• Like
C  Documents And Settings Ib202643 ConfiguracióN Local Datos De Programa Mozilla Firefox Profiles R8tamevl
Upcoming SlideShare
Loading in...5

Thanks for flagging this SlideShare!

Oops! An error has occurred.

C Documents And Settings Ib202643 ConfiguracióN Local Datos De Programa Mozilla Firefox Profiles R8tamevl


Human Development Report 2009

Human Development Report 2009

Published in News & Politics
  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
  • That was quite a bit of information to assimilate. Good info, just a lot of it. http://www.tinasemporium.com
    Are you sure you want to
    Your message goes here
    Be the first to like this
No Downloads


Total Views
On SlideShare
From Embeds
Number of Embeds



Embeds 0

No embeds

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

    No notes for slide


  • 1. Human Development Report 2009 Overcoming barriers: Human mobility and development Published for the United Nations Development Programme (UNDP)
  • 2. Copyright © 2009 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 any means, electronic, mechanical, photocopying, recording or otherwise, without prior permission. ISBN 978-0-230-23904-3 First published in 2009 by Palgrave Macmillan Houndmills, Basingstoke, Hampshire RG21 6XS and 175 Fifth Avenue, New York, NY 10010 Companies and representatives throughout the world Palgrave Macmillan in the UK is an imprint of Macmillan Publishers Limited, registered in England, company number 785998, of Houndmills, Basingstoke, Hampshire RG21 6XS. Palgrave Macmillan in the US is a division of St Martin’s Press LLC, 175 Fifth Avenue, New York, NY 10010. Palgrave Macmillan is the global academic imprint of the above companies and has companies and representatives throughout the world. Palgrave® and Macmillan® are registered trademarks in the United States, the United Kingdom, Europe and other countries. A catalogue record for this book is available from the British Library and the Library of Congress. Printed in Canada by the Lowe–Martin Group. Cover is printed on FSC 10pt Kallima coated one-side paper. Text pages are printed on FSC 60# Rolland Opaque30 smooth text that is 30% post-consumer waste. FSC supports responsible forest management worldwide. Printed with vegetable-based inks and produced by means of environmentally-compatible technology. [FSC LOGO WILL BE INSERTED HERE] Editing: Green Ink Design and Layout: ZAGO For a list of any errors or omissions found subsequent to printing, please visit our website at http://hdr.undp.org
  • 3. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Team Team for the preparation of the Human Development Report 2009 Director Jeni Klugman Research Led by Francisco R. Rodríguez, comprising Ginette Azcona, Matthew Cummins, Ricardo Fuentes Nieva, Mamaye Gebretsadik, Wei Ha, Marieke Kleemans, Emmanuel Letouzé, Roshni Menon, Daniel Ortega, Isabel Medalho Pereira, Mark Purser and Cecilia Ugaz (Deputy Director until October 2008). Statistics Led by Alison Kennedy, comprising Liliana Carvajal, Amie Gaye, Shreyasi Jha, Papa Seck and Andrew Thornton. National HDR and network Eva Jespersen (Deputy Director HDRO), Mary Ann Mwangi, Paola Pagliani and Timothy Scott. Outreach and communications Led by Marisol Sanjines, comprising Wynne Boelt, Jean-Yves Hamel, Melissa Hernandez, Pedro Manuel Moreno and Yolanda Polo. Production, translation, budget and operations, administration Carlotta Aiello (production coordinator), Sarantuya Mend (operations manager), Fe Juarez-Shanahan and Oscar Bernal. iii
  • 4. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Foreword Foreword Migration not infrequently gets a bad press. Negative stereotypes portraying migrants as ‘stealing our jobs’ or ‘scrounging off the taxpayer’ abound in sections of the media and public opinion, es- pecially in times of recession. For others, the word ‘migrant’ may evoke images of people at their most vulnerable. This year’s Human Development Report, Overcoming Barriers: Human Mobility and Development, challenges such stereotypes. It seeks to broaden and rebalance perceptions of migration to reflect a more complex and highly variable reality. This report breaks new ground in applying a The report suggests that the policy response human development approach to the study of to migration can be wanting. Many govern- migration. It discusses who migrants are, where ments institute increasingly repressive entry they come from and go to, and why they move. It regimes, turn a blind eye to health and safety looks at the multiple impacts of migration for all violations by employers, or fail to take a lead who are affected by it—not just those who move, in educating the public on the benefits of but also those who stay. immigration. In so doing, the report’s findings cast new By examining policies with a view to ex- light on some common misconceptions. For ex- panding people’s freedoms rather than con- ample, migration from developing to developed trolling or restricting human movement, this countries accounts for only a minor fraction of report proposes a bold set of reforms. It argues human movement. Migration from one develop- that, when tailored to country-specific contexts, ing economy to another is much more common. these changes can amplify human mobility’s Most migrants do not go abroad at all, but in- already substantial contributions to human stead move within their own country. development. Next, the majority of migrants, far from The principal reforms proposed centre being victims, tend to be successful, both before around six areas, each of which has important they leave their original home and on arrival and complementary contributions to make to in their new one. Outcomes in all aspects of human development: opening up existing entry human development, not only income but also channels so that more workers can emigrate; education and health, are for the most part posi- ensuring basic rights for migrants; lowering the tive—some immensely so, with people from the transaction costs of migration; finding solutions poorest places gaining the most. that benefit both destination communities and Reviewing an extensive literature, the report the migrants they receive; making it easier for finds that fears about migrants taking the jobs people to move within their own countries; and or lowering the wages of local people, placing an mainstreaming migration into national develop- unwelcome burden on local services, or costing ment strategies. the taxpayer money, are generally exaggerated. The report argues that while many of these When migrants’ skills complement those of local reforms are more feasible than at first thought, people, both groups benefit. Societies as a whole they nonetheless require political courage. There may also benefit in many ways—ranging from ris- may also be limits to governments’ ability to ing levels of technical innovation to increasingly make swift policy changes while the recession diverse cuisine to which migrants contribute. persists. v
  • 5. Foreword HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development This is the first Human Development discourse on migration and inform the work of Report for which as Administrator I am writ- development practitioners and policy makers ing the foreword. Like all such reports, this is around the world. an independent study intended to stimulate debate and discussion on an important issue. It is not a statement of either United Nations or UNDP policy. At the same time, by highlighting human mobility as a core component of the human Helen Clark development agenda, it is UNDP’s hope that Administrator the following insights will add value to ongoing United Nations Development Programme The analysis and policy recommendations of this report do not necessarily reflect the views of the United Nations Development Programme, its Executive Board or its Member States. The report is an independent publication commissioned by UNDP. It is the fruit of a collaborative effort by a team of eminent advisers and the Human Development Report team. Jeni Klugman, Director of the Human Development Report Office, led the effort. vi
  • 6. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Acknowledgements Acknowledgements This report is the fruit of the efforts, contribu- Statistics; UNHCR; Treaty Section, United tions and support of many people and organiza- Nations Office of Legal Affairs; UNRWA; the tions. I would like to thank Kemal Derviş for World Bank; and WHO. the opportunity to take on the daunting task of The report benefited greatly from intel- Director of the Human Development Report, lectual advice and guidance provided by an and the new UNDP Administrator, Helen academic advisory panel. The panel comprised Clark, for advice and support. Coming back to Maruja Asis, Richard Black, Caroline Brettell, the office after its 20 years of growth and success Stephen Castles, Simon Commander, Jeff Crisp, has been a tremendously rewarding experience, Priya Deshingkar, Cai Fang, Elizabeth Ferris, and I would like to especially thank my fam- Bill Frelick, Sergei Guriev, Gordon Hanson, ily, Ema, Josh and Billy, for their patience and Ricardo Hausmann, Michele Klein-Solomon, support throughout. The dedication and hard Kishore Mahbubani, Andrew Norman Mold, work of the whole HDR team, listed earlier, was Kathleen Newland, Yaw Nyarko, José Antonio critical. Among those who provided important Ocampo, Gustav Ranis, Bonaventure Rutinwa, strategic advice and suggestions, which were es- Javier Santiso, Maurice Schiff, Frances Stewart, pecially critical in pulling the report together, Elizabeth Thomas-Hope, Jeffrey Williamson, were Oliver Bakewell, Martin Bell, Stephen Castles, Ngaire Woods and Hania Zlotnik. Joseph Chamie, Samuel Choritz, Michael Clemens, From the outset, the process involved a Simon Commander, Sakiko Fukuda-Parr, Hein range of participatory consultations designed de Haas, Frank Laczko, Loren Landau, Manjula to draw on the expertise of researchers, civil Luthria, Gregory Maniatis, Philip Martin, Douglas society advocates, development practitioners Massey, Saraswathi Menon, Frances Stewart, and policy makers from around the globe. This Michael Walton and Kevin Watkins. included 11 informal stakeholder consultations Background studies were commissioned on a held between August 2008 and April 2009 range of thematic issues and published online in in Nairobi, New Delhi, Amman, Bratislava, our Human Development Research Papers series, Manila, Sydney, Dakar, Rio de Janeiro, Geneva, launched in April 2009, and are listed in the bib- Turin and Johannesburg, involving almost 300 liography. A series of 27 seminars that were held experts and practitioners in total. The support between August 2008 and April 2009 likewise of UNDP country and regional offices and provided important stimulus to our thinking and local partners was critical in enabling these the development of ideas, and we would again consultations. Several events were hosted by thank those presenters for sharing their research key partners, including the IOM, the ILO and and insights. We would also like to acknowledge the Migration Policy Institute. Additional aca- the contribution of the national experts who par- demic consultations took place in Washington ticipated in our migration policy assessment. D.C. and Princeton, and HDRO staff partici- The data and statistics used in this report pated in various other regional and global fora, draw significantly upon the databases of other including the Global Forum on Migration and organizations to which we were allowed gener- Development (GFMD) in Manila, preparatory ous access: Andean Development Corporation; meetings for the Athens GFMD, and many con- Development Research Centre on Migration, ferences and seminars organized by other UN University of Sussex; ECLAC; International agencies (e.g. ILO, UNDESA and UNITAR), Migration Institute, Oxford; Inter-Parliamentary universities, think-tanks and non-governmental Union; Internal Displacement Monitoring organizations. Participants in a series of Human Centre; the Department of Statistics and the Development Network discussions provided International Migration Programme of the wide-ranging insights and observations on the ILO; IOM; Luxembourg Income Study; OECD; linkages between migration and human devel- UNICEF; UNDESA, Statistics Division and opment. More details on the process are avail- Population Division; UNESCO Institute for able at http://hdr.undp.org/en/nhdr. vii
  • 7. Acknowledgements HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development A UNDP Readers Group, comprising repre- Luc Gregoire, Madi Musa, Uladzimir Shcherbau sentatives of all the regional and policy bureaux, and Oscar Yujnovsky. Margaret Chi and provided many useful inputs and suggestions on Solaiman Al-Rifai of the United Nations Office the concept note and report drafts, as did a num- for Project Services provided critical administra- ber of other colleagues who provided inputs and tive support and management services. advice. We would especially thank Amat Alsoswa, The report also benefited from the dedicated Carolina Azevedo, Barbara Barungi, Tony work of a number of interns, namely Shreya Bislimi, Kim Bolduc, Winifred Byanyima, Ajay Basu, Vanessa Alicia Chee, Delphine De Quina, Chhibber, Samuel Choritz, Pedro Conceição, Rebecca Lee Funk, Chloe Yuk Ting Heung, Awa Dabo, Georgina Fekete, Priya Gajraj, Enrique Abid Raza Khan, Alastair Mackay, Grace Parker, Ganuza, Tegegnework Gettu, Rebeca Grynspan, Clare Potter, Limon B. Rodriguez, Nicolas Roy, Sultan Hajiyev, Mona Hammam, Mette Bloch Kristina Shapiro and David Stubbs. Hansen, Mari Huseby, Selim Jahan, Bruce Jenks, We thank all of those involved directly or Arun Kashyap, Olav Kjoren, Paul Ladd, Luis indirectly in guiding our efforts, while acknowl- Felipe López-Calva, Tanni Mukhopadhyay, B. edging sole responsibility for errors of commis- Murali, Theodore Murphy, Mihail Peleah, Amin sion and omission. Sharkawi, Kori Udovicki, Mourad Wahba and Caitlin Wiesen for comments. A team at Green Ink, led by Simon Chater, provided editing services. The design work was carried out by Zago. Guoping Huang developed some of the maps. The production, translation, distribution and promotion of the report ben- efited from the help and support of the UNDP Jeni Klugman Office of Communications, and particularly of Director Maureen Lynch. Translations were reviewed by Human Development Report 2009 viii
  • 8. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Acronymes Acronyms Acronyms CEDAW United Nations Convention on the Elimination of All Forms of Discrimination against Women CMW United Nations International Convention on the Protection of the Rights of All Migrant Workers and Members of their Families CRC United Nations Convention on the Rights of the Child ECD Early childhood development ECLAC Economic Commission for Latin America and the Caribbean ECOWAS Economic Community of West African States EIU Economist Intelligence Unit EU European Union GATS General Agreement on Trade in Services GDP Gross domestic product GCC Gulf Cooperation Council HDI Human Development Index HDR Human Development Report HDRO Human Development Report Office ILO International Labour Organization IOM International Organization for Migration MERCOSUR Mercado Común del Sur (Southern Common Market) MIPEX Migrant Integration Policy Index NGO Non-governmental organization OECD Organisation for Economic Co-operation and Development PRS Poverty Reduction Strategy PRSP Poverty Reduction Strategy Paper TMBs Treaty Monitoring Bodies UNDESA United Nations Department of Economic and Social Affairs UNDP United Nations Development Programme UNESCO United Nations Educational, Scientific and Cultural Organization UNHCR Office of the United Nations High Commissioner for Refugees UNICEF United Nations Children’s Fund UNODC United Nations Office on Drugs and Crime UNRWA United Nations Relief and Works Agency for Palestine Refugees in the Near East USSR Union of Soviet Socialist Republics WHO World Health Organization WTO World Trade Organization ix
  • 9. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Contents Contents Foreword v CHAPTER 4 Acknowledgements vii Acronyms ix Impacts at origin and destination 71 4.1 Impacts at places of origin 71 OVERVIEW 1 4.1.1 Household level effects 71 4.1.2 Community and national level economic effects 76 How and why people move 1 4.1.3 Social and cultural effects 79 Barriers to movement 2 4.1.4 Mobility and national development strategies 82 The case for mobility 3 4.2 Destination place effects 83 Our proposal 3 4.2.1 Aggregate economic impacts 84 The way forward 5 4.2.2 Labour market impacts 85 4.2.3 Rapid urbanization 86 CHAPTER 1 4.2.4 Fiscal impacts 87 Freedom and movement: how mobility can foster human development 9 4.2.5 Perceptions and concerns about migration 89 1.1 Mobility matters 9 4.3 Conclusions 92 1.2 Choice and context: understanding why people move 11 1.3 Development, freedom and human mobility 14 CHAPTER 5 1.4 What we bring to the table 16 Policies to enhance human development outcomes 95 5.1 The core package 96 CHAPTER 2 5.1.1 Liberalizing and simplifying regular channels 96 People in motion: who moves where, when and why 21 5.1.2 Ensuring basic rights for migrants 99 2.1 Human movement today 21 5.1.3 Reducing transaction costs associated with movement 102 2.2 Looking back 28 5.1.4 Improving outcomes for migrants and destination communities 104 2.2.1 The long-term view 28 5.1.5 Enabling benefits from internal mobility 106 2.2.2 The 20th century 30 5.1.6 Making mobility an integral part of national development strategies 108 2.3 Policies and movement 33 5.2 The political feasibility of reform 108 2.4 Looking ahead: the crisis and beyond 40 5.3 Conclusions 112 2.4.1 The economic crisis and the prospects for recovery 41 2.4.2 Demographic trends 43 Notes 113 2.4.3 Environmental factors 45 Bibliography 119 2.5 Conclusions 46 STATISTICAL ANNEX CHAPTER 3 Tables 143 Reader’s guide 203 How movers fare 49 Technical note 208 3.1 Incomes and livelihoods 49 Definition of statistical terms and indicators 209 3.1.1 Impacts on gross income 50 Country classification 213 3.1.2 Financial costs of moving 53 3.2 Health 55 3.3 Education 57 3.4 Empowerment, civic rights and participation 60 3.5 Understanding outcomes from negative drivers 62 3.5.1 When insecurity drives movement 62 3.5.2 Development-induced displacement 64 3.5.3 Human trafficking 65 3.6 Overall impacts 67 3.7 Conclusions 68 xi
  • 10. Contents HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development BOXES 2.13 Working-age population will increase in developing regions: Projections of working-age population by region, 2010–2050 44 1.1 Estimating the impact of movement 12 3.1 Movers have much higher incomes than stayers: 1.2 How movement matters to the measurement of progress 14 Annual income of migrants in OECD destination countries and 1.3 Basic terms used in this report 15 GDP per capita in origin countries, by origin country HDI category 50 1.4 How do the poor view migration? 16 3.2 Huge salary gains for high-skilled movers: Gaps in average 2.1 Counting irregular migrants 23 professional salaries for selected country pairs, 2002–2006 50 2.2 Conflict-induced movement and trafficking 26 3.3 Significant wage gains to internal movers in Bolivia, especially the 2.3 Migration trends in the former Soviet Union 31 less well educated: Ratio of destination to origin wages for internal 2.4 Global governance of mobility 39 migrants in Bolivia, 2000 51 3.1 China: Policies and outcomes associated with internal migration 52 3.4 Poverty is higher among migrant children, but social transfers can help: 3.2 Independent child migrants 59 Effects of transfers on child poverty in selected countries, 1999–2001 53 3.3 The next generation 60 3.5 Costs of moving are often high: Costs of intermediaries in selected 3.4 Enforcement mechanisms in Malaysia 62 corridors against income per capita, 2006–2008 54 4.1 How cell-phones can reduce money transfer costs: the case of Kenya 74 3.6 Moving costs can be many times expected monthly earnings: 4.2 The 2009 crisis and remittances 75 Costs of movement against expected salary of low-skilled 4.3 Impacts of skills flows on human development 77 Indonesian workers in selected destinations, 2008 54 4.4 Mobility and the development prospects of small states 80 3.7 The children of movers have a much greater chance of surviving: 4.5 Mobility and human development: some developing country perspectives 82 Child mortality at origin versus destination by origin country 5.1 Opening up regular channels—Sweden and New Zealand 97 HDI category, 2000 census or latest round 55 5.2 Experience with regularization 98 3.8 Temporary and irregular migrants often lack access to health care 5.3 Reducing paperwork: a challenge for governments and partners 103 services: Access to health care by migrant status in developed 5.4 Recognition of credentials 105 versus developing countries, 2009 57 5.5 When skilled people emigrate: some policy options 109 3.9 Gains in schooling are greatest for migrants from low-HDI countries: Gross total enrolment ratio at origin versus destination by origin country HDI category, 2000 census or latest round 58 FIGURES 3.10 Migrants have better access to education in developed countries: Access to public schooling by migrant status in developed versus 2.1 Many more people move within borders than across them: developing countries, 2009 58 Internal movement and emigration rates, 2000–2002 22 3.11 Voting rights are generally reserved for citizens: Voting rights in local 2.2 The poorest have the most to gain from moving… elections by migrant status in developed versus developing countries, Differences between destination and origin country HDI, 2000–2002 23 2009 61 2.3 … but they also move less: Emigration rates by HDI and income 25 3.12 School enrolment among refugees often exceeds that of host 2.4 An increasing share of migrants come from developing countries: Share communities in developing countries: Gross primary enrolment ratios— of migrants from developing countries in selected developed countries 32 refugees, host populations and main countries of origin, 2007 64 2.5 Sources and trends of migration into developing countries: Migrants as a 3.13 Significant human development gains to internal movers: share of total population in selected countries, 1960–2000s 33 Ratio of migrants’ to non-migrants’ estimated HDI in selected 2.6 Internal migration rates have increased only slightly: Trends in lifetime developing countries, 1995–2005 67 internal migration intensity in selected countries, 1960–2000s 34 3.14 Migrants are generally as happy as locally-born people: Self-reported 2.7 Global income gaps have widened: Trends in real per capita GDP, happiness among migrants and locally-born people around the world, 1960–2007 35 2005/2006 68 2.8 Welcome the high-skilled, rotate the low-skilled: Openness to legal 4.1 The global recession is expected to impact remittance flows: Projected immigration in developed versus developing countries, 2009 36 trends in remittance flows to developing regions, 2006–2011 75 2.9 Enforcement practices vary: Interventions and procedures regarding 4.2 Skilled workers move similarly across and within nations: Population irregular migrants, 2009 37 and share of skilled workers who migrate internally and internationally 78 2.10 Cross-country evidence shows little support for the ‘numbers versus 4.3 Support for immigration is contingent on job availability: rights’ hypothesis: Correlations between access and treatment 38 Attitudes towards immigration and availability of jobs, 2005/2006 90 2.11 Unemployment is increasing in key migrant destinations: 4.4 When jobs are limited, people favour the locally born: Public opinion Unemployment rates in selected destinations, 2007–2010 41 about job preferences by destination country HDI category, 2005/2006 91 2.12 Migrants are in places hardest hit by the recession: Immigrants’ location 4.5 Many people value ethnic diversity: Popular views about the value and projected GDP growth rates, 2009 42 of ethnic diversity by destination country HDI category, 2005/2006 92 xii
  • 11. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Contents 5.1 Ratification of migrants’ rights convention has been limited: STATISTICAL ANNEX TABLES Ratification of selected agreements by HDI category, as of 2009 100 5.2 Support for opportunity to stay permanently: A Human movement: snapshots and trends 143 Preferences for temporary versus permanent migration, 2008 110 B International emigrants by area of residence 147 C Education and employment of international migrants in OECD countries (aged 15 years and above) 151 MAPS D Conflict and insecurity-induced movement 155 E International financial flows: remittances, official development 1.1 Borders matter: HDI in United States and Mexican border localities, 2000 10 assistance and foreign direct investment 159 1.2 Migrants are moving to places with greater opportunities: Human F Selected conventions related to human rights and migration development and inter-provincial migration flows in China, 1995–2000 11 (by year of ratification) 163 2.1 Most movement occurs within regions: Origin and destination of G Human development index trends 167 international migrants, circa 2000 24 H Human development index 2007 and its components 171 3.1 Conflict as a driver of movement in Africa: Conflict, instability and I1 Human and income poverty 176 population movement in Africa 63 I2 Human and income poverty: OECD countries 180 4.1 Remittances flow primarily from developed to developing regions: J Gender-related development index and its components 181 Flows of international remittances, 2006–2007 73 K Gender empowerment measure and its components 186 L Demographic trends 191 M Economy and inequality 195 TABLES N Health and education 199 2.1 Five decades of aggregate stability, with regional shifts: Regional distribution of international migrants, 1960–2010 30 2.2 Policy makers say they are trying to maintain existing immigration levels: Views and policies towards immigration by HDI category, 2007 34 2.3 Over a third of countries significantly restrict the right to move: Restrictions on internal movement and emigration by HDI category 40 2.4 Dependency ratios to rise in developed countries and remain steady in developing countries: Dependency ratio forecasts of developed versus developing countries, 2010–2050 45 4.1 PRSs recognize the multiple impacts of migration: Policy measures aimed at international migration in PRSs, 2000–2008 83 xiii
  • 12. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Overview Overview Consider Juan. Born into a poor family in rural Mexico, his family struggled to pay for his health care and education. At the age of 12, he dropped out of school to help support his family. Six years later, Juan followed his uncle to Canada in pursuit of higher wages and better opportunities. Life expectancy in Canada is five years higher When people move they embark on a journey than in Mexico and incomes are three times of hope and uncertainty whether within or across greater. Juan was selected to work temporarily international borders. Most people move in search in Canada, earned the right to stay and eventu- of better opportunities, hoping to combine their ally became an entrepreneur whose business now own talents with resources in the destination employs native-born Canadians. This is just one country so as to benefit themselves and their im- case out of millions of people every year who find mediate family, who often accompany or follow new opportunities and freedoms by migrating, them. If they succeed, their initiative and efforts benefiting themselves as well as their areas of ori- can also benefit those left behind and the society gin and destination. in which they make their new home. But not all Now consider Bhagyawati. She is a mem- do succeed. Migrants who leave friends and family ber of a lower caste and lives in rural Andhra may face loneliness, may feel unwelcome among Pradesh, India. She travels to Bangalore city people who fear or resent newcomers, may lose with her children to work on construction their jobs or fall ill and thus be unable to access sites for six months each year, earning Rs 60 the support services they need in order to prosper. (US$1.20) per day. While away from home, The 2009 HDR explores how better poli- her children do not attend school because it is cies towards human mobility can enhance too far from the construction site and they do human development. It lays out the case for not know the local language. Bhagyawati is not governments to reduce restrictions on move- entitled to subsidized food or health care, nor ment within and across their borders, so as to does she vote, because she is living outside her expand human choices and freedoms. It argues registered district. Like millions of other inter- for practical measures that can improve pros- nal migrants, she has few options for improving pects on arrival, which in turn will have large her life other than to move to a different city in benefits both for destination communities and search of better opportunities. for places of origin. Our world is very unequal. The huge differ- ences in human development across and within How and why people move countries have been a recurring theme of the Discussions about migration typically start from Human Development Report (HDR) since the perspective of flows from developing coun- it was first published in 1990. In this year’s re- tries into the rich countries of Europe, North port, we explore for the first time the topic of America and Australasia. Yet most movement in migration. For many people in developing the world does not take place between develop- countries moving away from their home town ing and developed countries; it does not even take or village can be the best—sometimes the place between countries. The overwhelming ma- only—option open to improve their life chances. jority of people who move do so inside their own Human mobility can be hugely effective in rais- country. Using a conservative definition, we esti- ing a person’s income, health and education mate that approximately 740 million people are prospects. But its value is more than that: being internal migrants—almost four times as many as able to decide where to live is a key element of those who have moved internationally. Among human freedom. people who have moved across national borders, 1
  • 13. Overview HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Most migrants, internal just over a third moved from a developing to a de- Barriers to movement and international, reap veloped country—fewer than 70 million people. The share of international migrants in the gains in the form of Most of the world’s 200 million international world’s population has remained remark- higher incomes, better migrants moved from one developing country to ably stable at around 3 percent over the past access to education another or between developed countries. 50 years, despite factors that could have been and health, and Most migrants, internal and international, expected to increase f lows. Demographic improved prospects for reap gains in the form of higher incomes, bet- trends—an aging population in developed their children ter access to education and health, and improved countries and young, still-rising populations in prospects for their children. Surveys of migrants developing countries—and growing employ- report that most are happy in their destination, ment opportunities, combined with cheaper despite the range of adjustments and obstacles communications and transport, have increased typically involved in moving. Once established, the ‘demand’ for migration. However, those migrants are often more likely than local resi- wishing to migrate have increasingly come up dents to join unions or religious and other against government-imposed barriers to move- groups. Yet there are trade-offs and the gains ment. Over the past century, the number of from mobility are unequally distributed. nation states has quadrupled to almost 200, People displaced by insecurity and conflict creating more borders to cross, while policy face special challenges. There are an estimated changes have further limited the scale of mi- 14 million refugees living outside their country gration even as barriers to trade fell. of citizenship, representing about 7 percent of the Barriers to mobility are especially high for world’s migrants. Most remain near the country people with low skills, despite the demand for they fled, typically living in camps until condi- their labour in many rich countries. Policies tions at home allow their return, but around half generally favour the admission of the better a million per year travel to developed countries educated, for instance by allowing students to and seek asylum there. A much larger number, stay after graduation and inviting professionals some 26 million, have been internally displaced. to settle with their families. But governments They have crossed no frontiers, but may face spe- tend to be far more ambivalent with respect to cial difficulties away from home in a country riven low-skilled workers, whose status and treatment by conflict or racked by natural disasters. Another often leave much to be desired. In many coun- vulnerable group consists of people—mainly tries, agriculture, construction, manufacturing young women—who have been trafficked. Often and service sectors have jobs that are filled by duped with promises of a better life, their move- such migrants. Yet governments often try to ro- ment is not one of free will but of duress, some- tate less educated people in and out of the coun- times accompanied by violence and sexual abuse. try, sometimes treating temporary and irregular In general, however, people move of their workers like water from a tap that can be turned own volition, to better-off places. More than on and off at will. An estimated 50 million peo- three quarters of international migrants go to a ple today are living and working abroad with ir- country with a higher level of human develop- regular status. Some countries, such as Thailand ment than their country of origin. Yet they are and the United States, tolerate large numbers significantly constrained, both by policies that of unauthorized workers. This may allow those impose barriers to entry and by the resources individuals to access better paying jobs than at they have available to enable their move. People home, but although they often do the same work in poor countries are the least mobile: for exam- and pay the same taxes as local residents, they ple, fewer than 1 percent of Africans have moved may lack access to basic services and face the risk to Europe. Indeed, history and contemporary of being deported. Some governments, such as evidence suggest that development and migra- those of Italy and Spain, have recognized that tion go hand in hand: the median emigration unskilled migrants contribute to their societies rate in a country with low human development and have regularized the status of those in work, is below 4 percent, compared to more than 8 per- while other countries, such as Canada and New cent from countries with high levels of human Zealand, have well designed seasonal migrant development. programmes for sectors such as agriculture. 2
  • 14. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Overview While there is broad consensus about the through flows of money, knowledge and ideas. Large gains to value of skilled migration to destination coun- Because migrants tend to come in large num- human development tries, low-skilled migrant workers generate much bers from specific places—for example, Kerala can be achieved controversy. It is widely believed that, while in India or Fujian Province in China—commu- by lowering the these migrants fill vacant jobs, they also displace nity-level effects can typically be larger than na- barriers to movement local workers and reduce wages. Other concerns tional ones. However, over the longer term, the and improving the posed by migrant inflows include heightened flow of ideas from human movement can have treatment of movers risk of crime, added burdens on local services far-reaching effects on social norms and class and the fear of losing social and cultural cohe- structures across a whole country. The outflow sion. But these concerns are often exaggerated. of skills is sometimes seen as negative, particu- While research has found that migration can, in larly for the delivery of services such as educa- certain circumstances, have negative effects on tion or health. Yet, even when this is the case, the locally born workers with comparable skills, the best response is policies that address underlying body of evidence suggests that these effects are structural problems, such as low pay, inadequate generally small and may, in some contexts, be financing and weak institutions. Blaming the entirely absent. loss of skilled workers on the workers themselves largely misses the point, and restraints on their The case for mobility mobility are likely to be counter-productive— This report argues that migrants boost eco- not to mention the fact that they deny the basic nomic output, at little or no cost to locals. human right to leave one’s own country. Indeed, there may be broader positive effects, for However, international migration, even if instance when the availability of migrants for well managed, does not amount to a national childcare allows resident mothers to work out- human development strategy. With few excep- side the home. As migrants acquire the language tions (mainly small island states where more and other skills needed to move up the income than 40 percent of inhabitants move abroad), ladder, many integrate quite naturally, making emigration is unlikely to shape the development fears about inassimilable foreigners—similar prospects of an entire nation. Migration is at best to those expressed early in the 20th century in an avenue that complements broader local and America about the Irish, for example—seem national efforts to reduce poverty and improve equally unwarranted with respect to newcom- human development. These efforts remain as ers today. Yet it is also true that many migrants critical as ever. face systemic disadvantages, making it difficult At the time of writing, the world is undergo- or impossible for them to access local services on ing the most severe economic crisis in over half a equal terms with local people. And these prob- century. Shrinking economies and layoffs are af- lems are especially severe for temporary and ir- fecting millions of workers, including migrants. regular workers. We believe that the current downturn should In migrants’ countries of origin, the impacts be seized as an opportunity to institute a new of movement are felt in higher incomes and deal for migrants—one that will benefit work- consumption, better education and improved ers at home and abroad while guarding against a health, as well as at a broader cultural and so- protectionist backlash. With recovery, many of cial level. Moving generally brings benefits, most the same underlying trends that have been driv- directly in the form of remittances sent to im- ing movement during the past half-century will mediate family members. However, the benefits resurface, attracting more people to move. It is are also spread more broadly as remittances are vital that governments put in place the necessary spent—thereby generating jobs for local work- measures to prepare for this. ers—and as behaviour changes in response to ideas from abroad. Women, in particular, may Our proposal be liberated from traditional roles. Large gains to human development can be The nature and extent of these impacts de- achieved by lowering the barriers to movement pend on who moves, how they fare abroad and and improving the treatment of movers. A bold whether they stay connected to their roots vision is needed to realize these gains. This 3
  • 15. Overview HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development The two most report sets out a case for a comprehensive set of conditional on local demand. Experience important dimensions reforms that can provide major benefits to mi- suggests that good practices here include: en- of the mobility agenda grants, communities and countries. suring immigrants have the right to change that offer scope for Our proposal addresses the two most im- employers (known as employer portability), better policies are portant dimensions of the mobility agenda offering immigrants the right to apply to admissions and that offer scope for better policies: admissions extend their stay and outlining pathways to treatment and treatment. The reforms laid out in our eventual permanent residence, making pro- proposed core package have medium- to long- visions that facilitate return trips during the term pay-offs. They speak not only to destina- visa period, and allowing the transfer of accu- tion governments but also to governments of mulated social security benefits, as adopted origin, to other key actors—in particular the in Sweden’s recent reform. private sector, unions and non-governmental Destination countries should decide on the organizations—and to individual migrants desired numbers of entrants through political themselves. While policy makers face common processes that permit public discussion and the challenges, they will of course need to design balancing of different interests. Transparent and implement different migration policies mechanisms to determine the number of en- in their respective countries, according to na- trants should be based on employer demand, tional and local circumstances. Certain good with quotas according to economic conditions. practices nevertheless stand out and can be At destination, immigrants are often treated more widely adopted. in ways that infringe on their basic human We highlight six major directions for re- rights. Even if governments do not ratify the form that can be adopted individually but that, international conventions that protect migrant if used together in an integrated approach, can workers, they should ensure that migrants have magnify their positive effects on human devel- full rights in the workplace—to equal pay for opment. Opening up existing entry channels equal work, decent working conditions and so that more workers can emigrate, ensuring collective organization, for example. They may basic rights for migrants, lowering the trans- need to act quickly to stamp out discrimina- action costs of migration, finding solutions tion. Governments at origin and destination that benefit both destination communities can collaborate to ease the recognition of cre- and the migrants they receive, making it easier dentials earned abroad. for people to move within their own countries, The current recession has made migrants par- and mainstreaming migration into national ticularly vulnerable. Some destination country development strategies—all have important governments have stepped up the enforcement and complementary contributions to make to of migration laws in ways that can infringe on human development. migrants’ rights. Giving laid-off migrants the The core package highlights two avenues for opportunity to search for another employer opening up regular existing entry channels: (or at least time to wrap up their affairs before • We recommend expanding schemes for departing), publicizing employment outlooks— truly seasonal work in sectors such as agri- including downturns in source countries—are culture and tourism. Such schemes have al- all measures that can mitigate the disproportion- ready proved successful in various countries. ate costs of the recession borne by both current Good practice suggests that this interven- and prospective migrants. tion should involve unions and employers, For international movement, the transaction together with the destination and source costs of acquiring the necessary papers and meet- country governments, particularly in design- ing the administrative requirements to cross na- ing and implementing basic wage guaran- tional borders are often high, tend to be regressive tees, health and safety standards and provi- (proportionately higher for unskilled people and sions for repeat visits as in the case of New those on short-term contracts) and can also have Zealand, for example. the unintended effect of encouraging irregular • We also propose increasing the number of movement and smuggling. One in ten countries visas for low-skilled people, making this have passport costs that exceed 10 percent of per 4
  • 16. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Overview capita income; not surprisingly, these costs are this potential and to integrate migration with While not a negatively correlated with emigration rates. Both other aspects of national development policy. A substitute for broader origin and destination governments can simplify critical point that emerges from experience is the development efforts, procedures and reduce document costs, while importance of national economic conditions and migration can be a vital the two sides can also work together to improve strong public-sector institutions in enabling the strategy for households and regulate intermediation services. broader benefits of mobility to be reaped. and families seeking to It is vital to ensure that individual migrants diversify and improve settle in well on arrival, but it is also vital that The way forward their livelihoods the communities they join should not feel un- Advancing this agenda will require strong, en- fairly burdened by the additional demands lightened leadership coupled with a more deter- they place on key services. Where this poses mined effort to engage with the public and raise challenges to local authorities, additional fis- their awareness about the facts around migration. cal transfers may be needed. Ensuring that For origin countries, more systematic consid- migrant children have equal access to educa- eration of the profile of migration and its ben- tion and, where needed, support to catch up efits, costs and risks would provide a better basis and integrate, can improve their prospects and for integrating movement into national develop- avoid a future underclass. Language training is ment strategies. Emigration is not an alternative key—for children at schools, but also for adults, to accelerated development efforts at home, but both through the workplace and through spe- mobility can facilitate access to ideas, knowledge cial efforts to reach women who do not work and resources that can complement and in some outside the home. Some situations will need cases enhance progress. more active efforts than others to combat dis- For destination countries, the ‘how and crimination, address social tensions and, where when’ of reforms will depend on a realistic look relevant, prevent outbreaks of violence against at economic and social conditions, taking into immigrants. Civil society and governments account public opinion and political constraints have a wide range of positive experience in at local and national levels. tackling discrimination through, for example, International cooperation, especially through awareness-raising campaigns. bilateral or regional agreements, can lead to bet- Despite the demise of most centrally planned ter migration management, improved protection systems around the world, a surprising number of migrants’ rights and enhanced contributions of governments—around a third—maintain de of migrants to both origin and destination coun- facto barriers to internal movement. Restrictions tries. Some regions are creating free-movement typically take the form of reduced basic service zones to promote freer trade while enhancing provisions and entitlements for those not regis- the benefits of migration—such as West Africa tered in the local area, thereby discriminating and the Southern Cone of Latin America. The against internal migrants, as is still the case in expanded labour markets created in these regions China. Ensuring equity of basic service provi- can deliver substantial benefits to migrants, their sion is a key recommendation of the report as families and their communities. regards internal migrants. Equal treatment is There are calls to create a new global regime to important for temporary and seasonal workers improve the management of migration: over 150 and their families, for the regions where they go countries now participate in the Global Forum to work, and also to ensure decent service provi- on Migration and Development. Governments, sion back home so that they are not compelled to faced with common challenges, develop com- move in order to access schools and health care. mon responses—a trend we saw emerge while While not a substitute for broader develop- preparing this report. ment efforts, migration can be a vital strategy for Overcoming Barriers fixes human develop- households and families seeking to diversify and ment firmly on the agenda of policy makers who improve their livelihoods, especially in develop- seek the best outcomes from increasingly com- ing countries. Governments need to recognize plex patterns of human movement worldwide. 5
  • 17. Freedom and movement: how mobility can foster human development 1
  • 18. The world distribution of opportunities is extremely unequal. This inequality is a key driver of human movement and thus implies that movement has a huge potential for improving human development. Yet movement is not a pure expression of choice—people often move under constraints that can be severe, while the gains they reap from moving are very unequally distributed. Our vision of development as promoting people’s freedom to lead the lives they choose recognizes mobility as an essential component of that freedom. However, movement involves trade-offs for both movers and stayers, and the understanding and analysis of those trade-offs is key to formulating appropriate policies.
  • 19. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 1 Freedom and movement: how mobility can foster human development Every year, more than 5 million people cross international borders to go and live in a developed country.1 The number of people who move to a developing nation or within their country is much greater, al- though precise estimates are hard to come by.2 Even larger numbers of people in both destination and source places are affected by the movement of others through flows of money, knowledge and ideas. For people who move, the journey almost always and its HDI. The lowest HDI in a United States entails sacrifices and uncertainty. The possible border county (Starr County, Texas) is above costs range from the emotional cost of separa- even the highest on the Mexican side (Mexicali tion from families and friends to high monetary Municipality, Baja California).4 This pattern fees. The risks can include the physical dangers suggests that moving across national borders of working in dangerous occupations. In some can greatly expand the opportunities available cases, such as those of illegal border crossings, for improved well-being. Alternatively, consider movers face a risk of death. Nevertheless, mil- the direction of human movements when re- lions of people are willing to incur these costs or strictions on mobility are lifted. Between 1984 risks in order to improve their living standards and 1995, the People’s Republic of China pro- and those of their families. gressively liberalized its strict regime of inter- A person’s opportunities to lead a long and nal restrictions, allowing people to move from healthy life, to have access to education, health one region to another. Massive flows followed, care and material goods, to enjoy political free- largely towards regions with higher levels of doms and to be protected from violence are all human development. In this case the patterns strongly influenced by where they live. Someone again suggest that opportunities for improved born in Thailand can expect to live seven more well-being were a key driving factor (map 1.2).5 years, to have almost three times as many years These spatial impressions are supported by of education, and to spend and save eight times more rigorous research that has estimated the as much as someone born in neighbouring effect of changing one’s residence on well-being. Myanmar.3 These differences in opportunity These comparisons are inherently difficult be- create immense pressures to move. cause people who move tend to have different characteristics and circumstances from those who 1.1 Mobility matters do not move (box 1.1). Recent academic studies Witness for example the way in which human that carefully disentangle these complex relations development outcomes are distributed near na- have nonetheless confirmed very large gains from tional boundaries. Map 1.1 compares human moving across international borders. For example, development on either side of the United States– individuals with only moderate levels of formal Mexico border. For this illustration, we use the education who move from a typical developing Human Development Index (HDI)—a sum- country to the United States can reap an annual mary measure of development used throughout income gain of approximately US$10,000— this report to rank and compare countries. A roughly double the average level of per capita pattern that jumps out is the strong correlation income in a developing country.6 Background between the side of the border that a place is on research commissioned for this report found that 9
  • 20. 1 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Map 1.1 Borders matter HDI in United States and Mexican border localities, 2000 HDI, 2000 0.636 – 0.700 0.701 – 0.765 0.766 – 0.830 0.831 – 0.895 0.896 – 0.950 Mexicali: HDI = 0.757 Starr: HDI = 0.766 Source: Anderson and Gerber (2007a). a family who migrates from Nicaragua to Costa exploitative working conditions sometimes as- Rica increases the probability that their child will sociated with domestic work and the lack of re- be enrolled in primary school by 22 percent.7 dress mechanisms can trap migrant women in a These disparities do not explain all movement. vicious circle of poverty and HIV vulnerability.8 An important part of movement occurs in response The same study found that many countries test to armed conflict. Some people emigrate to avoid migrants for HIV and deport those found to political repression by authoritarian states. Moving carry the virus; few source countries have re-in- can provide opportunities for people to escape the tegration programs for migrants who are forced traditional roles that they were expected to fulfil in to return as a result of their HIV status.9 their society of origin. Young people often move in Movement across national borders is only search of education and broader horizons, intend- part of the story. Movement within national ing to return home eventually. As we discuss in borders is actually larger in magnitude and has more detail in the next section, there are multiple enormous potential to enhance human devel- drivers of, and constraints on, movement that ac- opment. This is partly because relocating to an- count for vastly different motives and experiences other country is costly. Moving abroad not only among movers. Nevertheless, opportunity and as- involves substantial monetary costs for fees and piration are frequently recurring themes. travel (which tend to be regressive—see chapter Movement does not always lead to better 3), but may also mean living in a very different human development outcomes. A point that we culture and leaving behind your network of emphasize throughout this report is that vast friends and relations, which can impose a heavy inequalities characterize not only the freedom if unquantifiable psychological burden. The lift- to move but also the distribution of gains from ing of what were often severe barriers to internal movement. When the poorest migrate, they movement in a number of countries (including often do so under conditions of vulnerability but not limited to China) has benefited many that reflect their limited resources and choices. of the world’s poorest people—an impact on The prior information they have may be limited human development that would be missed if or misleading. Abuse of migrant female do- we were to adopt an exclusive focus on interna- mestic workers occurs in many cities and coun- tional migration. tries around the world, from Washington and The potential of enhanced national and inter- London to Singapore and the Gulf Cooperation national mobility to increase human well-being Council (GCC) states. Recent research in leads us to expect that it should be a major focus the Arab states found that the abusive and of attention among development policy makers 10
  • 21. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 1 and researchers. This is not the case. The academic Refugees (UNHCR), and on discussions in such literature dealing with the effects of migration arenas as the Global Forum on Migration and is dwarfed by research on the consequences of Development, we argue that migration deserves international trade and macroeconomic poli- greater attention from governments, interna- cies, to name just two examples.10 While the tional organizations and civil society.12 This is international community boasts an established not only because of the large potential gains to institutional architecture for governing trade the world as a whole from enhanced movement, and financial relations among countries, the but also because of the substantial risks faced by governance of mobility has been well character- many who move—risks that could be at least ized as a non-regime (with the important excep- partly offset by better policies. tion of refugees).11 This report is part of ongoing efforts to redress this imbalance. Building on 1.2 Choice and context: the recent work of organizations such as the understanding why people move International Organization for Migration There is huge variation in the circumstances sur- (IOM), the International Labour Organization rounding human movement. Thousands of Chin (ILO), the World Bank and the Office of have emigrated to Malaysia in recent years to es- the United Nations High Commissioner for cape persecution by Myanmar’s security forces, Map 1.2 Migrants are moving to places with greater opportunities Human development and inter-provincial migration flows in China, 1995–2000 HDI, 1995 0.000 – 0.600 0.601 – 0.700 0.701 – 0.800 0.801 – 1 Number of migrants, 1995–2000 > 2,500,000 No data 1,000,000– 2,500,000 150,000–1,000,000 Source: UNDP (2008a) and He (2004). 11
  • 22. 1 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Box 1.1 Estimating the impact of movement Key methodological considerations affect the measurement of both or on particular skill groups. These may still be subject to selection returns to individuals and effects on places reported in the exten- bias associated with individual choices of location. A key issue, dis- sive literature on migration. Obtaining a precise measure of impacts cussed in chapter 4, is whether the migrants’ skills substitute for or requires a comparison between the well-being of someone who mi- complement those of local people; determining this requires accurate grates and their well-being had they stayed in their original place. measures of these skills. The latter is an unknown counterfactual and may not be adequately One increasingly popular approach seeks to exploit quasi- or proxied by the status of non-migrants. Those who move internation- manufactured randomization to estimate impacts. For example, New ally tend to be better educated and to have higher levels of initial Zealand’s Pacific Access Category allocated a set of visas randomly, income than those who do not, and so can be expected to be better allowing the impact of migration to be assessed by comparing lottery off than those who stay behind. There is evidence that this phenom- winners with unsuccessful applicants. enon—known technically as migrant selectivity—is also present in There is also an important temporal dimension. Migration has internal migration (see chapter 2). Comparisons of groups with similar high upfront costs and the gains may take time to accrue. For ex- observable characteristics (gender, education, experience, etc.) can ample, returns in the labour market tend to improve significantly be more accurate but still omit potentially important characteristics, over time as country-specific skills are learned and recognized. A such as attitudes towards risk. migrant’s decision to return is an additional complication, affecting There are a host of other methodological problems. Difficulties the period over which impacts should be measured. in identifying causality plague estimates of the impact of remittances Finally, as we discuss in more detail in the next chapter, migration on household consumption. Understanding how migration affects analysis faces major data constraints. Even in the case of rich coun- labour markets in the destination place is also problematic. Most tries, comparisons are often difficult to make for very basic reasons, studies have tried to look at the impact on wages at the regional level such as differences in the definition of migrants. Source: Clemens, Montenegro and Pritchett (2008), McKenzie, Gibson and Stillman (2006). but live under constant fear of detection by civil- migration, to take one example, covers cases rang- ian paramilitary groups.13 More than 3,000 people ing from Tajik workers in the Russian Federation are believed to have drowned between 1997 and construction industry, impelled to migrate by 2005 in the Straits of Gibraltar while trying to harsh economic conditions in a country where enter Europe illegally on makeshift boats.14 These most people live on less than US$2 a day, to highly experiences contrast with those of hundreds of coveted East Asian computer engineers recruited poor Tongans who have won a lottery to settle in by the likes of Motorola and Microsoft. New Zealand, or of the hundreds of thousands of Conventional approaches to migration Poles who moved to better paid jobs in the United tend to suffer from compartmentalization. Kingdom under the free mobility regime of the Distinctions are commonly drawn between mi- European Union introduced in 2004. grants according to whether their movement is Our report deals with various types of move- classed as forced or voluntary, internal or interna- ment, including internal and international, tem- tional, temporary or permanent, or economic or porary and permanent, and conflict-induced. non-economic. Categories originally designated The usefulness of casting a broad net over all to establish legal distinctions for the purpose of of these cases might be questioned. Are we not governing entry and treatment can end up play- talking about disparate phenomena, with widely ing a dominant role in conceptual and policy different causes and inherently dissimilar out- thinking. Over the past decade, scholars and pol- comes? Wouldn’t our purpose be better served icy makers have begun to question these distinc- if we limited our focus to one type of migration tions, and there is growing recognition that their and studied in detail its causes, consequences proliferation obscures rather than illuminates the and implications? processes underlying the decision to move, with We don’t think so. While broad types of potentially harmful effects on policy-making.15 human movement do vary significantly in their In nearly all instances of human movement drivers and outcomes, this is also true of more spe- we can see the interaction of two basic forces, cific cases within each type. International labour which vary in the degree of their influence. On 12
  • 23. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 1 the one hand we have individuals, families and to conceptualize migration flows focused on dif- Theories that sometimes communities, who decide to move of ferences in living standards, in recent years there emphasize purely their own free will in order to radically alter their has been growing understanding that these differ- economic factors fail circumstances. Indeed, even when people are im- ences only partly explain movement patterns.20 In to capture the broader pelled to move by very adverse conditions, the particular, if movement responds only to income social framework in choices they make almost always play a vital role. differentials, it is hard to explain why many suc- which decisions to Research among Angolan refugees settling in cessful migrants choose to return to their country migrate are taken northwest Zambia, for example, has shown that of origin after several years abroad. Furthermore, many were motivated by the same aspirations if migration were purely determined by wage dif- that impel those who are commonly classified as ferences, then we would expect to see large move- economic migrants.16 Similarly, Afghans fleeing ments from poor to rich countries and very little conflict go to Pakistan or Iran via the same routes movement among rich countries—but neither of and trading networks established decades ago for these patterns holds in practice (chapter 2). the purposes of seasonal labour migration.17 These observed patterns led to several strands On the other hand, choices are rarely, if ever, of research. Some scholars recognized that a unconstrained. This is evident for those who focus on the individual distracts from what is move to escape political persecution or economic typically a family decision and indeed strategy deprivation, but it is also vital for understanding (as when some family members move while oth- decisions where there is less compulsion. Major ers stay at home). 21 The need to go beyond the factors relating to the structure of the economy assumption of perfectly competitive markets and of society, which are context-specific but also also became increasingly evident. In particu- change over time, frame decisions to move as well lar, credit markets in developing countries are as to stay. This dynamic interaction between indi- highly imperfect, while household livelihoods vidual decisions and the socio-economic context often depend on such volatile sectors as agri- in which they are taken—sometimes labelled in culture. Sending a family member elsewhere sociological parlance the ‘agency–structure inter- allows the family to diversify against the risk action’—is vital for understanding what shapes of bad outcomes at home.22 Other researchers human behaviour. The evolution over time of key emphasized how structural characteristics and structural factors is dealt with in chapter 2. long-run trends in both origin and destination Consider the case of the tens of thousands of places—often labelled ‘push’ and ‘pull’ factors— Indonesian immigrants who enter Malaysia every shape the context in which movement occurs. year. These flows are driven largely by the wide Movement, for example, can result from grow- income differentials between these countries. But ing concentration in the ownership of assets such the scale of movement has also grown steadily as land, making it difficult for people to subsist since the 1980s, whereas the income gap be- through their traditional modes of production.23 tween the two countries has alternately widened It was also recognized that the opportunities and narrowed over the same period.18 Broader available to migrants are constrained by barriers socio-economic processes have clearly played a to entry, as we discuss in chapters 2 and 3, and part. Malaysian industrialization in the 1970s by the way in which labour markets function, as and 1980s generated a massive movement of shown by the considerable evidence that both in- Malays from the countryside to the cities, creat- ternational and internal migrants are channelled ing acute labour scarcity in the agricultural sector into lower-status and worse-paid occupations. at a time when the commercialization of farming Most importantly, theories that empha- and rapid population growth were producing a size purely economic factors fail to capture the surplus of agricultural labour in Indonesia. The broader social framework in which decisions are fact that most Indonesians are of similar ethnic, taken. For example, young men among the lower linguistic and religious backgrounds to Malays caste Kolas in the Central Gujarat region of India doubtless facilitated the flows.19 commonly seek factory jobs outside their village Recognition of the role of structural factors in order to break away from subordinate caste in determining human movement has had a deep relations. This occurs despite the fact that fac- impact on migration studies. While early attempts tory wages are not higher, and in some cases are 13
  • 24. 1 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development lower, than what they would earn as agricultural development as the expansion of people’s free- day labourers at home. 24 Escaping traditional hi- doms to live their lives as they choose. This con- erarchies can be an important factor motivating cept—inspired by the path-breaking work of migration (chapter 3). Nobel laureate Amartya Sen and the leadership Moreover, the relationship between move- of Mahbub ul Haq and also known as the ‘ca- ment and economics is far from unidirectional. pabilities approach’ because of its emphasis on Large-scale movements of people can have pro- freedom to achieve vital ‘beings and doings’— found economic consequences for origin and has been at the core of our thinking since the destination places, as we will discuss in detail in first Human Development Report in 1990, and chapter 4. Even the way in which we think about is as relevant as ever to the design of effective basic economic concepts is affected by the move- policies to combat poverty and deprivation. 25 ment of people, as can be illustrated by the issues The capabilities approach has proved powerful raised for the measurement of per capita incomes in reshaping thinking about topics as diverse as and economic growth (box 1.2). gender, human security and climate change. Using the expansion of human freedoms and 1.3 Development, freedom and capabilities as a lens has significant implications human mobility for how we think about human movement. This Our attempt to understand the implications of is because, even before we start asking whether human movement for human development be- the freedom to move has significant effects on in- gins with an idea that is central to the approach comes, education or health, for example, we rec- of this report. This is the concept of human ognize that movement is one of the basic actions Box 1.2 How movement matters to the measurement of progress Attempts to measure the level of development of a country rely on the Philippines, regardless of where they now live. This new measure various indicators designed to capture the average level of well- has a significant impact on our understanding of human well-being. being. While a traditional approach uses per capita income as a proxy In 13 of the 100 nations for which we can calculate this measure, the for economic development, this report has promoted a more com- HDI of their people is at least 10 percent higher than the HDI of their prehensive measure: the Human Development Index (HDI). However, country; for an additional nine populations, the difference is between both of these approaches are based on the idea of evaluating the 5 and 10 percent. For 11 of the 90 populations for which we could well-being of those who reside in a given territory. calculate trends over time, the change in HDI during the 1990–2000 As researchers at the Center for Global Development and period differed by more than 5 percentage points from the average Harvard University have recently pointed out, these approaches to change for their country. For example, the HDI of Ugandans went up measuring development prioritize geographical location over people by nearly three times as much as the HDI of Uganda. in the evaluation of a society’s progress. Thus, if a Fijian moves to Throughout the rest of this report, we will continue to adopt New Zealand and her living standards improve as a result, traditional the conventional approach for reasons of analytical tractability and measures of development will not count that improvement as an in- comparability with the existing literature. We also view these two crease in the development of Fiji. Rather, that person’s well-being will measures as complements rather than substitutes: one captures now be counted in the calculation of New Zealand’s indicator. the living standards of people living in a particular place, the other In background research carried out for this report, we dealt with of people born in a particular place. For example, when we anal- this problem by proposing an alternative measure of human devel- yse human development as a cause of human movement, as we opment. We refer to this as the human development of peoples (as do throughout most of this report, then the country measure will be opposed to the human development of countries), as it captures the more appropriate because it will serve as an indicator of how living level of human development of all people born in a particular country. standards differ across places. For the purposes of evaluating the For instance, instead of measuring the average level of human devel- success of different policies and institutions in generating well-being opment of people who live in the Philippines, we measure the aver- for the members of a society, however, there is a strong case for age level of human development of all individuals who were born in adopting the new measure. Source: Ortega (2009) and Clemens and Pritchett (2008). 14
  • 25. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 1 that individuals can choose to take in order to live as well as the act of movement itself, we rec- realize their life plans. In other words, the ability ognize the importance of the conditions under to move is a dimension of freedom that is part of which people are, or are not, able to choose their development—with intrinsic as well as potential place of residence. Much conventional analysis instrumental value. of migration centres on studying the effect of The notion that the ability to change one’s movement on well-being. Our concern, however, place of residence is a fundamental component is not only with movement in itself but also with of human freedom has been traced back to clas- the freedom that people have to decide whether sical philosophy in several intellectual traditions. to move. Mobility is a freedom—movement is Confucius wrote that “good government obtains the exercise of that freedom.29 when those who are near are made happy, and We understand human mobility as a posi- those who are far off are attracted to come,”26 tive and not only a negative freedom. In other while Socrates argued that “anyone who does words, the absence of formal restrictions on the not like us and the city, and who wants to em- movement of people across or within borders igrate to a colony or to any other city, may go does not in itself make people free to move if where he likes, retaining his property.”27 In 1215, England’s Magna Carta guaranteed the freedom “to go out of our Kingdom, and to return safely Box 1.3 Basic terms used in this report and securely, by land or water.” More recently, American philosopher Martha Nussbaum ar- gued that mobility is one of a set of basic human Human Development Index (HDI) A composite index measuring average functional capabilities that can be used to assess achievement in three basic dimensions of human development: a long the effective freedom that individuals have to and healthy life, access to knowledge and a decent standard of living. carry out their life plans.28 Yet world history is replete with the experi- Developed/developing We call countries that have achieved an HDI of 0.9 ences of societies that severely limited human or higher developed, and those that have not developing. development by restricting movement. Both feudalism and slavery were predicated on the Low/medium/high/very high HD A classification of countries based on physical restriction of movement. Several re- the value of the HDI according to the most recent data. The ranges are pressive regimes in the 20th century relied on 0–0.499 for low HDI, 0.500–0.799 for medium HDI, 0.800–0.899 for high the control of internal movement, including the HDI and greater than 0.900 for very high HDI. Pass Laws of South African apartheid and the propiska system of internal passports in Soviet Internal migration Human movement within the borders of a country, Russia. The subsequent demise of such restric- usually measured across regional, district or municipal boundaries. tions contributed to dramatic expansions in the freedoms enjoyed by these countries’ peoples. International migration Human movement across international borders, Our report seeks to capture and examine the resulting in a change of country of residence. full set of conditions that affect whether individu- als, families or communities decide to stay or to Migrant An individual who has changed her place of residence either by move. These conditions include people’s resources crossing an international border or by moving within her country of origin or entitlements as well as the way in which dif- to another region, district or municipality. An emigrant is a migrant viewed ferent constraints—including those associated from the perspective of the origin country, while an immigrant is a migrant with policies, markets, security, culture and val- viewed from the perspective of the destination country. While sometimes ues—determine whether movement is an option the term ‘migrant’ (as opposed to ’immigrant’) has been reserved for for them. People’s ability to choose the place they temporary migration, we do not adopt such a distinction in this report. call home is a dimension of human freedom that we refer to as human mobility. Box 1.3 defines this Human mobility The ability of individuals, families or groups of people to and other basic terms used in this report. choose their place of residence. The distinction between freedoms and actions is central to the capabilities approach. Human movement The act of changing one’s place of residence. By referring to the capability to decide where to 15
  • 26. 1 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development they lack the economic resources, security and perhaps impossible, for her to remain at home. networks necessary to enjoy a decent life in their These circumstances restrict the scope of her new home, or if informal constraints such as dis- choices, reducing her freedom to choose where crimination significantly impede the prospects to live. The induced movement may very well of moving successfully. coincide with a further deterioration in her liv- Let us illustrate the implications of this ap- ing conditions, but this does not mean that the proach with a couple of examples. In the case of movement is the cause of that deterioration. In human trafficking, movement comes together fact, if she were not able to move, the outcome with brutal and degrading types of exploita- would probably be much worse. tion. By definition, trafficking is an instance of If it is tempting to view the distinction be- movement in which freedoms become restricted tween mobility and movement as somewhat by means of force, deception and/or coercion. academic, we should take this opportunity to Commonly, a trafficked individual is not free emphasize that freedom to choose where to live to choose to abort the trip, to seek alternative emerged as an important theme in research to employment once she gets to her destination, or find out what poor people think about migration to return home. A trafficked person is physically (box 1.4). In the end, their views matter more moving, but doing so as a result of a restriction on than those of the experts, since it is they who her ability to decide where to live. From a capa- must take the difficult decision as to whether or bilities perspective, she is less, not more, mobile. not to risk a move. Alternatively, consider the case of some- one who has to move because of the threat of 1.4 What we bring to the table political persecution or because of degraded Putting people and their freedom at the centre environmental conditions. In these cases exter- of development has implications for the study of nal circumstances have made it more difficult, human movement. In the first place, it requires Box 1.4 How do the poor view migration? In recent years there has been growing interest in the use of qualita- Seasonal internal migration was the most common type of mi- tive methods to understand how people living in poverty view their gration discussed in focus groups with the poor. When international situation, as indicated by the landmark World Bank study Voices of migration was discussed, it was described as something for the bet- the Poor, published in 2000. In preparing the current report we com- ter off. For instance, participants in the Jamaica study said that the missioned research to investigate relevant findings of Participatory better off, unlike the poor, have influential contacts that help them Poverty Assessments—large-scale studies that combine qualitative acquire the necessary visas to travel and work abroad. Similarly, in and quantitative research methods to study poverty from the point of Montserrat participants described how the more educated and finan- view of the poor. What emerged is that moving is commonly described cially better off were able to leave the country after the 1995 volcano by the poor as both a necessity—part of a coping strategy for families eruption, while the less well off stayed on despite the devastation. experiencing extreme hardship—and an opportunity—a means of ex- Participatory Poverty Assessments give us a good picture of panding a household’s livelihoods and ability to accumulate assets. how poor people see movement but may be uninformative about In Niger, two thirds of respondents indicated that in order to cope how others have managed to move out of poverty, as these assess- with lack of food, clothing or income they had left their homes and ments are by design limited to people who are still poor. A more re- looked for livelihoods elsewhere. Some households reported mem- cent study of 15 countries carried out by the World Bank examines bers leaving in search of paid work, particularly to reduce pressures pathways out of poverty. In these studies, the ability to move evolved on dwindling food supplies in times of scarcity. In the villages of Ban as a common theme in conversations about freedom. In Morocco, Na Pieng and Ban Kaew Pad, Thailand, participants described mi- young women expressed frustration with traditional restrictions that gration as one of the ways in which a family’s socio-economic status limit women’s ability to travel without a male escort or search for could be enhanced. For these communities, remittances from abroad employment outside the home. Men described the ability to migrate enabled those left behind to invest in commercial fishing and thus as both a freedom and a responsibility, because with the freedom to expand the family’s standing and influence. move comes the responsibility to remit. Source: Azcona (2009), Narayan, Pritchett, and Kapoor (2009), World Bank (2000), World Bank (2003), and ActionAid International (2004). 16
  • 27. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 1 us to understand what makes people less or more internal movements.30 People living and working While there is mobile. This means considering why people with irregular status are often denied a whole host considerable intrinsic choose to move and what constraints encourage of basic entitlements and services and lead their value to mobility, its them or deter them from making that choice. In lives in constant fear of arrest and deportation. instrumental value chapter 2, we look at both choices and constraints Understanding the effects of movement requires for furthering other by studying the macro patterns of human move- the systematic analysis of these multiple dimen- dimensions of human ment over space and time. We find that these sions of human development in order to gain a development can patterns are broadly consistent with the idea that better sense of the nature and extent of these trade- also be of enormous people move to enhance their opportunities, but offs, as well as the associated policy implications. significance that their movement is strongly constrained by More complex trade-offs occur when mov- policies—both in their place of origin and at ers have an effect on the well-being of non- their destinations—and by the resources at their movers. Indeed, the perception that migration disposal. Since different people face different generates losses for those in destination coun- constraints, the end result is a process character- tries has been the source of numerous debates ized by significant inequalities in opportunities among policy makers and academics. Chapter to move and returns from movement. 4 focuses on these debates. The evidence we We explore how these inequalities interact present strongly suggests that fears about the with policies in chapter 3. While, as we have negative effects of movement on stayers (both emphasized in this introductory chapter, there at source and destination) are frequently over- is considerable intrinsic value to mobility, its stated. However, sometimes these concerns are instrumental value for furthering other dimen- real and this has significant implications for the sions of human development can also be of enor- design of policy. mous significance. But while people can and do If movement is constrained by policies and expand other freedoms by moving, the extent resources, yet enhanced mobility can signifi- to which they are able to do so depends greatly cantly increase the well-being of movers while on the conditions under which they move. In often also having positive effects on stayers, what chapter 3 we look at the outcomes of migration should policy towards human movement look in different dimensions of human development, like? In chapter 5, we argue that it should look including incomes and livelihoods, health, edu- very different from what we see today. In par- cation and empowerment. We also review the ticular, it should be redesigned to open up more cases in which people experience deteriorations opportunities for movement among low-skilled in their well-being during movement—when workers and to improve the treatment of movers this is induced by trafficking or conflict, for ex- at their destinations. ample—and argue that these cases can often be We do not advocate wholesale liberalization traced back to constraints on the freedom of in- of international mobility. This is because we dividuals to choose where they live. recognize that people at destination places have A key point that emerges in chapter 3 is that a right to shape their societies, and that borders human movement can be associated with trade- are one way in which people delimit the sphere of offs—people may gain in some and lose in other their obligations to those whom they see as mem- dimensions of freedom. Millions of Asian and bers of their community. But we also believe that Middle Eastern workers in the GCC states ac- people relate to each other in myriad ways and cept severe limitations on their rights as a condi- that their moral obligations can operate at differ- tion for permission to work. They earn higher pay ent levels. This is primarily because individuals than at home, but cannot be with their families, don’t belong to just one society or group. Rather obtain permanent residence or change employ- than being uniquely or solely defined by their re- ers. Many cannot even leave, as their passports are ligion, race, ethnicity or gender, individuals com- confiscated on entry. For many people around the monly see themselves through the multiple prisms world the decision to move involves leaving their of a set of identities. As Amartya Sen has power- children behind. In India, seasonal workers are in fully put it, “A Hutu labourer from Kigali… is not practice excluded from voting in elections when only a Hutu, but also a Kigalian, a Rwandan, an these are scheduled during the peak period of African, a labourer and a human being.”31 17
  • 28. 1 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development We see mobility The responsibilities of distributive justice the distribution of opportunities in our world is as vital to human are overlapping and naturally intersect national highly unequal. We go on to argue that this fact development and boundaries; as such, there is no contradiction has significant implications for understanding movement as a natural between the idea that societies may design insti- why and how people move and how we should expression of people’s tutions with the primary purpose of generating reshape policies towards human movement. Our desire to choose how just outcomes among their members, and the critique of existing policies towards migration is and where to lead idea that the members of that same society will directed at the way in which they reinforce those their lives share an obligation to create a just world with inequalities. As noted in the 1997 HDR, it is and for their fellow humans outside that soci- precisely because “the principles of free global ety. There are many ways in which these obli- markets are applied selectively” that “the global gations are articulated: the creation of charities market for unskilled labour is not as free as the and foundations, the provision of development market for industrial country exports or capi- aid, assistance in building national institutions, tal”.33 Our emphasis on how migration enhances and the reform of international institutions so cultural diversity and enriches people’s lives by as to make them more responsive to the needs of moving skills, labour and ideas builds on the poor countries are just some of them. However, analysis of the 2004 HDR, which dealt with the our analysis, which informs the recommenda- role of cultural liberty in today’s diverse world.34 tions in chapter 5, suggests that reducing restric- At the same time, the agenda of human de- tions on the entry of people—in particular of velopment is evolving, so it is natural for the low-skilled workers and their families—into treatment of particular topics to change over better-off developed and developing countries is time. This report strongly contests the view— one relatively effective way of discharging these held by some policy makers and at times echoed obligations. in past reports—that the movement of people Our report’s policy recommendations are not should be seen as a problem requiring corrective only based on our view of how the world should action.35 In contrast, we see mobility as vital to be. We recognize that the formulation of policies human development and movement as a natural towards human movement must contend with expression of people’s desire to choose how and what can at times look like formidable political where to lead their lives. opposition to greater openness. However, hav- While the potential of increased mobility for ing considered issues of political feasibility, we increasing the well-being of millions of people argue that a properly designed programme of lib- around the world is the key theme of this re- eralization—designed so as to respond to labour port, it is important to stress at the outset that market needs in destination places while also enhanced mobility is only one component of a addressing issues of equity and non-discrimina- strategy for improving human development. We tion—could generate significant support among do not argue that it should be the central one, voters and interest groups. nor are we arguing that it should be placed at the Our analysis builds on the contributions to same level in the hierarchy of capabilities as, say, thinking about human development that have adequate nourishment or shelter. Neither do we been made since the concept was introduced in believe mobility to be a replacement for national the 1990 HDR. That report devoted a full chap- development strategies directed toward invest- ter to urbanization and human development, ing in people and creating conditions for people reviewing the failed experiences of policies de- to flourish at home. Indeed, the potential of mo- signed to reduce internal migration and con- bility to improve the well-being of disadvantaged cluding: “[A]s long as differences exist between groups is limited, because these groups are often rural and urban areas, people will move to try least likely to move. Yet while human mobility is to take advantage of better schools and social not a panacea, its largely positive effects both for services, higher income opportunities, cultural movers and stayers suggest that it should be an amenities, new modes of living, technological in- important component of any strategy to generate novations and links to the world.”32 Like other sustained improvements in human development HDRs, this one begins with the observation that around the world. 18
  • 29. People in motion: who moves where, when and why 2
  • 30. This chapter examines human movement across the world and over time. The patterns are consistent with the idea that people move to seek better opportunities, but also that their movement is strongly constrained by barriers—most importantly, by policies at home and at destination and by lack of resources. Overall, the share of people going to developed countries has increased markedly during the past 50 years, a trend associated with growing gaps in opportunities. Although these flows of people are likely to slow temporarily during the current economic crisis, underlying structural trends will persist once growth resumes and are likely to generate increased pressures for movement in the coming decades.
  • 31. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 2 People in motion: who moves where, when and why The aim of this chapter is to characterize human movement gen- erally—to give an overview of who moves, how, why, where and when. The picture is complex and our broad brushstrokes will in- evitably fail to capture specifics. Nevertheless, the similarities and commonalities that emerge are striking, and help us understand the forces that shape and constrain migration. We start by examining the key features of times greater than those who emigrate. 2 Using movement—its magnitude, composition and the regional patterns found in these data, we directions—in section 2.1. Section 2.2 considers estimate that there are about 740 million inter- how movement today resembles or differs from nal migrants in the world—almost four times as movement in the past. Our examination sug- many as those who have moved internationally. gests that movement is largely shaped by policy By comparison, the contemporary figure for constraints, an issue that we discuss in detail in international migrants (214 million, or 3.1 per- the third section (2.3). In the last section (2.4), cent of the world’s population) looks small. Of we turn to the future and try to understand how course this global estimate is dogged by a num- movement will evolve in the medium to longer ber of methodological and comparability issues, term, once the economic crisis that started in but there are good reasons to believe that the 2008 is over. order of magnitude is right.3 Box 2.1 deals with one of the most frequently voiced concerns about 2.1 Human movement today the international data on migration, namely the Discussions about migration commonly start extent to which they capture irregular migration with a description of flows between developing is discussed below. and developed countries, or what sometimes Even if we restrict attention to international are loosely—and inaccurately—called ‘South– movements, the bulk of these do not occur be- North’ flows. However, most movement in the tween countries with very different levels of devel- world does not take place between developing opment. Only 37 percent of migration in the world and developed countries. Indeed, it does not even is from developing to developed countries. Most take place between countries. The overwhelming migration occurs within countries in the same majority of people who move do so within the category of development: about 60 percent of mi- borders of their own country. grants move either between developing or between One of the reasons why this basic reality of developed countries (the remaining 3 percent human movement is not better known lies in se- move from developed to developing countries).4 vere data limitations. Background research con- This comparison relies on what is inevitably ducted for this report sought to overcome this a somewhat arbitrary distinction between coun- knowledge gap by using national censuses to cal- tries that have achieved higher levels of develop- culate the number of internal migrants on a con- ment and those that have not. We have classified sistent basis for 24 countries covering 57 percent countries that have attained an HDI greater of the world’s population (figure 2.1).1 Even with than or equal to 0.9 (on a scale of 0 to 1) as de- a conservative definition of internal migration, veloped and those that have not as developing (see which counts movement across only the largest box 1.3). We use this demarcation throughout zonal demarcations in a country, the number of this report, without intending any judgement of people who move internally in our sample is six the merits of any particular economic or political 21
  • 32. 2 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Figure 2.1 Many more people move within borders than across them system or seeking to obscure the complex inter- Internal movement and emigration rates, 2000–2002 actions involved in increasing and sustaining human well-being. The countries and territories thereby classified as developed feature many that Ghana would normally be included in such a list (all Kenya Western European countries, Australia, Canada, Japan, New Zealand and the United States), but also several that are less frequently labelled as de- Africa Rwanda veloped (Hong Kong (China), Singapore and the South Africa Republic of Korea, in East Asia; Kuwait, Qatar Uganda and the United Arab Emirates, in the Gulf re- gion). However, most Eastern European econo- Argentina mies, with the exception of the Czech Republic Brazil and Slovenia, are not in the top HDI category (see Statistical Table H). Chile One obvious reason why there is not more movement from developing to developed coun- Colombia tries is that moving is costly, and moving long Costa Rica distances is costlier than undertaking short Americas journeys. The higher expense of international Ecuador movement comes not only from transport costs Mexico but also from the policy-based restrictions on crossing international borders, which can be Panama overcome only by those who have enough re- sources, possess skills that are sought after in the United States new host country, or are willing to run very high Venezuela risks. Nearly half of all international migrants move within their region of origin and about 40 Cambodia percent move to a neighbouring country. The China proximity between source and destination coun- tries, however, is not solely geographical: nearly India 6 out of 10 migrants move to a country where the major religion is the same as in their country Asia Indonesia of birth, and 4 out of 10 to a country where the Malaysia dominant language is the same.5 The pattern of these inter- and intra-regional Philippines movements is presented in map 2.1, where the Belarus absolute magnitudes are illustrated by the thick- ness of the arrows, the size of each region has been Europe Portugal represented in proportion to its population, and the colouring of each country represents its HDI Spain category. Intra-regional movement dominates. To | | | | | | take one striking example, intra-Asian migration 0 5 10 15 20 25 accounts for nearly 20 percent of all international migration and exceeds the sum total of move- Lifetime internal migration intensity (%) ments that Europe receives from all other regions. Emigration rate (%) The fact that flows from developing to de- veloped countries account for only a minor- Source: Bell and Muhidin (2009) and HDR team estimates based on Migration DRC (2007) database. ity of international movement does not mean Note: All emigration data are from the Migration DRC (2007) database and cover 2000–2002. The internal migration rates are based on census data from 2000 to 2002, except for Belarus (1999), Cambodia (1998), Colombia (2005), Kenya (1999) and the Philippines (1990). that differences in living standards are unim- portant. Quite the contrary: three quarters of 22
  • 33. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 2 Box 2.1 Counting irregular migrants The only comprehensive estimates of the number of foreign-born people percent. Other researchers estimated under-coverage rates in Los in the world come from the United Nations Department of Economic and Angeles during the 2000 Census at 10–15 percent. Thus it appears Social Affairs (UNDESA) and cover approximately 150 United Nations that the official count in the United States misses 1–1.5 million irregu- (UN) member states. These estimates are primarily based on national lar migrants, or 0.5 percent of the country’s population. censuses, which attempt to count the number of people residing in a par- Few studies of the undercount of migrants have been conducted ticular country at a given moment, where a resident is defined as a per- in developing countries. One exception is Argentina, where a recent son who “has a place to live where he or she normally spends the daily study found an underestimation of the migrant stock equivalent to period of rest.” In other words, national censuses attempt to count all 1.3 percent of the total population. In other developing countries, residents, regardless of whether they are regular or irregular. the undercount rates could be much higher. Estimates of the num- However, there are good reasons to suspect that censuses sig- ber of irregular migrants for a number of countries—including the nificantly undercount irregular migrants, who may avoid census inter- Russian Federation, South Africa and Thailand—range from 25 to 55 viewers for fear that they will share information with other government percent of the population. However, there is huge uncertainty about authorities. House owners may conceal the fact that they have illegal the true number. According to the migration experts surveyed by the units rented to irregular migrants. Migrants may also be more mobile HDR team, irregular migration was estimated to average around one and thus harder to count. third of all migration for developing countries. An upper bound for Studies have used a variety of demographic and statistical meth- the number of migrants omitted from international statistics can be ods to assess the magnitude of the undercount. In the United States, obtained by assuming that none of these migrants are captured by the Pew Hispanic Center has developed a set of assumptions con- country censuses (that is, an undercount of 100 percent); in that case, sistent with census-based studies and historical demographic data the resulting underestimation in the global statistics for developing from Mexico that estimate the undercount to be approximately 12 countries would be around 30 million migrants. Source: UN (1998), Passel and Cohn (2008), Marcelli and Ong (2002), Comelatto, Lattes, and Levit (2003). See Andrienko and Guriev (2005) for the Russian Federation, and Sabates-Wheeler (2009) for South Africa and Martin (2009b) for Thailand. international movers move to a country with a report compared the HDI of migrants at home higher HDI than their country of origin; among and destination and found that the differences— those from developing countries, this share ex- in both relative and absolute terms—are inversely ceeds 80 percent. However, their destinations are related to the HDI of the country of origin.7 often not developed countries but rather other developing countries with higher living stan- dards and/or more jobs. Figure 2.2 The poorest have the most to gain from moving… The difference between human development Differences between destination and origin country HDI, at origin and destination can be substantial. 2000–2002 Figure 2.2 illustrates this difference—a magni- tude that we loosely call the human development Latin America and 0.3 ‘gains’ from migration—plotted against the ori- the Caribbean Average difference at destination by region gin country’s HDI.6 If migrants were on average emigrating to countries with the same level of 0.2 Asia human development as their origin countries, Africa this magnitude would be zero. In contrast, the difference is positive and generally large for all 0.1 but the most developed countries. The fact that Europe the average gain diminishes as human develop- ment increases shows that it is people from the 0 | | | | | 0.2 0.4 0.6 0.8 1 poorest countries who, on average, gain the most from moving across borders. Origin country HDI Oceania That movers from low-HDI countries have –0.1 North America the most to gain from moving internation- ally is confirmed by more systematic studies. Source: HDR team estimates based on Migration DRC (2007) database. Note: Averages estimated using Kernel density regressions. Background research commissioned for this 23
  • 34. 2 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Map 2.1 Most movement occurs within regions Origin and destination of international migrants, circa 2000 8.22 0.53 9.57 8.53 0.84 Europe 2.44 15.69 North America 31.52 1.30 1.33 1.34 0.35 Asia 1.07 3.1 35.49 1.24 7.25 1.29 19.72 0.06 0.22 0.75 Latin America 13.18 0.14 and the Caribbean 3.13 3.54 0.08 0.30 Africa 1.65 0.31 0.25 0.02 0.13 0.73 Oceania 0.35 Human Development Index, 2007 0.01 Regions Number of migrants (in millions) Very high North America High Europe Intra- Medium Oceania regional Low Latin America and the Caribbean migration Asia The size of countries is proportional to 2007 population. Africa Source: HDR team estimates based on Migration DRC (2007) database. Migrants from low-HDI countries had the most from where they were born and fewer than 1 per- to gain—and indeed on average saw a 15-fold in- cent of Africans live in Europe. Several scholars crease in income (to US$15,000 per annum), a have observed that if we correlate emigration doubling in education enrolment rate (from 47 rates with levels of development, the relation- to 95 percent) and a 16-fold reduction in child ship resembles a ‘hump’, whereby emigration mortality (from 112 to 7 deaths per 1,000 live rates are lower in poor and rich countries than births). Using comparable surveys in a number of among countries with moderate levels of devel- developing countries, the study also found that opment.9 This is illustrated in figure 2.3, which self-selection—the tendency for those who move shows that the median emigration rate in coun- to be better off and better educated—accounted tries with low levels of human development is for only a fraction of these gains. Analysis of only about one third the rate out of countries bilateral migration flows across countries, pre- with high levels of human development.10 When pared as background research for this report, we restrict the comparison to out-migration to confirmed the positive effect on emigration of all developed countries, the relationship is even components of human development at destina- stronger: the median emigration rate among tion, while finding that income differences had countries with low human development is less the most explanatory power.8 These patterns are than 1 percent, compared to almost 5 percent discussed in detail in the next chapter. out of countries with high levels of human de- Paradoxically, despite the fact that people velopment. Analysis of bilateral migration flows moving out of poor countries have the most to prepared as background research for this report gain from moving, they are the least mobile. confirmed that this pattern holds, even when For example, despite the high levels of attention controlling for characteristics of origin and des- given to emigration from Africa to Europe, only tination countries such as life expectancy, years 3 percent of Africans live in a country different of schooling and demographic structure.11 24
  • 35. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 2 Evidence that poverty is a constraint to emi- The aggregate facts just surveyed tell us where gration has also been found in household-level migrants come from and go to, but they do not analysis: a study of Mexican households, for ex- tell us who moves. While severe data limitations ample, found that the probability of migration impede presentation of a full global profile of mi- increased with higher income levels for house- grants, the existing data nonetheless reveal some hold incomes lower than US$15,000 per annum interesting patterns. (figure 2.3, panel B). A commissioned study Approximately half (48 percent) of all in- found that during the Monga or growing sea- ternational migrants are women. This share has son in Bangladesh, when people’s cash resources been quite stable during the past five decades: it are lowest, a randomized monetary incentive stood at 47 percent in 1960. This pattern con- significantly increased the likelihood of migrat- trasts with that of the 19th century, when the ing.12 The magnitude of the effect was large: giv- majority of migrants were men.14 Yet despite ing emigrants an amount equivalent to a week’s recent references to the ‘feminization’ of migra- wages at destination increased the propensity tion, it appears that numerical gender balance to migrate from 14 to 40 percent. These results was largely reached some time ago. However, the shed strong doubts on the idea, often promoted aggregate stability hides trends at the regional in policy circles, that development in countries level. While the share of women going to the of origin will reduce migratory flows. European Union has increased slightly from 48 While many migrant families do improve their standard of living by moving, this is not al- ways the case. As discussed in chapter 3, move- Figure 2.3 … but they also move less ment often coincides with adverse outcomes Emigration rates by HDI and income when it occurs under conditions of restricted choice. Conflict-induced migration and traffick- Panel A: Median emigration rates by origin country HDI group ing are not a large proportion of overall human Low HDI To developing countries movement, but they affect many of the world’s To developed countries poorest people and are thus a special source of Medium HDI concern (box 2.2). High HDI Another key fact about out-migration pat- Very high HDI terns is their inverse relation to the size of a | | | | | | country’s population. For the 48 states with 0 2 4 3 8 10 populations below 1.5 million—which include Median emigration rate (%) 1 low-, 21 medium-, 12 high- and 11 very high- HDI countries—the average emigration rate is Source: HDR team estimates based on Migration DRC (2007) and UN (2009e). 18.4 percent, considerably higher than the world average of 3 percent. Indeed, the top 13 emigra- Panel B: Probability of emigration by income level in Mexican households tion countries in the world are all small states, with Antigua and Barbuda, Grenada, and Saint 1.6 Kitts and Nevis having emigration rates above 40 1.4 percent. The simple correlation between size and 1.2 Probability of migration (%) emigration rates is –0.61. In many cases, it is re- 1.0 moteness that leads people born in small states to 0.8 move in order to take advantage of opportunities 0.6 elsewhere—the same factor that drives much of 0.4 the rural to urban migration seen within coun- 0.2 tries. Cross-country regression analysis confirms that the effect of population size on emigration is 0.0 | | | | | 0 5 10 15 20 higher for countries that are far from world mar- Income per capita (US$ thousands) kets—the more remote a small country is, the more people decide to leave.13 The implications Source: Meza and Pederzini (2006). of these patterns are discussed in box 4.4. 25
  • 36. 2 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development to 52 percent, that same share has dropped from It is important not to overemphasize the dis- 47 to 45 percent in Asia. tinction between categories of migrants, as many Of course, the relatively equal gender shares of migrants shift between categories. Indeed, the mi- the migrant population may hide significant dif- gration regime in many countries can perhaps best ferences in the circumstances of movement and be understood through the analogy of the multi- the opportunities available.15 At the same time, ple doors of a house. Migrants can enter the house a growing literature has challenged conventional through the front door (permanent settlers), the views about the subordinate role of women in side door (temporary visitors and workers) or the migration decisions.16 For example, a qualitative back door (irregular migrants). However, once study of decisions taken by Peruvian couples mov- inside a country, these channels often merge, as ing to Argentina found that many of the women when temporary visitors become immigrants or moved first by themselves, because they were able slip into unauthorized status, those with irregular to secure jobs more rapidly than their partners, status gain authorization to remain, and people who would later follow with the children.17 with permanent status decide to return. The data also show very large temporary flows This analogy is particularly useful for un- of people. In the countries of the Organisation derstanding irregular migration. Overstaying is for Economic Co-operation and Development an important channel through which migrants (OECD), temporary migrants typically repre- become irregular, particularly in developed coun- sent more than a third of arrivals in a given year. tries. In fact, the distinction between regular and However, since most leave after a short period irregular is much less clear-cut than is often as- while others transit towards more permanent sumed. For example, it is common for people to arrangements, the number of people on tempo- enter a country legally, then work despite lacking rary visas at any given moment is much smaller a permit to do so.20 In some island states, such than the aggregate flows suggest. Indeed, 83 per- as Australia and Japan, overstaying is practically cent of the foreign-born population in OECD the only channel to irregular entry; even in many countries has lived there for at least five years.18 European countries, overstay appears to account Almost all temporary migrants come for work- for about two thirds of unauthorized migra- related reasons. Some enter into ‘circular’ ar- tion. In OECD countries, people with irregular rangements, whereby they repeatedly enter and residence or work status tend to be workers with leave the destination country to carry out sea- low levels of formal education. 21 The best esti- sonal or temporary work, effectively maintaining mates of the number of irregular migrants in the two places of residence.19 United States amount to about 4 percent of the Box 2.2 Conflict-induced movement and trafficking People affected by conflict and insecurity can suffer some of the million in Sudan, 2.8 million in Iraq and 1.4 million in the Democratic worst human development outcomes of all migrants. The number of Republic of the Congo. people who move as a result of conflict is significant: at the begin- It is much harder to ascertain the magnitude of human traffick- ning of 2008, there were around 14 million refugees falling under the ing. In fact, there are no accurate estimates of the stocks and flows mandate of either UNHCR or the United Nations Relief and Works of people who have been trafficked. Among the reasons for this are Agency for Palestine Refugees in the Near East (UNRWA), accounting the fact that trafficking data are commonly mixed with data on other for roughly 7 percent of all international migration. The vast majority forms of illegal migration or migrant exploitation, the inherent chal- of refugees originate in and relocate to the poorer countries of the lenges in distinguishing between what is voluntary and forced, and world: in Asia and Africa refugees account respectively for 18 and 13 the very nature of human trafficking as a clandestine and criminal percent of all international migrants. activity. Many of the frequently cited figures are disputed by the coun- Even more individuals displaced by violence and conflict relo- tries concerned, and there is a significant gap between estimated cate within the borders of their country. It is estimated that, in 2009, numbers and identified cases. internally displaced persons number some 26 million, including 4.9 Source: IDMC (2009b), Carling (2006), Kutnick, Belser, and Danailova-Trainor (2007), de Haas (2007) and Lazcko (2009). 26
  • 37. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 2 population or 30 percent of total migrants.22 A facilitate movement. In other words, poorer Movement both within recent research project funded by the European people may decide to take the risk of migrat- and between nations is Commission estimated that in 2005 irregular ing as they hear news of others’ success and predominantly driven migrants accounted for 6–15 percent of the total become more confident that they will receive by the search for better stock of migrants, or about 1 percent of the popu- the support they need in order to succeed them- opportunities lation of the European Union.23 Some of these selves. Other commissioned research generated migrants are counted in official estimates of mi- education profiles for internal migrants across gration, but many are not (box 2.1). 34 developing countries. This showed that mi- The over-representation of skilled, working- grants were more likely than non-migrants to age people in migrant populations is one aspect complete secondary school, reflecting both se- of migrant selectivity. Not only do migrants tend lectivity and better outcomes among migrant to have higher income-earning capacity than children (chapter 3). 28 non-migrants but they often also appear to be What else do we know about the relation- healthier and more productive than natives of ship between internal and international migra- the destination country with equivalent educa- tion? Internal migration, particularly from rural tional qualifications. Migrant selectivity usually to urban areas, can be a first step towards inter- reflects the effect of economic, geographical or national migration, as found by some studies policy-imposed barriers that make it harder for in Mexico, Thailand and Turkey, but this is far low-skilled people to move. This is most evident from being a universal pattern.29 Rather, emigra- in terms of formal education. Tertiary graduates, tion may foster subsequent internal migration in for example, make up 35 percent of working- the home country. In Albania, migration flows to age immigrants to the OECD but only about 6 Greece in the early 1990s generated remittances, percent of the working-age population in non- which helped to finance internal migration to OECD countries.24 Immigrants to the OECD urban centres; in India, international movers from developing countries tend to be of working from the state of Kerala have freed up positions age: for example, over 80 percent of those from in their areas of origin and their remittances sub-Saharan Africa fall into this group. 25 spurred a construction boom that has attracted What do we know about migrant selectivity low-skilled migrants from surrounding areas.30 in developing countries? When the migration Comparisons between internal and interna- process is more selective, individuals of work- tional migration can yield useful insights into ing age (who have higher earning capacity than the causes and implications of human movement. those out of the labour force) form a large pro- For example, background research for this report portion of movers. Using census data, we com- analysed the relationship between the size of the pared the age profiles of migrants to people in place of origin (as measured by its population) their countries of origin in 21 developing and 30 and skilled labour flows and found that the pat- developed countries. We found a significant dif- terns were broadly similar across countries as well ference between the age profile of immigrants in as within them. In particular, emigration rates for developed countries and that of their countries skilled workers are higher in small localities than of origin: 71 percent of migrants in developed in large ones, just as they are higher in small coun- countries are of working age, as opposed to 63 tries than in large ones.31 These patterns reflect percent of the population in their origin coun- the importance of human interaction in driving tries; in contrast, the difference is negligible in movement. Movement both within and between developing countries (63 versus 62 percent). nations is predominantly driven by the search for New evidence on internal migration paints better opportunities, and in many cases—in par- a more complex picture of migrant selectivity. ticular those involving skilled labour—oppor- In Kenya, for example, commissioned research tunities will be greater in places where there are found a positive relationship between measures other people with complementary skills. This is of human capital and migration, 26 which tends one of the reasons why people gravitate to urban to diminish with successive cohorts of migrants centres, and why high-skilled professionals often over time, 27 a result that is consistent with the move to cities and places where their profession is development of social and other networks that already well established.32 27
  • 38. 2 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Unfortunately, Despite our ability to establish these broad processes. Adding questions on country of birth migration data remain contours of movement, what we know is dwarfed or country of previous residence to the national weak. It is much easier by what we don’t know. Unfortunately, migra- census would be a low-cost way forward for many for policy makers to tion data remain weak. It is much easier for countries. Another would be the public release count the international policy makers to count the international move- of existing labour force data, including coun- movements of shoes ments of shoes and cell-phones than of nurses try of birth, as Brazil, South Africa, the United and cell-phones and construction workers. Most of our informa- States and some other countries already do. Yet than of nurses and tion is based on censuses, but these do not pro- another would be the inclusion of standard mi- construction workers vide time series of migration flows that would gration questions in household surveys in coun- enable trends to be recognized nor key data for tries where migration has grown in importance. assessing the impact of migration, such as the These improvements are worthy of government income and other characteristics of migrants at attention and increased development assistance. the time of admission. Population registers can produce such time series, but very few countries 2.2 Looking back have registers with that capacity. Policy makers We now consider how human movement has typically require information about migrant shaped world history. Doing so sheds light on admissions by type (e.g. contract workers, train- the extent to which earlier movements differed ees, family members, skilled professionals, etc.), from or were similar to those of today. It will so administrative data reflecting the number of also reveal the role of migration in the structural visas and permits granted to different types of transformation of societies, the forces that drive migrants are important. Yet none of these data migration and the constraints that frustrate it. sources can answer questions about the social or We then present a more detailed discussion of economic impact of international migration. the evolution of internal and international move- Advances have been made in recent years. ments during the 20th century, with a focus on The OECD, the UN, the World Bank and other the post-World War II era. The analysis of trends agencies have compiled and published census and during the past 50 years is key to understanding administrative databases that shed new light on the factors causing recent changes in migration some aspects of global flows of people. But pub- patterns and how we can expect these to con- lic data still cannot answer basic questions, such tinue evolving in the future. as: how many Moroccans left France last year? What are the occupations of Latin Americans 2.2.1 The long-term view who took up United States residency in 2004? Despite the widespread perception that inter- How has the number of Zimbabweans going national migration is associated with the rise to South Africa changed in recent years? How of globalization and trade in the late 20th cen- much return or circular migration occurs glob- tury, large-scale long-distance movements were ally, and what are the characteristics of those mi- prevalent in the past. At the peak of Iberian grants? For the most part, migration data remain rule in the Americas, more than half a million patchy, non-comparable and difficult to access. Spaniards and Portuguese and about 700,000 Data on trade and investment are vastly more de- British subjects went to the colonies in the tailed. Many aspects of human movement simply Americas.34 Through the brutal use of force, remain a blind spot for policy makers. 11–12 million Africans were sent as slaves across While some data limitations are difficult to the Atlantic between the 15th and late 19th cen- overcome—including the problem of accurately turies. Between 1842 and 1900, some 2.3 mil- estimating the number of irregular migrants— lion Chinese and 1.3 million Indians travelled others should be surmountable. A logical first as contract labourers to South-East Asia, Africa step is to ensure that national statistics offices fol- and North America.35 At the close of the 19th low international guidelines, such that every cen- century the fraction of foreign-born residents in sus contains core migration questions.33 Existing many countries was higher than today.36 surveys could be slightly expanded, or existing Going back further in time, we find human administrative data compiled and disseminated, movement has been a pervasive phenomenon to increase public information on migration throughout history, present in nearly every 28
  • 39. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 2 community for which historical or archaeo- emigrants over a decade accounted for 14 percent Population movements logical evidence is available. Recent DNA tests of the Irish population, 1 in 10 Norwegians, and have played a vital support previous fossil evidence that all human 7 percent of the populations of both Sweden and role in the structural beings evolved from a common ancestor from the United Kingdom. In contrast, the number transformation equatorial Africa, who crossed the Red Sea into of lifetime emigrants from developing countries of economies Southern Arabia approximately 50,000 years today is less than 3 percent of the total popula- throughout history ago.37 While encounters among different societ- tion of these countries. This historical episode ies often led to conflict, the peaceful coexistence was partly driven by falling travel costs: between of immigrants in foreign lands is also recorded. the early 1840s and the late 1850s, passenger An ancient Babylonian tablet from the 18th fares from Britain to New York fell by 77 per- century BCE, for example, talks about a com- cent in real terms.43 There were other determin- munity of migrants from Uruk who fled their ing factors in particular cases, such as the potato homes when their city was raided and, in their famine in Ireland. These population movements new home, met little resistance to their cultural had sizeable effects on both source and destina- practices, with their priests being allowed to tion countries. Workers moved from low-wage inhabit the same quarters as those venerating labour-abundant regions to high-wage labour- local gods.38 The idea that migrants should be scarce regions. This contributed to significant treated according to basic norms of respect is economic convergence: between the 1850s and found in many ancient religious texts. The Old World War I, real wages in Sweden rose from 24 Testament, for example, states that “the alien to 58 percent of those of the United States, while, living with you must be treated as one of your over the same period, Irish wages rose from 61 to native-born,” whereas the Koran requires the 92 percent of those in Great Britain. According faithful to move when their beliefs are in danger to economic historians, more than two thirds of and to give aman (refuge) to non-Muslims, even the wage convergence across countries that oc- if they are in conflict with Muslims.39 curred in the late 19th century can be traced to Population movements have played a vital role the equalizing effect of migration.44 in the structural transformation of economies Remittances and return migration were also throughout history, thereby contributing greatly very important in the past. Remittances were to development. Genetic and archaeological evi- sent by courier and through transfers and notes dence from the Neolithic period (9500–3500 via immigrant banks, mercantile houses, postal BCE) suggests that farming practices spread with services and, after 1900, by telegraph wire. It is the dispersal of communities after they had mas- estimated that the average British remitter in tered the techniques of cultivation.40 The British the United States in 1910 sent up to a fifth of Industrial Revolution both generated and was his income back home, and that about a quar- fuelled by rapid urban growth, driven mainly ter of European migration to the United States by movement from the countryside.41 The share around that time was financed through remit- of rural population has declined markedly in all tances from those already there.45 Return migra- economies that have become developed, falling tion was often the norm, with estimated rates of in the United States from 79 percent in 1820 to return from the United States ranging as high as below 4 percent by 1980, and even more rapidly 69 percent for Bulgaria, Serbia and Montenegro in the Republic of Korea, from 63 percent in 1963 and 58 percent for Italy.46 In Argentina, Italian to 7 percent in 2008.42 immigrants were often referred to as golondrinas An interesting episode from the standpoint (swallows) because of their tendency to return, of our analysis was that of the large flows from and a contemporary observer wrote that “the Europe to the New World during the second Italian in Argentina is no colonist; he has no half of the 19th century. By 1900, more than a house, he will not make a sustenance… his only million people were moving out of Europe each hope is a modest saving.”47 year, spurred by the search for better conditions These population movements were enabled in the face of hunger and poverty at home. The by a policy stance that was not only receptive to size of these flows is staggering by contemporary migration but in many cases actively encouraged standards. At its peak in the 19th century, total it. This is as true of origin countries, which often 29
  • 40. 2 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development subsidized passage in order to reduce pressures at in societies where intolerance of minorities was home, as it was of destination governments, which prevalent and socially accepted to a far larger invited people to come in order to consolidate set- extent than today.51 It is also a useful reminder tlements and take advantage of natural resources. that the barriers to migration that characterize For example, by the 1880s about half of migrants many developed and developing countries today to Argentina received a travel subsidy, while a law are much less an immutable reality than might passed in Brazil in 1850 allotted land to migrants at first be supposed. free of charge.48 More generally, the late 19th cen- tury was marked by the absence of the plethora 2.2.2 The 20th century of mechanisms to control international flows of The pro-migration consensus was not to last. people that subsequently emerged. Until the pas- Towards the end of the 19th century, many coun- sage of restrictive legislation in 1924, for example, tries introduced entry restrictions. The causes there was not even a visa requirement to settle per- were varied, from the depletion of unsettled land manently in the United States, and in 1905, only to labour market pressures and popular senti- 1 percent of the one million people who made the ment. In countries such as Argentina and Brazil transatlantic journey to Ellis Island were denied the policy shift occurred through the phasing out entry into the country.49 of subsidies; in Australia and the United States it One key distinction between the pre-World came through the imposition of entry barriers.52 War I period and today lies in the attitudes of Despite the introduction of these restrictions, destination governments. While anti-immigrant estimates from the early 20th century indicate sentiment could run high and often drove the that the share of international migrants in the erection of barriers to specific kinds of move- world’s population was similar if not larger than ment, the prevailing view among governments it is today. This is especially striking given the was that movement was to be expected and was relatively high transport costs at that time.53 ultimately beneficial to both origin and destina- There was nothing in the area of migration tion societies.50 This is all the more remarkable policy even remotely resembling the rapid mul- tilateral liberalization of trade in goods and movements of capital that characterized the Table 2.1 Five decades of aggregate stability, with regional shifts post-World War II period.54 Some countries en- Regional distribution of international migrants, 1960–2010 tered bilateral or regional agreements to respond 1960 2010 to specific labour shortages, such as the United States’ 1942 Mexican Farm Labour (Bracero) Total Share Total Share migrants of world Share of migrants of world Share of Program, which sponsored 4.6 million contracts migrants population migrants for work in the United States over a 22-year pe- (millions) (millions) population riod,55 the 1947 United Kingdom–Australia World 74.1 2.7% 188.0 2.8% (excluding the former Soviet Union and Assisted Passage Agreement, or the flurry of former Czechoslovakia) European labour movement agreements and BY REGION guest-worker programmes.56 But early enthusi- Africa 9.2 12.4% 3.2% 19.3 10.2% 1.9% asm for guest-worker programmes had fizzled Northern America 13.6 18.4% 6.7% 50.0 26.6% 14.2% out by the 1970s. The United States phased out Latin America and the Caribbean 6.2 8.3% 2.8% 7.5 4.0% 1.3% Asia 28.5 38.4% 1.7% 55.6 29.6% 1.4% its Bracero Program in 1964, and most Western GCC states 0.2 0.3% 4.6% 15.1 8.0% 38.6% European countries that had heavily relied on Europe 14.5 19.6% 3.5% 49.6 26.4% 9.7% guest-worker programmes ceased recruitment Oceania 2.1 2.9% 13.5% 6.0 3.2% 16.8% during the 1970s oil shock.57 BY HUMAN DEVELOPMENT CATEGORY This lack of liberalization is consistent with Very high HDI 31.1 41.9% 4.6% 119.9 63.8% 12.1% the observed stability in the global share of mi- OECD 27.4 37.0% 4.2% 104.6 55.6% 10.9% High HDI 10.6 14.2% 3.2% 23.2 12.3% 3.0% grants. As shown in table 2.1, this share (which Medium HDI 28.2 38.1% 1.7% 35.9 19.1% 0.8% excludes Czechoslovakia and the former Soviet Low HDI 4.3 5.8% 3.8% 8.8 4.7% 2.1% Union for comparability reasons—see below) has inched up from 2.7 to 2.8 percent between Source: HDR team estimates based on UN (2009d). Note: Estimates exclude the former Soviet Union and former Czechoslovakia. 1960 and 2010. The data nonetheless reveal a 30
  • 41. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 2 remarkable shift in destination places. The share picture of bilateral flows over time, but figure 2.4 in developed countries more than doubled, displays the evolution of the share of people from from 5 percent to more than 12 percent.58 An developing countries in eight developed econo- even larger increase—from 5 to 39 percent of the mies that have comparable information. In all population—occurred in the GCC countries, but one case (the United Kingdom), there were which have experienced rapid oil-driven growth. double-digit increases in the share of migrants In the rest of the world, however, the fraction of from developing countries.60 In many European foreign-born people has been stable or declining. countries, this shift is driven by the increase The declines are most marked in Latin America in migrants from Eastern European countries and the Caribbean, where international migra- classed as developing according to their HDI. tion has more than halved, but are also evident For example, during the 1960s only 18 percent in Africa and the rest of Asia. of developing country immigrants into Germany An important caveat is that these trends came from Eastern Europe; 40 years later that exclude two sets of countries for which it is dif- ratio was 53 percent. ficult to construct comparable time series on In developing countries, the picture is more international migrants, namely the states of the mixed, although data are limited. We can com- former Soviet Union, and the two components pare the source of migrants today and several of former Czechoslovakia. The independence of decades ago for a few countries, revealing some these new nations generated an artificial increase interesting contrasts (figure 2.5). In Argentina in the number of migrants, which should not be and Brazil, the decline in the share of foreign- interpreted as a real increase in the prevalence of born people was driven by a fall in those com- international movement (box 2.3).59 ing from the poorer countries of Europe, as Where are recent migrants to developed those countries experienced dramatic post- countries coming from? We do not have a full war growth while much of Latin America Box 2.3 Migration trends in the former Soviet Union When the Soviet Union broke up in 1991, 28 million people be- Any attempt to understand trends in the former Soviet Union came international migrants overnight—even if they hadn’t moved should use comparable territorial entities. One way to do this is to an inch. This is because statistics define an international migrant consider inter-republic migration before and after the break-up. In this as a person who is living outside their country of birth. These peo- approach, anyone who moved between two republics that would later ple had moved within the Soviet Union before 1991 and were now become independent nations would be considered an international classified as foreign-born. Without their knowing it, they were now migrant. Thus, a Latvian in St. Petersburg would be classified as an ‘statistical migrants’. international migrant both before and after 1991. At one level, the reclassification makes sense. A Russian in Minsk In background research for this report, Soviet census data were was living in the country of her birth in 1990; by the end of 1991, she used to construct such a series. Thus defined, the share of foreign-born was technically a foreigner. But interpreting the resulting increase in people in the republics of the USSR rose slightly from 10 percent in 1959 the number of migrants as an increase in international movement, to 10.6 percent in 1989. After 1990, there were divergent trends across as some authors have done, is mistaken. Hence we have excluded the different states. In the Russian Federation, which became some- them, together with migrants in the former Czechoslovakia, from the thing of a magnet in the region, the migrant stock increased from 7.8 to calculation of trends in table 2.1. 9.3 percent of the population. For Ukraine and the three Baltic states, Has human movement increased in the former Soviet Union since migrant shares declined, as large numbers of foreign-born people left. 1991? On the one hand, the relaxation of propiska controls increased In all the other states of the former Soviet Union, the absolute number of human mobility. On the other, the erection of national boundaries may migrants declined until 2000 and in most cases the migrant share of the have reduced the scope for movement. The picture is further com- population also declined. Thus, while 30.3 million foreign-born people plicated by the fact that many movements after 1991 were returns to lived in the territory of the Soviet Union at the time of its dissolution, the the region of origin: for example, people of Russian origin returning aggregate number fell to 27.4 million in 2000 and to 26.5 million in 2005, from central Asia. as many in the post-Soviet space chose to return home. Source: Heleniak (2009), UN (2002), Zlotnik (1998), and Ivakhnyuk (2009). 31
  • 42. 2 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Figure 2.4 An increasing share of migrants come from developing countries A levelling off or even a decline in internal Share of migrants from developing countries in selected migration flows is to be expected in developed developed countries and high-HDI countries, where past flows were associated with rapid urbanization that has now abated. But in many developing countries Australia 1960–1969 urbanization has not slowed and is expected to 1990–2004 continue. In fact, estimates from UNDESA sug- Belgium gest that the urban share of the world’s popula- tion will nearly double by 2050 and will increase from 40 percent to over 60 percent in Africa. Canada Urbanization is spurred in part by natural popu- lation growth in urban areas, alongside migration Germany from rural areas and from abroad. Although it is difficult to determine the precise contributions of these different sources, it is clear that migration is New Zealand an important factor in many countries.61 Urbanization can be associated with major Sweden challenges to city dwellers and the government authorities responsible for urban planning and service provision. The most visible of these chal- United Kingdom lenges is the 2 billion people—40 percent of urban residents—who are expected to be living United States in slums by 2030.62 As is well known, living con- ditions are often very poor in the slums, with in- | | | | | | | | | adequate access to safe water and sanitation and 0 10 20 30 40 50 60 70 80 insecure land tenure. As we discuss in chapters Share of all migrants (%) 4 and 5, it is important that urban local authori- ties be accountable to residents and adequately Source: HDR team estimates based on UN (2006a). financed to tackle these challenges, since local planning and programmes can play a critical role in improving matters. stagnated. In contrast, the rise in the immigra- In sum, the period since 1960 has been tion rate in Costa Rica was driven by large flows marked by a growing concentration of migrants of Nicaraguan migrants, while the reduction in in developed countries against a background of Mali reflects significant declines in immigration aggregate stability in overall migration. How do from Burkina Faso, Guinea and Mauritania. we explain these patterns? Our research shows Many countries have experienced increases that three key factors—trends in income, popu- in internal migration, as shown in figure 2.6. lation and transport costs—tended to increase However, this trend is far from uniform. For movement, which simultaneously faced an in- the 18 countries for which we have comparable creasingly significant constraint: growing legal information over time, there is an increasing and administrative barriers. trend in 11 countries, no clear trend in four, Divergence in incomes across regions, com- and a decline in two developed countries. The bined with a general increase in incomes around average rate of increase for this set of countries most of the world, is a major part of the expla- is around 7 percent over a decade. However, our nation of movement patterns. The evolution of research has also found that the share of recent income inequalities shows remarkable diver- migrants (defined as those who have moved gence between most developing and developed between regions in the past five years) has not regions, even if the East Asia–Pacific and South increased in most countries in our sample, indi- Asia regions have seen a mild convergence (figure cating a possible stabilization of internal migra- 2.7, panel A).63 China presents an exception to tion patterns. the broad pattern of lack of convergence, with 32
  • 43. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 2 national per capita income rising from 3 to 14 examination of the role that these barriers play percent of the developed country average be- in shaping and constraining movement today. tween 1960 and 2007.64 Overall, the data indi- cate that income incentives to move from poor to 2.3 Policies and movement rich countries have strongly increased.65 Since the emergence of modern states in the 17th Attempts to account for this divergence have century, the international legal system has been generated a vast literature, in which differences built on the bedrock of two principles: sovereignty in labour and capital accumulation, technologi- and territorial integrity. Within this system, cal change, policies and institutions have all been which includes a series of norms and constraints investigated.66 Whatever the ultimate driving imposed by international law, governments police forces, one of the key contributing factors has their country’s borders and enforce their right to been differing population growth rates. As is restrict entry. This section discusses the different well known, between 1960 and 2010 the spatial ways in which government policy determines how demographic composition of the world popula- many people to admit, where these people come tion shifted: of the additional 2.8 billion work- from, and what status is accorded to them. ing-aged people in the world, 9 out of 10 were in developing countries. Because labour became much more abundant in developing countries, Figure 2.5 Sources and trends of migration into developing countries wage differentials widened. This meant that Migrants as a share of total population in selected countries, moving to developed countries became more at- 1960–2000s tractive and patterns of movement shifted as a result, despite—as we shall see—the raising of Argentina 1970 high barriers to admission. At the same time, 2001 average income levels in the world as a whole Brazil were increasing, as shown in panel B of figure 1960 2000 2.7 (even if some developing regions also saw pe- riods of decline). Since poverty is an important Costa Rica 1960 constraint on movement, higher average incomes 2000 made long-distance movement more feasible. In India other words, as incomes rose, poorer countries 1961 2001 moved up the ‘migration hump’, broadening Indonesia the pool of potential migrants to developed 1971 countries. 1990 Recent declines in transport and communi- Thailand cation costs have also increased movement. The 1970 2000 real price of air travel fell by three fifths between Turkey 1970 and 2000, while the cost of communica- 1965 tions fell massively.67 The real cost of a 3-minute 2000 telephone call from Australia to the United Mali Kingdom fell from about US$350 in 1926 to 1976 1998 US$0.65 in 2000—and, with the advent of in- Rwanda ternet telephony, has now effectively fallen to 1978 zero. 68 Such trends have made it easier than ever 2002 before for people to reach and establish them- | | | | | selves in more distant destinations. 0 2 4 6 8 Given these drivers, we would expect to see Share of population (%) significant growth in international migration in From developing countries From developed countries recent decades. However, this potential has been Origin unknown constrained by increased policy barriers to move- ment, especially against the entry of low-skilled Source: HDR team estimates based on Minnesota Population Center (2008) and national census data for indicated years. applicants. We turn now to a more in-depth 33
  • 44. 2 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Figure 2.6 Internal migration rates have increased only slightly While there is a wealth of qualitative coun- Trends in lifetime internal migration intensity in selected try-level analysis of policies—especially for countries, 1960–2000s developed countries—severe data limitations 35 impede comparisons of policy across countries. Measurement is intrinsically difficult because the United States rules take many forms and are enforced in dif- 30 ferent ways and to varying degrees, with results that are generally not amenable to quantification. 25 In contrast to most aspects of economic policy, for example, national statistical bureaux do not Malaysia measure the effects of migration policy in ways 20 Costa Rica Mexico that are consistent across countries. Most of the measures used in this report have been developed by international research and non-governmental Lifetime migration intensity (%) 15 Brazil Kenya organizations (NGOs), not by national public- 10 Rwanda sector agencies. The measure that covers the largest number of countries and the longest time span comes from 5 India a periodic survey of policy makers conducted by UNDESA, in which governments report their 0 | | | | | | views and responses to migration. The survey 1960 1970 1980 1990 2000 2010 covers 195 countries and reflects the views of Source: Bell and Muhidin (2009). policy makers regarding the level of immigration and whether their policy is to lower, maintain or raise future levels. While it is a self-assessment, and official intentions rather than practice are in- Table 2.2 Policy makers say they are trying to maintain existing dicated, some interesting patterns emerge (table immigration levels 2.2). In 2007, some 78 percent of respondent Views and policies towards immigration by HDI category, 2007 governments viewed current immigration levels as satisfactory, while 17 percent felt them to be Government’s view on immigration Policy on immigration too high and 5 percent too low. A similar picture HDI Satis- Lower Maintain Raise No inter- emerges when governments are asked to describe categories Too high factory Too low Total levels levels levels vention Total their policies. On both questions, developed country governments appear to be more restric- VERY HIGH HDI tive than those of developing countries. No. of Countries 7 26 6 39 7 24 7 1 39 These patterns indicate a significant gap Percent (%) 18 67 15 100 18 62 18 3 100 between the policies that the public appears to HIGH HDI favour in most countries—namely greater re- No. of Countries 6 40 1 47 9 37 1 0 47 strictions on immigration—and actual policies, Percent (%) 13 85 2 100 19 79 2 0 100 which in fact allow for significant amounts of MEDIUM HDI immigration.69 While explanations for this gap No. of Countries 17 62 4 83 18 47 3 15 83 are complex, several factors likely come into play. Percent (%) 20 75 5 100 22 57 4 18 100 The first is that opposition to immigration is LOW HDI not as monolithic as first appears, and voters often No. of Countries 4 22 0 26 4 6 0 16 26 have mixed views. As we show below, in many Percent (%) 15 85 0 100 15 23 0 62 100 countries, concerns about adverse employment or TOTAL fiscal effects are mixed with the recognition that No. of Countries 34 150 11 195 38 114 11 32 195 tolerance of others and ethnic diversity are posi- Percent (%) 17 77 6 100 19 58 6 16 100 tive values. Second, organized groups such as la- Source: UN (2008b). bour unions, employer organizations and NGOs can have a significant effect on the formulation 34
  • 45. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 2 of public policies; in many cases these groups do developing countries were much more likely to not advocate for tight restrictions to immigra- have temporary regimes for low-skilled workers. tion. Third, many governments implicitly toler- Rules concerning changes in visa status ate irregular migration, suggesting that policy and family reunion differ widely across coun- makers are aware of the high economic and social tries.74 Some temporary schemes offer a path to costs of a crackdown. For example, in the United States employers are not required to verify the authenticity of immigration documents, but Figure 2.7 Global income gaps have widened must deduct federal payroll taxes from migrants’ Trends in real per capita GDP, 1960–2007 pay: through this mechanism, illegal immigrant workers provide around US$7 billion annually to Panel A: Ratio of income of developing countries to income of developed countries the US Treasury.70 50 For the purposes of this report, we sought to address existing gaps in knowledge by working Ratio of regional to developed per capita GDP (%) with national migration experts and the IOM to 40 conduct an assessment of migration policies in 28 Eastern Europe countries.71 The key value added of this exercise lies in the coverage of developing countries (half 30 the sample), which have typically been excluded Latin America from such assessments in the past, and the rich and the Caribbean information we collected on aspects such as ad- 20 missions regimes, treatment and entitlements, China and enforcement. Asia, excl. China 10 Africa Comparing the migration policy regimes of Oceania developed and developing countries reveals strik- ing differences as well as similarities. Some of the 0 restrictions commonly noted (and criticized) 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 in developed countries are also present in many developing countries (figure 2.8). The regimes in Panel B: Real per capita income of developing countries by region both groups of countries are biased in favour of high-skilled workers: 92 percent of developing 12 Eastern Europe and all of developed countries in our sample were open to temporary skilled migrants; for perma- 10 nent skilled migration, the corresponding figures were 62 and 93 percent. In our country sample, Latin America 38 percent of developing and half of developed and the Caribbean 8 countries were closed to permanent migration of unskilled workers.72 Temporary regimes have long been used 6 and most countries provide such permits. These GDP per capita (US$ thousands) programmes stipulate rules for the time-bound China admission, stay and employment of foreign 4 Asia, excl. China workers. The H1B visas of the United States, for instance, grant temporary admission to high- Africa skilled workers for up to six years, while H2B 2 Oceania visas are available for low-skilled seasonal work- ers for up to three. Similarly, Singapore’s im- migration policy has Employment Passes—for 0 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 skilled professionals—and a Work Permit or R-Pass for unskilled or semi-skilled workers.73 Source: HDR team estimates based on World Bank (2009b) and Heston, Summers, and Aten (2006). Among the countries in our policy assessment, 35
  • 46. 2 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Figure 2.8 Welcome the high-skilled, rotate the low-skilled Kafala programmes are restrictive on several Openness to legal immigration in developed versus counts, including family reunification. Human developing countries, 2009 rights abuses—including non-payment of wages Panel A: Permanent immigration and sexual exploitation of domestic workers— are well documented, especially among the in- High-skilled creasing share of migrants originating in the Developed Indian subcontinent.77 Developing In recent years, some countries in the region have taken moderate steps in the direction of Developed Low-skilled reforming their immigration regimes. Saudi Arabia recently passed a series of regulations Developing facilitating the transfer of workers employed by | | | | | | companies providing services (e.g. maintenance) 0 20 40 60 80 100 Share of countries in sample (%) to government departments.78 Other initiatives have also been implemented to monitor the liv- ing and working conditions of foreign migrants. Panel B: Temporary immigration In the United Arab Emirates, the Ministry of High-skilled Labour has introduced a hotline to receive com- Developed plaints from the general public. In 2007, the Developing authorities inspected 122,000 establishments, resulting in penalties for almost 9,000 violations Low-skilled of workers’ rights and of legislation on working Developed conditions. However, more ambitious proposals Developing for reform, such as Bahrain’s proposal in early | | | | | | 2009 to abolish the kafala system, have floun- 0 20 40 60 80 100 dered, reportedly in the face of intense political Share of countries in sample (%) opposition by business interests.79 In some developed countries—including Open Partially closed Totally closed Australia, Canada and New Zealand—the pref- Source: Klugman and Pereira (2009). erence for high-skilled workers is implemented through a points system. The formulae take into account such characteristics as education, occu- long-term or even permanent residence and allow pation, language proficiency and age. This con- foreign workers to bring in their dependents. An fers some objectivity to what otherwise might example is the US’s H2B visas, although their seem an arbitrary selection process, although annual number is capped at a low level and the other countries attract large numbers of gradu- dependents are not entitled to work. Other gov- ates without a point-based system.80 ernments explicitly prohibit status change and Points systems are uncommon in developing family reunion, or severely restrict them. countries. Formal restrictions on entry include The temporary worker or kafala (literally requirements such as a previous job offer and, in meaning ‘guaranteeing and taking care of ’ in some cases, quotas. One aspect on which devel- Arabic) programmes of the GCC countries are a oping countries appear to be relatively restrictive special case.75 Under these programmes, foreign is family reunification. About half the develop- migrant workers receive an entry visa and resi- ing countries in our sample did not allow the dence permit only if a citizen of the host country family members of temporary immigrants to sponsors them. The khafeel, or sponsor-employer, come and work—as opposed to one third of de- is responsible financially and legally for the veloped countries. worker, signing a document from the Labour Family reunification and marriage migration Ministry to that effect.76 If the worker is found represent a significant share of inflows into virtu- to have breached the contract, they have to leave ally all OECD countries. Indeed, some countries the country immediately at their own expense. are dominated by flows linked to family ties, as in 36
  • 47. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 2 France and the United States, where these account Figure 2.9 Enforcement practices vary for 60 and 70 percent of annual flows respectively. Interventions and procedures regarding irregular migrants, 2009 While it is common to distinguish between fam- Interventions to detect irregular migrants ily reunification and labour migration, it is impor- tant to note that family migrants often either have Border control or can acquire authorization to work. Raids by law Of course the stated policy may differ from enforcement agencies what happens in practice. Significant variations exist in migration law enforcement across coun- Random police checks tries (figure 2.9). In the United States, research Information gained has found that border enforcement varies over from service providers (e.g. schools) the economic cycle, increasing during recessions Procedure after detecting irregular migrants and easing during expansions.81 In South Africa, deportations more than doubled between 2002 None and 2006 without a change in legislation, as the police force became more actively involved in Migrant is fined enforcement.82 Our policy assessment suggested that while developing countries were somewhat Employer is fined less likely to enforce border controls and less likely to detain violators of immigration laws, Migrant is detained other aspects of enforcement including raids by law enforcement agencies and random checks, as Migrant is deported well as fines, were at least as frequent as in de- | | | | | veloped countries.  Lower institutional capacity 0 1 2 3 4 may explain part of this variation.  Even after Average score (1 = Never or rare; 5 = Almost always) detection, developing countries are reportedly Developed countries Developing countries more likely to do nothing or simply to impose fines on irregular migrants.  In some countries, Source: Klugman and Pereira (2009). courts weigh family unity concerns and the strength of an immigrant’s ties to a country in deportation procedures.83 Further discussion of Policy Index (MIPEX) measures policies to in- the role of enforcement in immigration policies tegrate migrants in six policy areas (long-term is provided in chapter 5. residence, family reunion, citizenship, political One question that emerges from these rules participation, anti-discrimination measures and on entry and treatment, which can be investi- labour market access). gated using cross-country data, is whether there Our analysis suggests that there is no system- is a ‘numbers versus rights’ trade-off. It is pos- atic relation between various measures of rights sible that countries will open their borders to and migrant numbers (figure 2.10). Comparison a larger number of immigrants only if access to with the EIU index (panel A), which has a some basic rights is limited. This could arise if, broader sample of developed and developing for example, immigration is seen to become too countries, reflects essentially no correlation be- costly, so that neither voters nor policy makers tween the number of migrants and their access to will support it.84 Data on the treatment of im- basic rights, suggesting that the various regimes migrants allow us to empirically examine this governing such access are compatible with both question. The Economist Intelligence Unit high and low numbers of migrants. Restricting (EIU) has created an accessibility index for 61 the analysis to the smaller sample of countries countries (34 developed, 27 developing) that covered by the MIPEX allows us to take advan- summarizes official policy in terms of ease of tage of OECD data, which distinguish the share hiring, licensing requirements, ease of fam- of immigrants with low levels of formal educa- ily reunification and official integration pro- tion from developing countries. Again, we find grammes for migrants. The Migrant Integration essentially no correlation (panel B). For example, 37
  • 48. 2 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Figure 2.10 Cross-country evidence shows little support for the countries like Poland and Ireland have very low ‘numbers versus rights’ hypothesis shares of low-skilled workers from developing Correlations between access and treatment countries, yet score poorly in the MIPEX. We have also found that countries that have seen in- Panel A: Foreign-born migrants and EIU accessibility score, 2008 creases in their migrant shares over time did not 85 Australia curtail the rights and entitlements provided to Canada Australia immigrants. 85 For example, between 1980 and Singapore Canada Singapore 2005 the share of immigrants in Spain increased New Zealand New Zealand from 2 to 11 percent; during the same period the 75 Portugal Israel Spanish government extended the provision of Israel Belgium PortugalUnited States emergency and non-emergency health care to ir- Hong Kong (China) Chile Belgium SpainUnited States Hong Kong (China) regular migrants.86 United Kingdom Spain Thailand Chile Ireland United Kingdom Similar results were found in our policy as- ThailandBrazil 65 Peru Ireland Switzerland sessment, which allowed us to distinguish be- Brazil Poland Germany Nigeria Peru Venezuela Hungary Poland Malaysia Switzerland tween different components of migration policy. Germany Nigeria Argentina Venezuela Italy In fact, if there was any indication of a correla- Migrant accessibility score Ukraine Malaysia Kazakhstan Hungary Mexico Argentina Italy Ukraine d’Ivoire China India Mexico Côte Kazakhstan Russian Federation tion, it was often the opposite of that proposed Romania Turkey China France d’Ivoire Côte India Romania Turkeyof Korea Russian Federation Greece by the numbers versus rights hypothesis. What 55 Republic France Qatar South AfricaGreece Republic of Korea Jordan the data reveal is that, in general, across many Qatar Japan South Africa Latvia Jordan United Arab Emirates measures, developing countries have lower me- Japan Latvia Kuwait United Arab Emirates dian shares of foreign-born workers and lower Ghana Kuwait Ghana Estonia protection of migrant rights. Developed coun- 45 Iran Estonia Saudi Arabia tries, which have more migrants, also tend to | | Saudi Arabia | | | 0 Iran 20 40 60 80 have rules that provide for better treatment of Share of migrants in population (%) migrants. For example, India has the lowest score on provision of entitlements and services to in- ternational migrants in our assessment, but has Panel B: Low-skilled foreign-born migrants in OECD and MIPEX aggregate score an immigrant share of less than 1 percent of the Sweden population; Portugal has the highest score while Sweden having an immigrant share of 7 percent. 85 Policies towards migration are not deter- Portugal mined solely at the national level. Supra-national Portugal agreements, which can be bilateral or regional 75 in nature, can have significant effects on mi- Finland Netherlands Belgium Canada gration flows. Regional agreements have been Belgium Finland Netherlands Italy Norway Canada established under various political unions, such 65 Italy United Norway Kingdom as the Economic Community of West African Index of migrant integration (MIPEX) Spain Spain United Kingdom States (ECOWAS), the European Union and the Luxembourg France Mercado Común del Sur (MERCOSUR), while Ireland Germany 55 Ireland Luxembourg France Germany Switzerland a good example of a bilateral agreement is that of Hungary Czech Republic Switzerland the Trans-Tasman Travel Arrangement between Poland Hungary Czech Republic Denmark Australia and New Zealand. These agreements 45 Poland Denmark have had significant effects on migration flows Slovakia Greece Slovakia Greece Austria between signatory countries. They are most Austria likely to allow freedom of movement when par- 35 ticipating member states have similar economic | | | | | | 0 2 4 6 8 10 conditions and when there are strong political Low-skilled migrants from developing countries as share of population (%) or other motivations for socio-economic integra- tion. For the countries in our policy assessment, Source: UN (2009d), The Economist Intelligence Unit (2008), OECD (2009a) and Migration Policy Group and British Council (2007). about half of the special mobility agreements of developed countries were with other developed 38
  • 49. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 2 Box 2.4 Global governance of mobility Beyond a well-established convention on refugees, international negotiations. International discussions have also been characterized mobility lacks a binding multilateral regime. The ILO has long had by lack of cooperation among sending countries. These obstacles conventions on the rights of migrant workers, but they are heavily have so far defied the best efforts of international organizations and undersubscribed (chapter 5). The IOM has expanded beyond its his- a handful of governments to promote cooperation and binding inter- toric role in the post-war repatriation of refugees towards a more gen- national commitments. eral mission to improve migration management and has increased Further liberalization is currently being canvassed in the Doha its membership, but it is outside the UN system and remains largely Round of trade negotiations, which began in 2000 but have long oriented towards service provision to member states on a project since stalled. Existing commitments under GATS are limited, refer- basis. Under the General Agreement on Trade in Services (GATS) of ring mainly to high-skilled workers. GATS also excludes “measures the World Trade Organization (WTO), some 100 member states have affecting natural persons seeking access to the employment market made commitments to the temporary admission of foreign nationals of another country [or] measures regarding citizenship, residence, or who provide services, but these mostly involve business visitor visas employment on a permanent basis”. Nor does GATS apply to perma- for up to 90 days and fixed-term intra-company transfers involving nent migration: most WTO members limit service providers to less high-skilled professionals. than five years in their country. The lack of multilateral cooperation on migration has been at- During the Doha Round it became clear that developing countries tributed to several related factors. In contrast to trade negotiations, want to liberalize the movement of natural persons, whereas indus- where countries negotiate over the reciprocal reduction of barriers to trial countries prefer trade in services. It could be argued that the each other’s exports, developing countries are in a weaker bargain- importance of GATS to labour migration does not lie in the relatively ing position on the migration front. Most migrants from developed small amount of additional mobility facilitated thus far, but rather in countries go to other developed countries, so there is little pressure the creation of an institutional framework for future negotiations. from developed country governments to open channels for entering However, better progress might be made if the WTO took a more developing countries. This asymmetry, as well as the political sen- inclusive and people-centred approach, which allowed greater par- sitivity of the migration issue in most developed destination coun- ticipation by other stakeholders and linked more closely with existing tries, has led to a lack of leadership from these states in international legal regimes for the protection of human rights. Source: Castles and Miller (1993), Neumayer (2006), Leal-Arcas (2007), Charnovitz (2003), p.243, Mattoo and Olarreaga (2004), Matsushita, Schoenbaum, and Mavroidis (2006), Solomon (2009), and Opeskin (2009). countries, while more than two thirds of those for slow implementation range from inconsis- of developing countries were with other de- tency between the protocol and national laws, veloping countries. There are examples where regulations and practices to border disputes and mobility is granted only to some workers, such full-scale wars which have often led to the expul- as the higher skilled. For example, the migra- sion of foreign citizens. 87 tion system of the North American Free Trade We also find restrictions on human move- Agreement (NAFTA) covers only nationals of ment within nations as well as on exit. One Canada, Mexico and the United States who have source of data on these restrictions is the NGO a B.A. degree and a job offer in another member Freedom House, which has collected informa- country. Box 2.4 briefly overviews the multilat- tion on formal and informal restrictions on eral arrangements related to human movement. foreign and internal travel as a component of its However, there can be large differences be- assessment of the state of freedom in the world.88 tween the letter of these agreements and actual The results are striking, particularly given that practice, particularly in countries where the the Universal Declaration of Human Rights rule of law is weak. For example, despite the guarantees the right to move freely within one’s provisions for comprehensive rights of entry, country and to exit and return to one’s own residence and establishment provided for in the country: over a third of countries in the world ECOWAS agreement signed in 1975 (which impose significant restrictions on these freedoms were to be implemented in three phases over a (table 2.3). 15-year period), only the first phase of the proto- Formal restrictions on internal movement are col—elimination of the need for visas for stays present in many countries with a legacy of central up to 90 days—has been achieved. The reasons planning, including Belarus, China, Mongolia, 39
  • 50. 2 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Table 2.3 Over a third of countries significantly restrict the right to move exclude migrants from access to the public ser- Restrictions on internal movement and emigration by HDI category vices and legal rights accorded to local people (chapter 3). Countries can limit exit by nationals from Restrictions on mobility, 2008 their territory by several means, ranging from Most Least formal prohibitions to practical barriers cre- HDI categories restrictive 1 2 3 restrictive Total ated by fees and administrative requirements. Exorbitant passport fees can make it all but VERY HIGH HDI impossible for a poor person to leave the coun- No. of Countries 0 3 1 3 31 38 try through regular channels: a recent study Percent (%) 0 8 3 8 81 100 found that 14 countries had passport fees HIGH HDI that exceeded 10 percent of annual per capita No. of Countries 2 4 4 10 27 47 income.92 In many countries, a labyrinth of Percent (%) 4 9 9 21 57 100 procedures and regulations, often exacerbated MEDIUM HDI by corruption, causes excessive delays and No. of Countries 2 13 24 27 16 82 compounds the costs of leaving. For example, Percent (%) 2 16 29 33 20 100 Indonesian emigrants have to visit numerous LOW HDI government offices in order to acquire the nec- No. of Countries 2 5 13 5 0 25 essary paperwork to leave. Not surprisingly, Percent (%) 8 20 52 20 0 100 these exit restrictions are negatively correlated TOTAL with emigration rates.93 No. of Countries 6 25 42 45 74 192 A handful of countries have formal restric- Percent (%) 3 13 22 23 39 100 tions on exit. These are strictly enforced in Source: Freedom House (2009). Cuba and the Democratic People’s Republic of Korea, and are in place in China, Eritrea, Iran, Myanmar, and Uzbekistan.94 Eritrea, for the Russian Federation and Viet Nam. 89 These example, requires exit visas for citizens and restrictions are costly, time-consuming and foreign nationals and has reportedly denied cumbersome to maintain, as are informal bar- the exit visas of children whose parents (living riers, albeit to a lesser extent. Although many abroad) had not paid the 2 percent tax on for- people in these countries are able to travel eign income.95 Twenty countries restrict the exit without the proper documentation, they later of women—including Myanmar, Saudi Arabia find that they cannot access services and jobs and Swaziland—while eight impose age-specific without them. In several countries, corruption restrictions related to the travel of citizens of is a key impediment to internal movement. military service age.96 Checkpoints on local roads, where bribes are levied, are commonplace in parts of sub-Saha- 2.4 Looking ahead: the crisis and ran Africa. For instance, in Côte d’Ivoire, peo- beyond ple living in northern areas controlled by rebel The future of the global economy is a central groups were routinely harassed and forced to concern for policy makers. Like everyone else, pay US$40–60 when attempting to travel south we hold no crystal ball, but we can examine the to government-controlled areas.90 Examples of impacts and implications of the current crisis corruption were also reported from Myanmar, as the basis for identifying probable trends for the Russian Federation and Viet Nam, where the coming decades. Demographic trends, in bribes were required to process applications for particular, can be expected to continue to play changes in place of residence. In several South a significant role in shaping the pressures for Asian countries, migrants living in urban movement between regions, as we have seen slums face constant threats of clearance, evic- over the past half-century. But new phenomena tion and rent-seeking from government offi- such as climate change are also likely to come cials.91 Internal movement is also impeded by into play, with effects that are much more dif- regulations and administrative procedures that ficult to predict. 40
  • 51. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 2 2.4.1 The economic crisis and the Figure 2.11 Unemployment is increasing in key migrant destinations prospects for recovery Unemployment rates in selected destinations, 2007–2010 Many people are now suffering the consequences of the worst economic recession in post-war his- tory. At the time of writing, world GDP was 12 expected to fall by approximately 1 percent in 2009, marking the first contraction of global Germany France 10 output in 60 years.97 This year’s contraction in United States Canada developed countries is much larger, approach- Italy ing 4 percent. However, initial optimism that 8 Australia emerging economies might be able to ‘decouple’ United Kingdom from the financial crisis has been dampened by 6 Hong Kong (China) mounting evidence that they too are, or will be, Unemployment rate (%) hard hit. Asian countries have suffered from col- 4 lapsing export demand, while increases in the cost of external credit have adversely affected Central and Eastern Europe. African countries 2 are battling with collapsing commodity prices, the drying up of capital liquidity, a sharp de- 0 | | | | cline in remittances and uncertainty concern- 2007 2008 2009* 2010* ing future flows of development aid. Some of the largest emerging economies, such as Brazil and the Russian Federation, will dip into negative * Forecasts Source: Consensus Economics (2009a,b). growth, while others, notably China and India, will see severe slowdowns.98 Typical recessions do not have a large impact topped 28 percent among migrants.103 The places on long-run economic trends.99 However, it is now hit hardest by the crisis thus far are those where clear that this is anything but a typical recession. most migrants live—the more developed econo- As such it is likely to have long-lasting and maybe mies. The negative correlation between numbers even permanent effects on incomes and employ- of immigrants and economic growth suggests ment opportunities, which are likely to be expe- that migrants are likely to be badly affected not rienced unequally by developing and developed only in OECD countries but also in the Gulf, countries.100 For example, the recession set off by East Asia and South Africa (figure 2.12).104 the Federal Reserve’s increase of interest rates in A jobs crisis is generally bad news for mi- 1980 lasted just 3 years in the United States, but grants. Just as economies tend to call on people the ensuing debt crisis led to a period of stagna- from abroad when they face labour shortages, so tion that became known as the ‘lost decade’ in they tend to lay off migrants first during times Africa and Latin America, as the terms of trade of of recession. This is partly because, on average, countries in these regions deteriorated by 25 and migrants have a profile typical of workers who 37 percent respectively. As commodity prices have are most vulnerable to recessions—that is, they fallen significantly from the peak levels of 2008, a are younger, have less formal education and less similar scenario is probable this time round. work experience, tend to work as temporary The financial crisis has quickly turned into labourers and are concentrated in cyclical sec- a jobs crisis (figure 2.11). The OECD unem- tors.105 Even controlling for education and gen- ployment rate is expected to hit 8.4 percent in der, labour force analysis in Germany and the 2009.101 That rate has already been exceeded in United Kingdom found that migrants are much the United States, which by May 2009 had lost more likely to lose their job during a downturn nearly six million jobs since December 2007, than non-migrants.106 Using quarterly GDP and with the total number of jobless rising to 14.5 unemployment data from 14 European coun- million.102 In Spain, the unemployment rate tries between 1998 and 2008, we also found climbed as high as 15 percent by April 2009 and that, in countries that experienced recessions, 41
  • 52. 2 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Figure 2.12 Migrants are in places hardest hit by the recession Immigrants’ location and projected GDP growth rates, 2009 8 China India 4 Peru Indonesia Uruguay Bolivia Honduras Dominican Guatemala Colombia Republic Poland Projected per capita GDP growth rate, 2009 Venezuela 0 EI Salvador Chile Malaysia Bulgaria Australia Paraguay Brazil Argentina Norway Canada Romania Thailand Switzerland Russian Federation Czech Spain France Ecuador Belgium Netherlands Republic Mexico Italy United United Germany States Republic of Korea –4 Turkey Hong Kong (China) Kingdom Hungary Singapore Japan Lithuania –8 Estonia Ukraine Latvia –12 | | | | 10 12 14 16 Total number of immigrants (log scale) Source: HDR team estimates based on Consensus Economics (2009a,b,c,d) and UN (2009d). the unemployment rate of migrants tended reducing the number of visa slots, especially for to increase faster than that of other groups. low-skilled workers but also for skilled work- Within the OECD, migrants were concentrated ers. In some cases these measures are seen as a in highly cyclical sectors that have suffered the short-term response to circumstances and have largest job losses—including manufacturing, involved marginal adjustments rather than out- construction, finance, real estate, hotels and res- right bans (e.g. Australia plans to reduce its an- taurants—sectors that employ more than 40 per- nual intake of skilled migrants by 14 percent).108 cent of immigrants in almost every high-income But there is also a populist tone to many of the OECD country.107 The decline in remittances announcements and provisions. For example, from migrants is likely to have adverse effects the United States economic stimulus pack- on family members in countries of origin, as we age restricts H1B hires among companies that discuss in greater detail in chapter 4. receive funds from the Troubled Asset Relief Several factors come into play in determin- Program;109 the Republic of Korea has stopped ing how the crisis affects—and will affect—the issuing new visas through its Employment movement of people. They include immediate Permit System; and Malaysia has revoked more prospects at home and abroad, the perceived than 55,000 visas for Bangladeshis in order to risks of migrating, staying or returning, and the boost job prospects for locals.110 increased barriers that are likely to come into There is some evidence of a decline of flows place. Several major destination countries have into developed countries during 2008, as the introduced incentives to return (bonuses, tickets, crisis was building. In the United Kingdom, lump sum social security benefits) and increased applications for National Insurance cards from restrictions on entry and stay. Some govern- foreign-born people fell by 25 percent.111 Data ments are discouraging foreign recruitment and from surveys carried out by the US Census 42
  • 53. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 2 Bureau show a 25 percent decline in the flow of as the demographic trends, will persist regardless Current forecasts Mexican migrants to the United States in the of the direction taken by the recession. are that the world’s year ending in August 2008.112 These trends can population will grow by be expected to continue in 2009 and 2010, as the 2.4.2 Demographic trends a third over the next full effect of the crisis plays out in rising domestic Current forecasts are that the world’s population four decades unemployment. There are reasons to be sceptical, will grow by a third over the next four decades. however, that major return flows will emerge. Virtually all of this growth will be in developing As the experience of European guest-worker countries. In one in five countries—including programmes in the 1970s demonstrates, the size Germany, Japan, the Republic of Korea and the of return flows are affected by the prospects of Russian Federation—populations are expected re-entry to the host country, the generosity of to shrink; whereas one in six countries—all of the host country’s welfare system, the needs of them developing and all but three of them in family members and conditions back home—all Africa—will more than double their popula- of which tend to encourage migrants to stay put tions within the next 40 years. Were it not for and ride out the recession. migration, the population of developed coun- Whether this crisis will have major structural tries would peak in 2020 and fall by 7 percent in effects on migration patterns is not yet clear. the following three decades. The trend evident Evidence from previous recessions shows that over the past half century—the fall in the share the outcomes have varied. A historical review of of people living in Europe and the increase in several countries—Argentina, Australia, Brazil, Africa—is likely to continue.118 Canada, the United States and the United Aging of populations is a widespread phe- Kingdom—showed that, between 1850 and nomenon. By 2050, the world as a whole and 1920, declines in domestic wages led to tighter every continent except Africa are projected to restrictions on immigration.113 Several schol- have more elderly people (at least 60 years of ars have argued that the 1973 oil crisis, which age) than children (below 15). This is a natural heralded a prolonged period of economic stag- consequence of the decline in death rates and nation, structural unemployment and lower de- the somewhat slower decline in birth rates that mand for unskilled workers in Europe, affected has occurred in most developing countries, a migration patterns as a wealthier Middle East well-known phenomenon known as the ‘demo- emerged as the new destination hub.114 During graphic transition’. By 2050, the average age in the 1980s, the collapse of Mexican import sub- developing countries will be 38 years, compared stitution set in motion an era of mass migration to 45 years in developed countries. Even this to the United States that was unintentionally ac- seven-year difference will have marked effects. celerated by the 1986 United States immigration The global working-age population is expected reform.115 In contrast, there is little evidence that to increase by 1.1 billion by 2050, whereas the the East Asian financial crisis of the late 1990s working-age population in developed countries, had a lasting impact on international migration even assuming a continuation of current migra- flows.116 tion flows, will decline slightly. Over the next At this stage it is impossible to predict the 15 years, new entrants to the labour force in de- type and magnitude of the structural changes veloping countries will exceed the total number that will emerge from the current crisis with any of working-age people currently living in devel- confidence. Some commentators have argued oped countries (figure 2.13). As in the past, these that the origin of the crisis and its fierce con- trends will put pressure on wages and increase the centration in certain sectors in developed coun- incentives for moving among potential employees tries may strengthen the position of developing in poor countries—and for seeking out workers countries, particularly in Asia, even leading to from abroad among employers in rich countries. a radically different configuration of the global This process affects the dependency ratio— economy.117 However, there are also reasons for that is, the ratio of elderly and young to the expecting a revival of pre-crisis economic and working-age population (table 2.4). For every structural trends once growth resumes. It is cer- 100 working-age people in developed countries, tainly true that deeper long-term processes, such there are currently 49 who are not of working 43
  • 54. 2 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development age, roughly half of whom are children or elderly. shrinks it becomes more difficult to maintain In contrast, in developing countries, the ratio is expenditure levels. higher, at 53, but three quarters of the depen- These demographic trends argue in favour dents are children. Over the next 40 years, as of relaxing the barriers to the entry of migrants. the effect of lower birth rates is felt and the pro- However, we do not suggest that migration is portion of children falls as they reach working the only possible solution to these challenges. age, the dependency ratio will remain roughly Greater labour scarcity can lead to a shift in stable in developing countries, reaching just 55 specialization towards high-technology and by 2050. However, the proportion of elderly will capital-intensive industries, and technological rise markedly in developed countries, so that innovations are possible for services that were there will then be 71 non-working-age people for traditionally labour-intensive, such as care of every 100 of working age, a significantly higher the old. The sustainability of pensions and health fraction than today. These dependency ratios care systems can also be addressed, at least in would increase even more rapidly without the part, by increases in the retirement age and in moderate levels of immigration included in these social security contributions.119 Growing depen- scenarios: if developed countries were to become dency ratios will occur sooner or later in all coun- completely closed to new immigration, the ratio tries undergoing demographic transitions—and would rise to 78 by 2050. migrants themselves grow old. Nevertheless, the As is well known, this scenario makes it growing labour abundance of developing coun- much more difficult for developed countries to tries suggests that we are entering a period when pay for the care of their children and old people. increased migration to developed countries will Publicly funded education and health systems benefit not only migrants and their families but are paid with taxes levied on the working popu- will also be increasingly advantageous for the lation, so that as the share of potential taxpayers populations of destination countries. Figure 2.13 Working-age population will increase in developing regions Projections of working-age population by region, 2010–2050 North America Europe Asia 0.23 0.27 billion 0.50 0.38 billion 2.80 3.40 billion +16% –23% +22% Latin America and the Carribean Africa Oceania 0.39 0.49 billion 0.58 1.3 billion 0.02 0.03 billion +26% +125% +31% 2010 2050 Source: HDR team calculations based on UN(2009e). 44
  • 55. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 2 2.4.3 Environmental factors Table 2.4 Dependency ratios to rise in developed countries and The environment can be a key driver of human remain steady in developing countries movement. From nomadic pastoralists, who fol- Dependency ratio forecasts of developed versus developing low the favourable grazing conditions that arise countries, 2010–2050 after rain, to the people displaced by natural di- Developed countries Developing countries sasters such as the Indian Ocean tsunami and Hurricane Katrina, environmental conditions Year Baseline scenario No Migration scenario Baseline scenario No Migration scenario have been intimately linked to movements of people and communities throughout human his- 2010 49 50 53 53 2020 55 56 52 52 tory. Some are now expecting that the continu- 2030 62 65 52 52 ing warming of the earth will generate massive 2040 68 74 53 53 population shifts. 2050 71 78 55 54 Climate change is projected to increase en- vironmental stress in already marginal lands Source: UN (2009e). and to raise the frequency of natural hazards. Continued greenhouse gas emissions are likely to be associated with changes in rainfall pat- exposed to the risk of major climatic events and terns, desertification, more frequent storms and do not take into account the adaptation mea- rises in sea level, all of which have implications sures that individuals, communities and gov- for human movement.120 Changing rainfall pat- ernments may undertake.125 It is thus difficult to terns, for example, will affect the availability of know whether such inevitably crude estimates water and hence the production of food, possibly facilitate or obstruct reasoned public debate. increasing food prices and the risk of famine. The effect of climate change on human set- Existing estimates indicate that several de- tlement depends partly on how change comes veloping areas will be strongly affected by cli- about—as discrete events or a continuous pro- mate change, although the range of estimates cess. Discrete events often come suddenly and is still very wide and predictions are subject to dramatically, forcing people to move quickly to considerable uncertainty. At one extreme, by more secure places. Continuous processes, on 2020 it is expected that the yields from rainfed the other hand, are associated with slow-onset agriculture in Southern Africa could be halved changes like sea level rise, salinization or erosion by drought.121 Over the medium term, as glacial of agricultural lands and growing water scarcity. water banks run down, river flows are expected In many cases, continuous change leads commu- to diminish, severely affecting irrigated agricul- nities to develop their own adaptation strategies, ture, especially around major massifs such as the of which migration—whether seasonal or per- Himalayas. manent—may be only one component. Under Rises in sea level most directly affect people these conditions movement typically takes the in coastal areas. One scenario suggests that 145 form of income diversification by the house- million people are presently at risk from a rise hold, with some household members leaving and of one meter, three quarters of whom live in others staying behind.126 This pattern has been East and South Asia.122 In some cases, rises will observed, for example, among Ethiopian house- imply the relocation of entire communities. The holds hit by severe and recurrent droughts.127 government of the Maldives, for example, is con- Given the uncertainty as to whether climate sidering buying land in other countries as a safe change will occur through a continuous process haven, given the probability that their archipel- or discrete events, the extent and type of result- ago will become submerged.123 ing adaptation and movement are difficult to Some estimates of the numbers of people predict. Moreover, environmental factors are not who will be forced to move as a result of climate the sole determinants of movement but interact change have been presented, ranging from 200 with livelihood opportunities and public policy million to 1 billion.124 Regrettably, there is little responses. It is often the case that natural disas- hard science backing these numbers. For the ters do not lead to out-migration of the most vul- most part, they represent the number of people nerable groups, because the poorest usually do 45
  • 56. 2 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Movement largely not have the means to move and natural disasters especially those in host countries. At present, reflects people’s further impair their ability to do so. Empirical policy makers in countries with large migrant need to improve studies in Mexico have found that the effects populations face conflicting pressures: signifi- their livelihoods… of changes in rainfall on migration patterns cant levels of resistance to increased immigration this movement is are determined by socio-economic conditions in public opinion on the one hand, and sound constrained by policy and the ability to finance the cost of moving.128 economic and social rationales for the relaxation and economic barriers Background research on migration patterns in of entry barriers on the other. Nicaragua during Hurricane Mitch, carried out How can we expect policies to evolve in the for this report, found that rural families in the next few decades? Will they evolve in ways that bottom two wealth quintiles were less likely to enable us to realize the potential gains from mo- migrate than other families in the aftermath of bility, or will popular pressures gain the upper Hurricane Mitch.129 hand? Will the economic crisis raise protection- More fundamentally, what happens in the ist barriers against immigration, or will it serve future is affected by the way we consume and as an opportunity to rethink the role of move- use our natural resources today. This was the ment in fostering social and economic progress? key message of the 2007/2008 HDR (Fighting History and contemporary experience provide climate change: Human solidarity in a divided contrasting examples. Acute labour scarcity world): catastrophic risks for future generations made the Americas very open to migration dur- can be avoided only if the international commu- ing the 19th century and allowed rapid rates of nity acts now. The demand for increased energy economic development despite widespread in- in developing countries, where many people still tolerance and xenophobia. This is analogous in lack access to electricity, can be met while re- some ways to the situation in the GCC states ducing total carbon emissions. The use of more today. However, the tendency to blame outsiders energy-efficient technologies that already exist for society’s ills is accentuated during economic in developed countries needs to be expanded downturns. Recent incidents across a range in developing countries, while creating the next of countries—from the Russian Federation to generation of still more efficient technologies South Africa to the United Kingdom—could and enabling developing countries to leapfrog presage a growing radicalization and closing off through to these better solutions. At the same to people from abroad.131 time, energy consumption in developed coun- Yet none of these outcomes is predetermined. tries needs to be rationalized. The policy options Leadership and action to change the nature of for encouraging a transition to a low-carbon en- public debate can make a crucial difference. ergy mix include market-based incentives, new Shifting attitudes towards internal migrants in standards for emissions, research to develop the United States during the Great Depression new technologies and improved international provide us with a compelling example. As a cooperation.130 result of severe drought in the nation’s south- ern Midwest region, an estimated 2.5 million 2.5 Conclusions people migrated to new agricultural areas dur- Three key findings have emerged from this chap- ing the 1930s. There they met fierce resistance ter’s analysis of global trends in human move- from some residents, who saw these migrants as ment. First, movement largely reflects people’s threats to their jobs and livelihoods. It was in this need to improve their livelihoods. Second, this context that John Steinbeck wrote The Grapes of movement is constrained by policy and eco- Wrath, one of the most powerful indictments nomic barriers, which are much more difficult of the mistreatment and intolerance of internal for poor people to surmount than for the rela- migrants ever written. Steinbeck’s novel sparked tively wealthy. Third, the pressure for increased a national debate, leading to a congressional in- flows will grow in the coming decades in the face vestigation into the plight of migrant workers of divergent economic and demographic trends. and ultimately to a landmark 1941 decision by Ultimately, how these structural factors will the Supreme Court establishing that states had affect the flow of people in the future depends no right to interfere with the free movement of critically on the stance taken by policy makers, people within the United States.132 46
  • 57. How movers fare 3
  • 58. Movers can reap large gains from the opportunities available in better-off places. These opportunities are shaped by their underlying resources—skills, money and networks—and are constrained by barriers. The policies and laws that affect decisions to move also affect the process of moving and the outcomes. In general, and especially for low-skilled people, the barriers restrict people’s choices and reduce the gains from moving.
  • 59. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 3 How movers fare People are motivated to move by the prospects of improved access to work, education, civil and political rights, security and health care. The majority of movers end up better off—sometimes much better off—than before they moved. The gains are potentially high- est for people who move from poor to the wealthiest countries, but this type of movement is only a small share of total flows. Available evidence suggests that people who move to emerging and develop- ing countries, as well as within countries, also tend to gain. However, movement does not necessarily yield This chapter examines how movement affects a direct positive impact on the well-being of those who move, why gains are unevenly distrib- everyone. Moving is risky, with uncertain out- uted and why some people win while others lose comes and with the specific impacts determined out. There may well be trade-offs, such as loss of by a host of contextual factors. For both inter- civic rights, even where earnings are higher. The nal and international mobility, different as- costs of moving also need to be taken into ac- pects of the process—including the proximate count. We review evidence about these impacts causes of moving and the resources and capa- in turn, to highlight the main findings from a bilities that people start out with—profoundly vast literature and experience. affect outcomes. Those who are forced to flee The key related question of how moving af- and leave behind their homes and belongings fects those who don’t move, in source and des- often go into the process with limited freedom tination places, is addressed in chapter 4. These and very few resources. Likewise, those who distinct areas of focus are of course inextricably are moving in the face of local economic crisis, linked—successful migrants tend to share their drought or other causes of desperate poverty, success with those who stay at home, while the may not know what capabilities they will have; policy responses of destination places affect how they only know that they cannot remain. Even non-movers, as well as movers, fare. Home and migrants who end up well off after a move often host-country impacts are interconnected. Socio- start out with very restricted capabilities and economic mobility in a host country and the high uncertainty. ability to move up the ladder in the homeland The human development outcomes of are often two sides of the same coin. moving are thus profoundly affected by the conditions under which people move. These 3.1 Incomes and livelihoods conditions determine what resources and ca- It is important to recall at the outset that esti- pabilities survive the move. Those who go to mating the impacts of migration is fraught with an embassy to collect a visa, buy a plane ticket difficulties, as we saw in box 1.1. The main prob- and take up a position as a student in, say, the lem is that movers may differ from non-movers United Kingdom, arrive at their destination in in their basic characteristics, so straight compari- much better shape than someone who is traf- sons can be misleading and the identification of ficked—arriving with no papers, no money causal relationships is problematic. and in bondage. The distance travelled (geo- That said, the most easily quantifiable im- graphical, cultural and social) is also impor- pacts of moving can be seen in incomes and tant. Travelling to a country where one does consumption. We begin with these, then turn not speak the language immediately devalues to review the costs of moving, which must be one’s knowledge and skills. subtracted from the gross benefits. 49
  • 60. 3 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Figure 3.1 Movers have much higher incomes than stayers 3.1.1 Impacts on gross income Annual income of migrants in OECD destination countries The evidence consistently reflects very large av- and GDP per capita in origin countries, by origin country HDI category erage income gains for movers. Commissioned research found large differences in income be- tween stayers and movers to OECD countries, Low HDI with the biggest differences for those moving Difference: US$13,736 from low-HDI countries (figure 3.1). Migrant Medium HDI workers in the United States earn about four Difference: US$12,789 times as much as they would in their develop- ing countries of origin,1 while Pacific Islanders High HDI in New Zealand increased their net real wages Difference: US$9,431 by a factor of three. 2 Evidence from a range of Very high HDI countries suggests that income gains increase Difference: US$2,480 over time, as the acquisition of language skills | | | | | | | leads to better integration in the labour market.3 0 5 10 15 20 25 30 Gains arise not only when people move to (US$ thousands) OECD countries. Thai migrants in Hong Kong GDP per capita at origin Migrants’ income in OECD destination countries (China) and Taiwan (Province of China), for ex- ample, are paid at least four times as much as they Source: Ortega (2009). would earn as low-skilled workers at home.4 In Tajikistan, when the average monthly wage was only US$9, seasonal earnings of US$500–700 Figure 3.2 Huge salary gains for high-skilled movers in the Russian Federation could cover a family’s Gaps in average professional salaries for selected country annual household expenses in the capital city, pairs, 2002–2006 Dushanbe.5 However, these average gains are unevenly distributed, and the costs of moving Physician Côte d’Ivoire also detract from the gross gains. France Gains can be large for the high-skilled as Zambia well as the low-skilled. The wages of Indian soft- Canada ware engineers in the late 1990s, for example, Nurse were less than 30 percent of their United States Malawi counterparts, so those who were able to relocate South Africa to this country reaped large gains. 6 Figure 3.2 Ghana United Kingdom illustrates the wage gaps, adjusted for purchas- ing power parity, between high-skilled profes- Software engineer – manager India sionals in selected pairs of countries. A doctor United States from Côte d’Ivoire can raise her real earnings Software engineer – developer by a factor of six by working in France. Beyond India salaries, many are also often motivated by fac- United States tors such as better prospects for their children, Professor – entry level improved security and a more pleasant working China environment.7 Australia Internal migrants also tend to access bet- Professor – top level India ter income-earning opportunities and are United Kingdom able to diversify their sources of livelihood. | | | | | | | Commissioned research found that internal 0 20 40 65 80 100 120 Annual salary (US$ thousands) migrants in Bolivia experienced significant real income gains, with more than fourfold in- Origin country Destination country creases accruing to workers with low education Source: Source: Clemens (2009b). levels moving from the countryside to the cities (figure 3.3). We also found that in 13 out of 16 50
  • 61. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 3 countries internal migrants had higher incomes Figure 3.3 Significant wage gains to internal movers in Bolivia, than non-migrants.8 In Brazil and Panama, a se- especially the less well educated ries of studies controlling for education found Ratio of destination to origin wages for internal migrants income gains for indigenous groups who move.9 in Bolivia, 2000 Studies across a range of countries suggest that 5 years of Rural to urban internal migration has enabled many house- schooling migrants holds to lift themselves out of poverty, as dis- Urban to urban 11 years of cussed further in the next chapter. schooling migrants The segmentation of labour markets in de- 16 years of veloping countries affects how movers fare. schooling Sometimes this can be traced to administrative | | | | | | 0 1 2 3 4 5 restrictions, as in the hukou system in China Ratio of destination to origin wages (box 3.1) and the ho khau system in Viet Nam. However, segmentation is also widespread in Source: Molina and Yañez (2009). other regions, including South Asia, Africa and Latin America, through barriers that, while not imposed by law, are nonetheless deeply en- in the personal service sector.15 Difficulties trenched through social and cultural norms.10 are compounded where migrant women are For example, rural–urban migrants in India are excluded from normal worker protections, as predominantly employed in industries such as is the case for domestic workers in the GCC construction, brick kilns, textiles and mining, states.16 Although practices are changing in which entail hard physical labour and harsh some countries (e.g. Saudi Arabia and the working and living environments; in Mongolia, United Arab Emirates), migrants are legally rural–urban migrants typically work in infor- prohibited from joining local unions, and even mal activities which are temporary, strenuous when this is allowed, they may face resistance and without legal protection.11 In Asia, recent and hostility from other workers.17 NGOs may low-skilled migrants from rural areas tend to oc- provide services and protection to migrants, cupy the lowest social and occupational rungs of but their coverage tends to be limited. urban society and are treated as outsiders. Labour market discrimination can be a As we saw in chapter 2, most movers from major obstacle to migrants. This is reflected in low-HDI countries are living and working in low call-back rates to job applications where the other low- or medium-HDI countries, in part applicant has a foreign-sounding surname.18 Yet because barriers to admission are often lower the picture is often complex, and ethnicity, gen- and the costs of moving are less. At the same der and legal status may all come into play. In the time, the conditions may well be more difficult United Kingdom, some studies have found dis- than in rich countries and there are risks of both crimination in hiring migrants in terms of lower exploitation and expulsion. employment rates and payments, whereas other Labour market opportunities for migrant studies found that people with Chinese, Indian women from developing countries tend to be and Irish backgrounds tended to have employ- highly concentrated in care activities, paid do- ment situations at least as good as white British mestic work and the informal sector.12 Such workers.19 Our analysis of the 2006 European women may become trapped in enclaves. For Social Survey reveals that the vast majority of example, in New York City, Hispanic-owned migrants (more than 75 percent) in this region firms were found to provide low wages, few did not report feeling discriminated against. benefits and limited career opportunities to However, in the much larger country sample Dominican and Colombian women, reinforc- provided by the World Values Survey, there was ing their social disadvantages.13 Similar results widespread support among locally born people were found among Chinese migrant women for the proposition, “Employers should give workers.14 Most Peruvian and Paraguayan priority to natives when jobs are scarce”, albeit women in Argentina (69 and 58 percent respec- with considerable differences across countries tively) work for low pay on an informal basis (see section 4.2.5). 51
  • 62. 3 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Box 3.1 China: Policies and outcomes associated with internal migration Modelled after the Soviet propiska system, albeit with roots dating Occupational hazards are high—migrants accounted for about 75 back to ancient times, China’s Residence Registration System oper- percent of the 11,000 fatalities in 2005 in the notoriously dangerous ates through a permit (hukou), needed to gain access to farmland in mining and construction industries. agricultural areas and to social benefits and public services in urban Access to services. Children who move with temporary sta- areas. Until the mid-1980s, the system was administered strictly and tus pay additional fees and are denied access to elite schools. An movement without a hukou was forbidden. Since then, China has lib- estimated 14–20 million migrant children lack access to schooling eralized movement but formally maintained the hukou system. altogether. Their drop-out rates at primary and secondary schools As in other areas of reform, China chose a gradual and partial exceed 9 percent, compared to close to zero for locals. Access to approach. Beginning in the mid-1880s, it allowed people to work out- basic health services is limited. Even in Shanghai, one of the better side their place of residence without a hukou, but did not allow them cities in terms of providing social services to migrants, only two thirds access to social benefits, public services or formal-sector jobs. A of migrant children were vaccinated in 2004, compared to universal two-tier migration system analogous to the points system in some rates for local children. When migrants fall ill, they often move back developed countries was designed: changes in permanent residency to rural areas for treatment, due to the costs of urban health care. are permitted for the well educated, but only temporary residence Participation. Many migrants remain marginalized in destination is granted for less-educated rural migrants. Many city governments places due to institutional barriers. They have few channels for ex- have offered ‘blue-stamp’ hukou to well-off migrants who were able pressing their interests and protecting their rights in the work place. to make sizeable investments. Almost 8 out of 10 have no trade union, workers’ representative con- The evidence suggests that the human development gains for ference, labour supervisory committees or other labour organization, internal migrants and their families have been limited by the persis- compared to one fifth of locally born people. Long distances also tence of the hukou system, along the dimensions illustrated below: hinder participation: in a survey of migrants in Wuhan City, only 20 Income gains. In 2004, on average, rural–urban migrants earned percent had voted in the last village election, mainly because they RMB780 (US$94) per month, triple the average rural farm income. lived too far away from polling stations. However, due to the segmentation created by the hukou system, tem- Discussions about hukou reform are reportedly ongoing, while some porary migrants typically move to relatively low-paid jobs, and their regional governments have further liberalized their systems. Legislative poverty incidence is double that of urban residents with hukou. reforms in 1997 significantly improved the rights of all workers—includ- Working conditions. Low-skilled migrants tend to work in informal ing migrants, and measures to provide portable pensions for migrant jobs that have inadequate protection and benefits. According to one workers were announced in 2008. Other signs of change come from survey in three provinces, migrants’ work hours are 50 percent longer Dongguan, Guangdong, for example, where migrants are now referred than locals, they are often hired without a written contract and fewer to as ‘new residents’ and the Migrants and Rental Accommodation than 1 in 10 have old-age social security and health insurance, com- Administration Office was relabelled the ‘Residents Service Bureau’. pared to average coverage of over 70 percent in China as a whole. Source: Avenarius (2007), Gaige (2006), Chan, Liu, and Yang (1999), Fan (2002), Meng and Zhang (2001), Cai, Du, and Wang (2009), Huang (2006), Ha, Yi, and Zhang (2009b), Fang and Wang (2008), and Mitchell (2009). One problem facing many migrants on ar- in Canada, despite the points system, this prob- rival is that their skills and credentials go un- lem is estimated to drain US$1.7 billion a year recognized. 20 Coupled with language and other from the economy. 22 In response, the Canadian social barriers, this means that they tend to government has launched programmes to speed earn much less than similarly qualified local up the recognition of credentials earned abroad. residents. 21 The extent of this problem seems Incomes do not depend solely on labour to vary across sectors. Information technology market earnings. In countries with established firms tend to be more flexible on credentials, for welfare systems, social transfers reduce poverty example, whereas public-sector organizations rates among disadvantaged groups through un- are often more closed. The failure to fully deploy employment benefits, social assistance and pen- their skills can cause new immigrants to incur sions. Whether or not a programme benefits significant costs. The Migration Policy Institute migrant families depends on the design and recently estimated that up to 20 percent of col- rules of the system. There are obvious differ- lege-educated migrants in the United States were ences across countries in the generosity of these unemployed or working in low-skilled jobs, and programmes, as their scale tends to be more 52
  • 63. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 3 limited in developing countries due to budget- in France and the United Kingdom respectively. ary constraints. Since most developing coun- The redistributive effect of social welfare in these tries do not have extensive systems in place, the countries is significant, since transfers more than question of equality of access does not arise. The halve these rates for both migrant and locally focus here is therefore on developed countries. born children (figure 3.4).24 In contrast, in the Our policy assessment found that nearly all United States the poverty-reducing effect of so- developed countries in the sample granted per- cial transfers for both local and migrant families manent migrants access to unemployment ben- is negligible, because transfers overall are rela- efits and family allowances. However, people tively small. At the same time it is notable that in with temporary status are less likely to be able Australia, Germany and the United States rates of to access assistance. Some countries, includ- market–income poverty are much lower than in ing Australia and New Zealand, have imposed France and the United Kingdom, suggesting that waiting periods before various benefits can be migrant families are doing better in the labour accessed. And in efforts to avoid welfare depen- market in those countries. dency, countries such as France and Germany require that applications for family reunifica- 3.1.2 Financial costs of moving tion demonstrate that the applicant has stable The gross income gains reported in the litera- and sufficient income to support all family ture typically do not account for the monetary members without relying on state benefits. costs of moving. These costs arise from various The Luxembourg Income Study and sources, including official fees for documents the European Survey of Income and Living and clearances, payments to intermediaries, Conditions allow estimates of the effects of social travel expenses and, in some cases, payments of transfers on poverty among families with chil- bribes. The costs appear regressive, in that fees dren.23 For all 18 countries in the sample, migrant for unskilled workers are often high relative to families are more likely to be poor than locally expected wages abroad, especially for those on born families. Based on market incomes before temporary contracts. 25 social transfers, poverty rates among children ex- Substantial costs may arise for those with- ceed 50 and 40 percent among migrant families out basic documents. Around the world, an Figure 3.4 Poverty is higher among migrant children, but social transfers can help Effects of transfers on child poverty in selected countries, 1999–2001 Australia 32 20 56 France 13 United States 33 20 19 6 Share of migrant children in poverty before social transfers (%) 8 Share of migrant children in poverty 16 15 after social transfers (%) 23 22 Germany Share of non-migrant children in poverty after social transfers (%) 43 United Kingdom Source: Smeeding, Wing, and Robson (2008). 53
  • 64. 3 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Figure 3.5 Costs of moving are often high estimated 48 million children, often from very Costs of intermediaries in selected corridors against income poor families, lack a birth certificate. The main per capita, 2006–2008 reason is the fee for obtaining such documents and related factors such as distance to the regis- Viet Nam to Japan (6 years, 5 months and 4 days) tration centre. 26 Lengthy application processes and, in some countries, payments of bribes for routine ser- Bangladesh to Saudi Arabia (5 years, 2 months and 3 days) vices can make applying for vital records and basic travel documents very expensive. 27 In the Democratic Republic of the Congo passport China to Australia (3 years, 10 months and 16 days) applicants can expect to pay up to US$500 (70 percent of average annual income) in bribes. 28 Other countries with limited bureaucratic ca- Colombia to Spain (1 year, 8 months and 3 days) pacity and corruption in the issuance of travel documents reportedly include Azerbaijan, India and Uzbekistan. 29 India to United Kingdom (1 year, 3 months) Intermediaries, also known as ‘middlemen’, perform a specific function in the global labour market. They help to overcome information gaps Philippines to Singapore (8 months and 26 days) and meet administrative requirements (such as having a job offer prior to visa application) and sometimes lend money to cover the upfront costs = Origin country annual GNI per capita of the move. There are a large number of agen- cies: in the Philippines alone there are nearly 1,500 licensed recruitment agencies, while India Source: Bangladesh to Saudi Arabia: Malek (2008); China to Australia: Zhiwu (2009); Colombia to Spain: Grupo de Investigación en Movilidad Humana (2009); Philippines to Singapore: TWC (2006); Viet Nam to Japan: van Thanh (2008). has close to 2,000.30 The cost of intermediary services appears to vary enormously, but often exceeds per capita income at home (figure 3.5). Figure 3.6 Moving costs can be many times expected The example of Indonesia illustrates how monthly earnings the costs can vary by destination, with moves to Costs of movement against expected salary of low-skilled Malaysia and Singapore costing about six months’ Indonesian workers in selected destinations, 2008 expected salary and to Taiwan a full year (figure 3.6). Legal caps on fees charged by recruiters Hong Kong (China) are generally ignored, as migrants routinely pay much more.31 The difference between wages at home and expected wages abroad is perhaps the Taiwan (Province of China) most important determinant of the price of in- termediary services. Where relatively few jobs are available, intermediaries who are in a position to Malaysia allocate these slots are able to charge additional rents. There are cases of abuse and fraud, where prospective movers pay high recruitment fees only Singapore to find later on (at the destination) that the work contract does not exist, there have been unilat- eral changes to the contract, or there are serious violations related to personal safety and working 12 months = Monthly expected salary conditions.32 Some migrants report that employ- ers confiscate their passports, mistreat their em- ployees and deny access to medical care.33 Source: The Institute for ECOSOC Rights (2008). Extensive regulations and official fees can encourage irregularity. For Russian employers, 54
  • 65. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 3 Figure 3.7 The children of movers have a much greater chance of surviving Child mortality at origin versus destination by origin country HDI category, 2000 census or latest round Low HDI Medium HDI High HDI Very high HDI (100 versus 7 per thousand) (50 versus 7 per thousand) (16 versus 7 per thousand) (5 versus 7 per thousand) Child mortality at origin Child mortality at destination Source: Ortega (2009). the administrative procedure to apply for a li- as frequent.37 However, the links between cense to hire a foreign worker is reportedly so migration and health are complex. Migrants’ time-consuming and corrupt that it frequently health depends on their personal history be- leads to evasion and perpetuates irregular em- fore moving, the process of moving itself, and ployment practices.34 In Singapore, employers of the circumstances of resettlement. Destination low-skilled migrants must pay a levy, which they governments often rigorously screen applicants in turn deduct from workers’ wages.35 Under for work visas, so successful applicants tend to agreements between Thailand, Cambodia and be healthy.38 Nevertheless, irregular migrants the Lao People’s Democratic Republic, recruit- may have specific health needs that remain ment fees are equivalent to 4–5 months’ salary, unaddressed. processing time averages about four months and Moving to more developed countries can 15 percent of wages are withheld pending the improve access to health facilities and profes- migrant’s return home. In contrast, smugglers in sionals as well as to health-enhancing factors these corridors reportedly charge the equivalent such as potable water, sanitation, refrigeration, of one month’s salary. Given these cost differ- better health information and, last but not least, ences, it is not surprising that only 26 percent higher incomes. Evidence suggests that migrant of migrant workers in Thailand were registered families have fewer and healthier children than in 2006.36 they would have had if they had not moved.39 Recent research conducted in the United States 3.2 Health using panel data, which tracks the same indi- This section reviews the impacts of movement viduals over time, found that health outcomes on the health of those who move. Gaining bet- improve markedly during the first year after ter access to services, including health care, immigration.40 may be among the key motivations for moving. Our commissioned study found a 16-fold re- Among top high-school graduates from Tonga duction in child mortality (from 112 to 7 deaths and Papua New Guinea, ‘health care’ and ‘chil- per 1,000 live births) for movers from low-HDI dren’s education’ were mentioned more often countries (figure 3.7). Of course these gains are than ‘salary’ as reasons for migrating, and an- partly explained by self-selection.41 Nonetheless, swers such as ‘safety and security’ were almost the sheer size of these differences suggests that 55
  • 66. 3 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Barriers to health similar outcomes would have been very difficult may be abusive and/or hazardous.49 In India, for services arise due to to realize at home. For comparison, as reported example, many internal migrants work in dan- financial constraints in the 2006 HDR, families in the richest quin- gerous construction jobs, while working condi- as well as status, tile in Burkina Faso had a child mortality rate of tions in the leather industry expose the mainly cultural and language about 150 deaths per 1,000 live births. migrant workers to respiratory problems and differences Not surprisingly, given the poor health ser- skin infections.50 Yet these jobs are well paid vices, water quality and sanitation in rural areas, compared to what was available at home, and studies suggest that migrants to urban centres interviews in rural Bihar indicate that such jobs significantly improve their chances of survival are highly sought after.51 relative to rural residents.42 The size of this effect Not all types of migrants have the same ac- has been correlated with duration of stay, which cess to health care.52 Permanent migrants often was itself associated with higher incomes and have greater access than temporary migrants, improved knowledge and practices. Sometimes and the access of irregular migrants tends to be migrants use health care services more than much more restricted (figure 3.8). Movement urban locals, suggesting that the availability of sometimes deprives internal migrants of ac- these may have motivated their move in the first cess to health services if eligibility is linked to place. However, the health outcomes associated authorized residence, as in China. In contrast, with urbanization are variable: a broader study permanent migrants, especially the high-skilled, found that internal migrants’ outcomes were tend to enjoy relatively good access, while in worse than those of urban natives, due to their some countries health care is open to all mi- socio-economic disadvantage, and our commis- grants, regardless of their legal status, as is the sioned research found that internal migrants case in Portugal and Spain. In the United Arab had higher life expectancy than non-migrants Emirates coverage varies by emirate, but both in only half of the countries studied.43 Abu Dhabi and Dubai have compulsory insur- Detailed studies in a number of OECD ance schemes to which employers must contrib- countries have found that migrants’ initial ute on behalf of their workers. In Canada all health advantage tends to dissipate over time.44 residents are entitled to national health insur- This is believed to reflect the adoption of poorer ance, and the provincial authorities determine health behaviour and lifestyles as well as, for who qualifies as a resident. some, exposure to the adverse working, hous- In practice, barriers to health services arise ing and environmental conditions that often due to financial constraints as well as status, characterize low-income groups in industrial cultural and language differences,53 especially countries. Separation from family and social for irregular migrants. In France, Germany and networks and uncertainty regarding job secu- Sweden there is a ‘responsibility to report’ the rity and living conditions can affect health. In treatment of an irregular migrant, which can several studies, migrants have reported higher lead to a lack of trust between providers and pa- incidence of stress, anxiety and depression than tients and deter migrants from seeking care.54 residents,45 outcomes that were correlated with If single female migrants in the GCC states are worse economic conditions, language barriers, found to be pregnant, they are deported.55 irregular status and recent arrival. Conversely, In less-wealthy destination countries there is other studies have found positive effects of mi- a tension between the ideal of granting health gration on mental health, associated with better care access to irregular migrants and the real- economic opportunities.46 ity of resource constraints. In South Africa Poor housing conditions and risky occupa- many non-nationals report not being able to tions can increase accidents and compromise access antiretroviral drugs against AIDS be- health, which may be worse for irregular mi- cause facilities deny treatment on the basis of grants.47 There are well-documented  inequali- ‘being foreign’ or not having a national identity ties in health care and status between vulnerable booklet.56 Given that South Africa has one of migrant groups and host populations in devel- the highest HIV prevalence rates in the world, oped countries.48 The health of child migrants combined with improved but still limited ac- can also be affected by their type of work, which cess to antiretrovirals, it is not surprising that 56
  • 67. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 3 irregular migrants represent a low priority. Figure 3.8 Temporary and irregular migrants often lack access But more positive examples are found in other to health care services parts of the world. Thailand, for example, pro- Access to health care by migrant status in developed versus vides antiretroviral treatment to migrants from developing countries, 2009 Cambodia and Myanmar, with support from the Global Fund on AIDS, Tuberculosis and Panel A: Preventive care Malaria. Thailand also provides migrants with access to health insurance, and efforts are under Developed countries Permanent way to reach irregular migrants. Temporary 3.3 Education Irregular Education has both intrinsic value and brings Humanitarian instrumental gains in income-earning potential and social participation. It can provide the lan- Developing countries guage, technical and social skills that facilitate Permanent economic and social integration and intergener- Temporary ational income gains. Movement is likely to en- Irregular hance educational attainment, especially among children. Many families move with the specific Humanitarian objective of having their children attend better | | | | | | 0 20 40 60 80 100 and/or more advanced schools. In many rural Share of countries in sample (%) areas in developing countries education is avail- able only at primary level and at a lower quality than in urban areas, providing an additional motive for rural–urban migration.57 Similarly, Panel B: Emergency care international migration for educational pur- poses—school migration—is rising.58 Developed countries In this section we review the evidence con- Permanent cerning school completion levels at places of Temporary origin and at destinations, whether migrant Irregular children can access state schools and how well Humanitarian they perform relative to children born locally. School enrolments can change for a number Developing countries of reasons when a family relocates. Higher in- Permanent comes are part of the story, but other factors, Temporary such as the availability of teachers and schools, Irregular the quality of infrastructure and the cost of transport, may be important as well. A natu- Humanitarian ral starting point when measuring education | | | | | | gains is a comparison of enrolment rates. These 0 20 40 60 80 100 Share of countries in sample (%) present a striking picture of the advantages of moving (figure 3.9), with the crude differences being largest for children from low-HDI coun- Only available for citizens or not available tries. Two familiar notes of caution should be Available for migrants with conditions sounded, however: these results may be over- Immediately available for migrants estimated due to positive selection; and mere enrolment guarantees neither a high-quality education nor a favourable outcome from schooling.59 The importance of early stimulation to the Source: Klugman and Pereira (2009). physical, cognitive and emotional development 57
  • 68. 3 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Figure 3.9 Gains in schooling are greatest for migrants However, due to traditional norms, language from low-HDI countries and cultural barriers and sometimes uncertain Gross total enrolment ratio at origin versus destination by legal status, these children are generally less origin country HDI category, 2000 census or latest round likely to enrol in formal ECD programmes, de- spite the fact that authorities in Europe and the United States often actively reach out to migrant children.62 Thailand is among those developing countries that seek to extend informal ECD to migrants, in border areas in the north. Similar arrangements can be found in some other coun- tries; programmes in the Dominican Republic Low HDI Medium HDI High HDI Very high HDI serve Haitian children, for example. (47% versus 95%) (66% versus 92%) (77% versus 92%) (92% versus 93%) In some countries migrant children may not have access to state schools or their parents may be asked to pay higher fees. Our policy assess- Enrolment ratio at origin Enrolment ratio at destination ment found that developed countries are more likely to allow immediate access to schooling for all types of migrant—permanent, temporary, humanitarian and irregular (figure 3.10). Yet a Source: Ortega (2009). Note: Gross total enrolment includes primary, secondary and tertiary education. third of developed countries in our sample, in- cluding Singapore and Sweden,63 did not allow access to children with irregular status, while Figure 3.10 Migrants have better access to education in the same was true for over half the developing developed countries countries in the sample, including Egypt and Access to public schooling by migrant status in India. Some specific cases: in the United Arab developed versus developing countries, 2009 Emirates children with irregular migrant sta- tus do not have access to education services; in Developed countries Belgium education is free and a right for every Permanent Only available for person, but not compulsory for irregular chil- citizens or not Temporary available dren; in Poland education for children between Irregular 6 and 18 years is a right and is compulsory, Available for migrants with but children with irregular status cannot be Humanitarian conditions counted for funding purposes, which may lead Developing countries Immediately the school to decline to enrol such children.64 available for Permanent migrants Poverty and discrimination (formal and in- Temporary formal) can inhibit access to basic services. Even if children with irregular status have the right Irregular to attend a state school, there may be barriers Humanitarian to their enrolment. In several countries (e.g. | | | | | | France, Italy, the United States), fears that their 0 20 40 60 80 100 irregular situation will be reported have been Share of countries in sample (%) found to deter enrolment. 65 In South Africa Source: Klugman and Pereira (2009). close to a third of school-age non-national children are not enrolled, for a combination of reasons including inability to pay for fees, of children, and the associated importance transport, uniforms and books, and exclusion of early childhood development (ECD) pro- by school administrators, while those in school grammes, is well established. 60 Research from regularly report being subjected to xenophobic Germany indicates that ECD can bring the comments by teachers or other students.66 children of migrants to par with native children The steepest challenges appear to be faced with the same socio-economic background. 61 by two groups: children who migrate alone, 58
  • 69. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 3 who tend to have irregular status (box 3.2), Box 3.2 Independent child migrants and children who migrate within and between developing countries with their parents, on a Trafficking and asylum-seeking are often depicted as accounting for most of the temporary basis. The first group is unlikely to independent movement of children. However, evidence with a long historic record be able to access education at all, due to social confirms that children also move in search of opportunities for work and education. and cultural isolation, strenuous and hazardous The Convention on the Rights of the Child goes some way to recognizing children as work, extreme poverty, poor health conditions agents, decision makers, initiators and social actors in their own right. However, the and language barriers. 67 As regards the second literature and policy responses to children’s mobility have largely focused on welfare group, qualitative studies in Viet Nam and and protection from harm, and tended to neglect policies of inclusion, facilitation and Pakistan have found that seasonal migration non-discrimination. disrupts their education.68 For instance, the Rac As for other types of movement, the effect of independent child migration is Lai minority in Viet Nam migrate with their context-specific. Some studies have found a significant link between non-attendance children to isolated mountainous areas during at school and the propensity to migrate to work among rural children, while others the harvest season and their children do not at- find that migration is positively associated with education. A recent study using cen- tend school during this period.69 sus data in Argentina, Chile and South Africa shows that independent child migrants Even if migrant children gain access to bet- had worse shelter at destination, whereas dependent child migrants were similar to ter schools than would have been available to non-migrants in their type of shelter. Over a fifth of international independent child them at origin, they do not all perform well migrants aged 15–17 years in these countries were employed, compared to fewer than in examinations in comparison with their lo- 4 percent of non-migrant dependent children. Many live with relatives or employers, cally born peers. In the 21 OECD and 12 non- but shelter and security can be important concerns. Children may be less able than OECD countries covered by the Programme adults to change jobs, find it harder to obtain documents even when eligible, may be for International Student Assessment,70 which more likely to suffer employer violence or encounters with the police, and may be more tested performance in science, pupils who were easily cheated by employers and others. migrants tended to perform worse in this sub- Source: Bhabha (2008) and Yaqub (2009). ject than locally born children. However, for- eign-born pupils perform as well as their native peers in Australia, Ireland and New Zealand, children are often disadvantaged by the absence as well as in Israel, Macao (China) the Russian of a parent. Federation and Serbia. Likewise, pupils from In OECD countries migrant pupils generally the same country of origin performed differently attend schools with teachers and educational across even neighbouring countries: for example, resources of similar quality to those attended migrant pupils from Turkey perform better in by locally-born pupils, although there are some mathematics in Switzerland than in Germany.71 exceptions, including Denmark, Greece, the The next generation—children of migrants who Netherlands and Portugal. In some cases, the are born in the destination place—generally do quality of schools that migrant children attend better, but with exceptions, including Denmark, is below national standards, but this is more Germany and the Netherlands. often related to local income levels generally Part of the educational disadvantage of chil- than to migrant status in particular. Studies on dren in migrant families can be traced to low school segregation in the United States suggest parental education and low income. Children that children from migrant families have worse whose parents have less than full second- test scores if they attend minority, inner-city ary completion—which tends to be the case schools.73 Studies from the Netherlands and in migrant households in France, Germany, Sweden find that clustering migrant children Switzerland and the United States—typically and separating them from other children is det- complete fewer years of school. However, while rimental to school performance.74 Even if they many migrant families live away from relatives are not at a disadvantage with regard to instruc- and social networks, a study of migrant children tional materials and equipment, migrant pupils in eight developed countries found that they are may need special services, such as local language generally more likely than local children to grow instruction. up with both parents.72 This counters a belief Our interest in schooling is partly due to sometimes found in the literature that migrant its value in improving the prospects of future 59
  • 70. 3 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Box 3.3 The next generation People who move are often motivated by the prospect of better lives they have higher employment rates compared to migrants in the for their children. And indeed the children of migrants can represent same age group, but they are at a disadvantage compared to those a key population group requiring the attention of policy makers. In without a migrant background. In some European countries youth Brussels, for example, they represent over 40 percent of the school- unemployment rates are worse among the children of migrants. age population, while in New York they are half and in Los Angeles Limited access to informal networks and discrimination (whether County almost two thirds. origin- or class-based) can contribute to these disparities. Obtaining a good education is critical to future prospects. Some children of migrants encounter racism, often linked to Evidence suggests that the children of migrants typically perform limited job opportunities. Studies in the United States, for exam- better than their parents, but do not fully catch up with children with- ple, have suggested that there is a risk of ‘segmented assimilation’, out a migrant background, even after controlling for socio-economic meaning that the contacts, networks and aspirations of children of characteristics. There are exceptions, however, including Australia immigrants are limited to their own ethnic group, but also that this and Canada, where school performance is close to or exceeds that risk varies across groups. Teenage children of Mexican migrants of native peers. Countries with education systems that involve early have been found to be at higher risk of dropping out of school, going streaming, such as Germany and the Netherlands, appear to have to prison or becoming pregnant. The same studies suggest that eco- the biggest gaps in school performance. nomic and social resources at the family and community levels can How the children of migrants fare in the labour market also tends help to overcome these risks and avert the rise of an underclass of to differ across countries and groups. Recent findings suggest that disaffected youth. Source: Crul (2007), OECD (2007), Castles and Miller (1993), and Portes and Zhou (2009). generations. Some evidence on the extent to Movement can allow rural women to gain which this happens is presented in box 3.3. autonomy. Empowerment tends to occur when migration draws women from rural to urban 3.4 Empowerment, civic rights areas, separating them from other family mem- and participation bers and friends and leading them to take paid Moving has the potential to affect not only mate- work outside the home.77 Qualitative studies rial well-being but also such things as bargaining in Ecuador, Mexico and Thailand have dem- power, self-respect and dignity. Empowerment, onstrated such effects. For the women in these defined as the freedom to act in pursuit of per- studies, returning to the old rural way of life was sonal goals and well-being,75 can be enhanced an unthinkable proposition.78 Higher labour through movement. However, the reception in force participation and greater autonomy have the host country obviously matters, especially also been found among Turkish women who when migrants face local hostility, which can emigrated.79 It is not only women who seek to even lead to outbreaks of violence. challenge traditional roles when they move: Human development is concerned with the young migrant men can be similarly empow- full range of capabilities, including social free- ered to challenge patriarchal structures within doms that cannot be exercised without political the family.80 and civic guarantees. These form part of the di- But such positive outcomes are not inevita- mension of freedom that some philosophers have ble. Some migrant communities become caught labelled “the social bases of self-respect”.76 They in a time warp, clinging to the cultural and so- can be just as important as gains in income and cial practices that prevailed in the home country may be associated with these gains, but are often at the time of migration, even if the country has held in check by deep-seated social, class and ra- since moved on. 81 Or the migrant communi- cial barriers. In many countries the attitude to- ties may develop radically conservative ideas wards migration is negative, which can diminish and practices, as a way to isolate them from the migrants’ sense of dignity and self-respect. This host culture. This can lead to alienation and, is not a new phenomenon: in the 19th century, occasionally, to extremism. There is a complex the Irish faced the same prejudices in the United dynamic between cultural and community Kingdom, as did the Chinese in Australia. traditions, socio-economic circumstances and 60
  • 71. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 3 public policies. Recent micro-analysis for 10 for many refugees fleeing from conflict, even if Latin American countries found that internal their legal situation remains tenuous while they migrants of indigenous origin still faced dis- are seeking asylum. Our analysis of determi- crimination in urban areas, even while they nants of flows between pairs of countries shows gained greater access to services than they had that the level of democracy in a country has a in their rural area.82 Another study found that positive, significant effect on migrant inflows. 85 Bolivian women in Argentina were discrimi- Yet even countries with strong legal tradi- nated against, had only limited employment tions are tested when routine police work in- opportunities and continued to occupy subor- volves the enforcement of migration law. As dinate social positions. 83 we saw in chapter 2, countries vary in their Participation and civic engagement are im- enforcement practices. In some countries, ir- portant aspects of empowerment. Our analysis regular migrants may be seen as easy targets by using the World Values Survey suggests that corrupt officials. In South Africa police hoping people with a migrant background are more to extort bribes often destroy or refuse to rec- likely to participate in a range of civic associa- ognize documents in order to justify arrest. 86 tions. Compared to people who do not have Mongolian migrants in the Czech Republic also a migrant parent, they are more likely to be a report paying fines during police raids, regard- member of, and also tend to have more confi- less of whether they are authorized or not. 87 In dence in, a range of organizations, such as sport, Malaysia migrants have sometimes been subject recreational, art and professional organizations. to informal enforcement mechanisms, which Research also suggests that political participa- have led to complaints of abuse (box 3.4). tion increases with the ability to speak the host As we shall see in chapter 4, people in des- country’s language, with duration of stay, educa- tination places often have concerns about the tion in the destination country, connections to economic, security and cultural impacts of im- social networks and labour markets, and when migration. In some cases, xenophobia arises. institutional barriers to registering and voting This appears to be most likely where extrem- are lower.84 ists foment fears and insecurities. Outbreaks Institutional factors matter, especially civic of violence towards migrants can erupt—such and electoral rights. Our policy assessment found that voting in national elections was largely restricted to citizens, although several Figure 3.11 Voting rights are generally reserved for citizens developed countries allow foreigners to vote Voting rights in local elections by migrant status in in local elections (figure 3.11). The Migrant developed versus developing countries, 2009 Integration Policy Index (MIPEX), which as- Developed countries sesses the opportunities for migrants to par- Permanent ticipate in public life in terms of collective Temporary associations, voting in, and standing for local elections and support provided to migrant asso- Developing countries ciations, found policies in Western Europe to be Permanent favourable to participation, but those in Central, Temporary Eastern and South-Eastern Europe were less so. In Sweden any legal resident who has lived in | | | | | | 0 20 40 60 80 100 the country for three years can vote in regional Share of countries in sample (%) and local elections and stand for local elections, while in Spain foreigners can vote in local elec- Only available for citizens or not available tions as long as they are registered as residents with their local authority. Available for migrants with conditions Many people move at least partly to enjoy greater physical and personal security, and to places where the rule of law and government ac- Source: Klugman and Pereira (2009). countability are better. This is obviously the case 61
  • 72. 3 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Box 3.4 Enforcement mechanisms in Malaysia such as the Asian tsunami. Moving under these circumstances can expose people to risk, in- As one of the most robust economies in South-East Asia, Malaysia has attracted crease their vulnerability and erode their capa- many migrant workers (officially measured at around 7 percent of the population in bilities. But of course in these cases it is not the 2005). The Malaysian labour force at the end of 2008 was almost 12 million, about migration per se but the underlying drivers that 44 percent of the 27 million residents, and included about 2.1 million legal migrants cause such deterioration in outcomes. This sec- from Bangladesh, Indonesia and other Asian countries. The Malaysian government tion reviews the outcomes associated with three has tended to tolerate unauthorized migration, while regularizations have sometimes broad drivers: conflict, development-induced been coupled with a ban on new entries and stepped up enforcement. displacement and trafficking. Since 1972, Malaysia’s People’s Volunteer Corps (Ikatan Relawan Rakyat or RELA) has helped to enforce laws, including immigration laws. RELA volunteers, who 3.5.1 When insecurity drives number about 500,000, are allowed to enter workplaces and homes without warrants, movement to carry firearms and to make arrests after receiving permission from RELA leaders. People who flee insecurity and violence typically Migrant activists say that RELA volunteers have become vigilantes, planting evidence see an absolute collapse in their human devel- to justify arrests of migrants and using excessive force in their policing. The govern- opment outcomes. But migration nonetheless ment has recently announced its intention to curb abuses and is currently looking into protects them from the greater harm they would ways of improving RELA by providing training to its members. doubtless come to if they were to stay put. Several Source: Crush and Ramachandran (2009), Vijayani (2008) and Migration DRC (2007). forms of protection are available for refugees, es- pecially for those covered by the 1951 Refugee Convention—which defines the criteria under as those in Malaysia and South Africa in 2008 which individuals may be granted asylum by its and Northern Ireland in 2009, for example— signatory countries and sets out their associated with serious repercussions for both the indi- rights—and thus under the UNHCR mandate. viduals involved and the societies as a whole. 88 This protection has allowed millions of people Experience suggests that such outbreaks typi- to move to new safe and secure environments. cally occur where political vacuums allow Contemporary conflicts are increasingly as- unscrupulous local actors to manipulate under- sociated with large population movements, in- lying social tensions.89 cluding deliberate displacement of civilians as a Ironically, although intolerance often results weapon of war.91 While some are able to flee to in resistance to social contact, evidence suggests more distant places in North America, Western that increased social contact between migrants Europe and Australasia, most displaced people and non-migrants can improve levels of toler- relocate within or near their country of origin. ance for migrant groups and counter existing Even if camps host only about a third of those biases.90 Clearly, moderate politicians, govern- displaced by conflicts,92 these settlements have ment authorities and NGOs all have a critical come to symbolize the plight of people in poor, part to play in designing and delivering policies conflict-affected regions. A contemporary ex- and services that facilitate integration and avert ample is the people of Darfur, Sudan, who fled escalated tensions. Having legislation on the their villages in the wake of attacks that de- books is not enough: it must be accompanied by stroyed their cattle and crops, wells and homes, leadership, accountability and informed public to join what was already the largest displaced debate (chapter 5). population in the world in the wake of the long- running war in southern Sudan. 3.5 Understanding outcomes When the poor and destitute flee combat from negative drivers zones, they run severe risks. Conflict weakens Some people move because their luck im- or destroys all forms of capital and people are proves—they win the green card lottery, or a cut off from their existing sources of income, friend or relative offers a helping hand to take services and social networks, heightening their up a new opportunity in the city. But many vulnerability. After flight, those displaced may others move in response to difficult circum- have escaped the most direct physical threats, stances—economic collapse and political unrest but still face a range of daunting challenges. in Zimbabwe, war in Sudan, natural disasters Security concerns and local hostility rank high 62
  • 73. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 3 among their problems, especially in and around Map 3.1 Conflict as a driver of movement in Africa camps.93 In civil wars, the internally displaced Conflict, instability and population movement in Africa may face harassment from government and ani- mosity from local people. Nevertheless, it is important to bear in mind Tunisia that conflict and insecurity drive only a small Morocco share of all movement—about one tenth of in- ternational movement and around one twen- Algeria Libyan Arab Egypt tieth of internal movement. There are regional Western Jamahiriya 0.3 differences: Africa has been more extensively af- Sahara fected, conflict being associated with about 13 20.0 Mauritania percent of international movement on the con- Mali Niger 3.7 0.3 tinent. Map 3.1 shows the location of conflicts Senegal 0.6 Chad Eritrea Sudan and major flows of people displaced within and Gambia Burkina Faso Djibouti 16.6 Guinea- across borders in Africa. While the map paints Bissau Guinea Ghana Benin Nigeria 0.6 a sombre picture, we underline that the vast Sierra Leone Togo 3.5 Central 2.5 Ethiopia Cameroon African Republic 2.7 majority of migration in Africa is not conflict- Liberia Côte 24.9 1.2 Somalia d’Ivoire induced and that most Africans move for the Equatorial Guinea 9.4 23.8 same reasons as everyone else.94 Sao Tome and Principe Gabon Congo Uganda Kenya 2.6 Rwanda Beyond continuing insecurity, trying to 7.0 0.4 Burundi earn a decent income is the single greatest chal- Congo, 0.1 (Dem. Rep. of the) Tanzania lenge that displaced people encounter, especially 0.3 where they lack identity papers.95 In commis- sioned case studies,96 Uganda was the only one Angola of six countries where refugees were legally al- Zambia Malawi Mozambique lowed to move around freely, to accept work and Zimbabwe Madagascar to access land. About 44 percent of Uganda’s Namibia working-age camp population was employed, Botswana whereas in all five other countries the figure was Swaziland below 15 percent. Even if the displaced are per- Lesotho mitted to work, opportunities are often scarce. South Africa The human development outcomes of those driven to move by insecurity vary considerably. While the UN Guiding Principles on Internal Displacement have raised awareness, internally Recent conflict zones Internally displaced persons (end of 2008) Burundi 100,000 displaced people—80 percent of whom are Ongoing UN peacekeeping missions (2009) Central African Republic 108,000 Refugee flows in 2007 (in thousands) women and children—do not benefit from the 23.8 Chad 180,000 same legal rights as refugees.97 Roughly half Congo Up to 7,800 Number of refugees (end of 2008) the world’s estimated 26 million internally Congo, DRC 1,400,000 0–1,000 Côte d’Ivoire At least 621,000 displaced people receive some support from 1,000–10,000 Ethiopia 200,000–300,000 UNHCR, IOM and others, but sovereignty Kenya 300,000–600,000 10,000–100,000 is often invoked as a justification for restrict- Liberia Undetermined 100,000–523,032 ing international aid efforts. In 2007, Sudan, Rwanda Undetermined Senegal 10,000–70,000 Myanmar and Zimbabwe each had more than Somalia 1,300,000 500,000 crisis-affected people who were be- Sudan 4,900,000 yond the reach of any humanitarian assistance.98 Uganda 869,000 Even in less extreme cases, malnutrition, poor Zimbabwe 570,000–1,000,000 access to clean water and health care, and lack of documentation and property rights are typi- Source: UNHCR (2008) and IDMC (2008). cal among the internally displaced. However, Note: This map illustrates refugee flows based on official UNHCR data and misses important flows associated with instability, as in the case of Zimbabweans fleeing to South Africa for example. some governments have made concerted efforts 63
  • 74. 3 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development to improve the rights and living conditions of refugees than for the general population (figure their internally displaced populations.99 3.12). These patterns reflect both the effects of The situation of international refugees also humanitarian assistance in camps and the gen- varies, but can be bleak, especially in cases of erally poor human development conditions and protracted conflict, such as Palestine. Such indicators prevailing in countries that host the cases account for roughly half of all refugees. bulk of refugees. Our commissioned analysis confirmed overall As noted above, most refugees and internally weak human development outcomes, alongside displaced people do not end up in camps at all, some heterogeneity across groups and countries. or at least not for long. For example, less than a The incidence of sexual and gender-based vio- third of Palestinian refugees live in UNRWA- lence is high. Paradoxically, however, women in administered camps.101 On average, those who Burundi and Sri Lanka were reportedly empow- relocate to urban centres seem to be younger and ered as they adopted new social roles as protec- better educated, and may enjoy better human de- tors and providers for their families.100 velopment outcomes than those living in camps. Education and health indicators in refugee Others, usually the better off, may be able to flee camps are sometimes superior to those of sur- to more distant and wealthier countries, some- rounding local populations. Our study found times under special government programmes. that the share of births attended by skilled Only a minority of asylum seekers succeed medical personnel in camps surveyed in Nepal, in obtaining either refugee status or residency, Tanzania and Uganda was significantly higher and those whose request is denied can face pre- than among these countries’ population as a carious situations.102 Their experience depends whole. Similarly, education indicators—such on the policies of the destination country. as gross primary enrolment ratios and pupil-to- Developed countries in our policy assessment teacher ratios—were better among camp-based allowed humanitarian migrants access to emer- gency services, but more restricted access to preventive services, whereas in the developing Figure 3.12 School enrolment among refugees often exceeds countries in our sample, access to public health that of host communities in developing countries services was even more restricted (figure 3.8). Gross primary enrolment ratios: refugees, host populations Finding durable long-term solutions to the and main countries of origin, 2007 problem in the form of sustainable return or successful local integration has proved a major Kenya Refugees challenge. In 2007, an estimated 2.7 million in- Population of main ternally displaced people and 700,000 refugees, origin country representing about 10 and 5 percent of stocks Uganda Population of respectively, returned to their areas of origin.103 asylum country Perhaps the Palestinian case, more than any Bangladesh other, illustrates the hardships faced by refu- gees when conflict is protracted, insecurity is rampant and local economic opportunities are Tanzania almost non-existent.104 In other cases, gradual integration into local Nepal communities, sometimes through naturaliza- tion, has taken place in a number of develop- ing and developed countries, although refugees Thailand tend to be relatively disadvantaged, especially as regards labour market integration.105 | | | | | | | | | 0 20 40 60 80 100 120 140 160 Gross enrolment ratio (%) 3.5.2 Development-induced displacement Source: de Bruijn (2009), UNHCR (2008) and UNESCO Institute for Statistics (2008b). Outcomes may also be negative when people are displaced by development projects. The classic 64
  • 75. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 3 case of this occurs when large dams are built This issue was addressed in the Guiding Above all, trafficking to provide urban water supplies, generate elec- Principles on Internal Displacement mentioned is associated with tricity or open up downstream areas for irriga- above. The principles provide that, during the restrictions on human tion. Agricultural expansion is another major planning stage, the authorities should explore freedom and violations cause, as when pastoralists lose traditional riv- all viable options for avoiding displacement. of basic human rights erine grazing lands when these are developed Where it cannot be avoided, it is up to the au- for irrigated cash crops. Infrastructure projects thorities to make a strong case for it, stating why such as roads, railways or airports may also dis- it is in the best interests of the public. The sup- place people, while the energy sector—mining, port and participation of all stakeholders should power plants, oil exploration and extraction, be sought and, where applicable, agreements pipelines—may be another culprit. Parks and should stipulate the conditions for compensa- forest reserves may displace people when man- tion and include a mechanism for resolving aged in a top-down style rather than by local disputes. In all instances, displacement should communities. not threaten life, dignity, liberty or security, These types of investment generally expand and should include long-term provisions for ad- most people’s opportunities—in terms of pro- equate shelter, safety, nutrition and health for viding yield-increasing technology, links to those displaced. Particular attention should be markets and access to energy and water, among given to the protection of indigenous peoples, other things.106 But how the investments are minorities, smallholders and pastoralists. designed and delivered is critical. By the 1990s These principles can help inform devel- it was recognized that such interventions could opment planners as to the social, economic, have negative repercussions for the minority of cultural and environmental problems that people directly affected, and were criticized on large- and even small-scale development proj- social justice and human rights grounds.107 One ects can create. Incorporating such analysis in vocal critic has been the World Commission on planning processes, as has been done for some Dams, which has stated that, “impoverishment major sources of development finance—includ- and disempowerment have been the rule rather ing the World Bank, which has an Involuntary than the exception with respect to resettled Resettlement Policy—has been an important people around the world,”108 and that these out- step forward.111 Such policies allow for rights of comes have been worst for indigenous and tribal appeal by aggrieved parties through inspection peoples displaced by big projects. panels and other mechanisms. Approaches of Among the impacts observed in indigenous this kind can enable favourable human develop- communities are loss of assets, unemployment, ment outcomes for the majority while helping to debt bondage, hunger and cultural disintegra- mitigate the risks borne by the displaced minor- tion. There are many such examples, which ity, though the challenges remain large. have been well documented elsewhere.109 The India Social Institute estimates that there are 3.5.3 Human trafficking about 21 million development-induced dis- The images associated with trafficking are often placed persons in India, many of whom belong horrendous, and attention tends to focus on its to scheduled castes and tribal groups. In Brazil association with sexual exploitation, organized the construction of the Tucuruí Dam displaced crime, violent abuse and economic exploitation. an estimated 25,000 to 30,000 people and sig- Human trafficking not only adversely affects nificantly altered the lifestyle and livelihood individuals but can also undermine respect for means of the Parakanã, Asurini and Parkatêjê whole groups. However, the increasing focus on indigenous groups. Poor resettlement planning this phenomenon has not yet provided a reliable split up communities and forced them to relo- sense of either its scale or its relative importance cate several times, often in areas that lacked the in movements within and across borders (chap- necessary infrastructure to serve both the needs ter 2). of a growing migrant population (pulled in by Above all, trafficking is associated with re- construction jobs) and those displaced by the strictions on human freedom and violations of project.110 basic human rights. Once caught in a trafficking 65
  • 76. 3 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Trafficking can be network, people may be stripped of their travel state to its citizens and to authorized migrant most effectively documents and isolated, so as to make escape workers.115 More generally, of course, trafficking combated through difficult if not impossible. Many end up in debt can be most effectively combated through better better opportunities and bondage in places where language, social and opportunities and awareness at home—the abil- awareness at home— physical barriers frustrate their efforts to seek ity to say ‘no’ to traffickers is the best defence. the ability to say ‘no’ to help. In addition, they may be reluctant to iden- Difficulties in distinguishing trafficking traffickers is the best tify themselves, since they risk legal sanctions from other types of exploitation, as well as chal- defence or criminal prosecution. People trafficked into lenges involved in defining exploitative practices, sex work are also at high risk of infection from further complicate the rights of trafficked peo- HIV and other sexually transmitted diseases.112 ple. Problems can arise over enforcement. It ap- One basic constraint in assessing the im- pears that trafficking is sometimes very broadly pacts of trafficking relates to data. The IOM’s interpreted to apply to all migrant women who Counter Trafficking Module database contains engage in sex work. This can be used to justify data on fewer than 14,000 cases that are not a their harassment and deportation, making them representative sample, and the same applies to even more vulnerable to exploitation. And once the database of the United Nations Office on identified, they are virtually always deported or Drugs and Crime (UNODC).113 The picture referred to assistance programmes conditional that emerges from these data, alongside existing on cooperation with law enforcement. studies and reports, suggests that most people Anti-trafficking initiatives have burgeoned who are trafficked are young women from mi- in recent years. Interventions to reduce vulner- nority ethnic groups. This is confirmed by other ability in potential source communities, such as sources—for example a study in South-eastern awareness campaigns and livelihood projects, Europe, which found that young people and have been undertaken. Assistance programmes ethnic minorities in the rural areas of post-con- have also provided counselling, legal aid and flict countries were vulnerable to trafficking, as support for return and reintegration. Some of they tended to experience acute labour market these programmes are proving successful, such exclusion and disempowerment.114 However, as the use of entertainment and personal stories this picture may be biased, since it is possible as community awareness tools in Ethiopia and that males are less willing to self-report for fear Mali, or door-to-door mass communication they will be refused victim status. In addition campaigns as in the Democratic Republic of to social and economic exclusion, violence and the Congo.116 Other initiatives, however, have exploitation at home or in the home community led to counterproductive and sometimes even increase vulnerability to trafficking. So too does disastrous outcomes, including prejudicial limi- naïve belief in promises of well-paid jobs abroad. tations on women’s rights. In Nepal, for exam- Sexual exploitation is the most commonly ple, prevention messages discouraged girls and identified form of human trafficking (about women from leaving their villages, while HIV 80 percent of cases in the UNODC database), awareness campaigns stigmatized returnees.117 with economic exploitation comprising most of Anti-trafficking initiatives clearly raise very the balance. For women, men and children traf- complex and difficult challenges, which need to ficked for these and other exploitative purposes, be carefully handled. bonded labour, domestic servitude, forced mar- The lines between traffickers on the one riage, organ removal, begging, illicit adoption hand and recruiters and smugglers on the other and conscription have all been reported. can be blurred. For example, the business of re- Alongside the lack of power and assets of the cruitment expands to include numerous layers individuals involved, the negative human devel- of informal sub-agents. These sub-agents, work- opment outcomes of trafficking can be partly ing under the umbrella of legitimate recruiters associated with the legal framework of destina- can reduce accountability and increase costs. tion countries. Restrictive immigration controls The risks of detention and deportation are high. mean that marginalized groups tend to have ir- Smuggling costs in some cases include bribing regular status and so lack access to the formal la- corrupt border officials and manufacturing false bour market and the protections offered by the documents.118 66
  • 77. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 3 3.6 Overall impacts Figure 3.13 Significant human development gains to internal movers We have studied the discrete impacts of migra- Ratio of migrants’ to non-migrants’ estimated HDI in selected tion on incomes, health, education and aspects developing countries, 1995–2005 of empowerment and agency—and looked at the negative outcomes that can occur when people Guinea move under duress. Differences in the HDI are a simple way to capture overall changes. Madagascar Our background research found very large Uganda average differences between the HDI of mi- grants and that of non-migrants, moving in- Indonesia ternally and across borders. We found that, Viet Nam on average, migrants to OECD countries had an HDI about 24 percent higher than that of Côte d’Ivoire people who stayed in their respective countries of origin.119 But the gains are large not only for Ghana those who move to developed countries: we also Kyrgyzstan found substantial differences between inter- nal migrants and non-migrants.120 Figure 3.13 Paraguay shows that, in 14 of the 16 developing countries Cameroon covered by this analysis, the HDI for internal migrants is higher than that of non-migrants. Bolivia In some cases the differences are substantial. For internal movers in Guinea, for example, the Nicaragua HDI for migrants is 23 percent higher than for Colombia non-migrants—only one percentage point lower than for migrants to OECD countries. If these Peru migrants were thought of as a separate country, Zambia they would be ranked about 25 places higher than non-migrants in the global HDI. Guatemala There are two major exceptions to the over- all pattern of improved well-being from in- | | | | | | 0.6 0.8 1.0 1.2 1.4 1.6 ternal movement: in Guatemala and Zambia HDI ratio internal migrants appear to do worse than non- migrants. Both these cases underline the risks that accompany migration. In Guatemala most Source: Harttgen and Klasen (2009). movers were displaced by violence and civil war in the 1980s and early 1990s, while in Zambia migrants faced extreme urban poverty follow- ing the successive economic shocks that have levels of happiness and health were very similar hit this country over the past 20 years. In a few among migrants and non-migrants: 84 percent other cases—Bolivia and Peru, for example— of migrants felt happy (compared to 83 percent the overall human development outcome ap- of non-migrants), while 72 percent felt that pears marginal despite sizeable income gains, their health was good or very good (compared suggesting poor access to services as a factor in- to 70 percent of non-migrants); only 9 percent hibiting well-being. However, these exceptional were ‘not satisfied’ with life (compared to 11 cases serve to emphasize the norm, which is that percent of non-migrants). The share of migrants most movers are winners. reporting that they felt quite or very happy was These findings for international movers are highest in developed countries. Similar shares borne out by evidence on migrants’ own sense of of foreign and locally born respondents—more well-being (figure 3.14). We analysed data for 52 than 70 percent—felt that they have ‘freedom countries in 2005 and found that self-reported and choice over their lives’.121 67
  • 78. 3 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Figure 3.14 Migrants are generally as happy as locally-born people 3.7 Conclusions Self-reported happiness among migrants and locally-born The complex effects associated with movement people around the world, 2005/2006 are difficult to summarize simply. The broad findings presented in this chapter underline the role of movement in expanding human free- Locally born Total doms that was outlined in chapter 1. We saw that people who move generally do enhance Foreign born their opportunities in at least some dimensions, Total with gains that can be very large. However, we also saw that the gains are reduced by policies Low HDI at home and destination places as well as by the constraints facing individuals and their fami- Medium HDI lies. Since different people face different oppor- High HDI tunities and constraints, we observed significant inequalities in the returns to movement. The Very high HDI cases in which people experience deteriorations in their well-being during or following the pro- | | | | | | cess of movement—conflict, trafficking, natural 0 20 40 60 80 100 Percent of responses disasters, and so on—were associated with con- straints that prevent them from choosing their place in life freely. “All things considered, would you say you are:” A key point that emerged is that human Very happy Quite happy Not very happy Not at all happy movement can also be associated with trade- offs—people may gain in some and lose in other dimensions of freedom. However, the losses can Source: HDR team estimates based on WVS (2006). be alleviated and even offset by better policies, as we will show in the final chapter. 68
  • 79. Impacts at origin and destination 4
  • 80. Movement has multiple impacts on other people besides those who move—impacts that critically shape its overall effects. This chapter explores impacts in the country of origin and in the host country while underlining their interconnectedness. Families with members who have moved elsewhere in the country or abroad tend to experience direct gains, but there can also be broader benefits, alongside concerns that people’s departure is a loss to origin communities. As regards impacts on places of destination, people often believe that these are negative—because they fear that newcomers take jobs, burden public services, create social tensions and even increase criminality. The evidence suggests that these popular concerns are exaggerated and often unfounded. Still, perceptions matter—and these warrant careful investigation to help frame the discussion of policy.
  • 81. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 4 Impacts at origin and destination Among people who do not move but can be affected by movement are the families of movers and communities at places of origin and destination. The multiple impacts of movement in these differ- ent places are critical in shaping the overall human development effects of movement; this chapter addresses each in turn. At places of origin, impacts can be seen on in- investigate the vast empirical literature on these come and consumption, education and health, issues, which reveals that these fears are exagger- and broader cultural and social processes. These ated and often unfounded. Nevertheless, these impacts are mostly favourable, but the concern perceptions matter because they affect the po- that communities lose out when people move litical climate in which policy decisions about needs to be explored. Our review of the evi- the admission and treatment of migrants are dence shows that impacts are complex, context- made—fears may stoke the flames of a broader specific and subject to change over time. The hostility to migrants and allow political extrem- nature and extent of impacts depend on who ists to gain power. Indeed, historical and con- moves, how they fare abroad and their procliv- temporary evidence suggests that recessions are ity to stay connected, which may find expression times when such hostility can come to the fore. in flows of money, knowledge and ideas, and in We end this chapter by tackling the thorny issue the stated intention to return at some date in the of public opinion, which imposes constraints on future. Because migrants tend to come in large the policy options explored in the final chapter. numbers from specific places—e.g. Kerala in India and Fujian Province in China—impacts 4.1 Impacts at places of origin on local communities may be more pronounced Typically, only a small share of the total population than national impacts. Yet the flow of ideas can of an origin country will move. The exceptions— also have far-reaching effects on social norms countries with significant shares abroad—are and class structures, rippling out to the broader often small states, including Caribbean nations community over the longer term. Some of these such as Antigua and Barbuda, Grenada, and Saint impacts have traditionally been seen as negative, Kitts and Nevis. In these cases the share can ex- but a broader perspective suggests that a more ceed 40 percent. The higher the share, the more nuanced view is appropriate. In this light we also likely it is that impacts on people who stay will examine the extent to which national develop- be more pervasive and more profound. While the ment plans, such as poverty reduction strategies discussion below focuses on developing countries, (PRSs), reflect and frame efforts of developing it is important to bear in mind that, as shown in countries to promote gains from mobility. chapter 2, emigration rates for low-HDI countries Much academic and media attention has are the lowest across all country groupings. been directed to the impacts of migrants on In general, the largest impacts at places of places of destination. One widespread belief is origin are felt by the households with an absent that these impacts are negative—newcomers are migrant. However, the community, the region seen as ‘taking our jobs’ if they are employed, liv- and even the nation as a whole may be affected. ing off the taxpayer by claiming welfare benefits We now look at each of these in turn. if they are not employed, adding an unwanted extra burden to public services in areas such as 4.1.1 Household level effects health and education, creating social tensions In many developing countries, movement is a with local people or other immigrant groups household strategy aimed at improving not only and even increasing criminal behaviour. We the mover’s prospects but those of the extended 71
  • 82. 4 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Despite these financial family as well. In return for supporting the An important function of remittances is to di- rewards, separation move, the family can expect financial remit- versify sources of income and to cushion families is typically a painful tances when the migrant is established—trans- against setbacks such as illness or larger shocks decision incurring high fers that typically far outweigh the initial outlay caused by economic downturns, political con- emotional costs for or what the mover might have hoped to earn in flicts or climatic vagaries.5 Studies in countries as both the mover and the place of origin. These transfers can in turn diverse as Botswana, El Salvador, Jamaica and the those left behind be used to finance major investments, as well as Philippines have found that migrants respond to immediate consumption needs. weather shocks by increasing their remittances, Despite these financial rewards, separation is although it is difficult to establish whether these typically a painful decision incurring high emo- effectively serve as insurance. Recent examples tional costs for both the mover and those left include the 2004 Hurricane Jeanne in Haiti, the behind. In the words of Filipina poet Nadine 2004 tsunami in Indonesia and Sri Lanka and Sarreal: the 2005 earthquake in Pakistan.6 In a sample Your loved ones across that ocean of poor countries, increased remittances were Will sit at breakfast and try not to gaze found to offset some 20 percent of the hurricane Where you would sit at the table damage experienced,7 while in the Philippines Meals now divided by five about 60 percent of declines in income due to Instead of six, don’t feed an emptiness.1 rainfall shocks were offset.8 In El Salvador crop failure caused by weather shocks increased the The fact that so many parents, spouses and probability of households sending a migrant to partners are willing to incur these costs gives an the United States by 24 percent.9 idea of just how large they must perceive the re- Migrants can provide this kind of protection wards to be. if their incomes are large enough and do not vary Financial remittances are vital in improving the in tandem with their families’. This depends on livelihoods of millions of people in developing the nature and breadth of the shock, as well as countries. Many empirical studies have confirmed the location of the migrant. For example, remit- the positive contribution of international remit- tances may not provide much insurance against tances to household welfare, nutrition, food, health the effects of the current global economic reces- and living conditions in places of origin.2 This con- sion, as migrant workers almost everywhere suf- tribution is now well recognized in the literature on fer retrenchment just when their families most migration and reflected in the increasingly accurate need support (box 4.2). Remittances to develop- data on international remittances published by the ing countries are expected to fall from US$308 World Bank and others, illustrated in map 4.1. billion in 2008 to US$293 billion in 2009.10 Even those whose movement was driven by conflict Even when the total volume of remittances can be net remitters, as illustrated at various points is large, their direct poverty-reducing impact in history in Bosnia and Herzegovina, Guinea- depends on the socio-economic background of Bissau, Nicaragua, Tajikistan and Uganda, where those who moved. Within the Latin America remittances helped entire war-affected communi- region, for example, a recent study found that ties to survive.3 in Mexico and Paraguay remittance-receiving In some international migration corridors, households were primarily from the bottom of money transfer costs have tended to fall over the income and education distribution, whereas time, with obvious benefits for those sending the opposite pattern was found in Peru and and receiving remittances.4 Recent innovations Nicaragua.11 More generally, however, restric- have also seen significant falls in costs at the na- tions imposed by the limited opportunities of tional level, as in the case of Kenya described in the low-skilled to move across borders mean that box 4.1. With the reduction in money transfer remittances do not tend to flow directly to the costs, families who once relied on relatives and poorest families,12 nor to the poorest countries.13 close family friends or who used informal ave- Take China, for example: because migrants nues such as the local bus driver to remit are now generally do not come from the poorest house- opting to send money through banks, money holds, the aggregate poverty impact of internal transfer companies and even via cell-phones. migration is limited (an estimated 1 percent 72
  • 83. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 4 Map 4.1 Remittances flow primarily from developed to developing regions Flows of international remittances, 2006–2007 30.1 0.9 17.3 0.3 5.3 4.4 2.2 42.0 52.5 Europe 15.9 North America 2.9 Asia 0.4 0.02 1.2 10.3 0.2 3.1 1.6 Africa 0.3 36.3 Latin America 2.2 1.9 & Caribbean 0.3 0.01 1.1 4.0 2.8 1.5 Oceania 0.02 0.9 0.1 3.6 0.08 0.5 0.02 Regions Remittances, 2006 (in US$ billions) Remittances as a share of GDP, 2007 North America no data 10.0%–14.9% Europe Intra- 0.0%–0.4% 15.0%–19.9% Oceania regional 0.5%–0.9% 20.0%–24.9% Latin America and the Caribbean remittances 1.0%–4.9% 25.0%–29.9% Asia 5.0%–9.9% >30% Africa Source: HDR team data based on Ratha and Shaw (2006) and World Bank (2009b). reduction), although this still translates into al- for this report, using panel data and controlling most 12 million fewer poor.14 At the same time, for selection bias, examined the cases of Indonesia some migrants do come from poor households between 1994 and 2000 and Mexico between and significant remittances sometimes flow to 2003 and 2005. In Indonesia, where almost non-family members, which allows for broader half of all households had an internal migrant, benefits—as has been found for Fiji and Jamaica, poverty rates for non-migrants were essentially for example.15 stable for the period (which included the East The poverty-reducing effects of internal mi- Asian financial crisis), falling slightly from 40 gration, which have been demonstrated by stud- to 39 percent, but declined rapidly for migrants, ies in a diverse range of national situations, may from 34 to 19 percent. In Mexico, where about 9 be even more significant. In Andhra Pradesh and percent of households had an internal migrant, Madhya Pradesh in India poverty rates in house- poverty rates rose sharply from 25 to 31 percent holds with a migrant fell by about half between for non-migrants for the period (which included 2001/02 and 2006/07,16 and similar results were the 2001/02 recession), but only slightly, from 29 found for Bangladesh.17 Large gains have also to 30 percent, for migrants. In both countries, at been reported from panel data, tracking individ- the outset households with a migrant made up uals over time, in the Kagera region of Tanzania less than half of the top two wealth quintiles, but between 1991 and 2004.18 Research conducted over time this share rose to nearly two thirds.19 73
  • 84. 4 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Box 4.1 How cell-phones can reduce money transfer costs: the case of Kenya For many people in remote rural areas of developing countries, the A recent survey of users across Kenya found that, in just two costs of receiving money remain high: recipients typically have to years, M-PESA has expanded rapidly. It is now used by some 6 travel long distances to a regional or national capital to collect cash, million people or 17 percent of the population—out of 26 percent or the cash has to be hand-delivered by an intermediary, who may who are cell-phone owners—and is supported by a network of take a sizeable margin. more than 7,500 agents. Transfers can be made from the port city The rapid diffusion of cell-phone technology over the past de- of Mombasa to Kisumu on the shores of Lake Victoria, or from cade has led to the development of innovative money transfer sys- Nairobi in the south to Marsabit in the north—both two-day bus tems in several countries. For example, in Kenya, a leading cell-phone trips—with the push of a few buttons and at a cost of less than a company, Safaricom, teamed up with donors to pilot a system that dollar. By mid-2008, the volume of money sent had reached some subsequently led to the launch in 2007 of M-PESA (meaning ‘Mobile- 8 percent of GDP, mostly in the form of a large number of relatively Cash’). Anyone with a cell-phone can deposit money in an account small transactions. and send it to another cell-phone user, using M-PESA agents distrib- uted across the country. Source: Jack and Suri (2009). One dimension of movement that appears to communities. Improvements in nutrition and affect remittance flows is gender. Evidence sug- other basic consumption items greatly enhance gests that women tend to send a larger propor- human capital and hence future incomes. 25 tion of their incomes home, on a more regular Similarly, spending on schooling is often a pri- basis, though their lower wages often mean that ority for families receiving remittances, because the absolute amounts are smaller.20 it increases the earning power of the next genera- There is also a temporal dimension to these tion. Second, most types of spending, especially flows. Over time, the knock-on effects of remit- on labour-intensive goods and services such as tances may substantially broaden the impacts housing and other construction, will benefit the on poverty and inequality.21 The poor may gain local economy and may have multiplier effects. 26 when remittances are spent in ways that gener- All of these effects are positive. ate local employment, such as building houses, or Families with migrants appear more likely when businesses are established or expanded. 22 to send their children to school, using cash from Some studies have found that remittance re- remittances to pay fees and other costs. This re- cipients exhibit greater entrepreneurship and a duces child labour. And, once there, the children higher marginal propensity to invest than house- of migrants are more likely to finish school, as the holds without a migrant. 23 Positive investment better prospects associated with migration affect effects can take decades to materialize in full, social norms and incentives. 27 In Guatemala in- however, and are complex and far from auto- ternal and international migration is associated matic. The lag may reflect delays in the sending with increased educational expenditures (45 and of remittances as migrants adapt to their new 48 percent respectively), especially on higher lev- homes, or political and economic conditions els of schooling. 28 In rural Pakistan temporary in places of origin—such as a poor climate for migration can be linked with increased enrol- investment—which can inhibit or deter trans- ment rates and declines in school dropout rates fers.24 Lastly, remittances can also create a store that exceed 40 percent, with larger effects for of capital to fund further migration, years after girls than for boys.29 In our own commissioned the first family member has left. research, similar results were found in Mexico, Some commentators discount the impor- where children in households with an internal tance of remittances because they are partly migrant had a 30–45 percent higher probability spent on consumption. This critique is mis- of being in an appropriate grade for their age.30 taken, for two broad reasons. First, consumption The prospect of moving can strengthen in- can be inherently valuable and often has long- centives to invest in education.31 This has been term, investment-like effects, especially in poor predicted in theory and shown in practice in 74
  • 85. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 4 some countries. Emigration of Fijians to high- Box 4.2 The 2009 crisis and remittances skilled jobs in Australia, for example, has en- couraged the pursuit of higher education in Fiji. The 2009 economic crisis, which began in major destination countries and has now This effect is so large that, while roughly a third gone global, has shrunk flows of remittances to developing countries. There is already of the Indo-Fijian population has emigrated in evidence of significant declines in flows to countries that depend heavily on remit- the past three decades and skilled workers are tances, including Bangladesh, Egypt, El Salvador and the Philippines. over-represented among emigrants, the absolute Countries and regions vary in their exposure to the crisis via remittance effects. number of skilled Indo-Fijian workers in Fiji has Remittances to Eastern European and Central Asian countries are forecast to suffer greatly increased.32 A number of governments, the biggest drop in both relative and absolute terms, partly reflecting the reversal of including the Philippines, have deliberately the rapid expansion that had followed European Union accession and the economic sought to promote work abroad in part by facili- boom in the Russian Federation. In Moldova and Tajikistan, where remittance shares tating the generation of skills at home.33 of GDP are the highest in the world (45 and 38 percent respectively), flows are pro- The impacts of migration prospects on jected to shrink by 10 percent in 2009. El Salvador is facing a significant decline in schooling incentives are shaped by the con- remittances, which account for over 18 percent of its GDP. text and the prospects themselves. In Mexico, About three quarters of remittances to sub-Saharan Africa come from the United for instance, where low-skilled, often ir- States and Europe, which have been badly affected by the downturn (chapter 2). It regular migration predominates, boys were remains to be seen whether these sources will prove more or less resilient than official more likely to drop out of school to take up development aid and private investment flows. this option. 34 In our commissioned study of Chinese census data at the provincial level, Source: Ratha and Mohapatra (2009a,b). investments in schooling in rural source com- munities responded to the skills needed for job opportunities outside the province. Thus, Figure 4.1 The global recession is expected to impact remittance flows where internal migrants had secondary educa- Projected trends in remittance flows to developing regions, tion, this generally encouraged the completion 2006–2011 of higher levels by children remaining in the community, whereas in provinces where mi- 70 grants tended to have completed only middle East Asia and the Pacific South Asia school, this was associated with lower high 60 Latin America school completion rates.35 and the Carribean The health outcomes of people who do not Remittance flows (in US$ billions) 50 Central and Eastern Europe move may be affected by migration, through and the CIS effects on nutrition, living conditions, higher 40 incomes and the transmission of knowledge Arab States and practices. There is evidence that the higher 30 incomes and better health knowledge associ- ated with migration have a positive influence on 20 Sub-Saharan Africa infant and child mortality rates.36 However, in Mexico at least, it was found that longer term 10 health outcomes may be adversely affected, be- | | | | | | 2006 2007 2008* 2009** 2010** 2011** cause levels of preventive health care (e.g. breast feeding and vaccinations) were lower when at * Estimate ** Forecasts least one parent had migrated.37 This may be Source: Ratha and Mohapatra (2009b) and The Economist Intelligence Unit (2009). associated with the higher work burden and/ Note: These regional groupings include all developing countries as per UNDP Regional Bureaux’ classification. For the complete list of countries in each region see ‘Classification of Countries’ in the Statistical Annex. or reduced levels of knowledge associated with single parenting or families with fewer adults. Moreover, when infectious diseases can be con- Offsetting the potential gains in consump- tracted in destination places, return travel can tion, schooling and health, children at home bring significant health risks to families at home. can be adversely affected emotionally by the The risks of HIV and other sexually transmitted process of migration. One in five Paraguayan diseases can be especially high.38 mothers residing in Argentina, for example, has 75
  • 86. 4 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development The effects of young children in Paraguay.39 Studies investigat- 4.1.2 Community and national level skills flows are less ing the possible impacts have found that these economic effects detrimental for origin depend on the age of the child when the separa- Beyond its direct impacts on families with mi- communities than is tion occurs (in the first years of life the impact grants, movement may have broader effects. often assumed may be greater), on the familiarity and attitude Migration-driven processes of social and cultural of the adult in whose care the child is left, and change can have significant impacts on entrepre- on whether the separation is permanent or tem- neurship, community norms and political trans- porary.40 The advent of cheap and easy commu- formations—impacts that are often felt down the nication, for example by cell-phone and Skype, generations. For example, Kenya, and indeed most has eased the separation of family members and of Africa, may be affected today and in the future has greatly helped the maintenance of ties and by Barack Obama Senior’s decision, taken five relationships in recent years. decades ago, to study in the United States. Most Movement can affect gender relations at of these effects are highly positive. However, one home.41 When women move, this can change tra- concern that needs to be addressed is the outflow ditional roles, especially those surrounding the of skills from source communities. care of children and the elderly.42 When men mi- Fears that the mobility of the skilled harms grate, rural women can be empowered by their ab- the economy of origin countries have long been sence: field studies conducted in Ecuador, Ghana, voiced, though the debate has become more nu- India, Madagascar and Moldova all found that, anced in recent years.48 The concerns surface with male migration, rural women increased their regularly in a range of small states and poorer participation in community decision-making.43 countries, but also extend to such countries as Norms adopted in a migrant’s new home—such as Australia, which sees many of its graduates go a higher age of marriage and lower fertility, greater abroad. This issue has, over the past few decades, educational expectations of girls, and labour force spawned a range of proposals, which are reviewed participation—can filter back to the place of ori- in chapter 5. But an important underlying point gin. This diffusion process may be accelerated in is that mobility is normal and prevalent, even in cases where the social and cultural gap between prosperous societies (chapter 2). Skilled people, sending and receiving countries is large.44 This has like everyone else, move in response to a perceived been confirmed by recent findings regarding the lack of opportunities at home and/or better op- transfer of fertility norms from migrants to the portunities elsewhere, for both themselves and extended family and friends at places of origin: their children. Attempts to curtail these move- lower numbers of children at the national level ments without addressing underlying structural become the norm in both places.45 causes are unlikely to be effective. There are also Overall, however, the evidence about impacts reasons to believe that the effects of skills flows on traditional gender roles is mixed. For example, are less detrimental for origin communities than where the lives of migrants’ wives at home remain is often assumed, as argued in box 4.3. largely confined to housekeeping, child-rearing One traditional concern has been that the and agricultural work, little may change—except departure of able-bodied youth leads to labour that their workloads increase. Gains in authority shortages and declines in output, particularly in may be temporary if male migrants resume their agriculture.49 In Indonesia, for example, com- position as head of the household on return, as munities faced shortages of labour for coopera- has been reported from Albania and Burkina tive farm work.50 However, in many developing Faso, for example.46 countries, movements of labour from agriculture  The transmission of norms may extend to par- to urban areas can be an important part of struc- ticipation in civic affairs. Recent studies in six Latin tural transformation. And to the extent that a American countries have found that individuals shortage of capital, not labour, constrains growth with greater connections to international migrant in most developing countries, remittances can be networks participate more in local community af- an important source of rural investment finance. fairs, are more supportive of democratic principles Migration can be a strong force for conver- and are also more critical of their own country’s gence in wages and incomes between source and democratic performance.47 destination areas. This is because, as mobility 76
  • 87. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 4 Box 4.3 Impacts of skills flows on human development The emigration of people with university degrees has attracted much worker leaves, the community is hardly affected. If, for example, popular and academic attention, especially because the shortage of teachers often do not show up to work, the direct impacts of their de- skills is acute in many poor countries. The evidence suggests that parture are unlikely to be large. While this should not weaken the drive improving local working conditions in order to make staying at home to address these underlying sources of inefficiency and waste, the more attractive is a more effective strategy than imposing restric- fact that staff may not currently be serving their communities is not a tions on exit. point that can simply be wished away in the debate about skills flows. It is important to recognize that the dreadful quality of key service Like other migrants, skilled people abroad often bring benefits to provision in some poor countries cannot be causally traced to the their countries of origin, through remittances and the development of emigration of professional staff. Systematic analysis of a new data- networks. As shown in figure 3.2, the absolute gain in income from base on health worker emigration from Africa confirms that low health migration can be huge, so that if only a fraction of the difference is re- staffing levels and poor public health conditions are major problems, mitted, the benefits to the home country can be considerable. Some but tend to reflect factors unrelated to the international movement research has suggested that the share of foreign direct investment in of health professionals—namely weak incentives, inadequate re- a developing country is positively correlated with the number of that sources, and limited administrative capacity. Migration is more accu- country’s graduates present in the investing country. Other studies rately portrayed as a symptom, not a cause, of failing health systems. have found that the more high-skilled emigrants from one country live The social cost associated with skilled emigration should not be in another, the more trade occurs between those countries. overestimated. Where graduate unemployment is high, as it often is Last but not least, significant numbers of skilled emigrants do re- in poor countries, the opportunity cost of departure may not be large. turn—a recent estimate suggested that about half do so, usually after If a highly productive but modestly paid worker leaves a community, about five years. Recent literature has also emphasized the increas- it suffers a significant loss; but if an equally skilled but unproductive ing importance of circular movement as transnational networks grow. Source: Clemens (2009b), Banerjee and Duflo (2006), Javorcik, Ozden, Spatareanu, and Neagu (2006), Rauch (1999), Felbermayr and Toubal (2008), Findlay and Lowell (2001) and Skeldon (2005). increases between two regions, their labour mar- The movement of skilled people happens kets become more integrated and large differences not only across but also within borders, as peo- in wages become more difficult to sustain. There ple move towards better opportunities. This is is considerable historical evidence, reviewed in illustrated in figure 4.2, which compares move- chapter 2, that enhanced mobility is associated ment within Brazil, Kenya, Philippines and with the reduction of wage disparities between United States to international rates. The strik- countries. Inequalities within countries can fol- ing result is that we find very similar patterns of low a bell-shaped pattern over time: progress in migration of skilled workers within and across some areas creates wealth and thus increases in- nations. In particular, the tendency for a higher equality, which encourages migration, which over proportion of skilled workers to emigrate from time in turn tends to reduce inequality. Studies small states is echoed in a similar tendency to have associated greater internal labour mobility migrate more from small localities. This sug- with a reduction in inter-regional income dispari- gests that the policy options explored in discus- ties in Brazil, India, Indonesia and Mexico.51 sions of local development—such as increased Interestingly, emigration rates for skilled incentives and improved working conditions— workers are substantially higher among women may also be relevant to policy-making related to than men in most developing countries.52 the emigration of skilled professionals abroad. Women with tertiary degrees are at least 40 per- More broadly, the economic effects of mi- cent more likely than male graduates to emigrate gration at the national level in countries of to OECD countries from a wide range of coun- origin are complex and, for the most part, dif- tries, including Afghanistan, Croatia, Ghana, ficult to measure. Networks may arise that fa- Guatemala, Malawi, Papua New Guinea, Togo, cilitate the diffusion of knowledge, innovation Uganda and Zambia. While this could reflect and attitudes and so promote development in various factors, structural and/or cultural barri- the medium to longer term. There is a host of ers to professional achievement at home seem the anecdotal evidence indicating that migrants most likely explanation.53 support productive activities in their countries 77
  • 88. 4 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development of origin, through technology transfer, the re- skilled workers abroad than have India and patriation of enhanced skills and exposure to Israel, for example.56 better working and management practices.54 Almost all the quantitative macro studies The Chinese government has pursued links on effects at the national level have focused with Chinese studying abroad to help pro- more narrowly on the scale and contribution mote academic excellence in its universities. of remittances. In 2007 the volume of officially Similarly, India’s ‘argonauts’—young graduates recorded remittances to developing countries who helped fuel the country’s high-tech boom was about four times the size of total official in the early 2000s—brought to their jobs the development aid.57 At this scale, remittances ideas, experience and money they had accu- are likely to be making a strong contribution mulated in the United States and elsewhere.55 to foreign exchange earnings relative to other The entire software industry model changed as sources in individual countries. In Senegal, firms increasingly outsourced production to for instance, remittances in 2007 were 12 India or based themselves there. In this case, times larger than foreign direct investment. skilled migration brought significant external Remittances represent a significant share of and dynamic effects, which benefit both work- GDP in a range of small and poor states, with ers and the industry in the place of origin. Tajikistan topping the list at 45 percent; for all The spread of new industries via international the countries in the top 20 remittance receiv- networks of skilled professionals can be rapid ing countries, the share exceeded 9 percent in and unpredictable, can find niches even amidst 2007; and in more than 20 developing coun- otherwise low levels of overall development, and tries, remittances exceed the earnings from the depends crucially on the openness of the busi- main commodity export. ness and political environment at home. It ap- However, two major qualifications should pears that countries such as the Islamic Republic be attached to these findings. First, the vast of Iran, Viet Nam and the Russian Federation, bulk of these flows do not go to the poorest which have more closed systems, have benefited countries. Of the estimated inflows of remit- less in high-tech business formation via their tances in 2007, less than 1 percent went to Figure 4.2 Skilled workers move similarly across and within nations Population and share of skilled workers who migrate internally and internationally Share of skilled workers who migrated (%) 1 0.8 0.6 0.4 US states Kenyan districts 0.2 Brazilian states Filipino provinces Countries 0 | | | | | | | | | | | 0 15 20 Total population (in millions, log scale) Source: Clemens (2009b). Note: Shares represented using Kernel density regressions. 78
  • 89. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 4 countries in the low-HDI category. So, for this on such things as kinship, skin colour, ethnic The ideas, practices, group, remittances are only about 15 percent group or religion, which are associated with un- identities and social of their official development aid. By contrast, equal access to land and other resources. capital that flow in Latin America and the Caribbean remit- The ideas, practices, identities and social back to families and tances in 2007 amounted to about 60 percent capital that flow back to families and communi- communities at origin of the combined volume of all foreign direct ties at origin are known as social remittances.64 are known as social investment and aid. Second, studies that have These remittances can arise through visits and remittances sought to trace the impacts of remittances on through rapidly improving communications. the long-term growth of the recipient country The case of the Dominican village of Miraflores, suggest that these impacts are generally small, where two thirds of families sent members to although the findings are mixed.58 This stems in Boston in the 1990s, shows the impacts on gen- part from the fact that the development impact der dynamics. Women’s roles changed, not only of remittances is ultimately contingent on local in Boston, where they went out to work, but also institutional structures.59 in the Dominican Republic, where they enjoyed Concerns have been expressed that remit- a more equal distribution of household tasks and tances create a form of ‘resource curse’, contrib- greater empowerment generally. Another exam- uting to undesirable currency appreciation and ple comes from Pakistanis at the Islamic Center thereby hampering competitiveness. Here again, of New England in  the United States, where however, the evidence is mixed.60 Moreover, re- women pray and run the mosque alongside mittances go to individuals and families and are men. News of these changes has travelled back thus distributed more widely than rents from to Karachi in Pakistan, where some women still natural resources, which flow only to govern- prefer traditional approaches but others are try- ments and a handful of companies and thus ing to create new spaces where women can pray can tend to exacerbate corruption. One positive and study together. Health is another area where macroeconomic feature of remittances is that social remittances have an impact. As a result of they tend to be less volatile than either official exposure abroad, visiting or returning migrants development aid or foreign direct investment, may bring back practices such as drinking safe although still subject to cyclical fluctuations, as water, keeping animals out of living spaces, or seen in 2009 (box 4.2).61 going for annual medical check-ups. In general, ‘remittance-led development’ The social and cultural effects of migration would not appear to be a robust growth strat- are not always positive, however. A counter- egy. Like flows of foreign aid, remittances alone example is the deportation of youth from the cannot remove the structural constraints to United States back to Central America, which economic growth, social change and better gov- has been likened to the export of gangs and gang ernance that characterize many countries with cultures.65 Although detailed data and analysis low levels of human development. That said, for are not available, a recent regional report found some small states, particularly those facing addi- that the distinction between home-grown gangs tional challenges related to remoteness, mobility (pandillas) and those exported from the United may be integral to an effective overall strategy for States (maras) is not always clear.66 In either case, human development (box 4.4). programmes that target at-risk individuals and communities with a view to preventing youth 4.1.3 Social and cultural effects and gang violence are needed, alongside inter- Mobility can have profound consequences for governmental cooperation and greater support social, class and ethnic hierarchies in origin and funding for reintegration programmes.67 communities if lower status groups gain ac- For many young people all over the world, cess to substantially higher income streams. spending time abroad is considered a normal This is illustrated by the cases of the Maya in part of life experience and migration marks the Guatemala62 and the Haratin, a group of mainly transition to adulthood. Field studies in Jordan, black sharecroppers, in Morocco. 63 These are Pakistan, Thailand and Viet Nam have found welcome changes, which can disrupt traditional, that migration was a means of enhancing a fam- caste-like forms of hereditary inequality based ily’s social status in the local community. It is 79
  • 90. 4 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Box 4.4 Mobility and the development prospects of small states As noted in chapter 2, it is striking that the countries with the high- (2003) was that of developing a plan for 1,000 workers to go abroad an- est rates of emigration are small states. These rates often coincide nually. However, others (Djibouti, Gambia, Guyana and Maldives) refer with underdevelopment. For poorer small states, the disadvantages to emigration only as a problem. Some see negative aspects, such of being small include over-dependence on a single commodity or as exposure to downturns in remittances (Cape Verde) and increased sector and vulnerability to exogenous shocks. Small countries cannot inequality (Bhutan). Dominica’s PRS saw emigration both as a cause easily take advantage of economies of scale in economic activity and of poverty and as contributing to poverty reduction. in the provision of public goods, and often face high production costs Small states can make migration a strategic element of develop- and consumer prices. In the case of small island states, remoteness ment efforts in several ways, some of which involve regional agree- is an additional factor, raising transport costs and times and making ments. Some countries focus on temporary employment abroad. it difficult to compete in external markets. All these factors encour- Others emphasize the creation of skills, sometimes in concert with age out-migration. neighbours. Mauritius has actively encouraged temporary employ- The financial benefits associated with migration are relatively ment abroad as a way of acquiring skills and capital that migrants large for small states. In 2007, remittances averaged US$233 per cap- can use to set up their own business on return. Supported by do- ita, compared to a developing country average of US$52. The annual nors, the government has established a programme that provides highest flows relative to GDP are found in the Caribbean, with remit- technical and financial support to returning migrants. The Lesotho tances accounting for 8 percent of GDP. However, most small states Development Vision 2020 focuses on generating jobs at home by are not among the countries with the highest GDP shares of remit- attracting foreign direct investments, while recognizing the role of tances, so they are not especially exposed to shocks from this source. work abroad, especially in neighbouring South Africa. Its PRS sets At the same time, the benefits of migration for small states go well out reform measures that include automation and decentralization beyond the monetary value of remittances. Moving opens up oppor- of immigration services, establishment of a one-stop shop for ef- tunities for labour linkages, which can enhance integration with eco- ficient processing of immigration and work permits, and anti-cor- nomic hubs. Temporary labour migration can be a way of balancing ruption measures in the Department of Immigration. Development the economic needs of both the origin and destination sides, of pro- strategies can take broader measures to deal with the challenges of viding opportunities for low-skilled workers and of enabling broader remoteness. For example, in the South Pacific, regional universities benefits at home through the repatriation of skills and business ideas. and vocational training have facilitated mobility, and several states To the extent that smallness overlaps with fragility and, in some coun- have entered into migration agreements with their neighbours. tries, instability, migration can be a safety valve to mitigate the risk Emigrants from small states have similar profiles to migrants gen- of conflict, as well as a diversification strategy over the longer term. erally, in that they tend to have more skills and resources than people Some small states have integrated emigration into their develop- who stay. In Mauritius, for example, the total emigration rate is 12.5 ment strategies, mainly to meet the challenge of job creation. Our com- percent, but about 49 percent for graduates. Overall, however, there missioned review of PRSs showed that many small states (Bhutan, is no significant difference in the net supply of skills, measured by the Cape Verde, Dominica, Guinea-Bissau, Sao Tome and Principe, and number of doctors per 10,000 population, between small and large Timor-Leste) mention positive elements of international migration in states. In terms of simple averages, the number of doctors is actually terms of impact on development and/or poverty reduction. Among higher for small states, at 23 per 10,000 compared to 20 per 10,000 the goals in Timor-Leste’s Poverty Reduction Strategy Paper (PRSP) on average for all countries. Source: Luthria (2009), Winters and Martin (2004), Black and Sward (2009), Seewooruthun (2008), Government of Lesotho (2004), Winters, Walmsley, Wang, and Grynberg (2003), Amin and Mattoo (2005), Koettl (2006) and Pritchett (2006). thus not surprising that the probability of mi- out-migration, such as the Philippines, as well as gration increases for those with links to people in West and Southern Africa. A study in Nigeria already abroad. found that two out five undergraduate students Sometimes a ‘culture of migration’ emerges, were more interested in leaving Nigeria as a way in which international migration is associated of gaining social status than in seeking gainful with personal, social and material success while employment at home.69 This can also be seen staying home smacks of failure.68 As the social with respect to internal migration: a recent study network grows, the culture is further engrained from Ethiopia suggests that shifting preferences and migration becomes the norm, particularly and aspirations as a result of education could lead among the young and able. This has been ob- people to migrate out of rural areas, irrespective served in cases where there has been large-scale of the earning potential that migration may 80
  • 91. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 4 provide.70 The culture can acquire its own self- goods such as local church or soccer field resto- Evidence that perpetuating momentum, as illustrated by the rations. Sometimes these are co-financed—the emigrants have spurred Irish, who continued to emigrate at the height of most famous example being Mexico’s Tres Por the improvement of the Celtic Tiger boom. Uno programme, which aims to increase col- political institutions in In West Africa, migration is often not merely lective remittances by assuring migrant asso- their home countries is a vehicle for economic mobility but is also consid- ciations that, for every peso they invest in local accumulating ered a process through which a boy attains matu- development projects, the federal, municipal and rity.71 For some groups in Mali, Mauritania and local government will put in three. The amount Senegal, migration is a rite of passage: it is through transferred as collective remittances remains the knowledge and experience acquired from only a fraction of that sent back individually to travel that young adolescent males become men.72 families, so the potential development impact In the Soninke village of Kounda in Mali mobil- of such programmes should not be overstated.79 ity distinguishes males and females.73 Masculinity For example, it has been estimated that, since involves the freedom to move, whereas women 1990, Filipinos in the United States have do- in the village are to a large extent fixed inside the nated US$44 million in financial and material household. Men who do not migrate and remain assistance to charitable organizations in the economically dependent on their kin are consid- Philippines, an amount equivalent to only 0.04 ered to be immature youngsters and women refer percent of GDP in 2007.80 to them with a derogatory term, tenes, which Mobility can affect social and political life in means ‘being stuck like glue’. In Mali, the collo- countries of origin in a broader sense. Migrants quial French term used to describe migration is and their descendants may return and become aller en aventure, literally, to go on adventure. For directly involved in civic and political activities. the Soninke, being ‘on adventure’ implies being Alternatively, business investments, frequent re- ‘on the path to adulthood’. turn visits and/or collective initiatives can affect The effect of migration on income distribution patterns of participation by others at home. For and social inequality is primarily a function of se- example, in Lebanon, new political forces were lection—that is, who moves (see chapter 2).74 In formed, particularly after the 1989 Ta’ef Accord, general, money flows associated with international as returning migrants used the wealth earned migration tend to go to the better off, whereas, at abroad to engage in politics.81 least in the longer term, remittances from internal Evidence that emigrants have spurred the im- migrants tend to be more equalizing.75 This type of provement of political institutions in their home pattern has been found for Mexico and Thailand, countries is accumulating. Democratic reform has for example.76 Our commissioned research on been found to progress more rapidly in developing China also found that inequality initially rose countries that have sent more students to universi- with internal remittances, then fell.77 ties in democratic countries.82 Knowledge and ex- If it is the better off who tend to migrate, then pectations brought home by a group of Moroccans an appropriate response is to ensure access to returning from France have been found to shape basic services and opportunities at home as well basic infrastructure investments by the govern- as to facilitate the mobility of the poor. As we ment in their home region.83 However, if emi- argue in chapter 5, poor people should not have gration serves simply as a safety valve, releasing to move in order to be able to send their children political pressure, the incentives of the established to decent schools: they should have options at political elite to reform are diminished.84 home, alongside the possibility of moving. Just as migrants enrich the social fabric Collective remittances sent through home- of their adopted homes, so too they can act as town associations and other community groups agents of political and social change if they re- have arisen in recent decades.78 These usually turn with new values, expectations and ideas take the form of basic infrastructure projects, shaped by their experiences abroad. Sometimes such as the construction of roads and bridges, this has taken the form of supporting civil wars, the installation of drinking water and drainage as in the case of Sri Lanka’s diaspora, but in systems, the sinking of wells, the bringing of most cases engagement is more constructive.85 electricity and telephone lines, and other public Contemporary high-profile examples include 81
  • 92. 4 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Box 4.5 Mobility and human development: some developing country perspectives Several recent National Human Development Reports (NHDRs), in- overall human development impacts of migration in Mexico are cluding those of Albania, El Salvador and Mexico, have focused on complex and conditional on the profile and resources of different the development implications of mobility. In other countries NHDRs groups. For example, while migration tends to reduce education have considered how mobility influences selected aspects of de- inequality, especially for girls, it can also discourage investment in velopment, such as the role of civil society (Egypt), rural develop- higher education in communities where most migrants traditionally ment (Uganda), economic growth (Moldova), social cohesion (Côte go abroad for low-skilled jobs. d’Ivoire) and inequality (China). Different insights come from El Salvador, where emigrants repre- Mexico’s NHDR identifies inequality as the most robust deter- sent 14 percent of the population and the impact of migration is more minant of migratory flows, and movement as a factor that modi- visible at the macro level. The recent acceleration of migration is seen fies the availability of opportunities to others, including stayers. to have contributed to the country’s transition to a service economy, Drawing on the National Employment Survey, the average Mexican which has relied heavily on remittances and a mosaic of small busi- migrant is found to have slightly above-average schooling and in- nesses specialized in delivering goods and services to migrants and termediate income levels but comes from a marginalized munici- their families, including nostalgia products and communications. The pality, suggesting an initial set of capabilities coupled with lack of report suggests that migration allows some relatively poor people a opportunities as major driving factors. The report finds that the degree of upward mobility through their links to the global economy. Source: UNDP (2000; 2004a; 2005a,b; 2006a; 2007c,e; 2008c). Ellen Johnson-Sirleaf, President of Liberia and from recent National Human Development Africa’s first female head of state, and Joaquim Reports. The highlights are summarized in box 4.5. Chissano, former President of Mozambique and To gain insights into the link between na- now a respected elder statesman. Recognizing tional development strategies and migration the potential benefits of diaspora engagement, in a larger sample of countries, we commis- some governments have begun to actively reach sioned a study to review the role of migration out.86 For example, Morocco and Turkey have in Poverty Reduction Strategies (PRSs). These extended political and economic rights to emi- strategies are statements of development objec- grants and allowed dual citizenship.87 However, tives and policy, prepared by poorer countries whether these policies of engagement benefit whose views are often neglected in migration non-migrants or simply subsidize an elite group debates. PRSs are of interest since they also in- outside the country remains an open question. volve contributions from, or partnerships with, By improving its investment climate (presently civil society actors, are intended to be based on ranked first in Africa by the World Bank’s Doing quantitative and participatory assessments of Business Index), Mauritius has also attracted mi- poverty, and provide a sense of government pri- grants back; similar patterns have been seen in orities. 88 They are also important important be- India and Turkey, among other countries. cause international partners have committed to aligning their assistance to these national strate- 4.1.4 Mobility and national gies, given the importance of country ownership development strategies in development. To date, national development and poverty re- To date, Bangladesh’s PRS has perhaps the duction strategies in developing countries have most comprehensive treatment of migration and tended not to recognize the potential of mobil- development linkages. The most recent PRSs for ity, nor integrated its dynamics into planning Albania, the Kyrgyzstan and Sri Lanka also re- and monitoring. This is in part due to the range flect a major focus on migration-related issues. of other pressing priorities facing these coun- Many African countries acknowledge the role of tries, from improving systems of service deliv- remittances, the advantages of return and circu- ery, through building basic infrastructure, to lar migration of skilled expatriates and the value promoting broad-based growth. of knowledge transfer from such people. Several Country-level perspectives on the links be- strategies intend to attract development invest- tween mobility and development can be gleaned ments from wealthy members of the diaspora. 82
  • 93. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 4 Earlier analysis of the treatment of interna- 4.2 Destination place effects tional migration in PRSs was based in part on the Debates about migration often dwell on the eco- number of mentions of the word ‘migration’.89 nomic and social impacts on rich destination While simple, this indicator is not very meaning- countries. This report has deliberately sought to ful. It is nonetheless striking that there is no sig- redress this imbalance, by beginning with the nificant correlation in PRSs between the number migrants and their families, then focusing on of references to migration and various measures of the places they came from. However, that is not its possible importance for national development, to say that the impacts on people in destination such as share of the population living abroad, level communities are unimportant. of remittances and rate of urbanization.90 In many developed countries, the percentage PRSs have laid out a wide range of migra- of migrants in the total population has risen rap- tion-related policy initiatives, although these idly over the past 50 years. It is now estimated to are often not explicitly based on prior analy- be in double figures in more than a dozen OECD sis. In many cases the state of knowledge about countries.91 As noted in chapter 2 and shown in the relationship between the proposed initia- detail in Statistical Table A, the highest shares are tive and its expected development impact is found in Oceania (16 percent)—which includes weak, underlining the importance of better Australia and New Zealand, North America data and analysis. (13 percent) and Europe (8 percent). The shares In general, PRSs appear to recognize the range between only 1 and 2 percent in the three complexity of international migration, acknowl- major developing regions of Africa, Asia, and edging both its advantages—opportunities for Latin America and the Caribbean. The highest development and poverty reduction—and its country shares are recorded in the GCC states possible negative effects. Some tend to stress and in South-East Asia, including 63 percent in the positive—for example the most recent PRSs Qatar, 56 percent in the United Arab Emirates, of Ethiopia, Nepal, Senegal and Uzbekistan 47 percent in Kuwait and 40 percent in Hong frame emigration as an opportunity, without Kong (China). The real and perceived impacts of mentioning possible downsides. Most recent immigration are critical, not least because these strategies emphasize the role of remittances, perceptions shape the political climate in which including those of Bangladesh, Democratic policy reforms are debated and determined. Republic of the Congo, Ghana, the Lao People’s We begin this section by reviewing the eco- Democratic Republic, Liberia, Pakistan, Timor- nomic impacts of immigration as a whole, then Leste and Uzbekistan. focus more narrowly on the labour market and Several strategies articulate policies towards migration. We can distinguish between poli- cies that are broadly ‘proactive/facilitative’ and Table 4.1 PRSs recognize the multiple impacts of migration those focused on ‘regulation/control’ (table 4.1). Policy measures aimed at international migration in PRSs, Combating trafficking, preventing irregular 2000–2008 migration and modernizing and strengthen- No. of No. of No. of ing immigration and customs services feature Proactive/facilitative countries Proactive/facilitative countries Regulation/control countries frequently. It is striking how some of these Export labour 10 Facilitate remittances 9 Combat trafficking 19 policies echo those promoted by rich country Encourage female migration 1 Encourage legal remittance channels 3 Modernise customs 18 governments. Promote student mobility 3 Engage diasporas 17 Strengthen border control 17 To sum up, while the PRS framework gen- Sign bilateral agreements 9 Promote investment by diasporas 8 Combat illegal migration 12 erally has not been geared towards addressing Improve labour conditions abroad 6 Import skills 4 Promote refugee return 10 Pre-departure training 6 Participate in regional Tackle the ‘brain drain’ 9 migration policy per se, it could provide a useful cooperation programmes 8 tool for integrating migration and development Develop consular services 3 Promote more research/monitoring 8 Support return 7 issues. Fitting this dimension into an overall Regulate recruitment industry 2 Build institutional capacity 5 Sign readmission agreements 2 Facilitate portability of pensions 2 Combat HIV/AIDS amongst migrants 7 national strategy for development will require Promote refugee integration 7 Re-integrate trafficking victims 5 investments in data and analysis and in broad stakeholder consultation. These challenges are Source: Adapted from Black and Sward (2009). Note: 84 PRSs reviewed. discussed further in chapter 5. 83
  • 94. 4 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Migrants can bring fiscal impacts. For each of these types of impact full adjustment of per capita investment levels is broader economic there are important distributional issues—while plausible even in the short run. benefits, including there are overall gains, these are not evenly At the individual country level, at least in higher rates of distributed. the OECD countries, similar results have been innovation found—that is, increased migration has neu- 4.2.1 Aggregate economic impacts tral or marginally positive effects on per capita The impact of migration on aggregate growth income. For example, simulations following rates of destination countries has been much the European Union accessions of 2004 sug- discussed, but robust measurement is difficult. gest that output levels in the United Kingdom The data requirements and methodological and Ireland, which allowed large-scale inflows complexities, including the need to disentangle from the new member states of Eastern Europe, direct and indirect effects and work out their would be 0.5–1.5 percent higher after about a timing, all present challenges (see box 1.1). decade.94 In countries where migrants account Economic theory predicts that there should for a much higher share of the population and be significant aggregate gains from movement, labour force—for example in the GCC states— both to movers and to destination countries. the aggregate and sectoral contributions to the This is because migration, like international economy can be expected to be larger. However, trade, allows people to specialize and take ad- detailed empirical analysis is unfortunately not vantage of their relative strengths. The bulk of available. the gains accrue to the individuals who move, Migrants can bring broader economic ben- but some part goes to residents in the place of efits, including higher rates of innovation. destination as well as to those in the place of ori- Productivity gains in a number of destination gin via financial and other flows. In background places have been traced to the contributions of research commissioned for this report, estimates foreign students and scientists to the knowledge using a general equilibrium model of the world base. Data from the United States show that be- economy suggested that destination countries tween 1950 and 2000, skilled migrants boosted would capture about one-fifth of the gains from innovation: a 1.3 percent increase in the share a 5 percent increase in the number of migrants of migrant university graduates increased the in developed countries, amounting to US$190 number of patents issued per capita by a massive billion dollars.92 15 percent, with marked contributions from sci- To complement our review of the country- ence and engineering graduates and without any level studies, we commissioned research to adverse effects on the innovative activity of local construct a new dataset on migration flows people.95 and stocks, including consistent annual data Countries explicitly compete for talent at the on nature of employment, hours worked, capi- global level and the share of graduates among mi- tal accumulation and changes in immigration grants varies accordingly.96 The United States, in laws for 14 OECD destination countries and particular, has been able to attract migrant tal- 74 origin countries for each year over the pe- ent through the quality of its universities and riod 1980–2005.93 Our research showed that research infrastructure and its favourable patent- immigration increases employment, with no ing rules.97 In Ireland and the United Kingdom evidence of crowding out of locals, and that in- the share of migrants with tertiary education vestment also responds vigorously. These results exceeds 30 percent, while in Austria, Italy and imply that population growth due to migration Poland it is below 15 percent.98 Countries offer- increases real GDP per capita in the short run, ing more flexible entry regimes and more prom- one-for-one (meaning that a 1 percent increase ising long-term opportunities have done better in population due to migration increases GDP in attracting skilled people, whereas restrictions by 1 percent). This finding is reasonable, since in on duration of stay, visa conditions and career most instances annual migration flows are only development, as in Germany for example, limit a fraction of a percentage point of the labour uptake. This has led to discussions about a blue force of the receiving country. Moreover, these card or European Union-wide employment flows are largely predictable, implying that the permit—an idea that has received preliminary 84
  • 95. HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development 4 backing from the European Parliament and ap- lively academic debates on the subject, notably Migrants can facilitate proval by the European Council.99 Singapore in the United States. Yet it is striking that most higher labour force and Hong Kong (China), have explicit policies empirical studies in the OECD draw similar participation among to welcome foreign high-skilled professionals. conclusions, namely that the aggregate effect of locally born females These policies range from allowing immigrants immigration on the wages of local workers may to bring their families, through facilitating per- be positive or negative but is fairly small in the manent residence after defined waiting periods short and long run.103 In Europe, both multi- and (two years for Singapore, seven for Hong Kong single-country studies find little or no impact of (China)), to the option of naturalization.100 migration on the average wages of local people.104 Programmes to attract skilled labour can At the same time it must be recognized that be developed using a general points-based ap- wage responses to immigration are unlikely to proach, linked to labour market tests and/or be distributed evenly across all workers and will employer requirements (chapter 2). A centralized be most pronounced where locally born work- ‘manpower’ planning approach can be difficult ers compete with immigrants. The debates have to implement, especially in the face of structural clarified that it is not just the total number of change and economic shocks. Points-based migrants that matter but their skill mix as well. schemes, which have the virtue of simplicity, The kinds of skill that migrants bring affect the have been used by destination governments to fa- wages and employment opportunities of differ- vour high-skilled migrants or to attract workers ent segments of the local population, sometimes for occupations in short supply on the national in subtle ways. If the skills of migrant workers labour market, as in Australia’s General Skilled complement those of locally born workers, then Migration programme. both groups will benefit.105 If the skills match Migration can stimulate local employment exactly, then competition will be heightened, and businesses, but such effects are likely to be creating the possibility that locally born workers context-specific. Migrants also affect the level will lose out. However, this is not a foregone con- and composition of consumer demand, for ex- clusion: often the results are mixed, with some ample in favour of nostalgia goods, as well as individuals in both groups gaining while others locally available goods and services that are lose. Assessing these effects is problematic, be- close to homes and work-places. Our commis- cause measuring the degree to which different sioned study of such effects in California found groups’ skills complement or substitute for one evidence suggesting that an influx of immigrants another is difficult, particularly across interna- over the decade to 2000 into specific areas (se- tional borders.106 lected to capture the potential pool of custom- One striking example of complementarity is ers for different firms) was positively correlated how migrants can facilitate higher labour force with higher employment growth in some sectors, participation among locally born females.107 The especially in education services. The impact on availability of low-cost child care can free up the composition of demand was mixed: a higher young mothers, enabling them to go out and find share of migrants was associated with fewer a job. There is consensus in the literature that small firms and stand-alone retail stores, but low-skilled migrant labour generally comple- more large-scale discount retailers. At the same ments local labour in Europe.108 This may arise time, consistent with expectations, the study in part because migrants are more mobile than found that increased immigration was associated locally born workers—as in Italy, for example.109 with increased ethnic diversity of restaurants.101 More importantly, migrants are often willing to accept work that locals are no longer prepared to 4.2.2 Labour market impacts undertake, such as child care, care of the elderly There is controversy around the effects of mi- (much in demand in aging societies), domestic gration on employment and wages in the des- work, and restaurant, hotel and other hospitality tination country, especially for those with low industry work. levels of formal education. Public opinion polls As noted, the small average effect on pay may show that there is significant concern that im- mask considerable variation across types of local migration lowers wages.102 There have also been workers. There is a vast empirical literature on 85
  • 96. 4 HUMAN DEVELOPMENT REPORT 2009 Overcoming barriers: Human mobility and development Legal and institutional the effect of immigration on the distribution the enforcement of regulations is weak. Even in factors—both their of wages in developed countries. In the United countries with well-regulated labour markets, design and their States, estimates of the effect on the wages of un- workers with irregular status often tend to fall enforcement—matter skilled workers range from –9 to +0.6 percent.110 ‘under the radar’—the drowning of Chinese Locals with low levels of formal schooling may cockle gatherers in Morecambe Bay in the still have advantages over migrants due not only United Kingdom was a notorious case of lack to language but also to knowledge of local insti- of enforcement of health and safety standards. tutions, networks and technology, which enables Recent British research found that more general them to specialize in complementary and better- structural trends, particularly the increasing use paid tasks.111 of agency (temporary) labour contracts, which The imperfect substitutability of migrant and are associated with fewer rights for workers, are local labour is consistent with recent evidence significant factors shaping the pay and working suggesting that the workers affected most by conditions of migrant workers. There is wide- the entry of new migrants are earlier migrants. spread evidence of payment below the legal mini- They feel the brunt of any labour market adjust- mum wage, especially for younger migrants.114 ment, since newcomers primarily compete with Among emerging and developing econo- them. In the United Kingdom, for example, mies, empirical evidence on the labour market heightened competition among migrants in the impacts of immigration is sparse. A recent study early 2000s may have increased the difference of Thailand, which investigated whether places between the wages of locals and migrants by up with higher concentrations of migrants had to 6 percent.112 lower wages, found that a 10