 0.506 0.493 0.477 0.442 181 Sierra Leone .. 0.426 0.441 0.453 0.457 0.435 0.514 ­2 .. 1.84 1.17 1.12 182 Burkina Faso 0.328 .. 0.416 0.429 0.445 0.453 0.429 1.53 182 Eritrea 0.278 0.333 0.396 0.412 0.428 0.435 0.431 0.513 ­2 1.86 0.94 1.16 0.85 184 Mali 0.301 0.391 0.426 0.422 0.423 0.426 185 Burundi .. 0.298 0.375 0.401 0.413 0.420 0.421 0.507 1 .. 1.57 0.93 1.53 0.392 0.286 0.433 0.425 0.433 0.434 0.99 0.403 0.408 0.412 0.420 0.503 ­1 0.60 0.92 0.72 .. .. 0.402 0.422 0.427 0.427 .. 0.377 0.308 0.496 ­3 ­1.24 3.01 0.83 1.84 0.217 0.293 3.34 0.270 0.495 0 0.99 2.14 1.32 ..

.. 0.485 0 .. 1.90 1.32 1.69 0.231 0.470 3 2.92 3.81 1.66 .. 0.295 ­0.07 0.466 0 1.35 0.79 1.26

0.466 2 2.00 1.67 1.86

0.465 ­4 0.44 0.67 ..

0.463 ­18 1.44 ­0.94 0.59

0.461 ­3 .. 1.01 ..

0.459 0 2.24 1.24 0.70

0.446 3 2.79 1.51 2.61

0.438 ­1 2.74 1.45 1.74

0.434 3 2.74 1.84 ..

0.434 ­1 .. 0.02 ..

0.427 0 2.72 0.72 2.22

0.423 ­3 3.20 0.65 1.29

306 | HUMAN DEVELOPMENT REPORT 2019

Human Development Index (HDI) Change in Average annual HDI growth

Value (%)

HDI rank 1990 2000 2010 2013 2015 2016 2017 2018 2013­2018a 1990­2000 2000­2010 2010­2018 1990­2018 .. .. 0.414 186 South Sudan .. 0.425 0.439 0.428 0.418 0.401 0.413 ­10 .. .. ­0.35 .. TABLE 187 Chad 0.298 0.374 0.399 0.403 0.398 0.376 0.401 ­1 .. 2.29 0.89 .. 188 Central African Republic 0.320 0.307 0.355 0.351 0.362 0.372 0.373 0.381 ­1 ­0.41 1.44 0.89 0.62 2 189 Niger 0.213 0.253 0.319 0.345 0.360 0.365 0.377 ­1 1.75 2.34 2.09 2.06 OTHER COUNTRIES OR TERRITORIES .. .. Korea (Democratic People’s Rep. of) … .. Monaco … … … … .. .. Nauru … .. San Marino … … … … .. .. Somalia … Tuvalu … .. 0.890 … … Human development groups 0.779 0.823 0.746 0.568 0.630 … .. 0.630 … … Very high human development 0.436 0.497 0.505 High human development 0.352 0.386 … .. 0.683 … … Medium human development 0.516 0.571 Low human development 0.866 0.878 0.886 0.888 0.701 0.892 — 0.55 0.52 0.36 0.48 Developing countries 0.556 0.613 0.706 0.727 0.738 0.743 0.737 Regions 0.519 0.597 0.575 0.599 0.616 0.625 0.776 0.750 — 1.04 1.15 0.75 1.00 Arab States 0.652 0.667 0.473 0.490 0.499 0.501 0.758 East Asia and the Pacific 0.628 0.687 0.642 0.663 0.674 0.680 0.639 0.634 — 1.30 1.48 1.22 1.34 Europe and Central Asia 0.441 0.505 0.539 Latin America and the Caribbean 0.402 0.423 0.525 0.507 — 0.94 2.04 0.88 1.32 South Asia 0.350 0.399 0.722 Sub-Saharan Africa 0.595 0.642 0.686 — 1.02 1.19 0.82 1.02 Least developed countries 0.894 Small island developing states 0.785 0.834 0.676 0.688 0.695 0.699 0.729 0.703 — 0.99 0.98 0.49 0.84 Organisation for Economic 0.598 0.641 0.691 0.714 0.727 0.733 Cooperation and Development 0.735 0.759 0.770 0.772 0.741 — 1.42 1.48 0.87 1.28 World 0.731 0.748 0.754 0.756 0.585 0.607 0.624 0.634 0.779 — 0.23 0.97 0.72 0.64 0.498 0.521 0.532 0.535 0.485 0.504 0.516 0.520 0.759 — 0.90 0.62 0.46 0.68 0.702 0.708 0.717 0.719 0.642 — 1.36 1.48 1.18 1.35 0.873 0.883 0.889 0.892 0.697 0.713 0.722 0.727 0.541 — 0.50 1.65 1.03 1.06

0.528 — 1.30 1.98 1.08 1.48

0.723 — 0.77 0.91 0.35 0.70

0.895 — 0.61 0.45 0.32 0.47

0.731 — 0.71 0.84 0.60 0.72

NOTES DEFINITIONS Average annual HDI growth: A smoothed (2019b), World Bank (2019a), Barro and Lee (2018) annualized growth of the HDI in a given period, and IMF (2019). For HDI values that are comparable across years and Human Development Index (HDI): A composite calculated as the annual compound growth rate. countries, use this table or the interpolated data at index measuring average achievement in three basic Column 9: Calculated based on data in columns http://hdr.undp.org/en/data, which present trends dimensions of human development— a long and MAIN DATA SOURCES 4 and 8. using consistent data. healthy life, knowledge and a decent standard of living. See Technical note 1 at http://hdr.undp.org/ Columns 1­8: HDRO calculations based on data Columns 10­13: Calculated based on data in a A positive value indicates an improvement in rank. sites/default/files/hdr2019_technical_notes.pdf for from UNDESA (2019b), UNESCO Institute for columns 1, 2, 3 and 8. details on how the HDI is calculated. Statistics (2019), United Nations Statistics Division

TABLE 2 Human Development Index trends, 1990­2018 | 307 TABLE 3 Inequality-adjusted Human Development Index

Inequality- SDG 10.1

adjusted Inequality- Inequality- Income share Human Coefficient Inequality life Inequality adjusted Inequality adjusted Index (HDI) Inequality-adjusted of human in life expectancy in education in income

Difference (%)

Overall from HDI Poorest Richest Richest Gini Value (%) Value (%) Value (%) Value 40 percent 10 percent 1 percent coefficient

TABLE HDI rank 2018 2018 2018 2018 2018 2015­2020c 2018 2018d 2018 2018d 2018 2010­2017e 2010­2017e 2010­2017e 2010­2017e

3 VERY HIGH HUMAN DEVELOPMENT 0.954 0.889 6.8 0 6.7 3.0 0.929 4.4 0.879 12.7 0.860 23.1 22.3 8.4 27.5 2 Switzerland 0.946 0.882 6.8 ­1 6.6 3.5 0.945 1.9 0.879 14.5 0.825 20.3 25.2 11.9 32.3 4 Germany 0.942 0.865 8.2 ­6 8.0 3.4 0.923 3.5 0.885 16.9 0.793 20.9 25.4 12.8 31.8 6 Australia 0.939 0.861 8.3 ­7 8.1 3.8 0.905 2.7 0.920 17.7 0.765 20.7 24.8 11.1 31.7 8 Sweden 0.939 0.815 13.2 ­17 12.6 2.5 0.970 9.8 0.776 25.6 0.720 … .. 10 Netherlands 0.938 0.862 8.1 ­4 7.9 3.7 0.938 2.7 0.898 17.3 0.761 18.8 27.8 9.1 35.8 12 Finland 0.938 0.885 5.7 4 5.6 2.4 0.944 2.8 0.892 11.7 0.822 23.2 23.5 6.8 27.8 14 New Zealand 0.937 0.874 6.7 2 6.6 2.9 0.936 3.8 0.880 13.0 0.811 22.1 22.9 8.3 29.2 15 United States 0.935 0.810 13.3 ­14 12.8 2.5 0.952 11.0 0.745 25.0 0.750 .. .. 14.0 .. 18 Liechtenstein 0.933 0.870 6.8 2 6.7 3.1 0.926 4.9 0.862 12.1 0.826 22.8 23.0 6.2 28.2 20 Austria 0.930 0.873 6.1 4 6.0 3.6 0.901 3.0 0.892 11.4 0.829 23.3 23.8 12.8 28.2 22 Israel 0.925 0.876 5.3 7 5.2 3.0 0.921 2.3 0.894 10.4 0.816 23.4 22.4 7.3 27.1 24 Slovenia 0.922 0.841 8.8 ­4 8.5 4.6 0.915 2.7 0.867 18.2 0.751 18.9 25.3 13.6 34.0 26 Czechia 0.921 0.836 9.2 ­4 9.1 4.3 0.915 6.4 0.863 16.4 0.740 .. .. 8.2 .. 28 Malta 0.920 0.845 8.2 0 8.0 4.1 0.903 2.8 0.890 17.0 0.750 19.7 25.4 11.7 33.2 30 Estonia 0.920 0.797 13.4 ­13 12.8 6.3 0.848 5.5 0.849 26.6 0.702 15.2 30.6 20.2 41.5 32 Greece 0.919 0.849 7.6 3 7.6 3.6 0.912 7.7 0.824 11.4 0.814 22.6 22.2 6.7 27.7 34 Lithuania 0.917 … … … … .. .. 36 Andorra 0.915 0.882 3.6 15 3.6 2.9 0.963 1.6 0.836 6.3 0.851 20.3 f 24.7 f 10.4 32.1 f 36 Slovakia 0.914 0.843 7.7 3 7.5 3.7 0.910 3.0 0.845 15.9 0.780 21.1 23.8 8.2 30.5 40 Portugal 0.909 0.822 9.5 1 9.3 3.4 0.923 8.0 0.738 16.6 0.817 19.3 25.4 9.1 33.8 42 Chile 0.906 0.809 10.8 ­3 10.2 3.3 0.935 3.7 0.844 23.7 0.671 15.9 27.7 .. 38.9 43 Hungary 0.906 0.777 14.3 ­9 13.9 3.0 0.938 18.5 0.702 20.2 0.712 20.3 23.8 12.2 31.6 46 Croatia 0.902 0.858 4.8 11 4.7 2.9 0.914 2.2 0.874 9.1 0.792 24.1 21.0 6.7 25.4 48 Argentina 0.893 0.765 14.3 ­13 14.0 3.0 0.947 17.1 0.683 21.9 0.692 17.5 26.2 9.8 36.2 50 Belarus 0.891 0.850 4.6 12 4.5 3.0 0.884 1.4 0.880 9.2 0.789 24.4 22.1 9.5 25.9 52 Bulgaria 0.891 0.809 9.2 1 9.1 3.8 0.926 9.1 0.737 14.4 0.777 20.7 26.6 10.8 32.7 52 Romania 0.885 0.815 8.0 6 7.9 4.6 0.915 6.7 0.763 12.5 0.774 21.9 23.6 11.7 29.4 56 Barbados 0.883 0.776 12.1 ­4 11.8 3.1 0.944 11.0 0.706 21.3 0.700 18.0 25.7 7.5 35.4 57 Uruguay 0.882 0.818 7.2 9 7.0 3.6 0.869 2.1 0.862 15.5 0.730 20.0 24.4 7.0 32.7 0.873 0.788 9.7 1 9.6 3.6 0.902 11.0 0.722 14.3 0.751 20.0 27.4 8.6 34.0

0.872 0.766 12.2 ­5 11.9 3.5 0.922 12.8 0.727 19.5 0.671 17.7 26.2 10.8 36.0

0.872 0.801 8.1 4 8.0 4.3 0.862 5.2 0.821 14.4 0.727 21.3 24.6 12.5 30.8

0.869 0.775 10.9 ­1 10.5 5.5 0.810 4.3 0.852 21.8 0.673 17.7 28.6 7.0 37.4

0.866 … .. 5.2 0.843 18.2 0.606 … .. 22.8 ..

0.857 … … 10.0 0.637 … …

0.857 … .. 6.4 0.792 18.0 0.651 … .. 19.7 ..

0.857 0.804 6.2 8 6.1 5.0 0.839 1.6 0.811 11.7 0.764 23.1 20.9 5.2 26.5

0.854 0.776 9.1 3 8.8 5.4 0.803 2.6 0.849 18.5 0.686 19.4 26.1 7.6 34.2

0.850 0.742 12.7 ­6 12.4 3.5 0.918 15.8 0.639 18.1 0.697 18.7 27.3 7.4 35.5

0.848 … .. 5.7 0.872 11.8 0.583 … .. 29.0 ..

0.847 0.696 17.8 ­14 17.0 6.3 0.866 12.0 0.711 32.7 0.548 14.4 37.9 23.7 46.6

0.845 … .. 7.6 0.792 … … .. ..

0.845 0.777 8.0 8 7.8 4.2 0.836 3.2 0.790 16.1 0.711 21.1 23.8 7.7 30.4

0.838 … .. 5.5 0.831 22.7 0.570 … .. 18.0 ..

0.837 0.768 8.3 4 8.1 4.3 0.859 4.9 0.757 15.2 0.697 20.4 23.2 7.6 31.1

0.834 0.725 13.1 ­3 12.0 6.7 0.827 11.9 0.644 20.1 0.714 .. .. 19.5 ..

0.830 0.714 14.0 ­4 13.6 8.6 0.795 6.2 0.790 25.8 0.579 15.3 29.4 .. 40.6

0.824 0.743 9.9 1 9.6 7.1 0.749 3.1 0.807 18.7 0.679 18.0 29.7 20.2 37.7

0.817 0.765 6.4 6 6.3 4.4 0.803 3.7 0.806 10.8 0.692 24.1 21.3 .. 25.4

0.817 0.759 7.1 4 7.1 7.7 0.756 3.2 0.791 10.3 0.732 23.4 23.0 .. 27.5

0.816 0.714 12.5 0 12.1 6.1 0.793 6.3 0.754 23.9 0.607 17.8 28.8 8.4 37.4

0.816 0.746 8.6 5 8.5 3.6 0.842 7.4 0.738 14.6 0.667 20.8 25.7 6.4 31.9

0.816 0.725 11.1 2 10.8 6.3 0.806 5.3 0.722 20.7 0.656 16.9 24.7 6.8 35.9

0.814 … … 1.9 0.829 … …

0.813 0.675 17.0 ­10 15.9 8.7 0.830 5.5 0.730 33.6 0.509 … ..

0.808 … .. 5.9 0.802 22.1 0.487 … .. 19.9 ..

0.808 0.703 13.0 0 12.7 7.9 0.819 8.2 0.684 22.0 0.621 16.5 29.7 14.0 39.5

0.806 0.675 16.2 ­8 16.1 9.0 0.804 16.5 0.594 22.6 0.645 15.6 32.1 23.4 41.9

308 | HUMAN DEVELOPMENT REPORT 2019

Inequality- SDG 10.1

adjusted Inequality- Inequality- Income share Human Coefficient Inequality life Inequality adjusted Inequality adjusted Index (HDI) Inequality-adjusted of human in life expectancy in education in income

Difference (%)

Overall from HDI Poorest Richest Richest Gini Value (%) Value (%) Value (%) Value 40 percent 10 percent 1 percent coefficient

HDI rank 2018 2018 2018 2018 2018 2015­2020c 2018 2018d 2018 2018d 2018 2010­2017e 2010­2017e 2010­2017e 2010­2017e TABLE

60 Bahamas 0.805 … .. 6.8 0.771 6.3 0.694 … … 3 61 Malaysia 0.804 .. 62 Seychelles … .. 6.1 0.809 12.1 0.627 .. 15.9 31.3 14.5 41.0 63 Serbia 0.801 … .. 9.6 0.742 .. .. 29.3 0.590 15.2 39.9 .. 46.8 65 Iran (Islamic Republic of) 0.799 0.685 14.4 ­4 13.7 4.9 0.817 8.1 0.719 28.1 0.546 22.5 23.1 6.4 28.5 67 Panama 0.799 … .. 14.9 0.699 … … .. .. 69 Albania 0.797 0.706 11.5 5 11.3 9.2 0.789 5.0 0.706 19.7 0.631 16.6 30.9 16.3 40.0 71 Sri Lanka 0.796 0.688 13.7 0 13.6 9.4 0.765 13.2 0.634 18.2 0.671 19.2 29.0 7.1 35.8 73 Saint Kitts and Nevis 0.795 0.626 21.2 ­13 20.3 12.0 0.790 12.5 0.610 36.5 0.510 11.5 37.7 .. 49.9 75 Bosnia and Herzegovina 0.794 0.645 18.7 ­7 18.0 7.1 0.859 14.7 0.611 32.2 0.511 12.8 37.0 .. 48.3 77 Thailand 0.791 0.705 10.9 8 10.9 7.2 0.835 12.3 0.665 13.2 0.631 22.1 22.9 6.4 29.0 79 Brazil 0.786 0.692 12.0 5 11.6 7.9 0.759 3.2 0.828 23.6 0.526 17.4 28.9 .. 37.9 81 Armenia 0.780 0.686 12.1 4 11.8 7.0 0.813 7.4 0.700 21.0 0.567 17.7 32.9 .. 39.8 82 North Macedonia 0.778 … .. 5.1 0.857 10.9 0.704 … … 85 China 0.777 … … … … .. .. 87 Azerbaijan 0.776 … .. 5.8 0.824 … … .. .. 89 Dominican Republic 0.769 0.658 14.4 ­2 14.2 5.4 0.833 17.0 0.586 20.2 0.584 19.8 25.1 6.2 33.0 91 Tunisia 0.767 0.595 22.5 ­17 21.8 10.5 0.757 18.5 0.558 36.3 0.498 15.5 34.8 .. 43.4 93 Lebanon 0.765 0.635 16.9 ­4 16.7 7.9 0.807 18.3 0.543 23.8 0.585 18.4 28.4 20.2 36.5 94 Saint Vincent and the Grenadines 0.763 … .. 11.2 0.716 … … .. .. 96 Venezuela (Bolivarian Republic of) 0.761 0.574 24.5 ­23 23.8 10.9 0.763 23.8 0.525 36.7 0.473 10.6 41.9 28.3 53.3 98 Fiji 0.761 0.585 23.1 ­16 22.4 10.7 0.785 20.3 0.545 36.2 0.468 12.4 39.0 20.5 49.7 98 Suriname 0.760 0.685 9.9 9 9.7 8.7 0.772 2.9 0.737 17.4 0.565 20.8 28.4 .. 33.6 103 Belize 0.759 0.604 20.4 ­8 19.7 14.1 0.749 33.7 0.448 11.4 0.658 23.1 22.9 .. 27.6 105 Tonga 0.759 0.660 13.1 5 12.9 7.9 0.789 10.5 0.623 20.3 0.585 17.3 24.8 5.8 35.6 107 Moldova (Republic of) 0.759 0.612 19.4 ­5 19.1 10.8 0.776 18.1 0.567 28.3 0.521 14.4 32.3 .. 43.3 108 Uzbekistan 0.758 0.636 16.1 4 15.7 7.9 0.803 11.7 0.573 27.4 0.558 17.0 29.4 13.9 38.6 111 Indonesia 0.758 0.607 19.9 ­4 19.5 11.5 0.773 16.5 0.596 30.5 0.485 14.1 33.8 .. 44.7 113 South Africa 0.754 0.683 9.4 13 9.3 13.9 0.700 5.3 0.657 8.9 0.692 … .. 115 Gabon 0.750 0.701 6.5 21 6.5 7.4 0.740 3.6 0.768 8.5 0.605 24.5 21.2 .. 25.0 MEDIUM HUMAN DEVELOPMENT 0.745 0.584 21.5 ­8 21.4 17.0 0.688 19.1 0.532 28.1 0.545 13.9 35.4 .. 45.7 0.745 0.617 17.2 4 16.9 10.6 0.771 12.6 0.584 27.4 0.521 11.0 38.6 .. 51.2

0.739 0.585 20.8 ­4 20.2 9.0 0.791 32.8 0.442 18.9 0.573 20.1 25.6 .. 32.8

0.735 0.635 13.6 10 13.6 13.1 0.664 11.9 0.646 15.7 0.596 20.4 25.6 .. 32.3

0.730 … .. 7.4 0.839 6.2 0.566 .. .. 20.6 24.8 23.4 31.8

0.728 … .. 19.4 0.611 … .. 10.9 41.5 .. 53.3

0.728 … .. 11.3 0.715 … … .. ..

0.726 0.604 16.7 3 15.9 10.0 0.753 5.6 0.653 32.0 0.449 … ..

0.726 0.600 17.3 1 17.0 17.1 0.665 8.8 0.638 25.2 0.510 … ..

0.724 … … … … .. ..

0.724 … .. 14.9 0.620 … .. 18.8 29.7 .. 36.7

0.724 0.545 24.7 ­14 23.8 13.8 0.718 18.1 0.519 39.5 0.435 13.2 39.2 .. 48.8

0.720 0.557 22.7 ­9 21.9 12.8 0.692 15.6 0.551 37.3 0.453 … ..

0.723 0.617 14.7 11 14.7 10.6 0.748 15.4 0.574 17.9 0.547 20.3 27.5 16.1 33.7

0.720 0.558 22.6 ­8 21.6 11.1 0.745 15.9 0.582 37.9 0.400 … ..

0.719 0.568 21.0 ­5 20.4 6.0 0.848 29.3 0.399 25.8 0.541 17.4 g 29.9 g .. 38.4 g

0.717 … .. 10.4 0.700 4.5 0.736 .. .. 18.2 29.7 .. 37.6

0.712 0.582 18.2 1 17.8 15.3 0.666 10.1 0.599 28.1 0.495 16.8 31.3 .. 40.1

0.711 0.638 10.4 21 10.3 9.6 0.721 7.3 0.656 14.0 0.549 24.1 21.7 6.1 25.9

0.710 0.579 18.5 1 17.9 23.4 0.567 3.6 0.606 26.8 0.564 … ..

0.710 … .. 13.9 0.683 0.7 0.713 … …

0.708 … .. 9.1 0.737 … … .. ..

0.707 0.584 17.4 6 17.4 13.9 0.682 18.2 0.511 20.1 0.570 17.5 29.5 .. 38.1

0.707 … .. 10.0 0.736 4.9 0.666 .. .. 17.9 31.3 .. 38.7

0.705 0.463 34.4 ­17 31.4 19.2 0.545 17.3 0.596 57.7 0.305 7.2 50.5 19.2 63.0

0.703 0.533 24.2 ­6 24.1 22.5 0.611 20.0 0.552 29.7 0.449 13.6 31.7 .. 44.0

0.702 0.544 22.5 ­4 22.5 22.8 0.549 23.5 0.486 21.2 0.602 16.8 27.7 .. 38.0

0.700 0.492 29.7 ­8 28.7 11.6 0.705 38.1 0.376 36.5 0.449 21.9 27.8 19.1 31.8

0.698 … … 4.3 0.677 … …

TABLE 3 Inequality-adjusted Human Development Index | 309 TABLE 3 INEQUALITY-ADJUSTED HUMAN DEVELOPMENT INDEX

Inequality- SDG 10.1

adjusted Inequality- Inequality- Income share Human Coefficient Inequality life Inequality adjusted Inequality adjusted Index (HDI) Inequality-adjusted of human in life expectancy in education in income

Difference (%)

Overall from HDI Poorest Richest Richest Gini Value (%) Value (%) Value (%) Value 40 percent 10 percent 1 percent coefficient

TABLE HDI rank 2018 2018 2018 2018 2018 2015­2020c 2018 2018d 2018 2018d 2018 2010­2017e 2010­2017e 2010­2017e 2010­2017e

3 118 Viet Nam 0.693 0.580 16.3 8 16.2 12.9 0.741 17.6 0.515 18.1 0.511 18.8 27.1 .. 35.3 120 Iraq 0.690 0.597 13.5 16 13.5 12.0 0.730 11.9 0.582 16.6 0.500 19.2 25.2 15.8 33.7 122 Kyrgyzstan 0.689 0.552 19.8 3 19.4 15.9 0.653 29.7 0.389 12.7 0.664 21.9 23.7 22.0 29.5 124 El Salvador 0.676 … .. 13.0 0.756 .. .. 21.7 0.510 17.4 31.9 .. 39.5 126 Cabo Verde 0.674 0.610 9.5 23 9.5 11.3 0.700 5.0 0.697 12.2 0.465 23.6 23.3 .. 27.3 126 Nicaragua 0.670 0.546 18.5 4 18.3 19.0 0.620 10.7 0.537 25.1 0.490 … .. 130 Namibia 0.667 0.521 21.9 1 21.6 12.5 0.715 29.1 0.401 23.2 0.492 17.4 29.1 .. 38.0 132 Honduras 0.656 0.574 12.5 12 12.4 16.7 0.652 6.0 0.632 14.5 0.459 19.4 26.4 .. 34.0 134 Bhutan 0.651 … .. 12.2 0.713 23.7 0.410 … .. .. 47.2 135 Micronesia (Federated States of) 0.651 0.472 27.4 ­2 26.9 14.6 0.710 30.8 0.353 35.4 0.420 13.1 38.0 .. 48.3 138 Congo 0.651 0.501 23.0 1 22.7 13.1 0.726 25.7 0.420 29.2 0.414 14.3 37.2 .. 46.2 140 Lao People’s Democratic Republic 0.647 0.477 26.3 1 25.7 19.7 0.610 38.7 0.342 18.8 0.518 19.8 30.1 21.3 35.7 142 Ghana 0.645 0.417 35.3 ­14 33.6 22.1 0.520 25.0 0.437 53.6 0.321 8.6 47.3 .. 59.1 144 Equatorial Guinea 0.626 0.450 28.0 ­5 26.7 21.7 0.593 44.9 0.273 13.6 0.564 22.8 24.0 .. 28.7 146 Cambodia 0.623 0.464 25.5 0 25.0 13.3 0.735 26.6 0.369 34.9 0.369 11.0 37.7 .. 50.5 147 Nepal 0.623 … .. 24.7 0.557 … … .. .. 150 Cameroon 0.617 0.450 27.1 ­3 26.3 17.1 0.656 41.7 0.257 20.0 0.539 17.5 27.9 .. 37.4 152 Pakistan 0.614 0.465 24.3 4 23.6 17.3 0.666 37.7 0.320 15.7 0.472 21.0 26.8 .. 32.4 LOW HUMAN DEVELOPMENT 0.614 … .. 16.1 0.616 .. .. 26.4 0.402 16.2 29.7 .. 40.1 155 Papua New Guinea 0.609 0.507 16.7 10 16.7 17.0 0.641 18.3 0.463 14.9 0.438 21.1 24.2 .. 30.8 157 Rwanda 0.608 0.456 25.0 2 24.9 22.8 0.526 20.9 0.426 31.0 0.423 12.4 37.9 .. 48.9 159 Tanzania (United Republic of) 0.608 0.430 29.3 ­4 29.0 25.1 0.454 24.1 0.411 37.9 0.426 11.5 g 40.0 g .. 51.5 g 161 Mauritania 0.604 0.454 24.9 3 24.7 22.6 0.567 31.3 0.330 20.3 0.499 19.1 29.8 .. 36.4 162 Madagascar .. 37.6 163 Benin 0.597 … .. 14.4 0.663 .. .. 19.7 0.405 17.8 29.4 .. 43.5 164 Lesotho .. 57.1 165 Côte d’Ivoire 0.596 0.427 28.3 ­3 28.1 24.2 0.511 34.9 0.364 25.3 0.419 14.3 32.2 167 Togo 0.591 0.394 33.4 ­6 32.3 26.5 0.492 21.7 0.448 48.6 0.278 8.9 44.4 169 Haiti 0.588 … .. 34.6 0.386 … … .. .. 171 Djibouti 0.584 0.448 23.2 3 23.2 22.8 0.557 26.9 0.330 19.9 0.490 18.6 31.7 .. 38.1 173 Ethiopia 0.581 0.465 20.1 12 19.9 18.1 0.625 27.3 0.346 14.3 0.464 … .. 174 Guinea 0.579 0.426 26.3 0 26.2 22.5 0.553 22.9 0.406 33.1 0.345 16.5 31.6 .. 40.8 0.579 0.430 25.8 3 24.9 17.5 0.641 40.9 0.296 16.3 0.419 20.4 26.4 .. 32.8

0.574 0.392 31.8 ­2 31.7 32.0 0.427 34.3 0.327 28.9 0.432 15.0 f 32.3 f .. 42.7 f

0.563 0.371 34.1 ­6 34.1 33.5 0.398 33.0 0.378 35.9 0.338 13.0 35.0 .. 46.6

0.563 0.435 22.8 7 22.7 24.2 0.480 16.8 0.473 27.0 0.362 15.3 33.8 .. 43.2

0.560 0.386 31.1 ­1 30.2 29.9 0.508 43.5 0.230 17.2 0.494 21.1 28.9 .. 33.5

0.557 … .. 12.1 0.714 .. .. 19.4 0.366 18.4 29.2 .. 37.1

0.549 … .. 13.0 0.693 … … 14.7 ..

0.543 … .. 24.1 0.517 11.5 0.382 .. .. 15.1 g 31.0 g .. 41.9 g

0.538 0.294 45.3 ­22 44.2 28.9 0.483 47.6 0.249 56.0 0.212 13.6 33.7 .. 45.3

0.536 0.382 28.7 ­1 28.4 19.5 0.603 29.3 0.324 36.4 0.286 15.8 35.6 .. 43.7

0.534 0.349 34.6 ­5 34.5 37.1 0.332 38.1 0.301 28.2 0.426 15.1 g 32.7 g .. 43.0 g

0.528 0.397 24.9 7 24.9 25.3 0.517 27.0 0.309 22.4 0.391 18.5 31.0 .. 37.8

0.528 0.387 26.7 4 26.7 27.2 0.481 27.9 0.371 24.9 0.325 15.9 34.2 .. 42.8

0.527 0.358 32.1 1 31.8 30.0 0.481 40.8 0.230 24.6 0.413 19.9 24.9 .. 32.6

0.521 0.386 25.8 6 25.5 21.1 0.567 35.0 0.320 20.4 0.318 15.7 33.5 .. 42.6

0.520 0.327 37.1 ­6 36.9 34.9 0.415 43.7 0.268 32.0 0.315 12.8 37.6 .. 47.8

0.518 0.350 32.5 3 32.0 33.1 0.347 21.9 0.398 41.1 0.310 9.6 40.9 .. 54.2

0.516 0.331 35.8 ­3 35.0 33.3 0.384 47.4 0.232 24.4 0.409 15.9 31.9 17.1 41.5

0.514 0.347 32.5 2 31.6 21.2 0.578 46.0 0.190 27.7 0.381 16.4 31.0 .. 40.3

0.513 0.350 31.7 6 31.5 30.5 0.436 38.9 0.314 25.1 0.313 14.5 31.6 .. 43.1

0.507 0.332 34.6 1 34.3 27.4 0.504 42.5 0.195 33.0 0.372 18.5 g 26.7 g .. 35.4 g

0.503 0.299 40.5 ­7 40.0 32.2 0.455 37.3 0.279 50.4 0.211 15.8 31.2 .. 41.1

0.496 … .. 28.3 0.491 45.4 0.225 … …

0.495 … .. 23.4 0.549 .. .. 27.7 0.391 15.8 32.3 .. 41.6

0.485 0.346 28.7 5 28.6 25.1 0.505 28.4 0.328 32.4 0.250 16.2 38.1 .. 44.7

0.470 0.337 28.4 5 27.3 24.9 0.534 43.5 0.189 13.4 0.377 17.6 31.4 .. 39.1

0.466 0.293 37.2 ­8 36.4 28.5 0.459 49.3 0.195 31.5 0.279 19.0 28.7 .. 35.9

0.466 0.310 33.4 ­1 32.2 31.3 0.435 48.3 0.176 17.1 0.388 19.8 26.4 .. 33.7

0.465 0.314 32.3 2 31.8 29.8 0.472 42.9 0.241 22.7 0.273 18.8 27.1 .. 35.3

310 | HUMAN DEVELOPMENT REPORT 2019

Inequality- SDG 10.1

adjusted Inequality- Inequality- Income share Human Coefficient Inequality life Inequality adjusted Inequality adjusted Index (HDI) Inequality-adjusted of human in life expectancy in education in income

Difference (%)

Overall from HDI Poorest Richest Richest Gini Value (%) Value (%) Value (%) Value 40 percent 10 percent 1 percent coefficient

HDI rank 2018 2018 2018 2018 2018 2015­2020c 2018 2018d 2018 2018d 2018 2010­2017e 2010­2017e 2010­2017e 2010­2017e TABLE 0.187 21.8 0.315 177 Yemen 0.463 0.316 31.8 5 30.9 24.7 0.534 46.1 0.233 37.9 0.260 18.8 29.4 15.7 36.7 3 179 Congo (Democratic Republic of the) 0.461 0.288 37.5 ­5 37.4 32.3 0.396 41.9 12.8 42.0 .. 50.7 181 Sierra Leone 0.459 0.316 31.0 7 30.9 36.1 0.397 28.5 0.354 28.2 0.225 15.5 32.0 .. 42.1 182 Eritrea 0.446 0.309 30.7 4 30.7 29.8 0.434 33.8 0.257 28.4 0.265 11.8 45.5 .. 54.0 185 Burundi 0.438 0.282 35.7 ­3 34.6 39.0 0.322 46.9 0.214 17.7 0.326 19.8 26.9 .. 34.0 187 Chad 0.434 0.303 30.1 5 29.5 32.0 0.431 39.2 0.183 17.3 0.354 20.0 29.6 .. 35.3 189 Niger 0.434 … .. 21.4 0.556 … … .. .. 0.427 0.294 31.2 3 30.4 36.7 0.379 39.2 0.176 15.4 0.381 20.1 g 25.7 g .. 33.0 g .. Monaco 0.423 0.296 30.1 5 29.6 28.5 0.454 39.5 0.253 20.9 0.225 17.9 31.0 .. 38.6 .. San Marino 0.413 0.264 36.1 ­1 36.0 36.2 0.369 39.6 0.182 32.3 0.274 12.5 g 33.2 g .. 46.3 g .. Tuvalu 0.401 0.250 37.7 ­1 37.4 40.9 0.309 43.0 0.164 28.4 0.307 14.6 32.4 .. 43.3 0.381 0.222 41.6 ­1 41.3 40.1 0.302 34.5 0.231 49.2 0.157 10.3 f 46.2 f .. 56.2 f High human development 0.377 0.272 27.9 3 27.4 30.9 0.447 35.0 0.161 16.4 0.279 19.6 27.0 .. 34.3 Low human development … .. .. 11.5 0.709 … … .. .. Regions … … … … … East Asia and the Pacific … … … 23.9 0.592 … .. Latin America and the Caribbean … … … … … Sub-Saharan Africa … .. .. 38.9 0.348 … … .. .. Small island developing states … … .. 10.5 … 17.4 30.7 .. 39.1 Cooperation and Development 0.892 0.796 10.7 — 10.5 5.2 0.868 7.0 0.796 19.3 0.730 18.2 27.6 14.9 — World 0.750 0.615 17.9 — 0.634 0.470 25.9 — 17.6 10.0 0.764 14.8 0.563 27.9 0.541 16.6 31.1 .. — 0.507 0.349 31.1 — 0.686 0.533 22.3 — 25.4 20.5 0.604 36.3 0.342 19.6 0.502 19.4 29.9 .. —

30.9 30.4 0.442 37.4 0.261 25.0 0.368 16.4 32.1 .. —

22.2 16.6 0.655 25.6 0.435 24.3 0.532 17.6 30.8 .. —

0.703 0.531 24.5 — 24.2 15.0 0.679 32.5 0.386 25.0 0.571 20.6 26.9 .. — 0.741 0.618 16.6 — 0.779 0.688 11.7 — 16.3 9.8 0.766 13.5 0.550 25.6 0.560 17.2 29.5 .. — 0.759 0.589 22.3 — 0.642 0.476 25.9 — 11.6 9.7 0.753 8.3 0.682 16.8 0.634 19.9 26.7 .. — 0.541 0.376 30.5 — 0.528 0.377 28.6 — 21.7 11.6 0.754 19.5 0.553 34.1 0.491 13.1 37.3 .. — 0.723 0.549 24.0 — 25.3 20.2 0.611 37.5 0.340 18.4 0.520 19.9 29.7 .. — 0.895 0.791 11.7 — 0.731 0.584 20.2 — 30.4 29.7 0.445 34.0 0.308 27.6 0.387 15.4 33.8 .. —

28.4 26.3 0.510 36.3 0.275 22.5 0.383 17.6 31.1 .. —

23.6 16.6 0.665 19.7 0.503 34.3 0.496 … --

11.4 5.3 0.880 8.0 0.783 20.9 0.717 18.0 28.0 14.2 —

20.1 14.7 0.690 22.3 0.492 23.3 0.586 17.7 30.2 .. —

NOTES Overall loss: Percentage difference between the distribution based on data from household surveys Column 5: Calculated as the arithmetic mean of the IHDI value and the HDI value. listed in Main data sources. values in inequality in life expectancy, inequality a See http://hdr.undp.org/en/composite/IHDI for in education and inequality in income using the the list of surveys used to estimate inequalities. Difference from HDI rank: Difference in ranks on Income share: Percentage of income (or methodology in Technical note 2 (available at http:// the IHDI and the HDI, calculated only for countries consumption) that accrues to the indicated hdr.undp.org/sites/default/files/hdr2019_technical_ b Based on countries for which an Inequality-adjusted for which an IHDI value is calculated. population subgroups of population. notes.pdf). Human Development Index value is calculated. Coefficient of human inequality: Average Gini coefficient: Measure of the deviation of Column 6: Calculated based on abridged life tables c Calculated by HDRO from the 2015­2020 period inequality in three basic dimensions of human the distribution of income among individuals or from UNDESA (2019b). life tables from UNDESA (2019b). development. households within a country from a perfectly equal distribution. A value of 0 represents absolute Column 7: Calculated based on inequality in life d Data refer to 2018 or the most recent year Inequality in life expectancy: Inequality in equality, a value of 100 absolute inequality. expectancy and the HDI life expectancy index. available. distribution of expected length of life based on data from life tables estimated using the Atkinson MAIN DATA SOURCES Columns 8 and 10: Calculated based on data from e Data refer to the most recent year available inequality index. the Luxembourg Income Study database, Eurostat’s during the period specified. Column 1: HDRO calculations based on data from European Union Statistics on Income and Living Inequality-adjusted life expectancy index: HDI UNDESA (2019b), UNESCO Institute for Statistics Conditions, the World Bank’s International Income f Refers to 2008. life expectancy index value adjusted for inequality in (2019), United Nations Statistics Division (2019b), Distribution Database, the Center for Distributive, distribution of expected length of life based on data World Bank (2019a), Barro and Lee (2018) and IMF Labor and Social Studies and the World Bank’s g Refers to 2009. from life tables listed in Main data sources. (2019). Socio-Economic Database for Latin America and the Caribbean, ICF Macro Demographic and Health DEFINITIONS Inequality in education: Inequality in distribution Column 2: Calculated as the geometric mean of the Surveys and United Nations Children’s Fund Multiple of years of schooling based on data from household values in inequality-adjusted life expectancy index, Indicator Cluster Surveys using the methodology in Human Development Index (HDI): A composite surveys estimated using the Atkinson inequality index. inequality-adjusted education index and inequality- Technical note 2 (available at http://hdr.undp.org/ index measuring average achievement in three basic adjusted income index using the methodology in sites/default/files/hdr2019_technical_notes.pdf). dimensions of human development—a long and Inequality-adjusted education index: HDI Technical note 2 (available at http://hdr.undp.org/ healthy life, knowledge and a decent standard of education index value adjusted for inequality in sites/default/files/hdr2019_technical_notes.pdf). Column 9: Calculated based on inequality in living. See Technical note 1 at http://hdr.undp.org/ distribution of years of schooling based on data from education and the HDI education index. sites/default/files/hdr2019_technical_notes.pdf for household surveys listed in Main data sources. Column 3: Calculated based on data in columns details on how the HDI is calculated. 1 and 2. Column 11: Calculated based on inequality in Inequality in income: Inequality in income income and the HDI income index. Inequality-adjusted HDI (IHDI): HDI value adjusted distribution based on data from household surveys Column 4: Calculated based on IHDI values and for inequalities in the three basic dimensions of estimated using the Atkinson inequality index. recalculated HDI ranks for countries for which an Columns 12, 13 and 15: World Bank (2019a). human development. See Technical note 2 at http:// IHDI value is calculated. hdr.undp.org/sites/default/files/hdr2019_technical_ Inequality-adjusted income index: HDI income Column 14: World Inequality Database (2019). notes.pdf for details on how the IHDI is calculated. index value adjusted for inequality in income

TABLE 3 Inequality-adjusted Human Development Index | 311 TABLE 4 Gender Development Index

Gender Development Human Development SDG 3 SDG 4.3 SDG 4.6 SDG 8.5 Life expectancy Expected years Mean years Estimated gross national at birth of schooling of schooling income per capitaa

Value (years) (years) (years) (2011 PPP $)

Value Groupb Female Male Female Male Female Male Female Male Female Male

HDI rank 2018 2018 2018 2018 2018 2018 2018c 2018c 2018c 2018c 2018 2018

VERY HIGH HUMAN DEVELOPMENT 0.990 1 0.946 0.955 84.3 80.3 18.8 d 17.4 12.6 12.5 60,283 75,688 e 1 Norway 0.963 2 0.924 0.959 16.1 16.3 12.7 13.6 49,275 69,649 2 Switzerland 0.975 2 0.929 0.953 85.5 81.7 18.9 d 18.7 d 12.7 f 12.3 f 44,921 66,583 3 Ireland 0.968 2 0.923 0.953 17.0 17.2 13.7 14.6 38,470 55,649 4 Germany 0.963 2 0.919 0.954 83.7 80.4 16.4 16.6 11.6 12.5 43,852 79,385 e 4 Hong Kong, China (SAR) 0.975 1 0.926 0.949 22.6 d 21.6 d 12.7 f 12.6 f 35,900 52,359 6 Australia 0.966 2 0.921 0.954 83.6 78.8 20.4 d 18.0 d 12.3 f 12.7 f 39,246 55,824 6 Iceland 0.982 1 0.928 0.945 19.6 d 18.0 d 12.5 12.3 41,919 53,979 TABLE 8 Sweden 0.988 1 0.929 0.941 87.6 81.8 16.5 16.1 11.1 12.0 74,600 92,163 e 9 Singapore 0.967 2 0.916 0.947 18.3 d 17.8 11.9 12.5 40,573 59,536 4 10 Netherlands 0.980 1 0.920 0.938 85.3 81.3 19.8 d 18.4 d 12.7 12.4 41,026 56,732 11 Denmark 0.990 1 0.920 0.929 20.1 d 18.5 d 12.6 12.3 35,066 48,689 12 Finland 0.989 1 0.916 0.926 84.4 81.3 16.6 15.6 13.5 f 13.1 f 35,118 52,221 13 Canada 0.963 2 0.902 0.936 19.7 d 17.9 12.6 f 12.8 f 26,754 43,745 14 New Zealand 0.967 2 0.904 0.935 84.4 80.9 18.0 d 17.1 12.9 g 13.0 g 28,526 50,771 15 United Kingdom 0.991 1 0.915 0.923 16.9 15.7 13.5 13.4 44,465 68,061 15 United States 0.972 2 0.904 0.931 85.6 81.3 20.6 d 18.8 d 11.6 11.9 34,928 52,927 17 Belgium .. 13.4 16.1 18 Liechtenstein .. 1 .. .. 83.8 80.4 15.2 15.3 … .. 19 Japan 0.976 2 0.901 0.923 16.6 16.0 13.0 h 12.6 h 28,784 53,384 20 Austria 0.963 2 0.895 0.929 82.8 78.8 14.3 14.1 12.3 13.0 32,618 60,303 21 Luxembourg 0.970 2 0.893 0.921 16.6 15.4 11.8 g 12.6 g 53,006 77,851 e 22 Israel 0.972 3 0.891 0.917 84.6 78.9 15.8 16.9 13.0 13.0 24,616 42,792 22 Korea (Republic of) 0.934 1 0.870 0.932 18.2 d 16.7 11.5 12.9 23,228 50,241 24 Slovenia 1.003 1 0.902 0.899 84.3 80.3 18.2 d 17.5 12.2 12.3 28,832 35,487 25 Spain 0.981 1 0.882 0.899 17.6 16.1 9.7 10.0 28,086 42,250 26 Czechia 0.983 1 0.882 0.897 83.9 80.4 15.8 15.2 12.5 13.0 24,114 39,327 26 France 0.984 2 0.883 0.897 16.4 15.4 11.2 11.6 33,002 48,510 28 Malta 0.965 2 0.867 0.899 83.0 79.5 16.6 15.9 11.0 11.6 25,023 44,518 29 Italy 0.967 1 0.866 0.895 16.8 15.3 10.0 g 10.5 g 26,471 46,360 30 Estonia 1.016 1 0.886 0.872 81.4 76.3 15.1 14.3 13.4 f 12.6 f 22,999 38,653 31 Cyprus 0.983 2 0.865 0.880 17.1 17.5 12.0 12.2 27,791 38,404 32 Greece 0.963 1 0.854 0.887 83.8 79.1 17.3 15.6 10.3 10.8 19,747 30,264 32 Poland 1.009 2 0.874 0.867 16.9 16.1 12.3 12.3 21,876 33,739 34 Lithuania 1.028 2 0.880 0.856 .. .. 14.3 13.4 13.0 g 13.0 g 25,665 34,560 35 United Arab Emirates 0.965 .. 0.832 0.862 12.0 9.8 24,211 85,772 e 36 Andorra 5 87.5 81.3 .. .. 10.1 10.2 36 Saudi Arabia .. 1 .. .. 15.8 g 17.6 g 9.0 g 10.1 g .. .. 36 Slovakia 0.879 2 0.784 0.892 83.8 79.0 15.0 14.1 12.5 f 12.7 f 18,166 72,328 39 Latvia 0.992 1 0.852 0.859 16.7 15.3 13.1 f 12.5 f 23,683 38,045 40 Portugal 1.030 2 0.865 0.840 84.2 80.0 16.2 16.4 9.2 9.2 21,857 31,520 41 Qatar 0.984 2 0.843 0.856 14.1 11.1 11.1 9.3 23,627 32,738 42 Chile 1.043 1 0.873 0.837 84.4 81.1 16.8 16.3 10.3 10.6 57,209 127,774 e 43 Brunei Darussalam 0.962 1 0.828 0.860 14.8 14.0 9.1 h 9.1 h 15,211 28,933 43 Hungary 0.987 3 0.837 0.848 85.8 79.7 15.4 14.8 11.7 12.1 65,914 86,071 e 45 Bahrain 0.984 1 0.836 0.850 16.1 14.7 9.3 g 9.5 g 21,010 33,906 46 Croatia 0.937 3 0.800 0.854 83.9 78.4 15.7 14.3 10.9 g 12.0 g 18,422 52,949 47 Oman 0.989 1 0.832 0.842 15.5 14.1 10.6 9.4 19,441 26,960 48 Argentina 0.943 1 0.793 0.841 86.1 80.7 18.9 d 16.4 10.7 f 10.5 f 11,435 50,238 49 Russian Federation 0.988 1 0.818 0.828 15.9 15.2 11.9 g 12.1 g 12,084 23,419 50 Belarus 1.015 1 0.828 0.816 81.8 76.6 15.7 15.0 12.2 i 12.4 i 19,969 30,904 50 Kazakhstan 1.010 1 0.820 0.811 15.6 14.9 11.9 h 11.7 h 13,923 20,616 52 Bulgaria 0.999 2 0.814 0.815 85.4 79.6 15.0 14.6 11.9 11.8 16,492 28,197 52 Montenegro 0.993 1 0.812 0.818 15.3 14.7 10.7 g 12.0 g 15,621 23,905 52 Romania 0.966 .. 0.801 0.829 84.1 80.5 14.6 13.9 10.6 11.3 14,457 20,634 55 Palau 0.986 1 0.809 0.821 16.3 g 15.0 g 19,487 28,569 56 Barbados 1 85.4 81.1 16.6 g 13.8 g .. .. 57 Kuwait .. 1 .. .. 14.3 12.9 10.9 j 10.3 j .. .. 57 Uruguay 1.010 4 0.816 0.808 82.6 74.1 17.1 15.1 8.0 6.9 13,686 18,292 59 Turkey 0.999 .. 0.802 0.803 15.9 g 16.9 g 9.0 8.4 49,067 85,620 e 60 Bahamas 1.016 2 0.810 0.797 82.9 78.7 6.9 8.4 14,901 24,292 61 Malaysia 0.924 0.771 0.834 .. .. 11.7 g 11.4 g 15,921 34,137 84.5 79.6 13.8 13.1 10.0 10.3 22,830 34,288 … 20,820 33,279 0.972 0.792 0.815 82.4 74.6

81.2 70.1

79.2 77.1

76.6 73.8

80.8 73.8

79.9 70.1

84.7 78.8

81.9 79.0

82.4 77.6

77.0 74.6

80.1 73.1

78.3 76.3

81.5 75.1

80.1 75.9

79.9 73.1

77.6 66.9

79.4 69.4

77.3 68.8

78.5 71.4

79.2 74.3

79.4 72.5

80.4 77.7

76.5 74.7

81.4 74.0

80.3 74.4

75.9 71.5

78.2 74.1

312 | HUMAN DEVELOPMENT REPORT 2019

Gender Development Human Development SDG 3 SDG 4.3 SDG 4.6 SDG 8.5 Life expectancy Expected years Mean years Estimated gross national at birth of schooling of schooling income per capitaa

Value (years) (years) (years) (2011 PPP $)

Value Groupb Female Male Female Male Female Male Female Male Female Male

HDI rank 2018 2018 2018 2018 2018 2018 2018c 2018c 2018c 2018c 2018 2018 62 Seychelles … .. 77.3 69.8 16.2 14.7 63 Serbia 0.976 1 0.789 0.808 78.5 73.3 15.3 14.3 10.7 11.6 12,549 17,995 63 Trinidad and Tobago 1.002 1 0.798 0.796 13.8 g 12.0 g 11.1 i 10.9 i 22,266 34,878 65 Iran (Islamic Republic of) 0.874 5 0.727 0.832 76.1 70.8 14.6 14.8 9.9 10.1 5,809 30,250 66 Mauritius 0.974 2 0.782 0.803 15.5 14.4 9.3 h 9.5 h 14,261 31,385 67 Panama 1.005 1 0.794 0.790 77.7 75.4 13.3 12.1 10.4 h 9.9 h 16,106 24,788 68 Costa Rica 0.977 1 0.782 0.800 15.8 14.9 8.8 8.5 10,566 19,015 69 Albania 0.971 2 0.779 0.802 78.4 71.5 15.8 14.8 9.9 j 10.2 j 9,781 14,725 TABLE 70 Georgia 0.979 1 0.775 0.791 15.7 15.2 12.8 12.8 6,505 12,929 71 Sri Lanka 0.938 3 0.749 0.799 81.6 75.2 14.2 13.7 10.5 g 11.6 g 6,766 16,852 4 72 Cuba 0.948 3 0.753 0.794 14.8 13.9 11.8 g 11.7 g 5,035 10,625 73 Saint Kitts and Nevis .. 82.7 77.5 13.8 g 13.5 g 74 Antigua and Barbuda … .. 13.1 g 11.8 g … .. 75 Bosnia and Herzegovina .. 4 .. .. 80.2 76.8 13.9 k 13.5 k … .. 76 Mexico 0.924 2 0.735 0.796 14.6 14.0 8.6 10.9 8,432 17,123 77 Thailand 0.957 1 0.747 0.781 78.0 69.2 14.8 g 14.5 g 8.4 8.8 11,254 24,286 78 Grenada 0.995 .. 0.763 0.766 17.0 16.2 7.5 8.0 14,319 18,033 79 Brazil .. 1 .. .. 80.1 73.4 15.8 15.0 … .. 79 Colombia 0.995 1 0.757 0.761 14.9 14.3 8.1 g 7.6 g 10,432 17,827 81 Armenia 0.986 2 0.755 0.765 80.7 76.8 13.6 g 12.8 g 8.5 8.2 10,236 15,656 82 Algeria 0.972 5 0.746 0.767 14.9 g 14.5 g 11.8 11.8 6,342 12,581 82 North Macedonia 0.865 3 0.685 0.792 .. .. 13.6 13.3 7.7 i 8.3 i 4,103 22,981 82 Peru 0.947 2 0.737 0.778 14.1 13.7 9.2 i 10.2 i 9,464 16,279 85 China 0.951 2 0.738 0.776 78.0 75.7 14.1 g 13.7 g 8.7 9.7 8,839 15,854 85 Ecuador 0.961 1 0.741 0.771 15.7 g 14.1 g 7.5 j 8.3 j 12,665 19,410 87 Azerbaijan 0.980 3 0.748 0.763 79.7 74.8 12.4 12.5 8.9 9.1 7,319 12,960 88 Ukraine 0.940 1 0.728 0.774 15.2 g 14.8 g 10.2 10.8 9,849 20,656 89 Dominican Republic 0.995 1 0.745 0.749 77.8 72.1 14.8 13.5 11.3 j 11.3 j 6,064 10,232 89 Saint Lucia 1.003 2 0.744 0.742 14.2 g 13.6 g 8.3 7.6 11,176 18,974 91 Tunisia 0.975 5 0.734 0.753 80.7 73.2 15.8 14.4 8.8 8.2 9,085 14,046 92 Mongolia 0.899 2 0.689 0.767 14.8 g 13.7 g 6.4 g 7.9 g 4,737 16,722 93 Lebanon 1.031 5 0.746 0.724 74.9 70.1 11.4 11.6 10.5 g 9.9 g 9,666 11,931 94 Botswana 0.891 1 0.678 0.762 12.8 g 12.6 g 8.5 l 8.9 l 4,667 17,530 94 Saint Vincent and the Grenadines 0.990 .. 0.723 0.731 79.4 72.0 13.7 g 13.4 g 9.2 j 9.5 j 14,176 17,854 96 Jamaica .. 1 .. .. 13.9 g 12.4 g .. .. 8,615 14,780 96 Venezuela (Bolivarian Republic of) 0.986 1 0.719 0.729 79.9 74.3 13.8 g 11.8 g 10.0 g 9.5 g 6,326 9,559 98 Dominica 1.013 .. 0.728 0.719 10.7 10.0 6,655 11,546 98 Fiji … .. 78.4 71.2 … … 98 Paraguay .. 2 … .. 11.0 h 10.7 h 5,839 12,292 98 Suriname 0.968 2 0.710 0.734 77.9 75.5 13.2 g 12.2 g 8.5 8.4 8,325 15,001 102 Jordan 0.972 5 0.710 0.731 13.4 g 12.4 g 9.0 9.2 7,953 15,868 103 Belize 0.868 1 0.654 0.754 77.7 73.7 12.1 g 11.6 g 10.2 h 10.7 h 2,734 13,668 104 Maldives 0.983 3 0.713 0.725 13.4 12.9 9.9 i 9.7 i 5,665 8,619 105 Tonga 0.939 3 0.689 0.734 79.3 73.8 12.2 m 12.0 m 6.7 m 6.9 m 7,454 15,576 106 Philippines 0.944 1 0.692 0.733 14.4 g 13.9 g 11.3 h 11.2 h 3,817 7,747 107 Moldova (Republic of) 1.004 1 0.712 0.710 79.1 74.5 13.0 g 12.4 g 9.6 g 9.2 g 7,541 11,518 108 Turkmenistan 1.007 .. 0.714 0.709 11.8 11.4 11.6 11.5 5,886 7,861 108 Uzbekistan .. 3 .. .. 79.6 74.1 10.5 g 11.1 g .. .. 11,746 21,213 110 Libya 0.939 3 0.685 0.730 11.8 12.2 11.3 11.8 4,656 8,277 111 Indonesia 0.931 3 0.670 0.720 75.3 70.3 13.0 l 12.6 l 8.0 j 7.2 j 4,867 18,363 111 Samoa 0.937 .. 0.681 0.727 12.9 12.9 7.6 8.4 7,672 14,789 113 South Africa .. 1 .. .. 76.7 67.0 12.9 g 12.1 g .. .. 3,955 7,685 114 Bolivia (Plurinational State of) 0.984 3 0.698 0.710 14.0 13.3 10.0 10.5 9,035 14,554 115 Gabon 0.936 4 0.678 0.724 77.2 70.8 14.0 n 14.0 n 8.3 9.8 4,902 8,780 116 Egypt 0.917 5 0.669 0.729 12.5 l 13.3 l 7.5 m 9.2 m 11,238 20,183 MEDIUM HUMAN DEVELOPMENT 0.878 0.643 0.732 77.4 74.7 13.1 13.1 6.7 h 8.0 h 4,364 16,989 117 Marshall Islands .. 118 Viet Nam .. 1 78.5 74.5 119 Palestine, State of 1.003 5 120 Iraq 0.871 5 74.0 65.6 121 Morocco 0.789 5 0.833 80.8 77.1

72.0 66.2

75.0 70.2

76.0 72.8

76.1 68.4

69.2 65.6

76.3 72.2

74.9 68.4

76.2 72.7

77.7 71.6

80.5 77.2

72.8 68.9

75.4 67.1

76.1 67.5

71.6 64.6

75.8 69.9

75.3 71.2

67.4 60.5

74.2 68.4

68.3 64.2

74.2 69.6

… … 10.9 g 11.2 g .. .. 0.693 0.692 12.9 i 12.5 i 7.9 h 8.5 h 5,739 6,703 0.624 0.716 79.4 71.2 13.7 12.0 8.9 9.3 1,824 8,705 0.587 0.744 10.2 m 12.1 m 6.0 g 8.6 g 3,712 26,745 0.603 0.724 75.6 72.3 12.6 g 13.6 g 4.6 h 6.4 h 3,012 12,019

72.5 68.4

77.7 75.2

TABLE 4 Gender Development Index | 313 TABLE 4 GENDER DEVELOPMENT INDEX

Gender Development Human Development SDG 3 SDG 4.3 SDG 4.6 SDG 8.5 Life expectancy Expected years Mean years Estimated gross national at birth of schooling of schooling income per capitaa

Value (years) (years) (years) (2011 PPP $)

Value Groupb Female Male Female Male Female Male Female Male Female Male

HDI rank 2018 2018 2018 2018 2018 2018 2018c 2018c 2018c 2018c 2018 2018

122 Kyrgyzstan 0.959 2 0.656 0.684 75.5 67.3 13.6 13.2 11.0 i 10.8 i 2,192 4,465 123 Guyana 0.973 2 0.656 0.674 11.9 g 11.1 g 8.9 i 8.0 i 4,676 10,533 124 El Salvador 0.969 2 0.654 0.675 73.0 66.8 11.9 12.2 6.6 7.3 5,234 8,944 125 Tajikistan 0.799 5 0.561 0.703 10.9 g 12.3 g 10.1 m 11.2 m 1,044 5,881 126 Cabo Verde 0.984 1 0.644 0.655 77.6 68.2 12.1 11.6 6.0 6.5 5,523 7,497 126 Guatemala 0.943 3 0.628 0.666 10.5 10.8 6.4 6.5 4,864 9,970 126 Nicaragua 1.013 1 0.655 0.646 73.2 68.7 12.5 n 11.9 n 7.1 h 6.5 h 4,277 5,318 129 India 0.829 5 0.574 0.692 12.9 11.9 4.7 g 8.2 g 2,625 10,712 130 Namibia 1.009 1 0.647 0.641 76.0 69.3 12.7 m 12.5 m 7.3 h 6.6 h 8,917 10,497 131 Timor-Leste 0.899 5 0.589 0.655 12.0 g 12.8 g 3.6 m 5.3 m 5,389 9,618 TABLE 132 Honduras 0.970 2 0.611 0.630 76.9 71.1 10.6 9.8 6.6 6.6 3,214 5,305 132 Kiribati .. 12.2 g 11.4 g 4 134 Bhutan .. 5 .. .. 77.8 70.7 12.2 g 12.0 g … .. 135 Bangladesh 0.893 5 0.581 0.650 11.6 10.8 2.1 g 4.2 g 6,388 10,579 135 Micronesia (Federated States of) 0.895 .. 0.575 0.642 70.7 68.2 5.3 6.8 2,373 5,701 137 Sao Tome and Principe 5 .. .. 138 Congo .. 3 .. .. 66.2 60.4 12.8 g 12.6 g … .. 138 Eswatini (Kingdom of) 0.900 2 0.571 0.635 11.5 l 11.9 l 5.7 g 7.2 g 1,885 4,162 140 Lao People’s Democratic Republic 0.931 3 0.591 0.635 71.4 67.3 10.9 g 11.7 g 6.1 j 7.5 j 4,989 6,621 141 Vanuatu 0.962 .. 0.595 0.618 10.8 11.3 6.3 i 7.2 i 7,030 11,798 142 Ghana 0.929 4 0.581 0.625 77.4 72.8 10.9 g 11.7 g 4.8 h 5.6 h 5,027 7,595 143 Zambia 3 11.4 11.7 2,185 3,413 144 Equatorial Guinea … .. 72.1 64.0 11.6 m 12.5 m .. .. 3,287 4,889 145 Myanmar 0.912 2 0.567 0.622 6.4 h 7.9 h 3,011 4,164 146 Cambodia 0.949 4 0.575 0.606 71.8 71.1 .. .. 6.7 m 7.5 m 12,781 21,809 147 Kenya 3 10.5 10.1 3.9 k 7.2 k 3,613 8,076 147 Nepal .. 5 .. .. 74.3 70.6 10.9 g 11.8 g 5.0 m 4.9 m 3,129 4,089 149 Angola 0.953 4 0.566 0.594 10.3 g 10.9 g 4.1 h 5.7 h 2,619 3,490 150 Cameroon 0.919 5 0.557 0.606 69.5 66.1 12.7 11.7 6.0 h 7.2 h 2,113 3,510 150 Zimbabwe 0.933 4 0.553 0.593 11.0 m 12.7 m 3.6 h 6.4 h 4,720 6,407 152 Pakistan 0.897 5 0.549 0.612 72.6 67.8 11.9 13.6 4.0 m 6.4 m 2,724 3,858 153 Solomon Islands 0.902 .. 0.546 0.605 10.3 10.6 4.8 i 7.8 i 2,280 3,080 LOW HUMAN DEVELOPMENT 0.869 0.522 0.601 65.7 62.8 7.8 9.3 7.6 g 9.0 g 1,570 8,605 154 Syrian Arab Republic 0.925 5 0.540 0.584 9.7 g 10.7 g 3.8 6.5 1,569 2,469 155 Papua New Guinea 0.747 .. 0.464 0.622 64.0 55.3 156 Comoros 5 .. .. 157 Rwanda .. 3 .. .. 69.4 65.8 158 Nigeria 5 159 Tanzania (United Republic of) 0.795 3 72.0 68.8 159 Uganda .. 5 161 Mauritania 5 64.9 62.7 162 Madagascar 0.888 3 163 Benin 0.943 5 66.4 60.5 164 Lesotho 0.868 2 165 Côte d’Ivoire 0.936 5 59.6 57.4 166 Senegal 0.863 5 167 Togo 0.853 5 69.9 63.8 168 Sudan 0.946 5 169 Haiti 0.883 5 71.6 67.3 170 Afghanistan 1.026 5 171 Djibouti 0.796 .. 68.7 64.0 172 Malawi 0.873 3 173 Ethiopia 0.818 5 71.9 69.0 174 Gambia 0.837 5 174 Guinea 0.890 5 63.7 58.1 176 Liberia 0.723 5 177 Yemen 5 60.2 57.7 178 Guinea-Bissau .. .. 179 Congo (Democratic Republic of the) 0.930 5 62.6 59.5 180 Mozambique 0.844 4 181 Sierra Leone 0.832 5 68.1 66.2 0.899 74.7 71.2 0.458 0.457 0.575 77.8 66.6 8.7 g 8.8 g 4.6 o 5.6 o 656 4,779 0.844 65.6 63.0 3.9 h 5.4 h 3,248 4,106 0.901 0.504 0.568 11.1 g 11.4 g 0.882 0.520 0.551 65.9 62.4 11.2 11.2 3.9 m 5.9 m 1,812 3,030 0.492 0.567 8.6 i 10.1 i 0.509 0.544 70.8 66.5 7.7 8.1 3.9 g 4.9 g 1,708 2,218 0.484 0.561 10.4 g 11.5 g 0.479 0.562 55.2 53.5 8.5 8.5 5.3 m 7.6 m 4,313 5,838 0.504 0.533 10.3 10.4 0.486 0.550 66.8 63.2 11.4 13.8 5.6 h 6.4 h 2,436 3,175 0.522 0.509 11.1 10.3 0.445 0.559 65.2 60.7 8.2 10.0 4.8 m 7.4 m 1,272 2,247 0.476 0.545 9.4 8.6 0.459 0.561 66.3 63.1 11.4 13.7 3.7 h 5.5 h 2,018 5,462 0.457 0.546 7.7 8.3 0.477 0.536 68.3 65.1 9.6 l 10.4 l 6.4 l 5.8 l 1,119 1,690 0.411 0.568 7.9 12.5 63.0 59.9 6.0 g 6.9 g 3.0 j 4.4 j 1,863 2,407 .. .. 10.9 m 11.0 m 0.466 0.501 57.0 50.6 8.3 g 9.1 g 7.0 h 5.5 h 2,641 3,864 0.428 0.507 9.5 g 9.4 g 0.416 0.500 58.7 56.3 7.7 g 10.3 g 4.1 h 6.3 h 1,790 5,355 0.413 0.513 8.8 g 10.1 g 0.438 0.487 69.6 65.5 7.4 g 10.1 g 1.8 g 4.4 g 2,173 4,396 0.245 0.535 61.6 59.9 .. .. 3.3 m 6.6 m 1,200 1,989 .. .. 8.7 g 10.6 g 0.419 0.496 66.9 63.3 9.3 10.2 3.2 h 4.2 h 1,759 6,168 0.422 0.468 9.7 g 10.6 g 0.411 0.465 65.8 61.5 4.3 m 6.6 m 1,388 1,949

66.0 63.0 1.9 h 6.0 h 1,102 2,355

68.8 64.6 .. .. 2,900 4,232

66.9 60.7 4.1 h 5.1 h 925 1,400

68.2 64.4 1.6 m 3.9 m 1,333 2,231

63.2 60.4 3.0 m 4.3 m 800 2,190

61.7 60.5 1.5 m 3.9 m 1,878 2,569

65.1 62.3 3.5 h 5.9 h 1,051 1,030

67.8 64.4 1.9 j 4.4 j 168 2,679

59.9 56.0 .. .. 1,305 1,895

61.9 58.9 5.3 8.4 684 917

63.0 57.1 2.5 g 4.6 g 1,031 1,284

55.1 53.5 2.8 h 4.4 h 1,238 1,525

314 | HUMAN DEVELOPMENT REPORT 2019

Gender Development Human Development SDG 3 SDG 4.3 SDG 4.6 SDG 8.5 Life expectancy Expected years Mean years Estimated gross national at birth of schooling of schooling income per capitaa

Value (years) (years) (years) (2011 PPP $)

Value Groupb Female Male Female Male Female Male Female Male Female Male

HDI rank 2018 2018 2018 2018 2018 2018 2018c 2018c 2018c 2018c 2018 2018 0.875 5 182 Burkina Faso .. 0.403 0.461 61.9 60.4 8.7 9.1 1.0 m 2.1 m 1,336 2,077 182 Eritrea .. 5 .. .. 4.6 5.4 .. .. 1,403 2,011 184 Mali 0.807 1 68.2 63.8 6.8 8.6 1,311 2,618 185 Burundi 1.003 5 0.380 0.471 10.9 11.7 1.7 i 3.0 i 186 South Sudan 0.839 5 0.422 0.420 59.6 58.1 3.5 g 5.9 g 2.7 m 3.6 m 763 555 187 Chad 0.774 5 0.369 0.440 6.0 g 8.9 g 4.0 5.3 1,277 1,633 188 Central African Republic 0.795 5 0.347 0.449 63.0 59.4 6.2 g 8.9 g 1.3 m 3.6 m 1,377 2,056 189 Niger 0.298 0.335 0.421 5.8 7.2 3.0 h 5.6 h OTHER COUNTRIES OR TERRITORIES .. 0.130 0.435 59.1 56.1 1.4 g 2.7 g 622 935 .. .. 112 1,705 .. Korea (Democratic People’s Rep. of) .. .. 55.4 52.6 TABLE .. Monaco … Nauru .. .. 55.0 50.6 4 .. San Marino … Somalia .. 63.2 60.9 .. Tuvalu — Human development groups 0.979 — .. .. 75.5 68.4 10.4 g 11.3 g … .. 0.960 — .. .. Very high human development 0.845 — … … .. .. High human development 0.858 — 11.8 g 10.8 g Medium human development 0.918 … .. 15.6 14.6 … .. Low human development — Developing countries 0.856 — … … … .. Regions 0.962 — .. .. Arab States 0.953 — .. .. 58.8 55.4 … .. East Asia and the Pacific 0.978 — Europe and Central Asia 0.828 — … … .. .. Latin America and the Caribbean 0.891 — South Asia 0.869 — 0.880 0.898 82.4 76.7 16.7 16.1 12.0 12.1 30,171 50,297 Sub-Saharan Africa 0.967 0.732 0.763 Least developed countries — 0.571 0.676 77.8 72.7 14.0 13.6 8.0 8.6 10,460 18,271 Small island developing states 0.976 — 0.465 0.542 Organisation for Economic 0.941 0.653 0.711 70.9 67.8 11.9 11.5 5.0 7.8 2,787 9,528 World 63.0 59.7 8.5 9.9 3.8 5.8 1,928 3,232

73.2 69.1 12.2 12.2 6.7 8.1 6,804 14,040

0.634 0.740 73.8 70.2 11.7 12.3 6.4 7.8 5,338 25,343 0.725 0.754 0.757 0.794 77.8 72.9 13.5 13.3 7.5 8.3 11,385 17,728 0.747 0.764 0.570 0.688 77.5 70.8 14.4 14.7 9.9 10.5 10,588 20,674 0.507 0.569 0.489 0.562 78.6 72.3 14.9 14.1 8.6 8.5 9,836 18,004 0.718 0.743 71.1 68.5 12.0 11.6 5.0 8.0 2,639 10,693 0.882 0.903 0.707 0.751 62.9 59.4 9.3 10.4 4.8 6.6 2,752 4,133

66.9 63.2 9.3 10.2 3.9 5.7 1,807 3,462

74.0 69.8 13.1 12.6 8.5 9.0 12,022 19,066

83.0 77.7 16.6 16.0 11.9 12.1 31,016 50,530

74.9 70.4 12.7 12.6 7.9 9.0 11,246 20,167

NOTES m Updated by HDRO based on data from ICF Macro achievements between women and men (absolute purchasing power parity terms). See Technical Demographic and Health Surveys for 2006­2018. deviation from gender parity of more than 10 percent). note 3 at http://hdr.undp.org/sites/default/files/ a Because disaggregated income data are not hdr2019_technical_notes.pdf for details. available, data are crudely estimated. See n Updated by HDRO based on data from CEDLAS Human Development Index (HDI): A composite Definitions and Technical note 3 at http://hdr. and World Bank (2018). index measuring average achievement in three basic MAIN DATA SOURCES undp.org/sites/default/files/hdr2019_technical_ dimensions of human development— a long and notes.pdf for details on how the Gender o Updated by HDRO based on Syrian Center for healthy life, knowledge and a decent standard of Column 1: Calculated based on data in columns Development Index is calculated. Policy Research (2017). living. See Technical note 1 at http://hdr.undp.org/ 3 and 4. sites/default/files/hdr2019_technical_notes.pdf for b Countries are divided into five groups by absolute DEFINITIONS details on how the HDI is calculated. Column 2: Calculated based on data in column 1. deviation from gender parity in HDI values. Gender Development Index: Ratio of female Life expectancy at birth: Number of years a Columns 3 and 4: HDRO calculations based on c Data refer to 2018 or the most recent year available. to male HDI values. See Technical note 3 at newborn infant could expect to live if prevailing data from UNDESA (2019b), UNESCO Institute for http://hdr.undp.org/sites/default/files/hdr2019_ patterns of age-specific mortality rates at the time of Statistics (2019), Barro and Lee (2018), World Bank d In calculating the HDI value, expected years of technical_notes.pdf for details on how the Gender birth stay the same throughout the infant’s life. (2019a), ILO (2019) and IMF (2019). schooling is capped at 18 years. Development Index is calculated. Expected years of schooling: Number of years Columns 5 and 6: UNDESA (2019b). e In calculating the male HDI value, estimated gross Gender Development Index groups: Countries of schooling that a child of school entrance age national income per capita is capped at $75,000. are divided into five groups by absolute deviation can expect to receive if prevailing patterns of Columns 7 and 8: UNESCO Institute for Statistics from gender parity in HDI values. Group 1 comprises age-specific enrolment rates persist throughout the (2019), ICF Macro Demographic and Health Surveys, f Based on data from OECD (2018). countries with high equality in HDI achievements child’s life. UNICEF Multiple Indicator Cluster Surveys and between women and men (absolute deviation of OECD (2018). g Updated by HDRO based on data from UNESCO less than 2.5 percent), group 2 comprises countries Mean years of schooling: Average number of Institute for Statistics (2019). with medium to high equality in HDI achievements years of education received by people ages 25 and Columns 9 and 10: UNESCO Institute for between women and men (absolute deviation of 2.5­ older, converted from educational attainment levels Statistics (2019), Barro and Lee (2018), ICF Macro h Based on Barro and Lee (2018). 5 percent), group 3 comprises countries with medium using official durations of each level. Demographic and Health Surveys, UNICEF Multiple equality in HDI achievements between women and Indicator Cluster Surveys and OECD (2018). i Updated by HDRO based on data from United men (absolute deviation of 5­7.5 percent), group 4 Estimated gross national income per capita: Nations Children’s Fund (UNICEF) Multiple comprises countries with medium to low equality Derived from the ratio of female to male wages, Columns 11 and 12: HDRO calculations based on Indicator Cluster Surveys for 2006­2018. in HDI achievements between women and men female and male shares of economically active ILO (2019), UNDESA (2019b), World Bank (2019a), (absolute deviation of 7.5­10 percent) and group population and gross national income (in 2011 United Nations Statistics Division (2019b) and IMF j Updated by HDRO using Barro and Lee (2018) 5 comprises countries with low equality in HDI (2019).

k Based on data from the national statistical office.

l Based on cross-country regression.

TABLE 4 Gender Development Index | 315 TABLE 5 Gender Inequality Index

Gender Inequality Index SDG 3.1 SDG 3.7 SDG 5.5 SDG 4.6 Labour force Maternal Adolescent Share of seats Population with at least some mortality ratio birth rate in parliament secondary education

(deaths per 100,000 (births per 1,000 women (% ages 25 and older) (% ages 15 and older)

Value Rank live births) ages 15­19) (% held by women) Female Male Female Male

HDI rank 2018 2018 2015 2015­2020b 2018 2010­2018c 2010­2018c 2018 2018

VERY HIGH HUMAN DEVELOPMENT 0.044 5 5 5.1 41.4 96.1 94.8 60.2 66.7 1 Norway 0.037 96.4 97.2 2 Switzerland 0.093 1 5 2.8 29.3 90.2 d 86.3 d 62.6 74.1 3 Ireland 0.084 96.0 96.6 4 Germany 22 8 7.5 24.3 76.6 82.9 55.1 68.1 4 Hong Kong, China (SAR) .. 90.0 90.7 6 Australia 0.103 19 6 8.1 31.5 100.0 e 100.0 e 55.3 66.2 6 Iceland 0.057 88.8 89.0 8 Sweden 0.040 .. .. 2.7 .. 76.3 83.3 54.1 67.8 9 Singapore 0.065 86.6 90.1 10 Netherlands 0.041 25 6 11.7 32.7 89.2 89.4 59.7 70.5 11 Denmark 0.040 100.0 100.0 12 Finland 0.050 9 3 6.3 38.1 100.0 e 100.0 e 72.1 80.6 13 Canada 0.083 97.2 96.6 TABLE 14 New Zealand 0.133 2 4 5.1 46.1 82.9 85.7 61.1 67.6 15 United Kingdom 0.119 95.7 95.5 5 15 United States 0.182 11 10 3.5 23.0 82.6 87.1 60.5 76.3 17 Belgium 0.045 18 Liechtenstein 4 7 3.8 35.6 .. .. 58.0 68.9 19 Japan .. 95.2 d 92.2 d 20 Austria 0.099 2 6 4.1 37.4 100.0 100.0 58.1 65.9 21 Luxembourg 0.073 100.0 100.0 22 Israel 0.078 7 3 5.8 42.0 87.8 90.5 55.0 62.2 22 Korea (Republic of) 0.100 89.8 95.6 24 Slovenia 0.058 18 7 8.4 31.7 97.0 98.3 60.9 69.7 25 Spain 0.069 73.3 78.4 26 Czechia 0.074 34 11 19.3 38.3 99.8 99.8 64.6 75.7 26 France 0.137 81.0 86.3 28 Malta 0.051 27 9 13.4 28.9 74.3 82.2 57.1 67.8 29 Italy 0.195 75.6 83.0 30 Estonia 0.069 42 14 19.9 23.6 100.0 e 100.0 e 56.1 68.2 31 Cyprus 0.091 78.2 82.6 32 Greece 0.086 6 7 4.7 41.4 61.5 73.2 47.9 58.9 32 Poland 0.122 82.9 88.1 34 Lithuania 0.120 … 12.0 92.9 97.5 .. .. 35 United Arab Emirates 0.124 78.8 d 65.7 d 36 Andorra 0.113 23 5 3.8 13.7 71.5 73.3 51.4 70.7 36 Saudi Arabia 67.8 75.5 36 Slovakia .. 14 4 7.3 34.8 99.1 100.0 54.8 65.9 39 Latvia 0.224 100.0 e 99.1 e 40 Portugal 0.190 16 10 4.7 20.0 53.6 54.8 53.5 62.7 41 Qatar 0.169 73.5 66.1 42 Chile 0.081 24 5 9.6 27.5 79.0 80.9 59.2 69.1 43 Brunei Darussalam 0.202 69.5 d 70.6 d 43 Hungary 0.288 10 11 1.4 17.0 96.3 98.2 52.8 73.3 45 Bahrain 0.234 64.2 d 57.5 d 46 Croatia 0.258 12 9 3.8 20.0 94.5 96.9 53.4 62.7 47 Oman 0.207 73.4 63.7 48 Argentina 0.122 15 5 7.7 38.6 66.5 d 63.3 d 51.7 63.4 49 Russian Federation 0.304 96.3 95.7 50 Belarus 0.354 35 4 12.0 20.3 87.2 92.5 52.4 68.4 50 Kazakhstan 0.255 98.3 d 98.9 d 52 Bulgaria 0.119 8 8 4.7 35.7 94.2 96.2 50.3 60.0 52 Montenegro 0.203 88.0 97.5 52 Romania 0.218 44 9 12.9 11.9 87.2 93.1 43.3 66.2 55 Palau 0.119 96.9 97.3 56 Barbados 0.316 12 4 5.2 35.6 94.6 d 91.9 d 40.0 58.4 57 Kuwait 56.8 49.3 57 Uruguay .. 21 9 7.7 26.7 57.8 54.0 57.0 70.9 59 Turkey 0.256 44.3 66.0 60 Bahamas 0.245 20 7 4.6 17.9 88.0 91.0 57.3 67.2 0.275 0.305 31 3 7.2 18.7 45.3 60.7 0.353 30 3 10.5 25.5 48.9 65.5

33 10 10.9 21.3 56.4 66.7

26 6 6.5 22.5 51.2 93.4

… 32.1 .. ..

49 12 7.3 19.9 23.4 79.2

43 6 25.7 20.0 52.7 67.4

40 18 16.2 31.0 55.4 68.0

17 10 8.4 34.8 53.9 64.2

45 13 9.9 9.8 57.8 94.7

62 22 41.1 22.7 51.0 74.2

51 23 10.3 9.1 58.2 71.7

56 17 24.0 12.6 48.3 65.0

47 15 13.4 18.8 44.5 87.3

31 8 8.7 18.5 45.7 58.2

65 17 13.1 8.8 31.0 88.7

77 52 62.8 39.5 49.0 72.8

54 25 20.7 16.1 54.9 70.5

27 4 14.5 33.1 58.1 70.3

46 12 29.8 22.1 65.2 77.1

48 11 39.9 23.8 49.5 61.6

27 7 9.3 23.5 43.6 58.1

69 31 36.2 18.7 45.6 64.2

… 13.8 .. ..

55 27 33.6 27.5 61.9 69.6

53 4 8.2 3.1 57.5 85.3

59 15 58.7 22.3 55.8 73.8

66 16 26.6 17.4 33.5 72.6

76 80 30.0 21.8 67.6 82.0

316 | HUMAN DEVELOPMENT REPORT 2019

Gender Inequality Index SDG 3.1 SDG 3.7 SDG 5.5 SDG 4.6 Labour force Maternal Adolescent Share of seats Population with at least some mortality ratio birth rate in parliament secondary education

(deaths per 100,000 (births per 1,000 women (% ages 25 and older) (% ages 15 and older)

Value Rank live births) ages 15­19) (% held by women) Female Male Female Male

HDI rank 2018 2018 2015 2015­2020b 2018 2010­2018c 2010­2018c 2018 2018 0.274 79.8 d 81.8 d 61 Malaysia 58 40 13.4 15.8 .. .. 50.9 77.4 62 Seychelles .. HIGH HUMAN DEVELOPMENT .. .. 62.1 21.2 .. .. 63 Trinidad and Tobago 0.161 37 17 14.7 34.4 85.7 93.6 46.8 62.1 65 Iran (Islamic Republic of) 74.4 d 71.2 d 66 Mauritius 0.323 72 63 30.1 30.1 67.4 72.0 50.4 71.3 67 Panama 65.7 d 68.1 d 68 Costa Rica 0.492 118 25 40.6 5.9 74.8 d 68.4 d 16.8 71.2 69 Albania 53.8 52.3 70 Georgia 0.369 82 53 25.7 11.6 93.5 92.8 45.0 71.8 71 Sri Lanka 97.4 98.6 72 Cuba 0.460 108 94 81.8 18.3 82.6 d 83.1 d 52.5 80.5 73 Saint Kitts and Nevis 86.7 d 88.9 d 74 Antigua and Barbuda 0.285 61 25 53.5 45.6 45.7 74.6 TABLE 75 Bosnia and Herzegovina .. .. 76 Mexico 0.234 51 29 19.6 27.9 .. .. 47.2 64.9 5 77 Thailand 73.1 90.0 78 Grenada 0.351 75 36 46.4 16.0 58.4 61.1 57.8 78.7 79 Brazil 43.1 48.2 79 Colombia 0.380 86 30 20.9 5.8 .. .. 34.9 72.2 81 Armenia 61.0 57.7 82 Algeria 0.312 67 39 51.6 53.2 53.1 50.9 40.0 67.4 82 North Macedonia 96.9 97.6 82 Peru … .. 13.3 39.1 d 38.9 d .. .. 85 China 41.6 f 57.6 f 85 Ecuador … 42.8 31.4 57.4 68.5 .. .. 87 Azerbaijan 75.4 d 83.0 d 88 Ukraine 0.162 38 11 9.6 19.3 51.9 51.9 35.6 58.6 89 Dominican Republic 93.9 97.5 89 Saint Lucia 0.334 74 38 60.4 48.4 94.0 d 95.2 d 43.8 78.9 91 Tunisia 58.6 54.4 92 Mongolia 0.377 84 20 44.9 5.3 49.2 42.1 59.5 76.2 93 Lebanon 42.3 d 54.6 d 94 Botswana .. .. 27 29.2 39.3 91.2 86.3 .. .. 94 Saint Vincent and the Grenadines 54.3 g 55.6 g 96 Jamaica 0.386 89 44 59.1 15.0 89.6 d 90.3 d 54.0 74.4 96 Venezuela (Bolivarian Republic of) .. .. 98 Dominica 0.411 94 64 66.7 19.0 69.9 62.4 58.6 82.0 98 Fiji 71.7 66.6 98 Paraguay 0.259 57 25 21.5 18.1 .. .. 49.6 69.9 98 Suriname 78.3 d 70.2 d 102 Jordan 0.443 100 140 10.1 21.3 47.3 48.3 14.9 67.4 103 Belize 61.5 60.1 104 Maldives 0.145 36 8 15.7 38.3 82.0 d 85.9 d 42.7 67.5 105 Tonga 78.9 78.4 106 Philippines 0.381 87 68 56.9 27.7 44.9 d 49.3 d 69.9 84.7 107 Moldova (Republic of) 94.0 d 93.4 d 108 Turkmenistan 0.163 39 27 7.6 24.9 75.6 d 72.4 d 61.3 75.9 108 Uzbekistan 95.5 97.4 110 Libya 0.389 90 64 79.3 38.0 .. .. 56.6 81.8 111 Indonesia 99.9 99.9 111 Samoa 0.321 70 25 55.8 16.8 69.4 d 45.0 d 63.1 69.7 113 South Africa 44.5 53.2 114 Bolivia (Plurinational State of) 0.284 60 24 23.7 12.3 79.1 h 71.6 h 46.7 62.8 115 Gabon 75.0 78.2 116 Egypt 0.453 104 92 94.3 24.3 52.8 65.1 50.9 77.6 MEDIUM HUMAN DEVELOPMENT 65.6 d 49.8 d 117 Marshall Islands 0.333 73 48 40.5 20.7 59.2 d 71.2 d 60.2 75.3 119 Palestine, State of 0.300 63 62 7.8 31.3 24.1 69.9

0.322 71 44 31.0 17.1 53.3 66.7

0.362 79 15 14.5 4.7 23.5 70.9

0.464 111 129 46.1 9.5 66.2 78.6

.. .. 45 49.0 13.0 57.3 79.2

0.405 93 89 52.8 19.0 60.4 73.9

0.458 106 95 85.3 22.2 47.7 77.1

… .. 25.0 .. ..

0.357 78 30 49.4 19.6 38.1 76.1

0.482 117 132 70.5 16.0 56.9 84.1

0.465 112 155 61.7 25.5 39.2 64.2

0.469 113 58 25.9 15.4 14.1 64.0

0.391 91 28 68.5 11.1 53.3 81.4

0.367 81 68 7.8 5.9 41.9 82.0

0.418 96 124 14.7 7.4 45.3 74.1

0.425 98 114 54.2 29.1 45.7 74.1

0.228 50 23 22.4 22.8 38.9 45.6

.. .. 42 24.4 24.8 52.8 78.2

0.303 64 36 23.8 16.4 53.4 78.0

0.172 41 9 5.8 16.0 25.7 79.0

0.451 103 126 47.4 19.8 52.2 82.0

0.364 80 51 23.9 10.0 23.7 38.6

0.422 97 138 67.9 41.8 i 48.9 62.6

0.446 101 206 64.9 51.8 56.6 79.4

0.534 128 291 96.2 17.4 j 43.4 60.2

0.450 102 33 53.8 14.9 22.8 73.2

… .. 9.1 91.6 92.5 .. .. 66.2 d 77.7 d 0.314 68 54 30.9 26.7 60.0 62.2 72.7 82.5

.. .. 45 52.8 .. 19.3 71.1

TABLE 5 Gender Inequality Index | 317 TABLE 5 GENDER INEQUALITY INDEX

Gender Inequality Index SDG 3.1 SDG 3.7 SDG 5.5 SDG 4.6 Labour force Maternal Adolescent Share of seats Population with at least some mortality ratio birth rate in parliament secondary education

(deaths per 100,000 (births per 1,000 women (% ages 25 and older) (% ages 15 and older)

Value Rank live births) ages 15­19) (% held by women) Female Male Female Male

HDI rank 2018 2018 2015 2015­2020b 2018 2010­2018c 2010­2018c 2018 2018

120 Iraq 0.540 131 50 71.7 25.2 39.5 d 56.5 d 12.4 72.6 121 Morocco 0.492 29.0 d 35.6 d 122 Kyrgyzstan 0.381 118 121 31.0 18.4 98.6 d 98.3 d 21.4 70.4 123 Guyana 0.492 70.9 d 55.5 d 124 El Salvador 0.397 87 76 32.8 19.2 39.9 46.3 48.0 75.8 125 Tajikistan 0.377 98.8 d 87.0 d 126 Cabo Verde 0.372 118 229 74.4 31.9 28.7 31.2 41.2 73.6 126 Guatemala 0.492 38.4 37.2 126 Nicaragua 0.455 92 54 69.5 31.0 48.3 d 46.6 d 46.1 78.9 129 India 0.501 39.0 d 63.5 d 130 Namibia 0.460 84 32 57.1 20.0 40.5 d 41.9 d 27.8 59.7 132 Honduras .. 83 42 73.8 20.8 k .. .. 65.1 73.2 132 Kiribati 0.479 34.2 32.6 134 Bhutan 118 88 70.9 12.7 41.1 85.0 135 Bangladesh … TABLE 135 Micronesia (Federated States of) 0.436 105 150 85.0 45.7 7.6 17.5 50.7 83.7 137 Sao Tome and Principe 0.536 45.3 d 49.2 d 5 138 Congo 122 174 13.2 11.7 23.6 78.6 138 Eswatini (Kingdom of) … 140 Lao People’s Democratic Republic 0.547 108 265 63.6 39.7 31.5 45.8 56.2 65.9 141 Vanuatu 0.579 46.7 d 51.3 d 142 Ghana 0.579 .. 215 33.8 33.8 31.3 d 33.9 d 25.0 52.6 143 Zambia 0.463 35.0 d 46.0 d 144 Equatorial Guinea 116 129 72.9 21.1 47.2 83.7 145 Myanmar … 146 Cambodia 0.541 .. 90 16.2 6.5 55.7 d 71.1 d .. .. 147 Kenya 0.540 39.2 d 52.4 d 147 Nepal 99 148 20.2 15.3 58.2 74.5 149 Angola … 150 Cameroon 0.458 129 176 83.0 20.3 28.7 d 22.3 d 36.0 81.3 150 Zimbabwe 0.474 15.1 d 28.1 d 152 Pakistan 0.545 .. 100 13.9 0.0 l 29.8 d 37.3 d .. .. 153 Solomon Islands 0.476 29.0 d 44.2 d LOW HUMAN DEVELOPMENT 0.578 136 156 94.6 14.5 23.1 38.1 43.3 76.2 154 Syrian Arab Republic 0.566 32.7 40.9 155 Papua New Guinea 0.525 145 442 112.2 14.0 55.9 66.3 66.9 71.6 156 Comoros 0.547 26.7 47.3 157 Rwanda 145 389 76.7 12.1 41.4 65.9 158 Nigeria … 159 Tanzania (United Republic of) 110 197 65.4 27.5 76.8 79.7 161 Mauritania .. 78 49.4 0.0 l 61.5 79.6 163 Benin 133 319 66.6 12.7 63.6 71.5 165 Côte d’Ivoire 131 224 120.1 18.0 70.8 79.8 167 Togo .. 342 155.6 18.0 55.2 67.1 169 Haiti 106 178 28.5 10.2 47.7 77.3 171 Djibouti 114 161 50.2 19.3 75.2 87.6 173 Ethiopia 134 510 75.1 23.3 63.6 69.1 174 Guinea 115 258 65.1 33.5 81.7 84.4 177 Yemen 144 477 150.5 30.5 75.4 80.1

140 596 105.8 29.3 71.2 81.4

126 443 86.1 34.3 78.6 89.0

136 178 38.8 20.0 23.9 81.5

.. 114 78.0 2.0 62.4 80.3

0.547 136 68 38.6 13.2 37.1 d 43.4 d 12.0 70.3 52.7 0.0 l 9.9 d 15.2 d 0.740 161 215 65.4 6.1 46.0 47.6 39.1 55.7 … .. 335 107.3 5.8 12.9 d 17.9 d 37.4 50.7 118.4 37.2 0.412 95 290 118.8 34.3 .. .. 84.2 83.6 71.0 20.3 11.9 d 16.9 d .. .. 814 109.6 19.6 27.4 d 34.7 d 50.6 59.8 86.1 7.2 12.7 d 24.9 d 0.539 130 398 92.7 22.7 79.4 87.2 117.6 9.2 m .. .. 0.531 127 343 72.7 41.8 18.2 d 33.6 d 67.2 75.0 89.1 17.6 32.8 d 25.1 d 0.620 150 602 64.0 31.0 17.8 d 34.1 d 29.2 63.2 51.7 2.7 11.1 21.4 .. .. 353 69.0 27.4 j 27.6 d 54.0 d 83.6 89.3 18.8 26.2 15.3 d 19.6 d 0.613 148 405 132.7 16.7 26.9 d 39.9 d 69.2 73.3 66.7 37.3 13.2 d 36.9 d 0.546 135 487 78.2 10.3 59.8 74.9 135.3 21.9 .. .. 0.657 157 645 136.0 11.7 17.6 d 25.9 d 48.3 66.0 60.4 0.5 11.5 n 22.0 n 0.523 125 315 30.7 n 43.6 n 35.2 58.6

0.566 140 368 .. .. 76.1 79.3 18.5 d 39.6 d 0.560 139 311 19.9 d 35.5 d 24.5 70.3

0.620 150 359 63.3 72.8

0.575 143 396 48.7 82.1

.. .. 229 54.8 71.1

0.615 149 634 72.9 82.0

0.508 123 353 74.2 86.5

0.620 150 706 51.7 67.7

.. .. 679 64.1 65.1

0.651 155 725 54.7 57.5

0.834 162 385 6.0 70.8

318 | HUMAN DEVELOPMENT REPORT 2019

Gender Inequality Index SDG 3.1 SDG 3.7 SDG 5.5 SDG 4.6 Labour force Maternal Adolescent Share of seats Population with at least some mortality ratio birth rate in parliament secondary education

(deaths per 100,000 (births per 1,000 women (% ages 25 and older) (% ages 15 and older)

Value Rank live births) ages 15­19) (% held by women) Female Male Female Male

HDI rank 2018 2018 2015 2015­2020b 2018 2010­2018c 2010­2018c 2018 2018 178 Guinea-Bissau .. 549 104.8 13.7 .. .. 67.3 78.9 179 Congo (Democratic Republic of the) 0.655 36.7 65.8 180 Mozambique 0.569 156 693 124.2 8.2 14.0 27.3 60.8 66.5 181 Sierra Leone 0.644 19.9 d 32.9 d 182 Burkina Faso 0.612 142 489 148.6 39.6 6.0 n 12.1 n 77.5 79.6 184 Mali .. 153 1,360 112.8 12.3 .. .. 57.7 58.5 185 Burundi 0.676 7.3 f 16.4 f 186 South Sudan 0.520 147 371 104.3 11.0 7.5 d 11.0 d 58.5 75.1 188 Central African Republic .. .. 501 52.6 22.0 .. .. 74.1 87.1 189 Niger 0.701 1.7 n 10.3 n OTHER COUNTRIES OR TERRITORIES 0.682 158 587 169.1 8.8 13.4 d 31.1 d 61.3 80.9 0.647 4.3 d 8.9 d .. Korea (Democratic People’s Rep. of) 124 712 55.6 38.8 80.4 77.6 .. Monaco .. .. Nauru .. .. 789 62.0 26.6 71.8 74.3 TABLE .. San Marino .. .. Somalia .. 160 856 161.1 15.3 64.8 77.9 5 .. Tuvalu .. Human development groups .. 159 882 129.1 8.6 64.7 79.8

Very high human development 0.175 154 553 186.5 17.0 67.3 90.5 High human development 0.331 Medium human development 0.501 .. 82 0.3 16.3 .. .. 74.3 87.3 Low human development 0.590 Developing countries 0.466 … 33.3 … .. Arab States 0.531 … 10.5 … .. East Asia and the Pacific 0.310 Europe and Central Asia 0.276 … 26.7 … .. Latin America and the Caribbean 0.383 South Asia 0.510 .. 732 100.1 24.3 .. .. 19.1 74.3 Sub-Saharan Africa 0.573 Least developed countries 0.561 … 6.7 … .. Small island developing states 0.453 Organisation for Economic — 15 16.7 27.2 87.0 88.7 52.1 69.0 Cooperation and Development 0.182 World 0.439 — 56 33.6 24.4 68.9 74.5 53.9 75.6

— 198 34.3 20.8 39.5 58.7 32.3 78.9

— 557 101.1 21.3 17.8 30.3 58.2 73.1

— 231 46.8 22.4 55.0 65.8 46.6 76.6

— 148 46.6 18.3 45.9 54.9 20.4 73.8

— 62 22.0 20.3 68.8 76.2 59.7 77.0

— 25 27.8 21.2 78.1 85.8 45.2 70.1

— 68 63.2 31.0 59.7 59.3 51.8 77.2

— 176 26.1 17.1 39.9 60.8 25.9 78.8

— 550 104.7 23.5 28.8 39.8 63.5 72.9

— 434 T 94.4 22.5 25.3 34.9 57.3 78.8

— 192 57.5 24.6 59.0 61.5 51.0 70.2

— 14 20.5 30.1 84.8 87.7 51.6 68.5

— 216 T 42.9 24.1 62.8 71.2 48.0 74.9

NOTES j Refers to 2017. Maternal mortality ratio: Number of deaths due to actively looking for work, expressed as a percentage k Refers to 2013. pregnancy-related causes per 100,000 live births. of the working-age population. a Estimates modelled by the International Labour l In calculating the Gender Inequality Index, a value Organization. Adolescent birth rate: Number of births to women MAIN DATA SOURCES of 0.1 percent was used. ages 15­19 per 1,000 women ages 15­19. b Data are average annual estimates for m Refers to 2015. Column 1: HDRO calculations based on data in 2015­2020. n Updated by HDRO based on data from ICF Macro Share of seats in parliament: Proportion of seats columns 3­9. held by women in the national parliament expressed c Data refer to the most recent year available Demographic and Health Surveys for 2006­2018. as a percentage of total seats. For countries with a Column 2: Calculated based on data in column 1. during the period specified. T From original data source. bicameral legislative system, the share of seats is calculated based on both houses. Column 3: UN Maternal Mortality Estimation Group d Based on Barro and Lee (2018). DEFINITIONS (2017). Population with at least some secondary e Based on data from OECD (2018). Gender Inequality Index: A composite measure education: Percentage of the population ages 25 Column 4: UNDESA (2019b). reflecting inequality in achievement between women and older that has reached (but not necessarily f Updated by HDRO based on data from United and men in three dimensions: reproductive health, completed) a secondary level of education. Column 5: IPU (2019). Nations Children’s Fund Multiple Indicator Cluster empowerment and the labour market. See Technical Surveys for 2006­2018. note 4 at http://hdr.undp.org/sites/default/files/ Labour force participation rate: Proportion of Columns 6 and 7: UNESCO Institute for Statistics hdr2019_technical_notes.pdf for details on how the the working-age population (ages 15 and older) that (2019) and Barro and Lee (2018). g Based on cross-country regression. Gender Inequality Index is calculated. engages in the labour market, either by working or Columns 8 and 9: ILO (2019). h Based on data from the national statistical office.

i Excludes the 36 special rotating delegates

TABLE 5 Gender Inequality Index | 319 TABLE 6 Multidimensional Poverty Index: developing countries

SDG 1.2 SDG 1.2 SDG 1.1

Multidimensional Contribution of deprivation Population living below Population in multidimensional povertya in dimension to overall income poverty line Year and multidimensional povertya surveyb Headcount Population Population (%) (thousands) Inequality in severe vulnerable to

Intensity of among multidimensional multidimensional Standard National PPP $1.90 Health Education of living poverty line a day deprivation the poor poverty povertya

2007­2018 Value (%) year 2017 (%) Value (%) (%) (%) (%) (%) 2007­2018c 2007­2017c

Afghanistan 2015/2016 D 0.272d 55.9d 19,376d 19,865d 48.6d 0.020d 24.9d 18.1d 10.0d 45.0d 45.0d 54.5 .. Albania 2017/2018 D 0.003 39.1 ..e 0.1 5.0 28.3 55.1 Algeria 2012/2013 M 0.008 0.7 21 21 38.8 0.3 5.8 29.9 46.8 16.7 14.3 1.1 0.282 55.3 0.006 32.5 15.5 21.2 32.1 0.001 2.1 805 868 36.2 0.024 0.0 2.7 33.1 36.8 23.2 5.5 0.5 0.198 47.5 16.7 21.4 23.5 29.2 Angola 2015/2016 D 0.009f 51.1 14,725 15,221 34.2f ..e 0.0f 0.5f 96.0f 0.7f 46.8 36.6 30.1 0.017 39.8 0.016 0.6 8.4 39.5 20.9 Armenia 2015/2016 D 0.368 0.2 5 5 55.0 40.9 14.7 20.8 36.3 30.1 25.7 1.4 Bangladesh 2014 D 0.175g 46.8g ..e 14.7g 17.7g 24.2g 36.6g 0.094 41.7 66,468 68,663 46.0 0.007 7.1 15.7 21.6 26.6 47.3 24.3 14.8 0.008f 37.9f 0.025 0.1f 4.1f 79.7f 7.2f Barbados 2012 M 0.016 d,g,h 2.5f 7f 7f 42.5 d,g,h 0.016g 0.9 d,g,h 6.2 d,g,h 49.8 d,g,h 22.9 d,g,h 3.3f .. .. 0.519 61.9 0.014 64.8 7.4 20.0 40.6 Belize 2015/2016 M 0.403 4.3 16 16 54.3 0.002f 45.3 16.3 23.3 27.5 39.6 .. .. Benin 2017/2018 D 0.170 45.8 0.008 d,g,h 13.2 21.1 21.8 31.7 Bhutan 2010 M 0.243 66.8 7,672 7,465 53.5 0.027 25.6 17.3 23.2 28.2 42.9 40.1 49.5 0.465g 58.6g 0.022 54.7g 13.1g 27.8g 25.7g 0.533 37.3g 272g 302g 62.3 0.015 66.1 9.9 20.1 34.4 39.2g 8.2 1.5 0.016j,k 41.3 j,k 0.026 0.3 j,k 17.1 j,k 35.2 j,k 39.2 j,k TABLE Bolivia (Plurinational State of) 2008 D 0.020d 20.4 1,958 2,254 40.6d 0.028g 0.8d 6.2d 12.0d 39.5d 51.8 36.4 5.8 0.181 2.2f 80f 77f 48.5 0.026 16.1 22.3 20.8 31.6 6 Bosnia and Herzegovina 2011/2012 M 0.112 3.8 d,g,h 46.0 0.005j,k 9.4 21.3 23.4 20.2 13.1f 16.9 0.1 Brazil 2015 N h 0.389 7,913 d,g,h 8,041 d,g,h 52.5 0.009d 43.9 16.8 26.1 18.4 0.236 51.2 0.020 24.5 17.6 19.6 40.4 27.3d,g,h 26.5 4.8 0.015d 38.9d 0.013 0.5d 5.2d 29.1d 35.8d Burkina Faso 2010 D 0.018g 83.8 13,083 16,091 40.0g 0.020 0.8g 7.5g 40.8g 23.4g 39.4 40.1 43.7 0.019l 37.6l 0.019 0.6l 6.1l 39.8l 53.2l Burundi 2016/2017 D 0.032 74.3 8,067 8,067 41.3 0.006d 1.7 9.9 15.5 43.4 49.2 64.9 71.8 Cambodia 2014 D 0.081 42.3 0.007g 4.4 20.9 29.3 17.9 Cameroon 2014 M 0.489 37.2 5,679 5,952 58.5 0.004l 61.5 8.9 19.7 29.4 46.6 17.7 .. 0.066 44.3 0.009 4.7 17.5 31.0 22.2 0.286 45.3 10,081 10,903 51.7 0.009 32.0 21.8 28.2 34.4 48.6 37.5 23.8 0.138 79.4g 3,530g 3,697g 45.8 0.024 10.4 22.0 22.3 30.4 Central African Republic 2010 M 0.134 46.2 0.013 11.2 21.1 26.3 35.0 46.5g 62.0 66.3 0.336 54.3 0.018 37.7 17.2 18.7 38.7 Chad 2014/2015 D 0.372 85.7 12,002 12,765 55.3 0.016 40.4 19.2 21.3 33.9 45.5 46.7 38.4 China 2014 N i 0.014 3.9 j,k 53,688j,k 54,437j,k 41.8 0.013 0.7 5.8 31.5 18.7 0.200 4.8d 2,358d 2,378d 48.4 0.022 18.5 21.8 18.5 24.6 25.5 j,k 3.1 0.7 0.090m 46.4m 0.025 6.5m 22.3m 18.5m 33.0m Colombia 2015/2016 D 0.123 43.9 0.008 8.8 19.3 31.9 23.4 48.5d 27.0 3.9 0.028d 40.3d 0.019 1.2d 9.1d 23.2d 30.0d Comoros 2012 D 0.033 37.3 270 303 37.9 0.013m 1.3 5.2 33.1 60.9 47.6 42.4 17.9 0.018f 38.7f 0.014 0.8f 6.4f 42.1f 17.5f Congo 2014/2015 M 0.002 24.3 1,212 1,277 35.4 0.009d 0.0 0.7 37.5 53.5 56.4 46.5 37.0 0.002g 35.6g 0.005 0.0g 1.8g 90.4g 3.1g Congo (Democratic Republic of the) 2013/2014 D 0.178 74.0 54,590 60,230 46.0 13.3 34.9 24.9 14.6 55.5 63.9 76.6 0.008 36.3 ..e 0.0 8.3 52.8 13.0 Côte d’Ivoire 2016 M 0.108 46.1 10,916 11,192 47.0 ..e 9.6 21.2 21.5 39.7 40.0 46.3 28.2 0.146 3.9d 404d 418d 43.4 ..e 8.5 24.4 20.6 21.5 Dominican Republic 2014 M 0.320 4.5g 714g 746g 50.8 0.014 32.1 21.4 19.7 28.2 35.0d 30.5 1.6 Ecuador 2013/2014 N 0.007 5.2l 37.1 0.002 0.1 11.3 39.0 48.6 0.453 4,742l 5,038l 58.2 0.016 57.1 11.8 17.5 31.8 35.8g 23.2 3.2 0.243 46.2 0.010 18.5 28.5 20.7 23.1 Egypt 2014 D 0.003 34.4 0.019 0.0 4.8 80.7 15.1 7.0l 27.8 1.3 0.457 58.5 0.003 56.6 10.9 22.0 41.6 El Salvador 2014 M 0.261 7.9 494 501 51.5 0.023 26.3 18.6 20.2 33.1 41.1 29.2 1.9 Eswatini (Kingdom of) 2014 M 0.025f 39.2f 0.013 1.0f 4.7f 67.0f 14.1f Ethiopia 2016 D 0.004 19.2 249 263 37.4 ..e 0.1 3.7 9.2 42.4 52.8 63.0 42.0 0.042 41.7 0.024 1.6 19.2 24.0 20.9 0.002g 83.5 85,511 87,643 45.7g 0.019 0.1g 4.3g 24.4g 46.0g 50.8 23.5 27.3 0.085g 45.7g 0.008f 6.5g 13.2g 25.6g 42.1g Gabon 2012 D 14.8 261 301 ..e 46.8 33.4 3.4 0.007 Gambia 2013 D 55.2 1,027 1,160 ..e 37.5 48.6 10.1 Ghana 2014 D 0.017g Guatemala 2014/2015 D 30.1 8,109 8,671 47.2 23.4 13.3

28.9 4,694 4,885 38.7 59.3 8.7

Guinea 2016 M 61.9 7,668 7,867 42.6 55.2 35.3

Guinea-Bissau 2014 M 67.3 1,161 1,253 44.7 69.3 67.1 Guyana 2014 M 3.4 26 26 49.8 .. ..

Haiti 2016/2017 D 41.3 4,532 4,532 57.0 58.5 25.0 19.3m 1,642m 1,788m Honduras 2011/2012 D 48.5m 61.9 17.2 India 2015/2016 D Indonesia 2012 D 27.9 369,546 373,735 44.8 21.9 21.2 7.0d 17,452d 18,512d 46.8d 10.6 5.7

Iraq 2018 M 8.6 3,397 3,305 6.0 18.9 2.5

Jamaica 2014 N 4.7f 134f 135f 40.4f 19.9 .. Jordan 2017/2018 D 0.4 43 42 9.0 14.4 0.1

Kazakhstan 2015 M 0.5g 80g 82g 6.4g 2.5 0.0

Kenya 2014 D 38.7 17,801 19,223 60.5 36.1 36.8

Kyrgyzstan 2014 M 2.3 132 138 34.3 25.6 1.5

Lao People’s Democratic Republic 2017 M 23.1 1,582 1,582 38.8 23.4 22.7

Lesotho 2014 D 33.6 720 750 57.9 57.1 59.7

Liberia 2013 D 62.9 2,698 2,978 52.1 50.9 40.9 Libya 2014 P 2.0 124 127 12.4 .. ..

Madagascar 2008/2009 D 77.8 15,995 19,885 50.7 70.7 77.6

Malawi 2015/2016 D 52.6 9,520 9,799 56.2 51.5 70.3 Maldives 2016/2017 D Mali 2015 M 0.8 3 3 4.2 8.2 7.3

78.1 13,640 14,479 36.3 41.1 49.7

Mauritania 2015 M 50.6 2,115 2,235 46.6 31.0 6.0 Mexico 2016 N n Moldova (Republic of) 2012 M 6.3f 8,039f 8,141f 18.8f 43.6 2.5

0.9 38 38 48.4 9.6 0.1

Mongolia 2013 M 10.2 292 313 55.1 21.6 0.6 0.4g 2g 2g Montenegro 2013 M 18.6g 29.7g 24.0 0.0 Morocco 2011 P 6,101g 6,636g 32.3g 4.8 1.0

320 | HUMAN DEVELOPMENT REPORT 2019

SDG 1.2 SDG 1.2 SDG 1.1

Multidimensional Population living below Population in multidimensional povertya Contribution of deprivation income poverty line Year and surveyb Headcount Population Population in dimension to overall (%) (thousands) multidimensional povertya

Inequality in severe vulnerable to

Intensity of among multidimensional multidimensional Standard National PPP $1.90 Health Education of living poverty line a day deprivation the poor poverty povertya

2007­2018 Value (%) year 2017 (%) Value (%) (%) (%) (%) (%) 2007­2018c 2007­2017c

Mozambique 2011 D 0.411 72.5 18,069 21,496 56.7 0.023 49.1 13.6 17.2 32.5 50.3 46.1 62.4 Myanmar 2015/2016 D Namibia 2013 D 0.176 38.3 20,263 20,449 45.9 0.015 13.8 21.9 18.5 32.3 49.2 32.1 6.2 Nepal 2016 D Nicaragua 2011/2012 D 0.171 38.0 880 963 45.1 0.012 12.2 20.3 30.3 14.9 54.9 17.4 13.4 Niger 2012 D Nigeria 2016/2017 M 0.148 34.0 9,851 9,961 43.6 0.012 11.6 22.3 31.5 27.2 41.3 25.2 15.0 North Macedonia 2011 M Pakistan 2017/2018 D 0.074 16.3 956 1,011 45.2 0.013 5.5 13.2 11.1 36.5 52.4 24.9 3.2 Palestine, State of 2014 M Paraguay 2016 M 0.590 90.5 16,042 19,431 65.2 0.026 74.8 5.1 20.3 37.3 42.4 44.5 44.5 Peru 2012 D Philippines 2017 D 0.291 51.4 98,175 98,175 56.6 0.029 32.3 16.8 27.0 32.2 40.8 46.0 53.5 Rwanda 2014/2015 D Saint Lucia 2012 M 0.010f 2.5f 52f 53f 37.7f 0.007f 0.2f 2.9f 62.5f 17.0f 20.5f 22.2 5.2 Sao Tome and Principe 2014 M Senegal 2017 D 0.198 38.3 76,976 75,520 51.7 0.023 21.5 12.9 27.6 41.3 31.1 24.3 3.9 Serbia 2014 M Sierra Leone 2017 M 0.004 1.0 43 47 37.5 0.003 0.1 5.4 53.3 32.8 13.9 29.2 1.0 South Africa 2016 D South Sudan 2010 M 0.019 4.5 303 307 41.9 0.013 1.0 7.2 14.3 38.9 46.8 26.4 1.2 TABLE Sudan 2014 M Suriname 2010 M 0.053 12.7 3,818 4,072 41.6 0.009 2.9 12.5 20.3 23.7 56.0 21.7 3.4 6 Syrian Arab Republic 2009 P 7.3d 20.3d 31.0d 48.7d 21.6 7.8 Tajikistan 2017 D 0.024d 5.8d 6,081d 6,081d 41.8d 0.010d 1.3d Tanzania (United Republic of) 2015/2016 D Thailand 2015/2016 M 0.259 54.4 6,329 6,644 47.5 0.013 22.2 25.7 13.6 30.5 55.9 38.2 55.5 Timor-Leste 2016 D Togo 2013/2014 D 0.007f 1.9f 3f 3f 37.5f ..e 0.0f 1.6f 69.5f 7.5f 23.0f 25.0 4.7 Trinidad and Tobago 2011 M Tunisia 2011/2012 M 0.092 22.1 42 45 41.7 0.008 4.4 19.4 18.6 37.4 44.0 66.2 32.3 Turkmenistan 2015/2016 M Uganda 2016 D 0.288 53.2 8,428 8,428 54.2 0.021 32.8 16.4 22.1 44.9 33.0 46.7 38.0 Ukraine 2012 M Vanuatu 2007 M 0.001g 0.3g 30g 30g 42.5g ..e 0.1g 3.4g 20.6g 42.7g 36.8g 25.7 0.1 Viet Nam 2013/2014 M Yemen 2013 D 0.297 57.9 4,378 4,378 51.2 0.020 30.4 19.6 18.6 28.9 52.4 52.9 52.2 Zambia 2013/2014 D Zimbabwe 2015 D 0.025 6.3 3,505 3,549 39.8 0.005 0.9 12.2 39.5 13.1 47.4 55.5 18.9 Developing countries — Regions 0.580 91.9 9,248 11,552 63.2 0.023 74.3 6.3 14.0 39.6 46.5 82.3 42.7 Arab States — East Asia and the Pacific — 0.279 52.3 19,748 21,210 53.4 0.023 30.9 17.7 21.1 29.2 49.8 46.5 14.9 Europe and Central Asia — Latin America and the Caribbean — 0.041f 9.4f 49f 53f 43.4f 0.018f 2.5f 4.5f 45.7f 25.5f 28.8f .. .. South Asia — Sub-Saharan Africa — 0.029g 7.4g 1,539g 1,350g 38.9g 0.006g 1.2g 7.7g 40.7g 49.0g 10.2g 35.2 ..

0.029 7.4 664 664 39.0 0.004 0.7 20.1 47.8 26.5 25.8 31.3 4.8

0.273 55.4 30,814 31,778 49.3 0.016 25.9 24.2 21.1 22.9 56.0 28.2 49.1

0.003g 0.8g 541g 542g 39.1g 0.007g 0.1g 7.2g 35.0g 47.4g 17.6g 8.6 0.0

0.210 45.8 581 594 45.7 0.014 16.3 26.1 27.8 24.2 48.0 41.8 30.7

0.249 48.2 3,481 3,755 51.6 0.023 24.3 21.8 21.7 28.4 50.0 55.1 49.2

0.002g 0.6g 8g 9g 38.0g ..e 0.1g 3.7g 45.5g 34.0g 20.5g .. ..

0.005 1.3 144 153 39.7 0.006 0.2 3.7 25.7 50.2 24.1 15.2 0.3

0.001 0.4 23 23 36.1 ..e 0.0 2.4 88.0 4.4 7.6 .. ..

0.269 55.1 22,857 23,614 48.8 0.017 24.1 24.9 22.4 22.5 55.1 21.4 41.7

0.001d 0.2d 109d 106d 34.5d ..e 0.0d 0.4d 59.7d 28.8d 11.5d 2.4 0.1

0.174g 38.8g 85g 107g 44.9g 0.012g 10.2g 32.3g 21.4g 22.5g 56.2g 12.7 13.1

0.019d 4.9d 4,530d 4,677d 39.5d 0.010d 0.7d 5.6d 15.2d 42.6d 42.2d 9.8 2.0

0.241 47.7 12,199 13,475 50.5 0.021 23.9 22.1 28.3 30.7 41.0 48.6 18.8

0.261 53.2 8,317 9,102 49.1 0.017 24.2 22.5 23.7 22.5 53.7 54.4 57.5

0.137 31.8 5,018 5,257 42.9 0.009 8.0 27.4 27.3 12.3 60.4 72.3 21.4

0.114 23.1 1,279,663 1,325,994 49.4 0.018 10.5 15.3 25.8 29.5 44.7 21.3 14.2

0.076 15.7 48,885 52,251 48.4 0.018 6.9 9.4 26.2 35.3 38.6 25.2 4.6

0.024 5.6 110,775 113,247 42.3 0.009 1.0 14.9 27.4 35.6 37.0 6.6 2.1

0.004 1.1 1,237 1,240 37.9 0.004 0.1 3.6 52.8 23.3 23.9 11.9 0.6

0.033 7.5 38,067 39,324 43.1 0.011 2.0 7.7 35.4 25.7 38.9 31.5 4.1

0.142 31.0 542,492 548,048 45.6 0.016 11.3 18.8 29.2 27.9 42.9 22.9 17.5

0.315 57.5 538,206 571,884 54.9 0.022 35.1 17.2 22.2 29.6 48.1 43.7 44.7

NOTES j Missing indicator on housing. default/files/hdr2019_technical_notes.pdf for details on how the Population living below national poverty line: Percentage of k Child mortality was constructed based on deaths that occurred Multidimensional Poverty Index is calculated. the population living below the national poverty line, which is the a Not all indicators were available for all countries, so caution poverty line deemed appropriate for a country by its authorities. between surveys—that is, between 2012 and 2014. Child deaths Multidimensional poverty headcount: Population with a National estimates are based on population-weighted subgroup should be used in cross-country comparisons. When an indicator reported by an adult man in the household were taken into deprivation score of at least 33 percent. It is expressed as a share of estimates from household surveys. is missing, weights of available indicators are adjusted to total account because the date of death was reported. the population in the survey year, the number of people in the survey 100 percent. See Technical note 5 at http://hdr.undp.org/sites/ year and the projected number of people in 2017. Population living below PPP $1.90 a day: Percentage of the default/files/hdr2019_technical_notes.pdf for details. l Missing indicator on cooking fuel. population living below the international poverty line of $1.90 (in m Missing indicator on electricity. Intensity of deprivation of multidimensional poverty: Average purchasing power parity [PPP] terms) a day. b D indicates data from Demographic and Health Surveys, M n Multidimensional Poverty Index estimates are based on the deprivation score experienced by people in multidimensional poverty. MAIN DATA SOURCES indicates data from Multiple Indicator Cluster Surveys, N 2016 National Health and Nutrition Survey. Estimates based Inequality among the poor: Variance of individual deprivation indicates data from national surveys and P indicates data from on the 2015 Multiple Indicator Cluster Survey are 0.010 for scores of poor people. It is calculated by subtracting the deprivation Column 1: Refers to the year and the survey whose data were used Pan Arab Population and Family Health Surveys (see http://hdr. Multidimensional Poverty Index value, 2.6 for multidimensional score of each multidimensionally poor person from the average to calculate the country’s Multidimensional Poverty Index value and undp.org/en/faq-page/multidimensional-poverty-index-mpi for poverty headcount (%), 3,125,000 for multidimensional intensity, squaring the differences and dividing the sum of the its components. the list of national surveys). poverty headcount in year of survey, 3,200,000 for projected weighted squares by the number of multidimensionally poor people. multidimensional poverty headcount in 2017, 40.2 for intensity Columns 2­12: HDRO and OPHI calculations based on data on c Data refer to the most recent year available during the period of deprivation, 0.4 for population in severe multidimensional Population in severe multidimensional poverty: Percentage of household deprivations in health, education and standard of poverty, 6.1 for population vulnerable to multidimensional the population in severe multidimensional poverty—that is, those living from various household surveys listed in column 1 using the specified. poverty, 39.9 for contribution of deprivation in health, 23.8 for with a deprivation score of 50 percent or more. methodology described in Technical note 5 (available at http://hdr. contribution of deprivation in education and 36.3 for contribution undp.org/sites/default/files/hdr2019_technical_notes.pdf) and d Missing indicator on nutrition. of deprivation in standard of living. Population vulnerable to multidimensional poverty: Percentage Alkire, Kanagaratnam and Suppa (2019). Columns 4 and 5 also use of the population at risk of suffering multiple deprivations—that is, population data from UNDESA (2017b). e Value is not reported because it is based on a small number of DEFINITIONS those with a deprivation score of 20­33 percent. Columns 13 and 14: World Bank (2019a). multidimensionally poor people. Multidimensional Poverty Index: Percentage of the population Contribution of deprivation in dimension to overall that is multidimensionally poor adjusted by the intensity of the multidimensional poverty: Percentage of the Multidimensional f Missing indicator on child mortality. deprivations. See Technical note 5 at http://hdr.undp.org/sites/ Poverty Index attributed to deprivations in each dimension.

g Considers child deaths that occurred at any time because the

survey did not collect the date of child deaths.

h The methodology was adjusted to account for missing indicator

on nutrition and incomplete indicator on child mortality (the survey did not collect the date of child deaths).

i Based on data accessed on 7 June 2016.

TABLE 6 Multidimensional Poverty Index: developing countries | 321 dashboards DASHBOARD 1 Quality of human development HUMAN DEVELOPMENT REPORT 2019

Top third Middle third Bottom third

Three-colour coding is used to visualize partial grouping of countries by indicator. For each indicator countries are divided into three groups of approximately equal size (terciles): the top third, the middle third and the bottom third. Aggregates are colour coded using the same tercile cutoffs. See Notes after the table.

SDG 4.c SDG 4.a SDG 4.1 SDG 7.1 SDG 6.1 SDG 6.2 Quality of health Quality of standard of living

Schools with access Population Pupil­ Primary using at Population teacher school Primary Secondary using at ratio, teachers Rural least basic primary trained least basic school to teach Programme for International population drinking- sanitation Lost health Hospital Student Assessment Vulnerable with access water

expectancy Physicians beds (PISA) score employmenta to electricity sources

(pupils per (% of total

(%) (per 10,000 people) teacher) (%) (%) Mathematicsb Readingc Sciencec employment) (%)

HDI rank 2017 2010­2018d 2010­2015d 2013­2018d 2010­2018d 2010­2018d 2010­2018d 2015 2015 2015 2018 2017 2017 2017

VERY HIGH HUMAN DEVELOPMENT 14.7 46.3 39 9 .. 100 100 502 513 498 4.8 100 100 98 2 Switzerland 14.3 42.4 47 10 .. 100 100 521 492 506 9.0 100 100 100 4 Germany 13.9 30.9 28 … .. 504 521 503 10.9 100 97 91 6 Australia 13.8 42.1 83 12 … 506 509 509 5.9 100 100 99 8 Sweden … 14 97 99 95 548 527 523 5.9 100 .. .. 10 Netherlands 14.6 35.9 38 .. .. 100 100 494 503 510 10.7 100 100 100 12 Finland 13.8 39.7 32 10 … 488 482 473 8.0 100 100 99 14 New Zealand 14.1 54.0 26 12 … 494 500 493 6.2 100 100 99 15 United States 12.5 23.1 24 15 99 .. .. 564 535 556 9.8 100 100 100 18 Liechtenstein 13.9 35.1 47 e 12 .. 100 100 512 503 509 12.6 100 100 98 20 Austria 13.9 44.6 25 11 .. 100 100 511 500 502 5.1 100 100 100 22 Israel 14.3 38.1 44 13 .. 100 100 511 526 531 9.2 100 100 99 24 Slovenia 14.0 26.1 27 … .. 516 527 528 10.7 100 99 99 26 Czechia 15.3 30.3 28 15 … 495 509 513 12.4 100 100 100 28 Malta 14.4 28.1 28 15 … 492 498 509 13.0 100 100 99 30 Estonia 15.3 25.9 29 14 .. 100 100 470 497 496 3.8 100 99 100 32 Greece 14.5 33.2 62 11 .. 100 100 507 499 502 10.2 100 100 99 34 Lithuania … 8 … … .. 100 .. .. 36 Andorra 13.2 24.1 134 16 … 532 516 538 8.4 100 99 100 36 Slovakia 13.9 51.4 76 10 … 497 485 495 7.7 100 100 100 40 Portugal 14.7 30.3 48 8 … 486 481 483 6.3 100 100 98 42 Chile 14.0 32.2 31 12 .. 85 85 470 479 467 8.3 100 100 100 43 Hungary 13.2 23.7 115 16 .. 100 100 524 517 516 23.5 100 100 100 46 Croatia 15.3 30.0 46 14 .. 100 100 510 505 513 10.6 100 100 99 48 Argentina 13.2 40.7 30 13 .. 100 100 486 496 493 11.3 100 100 100 50 Belarus 14.9 43.1 65 19 … 492 487 493 14.0 100 100 99 13.4 32.3 65 18 .. 98 99 493 499 495 7.4 100 100 99

13.8 38.3 47 13 … 479 447 465 9.9 100 100 100

13.6 40.9 34 11 .. 70 88 490 485 481 17.0 100 99 99

14.2 34.7 50 11 .. 100 100 520 519 534 5.5 100 100 99

13.5 19.5 34 12 … 437 443 433 11.1 100 100 99

13.7 45.9 43 9 … 454 467 455 26.7 100 100 99

14.4 24.0 65 11 .. 100 100 504 506 501 16.3 100 100 99

14.3 43.4 73 13 … 478 472 475 9.5 100 98 93

13.9 23.9 12 25 100 .. .. 427 434 437 0.8 100 98 99

13.9 33.3 25 e 11 100 100 100 … .. 100 100 100 DASH 13.7 23.9 27 12 100 100 100 … 2.9 100 100 100 1 14.3 24.6 58 15 .. 100 100 475 453 461 12.0 100 100 98

14.1 31.9 58 11 .. 100 100 482 488 490 7.9 100 99 92

13.9 33.4 34 13 .. 100 100 492 498 501 12.3 100 100 100

14.7 0.0 12 12 49 e 100 100 402 402 418 0.1 100 100 100

13.8 10.8 22 18 … 423 459 447 24.1 100 100 100

12.1 17.7 27 10 85 … .. .. 6.0 100 100 96 f

14.3 32.3 70 11 .. 100 99 477 470 477 5.7 100 100 98

14.7 9.3 20 12 84 100 100 … 1.1 100 100 100

14.1 30.0 56 14 … 464 487 475 7.6 100 100 97

14.7 19.7 16 10 100 71 87 … 2.6 100 92 100

12.8 39.6 50 .. .. 38 56 456 g 475 g 475 g 21.5 100 99 h 94 h

13.7 40.1 82 21 … 494 495 487 5.3 100 97 90

13.6 40.8 110 19 100 100 100 … 3.4 100 96 98

12.9 32.5 67 20 100 … .. .. 25.8 100 96 98

DASHBOARD 1 Quality of human development | 323 DASHBOARD 1 QUALITY OF HUMAN DEVELOPMENT

SDG 4.c SDG 4.a SDG 4.1 SDG 7.1 SDG 6.1 SDG 6.2 Quality of health Quality of standard of living

Schools with access Population Pupil­ Primary using at Population teacher school Primary Secondary using at ratio, teachers Rural least basic primary trained least basic school to teach Programme for International population drinking- sanitation Lost health Hospital Student Assessment Vulnerable with access water

expectancy Physicians beds (PISA) score employmenta to electricity sources

(pupils per (% of total

(%) (per 10,000 people) teacher) (%) (%) Mathematicsb Readingc Sciencec employment) (%)

HDI rank 2017 2010­2018d 2010­2015d 2013­2018d 2010­2018d 2010­2018d 2010­2018d 2015 2015 2015 2018 2017 2017 2017

52 Bulgaria 13.4 39.9 68 18 … 441 432 446 8.3 100 99 86 52 Romania 13.7 23.3 40 … .. 418 427 411 13.3 100 97 98 56 Barbados 14.0 22.6 63 19 … 444 434 435 25.2 100 100 84 57 Uruguay .. 11.8 48 … … .. .. 100 100 100 60 Bahamas 11.9 24.9 58 14 80 … .. .. 15.8 100 98 97 62 Seychelles 14.9 25.8 20 9 79 … .. .. 1.1 100 100 100 63 Serbia 12.7 50.5 28 11 100 100 100 418 437 435 24.0 100 99 97 65 Iran (Islamic Republic of) 13.9 17.6 27 18 … 420 428 425 28.0 100 99 97 67 Panama 11.7 19.4 29 19 90 … .. .. 9.9 100 99 95 69 Albania 11.6 15.1 19 12 99 100 100 … 21.8 100 97 100 71 Sri Lanka 11.7 9.5 36 14 84 86 97 … .. 100 96 100 73 Saint Kitts and Nevis 13.7 31.3 57 14 56 … .. .. 27.1 100 86 98 75 Bosnia and Herzegovina 12.4 26.7 30 .. 88 e .. .. 417 427 425 18.1 100 98 93 77 Thailand 15.1 11.4 15 29 100 11 36 … 41.3 100 95 88 79 Brazil 13.4 20.2 34 18 100 35 94 … 16.3 100 100 96 81 Armenia 12.5 15.7 23 21 99 … .. .. 32.2 100 96 83 82 North Macedonia 12.2 11.5 12 12 94 22 51 400 427 420 20.1 99 100 98 85 China 13.7 12.0 29 18 … 413 405 427 54.9 100 91 98 87 Azerbaijan 12.4 51.0 26 9 95 e 100 100 404 401 411 49.2 100 98 90 89 Dominican Republic 12.3 9.6 36 23 85 … .. .. 38.9 97 89 96 91 Tunisia 11.9 81.9 52 9 100 … .. .. 8.0 100 95 93 93 Lebanon .. 25.2 23 14 72 100 100 … .. 100 99 i 92 i 94 Saint Vincent and the Grenadines 12.6 27.6 38 12 55 .. 91 … .. 100 97 88 96 Venezuela (Bolivarian Republic of) 14.3 20.0 35 17 … … 19.3 100 96 95 98 Fiji 12.3 22.5 15 27 97 39 53 408 423 416 26.9 100 99 91 98 Suriname 12.3 8.1 21 16 100 99 97 415 409 421 47.3 100 100 99 103 Belize 12.0 14.5 37 16 64 100 100 … .. 96 96 91 105 Tonga 13.4 21.5 22 20 .. 32 69 377 407 401 27.6 100 98 88 107 Moldova (Republic of) 12.1 20.8 15 24 95 39 70 390 425 416 46.8 98 97 90 13.0 29.0 42 … … .. 40.2 100 100 94

14.4 18.3 19 24 100 .. .. 360 350 376 26.8 100 94 88

13.7 28.7 44 14 … 371 352 384 19.1 100 93 99

12.5 12.7 16 17 95 41 74 387 398 397 50.9 84 91 74

11.7 17.9 42 17 .. 93 98 531 j 494 j 518 j 43.8 100 93 85

12.4 20.5 15 25 82 37 69 … 46.2 100 94 88

12.4 34.5 47 15 98 53 61 … 55.0 100 91 93

13.5 30.1 88 13 87 48 94 … 14.9 100 94 96

12.2 15.6 16 19 95 23 .. 328 358 332 40.2 100 97 84

12.2 1.1 e 13 15 89 99 100 … 29.3 99 98 88

14.0 12.7 23 16 100 58 .. 367 361 386 20.6 100 96 91

12.5 28.9 70 30 100 71 83 … 48.9 56 83 58

DASH 15.0 22.7 29 12 … 396 347 386 27.6 100 93 98 14.8 3.7 18 23 99 .. 86 … 25.3 24 90 77 1 12.2 6.6 26 14 84 100 100 … 17.9 100 95 87

12.1 13.2 17 22 96 84 73 … 35.7 99 91 87

12.1 .. 8 … … .. 32.9 100 96 94

12.0 10.8 38 13 66 100 93 … .. 100 97 f 78 f

13.2 8.4 23 20 90 … .. .. 43.3 91 94 95

13.3 13.7 13 .. 92 5 22 … 38.5 99 100 90

12.4 12.3 31 13 98 … .. .. 12.1 91 95 84

14.6 23.4 14 21 100 67 91 380 408 409 8.6 100 99 97

12.5 11.3 13 20 73 … .. .. 27.1 98 98 88

12.7 10.4 43 e 10 90 100 100 … 19.3 100 99 99

13.2 5.2 26 22 92 … .. .. 53.3 98 100 93

12.5 12.8 10 29 100 … .. .. 33.8 90 94 77

13.6 32.0 58 18 99 85 87 420 416 428 34.3 100 89 76

12.0 22.2 74 … … .. 23.6 100 99 99

324 | HUMAN DEVELOPMENT REPORT 2019

SDG 4.c SDG 4.a SDG 4.1 SDG 7.1 SDG 6.1 SDG 6.2 Quality of health Quality of standard of living

Schools with access Population Pupil­ Primary using at Population teacher school Primary Secondary using at ratio, teachers Rural least basic primary trained least basic school to teach Programme for International population drinking- sanitation Lost health Hospital Student Assessment Vulnerable with access water

expectancy Physicians beds (PISA) score employmenta to electricity sources

(pupils per (% of total

(%) (per 10,000 people) teacher) (%) (%) Mathematicsb Readingc Sciencec employment) (%)

HDI rank 2017 2010­2018d 2010­2015d 2013­2018d 2010­2018d 2010­2018d 2010­2018d 2015 2015 2015 2018 2017 2017 2017

108 Uzbekistan 12.4 23.7 40 21 99 91 90 … 40.1 100 98 100 111 Indonesia 14.8 21.6 37 … … .. 5.7 70 99 100 113 South Africa 12.3 3.8 12 16 .. .. 51 386 397 403 47.3 96 89 73 115 Gabon 13.2 3.4 … 14 23 … 31.0 96 97 98 MEDIUM HUMAN DEVELOPMENT 13.9 9.1 .. 30 … … 9.7 67 93 76 118 Viet Nam 12.5 16.1 11 19 58 … .. .. 58.1 75 93 61 120 Iraq 14.2 3.6 63 … … .. 31.5 49 86 47 122 Kyrgyzstan 13.9 7.9 16 24 74 48 49 … 21.3 100 99 94 124 El Salvador 12.6 4.6 27 .. .. 26 … .. .. 92 88 83 126 Cabo Verde 11.7 8.2 26 20 100 .. .. 495 487 525 54.5 100 95 84 126 Nicaragua 15.2 .. .. 25 100 57 72 … 22.9 100 .. .. 130 Namibia 16.0 8.2 14 … … .. 25.9 100 97 94 132 Honduras 14.6 7.3 11 28 100 79 89 … 48.8 100 87 89 134 Bhutan 12.8 18.8 45 25 95 41 44 … 33.9 100 87 97 135 Micronesia (Federated States of) 12.7 8.0 16 .. 70 … .. .. 56.8 89 96 86 138 Congo 12.2 15.7 13 28 95 36 40 … 36.1 100 97 87 140 Lao People’s Democratic Republic 12.8 17.0 48 22 100 … .. .. 45.2 99 81 97 142 Ghana 13.1 7.7 21 21 93 10 100 … 28.8 90 87 74 144 Equatorial Guinea 12.3 3.6 6 20 .. 9 44 … 34.5 89 94 65 146 Cambodia 12.7 10.1 9 .. 75 … .. .. 39.4 68 82 74 147 Nepal 13.9 7.8 7 35 70 … .. .. 76.7 89 93 60 150 Cameroon 14.1 3.7 e 27 e .. 96 … .. .. 24.8 29 83 35 152 Pakistan 13.6 7.2 59 … … .. 71.2 72 78 54 LOW HUMAN DEVELOPMENT 12.3 3.1 7 26 .. 16 … .. 40.5 72 95 81 155 Papua New Guinea 13.5 2.0 19 25 73 … … 100 72 48 157 Rwanda 13.4 3.7 17 35 100 46 … .. 71.3 97 97 69 159 Tanzania (United Republic of) 13.7 5.3 8 30 50 4 82 … 55.5 81 97 48 161 Mauritania 13.4 1.9 e 32 e 20 … … .. 77 79 88 163 Benin 12.9 3.2 29 31 27 … .. .. 46.9 45 84 43 13.7 1.2 .. .. 80 … .. .. 76.9 24 73 20

14.2 0.8 21 27 70 16 69 … 32.9 67 69 58

12.0 5.0 15 22 97 … .. .. 80.0 91 82 74

13.0 1.7 17 e 27 … … 70.8 53 91 34

13.0 1.8 9 27 60 8 20 … 68.9 65 81 18

12.8 0.9 20 42 99 6 … .. 77.8 14 60 26

13.9 4.0 21 23 37 … .. .. 55.8 6 65 66

12.6 8.6 9 23 98 0 5 … 59.5 60 82 64

13.2 1.7 8 42 100 … .. .. 50.8 86 79 59

12.6 2.0 14 31 97 e … .. .. 53.5 58 59 29

13.8 6.5 3 21 97 … .. .. 79.4 95 89 62

14.3 2.1 .. 50 47 3 17 … 67.1 0 56 50

13.4 0.9 13 45 81 .. 23 … 73.8 21 60 39 DASH

13.2 0.8 17 36 86 … .. .. 65.6 19 64 36 BOARD

13.2 9.8 6 45 82 … .. .. 59.3 54 91 60 1

12.9 2.0 14 26 74 .. 14 … 80.3 60 68 34

14.3 12.2 15 … … .. 34.4 78 97 91

13.2 0.5 .. 36 … … 78.3 50 41 13

12.5 1.7 22 19 55 8 11 … 64.6 74 80 36

12.9 1.3 16 e 58 93 25 33 … 68.7 24 58 67

14.3 3.8 .. .. 66 … .. .. 78.4 23 71 39

12.7 0.4 7 47 99 … .. .. 82.7 17 57 30

13.2 0.9 5 43 80 … .. .. 75.2 11 49 18

13.6 1.8 .. 36 85 … .. .. 52.8 0 71 48

12.8 1.8 2 41 15 … .. .. 85.3 0 54 11

13.6 1.6 5 44 68 … .. .. 88.0 17 66 16

13.9 0.7 .. 33 87 … .. .. 54.7 20 69 43

DASHBOARD 1 Quality of human development | 325 DASHBOARD 1 QUALITY OF HUMAN DEVELOPMENT

SDG 4.c SDG 4.a SDG 4.1 SDG 7.1 SDG 6.1 SDG 6.2 Quality of health Quality of standard of living

Schools with access Population Pupil­ Primary using at Population teacher school Primary Secondary using at ratio, teachers Rural least basic primary trained least basic school to teach Programme for International population drinking- sanitation Lost health Hospital Student Assessment Vulnerable with access water

expectancy Physicians beds (PISA) score employmenta to electricity sources

(pupils per (% of total

(%) (per 10,000 people) teacher) (%) (%) Mathematicsb Readingc Sciencec employment) (%)

HDI rank 2017 2010­2018d 2010­2015d 2013­2018d 2010­2018d 2010­2018d 2010­2018d 2015 2015 2015 2018 2017 2017 2017

165 Côte d’Ivoire 13.3 2.3 .. 42 100 … .. .. 72.4 37 73 32 167 Togo 13.5 0.7 3e 33 75 17 83 … 65.1 35 81 51 169 Haiti 13.2 0.5 7 40 73 … .. .. 77.4 19 65 16 171 Djibouti 14.7 4.1 8 … … .. 40.0 43 60 37 173 Ethiopia 13.3 2.3 7 … … .. 85.0 3 65 35 174 Guinea 16.4 2.8 5 44 … … 89.4 97 67 43 177 Yemen 11.9 2.2 14 29 100 … .. .. 47.3 26 76 64 179 Congo (Democratic Republic of the) 13.0 0.2 13 70 91 … .. .. 59.5 4 69 26 181 Sierra Leone 13.0 1.0 3 .. 85 e … .. .. 86.0 31 41 7 182 Eritrea 13.7 1.1 11 36 100 … .. .. 72.3 21 78 39 185 Burundi 13.0 0.8 3 47 75 … .. .. 89.9 9 62 23 187 Chad 15.7 0.4 8 27 47 .. 5 … 77.7 7 73 17 189 Niger 16.6 3.1 7 27 … … 45.4 69 63 59 13.3 2.0 10 e .. 39 … .. .. 78.4 9 67 21 .. Monaco 14.4 0.9 .. 33 95 … .. .. 79.7 0 43 20 .. San Marino 13.2 0.7 7 52 97 … .. .. 83.1 2 56 29 .. Tuvalu 13.7 0.3 .. 39 54 0 3 … 86.3 5 61 16 13.5 0.6 4 41 86 .. 3 … 86.4 10 48 19 High human development 13.1 .. 7 39 41 … .. .. 78.2 30 52 h 12 h Low human development 14.2 1.4 1 38 52 … .. .. 89.6 12 78 39 Regions 12.5 0.5 8 50 100 .. 1 … 94.7 2 61 46 East Asia and the Pacific 14.5 .. .. 47 44 … .. .. 87.3 21 41 11 Latin America and the Caribbean 14.2 0.5 .. 57 65 … .. .. 93.1 2 39 8 Sub-Saharan Africa 13.5 0.6 10 83 … … 93.6 15 46 h 25 h Small island developing states 13.0 0.5 3 36 66 … .. .. 89.0 11 50 14 Cooperation and Development 11.8 36.7 132 20 … … 65.9 52 95 83 .. 65.6 138 10 .. 100 100 … .. 100 100 100

.. 12.4 50 40 100 … … .. 99 66

.. 61.5 38 … … .. .. 100 100 100

12.5 0.2 9 … … .. 77.7 9 52 38

.. 9.2 .. 17 77 … … 100 99 84

14.0 30.4 55 14 .. — — — — — 10.3 100 99 98

12.3 16.5 32 19 .. — — — — — 40.2 98 94 85

13.6 7.3 9 33 75 — — — — — 68.6 82 90 60

13.9 2.1 .. 41 80 — — — — — 79.1 24 59 29

13.0 11.5 21 25 .. — — — — — 53.3 77 88 69

14.5 11.1 15 21 .. — — — — — 24.5 82 89 83

11.9 14.8 35 18 .. — — — — — 45.0 96 92 83

13.4 24.9 51 18 .. — — — — — 28.4 100 96 97

DASH 12.7 21.6 20 21 .. — — — — — 32.7 92 97 87 13.9 7.8 8 35 72 — — — — — 71.6 86 93 60 1 13.6 2.1 .. 39 80 — — — — — 74.9 22 61 30

13.6 2.5 7 37 76 — — — — — 73.7 38 64 34

12.6 22.2 25 18 94 — — — — — 40.1 60 82 67

14.0 28.9 50 15 .. — — — — — 11.8 100 99 99

13.2 14.9 28 23 .. — — — — — 45.1 79 90 73

326 | HUMAN DEVELOPMENT REPORT 2019

NOTES j Refers to the provinces of Beijing, Guangdong, Programme for International Student Population using at least basic sanitation Jiangsu and Shanghai. Assessment (PISA) score: Score obtained in services: Percentage of the population using at Three-colour coding is used to visualize partial testing of skills and knowledge of 15-year-old least basic sanitation services—that is, improved grouping of countries and aggregates by indicator. DEFINITIONS students in mathematics, reading and science. sanitation facilities that are not shared with other For each indicator countries are divided into three households. This indicator encompasses people groups of approximately equal size (terciles): the Lost health expectancy: Relative difference Vulnerable employment: Percentage of employed using basic sanitation services as well as those top third, the middle third and the bottom third. between life expectancy and healthy life expectancy, people engaged as unpaid family workers and own- using safely managed sanitation services. Improved Aggregates are colour coded using the same tercile expressed as a percentage of life expectancy at account workers. sanitation facilities include flush/pour flush toilets cutoffs. See Technical note 6 at http://hdr.undp.org/ birth. connected to piped sewer systems, septic tanks sites/default/files/hdr2019_technical_notes.pdf for Rural population with access to electricity: or pit latrines; pit latrines with slabs (including details about partial grouping in this table. Physicians: Number of medical doctors People living in rural areas with access to electricity, ventilated pit latrines); and composting toilets. (physicians), both generalists and specialists, expressed as a percentage of the total rural a Estimates modelled by the International Labour expressed per 10,000 people. population. It includes electricity sold commercially MAIN DATA SOURCES Organization. (both on grid and off grid) and self-generated Hospital beds: Number of hospital beds available, electricity but excludes unauthorized connections. Column 1: HDRO calculations based on data on life b Average score for Organisation for Economic expressed per 10,000 people. expectancy at birth and healthy life expectancy at Co-operation and Development (OECD) countries Population using at least basic drinking-water birth from IHME (2018). is 490. Pupil­teacher ratio, primary school: Average services: Percentage of the population using at number of pupils per teacher in primary education. least basic drinking-water services—that is, the Columns 2, 13 and 14: WHO (2019). c Average score for OECD countries is 493. population that drinks water from an improved Primary school teachers trained to teach: source, provided collection time is not more than 30 Columns 3 and 12: World Bank (2019a). d Data refer to the most recent year available Percentage of primary school teachers who have minutes for a round trip. This indicator encompasses during the period specified. received the minimum organized teacher training people using basic drinking-water services as well Columns 4­7: UNESCO Institute for Statistics (preservice or in-service) required for teaching at the as those using safely managed drinking-water (2019). e Refers to a year from 2007 to 2009. primary level. services. Improved water sources include piped water, boreholes or tubewells, protected dug wells, Columns 8­10: OECD (2017). f Refers to 2015. Schools with access to the Internet: Percentage protected springs, and packaged or delivered water. of schools at the indicated level with access to the Column 11: ILO (2019). g Refers to the adjudicated region of Ciudad Internet for educational purposes. Autónoma de Buenos Aires.

h Refers to 2016.

i Refers to 2013.

1

DASHBOARD 1 Quality of human development | 327 DASHBOARD 2 Life-course gender gap

Top third Middle third Bottom third

Three-colour coding is used to visualize partial grouping of countries by indicator. For each indicator countries are divided into three groups of approximately equal size (terciles): the top third, the middle third and the bottom third. Aggregates are colour coded using the same tercile cutoffs. See Notes after the table.

SDG 4.2 SDG 4.1 SDG 4.1 SDG 8.5 SDG 4.6 SDG 8.5 SDG 8.3 SDG 5.5 SDG 5.4 SDG 1.3

Gross enrolment ratio Population with Share of Youth at least some Total employment in Share of Time spent on unpaid Old-age Sex ratio unemployment secondary unemployment nonagriculture, seats in and care work recipients

HDI rank (male to Pre-primary Primary Secondary (female to (female to (female to (% of total (% held by (% of (female to (female to female 2013­2018c 2013­2018c 2013­2018c male ratio) male ratio) male ratio) employment in women) 24-hour day) male ratio) male ratio) VERY HIGH HUMAN DEVELOPMENT births) nonagriculture) 1 Norway 2018 2010­2018c 2018 2018 2008­2018c 2008­2018c 2013­2017c 2 Switzerland 2015­2020b 2018 4 Germany 1.06 1.00 1.00 0.96 0.72 1.01 0.81 47.9 41.4 15.3 1.2 0.87 6 Australia 1.05 0.99 0.99 0.96 0.96 0.99 1.11 46.6 29.3 16.8 1.6 1.04 8 Sweden 1.06 0.98 0.99 1.03 0.84 1.05 0.93 47.4 24.3 .. .. 0.61 10 Netherlands 1.05 0.99 0.99 0.95 0.74 0.99 0.84 46.9 31.5 15.9 d 1.6 d 1.00 12 Finland 1.08 0.99 0.98 0.96 0.87 0.92 0.83 49.4 .. 10.8 3.3 .. 14 New Zealand 1.06 0.96 1.00 0.89 0.80 0.99 1.04 46.8 32.7 .. .. 1.06 15 United States 1.05 1.02 1.00 1.00 0.67 1.00 0.93 48.2 38.1 .. .. 1.12 18 Liechtenstein 1.06 1.00 1.03 1.12 0.84 1.00 0.90 48.2 46.1 16.0 1.3 1.00 20 Austria 1.07 .. 1.00 0.99 1.92 0.92 1.17 45.1 23.0 … 22 Israel 1.05 1.00 1.00 1.02 0.84 0.96 1.17 46.4 35.6 14.7 e 1.6 e 1.00 24 Slovenia 1.06 0.99 0.99 1.03 0.76 1.00 1.08 47.9 37.4 15.6 e 1.4 e 1.02 26 Czechia 1.05 1.00 1.00 1.10 0.92 1.00 0.96 48.9 42.0 14.5 d 1.5 d 1.00 28 Malta 1.05 .. 1.00 1.01 0.78 1.00 0.93 47.7 31.7 14.6 1.5 1.00 30 Estonia 1.06 0.99 1.00 1.06 0.91 1.01 1.12 48.2 38.3 18.1 f 1.7 f 1.00 32 Greece 1.05 1.00 1.00 1.11 0.86 0.97 0.98 47.0 28.9 12.7 1.8 1.00 34 Lithuania 1.05 1.00 1.00 0.99 0.74 1.00 0.93 46.4 23.6 15.4 1.6 0.87 36 Andorra 1.05 1.00 1.00 1.12 0.91 0.95 1.02 46.0 41.4 15.9 f 1.6 f 1.00 36 Slovakia .. 1.06 0.96 0.78 … .. 12.0 … 40 Portugal 1.06 .. 1.00 1.01 0.85 1.03 0.88 43.9 13.7 14.4 d 4.7 d .. 42 Chile 1.06 0.99 1.00 0.96 1.02 1.00 0.98 46.9 34.8 18.3 d 1.9 d 0.99 43 Hungary 1.05 0.97 1.00 1.03 0.72 1.00 1.08 46.1 20.0 14.4 d 2.0 d 0.66 46 Croatia 1.05 1.00 1.01 1.02 0.97 0.97 1.03 47.3 27.5 … 48 Argentina 1.06 1.00 1.00 1.00 0.99 0.94 0.95 42.3 17.0 14.0 d 4.2 d 0.96 1.06 0.97 1.00 1.02 1.36 0.99 1.31 46.6 20.0 …

1.06 1.00 1.01 1.01 0.94 0.93 1.29 46.1 38.6 19.0 e 2.2 e 0.47

1.06 0.97 1.01 1.01 1.13 1.00 1.45 44.8 20.3 .. .. 1.00

1.05 1.00 0.99 1.01 0.93 0.94 1.01 47.4 35.7 15.8 1.7 1.00

1.06 1.03 1.04 1.04 0.85 0.90 1.00 39.8 11.9 .. .. 0.43

1.06 0.97 1.00 0.98 1.20 0.91 1.18 42.4 35.6 20.4 2.4 0.83

1.07 .. 1.00 1.01 0.73 1.00 0.86 49.5 26.7 17.2 d 1.6 d 1.00

1.07 0.99 1.00 0.99 0.59 0.95 1.01 47.0 17.9 .. .. 0.77

1.07 1.01 1.00 0.94 1.22 0.84 1.54 41.6 18.7 17.5 d 2.6 d ..

1.06 0.97 1.01 0.97 0.97 0.94 1.00 45.6 25.5 17.6 d 1.8 d 1.00

1.06 1.00 1.00 0.96 0.88 0.95 0.85 52.2 21.3 .. .. 1.00

1.05 1.08 0.97 0.94 2.00 1.20 4.41 14.9 22.5 …

… .. .. 0.97 .. .. 32.1 …

DASH 1.03 1.05 0.98 0.77 2.12 0.90 6.77 14.9 19.9 … 1.05 0.98 0.99 1.01 1.03 0.99 1.13 46.1 20.0 .. .. 1.00 2 1.07 0.99 1.00 0.99 1.06 1.01 0.76 52.0 31.0 .. .. 1.00

1.06 0.98 0.96 0.97 1.13 0.98 1.17 49.7 34.8 17.8 1.7 0.77

1.05 1.03 0.99 1.25 8.33 1.11 6.00 14.2 9.8 8.2 3.7 0.36

1.04 0.98 0.97 1.01 1.20 0.98 1.16 43.0 22.7 22.1 f 2.2 f 1.59

1.06 1.03 0.99 1.02 1.04 0.98 1.17 43.4 9.1 …

1.06 0.96 1.00 0.99 1.43 0.98 1.18 46.5 12.6 16.6 d 2.2 d 1.00

1.04 0.99 1.00 1.01 6.10 1.12 11.67 20.2 18.8 …

1.06 0.96 1.01 1.05 1.66 0.98 1.28 46.6 18.5 …

1.05 1.05 1.03 0.97 4.79 1.15 7.59 12.0 8.8 18.9 2.5 ..

1.04 1.01 1.00 1.04 1.34 1.05 1.27 41.2 39.5 23.4 2.5 ..

1.06 0.98 1.01 0.99 1.09 1.01 0.94 49.4 16.1 18.4 2.3 1.00

328 | HUMAN DEVELOPMENT REPORT 2019

SDG 4.2 SDG 4.1 SDG 4.1 SDG 8.5 SDG 4.6 SDG 8.5 SDG 8.3 SDG 5.5 SDG 5.4 SDG 1.3

Gross enrolment ratio Population with Share of Youth at least some Total employment in Share of Time spent on unpaid Old-age Sex ratio unemployment secondary unemployment nonagriculture, seats in and care work recipients

(male to Pre-primary Primary Secondary (female to (female to (female to (% of total (% held by (% of (female to (female to female 2013­2018c 2013­2018c 2013­2018c male ratio) male ratio) male ratio) employment in women) 24-hour day) male ratio) male ratio) births) nonagriculture) HDI rank 0.96 1.00 0.98 2018 2010­2018c 2018 2018 2008­2018c 2008­2018c 2013­2017c 2015­2020b 1.02 1.02 1.01 0.66 0.94 0.56 2018 33.1 50 Belarus 0.99 0.99 0.97 1.13 0.99 1.33 22.1 19.2 d 2.0 d .. 50 Kazakhstan 1.06 0.98 0.99 1.00 0.84 0.98 0.84 52.4 23.8 52 Bulgaria 1.07 1.00 0.99 0.99 0.84 0.90 1.05 48.6 23.5 17.9 d 3.0 d .. 52 Montenegro 1.06 1.09 0.96 1.05 0.99 0.94 0.77 47.9 18.7 52 Romania 1.07 1.04 0.98 1.04 1.00 44.1 13.8 18.5 e 2.0 e 1.00 55 Palau 1.06 1.00 1.00 1.08 .. 1.03 .. 44.1 27.5 56 Barbados 1.02 0.98 1.12 1.15 1.10 … 57 Kuwait .. 0.95 0.99 .. 4.18 1.07 5.11 .. 3.1 57 Uruguay 1.04 1.07 1.05 0.98 1.43 0.67 1.49 50.0 22.3 19.0 d 2.0 d 1.00 59 Turkey 1.05 1.04 1.01 1.06 1.39 0.97 1.42 31.8 17.4 60 Bahamas 1.05 1.03 1.01 1.05 1.59 0.98 1.28 46.9 21.8 … 61 Malaysia 1.05 1.07 1.13 .. 1.23 28.3 15.8 62 Seychelles 1.06 47.1 21.2 … HIGH HUMAN DEVELOPMENT 1.06 .. .. 39.9 63 Serbia 1.06 … 63 Trinidad and Tobago .. 65 Iran (Islamic Republic of) 19.9 2.4 1.04 67 Panama 19.2 5.2 .. 69 Albania … 71 Sri Lanka … 73 Saint Kitts and Nevis … 75 Bosnia and Herzegovina 1.07 1.00 1.00 1.01 1.17 0.92 1.14 45.2 34.4 19.2 2.2 .. 77 Thailand 1.04 … 1.05 1.05 1.11 43.2 30.1 … 79 Brazil 1.05 1.00 1.03 1.02 1.85 0.93 1.99 16.5 5.9 21.0 4.0 0.10 81 Armenia 1.04 1.00 1.02 1.07 1.55 0.96 2.10 38.5 11.6 … 82 North Macedonia 1.05 1.03 0.98 1.03 1.61 1.09 1.59 41.9 18.3 17.7 2.4 .. 85 China 1.05 1.00 1.01 1.05 1.47 1.03 1.51 40.7 45.6 21.3 f 2.6 f .. 87 Azerbaijan 1.09 0.99 0.97 0.94 0.82 1.01 0.90 39.4 27.9 21.7 d 6.3 d .. 89 Dominican Republic 1.07 .. 1.01 1.02 1.20 0.99 0.83 44.3 16.0 .. .. 0.92 91 Tunisia 1.04 0.97 0.99 1.05 1.76 0.99 2.33 32.5 5.8 … 93 Lebanon 1.06 1.00 0.95 1.03 0.92 0.98 1.19 42.3 53.2 … 94 Saint Vincent and the Grenadines … … .. .. 13.3 … 96 Venezuela (Bolivarian Republic of) 1.03 1.09 0.97 0.96 … .. 31.4 .. .. 0.95 98 Fiji 1.07 … 1.17 0.81 1.26 37.4 19.3 … 98 Suriname 1.05 1.02 1.01 1.09 1.09 0.96 1.03 40.1 48.4 28.1 f 3.0 f 0.84 103 Belize 1.06 0.99 1.00 0.96 1.68 0.89 1.17 47.5 5.3 11.8 g 3.2 g .. 105 Tonga 1.05 1.06 0.95 1.05 … .. 39.3 …

1.05 1.05 0.97 1.05 1.26 1.06 1.30 44.9 15.0 13.3 4.3 ..

1.05 .. 0.97 1.06 1.63 1.04 1.66 46.1 19.0 16.3 d 3.7 d 0.99

1.11 1.10 1.00 1.05 1.50 0.99 1.02 43.6 18.1 21.7 5.0 1.17

1.05 .. 0.95 .. 1.73 1.00 2.11 17.2 21.3 21.7 f 5.8 f ..

1.06 0.99 1.00 0.98 1.00 0.72 0.91 39.8 38.3 15.4 d 2.8 d ..

1.05 1.01 1.00 1.00 1.31 0.84 1.42 46.4 27.7 22.7 f 2.6 f ..

1.13 1.01 1.01 1.02 0.81 0.91 0.78 45.4 24.9 15.3 2.6 ..

1.05 1.05 1.01 1.03 1.64 1.00 1.56 42.5 38.0 19.8 4.4 ..

1.13 1.00 1.02 .. 1.27 0.96 1.39 44.0 16.8 25.4 2.9 1.51

1.06 0.97 1.02 0.98 0.88 0.99 0.77 49.3 12.3 …

1.05 1.02 0.93 1.08 2.07 1.08 1.95 42.8 24.3 16.7 4.4 ..

1.03 1.08 .. 1.01 1.23 1.17 1.26 48.6 20.7 …

1.06 1.00 0.97 1.11 1.12 0.78 1.75 25.3 31.3 …

1.03 1.00 0.98 .. 1.42 1.06 0.88 47.3 17.1 17.6 f 2.8 f ..

1.05 0.96 0.92 0.99 1.34 0.98 1.98 22.8 4.7 … DASH

1.03 1.04 0.97 .. 1.44 0.99 1.45 47.7 9.5 … BOARD

1.03 1.05 0.98 0.96 1.04 .. 0.82 47.5 13.0 .. .. 2 ..

1.05 1.01 .. 1.06 1.47 1.12 1.73 48.1 19.0 …

1.05 1.01 0.97 1.08 1.44 1.08 1.13 41.2 22.2 .. .. 0.72

.. 1.03 0.97 0.99 … .. 25.0 …

1.06 .. 0.99 .. 1.92 1.12 1.47 33.2 19.6 15.2 2.9 ..

1.05 1.01 .. .. 1.46 0.98 1.45 41.9 16.0 14.5 3.4 0.80

1.08 1.01 1.00 1.32 2.37 1.02 2.54 37.6 25.5 …

1.05 .. .. 1.03 1.64 0.96 1.73 16.5 15.4 …

1.03 1.05 0.95 1.05 2.83 1.01 2.83 42.9 11.1 …

1.07 1.00 1.00 .. 0.63 0.91 0.92 28.9 5.9 …

1.05 1.01 0.97 1.06 4.50 1.01 5.00 51.7 7.4 …

DASHBOARD 2 Life-course gender gap | 329 DASHBOARD 2 LIFE-COURSE GENDER GAP

SDG 4.2 SDG 4.1 SDG 4.1 SDG 8.5 SDG 4.6 SDG 8.5 SDG 8.3 SDG 5.5 SDG 5.4 SDG 1.3

Gross enrolment ratio Population with Share of Youth at least some Total employment in Share of Time spent on unpaid Old-age Sex ratio unemployment secondary unemployment nonagriculture, seats in and care work recipients

(male to Pre-primary Primary Secondary (female to (female to (female to (% of total (% held by (% of (female to (female to female 2013­2018c 2013­2018c 2013­2018c male ratio) male ratio) male ratio) employment in women) 24-hour day) male ratio) male ratio) births) nonagriculture) HDI rank 0.99 0.97 1.10 2018 2010­2018c 2018 2018 2008­2018c 2008­2018c 2013­2017c 2015­2020b 0.99 1.00 0.99 1.19 1.04 1.04 2018 106 Philippines 0.97 0.98 0.96 0.94 0.98 0.79 29.1 … 107 Moldova (Republic of) 1.06 0.96 0.98 0.99 0.55 .. 0.42 43.4 22.8 108 Turkmenistan 1.06 1.04 1.00 0.93 52.1 24.8 19.5 d 1.8 d .. 108 Uzbekistan 1.05 … 1.57 1.54 1.65 42.8 16.4 110 Libya 1.06 0.89 0.96 1.03 1.03 0.84 0.93 39.0 16.0 … 111 Indonesia 1.06 1.13 1.00 1.10 1.61 1.11 1.34 22.0 19.8 111 Samoa 1.05 1.00 0.96 1.09 1.22 0.96 1.17 40.1 10.0 … 113 South Africa 1.08 1.00 0.98 0.97 1.52 0.81 1.48 38.2 41.8 h 114 Bolivia (Plurinational State of) 1.03 1.35 1.32 2.01 44.6 51.8 … 115 Gabon 1.05 … 1.53 0.83 2.96 41.5 17.4 i 116 Egypt 1.03 0.99 1.00 0.98 25.1 14.9 … MEDIUM HUMAN DEVELOPMENT 1.06 17.4 117 Marshall Islands … 119 Palestine, State of 15.6 d 2.4 d .. 121 Morocco … 123 Guyana … 125 Tajikistan 22.4 d 9.2 d .. 126 Guatemala .. 0.93 1.02 1.10 .. 0.99 .. .. 9.1 … 129 India 1.12 0.98 1.00 .. 1.01 0.85 0.90 47.2 26.7 … 131 Timor-Leste 1.05 1.00 1.00 1.10 1.77 0.97 2.06 14.7 .. 17.8 d 6.0 d .. 132 Kiribati 1.07 … 1.97 0.70 1.71 13.0 25.2 … 135 Bangladesh 1.06 0.83 0.95 0.89 1.03 0.81 1.21 15.7 18.4 20.8 7.0 .. 137 Sao Tome and Principe 1.06 1.01 0.99 1.00 1.62 1.00 1.48 38.7 19.2 16.8 f 1.8 f .. 138 Eswatini (Kingdom of) 1.05 … 1.65 1.28 1.54 39.1 31.9 … 141 Vanuatu 1.05 1.01 0.97 0.99 1.24 0.86 0.76 49.0 31.0 22.7 2.9 .. 143 Zambia 1.07 0.86 0.99 0.90 0.90 1.14 0.84 20.6 20.0 … 145 Myanmar 1.03 1.02 0.93 1.10 1.10 0.92 1.08 50.2 20.8 j … 147 Kenya 1.05 1.02 0.97 0.95 1.82 1.03 1.68 43.3 12.7 17.8 7.5 0.50 149 Angola 1.05 … 1.99 1.04 1.36 51.1 45.7 … 150 Zimbabwe 1.10 0.93 1.17 1.02 1.32 0.61 1.57 16.7 11.7 … 153 Solomon Islands 1.01 1.05 0.97 .. 1.32 0.97 1.14 48.5 39.7 … 154 Syrian Arab Republic 1.05 1.02 0.97 1.08 2.03 .. 1.50 31.7 33.8 .. .. 1.13 156 Comoros 1.05 1.01 1.00 1.14 2.05 1.05 1.56 48.2 21.1 17.3 4.0 .. 158 Nigeria 1.06 .. 1.06 … .. .. 6.5 … 159 Uganda 1.04 1.06 1.00 1.10 1.48 0.43 1.76 32.2 15.3 15.0 2.5 ..

1.05 1.04 1.07 1.17 1.57 0.92 1.97 20.2 20.3 …

1.06 0.92 1.00 … .. .. 0.0 …

1.03 1.09 0.96 1.15 2.25 0.69 2.40 38.3 14.5 …

1.03 … 0.93 0.91 1.14 47.6 14.0 …

1.03 .. 0.92 0.98 1.10 0.93 1.15 40.9 12.1 …

1.05 1.03 0.97 0.93 0.94 0.76 0.86 47.0 27.5 10.4 d 4.2 d ..

1.07 0.97 0.98 1.06 1.10 .. 1.24 42.6 0.0 …

1.05 1.02 1.02 0.99 0.97 0.78 1.00 53.4 12.7 14.4 d 4.1 d ..

1.03 1.07 1.02 .. 0.99 0.75 0.92 39.5 18.0 .. .. 0.22

1.03 1.02 0.99 .. 1.08 .. 1.11 36.9 18.0 …

1.03 1.01 0.95 1.10 1.58 1.29 1.75 43.7 10.2 …

1.05 1.04 0.98 .. 0.86 0.54 0.75 48.5 19.3 .. .. 0.15

1.03 0.98 1.00 .. 0.99 0.80 0.98 41.4 23.3 …

1.07 0.94 1.06 1.11 0.62 0.66 0.73 34.6 33.5 …

DASH 1.03 0.88 0.86 0.63 0.99 0.61 1.10 43.6 30.5 … 1.03 1.02 0.90 0.86 1.19 0.80 1.34 41.8 29.3 14.6 d 3.1 d .. 2 1.02 1.02 0.98 0.98 1.23 0.84 1.23 42.5 34.3 …

1.09 0.87 0.86 0.81 1.57 0.57 2.04 10.0 20.0 …

1.07 1.02 0.99 .. 0.93 .. 0.80 42.3 2.0 …

1.05 0.96 0.97 1.00 2.55 0.86 3.43 12.8 13.2 …

1.08 0.99 0.91 0.73 0.58 0.66 0.38 45.4 0.0 …

1.05 1.03 0.96 1.06 0.79 .. 1.17 35.9 6.1 …

1.02 1.03 0.99 1.12 1.67 0.72 1.00 36.1 55.7 …

1.06 .. 0.94 0.90 0.97 .. 1.12 52.6 5.8 …

1.03 1.01 1.02 1.01 1.41 0.70 1.60 44.3 37.2 16.5 k 3.9 k ..

1.03 1.04 1.03 .. 1.41 0.79 1.50 39.2 34.3 …

330 | HUMAN DEVELOPMENT REPORT 2019

SDG 4.2 SDG 4.1 SDG 4.1 SDG 8.5 SDG 4.6 SDG 8.5 SDG 8.3 SDG 5.5 SDG 5.4 SDG 1.3

Gross enrolment ratio Population with Share of Youth at least some Total employment in Share of Time spent on unpaid Old-age Sex ratio unemployment secondary unemployment nonagriculture, seats in and care work recipients

(male to Pre-primary Primary Secondary (female to (female to (female to (% of total (% held by (% of (female to (female to female 2013­2018c 2013­2018c 2013­2018c male ratio) male ratio) male ratio) employment in women) 24-hour day) male ratio) male ratio) births) 2010­2018c nonagriculture) HDI rank 1.26 1.06 0.96 2018 2018 2018 2008­2018c 2008­2018c 2013­2017c 2015­2020b 1.09 1.00 1.01 1.19 0.51 1.42 2018 20.3 161 Mauritania 1.04 0.94 0.76 1.25 .. 1.20 19.6 … 162 Madagascar 1.05 1.05 0.97 1.36 1.10 1.10 31.2 163 Benin 1.03 1.01 0.91 0.75 1.38 0.54 1.30 53.7 7.2 … 164 Lesotho 1.04 1.12 1.16 1.09 1.57 1.31 1.55 55.6 22.7 165 Côte d’Ivoire 1.03 1.04 0.95 0.73 1.28 0.52 1.24 56.2 9.2 l … 166 Senegal 1.03 1.02 0.94 1.02 0.61 0.52 0.70 47.3 41.8 167 Togo 1.04 2.16 0.51 2.52 41.8 17.6 … 168 Sudan 1.02 … 1.59 0.78 1.49 53.6 31.0 169 Haiti 1.04 .. 0.69 0.57 1.76 0.67 2.18 16.8 2.7 … 170 Afghanistan 1.05 0.94 0.88 0.84 1.08 0.36 1.15 60.6 27.4 i 171 Djibouti 1.06 1.01 1.04 0.94 1.18 1.42 25.5 26.2 … 172 Malawi 1.04 0.95 0.91 0.96 1.80 .. 1.85 41.3 16.7 173 Ethiopia 1.03 1.07 1.09 1.92 0.68 1.88 39.5 37.3 … 174 Gambia 1.04 .. 0.82 .. 0.64 0.52 0.59 55.6 10.3 174 Guinea 1.03 1.01 0.92 0.66 1.57 0.71 1.05 38.7 21.9 … 176 Liberia 1.02 0.90 0.87 0.78 1.37 1.94 44.4 11.7 177 Yemen 1.05 .. 0.73 1.03 .. 1.08 48.7 0.5 … 178 Guinea-Bissau 1.05 1.07 .. 0.60 0.47 0.66 4.4 13.7 179 Congo (Democratic Republic of the) 1.03 .. 0.99 .. 0.89 0.56 1.06 44.4 8.2 … 180 Mozambique 1.03 1.10 0.93 0.64 0.42 0.69 36.1 39.6 181 Sierra Leone 1.02 0.99 1.01 0.91 2.31 .. 2.32 33.2 12.3 … 182 Burkina Faso 1.02 0.98 0.98 0.95 1.09 0.56 1.11 53.1 11.0 182 Eritrea 1.05 1.07 0.86 0.97 1.19 0.51 1.38 48.5 22.0 … 184 Mali 1.05 1.02 0.89 0.90 0.43 0.60 0.55 41.6 8.8 185 Burundi 1.05 0.95 1.00 0.81 0.87 0.50 1.21 45.2 38.8 19.3 d 2.9 d .. 186 South Sudan 1.03 0.93 0.71 1.02 1.14 1.37 24.1 26.6 187 Chad 1.04 1.03 0.78 0.54 1.12 .. 1.20 36.7 15.3 … 188 Central African Republic 1.03 1.06 0.76 0.46 0.17 0.45 0.50 39.9 8.6 189 Niger 1.03 0.87 0.66 0.68 41.9 17.0 … OTHER COUNTRIES OR TERRITORIES 1.05 0.73 51.4 .. 6.3 2.4 .. .. Korea (Democratic People’s Rep. of) 0.17 .. Monaco 0.43 … .. Nauru 0.48 .. San Marino … .. Tuvalu … High human development … Low human development .. .. 0.13 Regions … East Asia and the Pacific .. .. 0.11 Latin America and the Caribbean … Sub-Saharan Africa … Small island developing states … Cooperation and Development …

1.05 .. 1.00 1.01 0.80 .. 0.83 41.9 16.3 …

… … .. .. 33.3 …

.. 1.05 1.03 1.03 … .. 10.5 …

… … .. .. 26.7 …

1.03 … 1.12 .. 1.13 17.5 24.3 …

.. 1.04 0.97 1.25 … .. 6.7 …

1.05 0.99 1.00 0.99 1.08 0.98 1.15 44.3 27.2 — — 0.93

1.08 0.99 0.99 1.03 1.17 0.92 1.15 42.8 24.4 — — ..

1.08 0.96 1.08 1.00 1.32 0.67 1.51 22.8 20.8 — — ..

1.04 1.01 0.94 0.84 1.20 0.59 1.46 43.5 21.3 — — ..

1.07 0.98 1.01 0.99 1.24 0.84 1.30 36.8 22.4 — — ..

1.05 0.98 0.96 0.93 1.67 0.84 2.46 16.3 18.3 — — ..

1.10 0.99 0.99 1.02 0.90 0.90 0.81 44.8 20.3 — — .. DASH 1.06 0.98 1.00 0.98 1.17 0.91 1.09 40.0 21.2 — — .. 2 1.05 1.02 0.99 1.05 1.33 1.01 1.31 43.6 31.0 — — ..

1.09 0.94 1.09 1.00 1.41 0.66 1.74 17.0 17.1 — — ..

1.04 1.00 0.96 0.88 1.06 0.72 1.16 46.9 23.5 — — ..

1.04 1.00 0.96 0.92 1.32 0.72 1.52 36.6 22.5 — — ..

1.06 .. 0.95 1.00 1.55 0.96 1.48 44.1 24.6 — — ..

1.05 0.99 1.00 1.01 0.98 0.97 1.08 44.7 30.1 — — 0.91

1.07 0.98 1.01 0.99 1.20 0.88 1.24 39.2 24.1 — — ..

DASHBOARD 2 Life-course gender gap | 331 DASHBOARD 2 LIFE-COURSE GENDER GAP

NOTES i Refers to 2017. percentage of the female population ages 25 provision of services for own final use by household j Refers to 2013. and older that has reached (but not necessarily members or by family members living in other Three-colour coding is used to visualize partial k Refers to the population ages 5 and older. completed) a secondary level of education to the households. grouping of countries and aggregates by indicator. For l Refers to 2015. percentage of the male population ages 25 and older each indicator countries are divided into three groups with the same level of education achievement. Old-age pension recipients, female to male of approximately equal size (terciles): the top third, DEFINITIONS ratio: Ratio of the percentage of women above the middle third and the bottom third. Aggregates are Total unemployment rate, female to male ratio: the statutory pensionable age receiving an old-age colour coded using the same tercile cutoffs. Sex ratio Sex ratio at birth: Number of male births per Ratio of the percentage of the female labour force pension (contributory, noncontributory or both) to the at birth is an exception—countries are divided into female birth. population ages 15 and older that is not in paid percentage of men above the statutory pensionable two groups: the natural group (countries with a value employment or self-employed but is available for age receiving an old-age pension (contributory, of 1.04­1.07, inclusive), which uses darker shading, Gross enrolment ratio, female to male ratio: work and is actively seeking paid employment or noncontributory or both). and the gender-biased group (all others), which uses For a given level of education (pre-primary, primary, self-employment to the percentage of the male lighter shading. See Technical note 6 at http://hdr. secondary), the ratio of the female gross enrolment labour force population ages 15 and older that is not MAIN DATA SOURCES undp.org/sites/default/files/hdr2019_technical_notes. ratio to the male gross enrolment ratio. The gross in paid employment or self-employed but is available pdf for details about partial grouping in this table. enrolment ratio (female or male) is the total for work and is actively seeking paid employment or Column 1: UNDESA (2019b). enrolment in a given level of education, regardless of self-employment. a The natural sex ratio at birth is commonly age, expressed as a percentage of the official school- Columns 2­4: UNESCO Institute for Statistics assumed and empirically confirmed to be 1.05 age population for the same level of education. Share of employment in nonagriculture, (2019). male births to 1 female birth. female: Share of women in employment in the Youth unemployment rate, female to male nonagricultural sector, which comprises industry and Columns 5 and 7: HDRO calculations based on b Data are average annual estimates for ratio: Ratio of the percentage of the female labour services activities. ILO (2019). 2015­2020. force population ages 15­24 that is not in paid employment or self-employed but is available for Share of seats in parliament: Proportion of seats Column 6: HDRO calculations based on UNESCO c Data refer to the most recent year available work and is actively seeking paid employment or held by women in the national parliament, expressed Institute for Statistics (2019) and Barro and Lee during the period specified. self-employment to the percentage of the male as a percentage of total seats. For countries with a (2018). labour force population ages 15­24 that is not in bicameral legislative system, the share of seats is d Refers to the population ages 10 and older. paid employment or self-employed but is available calculated based on both houses. Column 8: ILO (2019). for work and is actively seeking paid employment or e Refers to the population ages 20­74. self-employment. Time spent on unpaid domestic chores and Column 9: IPU (2019). care work: The average daily number of hours f Refers to the population ages 12 and older. Population with at least some secondary spent on unpaid domestic and care work, expressed Column 10: United Nations Statistics Division education, female to male ratio: Ratio of the as a percentage of a 24-hour day. Unpaid domestic (2019a). g Refers to the population ages 6 and older. and care work refers to activities related to the Columns 11 and 12: HDRO calculations based on h Excludes the 36 special rotating delegates United Nations Statistics Division (2019a).

2

332 | HUMAN DEVELOPMENT REPORT 2019 DASHBOARD 3 Women’s empowerment HUMAN DEVELOPMENT REPORT 2019

Top third Middle third Bottom third

Three-colour coding is used to visualize partial grouping of countries by indicator. For each indicator countries are divided into three groups of approximately equal size (terciles): the top third, the middle third and the bottom third. Aggregates are colour coded using the same tercile cutoffs. See Notes after the table.

SDG 3.1 SDG 3.7, 5.6 SDG 5.6 SDG 5.3 SDG 5.3 SDG 5.2 SDG 5.2 SDG 5.5 SDG 1.3 Reproductive health and family planning Violence against girls and women

Child Violence against women Share of marriage ever experienceda Women graduates married in science, from science, with by age 18 technology, Prevalence engineering technology, account at genital and engineering financial mutilation/ mathematics Proportion cutting programmes and Female institution

Antenatal of births among at tertiary mathematics share of or with girls and level, female care attended Unmet women programmes employment mobile Mandatory

coverage, by skilled Contraceptive need for in tertiary in senior money- paid

at least health prevalence, family Intimate Nonintimate education who and middle service maternity one visit personnel any method planning are female management provider leave

(% of women (% of

ages 20­24 (% of girls female

(% of married or in-union who are and young (% of population

women of reproductive married or women ages female population ages 15

(%) (%) age, 15­49 years) in union) 15­49) ages 15 and older) (%) (%) (%) and older) (days)

HDI rank 2007­2017b 2013­2018b 2008­2018b 2008­2018b 2003­2018b 2004­2018b 2005­2019b 2005­2019b 2008­2018b 2008­2018b 2010­2018b 2017 2017

VERY HIGH HUMAN DEVELOPMENT .. 99.2 … .. 27.0 .. 9.9 28.4 33.5 100.0 .. 2 Switzerland .. .. 72.9 … .. .. 11.1 22.1 31.6 98.9 98 4 Germany .. 99.7 73.3 … 15.0 5.0 14.1 29.0 33.5 95.3 182 6 Australia .. 98.7 80.3 … 22.0 7.0 19.3 27.1 28.6 99.2 98 8 Sweden .. .. 74.8 … … .. .. 94.7 70 10 Netherlands 98.3 97.0 66.9 … 22.8 10.0 9.7 31.7 .. 99.2 .. 12 Finland .. 97.9 … .. 22.4 .. 10.3 35.2 43.1 .. 90 14 New Zealand … … 28.0 12.0 15.0 35.2 39.4 100.0 .. 15 United States .. 99.6 … .. 6.1 .. 22.3 33.7 .. 96.3 105 18 Liechtenstein .. .. 73.0 … 25.0 12.0 6.3 25.3 24.8 99.8 112 20 Austria .. 94.7 … .. 32.0 11.0 12.7 34.2 27.0 100.0 126 22 Israel .. 99.9 85.5 … 30.0 11.0 13.5 27.1 32.0 99.6 147 24 Slovenia 100.0 97.9 … … 11.6 31.4 .. 99.9 105 26 Czechia .. 96.3 … … 12.9 35.0 .. 99.3 112 c 28 Malta .. .. 84.0 … 29.0 7.0 17.5 38.1 34.2 96.1 42 30 Estonia .. 99.1 75.9 9.0 … .. 10.4 34.0 40.5 92.7 .. 32 Greece .. .. 66.8 … 24.0 8.0 7.9 27.5 33.5 98.8 105 34 Lithuania … … .. .. 33.8 40.7 … 36 Andorra .. 99.9 39.8 … … .. .. 98.1 98 36 Slovakia .. 98.4 65.7 … 13.0 4.0 14.3 25.9 28.9 98.4 112 40 Portugal … … 22.0 8.0 9.5 27.6 16.1 98.2 112 42 Chile … … … .. .. 93.7 105 43 Hungary .. 100.0 79.6 … .. .. 15.4 26.4 .. 94.7 90 46 Croatia … … 13.0 4.0 12.5 29.8 38.2 96.9 105

.. .. 70.9 … 13.0 3.0 12.7 29.7 31.9 91.6 112

.. 99.8 86.3 4.3 .. .. 21.0 4.0 13.5 35.4 26.6 78.6 196

.. 98.0 78.4 … 26.0 9.0 14.5 31.8 34.5 91.3 112

.. 99.7 … .. 15.0 5.0 8.6 28.1 27.6 97.0 126

.. 99.9 65.1 … 19.0 5.0 15.7 39.5 23.2 91.6 150

.. 99.2 … .. 20.0 9.0 16.4 38.3 33.2 98.4 140

99.2 96.0 … .. 15.0 2.0 10.4 42.2 22.4 90.0 126

.. 99.9 … .. 19.0 1.0 18.9 39.8 30.5 84.5 119

.. 99.8 62.3 … 13.0 2.0 15.3 44.1 39.5 88.0 140

.. 100.0 … .. 24.0 5.0 11.4 29.8 38.2 81.0 126

100.0 99.9 … … 17.3 43.5 12.2 76.4 45

.. 100.0 … … 4.7 … ..

97.0 99.7 24.6 … .. .. 17.2 41.7 .. 58.2 70 DASH

.. 98.5 … .. 23.0 4.0 12.0 35.6 30.4 83.1 238 BOARD

.. 99.9 … .. 32.0 7.0 10.0 31.9 43.2 92.5 112 3

.. 98.8 73.9 … 19.0 1.0 19.3 39.1 32.2 90.6 ..

90.8 100.0 37.5 12.4 4 … 14.5 41.9 .. 61.6 d 50

.. 99.7 76.3 … .. .. 6.8 18.8 .. 71.3 126

99.0 99.8 … … 23.6 51.9 37.0 .. 91

.. 99.7 61.6 … 21.0 3.0 11.7 31.5 37.1 72.2 168

100.0 99.7 … … 10.9 44.3 .. 75.4 60

.. 99.9 … .. 13.0 3.0 16.0 37.9 26.1 82.7 208

DASHBOARD 3 Women’s empowerment | 333 DASHBOARD 3 WOMEN’S EMPOWERMENT

SDG 3.1 SDG 3.7, 5.6 SDG 5.6 SDG 5.3 SDG 5.3 SDG 5.2 SDG 5.2 SDG 5.5 SDG 1.3 Reproductive health and family planning Violence against girls and women

married in science, from science, with Prevalence engineering technology, account at genital and engineering financial Proportion cutting programmes and Female institution

Antenatal of births among at tertiary mathematics share of or with care attended Unmet women programmes employment mobile Mandatory

coverage, by skilled Contraceptive need for in tertiary in senior money- paid

at least health prevalence, family Intimate Nonintimate education who and middle service maternity one visit personnel any method planning are female management provider leave

(% of women (% of

ages 20­24 (% of girls female

(% of married or in-union who are and young (% of population

women of reproductive married or women ages female population ages 15

(%) (%) age, 15­49 years) in union) 15­49) ages 15 and older) (%) (%) (%) and older) (days)

HDI rank 2007­2017b 2013­2018b 2008­2018b 2008­2018b 2003­2018b 2004­2018b 2005­2019b 2005­2019b 2008­2018b 2008­2018b 2010­2018b 2017 2017

47 Oman 98.6 99.1 29.7 17.8 4 … 39.8 52.8 .. 63.5 d 50 48 Argentina 49 Russian Federation 98.1 93.9 81.3 … 26.9 12.1 11.5 46.5 32.6 50.8 90 50 Belarus 50 Kazakhstan .. 99.7 68.0 8.0 … … 39.3 76.1 140 52 Bulgaria 52 Montenegro 99.7 99.8 72.1 7.0 3 … 15.4 26.7 .. 81.3 126 52 Romania 55 Palau 99.3 99.4 54.8 10.6 7 .. 16.5 1.5 14.8 32.9 .. 60.3 126 56 Barbados 57 Kuwait .. 99.8 … .. 23.0 6.0 12.3 38.3 39.3 73.6 410 57 Uruguay 59 Turkey 91.7 99.0 23.3 21.8 5 .. 17.0 1.0 .. .. 23.8 67.6 45 60 Bahamas 61 Malaysia 76.3 95.2 … .. 24.0 2.0 20.3 41.2 30.1 53.6 126 62 Seychelles HIGH HUMAN DEVELOPMENT 90.3 100.0 … .. 25.2 15.1 .. .. 35.5 .. .. 63 Trinidad and Tobago 93.4 99.0 59.2 19.9 11 … .. 40.5 .. .. 84 66 Mauritius 100.0 99.9 … … … 73.5 70 68 Costa Rica 97.2 99.7 79.6 .. 25 .. 16.8 .. 10.8 44.6 37.3 60.6 98 70 Georgia 97.0 98.0 73.5 5.9 15 .. 38.0 .. 14.2 34.7 16.3 54.3 112 72 Cuba 98.0 99.0 … … … .. 91 74 Antigua and Barbuda 97.2 99.5 52.2 … .. .. 18.1 38.6 .. 82.5 60 76 Mexico … … .. .. 8.5 38.9 43.8 .. 98 78 Grenada 98.3 98.4 58.4 14.9 3 .. 17.0 2.0 18.1 39.7 29.8 70.1 135 79 Brazil 95.1 100.0 79 Colombia 96.9 99.0 40.3 24.3 11 .. 30.2 19.0 … 73.6 98 81 Armenia 99.8 82 Algeria .. 94.2 77.4 5.7 17 … 32.1 30.1 .. 91.6 270 82 North Macedonia 93.4 98.7 82 Peru 98.1 99.8 63.8 12.5 … … 30.8 87.1 98 85 China 97.3 99.9 85 Ecuador 97.6 62.8 16.4 26 .. 14.4 .. 12.7 49.0 43.5 42.3 98 87 Azerbaijan 95.5 .. 88 Ukraine 98.5 99.9 77.8 7.6 21 .. 35.9 e .. 7.7 33.4 .. 60.9 120 89 Dominican Republic 100.0 100.0 89 Saint Lucia 100.0 100.0 46.0 15.1 12 .. 21.0 1.3 14.8 49.4 29.3 38.1 365 91 Tunisia 87.0 99.9 92 Mongolia 98.5 97.7 53.4 12.3 14 .. 6.0 2.7 15.8 43.7 .. 63.6 183 93 Lebanon 98.1 99.1 94 Botswana 100.0 98.9 61.7 7.5 10 … .. 40.3 25.6 73.4 84 94 Saint Vincent and the Grenadines 97.2 99.1 96 Jamaica 97.2 99.2 73.7 8.0 26 … 6.1 39.9 … 96 Venezuela (Bolivarian Republic of) 99.6 99.8 98 Dominica 92.7 96.6 … … … .. 91 98.6 99.9 97.0 93.1 … … 1.8 33.3 .. .. 91 96.5 99.9 96.4 45.8 9.0 4 .. 11.0 1.0 14.8 42.9 24.2 54.7 365 .. 99.8 91.7 99.9 66.9 13.0 26 .. 24.6 38.8 14.8 31.1 35.6 33.3 84 98.6 99.8 98.0 99.0 78.4 6.2 23 … 15.0 30.1 29.5 79.8 90 96.9 98.1 … … .. 8.2 35.4 .. .. 90 98.7 98.9 80.2 .. 26 .. 16.7 .. 10.7 36.6 .. 67.5 120 94.1 99.7 81.0 6.7 23 .. 33.3 .. 14.4 34.1 .. 42.5 126 99.5 98.6 97.7 97.6 57.1 12.5 5 .. 8.2 .. 8.4 32.8 .. 40.9 140 97.5 95.4 100.0 97.0 57.1 7.0 3 … 26.9 55.5 .. 29.3 98

40.2 17.2 7 .. 10.0 2.0 15.7 45.1 28.2 72.9 270

75.4 6.5 19 .. 31.2 .. 13.7 32.9 .. 34.4 98

84.5 … … .. .. 76.4 128

80.1 8.8 20 .. 40.4 .. 8.0 29.2 35.3 42.6 84

54.9 .. 11 .. 13.5 .. 16.4 40.1 .. 27.7 126

65.4 4.9 9 .. 26.0 5.0 12.5 27.4 .. 61.3 126

69.5 11.4 36 .. 28.5 .. 7.0 40.0 .. 54.1 98

DASH 55.5 17.0 8 … … .. 91 62.5 7.0 2 … 37.8 58.1 19.3 28.4 30 3 54.6 16.0 5 .. 31.2 14.0 11.9 33.7 40.0 95.0 120

54.5 .. 6 … 18.0 43.3 .. 32.9 70

52.8 9.6 … … .. 46.8 84

… … … .. 91

72.5 10.0 8 .. 27.8 23.0 … 77.8 f 56

75.0 … … .. .. 70.0 182

… … … .. 84

334 | HUMAN DEVELOPMENT REPORT 2019

SDG 3.1 SDG 3.7, 5.6 SDG 5.6 SDG 5.3 SDG 5.3 SDG 5.2 SDG 5.2 SDG 5.5 SDG 1.3 Reproductive health and family planning Violence against girls and women

married in science, from science, with Prevalence engineering technology, account at genital and engineering financial Proportion cutting programmes and Female institution

Antenatal of births among at tertiary mathematics share of or with care attended Unmet women programmes employment mobile Mandatory

coverage, by skilled Contraceptive need for in tertiary in senior money- paid

at least health prevalence, family Intimate Nonintimate education who and middle service maternity one visit personnel any method planning are female management provider leave

(% of women (% of

ages 20­24 (% of girls female

(% of married or in-union who are and young (% of population

women of reproductive married or women ages female population ages 15

(%) (%) age, 15­49 years) in union) 15­49) ages 15 and older) (%) (%) (%) and older) (days)

HDI rank 2007­2017b 2013­2018b 2008­2018b 2008­2018b 2003­2018b 2004­2018b 2005­2019b 2005­2019b 2008­2018b 2008­2018b 2010­2018b 2017 2017

98 Fiji 100.0 99.8 … .. 64.1 8.5 .. .. 38.6 .. 84 98 Suriname 98.7 97.3 68.4 12.1 22 .. 20.4 … .. 46.0 98 103 Belize 90.9 80.0 47.6 16.9 19 … … .. .. 105 Tonga 99.1 99.7 51.8 14.2 8 .. 19.0 … .. 26.6 70 106 Philippines 97.2 92.2 107 Moldova (Republic of) 51.4 22.2 34 .. 22.2 .. 11.7 41.8 41.7 52.3 f 98 108 Uzbekistan 99.1 95.6 34.7 28.6 4 .. 16.3 … 19.5 .. 60 111 Indonesia 99.0 .. 34.1 25.2 6 .. 39.6 6.3 … .. .. 113 South Africa 95.4 84.4 54.1 16.7 17 .. 14.8 .. 17.8 36.3 25.5 38.9 60 115 Gabon 98.8 99.7 59.5 9.5 12 .. 34.0 4.0 12.1 32.2 .. 44.6 126 MEDIUM HUMAN DEVELOPMENT 99.6 100.0 50.2 12.1 6 … … 35.5 .. 118 Viet Nam 99.4 100.0 .. .. 7 … … 36.0 126 120 Iraq 93.0 99.9 27.7 40.2 … … .. 59.6 98 122 Kyrgyzstan 95.4 93.6 61.0 14.8 11 .. 18.3 .. 12.2 37.1 19.4 51.4 90 124 El Salvador 93.3 82.5 26.9 34.8 11 .. 46.1 10.6 .. .. 41.6 .. 28 126 Cabo Verde 93.7 96.7 54.6 14.9 6 .. 21.3 .. 12.7 41.9 33.9 70.0 120 126 Nicaragua 90.1 71.3 66.5 23.2 20 .. 58.5 … 26.8 53.9 90 130 Namibia 94.7 .. 31.1 26.5 22 .. 48.6 5.0 … 53.7 98 132 Honduras 90.3 91.5 58.5 12.6 17 87.2 25.6 .. 7.7 36.9 .. 27.0 90 134 Bhutan 81.2 92.4 .. .. 26 .. 50.9 13.0 … .. .. 135 Micronesia (Federated States of) 95.8 93.8 75.7 6.1 11 .. 34.4 2.3 15.4 36.5 .. 30.4 180 138 Congo 99.4 99.6 57.2 10.9 15 … 11.7 44.9 17.8 15.9 84 140 Lao People’s Democratic Republic 77.7 95.6 52.8 13.3 28 7.4 … .. .. 19.5 98 142 Ghana 77.1 86.6 70.8 13.8 13 … 17.5 45.2 .. 16.8 98 144 Equatorial Guinea 98.4 98.4 42.0 19.1 12 .. 26.6 0.1 13.3 38.7 .. 38.9 126 146 Cambodia 90.7 85.7 33.9 28.0 30 … 5.2 27.2 35.4 .. 91 147 Nepal 96.0 99.9 72.0 11.1 26 .. 14.3 .. 9.4 23.5 32.7 24.4 112 150 Cameroon 78.8 94.8 29.3 16.5 9 .. 26.4 … .. 42.1 140

.. 92.6 .. .. 18 .. 12.6 .. 8.0 30.6 .. .. 60

91.3 69.2 60.6 13.9 30 .. 21.2 .. 5.4 34.7 34.5 42.1 84

94.7 89.6 80.4 5.8 35 .. 22.5 … .. 24.8 84

.. 81.4 53.5 12.9 27 .. 28.8 .. 27.7 43.9 13.0 76.6 182

96.6 88.2 56.1 17.5 7 .. 26.7 .. 8.1 41.9 48.2 80.7 84

84.4 56.7 26.1 25.3 15 .. 58.8 13.9 … .. 84

96.6 74.0 73.2 10.7 34 .. 27.8 .. 8.6 37.5 41.0 41.0 84

88.4 .. 22.3 28.0 20 .. 67.6 9.8 … .. 84

97.9 96.4 65.6 11.7 26 .. 15.1 5.8 … 27.7 f 56

63.9 67.8 62.3 12.0 59 .. 54.2 3.0 7.9 19.8 11.5 35.8 112

80.0 … .. .. 32.8 8.0 .. .. 18.2 .. ..

97.5 92.5 40.6 33.7 35 .. 27.9 … .. .. 98

93.5 91.2 30.1 17.9 27 … … 21.0 105

98.5 88.3 66.1 15.2 5 … .. .. 54.6 27.4 d 14

54.2 64.4 54.1 14.3 33 .. 15.3 5.3 8.6 25.2 23.4 31.9 105

75.6 89.4 49.0 24.2 21 .. 60.0 33.0 … .. 84

90.5 78.1 33.0 26.3 21 3.8 24.4 4.0 7.4 22.5 .. 53.7 84 DASH

95.7 63.3 49.0 21.1 31 .. 45.9 … 28.5 40.3 84 3 .. 84 91.3 .. 12.6 33.8 30 .. 56.9 … ..

80.7 60.2 52.2 16.2 16 .. 17.3 .. 47.3 64.9 31.5 26.0 98

95.3 89.0 56.3 12.5 19 .. 20.9 3.8 6.0 16.7 .. 21.5 90

93.7 61.8 60.5 14.9 23 21.0 40.7 .. 11.2 30.7 .. 77.7 90

83.6 58.0 52.6 23.7 40 .. 25.0 … 13.9 41.6 52

81.6 46.6 13.7 38.0 30 .. 34.8 .. 9.9 38.4 .. 22.3 f 90

82.8 64.7 34.4 18.0 31 1.4 51.1 5.0 … 30.0 98

DASHBOARD 3 Women’s empowerment | 335 DASHBOARD 3 WOMEN’S EMPOWERMENT

SDG 3.1 SDG 3.7, 5.6 SDG 5.6 SDG 5.3 SDG 5.3 SDG 5.2 SDG 5.2 SDG 5.5 SDG 1.3 Reproductive health and family planning Violence against girls and women

married in science, from science, with Prevalence engineering technology, account at genital and engineering financial Proportion cutting programmes and Female institution

Antenatal of births among at tertiary mathematics share of or with care attended Unmet women programmes employment mobile Mandatory

coverage, by skilled Contraceptive need for in tertiary in senior money- paid

at least health prevalence, family Intimate Nonintimate education who and middle service maternity one visit personnel any method planning are female management provider leave

(% of women (% of

ages 20­24 (% of girls female

(% of married or in-union who are and young (% of population

women of reproductive married or women ages female population ages 15

(%) (%) age, 15­49 years) in union) 15­49) ages 15 and older) (%) (%) (%) and older) (days)

HDI rank 2007­2017b 2013­2018b 2008­2018b 2008­2018b 2003­2018b 2004­2018b 2005­2019b 2005­2019b 2008­2018b 2008­2018b 2010­2018b 2017 2017

150 Zimbabwe 93.3 78.1 66.8 10.4 32 .. 37.6 .. 20.9 28.8 .. 51.7 98 153 Solomon Islands 73.1 69.3 34.2 17.3 21 .. 24.5 … 4.2 7.0 84 154 Syrian Arab Republic 88.5 86.2 29.3 34.7 21 .. 63.5 18.0 .. .. 25.1 .. 84 156 Comoros 87.7 .. 53.9 16.4 13 … 19.2 49.5 .. 19.6 d 120 158 Nigeria … .. 21 … .. .. 19.3 .. 0 159 Uganda 92.1 .. 19.4 31.6 32 .. 6.4 1.5 … 17.9 d 98 162 Madagascar 99.0 90.7 53.2 18.9 7 .. 37.1 .. 9.2 32.2 36.3 45.0 84 164 Lesotho 65.8 43.0 27.6 23.1 44 18.4 17.4 1.5 .. .. 28.9 27.3 84 165 Côte d’Ivoire 166 Senegal 91.4 63.5 38.4 22.1 31 10.0 46.2 … 17.3 42.2 84 167 Togo 168 Sudan 97.3 74.2 41.8 26.0 34 0.3 49.9 … .. 52.7 84 169 Haiti 170 Afghanistan 86.9 69.3 17.8 33.6 37 66.6 .. .. 29.4 28.9 .. 15.5 98 171 Djibouti 172 Malawi 82.1 44.3 47.9 16.4 41 … 13.6 28.1 24.5 16.3 98 173 Ethiopia 174 Gambia 82.8 78.1 15.5 32.3 26 9.2 23.8 .. 19.1 54.9 .. 28.6 98 174 Guinea 176 Liberia 95.2 77.9 60.2 18.4 17 … 4.5 23.4 .. 46.5 84 177 Yemen 178 Guinea-Bissau 93.2 73.7 23.3 26.5 27 36.7 25.9 … .. 35.6 98 179 Congo (Democratic Republic of the) 180 Mozambique 95.0 68.4 27.8 21.9 29 24.0 21.5 … .. 38.4 98 181 Sierra Leone 182 Burkina Faso 72.7 44.6 19.9 33.6 22 4.7 25.1 … .. 37.6 98 184 Mali 79.1 77.7 12.2 26.6 34 86.6 .. .. 27.8 47.2 .. 10.0 f 56 185 Burundi 186 South Sudan 91.0 41.6 34.3 38.0 15 .. 26.0 … .. 30.0 42 188 Central African Republic 58.6 58.8 22.5 24.5 35 .. 50.8 … 4.3 7.2 90 189 Niger OTHER COUNTRIES OR TERRITORIES 87.7 .. 19.0 .. 5 93.1 … .. .. 8.8 d 98

.. Korea (Democratic People’s Rep. of) 94.8 89.8 59.2 18.7 42 .. 37.5 … .. 29.8 56 .. Monaco .. Nauru 62.4 27.7 40.1 20.6 40 65.2 28.0 .. 7.6 17.3 21.1 29.1 90 .. San Marino .. Somalia 86.2 57.2 9.0 24.9 30 74.9 20.1 .. 53.1 45.7 33.7 .. 180 .. Tuvalu 84.3 55.3 8.7 27.6 51 96.8 … .. .. 19.7 98

95.9 61.1 31.2 31.1 36 44.4 38.5 2.6 .. .. 20.1 28.2 98

64.4 44.7 33.5 28.7 32 18.5 … .. .. 1.7 f 70

92.4 45.0 16.0 22.3 24 44.9 … … 60

88.4 80.1 20.4 27.7 37 .. 50.7 .. 11.0 25.1 .. 24.2 98

90.6 73.0 27.1 23.1 53 .. 21.7 .. 5.1 26.7 22.2 32.9 60

97.1 81.6 22.5 26.3 30 86.1 48.8 … .. 15.4 84

92.8 79.8 31.7 22.8 52 75.8 11.5 .. 7.0 15.1 .. 34.5 98

88.5 .. 8.4 27.4 41 83.0 .. .. 21.8 27.8 .. .. 60

75.6 67.3 15.6 17.2 50 82.7 35.5 … .. 25.7 98

99.2 85.1 28.5 29.7 19 .. 48.5 .. 10.4 18.2 .. 6.7 f 84

61.9 .. 4.0 26.3 52 … … 4.7 56

54.7 20.2 5.7 22.9 67 38.4 28.6 … .. 14.9 98

68.2 .. 15.2 27.0 68 24.2 29.8 … .. 9.7 98

82.8 39.7 11.0 15.0 76 2.0 .. .. 6.4 29.1 .. 10.9 98

100.0 99.5 78.2 7.0 … .. 22.2 19.3 …

94.5 … 27 .. 48.1 47.3 … .. ..

DASH … … … … 630 … .. 45 97.9 … .. .. 33.7 f .. 3 97.4 … 10 .. 36.8 … 36.7 .. ..

336 | HUMAN DEVELOPMENT REPORT 2019

SDG 3.1 SDG 3.7, 5.6 SDG 5.6 SDG 5.3 SDG 5.3 SDG 5.2 SDG 5.2 SDG 5.5 SDG 1.3 Reproductive health and family planning Violence against girls and women

Child Violence against women Share of marriage ever experienceda Share of graduates Women Women graduates married in science, from science, with by age 18 technology, Prevalence engineering technology, account at genital and engineering financial Proportion cutting programmes and Female institution

Antenatal of births among at tertiary mathematics share of or with care attended Unmet women programmes employment mobile Mandatory

coverage, by skilled Contraceptive need for in tertiary in senior money- paid

at least health prevalence, family Intimate Nonintimate education who and middle service maternity one visit personnel any method planning are female management provider leave

(% of women (% of

ages 20­24 (% of girls female

(% of married or in-union who are and young (% of population

women of reproductive married or women ages female population ages 15

(%) (%) age, 15­49 years) in union) 15­49) ages 15 and older) (%) (%) (%) and older) (days)

HDI rank 2007­2017b 2013­2018b 2008­2018b 2008­2018b 2003­2018b 2004­2018b 2005­2019b 2005­2019b 2008­2018b 2008­2018b 2010­2018b 2017 2017

Human development groups .. 98.9 69.1 … .. .. 13.2 33.5 — 86.8 112 Very high human development 96.3 97.7 High human development 75.4 … … .. — 65.4 116 Medium human development .. 78.1 Low human development 77.8 56.5 53.0 13.9 28 .. 30.7 .. 26.0 43.7 — 58.2 94 90.1 85.2 Developing countries 29.4 23.7 39 36.7 31.5 … — 26.1 86 60.5 15.0 27 … .. .. — 58.2 99 East Asia and the Pacific 86.5 88.5 47.9 15.8 20 … 19.0 48.1 — 27.0 75 Latin America and the Caribbean 95.8 96.6 77.2 … … .. -- .. 88 Sub-Saharan Africa 97.1 98.9 63.3 8.2 10 .. 27.8 .. 14.0 32.9 — 53.4 165 Small island developing states 97.1 95.1 74.5 .. 25 .. 23.8 .. 11.6 33.6 — 52.1 96 Cooperation and Development .. 78.8 52.9 13.3 29 .. 31.0 … — 65.0 110 81.8 60.6 34.0 22.3 36 30.3 31.4 … — 36.0 89

77.9 61.5 38.2 21.4 40 .. 38.3 … — 28.4 87

95.2 83.6 54.1 20.1 23 … .. .. — .. 79

.. 98.8 70.7 … .. .. 12.9 32.6 — 86.2 114

.. 87.0 61.9 … … .. — 64.6 107

NOTES DEFINITIONS Prevalence of female genital mutilation/cutting Women with account at financial institution or among girls and women: Percentage of girls and with mobile money-service provider: Percentage Three-colour coding is used to visualize partial Antenatal care coverage, at least one visit: women ages 15­49 who have undergone female of women ages 15 and older who report having an grouping of countries and aggregates by indicator. Percentage of women ages 15­49 attended at least genital mutilation/cutting. account alone or jointly with someone else at a bank For each indicator countries are divided into three once during pregnancy by skilled health personnel or other type of financial institution or who report groups of approximately equal size (terciles): the (doctor, nurse or midwife). Violence against women ever experienced, personally using a mobile money service in the past top third, the middle third and the bottom third. intimate partner: Percentage of the female 12 months. Aggregates are colour coded using the same tercile Proportion of births attended by skilled health population ages 15 and older that has ever Mandatory paid maternity leave: Number of days cutoffs. See Technical note 6 at http://hdr.undp.org/ personnel: Percentage of deliveries attended by experienced physical and/or sexual violence from an of paid time off work to which a female employee is sites/default/files/hdr2019_technical_notes.pdf for skilled health personnel (generally doctors, nurses intimate partner. entitled in order to take care of a newborn child. details about partial grouping in this table. or midwives) trained in providing lifesaving obstetric care—including giving the necessary supervision, Violence against women ever experienced, MAIN DATA SOURCES a Data collection methods, age ranges, sampled care and advice to women during pregnancy, labour nonintimate partner: Percentage of the female Column 1: UNICEF (2019b). women (ever-partnered, ever-married or all and the postpartum period, conducting deliveries population ages 15 and older that has ever Columns 2, 5 and 6: United Nations Statistics women) and definitions of the forms of violence on their own and caring for newborns. Traditional experienced sexual violence from a nonintimate Division (2019a). and of perpetrators vary by survey. Thus data are birth attendants, even if they receive a short training partner. Columns 3 and 4: UNDESA (2019a). not necessarily comparable across countries. course, are not included. Columns 7 and 8: UN Women (2019). Share of graduates in science, technology, Columns 9 and 10: UNESCO Institute for Statistics b Data refer to the most recent year available Contraceptive prevalence, any method: engineering and mathematics programmes at (2019). during the period specified. Percentage of married or in-union women of tertiary level, female: Share of female tertiary Column 11: ILO (2019). reproductive age (15­49 years) currently using any graduates in science, technology, engineering and Columns 12 and 13: World Bank (2019b). c Refers to 2015. contraceptive method. mathematics programmes among all female tertiary graduates. DASH d Refers to 2011. Unmet need for family planning: Percentage BOARD of married or in-union women of reproductive age Share of graduates from science, technology, e Refers to 2003. (15­49 years) who are fecund have an unmet need engineering and mathematics programmes 3 if they want to have no (more) births, or if they want in tertiary education who are female: Share of f Refers to 2014. to postpone or are undecided about the timing of female graduates among all graduates of tertiary their next birth, yet they are not using any method programmes in science, technology, engineering and of contraception. mathematics.

Child marriage, women married by age 18: Female share of employment in senior and Percentage of women ages 20­24 who were first middle management: Proportion of women in total married or in union before age 18. employment in senior and middle management.

DASHBOARD 3 Women’s empowerment | 337 DASHBOARD 4 Environmental sustainability

Top third Middle third Bottom third

Three-colour coding is used to visualize partial grouping of countries by indicator. For each indicator countries are divided into three groups of approximately equal size (terciles): the top third, the middle third and the bottom third. Aggregates are colour coded using the same tercile cutoffs. See Notes after the table.

SDG 12.c SDG 7.2 SDG 9.4 SDG 15.1 SDG 6.4 SDG 12.2 SDG 3.9 SDG 3.9 SDG 15.3 SDG 15.5

Fossil fuel Renewable Natural Household Unsafe water, 2018 resource and ambient sanitation energy energy Fresh water depletion air pollution and hygiene Degraded 0.940 withdrawals services land 0.974 consumption consumption Carbon dioxide emissions Forest area 0.925 0.983 (% of total (% of total (% of total 0.821 renewable water 0.825 energy final energy Per capita (kg per 2010 (% of total Change (% of total 0.861 (tonnes) US$ of GDP) land areaa) (%) resources) (% of GNI) land area) 0.993 consumption) consumption) (per 100,000 population) 0.860 2016 2016 0.943 HDI rank 2010­2015b 2015 2016 1990/2016 2007­2017b 2012­2017b 2016 2016 2015 0.972 0.990 VERY HIGH HUMAN DEVELOPMENT 57.0 57.8 6.8 0.11 33.2 ­0.1 0.8 4.4 9 0.2 .. 0.969 1 Norway 0.626 2 Switzerland 50.2 25.3 4.5 0.08 31.8 9.3 3.8 0.0 10 0.1 .. 0.783 3 Ireland 0.836 4 Germany 85.3 9.1 7.9 0.12 11.0 63.4 1.5 0.1 12 0.1 .. 0.986 4 Hong Kong, China (SAR) 0.993 6 Australia 78.9 14.2 8.9 0.21 32.7 1.0 16.5 0.0 16 0.6 .. 0.781 6 Iceland 0.894 8 Sweden 93.2 0.9 6.2 0.11 … … .. 0.987 9 Singapore 0.758 10 Netherlands 89.6 9.2 16.2 0.35 16.3 ­2.8 3.2 3.0 8 0.1 .. 0.733 11 Denmark 0.937 12 Finland 11.3 77.0 6.2 0.14 0.5 213.7 0.2 0.0 9 0.1 .. 0.843 13 Canada 0.971 14 New Zealand 25.1 53.2 3.9 0.08 68.9 0.8 1.6 0.2 7 0.2 .. 0.873 15 United Kingdom 0.883 15 United States 90.6 0.7 8.0 0.10 23.1 ­5.5 .. 0.0 26 0.1 .. 0.902 17 Belgium 0.986 18 Liechtenstein 93.5 5.9 9.2 0.20 11.2 9.4 9.8 0.3 14 0.2 .. 0.983 19 Japan 0.848 20 Austria 64.9 33.2 5.9 0.13 14.7 14.7 10.6 0.4 13 0.3 .. 0.971 21 Luxembourg 0.989 22 Israel 40.2 43.2 8.3 0.21 73.1 1.8 .. 0.1 7 0.1 c 1 0.863 22 Korea (Republic of) 0.917 24 Slovenia 74.1 22.0 14.9 0.35 38.2 ­0.4 1.2 0.7 7 0.4 .. 0.908 25 Spain 0.963 26 Czechia 59.7 30.8 6.5 0.19 38.6 5.1 1.6 0.5 7 0.1 .. 0.988 26 France 0.854 28 Malta 80.4 8.7 5.6 0.15 13.1 13.8 5.7 0.4 14 0.2 .. 0.826 29 Italy 0.755 30 Estonia 82.4 8.7 15.0 0.29 33.9 2.7 14.5 0.2 13 0.2 .. 0.825 31 Cyprus 0.930 32 Greece 75.9 9.2 8.1 0.20 22.6 .. 32.8 0.0 16 0.3 11 0.844 32 Poland 0.901 34 Lithuania .. 63.1 .. .. 43.1 6.2 … .. .. 0.885 35 United Arab Emirates 0.861 36 Andorra 93.0 6.3 9.0 0.24 68.5 0.0 18.9 0.0 12 0.2 .. 0.955 36 Saudi Arabia 0.972 36 Slovakia 65.7 34.4 7.2 0.17 46.9 2.6 4.5 0.1 15 0.1 .. 40 Portugal 80.6 9.0 14.6 0.16 35.7 .. 1.3 0.0 12 0.1 c 4 42 Chile 97.4 3.7 7.9 0.23 7.7 26.7 .. 0.1 15 0.2 .. 43 Hungary 81.0 2.7 11.6 0.33 63.4 ­4.1 .. 0.0 20 1.8 .. 46 Croatia 61.1 20.9 6.5 0.23 62.0 5.1 2.8 0.0 23 0.1 c 5 48 Argentina 73.0 16.3 5.1 0.16 36.9 33.6 28.7 0.0 10 0.2 18 50 Belarus 77.7 14.8 9.5 0.31 34.6 1.6 12.4 0.1 30 0.2 6

46.5 13.5 4.5 0.12 31.2 18.5 13.9 0.0 10 0.3 12

97.8 5.4 3.1 0.09 1.1 0.0 83.0 .. 20 0.1 c ..

79.9 16.5 5.4 0.16 31.8 23.2 17.9 0.0 15 0.1 13

13.1 27.5 12.4 0.47 51.3 ­1.4 13.4 0.2 25 0.1 c ..

92.9 9.9 5.4 0.24 18.7 7.2 28.0 0.0 20 0.3 19

82.6 17.2 5.9 0.25 31.7 23.8 14.0 0.1 28 0.1 c 16

90.3 11.9 7.7 0.31 30.9 6.5 17.5 0.4 38 0.1 5

68.0 29.0 3.7 0.14 34.8 12.3 11.3 .. 34 0.1 3

86.1 0.1 20.5 0.31 4.6 32.1 .. 4.0 55 0.1 c 1

.. 19.7 .. .. 34.0 0.0 … .. ..

99.9 0.0 16.3 0.33 0.5 0.0 871.7 7.9 84 0.1 4

64.1 13.4 5.6 0.19 40.4 1.0 1.1 0.0 34 0.1 c 4

56.7 38.1 3.4 0.15 54.0 5.8 0.6 0.0 41 0.1 c 13

77.0 27.2 4.6 0.17 34.6 ­7.8 11.8 0.1 10 0.2 32

100.0 0.0 29.8 0.27 0.0 0.0 .. 7.4 47 0.1 c 6

74.6 24.9 4.7 0.22 24.3 18.2 .. 6.5 25 0.2 1

100.0 0.0 15.1 0.21 72.1 ­8.0 .. 10.9 13 0.1 c ..

69.5 15.6 4.5 0.18 22.9 14.3 4.9 0.2 39 0.2 13

DASH 99.4 0.0 20.8 0.49 0.8 145.9 132.2 3.2 40 0.1 c .. 70.7 33.1 3.8 0.19 34.4 3.8 0.6 0.3 35 0.1 .. 4 100.0 0.0 14.1 0.37 0.0 0.0 .. 18.1 54 0.1 c 7

87.7 10.0 4.4 0.24 9.8 ­22.9 4.3 1.0 27 0.4 39

92.1 3.3 9.9 0.45 49.8 0.8 1.5 5.8 49 0.1 6

92.4 6.8 5.6 0.34 42.6 11.1 2.5 0.6 61 0.1 1

338 | HUMAN DEVELOPMENT REPORT 2019

SDG 12.c SDG 7.2 SDG 9.4 SDG 15.1 SDG 6.4 SDG 12.2 SDG 3.9 SDG 3.9 SDG 15.3 SDG 15.5

Fossil fuel Renewable Natural Household Unsafe water, 2018 resource and ambient sanitation 0.871 energy energy Fresh water depletion air pollution and hygiene Degraded 0.944 withdrawals services land 0.813 consumption consumption Carbon dioxide emissions Forest area 0.949 0.732 (% of total (% of total (% of total 0.914 renewable water 0.845 energy final energy Per capita (kg per 2010 (% of total Change (% of total 0.832 (tonnes) US$ of GDP) land areaa) (%) resources) (% of GNI) land area) 0.875 consumption) consumption) (per 100,000 population) 0.702 2016 2016 0.677 HDI rank 2010­2015b 2015 12.9 0.56 2016 1990/2016 2007­2017b 2012­2017b 2016 2016 2015 0.664 0.33 50 Kazakhstan 99.2 1.6 5.7 0.22 1.2 ­3.3 19.8 8.7 63 0.4 36 52 Bulgaria 3.4 0.17 52 Montenegro 71.0 17.7 3.4 35.4 17.6 26.4 0.7 62 0.1 .. 52 Romania .. 55 Palau 64.7 43.0 .. .. 61.5 32.1 .. 0.5 79 0.1 c 6 56 Barbados .. 0.33 57 Kuwait 72.5 23.7 22.8 0.09 30.1 8.4 3.0 0.5 59 0.4 2 57 Uruguay 1.8 0.18 59 Turkey .. 0.0 4.2 .. 87.6 … … 60 Bahamas .. 0.28 61 Malaysia .. 2.8 7.0 .. 14.7 0.0 .. 0.0 31 0.2 .. 62 Seychelles .. HIGH HUMAN DEVELOPMENT 93.7 0.0 0.4 81.2 .. 8.1 104 0.1 c 64 63 Trinidad and Tobago 46.3 58.0 10.7 134.1 .. 1.2 18 0.4 26 66 Mauritius 86.8 13.4 15.4 22.8 27.8 0.2 47 0.3 9 68 Costa Rica .. 1.2 51.4 0.0 .. 0.0 20 0.1 .. 70 Georgia 96.6 5.2 67.6 ­0.7 .. 3.1 47 0.4 16 72 Cuba .. 1.4 88.4 0.0 .. 0.0 49 0.2 12 74 Antigua and Barbuda 83.9 21.2 5.1 0.49 31.1 9.9 2.9 0.4 62 0.7 6 0.958 76 Mexico 99.9 0.3 15.3 0.52 46.0 ­1.9 8.8 6.9 39 0.1 .. 0.813 78 Grenada 99.0 0.9 7.1 0.39 6.6 17.8 .. 4.6 51 1.0 23 0.837 79 Brazil 79 Colombia 84.5 11.5 3.2 0.17 19.0 ­6.0 .. 0.0 38 0.6 27 0.396 81 Armenia 82 Algeria 80.7 21.2 2.5 0.12 61.9 ­8.7 0.9 0.1 26 1.9 14 0.733 82 North Macedonia 82 Peru 49.9 38.7 1.5 0.10 54.6 8.7 2.8 0.3 23 0.9 9 0.818 85 China 85 Ecuador 61.4 38.6 1.3 0.12 28.1 ­2.3 .. 1.1 68 0.2 8 0.844 87 Azerbaijan 88 Ukraine 72.2 28.7 2.2 0.26 40.6 2.6 2.9 0.7 102 0.2 6 0.864 89 Dominican Republic 89 Saint Lucia 50.5 52.9 1.0 0.09 32.9 ­9.7 .. 0.1 80 1.2 36 0.564 91 Tunisia 92 Mongolia 85.6 19.3 2.1 0.10 31.3 63.2 18.3 0.5 50 1.0 .. 0.651 93 Lebanon 94 Botswana .. 1.6 .. .. 42.3 0.0 51.3 … .. 0.731 94 Saint Vincent and the Grenadines 96 Jamaica .. 0.0 .. .. 22.3 ­4.9 8.5 .. 30 0.1 .. 0.888 96 Venezuela (Bolivarian Republic of) 98 Dominica 77.5 40.8 6.5 0.58 42.7 ­1.1 0.9 0.4 80 0.1 4 0.905 98 Fiji 98 Paraguay 90.4 9.2 3.6 0.21 33.9 ­5.5 18.6 2.2 37 1.1 47 0.677 98 Suriname 102 Jordan 79.8 22.9 3.5 0.23 32.2 17.3 13.1 1.6 61 3.5 21 0.795 103 Belize 104 Maldives .. 10.9 .. .. 50.0 0.0 7.1 .. 45 0.3 .. 0.763 105 Tonga 106 Philippines 59.1 43.8 2.0 0.15 58.9 ­9.9 0.7 1.9 30 1.0 27 0.902 76.7 23.6 1.8 0.14 52.7 ­9.2 0.5 3.4 37 0.8 7 0.737

74.6 15.8 1.7 0.21 11.7 ­0.8 36.7 2.9 55 0.2 2 0.846

100.0 0.1 3.1 0.23 0.8 17.8 77.8 9.3 50 1.9 1 0.904

79.4 24.2 3.3 0.26 39.6 10.3 8.6 1.2 82 0.1 .. 0.972

79.6 25.5 1.7 0.14 57.7 ­5.3 0.7 5.5 64 1.3 .. 0.724

87.7 12.4 6.4 0.47 22.4 33.6 20.9 0.9 113 0.6 27 0.744

86.9 13.8 2.1 0.21 50.2 ­5.0 .. 2.9 25 0.6 30 0.679

98.4 2.3 3.2 0.21 14.1 37.7 36.9 13.4 64 1.1 .. 0.912

75.3 4.1 4.4 0.62 16.7 4.4 5.6 1.0 71 0.3 25 0.946

86.6 16.5 2.2 0.15 41.7 82.5 30.4 1.6 43 2.2 .. 0.734

.. 2.1 .. .. 33.2 ­7.2 14.3 0.0 30 0.6 .. 0.842

88.9 12.6 2.2 0.21 6.8 63.5 103.3 1.6 56 1.0 13 0.974

93.2 3.4 5.9 0.53 8.0 ­0.6 1.3 22.8 156 1.3 13 0.948

97.6 3.6 3.5 0.30 13.4 4.9 40.2 0.0 51 0.8 .. 0.961

74.7 28.9 3.2 0.20 18.9 ­21.7 1.7 0.5 101 11.8 51 0.979

.. 5.8 .. .. 69.2 8.0 7.9 0.0 48 1.3 .. 0.772

81.0 16.8 2.5 0.31 30.9 ­2.8 12.5 0.3 25 0.6 .. 0.724

88.4 12.8 4.3 0.33 52.7 ­10.6 1.7 1.0 35 1.4 15 0.825

.. 7.8 .. .. 57.4 ­13.9 10.0 0.0 … 0.672

.. 31.3 .. .. 55.9 7.3 .. 0.8 99 2.9 .. 0.669

33.7 61.7 0.9 0.11 37.7 ­29.1 0.6 1.6 57 1.5 52 0.948

76.3 24.9 3.4 0.25 98.3 ­0.7 .. 28.1 57 2.0 21 0.983

97.6 3.2 2.5 0.31 1.1 ­0.6 96.4 0.1 51 0.6 4 0.963 DASH

.. 35.0 .. .. 59.7 ­15.8 .. 0.5 69 1.0 81 0.743 BOARD 0.3 .. 0.843 .. 1.0 .. .. 3.3 0.0 15.7 0.0 26 4

.. 1.9 .. .. 12.5 0.0 .. 0.0 73 1.4 .. 0.725

62.4 27.5 1.1 0.16 27.8 26.3 17.8 0.7 185 4.2 38 0.644

88.7 14.3 1.9 0.45 12.6 29.6 8.7 0.2 78 0.1 29 0.969

DASHBOARD 4 Environmental sustainability | 339 DASHBOARD 4 ENVIRONMENTAL SUSTAINABILITY

SDG 12.c SDG 7.2 SDG 9.4 SDG 15.1 SDG 6.4 SDG 12.2 SDG 3.9 SDG 3.9 SDG 15.3 SDG 15.5

Fossil fuel Renewable Natural Household Unsafe water, 2018 resource and ambient sanitation 0.975 energy energy Fresh water depletion air pollution and hygiene Degraded 0.969 withdrawals services land 0.969 consumption consumption Carbon dioxide emissions Forest area 0.754 (% of total (% of total (% of total 0.772 renewable water 0.870 energy final energy Per capita (kg per 2010 (% of total Change (% of total 0.961 (tonnes) US$ of GDP) land areaa) (%) resources) (% of GNI) land area) 0.909 consumption) consumption) (per 100,000 population) 2016 2016 HDI rank 2010­2015b 2015 12.2 0.79 2016 1990/2016 2007­2017b 2012­2017b 2016 2016 2015 108 Turkmenistan .. 0.0 2.7 0.96 8.8 0.0 .. .. 79 4.0 22 108 Uzbekistan 6.7 0.17 110 Libya 97.7 3.0 1.7 7.5 5.4 108.1 9.4 81 0.4 29 111 Indonesia .. 111 Samoa 99.1 2.0 .. 0.62 0.1 0.0 822.9 6.7 72 0.6 .. 113 South Africa 7.4 0.28 114 Bolivia (Plurinational State of) 66.1 36.9 1.8 0.10 49.9 ­23.8 11.0 1.9 112 7.1 21 115 Gabon 1.7 0.21 116 Egypt .. 34.3 2.2 60.4 31.5 .. 0.0 85 1.5 .. 117 Marshall Islands 86.8 17.2 7.6 0.0 30.2 2.7 87 13.7 78 119 Palestine, State of 84.2 17.5 50.3 ­13.2 0.4 5.8 64 5.6 18 121 Morocco 22.8 82.0 90.0 5.5 .. 10.5 76 20.6 16 123 Guyana 97.9 5.7 0.1 67.3 114.1 4.0 109 2.0 1 125 Tajikistan .. 11.2 .. .. 70.2 … … 0.839 126 Guatemala 69.8 35.0 2.0 0.35 48.1 67.1 .. 1.0 64 1.6 31 0.733 129 India .. 10.5 .. .. 1.5 1.0 42.8 … 15 0.780 131 Timor-Leste 96.0 0.8 3.8 0.24 1.9 3.4 42.9 10.9 75 3.0 26 0.799 132 Kiribati 88.5 11.3 1.6 0.22 12.6 13.5 35.7 0.3 49 1.9 19 0.887 135 Bangladesh 75.5 23.3 1.5 0.47 3.3 ­24.8 .. 6.3 111 0.8 24 0.984 137 Sao Tome and Principe .. 25.3 .. .. 83.9 ­0.9 0.5 13.3 108 3.6 16 0.922 138 Eswatini (Kingdom of) 48.4 24.4 1.1 0.14 12.6 ­30.9 .. 1.0 42 2.0 16 0.826 141 Vanuatu 46.0 44.7 0.6 0.20 3.0 1.9 .. 3.5 129 2.7 97 0.985 143 Zambia .. 26.6 .. .. 22.5 57.3 .. 0.5 99 4.1 17 0.890 145 Myanmar 37.4 63.7 1.0 0.14 32.7 ­26.2 .. 1.7 74 6.3 24 0.721 147 Kenya 40.7 48.2 0.8 0.17 25.9 ­31.0 0.9 2.9 56 2.2 .. 0.852 149 Angola 73.6 36.0 1.6 0.26 23.8 10.8 33.9 1.0 184 18.6 30 0.678 150 Zimbabwe 66.7 26.5 1.7 0.17 8.3 ­21.9 .. 2.6 145 18.3 19 0.966 153 Solomon Islands .. 18.2 .. .. 45.4 ­30.1 .. 29.7 140 9.9 .. 0.885 154 Syrian Arab Republic 52.5 51.5 1.0 0.23 40.0 ­45.0 .. 1.6 61 3.6 .. 0.743 156 Comoros .. 4.3 .. .. 15.0 0.0 .. 0.0 140 16.7 .. 0.760 158 Nigeria .. 86.9 .. .. 72.5 35.1 0.4 2.7 124 3.9 10 0.799 159 Uganda 73.8 34.7 0.5 0.14 11.0 ­4.5 2.9 0.6 149 11.9 65 0.760 162 Madagascar .. 1.2 .. .. 91.9 … 152 3.6 .. 0.686 .. 41.1 .. .. 55.8 ­4.3 1.9 0.0 162 11.4 .. 0.785

40.5 62.4 0.5 0.10 65.4 ­1.8 .. 31.4 131 38.7 10 0.983

.. 66.1 .. .. 34.3 25.1 .. 1.7 137 27.9 13 0.817

.. 59.3 .. .. 82.1 7.4 .. 6.3 188 11.3 .. 0.810

.. 36.1 .. .. 36.1 0.0 .. 0.0 136 10.4 .. 0.662

52.5 41.4 0.4 0.12 41.2 8.6 .. 11.4 204 18.8 14 0.844

10.6 88.0 0.2 0.06 65.2 ­8.2 .. 8.3 127 34.9 7 0.879

.. 7.8 .. .. 55.5 ­16.3 .. 22.9 178 22.3 19 0.813

44.3 61.5 0.4 0.08 43.6 ­27.3 .. 2.7 156 12.6 23 0.806

30.6 64.9 0.6 0.17 52.9 ­27.9 .. 1.0 150 6.5 33 0.816

17.4 72.7 0.3 0.11 7.8 ­5.8 13.1 2.5 78 51.2 40 0.797

15.5 85.3 0.3 0.13 25.4 ­24.7 .. 0.9 194 19.8 .. 0.825

48.3 49.6 0.7 0.12 46.3 ­5.3 .. 12.8 119 48.8 20 0.934

38.3 76.5 0.3 0.08 39.3 ­23.5 .. 2.5 208 45.2 0 0.836

29.1 81.8 0.7 0.35 35.5 ­38.0 17.9 3.1 133 24.6 36 0.789

61.6 46.5 0.8 0.17 1.9 ­43.5 74.4 0.8 174 19.6 5 0.859

.. 63.3 .. .. 77.9 ­6.2 .. 20.9 137 6.2 .. 0.767

97.8 0.5 1.5 0.77 2.7 32.1 .. .. 75 3.7 .. 0.943

.. 52.5 .. .. 74.1 ­0.2 .. 14.0 152 16.3 21 0.839

.. 45.3 .. .. 19.7 ­25.3 .. 1.8 172 50.7 22 0.764

.. 86.7 .. .. 19.7 53.1 .. 5.4 121 19.3 12 0.848

DASH 18.9 86.6 0.5 0.09 7.2 ­61.8 4.4 4.4 307 68.6 32 0.874 14.4 85.7 0.2 0.08 51.6 ­18.3 .. 2.2 139 38.4 .. 0.689 4 .. 89.1 .. .. 9.7 ­59.3 1.1 14.1 156 31.6 22 0.751

.. 32.2 .. .. 0.2 ­46.7 .. 12.4 169 38.6 3 0.977

.. 70.2 .. .. 21.4 ­9.1 .. 0.8 160 30.2 30 0.788

36.7 50.9 0.5 0.27 37.8 ­26.0 .. 1.8 205 59.7 53 0.910

340 | HUMAN DEVELOPMENT REPORT 2019

SDG 12.c SDG 7.2 SDG 9.4 SDG 15.1 SDG 6.4 SDG 12.2 SDG 3.9 SDG 3.9 SDG 15.3 SDG 15.5

Fossil fuel Renewable Natural Household Unsafe water, 2018 resource and ambient sanitation 0.953 energy energy Fresh water depletion air pollution and hygiene Degraded 0.888 withdrawals services land 0.943 consumption consumption Carbon dioxide emissions Forest area 0.854 0.933 (% of total (% of total (% of total 0.721 renewable water 0.837 energy final energy Per capita (kg per 2010 (% of total Change (% of total 0.816 (tonnes) US$ of GDP) land areaa) (%) resources) (% of GNI) land area) 0.808 consumption) consumption) (per 100,000 population) 0.842 2016 2016 0.981 HDI rank 2010­2015b 2015 .. .. 2016 1990/2016 2007­2017b 2012­2017b 2016 2016 2015 0.894 0.887 164 Lesotho .. 52.1 0.4 0.13 1.6 25.0 .. 5.1 178 44.4 20 0.884 165 Côte d’Ivoire 0.5 0.23 0.960 166 Senegal 26.5 64.5 0.3 0.19 32.7 1.7 1.4 2.2 269 47.2 14 0.891 167 Togo 0.5 0.11 0.825 168 Sudan 53.9 42.7 0.3 0.18 42.8 ­11.9 .. 1.0 161 23.9 6 0.911 169 Haiti 0.988 170 Afghanistan 17.8 71.3 .. .. 3.1 ­75.4 .. 13.4 250 41.6 12 0.907 171 Djibouti .. .. 0.981 172 Malawi 31.7 61.6 … .. 71.2 2.8 185 17.3 12 0.921 173 Ethiopia 0.1 0.07 0.931 174 Gambia 22.0 76.1 .. .. 3.5 ­17.1 10.3 1.2 184 23.8 .. 0.920 174 Guinea .. .. 0.943 176 Liberia .. 18.4 .. .. 2.1 0.0 .. 0.3 211 13.9 8 0.936 177 Yemen 0.3 0.15 178 Guinea-Bissau .. 15.4 .. .. 0.2 0.0 .. 0.7 159 31.3 .. 0.899 179 Congo (Democratic Republic of the) 0.0 0.03 0.759 180 Mozambique .. 83.6 0.3 0.23 33.2 ­19.7 .. 8.2 115 28.3 17 0.772 181 Sierra Leone .. .. 0.992 182 Burkina Faso 6.6 92.2 .. .. 12.5 .. 8.7 9.4 144 43.7 29 0.900 182 Eritrea 0.2 0.08 0.840 184 Mali .. 51.5 .. .. 48.4 10.8 .. 5.7 237 29.7 14 185 Burundi .. .. — 186 South Sudan .. 76.3 0.2 0.08 25.8 ­12.9 .. 13.0 243 44.6 11 — 187 Chad .. .. — 188 Central African Republic .. 83.8 .. .. 43.1 ­15.8 .. 19.2 170 41.5 29 — 189 Niger 0.1 0.11 — OTHER COUNTRIES OR TERRITORIES 98.5 2.3 1.0 0.0 .. 0.2 194 10.2 .. — .. Korea (Democratic People’s Rep. of) .. 86.9 69.8 ­11.5 .. 11.4 215 35.3 15 — .. Monaco — .. Nauru 5.4 95.8 67.2 ­5.0 .. 23.2 164 59.8 6 — .. San Marino — .. Somalia 12.6 86.4 48.0 ­13.0 0.7 1.3 110 27.6 .. — .. Tuvalu — Human development groups .. 77.7 43.1 ­0.3 .. 12.9 324 81.3 18 —

Very high human development .. 74.2 19.3 ­22.7 .. 15.0 206 49.6 19 — High human development — Medium human development 23.1 79.8 14.9 ­7.1 .. .. 174 45.6 35 Low human development Developing countries .. 61.5 3.8 ­30.7 .. 9.5 209 70.7 3 Arab States .. 95.7 10.9 ­2.9 .. 15.7 180 65.4 29 East Asia and the Pacific Europe and Central Asia 72.2 39.1 .. .. 1.3 14.0 165 63.3 .. Latin America and the Caribbean South Asia .. 89.4 3.8 ­29.2 .. 13.1 280 101.0 34 Least developed countries .. 76.6 35.6 ­1.8 .. 0.1 212 82.1 13 Small island developing states Organisation for Economic 24.1 78.9 0.9 ­41.9 5.1 11.9 252 70.8 7 World 62.1 23.1 1.0 0.25 40.7 ­40.2 .. .. 207 1.4 ..

.. 0.1 .. .. 0.0 0.0 … .. ..

… .. 0.0 0.0 … .. ..

.. 94.3 .. .. 10.0 ­24.1 .. 8.9 213 86.6 23

.. 0.0 .. .. 33.3 0.0 … .. ..

82.4 10.5 9.6 0.25 32.9 1.2 6.4 0.7 25 0.3 ..

84.9 15.8 4.7 0.36 31.6 ­4.3 5.9 1.5 94 1.9 25

69.0 39.8 1.3 0.23 30.9 ­7.7 .. 2.2 164 18.0 23

.. 81.0 .. .. 24.9 ­12.0 .. 6.4 202 46.5 16

80.5 23.5 3.1 0.32 27.1 ­6.4 .. 2.1 133 14.0 23

95.5 4.0 4.4 0.29 1.8 ­1.9 76.1 6.6 101 7.0 7

.. 15.9 .. .. 29.8 3.9 .. 1.1 115 2.2 ..

87.0 9.1 4.6 0.29 9.2 8.6 20.3 2.1 67 0.5 28

74.5 27.7 2.6 0.19 46.2 ­9.6 1.5 2.3 40 1.7 28

76.9 31.1 1.6 0.26 14.7 7.8 25.0 1.3 174 17.1 23

39.2 70.2 0.8 0.25 28.1 ­11.9 .. 6.1 187 47.8 22

.. 73.2 .. .. 29.1 ­11.3 .. 5.7 167 34.3 16

.. 17.8 .. .. 69.4 1.3 .. 1.5 92 8.9 ..

79.6 12.0 9.0 0.24 31.4 1.6 9.1 0.4 19 0.4 ..

80.6 18.2 4.3 0.27 31.2 ­3.0 7.7 1.1 114 11.7 20

4

DASHBOARD 4 Environmental sustainability | 341 DASHBOARD 4 ENVIRONMENTAL SUSTAINABILITY

NOTES Renewable energy consumption: Share of areas resulting from human intervention or natural that have experienced the reduction or loss of renewable energy in total final energy consumption. causes that are expected to regenerate. biological or economic productivity and complexity Three-colour coding is used to visualize partial Renewable sources include hydroelectric, resulting from a combination of pressures, including grouping of countries and aggregates by indicator. geothermal, solar, tides, wind, biomass and biofuels. Fresh water withdrawals: Total fresh water land use and management practices. For each indicator countries are divided into three withdrawn, expressed as a percentage of total groups of approximately equal size (terciles): the Carbon dioxide emissions: Human-originated renewable water resources. Red List Index: Measure of the aggregate top third, the middle third and the bottom third. carbon dioxide emissions stemming from the burning extinction risk across groups of species. It is based Aggregates are colour coded using the same tercile of fossil fuels, gas flaring and the production of Natural resource depletion: Monetary valuation on genuine changes in the number of species in each cutoffs. See Technical note 6 at http://hdr.undp.org/ cement. Carbon dioxide emitted by forest biomass of energy, mineral and forest depletion, expressed as category of extinction risk on the International Union sites/default/files/hdr2019_technical_notes.pdf for through depletion of forest areas is included. a percentage of gross national income (GNI). for Conservation of Nature Red List of Threatened details about partial grouping in this table. Data are expressed in tonnes per capita (based on Species. It ranges from 0, all species categorized midyear population) and in kilograms per unit of Mortality rate attributed to household and as extinct, to 1, all species categorized as least a This column is intentionally left without colour gross domestic product (GDP) in constant 2010 US ambient air pollution: Deaths resulting from concern. because it is meant to provide context for the dollars. exposure to ambient (outdoor) air pollution and indicator on change in forest area. household (indoor) air pollution from solid fuel use MAIN DATA SOURCES Forest area: Land spanning more than 0.5 hectare for cooking, expressed per 100,000 population. b Data refer to the most recent year available with trees taller than 5 metres and a canopy cover Ambient air pollution results from emissions from Columns 1, 2, 5 and 8: World Bank (2019a). during the period specified. of more than 10 percent or trees able to reach these industrial activity, households, cars and trucks. thresholds in situ. It excludes land predominantly Columns 3, 4, 11 and 12: United Nations Statistics c Less than 0.1. under agricultural or urban land use, tree stands in Mortality rate attributed to unsafe water, Division (2019a). agricultural production systems (for example, in fruit sanitation and hygiene services: Deaths DEFINITIONS plantations and agroforestry systems) and trees in attributable to unsafe water, sanitation and hygiene Column 6: HDRO calculations based on data on urban parks and gardens. Areas under reforestation focusing on inadequate wash services, expressed forest area from World Bank (2019a). Fossil fuel energy consumption: Percentage of that have not yet reached but are expected to reach per 100,000 population. total energy consumption that comes from fossil a canopy cover of 10 percent and a tree height of Column 7: FAO (2019b). fuels, which consist of coal, oil, petroleum and 5 metres are included, as are temporarily unstocked Degraded land: Rain-fed cropland, irrigated natural gas products. cropland, or range, pasture, forest and woodlands Columns 9 and 10: WHO (2019).

4

342 | HUMAN DEVELOPMENT REPORT 2019 DASHBOARD 5 Socioeconomic sustainability HUMAN DEVELOPMENT REPORT 2019

Top third Middle third Bottom third

Three-colour coding is used to visualize partial grouping of countries by indicator. For each indicator countries are divided into three groups of approximately equal size (terciles): the top third, the middle third and the bottom third. Aggregates are colour coded using the same tercile cutoffs. See Notes after the table.

SDG 17.4 SDG 9.5 SDG 10.1 SDG 5 SDG 10.1

Economic sustainability Social sustainability

Research Dependency expenditure versus

Gross Skilled Concentration and ratio military expenditure capital labour Overall loss in Gender Income share Adjusted Total index development Old age Military HDI value due Inequality of the poorest net savings debt service (exports) expenditure (65 and older) expenditurea Ratio of to inequalityc Indexc 40 percent

(% of exports education services and health income) (per 100 expenditure (% of labour to military Average annual change force) 15­64) (% of GNI) (value) (% of GDP) (% of GDP) expenditureb (%)

HDI rank 2015­2017d 2015­2017d 2015­2018d 2010­2018d 2018 2010­2017d 2030e 2010­2018d 2010­2016f 2010/2018g 2005/2018g 2005/2017

VERY HIGH HUMAN DEVELOPMENT 16.9 .. 27.6 84.3 0.368 2.0 31.9 h 1.6 11.7 0.3 ­3.7 0.3 2 Switzerland 16.4 .. 23.2 86.5 0.246 i 3.4 37.9 0.7 25.5 ­0.4 ­3.8 0.8 4 Germany 16.0 .. 25.4 84.9 0.269 1.2 27.8 0.3 32.7 ­2.3 ­4.0 0.4 6 Australia 14.1 .. 21.3 87.4 0.093 2.9 44.0 1.2 13.5 0.5 ­2.2 0.0 8 Sweden .. .. 21.7 77.0 0.286 0.8 43.2 … .. .. 10 Netherlands 5.6 .. 24.3 78.9 0.291 1.9 31.0 j 1.9 7.5 0.4 ­2.0 ­0.6 12 Finland 16.6 .. 22.6 74.5 0.461 2.1 31.8 .. .. ­1.7 ­4.2 0.2 14 New Zealand 19.1 .. 26.5 86.8 0.097 3.3 36.4 1.0 17.2 ­0.4 ­1.8 ­0.4 15 United States 36.8 .. 26.6 65.9 0.269 2.2 34.5 3.1 2.1 .. ­4.7 .. 18 Liechtenstein 18.4 .. 21.2 78.4 0.082 2.0 40.8 1.2 13.9 ­2.1 ­3.9 0.3 20 Austria 18.3 .. 22.7 78.7 0.101 2.9 37.1 1.2 15.5 ­0.7 ­2.9 ­0.5 22 Israel 10.2 .. 23.7 89.9 0.143 2.7 43.1 k 1.4 11.5 ­3.6 ­3.3 0.0 24 Slovenia 6.5 .. 23.1 91.8 0.147 1.5 36.7 1.3 13.0 0.2 ­2.9 ­0.3 26 Czechia 13.9 .. 23.5 82.2 0.175 1.3 33.3 1.2 13.2 .. ­2.2 .. 28 Malta 5.5 .. 17.2 83.6 0.111 1.7 34.8 1.8 8.5 ­1.9 ­3.2 0.3 30 Estonia 6.1 .. 20.6 96.4 0.099 2.7 32.5 3.2 6.2 2.2 ­2.4 ­0.4 32 Greece 12.0 .. 25.4 85.5 0.096 2.5 37.6 0.9 18.1 ­1.2 ­4.3 0.1 34 Lithuania … … … … 36 Andorra 7.3 .. 23.9 99.9 0.139 3.1 53.2 0.9 15.3 .. ­2.5 .. 36 Slovakia 14.1 .. 25.3 87.4 0.061 3.1 38.5 0.7 22.6 0.4 ­2.9 ­0.5 40 Portugal 20.9 .. 18.3 78.3 0.106 1.2 27.1 0.6 20.9 0.7 ­3.8 ­0.9 42 Chile 15.6 .. 20.8 90.6 0.223 4.3 22.5 4.3 2.3 ­1.8 ­3.2 0.7 43 Hungary 20.1 .. 30.2 85.7 0.175 4.2 38.2 2.6 4.7 ­1.8 ­3.5 0.1 46 Croatia 10.1 .. 21.9 91.1 0.177 2.0 41.8 1.0 14.4 ­3.9 ­3.9 ­0.2 48 Argentina 9.1 .. 21.9 66.9 0.096 1.2 39.8 l 1.3 10.6 5.9 ­2.8 ­1.2 10.3 .. 26.2 95.7 0.128 1.7 35.3 1.1 13.7 ­3.0 ­0.8 0.2

9.3 .. 23.5 84.8 0.089 2.2 40.4 2.3 7.5 ­0.1 ­4.8 ­0.5

.. .. 18.4 63.4 0.292 0.6 41.9 0.5 29.8 .. ­2.8 ­0.3

6.0 .. 18.0 69.6 0.053 1.3 45.8 1.3 10.8 0.1 ­4.7 ­0.6

15.2 .. 27.0 89.8 0.099 1.3 38.3 2.1 5.8 ­3.2 ­4.4 0.0

3.2 .. 19.1 85.1 0.401 0.5 27.0 m 1.6 7.9 ­2.0 ­3.0 ­0.9

­3.1 .. 13.1 78.3 0.295 1.0 42.5 2.4 .. 2.4 ­2.4 ­0.7

10.6 .. 20.7 95.1 0.063 1.0 37.0 2.0 5.2 ­3.2 ­2.1 0.9

.. .. 18.2 96.2 0.116 0.8 45.2 2.0 9.4 ­0.6 ­2.5 ­0.7

.. .. 22.4 52.8 0.276 1.0 6.4 5.6 .. .. ­6.2 ..

… .. 0.189 … … ..

13.4 .. 25.9 58.6 0.515 0.8 8.3 8.8 1.1 n .. ­5.1 ..

5.6 .. 23.6 95.5 0.216 0.8 32.7 1.2 10.3 ­1.0 0.1 0.2

6.0 .. 24.2 92.5 0.084 0.4 42.3 2.0 10.5 ­2.2 ­1.7 1.2

3.4 .. 17.5 54.1 0.080 1.3 44.3 1.8 7.7 0.8 ­4.3 0.4

26.8 .. 44.6 43.9 0.450 0.5 5.7 1.5 4.2 …

3.6 .. 22.7 70.3 0.325 0.4 26.0 1.9 7.2 1.0 ­1.9 1.4

34.6 .. 41.1 79.2 0.623 .. 14.4 2.4 1.9 …

13.2 .. 27.1 88.6 0.108 1.2 34.5 1.1 12.7 ­0.9 0.0 0.7

20.4 .. 32.9 19.3 0.372 0.1 7.1 3.6 1.6 .. ­2.8 ..

10.8 .. 21.4 91.5 0.071 0.8 40.5 1.5 6.9 ­5.7 ­1.9 0.6

­11.3 .. 31.3 .. 0.447 0.2 6.0 8.2 0.9 .. ­1.7 .. DASH

5.4 .. 20.8 65.8 0.227 0.5 19.7 0.9 16.1 ­3.6 ­0.4 2.0 BOARD 8.0 26.0 22.7 96.4 0.327 1.1 31.1 3.9 1.9 ­1.8 ­2.2 5

DASHBOARD 5 Socioeconomic sustainability | 343 DASHBOARD 5 SOCIOECONOMIC SUSTAINABILITY

SDG 17.4 SDG 9.5 SDG 10.1 SDG 5 SDG 10.1

Economic sustainability Social sustainability

Gross Skilled Concentration and ratio military expenditure capital labour Overall loss in Gender Income share Adjusted Total index development Old age Military HDI value due Inequality of the poorest net savings debt service (exports) expenditure (65 and older) expenditurea Ratio of to inequalityc Indexc 40 percent

(% of exports education services and health income) (per 100 expenditure (% of labour to military Average annual change (% of GNI) (value) (% of GDP) (% of GDP) expenditureb (%)

HDI rank 2015­2017d 2015­2017d 2015­2018d 2010­2018d 2018 2010­2017d 2030e 2010­2018d 2010­2016f 2010/2018g 2005/2018g 2005/2017

50 Belarus 21.2 11.8 27.5 98.6 0.183 0.6 32.5 1.3 8.9 ­3.9 .. 0.5 52 Bulgaria 5.8 47.9 26.6 74.0 0.601 0.1 17.4 1.0 6.8 ­5.9 ­3.4 3.1 52 Romania 14.8 21.3 20.7 88.8 0.092 0.8 37.2 1.7 7.4 1.4 ­1.1 ­0.3 56 Barbados .. 13.4 31.4 90.7 0.218 0.4 30.1 1.5 .. ­1.6 .. ­0.4 57 Uruguay 3.4 22.4 24.2 81.0 0.114 0.5 32.6 1.9 5.5 ­1.0 ­0.8 0.8 60 Bahamas .. .. 28.5 92.6 0.499 … … .. 62 Seychelles ­6.8 o .. 18.3 .. 0.160 .. 35.4 … ­2.0 .. 63 Serbia 14.6 .. 29.1 .. 0.303 0.1 10.0 5.1 .. .. ­2.5 .. 65 Iran (Islamic Republic of) 10.2 .. 16.5 26.4 0.226 0.4 27.0 2.0 7.6 ­2.4 ­2.0 1.7 67 Panama 11.4 40.2 29.2 44.2 0.076 0.9 18.5 2.5 4.6 ­3.9 ­3.5 0.2 69 Albania 7.1 .. 27.1 .. 0.423 .. 17.1 … ­0.1 .. 71 Sri Lanka 10.0 .. 23.6 66.9 0.218 1.3 14.7 p 1.0 6.1 .. ­1.2 1.5 73 Saint Kitts and Nevis .. .. 32.5 94.2 0.469 0.2 19.2 1.4 4.5 … 75 Bosnia and Herzegovina ­3.2 q 22.0 21.5 83.2 0.081 0.9 32.7 r 1.9 7.0 0.4 .. 2.0 77 Thailand … 71.9 0.348 0.1 24.1 0.8 .. .. ­0.6 .. 79 Brazil .. 0.4 34.7 18.0 s 0.523 0.3 14.1 2.7 3.9 .. ­0.3 1.0 81 Armenia ­6.4 19.8 19.1 61.1 0.219 0.2 26.7 t 0.2 57.3 .. 0.0 ­0.1 82 North Macedonia 25.3 .. 41.7 53.3 0.143 0.1 17.4 0.0 .. ­3.1 ­0.2 1.5 85 China 15.9 14.8 18.6 39.1 0.262 0.5 22.6 0.0 .. ­0.7 ­1.3 ­0.1 87 Azerbaijan 8.2 10.4 25.0 54.6 0.292 0.2 n 32.7 1.2 9.7 ­1.7 ­2.2 0.5 89 Dominican Republic 12.5 29.4 33.3 92.5 0.209 0.3 29.5 u 1.9 5.6 ­3.7 ­0.7 0.0 91 Tunisia 28.5 21.2 28.6 38.1 0.194 0.1 24.2 1.9 3.4 ­3.7 ­1.0 0.3 93 Lebanon .. .. 10.3 69.4 0.235 0.3 33.8 2.9 7.1 .. ­0.6 .. 94 Saint Vincent and the Grenadines … .. 0.283 … … .. 96 Venezuela (Bolivarian Republic of) … .. 0.416 .. 20.7 … .. .. 98 Fiji .. 15.6 21.7 85.0 0.100 0.2 37.5 1.1 .. ­3.7 .. 0.2 98 Suriname 7.5 14.0 23.0 40.9 0.137 0.5 15.2 0.5 16.5 0.9 ­1.7 1.6 103 Belize 14.0 4.7 25.0 38.0 0.079 0.8 29.6 1.3 5.4 ­2.5 0.6 1.2 105 Tonga .. 9.4 .. .. 0.208 .. 18.8 … .. ..

6.1 36.2 15.4 64.1 0.159 1.3 19.9 1.5 13.0 ­1.2 ­1.4 1.0

2.8 41.6 21.2 58.1 0.341 0.2 19.3 3.2 3.4 ­2.4 ­1.3 1.0

1.5 27.0 22.4 95.7 0.264 0.2 26.1 4.8 3.1 ­1.2 ­2.8 0.4

21.2 0.6 48.4 40.4 0.483 0.5 14.0 5.3 2.8 n .. ­1.6 ..

15.4 13.7 33.0 81.4 0.221 0.4 27.4 1.0 .. ­2.7 .. 3.3

7.1 21.7 21.7 82.8 0.295 0.1 17.5 1.2 6.9 ­4.6 ­1.3 2.0

20.1 7.6 44.3 .. 0.094 2.1 25.0 1.9 .. ­3.7 ­2.3 0.7

11.4 29.3 26.0 46.3 0.393 0.4 15.5 2.4 5.2 ­0.2 ­1.2 2.4

9.5 10.7 20.1 93.3 0.827 0.2 17.3 v 3.8 2.6 ­4.0 ­0.1 ..

3.5 20.7 18.8 98.3 0.140 0.4 30.2 w 3.8 3.2 ­2.5 ­1.8 0.9

17.3 15.6 24.4 43.8 0.188 .. 15.7 0.7 10.0 x ­1.7 ­0.4 1.2

­2.3 4.6 21.8 .. 0.268 .. 21.1 … .. ..

­9.6 17.2 19.8 54.9 0.137 0.6 19.0 2.1 6.0 ­2.2 ­0.9 1.3

­10.3 56.2 42.2 79.3 0.445 0.1 10.5 0.8 10.4 ­1.3 ­1.7 0.2

­16.9 70.6 17.2 .. 0.117 .. 17.9 5.0 2.4 …

26.6 2.5 29.4 34.0 0.891 0.5 8.6 2.8 5.1 y .. ­0.9 3.6

0.4 11.6 26.4 .. 0.524 .. 20.0 … .. ..

15.9 27.3 22.6 .. 0.498 .. 17.9 1.4 11.9 0.1 ­1.0 ..

7.2 q 57.4 24.8 42.3 0.734 0.1 15.0 0.5 11.2 y ­2.3 ­0.3 ..

.. 11.7 .. .. 0.409 … … ..

8.1 2.3 .. 62.5 0.220 .. 12.5 0.9 5.3 .. ­1.2 0.5

14.5 12.4 23.1 43.7 0.348 0.2 13.0 0.9 13.2 ­1.1 ­0.8 0.9

22.9 z .. 36.2 45.0 0.668 .. 15.1 .. .. ­0.8 ­0.8 ..

DASH 4.4 12.4 18.2 .. 0.163 0.3 8.2 4.7 2.0 ­2.9 ­1.3 1.2 ­0.9 9.7 17.9 43.5 0.311 .. 10.5 1.3 10.6 ­2.7 ­1.2 .. 5 .. 3.5 .. 32.7 0.617 .. 9.0 .. .. 4.4 ­1.2 ­0.1

9.3 aa 9.9 33.4 .. 0.297 .. 10.8 … ­1.1 0.4

344 | HUMAN DEVELOPMENT REPORT 2019

SDG 17.4 SDG 9.5 SDG 10.1 SDG 5 SDG 10.1

Economic sustainability Social sustainability

Gross Skilled Concentration and ratio military expenditure capital labour Overall loss in Gender Income share Adjusted Total index development Old age Military HDI value due Inequality of the poorest net savings debt service (exports) expenditure (65 and older) expenditurea Ratio of to inequalityc Indexc 40 percent

(% of exports education services and health income) (per 100 expenditure (% of labour to military Average annual change (% of GNI) (value) (% of GDP) (% of GDP) expenditureb (%)

HDI rank 2015­2017d 2015­2017d 2015­2018d 2010­2018d 2018 2010­2017d 2030e 2010­2018d 2010­2016f 2010/2018g 2005/2018g 2005/2017

106 Philippines 28.5 11.3 26.9 29.9 0.250 0.1 11.5 1.1 5.6 y ­0.5 ­0.7 0.3 108 Turkmenistan 14.7 10.7 25.3 60.0 0.189 0.3 24.6 ab 0.3 35.8 ­2.9 ­1.7 2.2 108 Uzbekistan 110 Libya .. .. 47.2 .. 0.645 .. 10.8 .. .. ­3.7 .. .. 111 Indonesia 111 Samoa .. .. 40.2 .. 0.349 0.2 11.3 3.6 … .. 113 South Africa 114 Bolivia (Plurinational State of) .. .. 29.8 n .. 0.798 .. 9.0 15.5 .. .. ­3.3 .. 116 Egypt 12.0 34.0 34.6 39.8 0.134 0.1 13.5 0.7 7.4 ­0.2 ­1.2 ­1.4 117 Marshall Islands .. 8.9 .. 66.6 0.366 .. 11.4 … ­1.6 0.5 119 Palestine, State of 0.6 12.2 18.0 51.2 0.151 0.8 9.9 1.0 13.1 1.3 0.0 ­0.2 121 Morocco 0.8 10.5 20.6 44.0 0.379 0.2 y 13.7 1.5 6.9 ­4.6 ­1.5 4.4 123 Guyana 8.9 aa 3.8 aa 21.4 35.5 0.546 0.6 y 6.4 1.5 4.5 0.8 ­0.7 0.5 125 Tajikistan 1.2 15.1 16.7 54.9 0.154 0.6 10.2 1.2 3.8 n 1.0 ­1.7 0.1 126 Guatemala .. .. 22.4 .. 0.752 … … .. 129 India 13.4 5.9 27.5 32.3 0.188 0.4 17.9 2.3 5.5 ­0.1 ­0.1 0.1 131 Timor-Leste .. .. 24.2 46.9 0.176 0.5 6.7 ac … .. 0.0 132 Kiribati ­7.0 .. 17.8 28.3 0.958 0.0 6.1 2.7 … ­0.6 135 Bangladesh 20.9 9.8 33.4 18.7 s 0.174 0.7 17.1 3.1 3.4 y .. ­1.2 0.3 137 Sao Tome and Principe 12.9 29.9 35.4 92.7 0.364 0.1 11.3 1.6 7.5 ­4.6 ­3.4 1.1 138 Eswatini (Kingdom of) 14.1 5.0 31.1 42.0 0.452 .. 16.1 1.7 6.8 ­0.1 ­0.6 .. 141 Vanuatu 6.4 20.2 20.4 37.4 0.213 0.1 16.3 1.0 10.5 ­2.6 ­1.4 2.9 143 Zambia 6.3 26.1 27.2 80.1 y 0.265 0.1 8.4 1.2 9.9 ­4.3 0.0 ­0.2 145 Myanmar 11.7 5.9 40.4 59.8 0.315 0.1 10.4 0.6 17.1 … 147 Kenya 1.9 28.6 12.1 18.1 0.136 0.0 9.5 0.4 20.4 ­2.3 ­1.1 1.4 149 Angola 14.4 19.8 22.9 30.5 0.221 0.1 12.0 0.6 20.0 ­0.8 ­1.2 0.8 150 Zimbabwe 16.3 10.1 31.0 17.6 0.139 0.6 12.5 2.4 3.1 ­5.4 ­1.6 ­0.5 153 Solomon Islands 4.5 .. 12.6 66.7 0.265 0.3 6.6 3.3 2.7 ­2.5 ­1.0 0.3 154 Syrian Arab Republic ­14.6 0.1 22.5 28.2 0.467 .. 8.2 0.6 6.9 ­2.0 .. 1.5 156 Comoros 19.5 23.9 25.5 24.3 0.222 0.0 10.0 1.7 8.8 ­2.1 ­0.5 3.2 158 Nigeria … 48.3 0.907 .. 10.1 … .. .. 159 Uganda 23.3 10.5 51.3 19.5 0.392 .. 11.1 … .. 0.4

24.5 5.5 31.2 25.8 0.404 .. 10.7 1.4 2.8 ­2.2 ­1.2 0.0

… 65.0 0.805 .. 9.7 … .. 0.6

.. 3.4 .. .. 0.688 .. 6.7 … .. 0.5

­40.4 3.2 18.2 .. 0.613 .. 5.9 2.5 1.3 ­2.7 ­0.5 ­1.4

0.8 2.2 11.7 17.9 0.331 0.3 6.0 1.5 8.1 ­2.2 ­0.5 ­0.5

­1.2 13.4 29.0 34.2 0.231 .. 8.5 0.2 29.7 0.1 ­1.2 ­0.9

20.8 q 2.1 26.4 .. 0.450 .. 7.0 … .. ..

­8.4 10.4 22.0 28.6 0.459 0.4 6.8 0.4 26.7 1.5 ­0.4 ­0.5

9.2 18.1 38.2 40.3 0.681 0.3 n 4.3 1.4 3.3 n 0.7 ­1.0 ­1.4

.. .. 15.1 .. 0.641 .. 3.5 0.2 … ..

23.1 5.2 32.8 17.5 0.216 .. 12.4 2.9 … ..

13.1 3.9 23.4 13.5 0.296 0.1 10.1 2.2 5.2 ­3.8 ­1.2 ..

­2.2 14.8 18.4 40.5 0.232 0.8 5.4 1.2 7.5 ­2.2 ­1.3 1.6

38.1 8.5 51.8 41.9 0.141 0.3 10.2 1.4 6.3 ­2.4 ­2.1 3.3

­16.3 13.4 24.1 10.2 0.933 .. 4.6 1.8 1.5 ­2.5 .. 4.5

4.5 10.7 22.4 19.8 0.336 .. 5.0 1.3 5.6 0.1 ­1.1 ­1.7

­22.2 8.4 12.6 13.0 0.325 .. 5.4 2.2 7.0 ­3.0 ­0.8 ..

12.7 22.8 16.4 27.9 0.204 0.2 8.3 4.0 1.5 ­0.2 ­0.7 ­0.2

.. 3.9 .. 18.7 0.676 .. 7.6 … .. 3.4

.. 3.1 z 27.8 x .. 0.235 0.0 9.4 4.1 2.2 y .. 0.0 ..

.. 27.1 .. 26.7 0.293 0.0 6.9 0.3 .. .. 0.7 ..

5.8 aa 1.9 17.5 .. 0.560 .. 6.3 .. .. 0.4 .. 2.1

­4.4 3.9 24.4 17.1 0.390 .. 7.3 1.2 8.0 ­2.8 ­1.2 2.1 DASH

1.4 6.8 15.5 35.2 0.783 0.2 x 5.2 0.5 .. ­2.1 .. ­1.1 BOARD

23.1 8.4 34.0 5.0 0.288 0.5 5.3 ad 1.2 7.3 ­1.5 ­0.7 0.2 5

­9.5 3.8 24.6 37.1 0.250 0.2 4.1 1.4 6.9 ­2.1 ­0.8 ­0.1

DASHBOARD 5 Socioeconomic sustainability | 345 DASHBOARD 5 SOCIOECONOMIC SUSTAINABILITY

SDG 17.4 SDG 9.5 SDG 10.1 SDG 5 SDG 10.1

Economic sustainability Social sustainability

Gross Skilled Concentration and ratio military expenditure capital labour Overall loss in Gender Income share Adjusted Total index development Old age Military HDI value due Inequality of the poorest net savings debt service (exports) expenditure (65 and older) expenditurea Ratio of to inequalityc Indexc 40 percent

(% of exports education services and health income) (per 100 expenditure (% of labour to military Average annual change (% of GNI) (value) (% of GDP) (% of GDP) expenditureb (%)

HDI rank 2015­2017d 2015­2017d 2015­2018d 2010­2018d 2018 2010­2017d 2030e 2010­2018d 2010­2016f 2010/2018g 2005/2018g 2005/2017

161 Mauritania ­10.3 13.2 55.3 5.8 0.308 .. 6.2 3.0 2.4 ­1.1 .. 1.5 163 Benin 7.7 3.2 15.2 18.5 0.213 0.0 6.4 0.6 12.1 ­1.4 .. ­1.5 165 Côte d’Ivoire ­3.4 4.2 25.8 17.1 0.346 .. 6.3 0.9 8.5 0.7 ­0.5 ­2.8 167 Togo 8.2 3.6 27.9 .. 0.288 0.0 8.7 1.8 13.2 n ­0.5 ­0.5 ­1.1 169 Haiti 16.6 17.6 19.8 25.5 0.361 .. 5.3 1.4 5.5 ­0.1 ­0.4 ­0.4 171 Djibouti 12.3 q 14.2 28.7 10.9 0.239 0.8 5.8 1.9 5.9 ­1.3 ­1.3 ­0.5 173 Ethiopia ­7.5 5.8 25.3 47.6 0.235 0.3 5.5 2.0 6.3 ­0.4 ­0.8 ­0.9 174 Guinea 0.2 4.2 19.3 22.8 0.440 .. 7.1 2.3 1.4 y .. ­1.2 .. 177 Yemen 17.6 1.5 29.0 9.4 0.508 .. 9.7 0.0 .. ­0.1 0.3 .. 179 Congo (Democratic Republic of the) 2.7 4.0 19.2 19.2 0.387 .. 5.1 1.0 15.1 .. ­1.1 .. 181 Sierra Leone ­1.8 11.1 57.8 .. 0.222 .. 9.4 3.7 n 3.2 x .. .. ­0.3 182 Eritrea ­16.7 5.7 13.4 17.6 0.558 .. 4.8 0.8 22.8 ­1.3 ­0.5 ­0.7 185 Burundi 9.3 20.8 34.1 6.8 0.288 0.6 6.4 0.6 12.4 ­2.2 ­1.3 ­2.2 187 Chad ­12.7 aa 16.9 17.0 12.3 0.449 0.1 4.8 1.1 4.9 ­0.6 ­0.4 2.9 189 Niger ­6.5 1.4 36.2 .. 0.493 .. 5.4 2.5 3.2 ­1.6 .. 2.4 ­99.0 3.5 13.0 21.1 0.394 .. 6.4 0.8 19.5 ­1.7 ­0.3 0.3 .. Monaco .. 14.6 .. 29.7 0.319 .. 5.4 4.0 2.5 n ­0.9 0.2 ­0.6 .. San Marino ­11.0 2.4 10.9 .. 0.875 .. 5.1 1.6 3.3 ­1.4 .. ­4.8 .. Tuvalu ­4.4 3.0 25.8 43.1 0.505 0.1 y 5.9 0.7 6.3 ­1.8 ­0.1 ­0.1 ­13.5 5.0 37.7 7.1 0.305 0.3 5.1 1.0 12.0 ­4.0 ­0.7 ­1.8 High human development ­33.5 3.8 18.5 15.2 0.255 .. 5.2 0.8 17.2 ­1.2 ­0.3 1.9 Low human development ­9.0 3.7 25.7 3.9 0.658 0.2 4.8 2.1 7.5 ­2.1 ­0.4 2.3 Regions .. .. 10.0 .. 0.319 .. 7.0 … .. .. East Asia and the Pacific ­2.3 q 4.5 23.8 4.7 0.670 0.3 4.5 2.9 2.7 ­2.3 ­0.3 2.4 Latin America and the Caribbean ­19.0 14.4 9.2 2.5 0.425 0.1 5.2 1.9 5.1 ­2.4 ­0.7 1.0 Sub-Saharan Africa .. .. 1.6 … 6.2 1.3 … .. Small island developing states .. .. 19.7 .. 0.774 0.3 4.7 2.1 1.4 ­0.5 .. ­1.7 Cooperation and Development .. .. 11.4 .. 0.313 .. 5.0 1.4 2.2 ­0.1 ­0.1 ­6.7 5.0 15.6 33.7 1.8 0.352 .. 5.2 2.5 4.6 ­2.2 ­0.6 2.6

… .. 0.255 .. 18.7 … .. ..

… 96.5 0.512 … … ..

… 55.7 … … .. ..

… .. 0.552 .. 5.6 … .. ..

… 50.1 0.554 … … ..

8.9 .. 22.1 84.7 — 2.3 33.2 2.3 7.0 ­1.1 ­2.4 —

16.2 12.9 36.5 .. — 1.5 20.4 1.7 .. ­2.5 ­1.2 —

13.2 10.0 28.1 21.6 — 0.5 11.4 2.3 3.3 ­3.9 ­1.2 —

2.7 8.9 21.9 22.2 — .. 5.7 1.0 4.0 ­1.7 ­0.6 —

14.9 13.7 33.5 32.5 — 1.3 14.7 2.1 4.5 ­2.8 ­0.9 —

10.4 16.8 27.0 41.1 — 0.6 9.7 5.5 1.7 ­1.3 ­1.0 —

19.7 9.0 41.6 .. — .. 21.7 1.8 .. ­3.0 ­0.8 —

9.7 31.8 28.1 71.8 — 0.6 20.1 2.4 4.6 ­3.5 ­2.1 —

6.8 24.0 20.1 54.6 — 0.7 17.8 1.2 10.8 ­1.4 ­1.1 —

17.1 10.7 30.3 20.0 — 0.5 11.9 2.5 3.0 ­4.5 ­1.2 —

­0.1 10.6 21.0 25.6 — 0.5 5.7 1.1 7.0 ­1.7 ­0.6 —

9.8 8.1 29.5 20.6 — .. 7.0 1.6 3.7 ­1.8 ­0.8 —

.. .. 24.0 44.3 — .. 17.1 .. .. ­2.1 — —

8.6 .. 21.9 81.9 — 2.4 34.1 2.1 7.8 ­0.6 ­2.3 —

10.9 14.8 26.2 46.3 — 2.0 18.0 2.2 6.7 ­2.6 ­0.8 —

5

346 | HUMAN DEVELOPMENT REPORT 2019

NOTES p Includes Sabah and Sarawak. roads, railways and the like, including schools, government agencies engaged in defence projects; q Refers to 2014. offices, hospitals, private residential dwellings and paramilitary forces, if these are judged to be trained Three-colour coding is used to visualize partial r Includes Kosovo. commercial and industrial buildings. Inventories are and equipped for military operations; and military grouping of countries and aggregates by indicator. s Includes only intermediate education. stocks of goods held by firms to meet temporary or space activities. For each indicator countries are divided into three t Includes Agalega, Rodrigues and Saint Brandon. unexpected fluctuations in production or sales as well groups of approximately equal size (terciles): the u Includes Abkhazia and South Ossetia. as goods that are work in progress. Net acquisitions Ratio of education and health expenditure top third, the middle third and the bottom third. v Includes Nagorno-Karabakh. of valuables are also considered capital formation. to military expenditure: Sum of government Aggregates are colour coded using the same tercile w Includes Crimea. Gross capital formation was formerly known as gross expenditure on education and health divided by cutoffs. See Technical note 6 at http://hdr.undp.org/ x Refers to 2007. domestic investment. military expenditure. sites/default/files/hdr2019_technical_notes.pdf for y Refers to 2009. details about partial grouping in this table. z Refers to 2010. Skilled labour force: Percentage of the labour Overall loss in HDI value due to inequality, aa Refers to 2012. force ages 15 and older with intermediate average annual change: Percentage change in a This column is intentionally left without colour ab Includes Transnistria. or advanced education, as classified by the overall loss in Human Development Index (HDI) value because it is meant to provide context for the ac Includes East Jerusalem. International Standard Classification of Education. due to inequality over 2010­2018, divided by the indicator on education and health expenditure. ad Includes Zanzibar. respective number of years. Concentration index (exports): A measure of b Data on government expenditure on health and DEFINITIONS the degree of product concentration in exports Gender Inequality Index, average annual education are available in tables 8 and 9 and at from a country (also referred to as the Herfindahl- change: Percentage change in Gender Inequality http://hdr.undp.org/en/data. Adjusted net savings: Net national savings plus Hirschmann Index). A value closer to 0 indicates Index value over 2005­2018, divided by the education expenditure and minus energy depletion, that a country’s exports are more homogeneously respective number of years. c A negative value indicates that inequality mineral depletion, net forest depletion, and carbon distributed among a series of products (reflecting declined over the period specified. dioxide and particulate emissions damage. Net a well diversified economy); a value closer to Income share of the poorest 40 percent, national savings are equal to gross national savings 1 indicates that a country’s exports are highly average annual change: Percentage change d Data refer to the most recent year available less the value of consumption of fixed capital. concentrated among a few products. of the income share of the poorest 40 percent of during the period specified. the population over 2005­2017, divided by the Total debt service: Sum of principal repayments Research and development expenditure: respective number of years. e Projections based on medium-fertility variant. and interest actually paid in currency, goods Current and capital expenditures (both public and or services on long-term debt; interest paid on private) on creative work undertaken systematically MAIN DATA SOURCES f Data refer to the most recent year for which all short-term debt; and repayments (repurchases and to increase knowledge, including knowledge of three types of expenditure (education, health charges) to the International Monetary Fund. It is humanity, culture and society, and the use of Columns 1­3, 6 and 8: World Bank (2019a). and military) are available during the period expressed as a percentage of exports of goods, knowledge for new applications. Research and specified. services and primary income. development covers basic research, applied research Column 4: ILO (2019). and experimental development. g The trend data used to calculate the change are Gross capital formation: Outlays on additions to Column 5: UNCTAD (2019). available at http://hdr.undp.org/en/data. the fixed assets of the economy plus net changes in Old-age dependency ratio: Ratio of the inventories. Fixed assets include land improvements population ages 65 and older to the population ages Column 7: UNDESA (2019b). h Includes Svalbard and Jan Mayen Islands. (such as fences, ditches and drains); plant, machinery 15­64, expressed as the number of dependants per and equipment purchases; and construction of 100 people of working age (ages 15­64). Columns 9 and 12: HDRO calculations based on i Includes Liechtenstein. data from World Bank (2019a). Military expenditures: All current and capital j Includes Christmas Island, Cocos (Keeling) expenditures on the armed forces, including Column 10: HDRO calculations based on the Islands and Norfolk Island. peacekeeping forces; defence ministries and other Inequality-adjusted HDI time series.

k Includes Åland Islands. Column 11: HDRO calculations based on the Gender Inequality Index time series. l Includes Canary Islands, Ceuta and Melilla.

m Includes Northern Cyprus.

n Refers to 2008.

o Refers to 2013.

5

DASHBOARD 5 Socioeconomic sustainability | 347 Developing regions

Arab States (20 countries or territories) Algeria, Bahrain, Djibouti, Egypt, Iraq, Jordan, Kuwait, Lebanon, Libya, Morocco, State of Palestine, Oman, Qatar, Saudi Arabia, Somalia, Sudan, Syrian Arab Republic, Tunisia, United Arab Emirates, Yemen East Asia and the Pacific (26 countries) Brunei Darussalam, Cambodia, China, Fiji, Indonesia, Kiribati, Democratic People’s Republic of Korea, Lao People’s Democratic Republic, Malaysia, Marshall Islands, Federated States of Micronesia, Mongolia, Myanmar, Nauru, Palau, Papua New Guinea, Philippines, Samoa, Singapore, Solomon Islands, Thailand, Timor-Leste, Tonga, Tuvalu, Vanuatu, Viet Nam Europe and Central Asia (17 countries) Albania, Armenia, Azerbaijan, Belarus, Bosnia and Herzegovina, Georgia, Kazakhstan, Kyrgyzstan, Republic of Moldova, Montenegro, North Macedonia, Serbia, Tajikistan, Turkey, Turkmenistan, Ukraine, Uzbekistan Latin America and the Caribbean (33 countries) Antigua and Barbuda, Argentina, Bahamas, Barbados, Belize, Plurinational State of Bolivia, Brazil, Chile, Colombia, Costa Rica, Cuba, Dominica, Dominican Republic, Ecuador, El Salvador, Grenada, Guatemala, Guyana, Haiti, Honduras, Jamaica, Mexico, Nicaragua, Panama, Paraguay, Peru, Saint Kitts and Nevis, Saint Lucia, Saint Vincent and the Grenadines, Suriname, Trinidad and Tobago, Uruguay, Bolivarian Republic of Venezuela South Asia (9 countries) Afghanistan, Bangladesh, Bhutan, India, Islamic Republic of Iran, Maldives, Nepal, Pakistan, Sri Lanka Sub-Saharan Africa (46 countries) Angola, Benin, Botswana, Burkina Faso, Burundi, Cameroon, Cabo Verde, Central African Republic, Chad, Comoros, Congo, Democratic Republic of the Congo, Côte d’Ivoire, Equatorial Guinea, Eritrea, Kingdom of Eswatini, Ethiopia, Gabon, Gambia, Ghana, Guinea, Guinea-Bissau, Kenya, Lesotho, Liberia, Madagascar, Malawi, Mali, Mauritania, Mauritius, Mozambique, Namibia, Niger, Nigeria, Rwanda, Sao Tomé and Príncipe, Senegal, Seychelles, Sierra Leone, South Africa, South Sudan, United Republic of Tanzania, Togo, Uganda, Zambia, Zimbabwe

Note: All countries listed in developing regions are included in aggregates for developing countries. Countries included in aggregates for Least Developed Countries and Small Island Developing States follow UN classifications, which are available at www.unohrlls.org. Countries included in aggregates for Organisation for Economic Co-operation and Development are listed at www.oecd.org/about/membersandpartners/list-oecd-member-countries.htm.

348 | HUMAN DEVELOPMENT REPORT 2019

Note: Statistical references relate to all statistical material in the 2019 Report, including the statistical tables posted at http://hdr.undp.org/en/human-development-report-2019.

Alkire, S., U. Kanagaratnam and N. Suppa. 2019. “The ------. 2018. Education at a Glance 2018. Paris. www. UNICEF (United Nations Children’s Fund). 2019a. UNICEF Global Multidimensional Poverty Index (MPI) 2019.” OPHI oecd-ilibrary.org/education/education-at-a-glance-2018_ Global Databases: Infant and Young Child Feeding: Exclu- MPI Methodological Note 47. University of Oxford, Oxford eag-2018-en. Accessed 15 June 2019. sive Breastfeeding, Predominant Breastfeeding. May Poverty and Human Development Initiative, Oxford, UK. 2019. New York. Palma, J. G. 2011. “Homogeneous Middles vs. Heterogene- Barro, R. J., and J.-W. Lee. 2018. Dataset of Education- ous Tails, and the End of the `Inverted-U’: The Share of the ------. 2019b. UNICEF Data. https://data.unicef.org. al Attainment, June 2018 Revision. www.barrolee.com. Rich is What It’s All About.” Cambridge Working Papers in Accessed 25 July 2019. Accessed 15 June 2019. Economics, 1111. Cambridge University, UK. www.econ .cam.ac.uk/research-files/repec/cam/pdf/cwpe1111.pdf. ------. Various years. Multiple Indicator Cluster Surveys. CEDLAS (Center for Distributive, Labor and Social Stud- Accessed 15 September 2013. New York. http://mics.unicef.org. Accessed 15 July 2019. ies) and World Bank. 2018. Socio-Economic Database for Latin America and the Caribbean. www.cedlas.econo.unlp. Syrian Center for Policy Research. 2017. Social Degrada- UNICEF (United Nations Children’s Fund), WHO (World edu.ar/wp/en/estadisticas/sedlac/estadisticas/. Accessed tion in Syria: The Conflict Impact on Social Capital. http:// Health Organization) and World Bank. 2019. Joint Child 15 July 2019. scpr-syria.org/publications/social-degradation-in-syria/. Malnutrition Estimates Expanded Database: Stunting. Accessed 15 July 2019. March 2019 Edition. New York. https://data.unicef.org/ CRED EM-DAT (Centre for Research on the Epidemiology topic/nutrition/malnutrition/. Accessed 26 July 2019. of Disasters). 2019. The International Disaster Database. UNCTAD (United Nations Conference on Trade and www.emdat.be. Accessed 25 June 2019. Development). 2019. Data Center. http://unctadstat. UN Inter-agency Group for Child Mortality Estimation. unctad.org. Accessed 15 August 2019. 2018. Child mortality estimates. www.childmortality.org. Eurostat. 2018. European Union Statistics on Income and Accessed 29 July 2019. Living Conditions. EUSILC UDB 2016—version 2 of UNDESA (United Nations Department of Economic and August 2016. Brussels. http://ec.europa.eu/eurostat/ Social Affairs). 2011. World Population Prospects: The United Nations Statistics Division. 2019a. Global SDG Indi- web/microdata/european-union-statistics-on-income-and- 2010 Revision. New York. www.un.org/en/d evelopment/ cators Database. https://unstats.un.org/sdgs/indicators/ living-conditions. Accessed 15 June 2019. desa/population/publications/trends/population- database/. Accessed 15 July 2019. prospects_2010_revision.shtml. Accessed 15 October FAO (Food and Agriculture Organization). 2019a. FAOSTAT 2013. ------. 2019b. National Accounts Main Aggregates Data- database. www.fao.org/faostat. Accessed 30 July 2019. base. http://unstats.un.org/unsd/snaama. Accessed 15 ------. 2017a. Trends in International Migrant Stock: The July 2019. ------. 2019b. AQUASTAT database. www.fao.org/ 2017 Revision. New York. www.un.org/en/development/ aquastat/en/. Accessed 2 July 2019. desa/population/migration/data/. Accessed 15 July 2019. UN Maternal Mortality Estimation Group (World Health Organization, United Nations Children’s Fund, Unit- Gallup. 2019. Gallup World Poll Analytics database. https:// ed Nations Population Fund and World Bank). 2017. ga.gallup.com. Accessed 7 May 2019. ------. 2017b. World Population Prospects: The 2017 Revi- Maternal mortality data. http://data.unicef.org/topic/ sion. New York. https://esa.un.org/unpd/wpp/. Accessed maternal-health/maternal-mortality/. Accessed 15 July ICF Macro. Various years. Demographic and Health Surveys. 30 April 2019. 2019. www.measuredhs.com. Accessed 15 July 2019. ------. 2018. World Urbanization Prospects: The 2018 Revi- UNODC (United Nations Office on Drugs and Crime). IDMC (Internal Displacement Monitoring Centre). 2019. sion. New York. https://esa.un.org/unpd/wup/. Accessed 2019. UNODC Statistics and Data. https://dataunodc Global Internal Displacement Database. www.internal- 23 July 2019. .un.org. Accessed 3 June 2019. displacement.org/database. Accessed 10 May 2019. ------. 2019a. World Contraceptive Use 2019. New UNOHCHR (United Nations Office of the High Commis- IHME (Institute for Health Metrics and Evaluation). 2018. York. www.un.org/en/development/desa/population/ sioner for Human Rights). 2019. Human rights treaties. Global Burden of Disease Collaborative Network. Glob- publications/dataset/contraception/wcu2019.asp. http://tbinternet.ohchr.org/_layouts/TreatyBodyExternal/ al Burden of Disease Study 2017 (GBD 2017) Disability- Accessed 3 May 2019. countries.aspx. Accessed 5 July 2019. Adjusted Life Years and Healthy Life Expectancy 1990­ 2017. Seattle, WA. http://ghdx.healthdata.org/record/ ------. 2019b. World Population Prospects: The 2019 Revi- UNRWA (United Nations Relief and Works Agency ihme-data/gbd-2017-dalys-and-hale-1990-2017. Accessed sion. New York. https://population.un.org/wpp/. Accessed for Palestine). 2019. “UNRWA in Figures 2018-2019.” 15 August 2019. 19 June 2019. Amman. www.unrwa.org/resources/about-unrwa/unrwa- figures-2018-2019. Accessed 25 June 2019. ILO (International Labour Organization). 2019. ILOSTAT UNECLAC (United Nations Economic Commission for database. www.ilo.org/ilostat. Accessed 17 June 2019. Latin America and the Caribbean). 2019. Preliminary UN Women (United Nations Entity for Gender Equality Overview of the Economies of Latin America and the and the Empowerment of Women). 2019. UN Women IMF (International Monetary Fund). 2019. World Eco- Caribbean 2018. Santiago. https://repositorio.cepal.org/ Global Database on Violence against Women. New York. nomic Outlook database. Washington, DC. www.imf. bitstream/handle/11362/44327/135/S1801218_en.pdf. http://evaw-global-database.unwomen.org. Accessed 19 org/e xternal/pubs/ft/weo/2019/01/weodata/index.aspx. Accessed 15 July 2019. April 2018. Accessed 15 July 2019. UNESCO (United Nations Educational, Scientific and WHO (World Health Organization). 2019. Global Health IPU (Inter-Parliamentary Union). 2019. Women in nation- Cultural Organization) Institute for Statistics. 2019. Observatory. www.who.int/gho/. Accessed 15 July 2019. al parliaments. www.ipu.org/wmn-e/classif-arc.htm. Data Centre. http://data.uis.unesco.org. Accessed 11 Accessed 11 April 2019. April 2019. WHO (World Health Organization) and UNICEF (United Nations Children’s Fund). 2019. Estimates of national ITU (International Telecommunication Union). 2019. ICT UNESCWA (United Nations Economic and Social Com- routine immunization coverage, 2018 revision (complet- Facts and Figures 2019. www.itu.int/en/ITU-D/Statistics/ mission for Western Asia). 2018. Survey of Economic ed July 2019). https://data.unicef.org/topic/child-health/ Pages/stat/. Accessed 8 August 2019. and Social Developments in the Arab Region 2017­2018. immunization/. Accessed 26 July 2019. Beirut. www.unescwa.org/publications/survey-e conomic- LIS (Luxembourg Income Study). 2019. Luxembourg Income social-development-arab-region-2017-2018. Accessed 15 World Bank. 2019a. World Development Indicators data- Study Project. www.lisdatacenter.org/data-access. July 2019. base. Washington, DC. http://data.worldbank.org. Accessed 19 August 2019. Accessed 15 July 2019. UNHCR (Office of the United Nations High Commissioner OECD (Organisation for Economic Co-operation and for Refugees). 2019. UNHCR Global Trends 2018. Gene- ------. 2019b. Gender Statistics database. Washington, DC. Development). 2017. PISA 2015 Results in Focus. Paris. va. www.unhcr.org/globaltrends2018/. Accessed 20 June http://data.worldbank.org. Accessed 3 July 2019. www.oecd.org/pisa/. Accessed 1 July 2019. 2019. World Inequality Database. 2019. World Inequality Data- base. http://wid.world. Accessed 15 August 2019.

Statistical references | 349 Human Development Reports 1990­2019

1990 Concept and Measurement of Human Development 1991 Financing Human Development 1992 Global Dimensions of Human Development 1993 People’s Participation 1994 New Dimensions of Human Security 1995 Gender and Human Development 1996 Economic Growth and Human Development 1997 Human Development to Eradicate Poverty 1998 Consumption for Human Development 1999 Globalization with a Human Face 2000 Human Rights and Human Development 2001 Making New Technologies Work for Human Development 2002 Deepening Democracy in a Fragmented World 2003 Millennium Development Goals: A Compact among Nations to End Human Poverty 2004 Cultural Liberty in Today’s Diverse World 2005 International Cooperation at a Crossroads: Aid, Trade and Security in an Unequal World 2006 Beyond Scarcity: Power, Poverty and the Global Water Crisis 2007/2008 Fighting Climate Change: Human Solidarity in a Divided World 2009 Overcoming Barriers: Human Mobility and Development 2010 The Real Wealth of Nations: Pathways to Human Development 2011 Sustainability and Equity: A Better Future for All 2013 The Rise of the South: Human Progress in a Diverse World 2014 Sustaining Human Progress: Reducing Vulnerability and Building Resilience 2015 Work for Human Development 2016 Human Development for Everyone 2019 Beyond Income, Beyond Averages, Beyond Today: Inequalities in Human Development in the 21st Century

350 | HUMAN DEVELOPMENT REPORT 2019 Key to HDI countries and ranks, 2018

Afghanistan170 Ghana142 Norway1

Albania69 Greece32 Oman47

Algeria82 Grenada78 Pakistan152

Andorra36 Guatemala126 Palau55

Angola149 Guinea174 Palestine, State of 119

Antigua and Barbuda 74 Guinea-Bissau178 Panama67

Argentina48 Guyana123 Papua New Guinea 155

Armenia81 Haiti169 Paraguay98

Australia6 Honduras132 Peru82

Austria20 Hong Kong, China (SAR) 4 Philippines106

Azerbaijan87 Hungary43 Poland32

Bahamas60 Iceland6 Portugal40

Bahrain45 India129 Qatar41

Bangladesh135 Indonesia111 Romania52

Barbados56 Iran (Islamic Republic of) 65 Russian Federation 49

Belarus50 Iraq120 Rwanda157

Belgium17 Ireland3 Saint Kitts and Nevis 73

Belize103 Israel22 Saint Lucia 89

Benin163 Italy29 Saint Vincent and the Grenadines 94

Bhutan134 Jamaica96 Samoa111

Bolivia (Plurinational State of) 114 Japan19 San Marino ..

Bosnia and Herzegovina 75 Jordan102 Sao Tome and Principe 137

Botswana94 Kazakhstan50 Saudi Arabia 36

Brazil79 Kenya147 Senegal166

Brunei Darussalam 43 Kiribati132 Serbia63

Bulgaria52 Korea (Democratic People’s Rep. of).. Seychelles62

Burkina Faso 182 Korea (Republic of) 22 Sierra Leone 181

Burundi185 Kuwait57 Singapore9

Cabo Verde 126 Kyrgyzstan122 Slovakia36

Cambodia146 Lao People’s Democratic Republic 140 Slovenia24

Cameroon150 Latvia39 Solomon Islands 153

Canada13 Lebanon93 Somalia..

Central African Republic 188 Lesotho164 South Africa 113

Chad187 Liberia176 South Sudan 186

Chile42 Libya110 Spain25

China85 Liechtenstein18 Sri Lanka 71

Colombia79 Lithuania34 Sudan168

Comoros156 Luxembourg21 Suriname98

Congo138 Madagascar162 Sweden8

Congo (Democratic Republic of the) 179 Malawi172 Switzerland2

Costa Rica 68 Malaysia61 Syrian Arab Republic 154

Côte d’Ivoire 165 Maldives104 Tajikistan125

Croatia46 Mali184 Tanzania (United Republic of) 159

Cuba72 Malta28 Thailand77

Cyprus31 Marshall Islands 117 Timor-Leste131

Czechia26 Mauritania161 Togo167

Denmark11 Mauritius66 Tonga105

Djibouti171 Mexico76 Trinidad and Tobago 63

Dominica98 Micronesia (Federated States of) 135 Tunisia91

Dominican Republic 89 Moldova (Republic of) 107 Turkey59

Ecuador85 Monaco.. Turkmenistan108

Egypt116 Mongolia92 Tuvalu..

El Salvador 124 Montenegro52 Uganda159

Equatorial Guinea 144 Morocco121 Ukraine88

Eritrea182 Mozambique180 United Arab Emirates 35

Estonia30 Myanmar145 United Kingdom 15

Eswatini (Kingdom of) 138 Namibia130 United States 15

Ethiopia173 Nauru.. Uruguay57

Fiji98 Nepal147 Uzbekistan108

Finland12 Netherlands10 Vanuatu141

France26 New Zealand 14 Venezuela (Bolivarian Republic of) 96

Gabon115 Nicaragua126 Viet Nam 118

Gambia174 Niger189 Yemen177

Georgia70 Nigeria158 Zambia143

Germany4 North Macedonia 82 Zimbabwe150 United Nations Development Programme ISBN: 978-92-1-126439-5 One United Nations Plaza New York, NY 10017

www.undp.org

In every country many people have little prospect for a they are as broad and multifaceted as life itself. In part better future. They are without hope, purpose or dignity, because the measures we rely on, and the data that underpin watching from society’s sidelines as they see others pulling them, are often inadequate. Yet important patterns repeat ahead to ever greater prosperity. Worldwide many have again and again. escaped extreme poverty. But even more have neither the opportunities nor the resources to control their lives. Far In every country the goalposts are moving. Inequality too often a person’s place in society is still determined by in human development is high or increasing in the areas ethnicity, gender or his or her parents’ wealth. expected to become more important in the future. There has been some progress worldwide in fundamental areas, such Inequalities. The evidence is everywhere. Inequalities do as escaping from poverty and receiving a basic education, not always reflect an unfair world, but when they have little though important gaps remain. Yet at the same time, to do with rewarding effort, talent or entrepreneurial risk- inequalities are widening higher up the ladder of progress. taking, they can be an affront to human dignity. Under the shadow of sweeping technological change and the climate A human development approach opens new windows crisis, such inequalities in human development hurt societies, on inequalities—why they matter, how they manifest weakening social cohesion and people’s trust in government, themselves and what to do about them—that help create institutions and each other. Most hurt economies, wastefully concrete action. The Report suggests the importance of preventing people from reaching their full potential at work realigning existing policy goals: emphasizing, for instance, and in life. They often make it harder for political decisions to the quality education at all ages—including at the preprimary reflect the aspirations of the whole of society and to protect level—in addition to focusing on primary and secondary the planet, if the few pulling ahead flex their power to shape enrolment rates. Many of these aspirations are already decisions in their interests. In the extreme, people can take reflected in the 2030 Agenda for Sustainable Development. It to the streets. also means addressing power imbalances that are at the heart of many inequalities, such as leveling the economic playing These inequalities in human development are a roadblock field through antitrust measures. In some cases, addressing to achieving the 2030 Agenda for Sustainable Development. inequalities means tackling social norms embedded deep They are not just about disparities in income and wealth. with a nation’s history and culture. Many policies comprise They cannot be accounted for simply by using summary options that would enhance both equity and efficiency. The measures of inequality that focus on a single dimension. And main reason why they often are not pursued may be linked they will shape the prospects of people that may live to see with the power of entrenched interests who do not stand to the 22nd century. This Report explores inequalities in human gain from change. development by going beyond income, beyond averages and beyond today. It asks what forms of inequality matter and The future of inequalities in human development in the 21st what drives them, recognizing that pernicious inequalities century is in our hands. But we cannot be complacent. The are generally better thought of as a symptom of broader climate crisis shows that the price of inaction compounds problems in a society and economy. It also asks what policies over time, as it feeds further inequality, which can in turn can tackle those drivers—policies that can simultaneously make action on climate more difficult. Technology is already help nations to grow their economies, improve human changing labour markets and lives, but not yet locked-in development and reduce inequality. is the extent to which machines may replace people. We are, however, approaching a precipice beyond which it will It is hard to get a clear picture of inequalities in human be difficult to recover. We do have a choice, and we must development and how they are changing. In part because exercise it now.

Documents associés

À lire aussi

• Lu pour vous

  Lu pour vous : Rapport 2020 des emplois du futur (Forum Economique Mondial)

  mai 2021·PDF

• Lu pour vous

  Lu pour vous : rapport du WEF de 2020 sur La transition énergétique dans le monde

  juin 2020·PDF

• Lu pour vous

  Lu pour vous : L’impact économique de Covid-19 en Tunisie

  juillet 2020·PDF

• Lu pour vous

  Lu pour vous : Doing Business 2019-2020

  décembre 2019·PDF

---