69530 M iddle E a s t a n d No r t h A f r ic a Re g io n v1 Labor Migration from North Africa Development Impact, Challenges, and Policy Options |  VOLUME 1: Main report  | Middle East and North Africa Region Labor Migration from North Africa Development Impact, Challenges, and Policy Options A project implemented A project funded by the by the World Bank European Union This report as well as the background research underlying the analysis and conclusions of this report constitute part of an EC- Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies, a program of migration-related research and activities to identify and support the implementation of projects, policies, regional arrangements, and institutional reforms that will maximize the benefits of international migration flows and reduce their costs. The views herein are those of the authors and should not be attributed to the World Bank, the European Commission, or the institutions and countries they represent. © 2010 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org E-mail: feedback@worldbank.org All rights reserved. This volume is a product of the Chief Economist’s Office of the Middle East and North Africa Region of the World Bank. The �ndings, interpretations, and conclusions expressed herein are those of the author(s) and do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundar- ies, colors, denominations, and other information shown on any map in this work do not imply any judgement on the part of the World Bank concerning the legal status of any territory or the endorse- ment or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The World Bank encourages dissemination of its work and will normally grant permission promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750-4470, www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Of�ce of the Publisher, World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522- 2422, e-mail: pubrights@worldbank.org. Cover photos: © World Bank/Yuri Kozyrev, Bill Lyons and Alan Gignoux. Table of Contents Acknowledgements.............................................................................................................................ix Abbreviations and Glossary. ..........................................................................................................xiii Overview............................................................................................................................................... xv Vue d’Ensemble du Programme de Recherche..........................................................................xxiii Chapter 1. A Heightened Urgency for Understanding International Migration...................1 Three Forces Heightening the Interest in North African Migration.................................3 A global economic downturn with strong effects on migration....................................4 Demographic forces increasing imperative for more well-managed migration........5 Recognition that the costs of not understanding migration are high..........................6 Chapter 2. Trends and Characteristics about North African Migration.................................9 Data Limitations for Understanding North African Migration..........................................9 North African Migrant Destinations and Nature of Migration. ........................................10 Remittances.......................................................................................................................12 The Skills Composition of North African Migrants..........................................................13 Recent Changes in North African Migration....................................................................16 The motivation behind migration decisions has evolved over time..........................16 Migration has become more feminized........................................................................16 North Africa has increasingly become a destination region for migrants (both permanent and transit)..............................................................................17 Managed and official migration has become increasingly undocumented..............18 Much about North African migration is unknown.....................................................19 Chapter 3. The Development Impact of North African Migration.........................................21 A Framework for Understanding the Development Impact of Migration. ......................22 The Evidence of the Development Impact of Migration.................................................25 The Impact of Migration on Growth.................................................................................26 Empirical estimation of impact of remittances on growth.........................................27 Measuring Migration’s Impact at the Household Level. ...................................................28 Data sources for analyzing remittance-receiving households in North Africa. ........29 Impacts on Poverty and Inequality...................................................................................30 A profile of migrant-sending, remittance-receiving households in North Africa.....31 The empirical results on poverty alleviation..............................................................35 Impacts on Household Labor Market Decisions..............................................................37 Some descriptive statistics of labor supply and employment among migrant and non-migrant families. .....................................................................37 iii   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options The empirical results on labor supply decisions. ........................................................41 Impacts on Decisions Affecting Children.........................................................................43 The empirical results on investments in children......................................................45 The Potential Effects of High-skilled Migration from North Africa................................46 Conditions that suggest a lower true cost of high-skilled migration. .........................50 Chapter 4. Further Contributions of Migrants to Host-Country Development, While Abroad and At Return......................................................................................57 The Role (or Potential Role) of North African Diaspora.................................................58 Description of North Africa’s diaspora.........................................................................59 Qualitative evidence of North African diaspora achievements.......................................60 Initiatives by national governments............................................................................64 Strengthening ties with North Africa’s diaspora.........................................................66 Return Migration...............................................................................................................68 Development impact of return migration....................................................................69 Information from returned migrants’ studies of North Africa...................................70 Return migration, reintegration, and entrepreneurship. ..........................................71 Determinants of return migration occupation mobility. ....................................71 Characteristics and Determinants of Entrepreneurship....................................72 Policy alternatives for North Africa and Europe. ........................................................75 Chapter 5. European Migration Policy, Labor Market Conditions, and Migration Outcomes.....................................................................................................77 Interactions Between Migration Policy, Migration Flows and Development Outcomes............................................................................................78 The architecture of migration. ......................................................................................78 Rich-country migration policy features. ......................................................................78 The Migration Architecture and Labor Market Conditions in Spain..............................80 The nexus between policy, markets, and migration outcomes. ..................................81 The impact of migration on economic growth............................................................82 The Migration Architecture of the Netherlands..............................................................86 From temporary to permanent residents....................................................................86 Doors are opened, and difficult to close.......................................................................87 How migration has adapted to changing policies. ......................................................88 What Spain and the Netherlands Tell Us about Migration Policy...................................90 Chapter 6. Demographic Forces and the Implications for North Africa. .............................93 Demographic Changes and the Opportunities for Arbitrage..........................................93 While demographics fit, the skills do not. .....................................................................95 Scenarios of How Rising Migration may Impact North Africa.........................................98 The results......................................................................................................................99 Effects on taxes.......................................................................................................99 Effects on GNI per capita..................................................................................... 100 Policy implications for North Africa.......................................................................... 102 Chapter 7. Further Agenda for Improving the Development Impact of North African Migration............................................................................................105 The Protection of North Africa’s Workers Abroad, and its Impacts on .......................106 Migration Flows Social protection plans to manage flows and protect workers................................. 108 iv Table of Contents North Africa’s Migration Information and Capacity Gap...............................................111 Data: Statistical capacity and instruments............................................................... 111 Turning data into information and knowledge........................................................ 114 Improving North Africa’s Environment for Migration Contributions . .........................115 to Development References. ........................................................................................................................................117 List of Boxes Box 2.1: Notes on the DLM database on migration by education level............................................14 Box 3.1: A note on measuring migration/remittance impact. ............................................................31 Box 4.1: Savoir et Développement: a European “Silicon Valley� Diaspora?.....................................63 Box 4.2: Unlikely reform initiated and sustained by diaspora members: electricity in rural area of Morocco............................................................................................................64 Box 4.3: Diaspora-led philanthropy transforms the way traditional charities operate in the Egyptian Coptic Church.............................................................................................65 Box 5.1: Migration and Economic Growth in Spain...........................................................................83 Box 6.1: CEDEFOP Definition of Occupational Groups....................................................................96 Box 7.1: Exportability and Portability explained.............................................................................106 Box 7.2: Modules on emigration to add to existing surveys............................................................114 List of Figures Figure 1.1: Middle East and North Africa Region Emigration 2005...................................................2 Figure 1.2: Rise in unemployment in North Africa from 1990 to 2000..............................................3 Figure 2.1: Estimates of the Stock of Egyptian migrants abroad, 1970–2006.................................12 Figure 2.2: Skills composition of North African migrants, 2000. ......................................................15 Figure 2.3: Growth in migrant population in OECD countries, 1990–000 (Total migrant stock, thousands)....................................................................................17 Figure 2.4: Moroccan immigration in Spain, 2000–2007 (Total migrant stock, thousands)...........18 Figure 3.1: Migration’s Development Impact Prism..........................................................................23 Figure 3.2: Use of Bank Credit by North African Firms. ...................................................................29 Figure 3.3: Distribution of Egyptian Migrants by Year of Departure and Country of Destination...................................................................................................................32 Figure 3.4: Unemployment in Morocco (2001) and Egypt (2006) .................................................38 Figure 3.5: Labor force participation rates in Morocco and Egypt by Gender, 2000 (Morocco) and 2006 (Egypt)..........................................................................................39 Figure 3.6: Male and female employment by sector, Morocco and Egypt.......................................40 Figure 3.7: Employment of women in Egypt, by household remittance status..............................41 Figure 3.8: Conceptual framework for gendered preferences over remittances ...........................44 Figure 3.9: Impact of Girls and Boys’ Demographic Share on Education . .....................................46 Figure 3.10: Probability that a foreign-educated bachelor degree holder enters a skilled occupation (2000 census)...............................................................................................49 Figure 3.11: Change in income per capita: worst-case and best-case scenarios...............................50 Figure 3.12: MENA laborers with secondary and tertiary education, as share of labor force..........52 and unemploye Figure 3.13: Unemployment and migration from North Africa, 2000................................................52 Figure 3.14: Percent of MENA eighth graders reaching high benchmarks in mathematics and science, 2007.............................................................................................................55 Figure 4.1: Stock of North African migrants in select OECD economies........................................60 v   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Figure 4.2: Stock of North African Tertiary-Educated Migrants in Select OECD economies........60 Figure 5.1: Difference between the proportion of “irregular residents� and the proportion of “irregular workers� in Spain by nationality, 2007....................................82 Figure 5.2: Evolution of the Working Age Population in Spain........................................................84 Figure 5.3: Immigration in the Netherlands, 1980–2007..................................................................89 Figure 5.4: Immigration in the Netherlands, 1960–1980..................................................................89 Figure 5.5: Share households with low income.................................................................................90 Figure 5.6: Educational Catch-Up Among Migrant Communities in the Netherlands....................91 Figure 6.1: Total dependency ratio (TDR) by region (1950–2050).................................................94 Figure 6.2: Aged dependency ratio (ADR) in North African and other MENA countries (1950–2050).....................................................................................................................94 Figure 6.3: Aged dependency ratio (ADR) by consolidated region.................................................95 Figure 6.4: Expected Employment Changes in Europe 2006–2015 by Occupational Groups.......97 Figure 6.5: Increase in Occupational Demand in Europe . ..............................................................97 Figure 6.6: Educational attainment of adult population in North Africa, 1960–2000.....................98 Figure 6.7: Tax rate in EU15 — “Skilled� vs “Unskilled� (in percent). ..........................................100 Figure 6.8: Tax rate in MENA — “Skilled� vs “Unskilled� (in percent).........................................100 Figure 6.9: GDP per capita in EU (percent deviation from the baseline) — “Skilled� vs “Unskilled�.................................................................................................................101 Figure 6.10: GDP per capita in MENA (percent deviation from the baseline) — “Skilled� vs “Unskilled�: with and without diaspora effect.........................................................102 Figure 6.11: GNI per cap. in Europe — Brain drain variants (in % dev. from baseline)................103 Figure 6.12: GNI per cap. in MENA — Brain drain variants (in % dev. from baseline). .................103 Figure 7.1: Social services for migrant. ............................................................................................107 Figure 7.2: Social protection regimes for migrants.........................................................................108 List of Tables Table 1.1: Projected workers’ remittances, 2008–2011 (credit, $U.S., billions) ............................4 Table 1.2: Remittances per capita and as a share of GDP, 2007.......................................................5 Table 1.3: Expected changes in world population by age group, 2010–2040..................................6 Table 2.1: Emigration from North African countries in 2000.........................................................11 Table 2.2: Location of North African emigrants in 2000.................................................................11 Table 2.3: Inward Remittance Flows, Per Capita Remittances, National ODA and FDI, 2007.....13 Table 2.4: Skilled emigration rates of North African countries in 2000.........................................15 Table 3.1: Channels of Development Impact over the Migration Lifecycle. ...................................24 Table 3.2: National Accounting: Impacts of Remittance Decline on GDP.....................................26 Table 3.3: Estimation results: Impact of remittances on per capita growth in North Africa........28 Table 3.4: Characteristics of remittance-receiving households in Morocco and Egypt................33 Table 3.5: Distribution of remittances in Morocco, “adjusted� quintiles. .......................................34 Table 3.6: Impact of Remittances on Poverty North Africa and Comparator Countries...............35 Table 3.7: Regional differences in the impact of remittances in Morocco.....................................36 Table 3.8: Employment status of Moroccan labor force age population, by education. ................39 Table 3.9: Employment patterns of Moroccan non-migrants, by location.....................................40 Table 3.10: Impacts of remittances on informality in Morocco: Multinomial Logit analysis...........42 Table 3.11: The association between receiving remittances/having a migrant and the probability of participating in labor markets and self-employment..............................43 Table 3.12: Public education spending is comparatively high in MENA .........................................48 Table 3.14: Conditions that influence the impact of skilled emigration on sending countries.......51 Table 3.15: Employment Status of Returned Migrants from the Maghreb, prior to Migration ......53 Table 3.16: GDP per laborer growth and growth of accumulation and productivity vi Table of Contents by region, 1970–2005 (% per year). ................................................................................54 Table 3.17: Skills shortages and the business environment (% of firms listing the respective area as the major obstacle to business). .......................................................56 Table 4.1: Share of migrants with tertiary education in the top 7 destination countries of the OECD.....................................................................................................................61 Table 4.2: Active foreign-born tertiary educated migrants in selected OECD countries, selected occupations (2000)...........................................................................................61 Table 4.3: Active foreign-born tertiary educated migrants in United States, selected occupations (2000)..........................................................................................................62 Table 4.4: Institutionalization of diaspora networks as search networks. ......................................67 Table 4.5: Composition of national samples for return migrants. ...................................................71 Table 4.6: Probit model of upward occupational mobility after migration with selection............73 Table 4.7: Probit model of becoming an entrepreneur after migration (marginal effects). ..........74 Table 5.1: The International Migration Architecture Examples.....................................................79 Table 5.2: Legislative development and regularization processes in Spain...................................81 Table 7.1: Data on emigration in statistical instruments. ..............................................................113 Table 7.2: 2008 statistical capacity indicator (out of 100). ...........................................................114 vii Acknowledgements This report represents the integrated findings activities on the functioning and roles of diaspo- of an EC-Funded World Bank Program of Inter- ras and returned migrants, and identification of national Migration from Middle East and North programs that have supported their contribution Africa and Poverty Reduction Strategies, a pro- to economic and social progress of countries of gram of migration-related research and activities origin; and, (v) the activation of a network of to identify and support the implementation of migration practitioners from EU and Maghreb projects, policies, regional arrangements, and countries, and establishment of a platform for institutional reforms that will maximize the ben- debate on findings, needs, and policy options efits of international migration flows and reduce uncovered in the action. their costs. The research program began with the premise that a well-managed international The EC-Funded World Bank Program of migration regime, based on cooperation between MENA International Migration was managed home and host countries, and more consistently by Jennifer Keller, Senior Economist, Middle integrated into development and poverty reduc- East and North Africa Region, Office of the tion strategies, could become a beneficial strat- Chief Economist. Individual components of the egy for sending countries, receiving countries, program were managed by a World Bank team the migrants, and the migrants’ households. consisting of Jennifer Keller, Sara Johansson Through a migration-related research project, Silva (consultant, Middle East and North Africa the program has sought to identify and support Reigion, Poverty and Economic Management the implementation of projects, policies, regional Unit), Edmundo Murrugarra (Senior Economist, arrangements, and institutional reforms that will Poverty Reduction and Economic Manage- maximize the benefits of international migration ment Network), and John Blomquist (Senior flows and reduce their costs. Economist, Middle East and North Africa Region, Human Development Unit). The overall work The EC-Funded World Bank Program of In- program was initiated under the guidance and ternational Migration from Middle East and North supervision of Mustapha K. Nabli while he was Africa and Poverty Reduction Strategies under- Chief Economist of the Middle East and North took research and activities across a spectrum Africa Region, and subsequently Ritva Reinikka, of migration related themes, including: (i) as- Director, Middle East and North Africa Region, sessment of the statistical and technical capacity Poverty and Economic Management Unit. for monitoring and analyzing the nature, causes, and consequences of migration flows; (ii) study The publications and outputs stemming from of the poverty, social, and labor market impact this three year research program, in addition of migration; (iii) identification of good practices to this integrated volume, include the follow- in the area of labor migration management and ing: (1) one technical report on statistical and social protection arrangements for migrant work- technical capacity for the analysis of migration ers and returnees; (iv) development of learning flows (An Assessment of Sources of Data on ix   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options International Migration in the Maghreb and Flore Gubert; Christophe Nordman, 2008); Egypt, Focusing on Population Censuses (9) two country studies on the impact of inter- and Household Surveys, Richard E. Bilsbor- national migration from Maghreb to EU coun- row and Ingrid Ivins; 2009); (2) one technical tries on the well-being of children and women report on the impact of remittances on economic (Remittances, Gender Roles and Children’s growth (Remittances and Growth in North Welfare Outcomes in Morocco, Catalina Her- Africa, Richard Adams, Jennifer Keller, Lili rera, Nora Dudwick and Edmundo Murrugarra, Mottaghi, Marie Alienor van den Bosch, 2009); 2008; Impact of International Migration and (3) two technical reports on the distributional Remittances on Child Schooling and Child impact of remittances on household welfare (Mo- Work: The Case of Egypt, Asmaa Elbadawy and rocco’s migration: the profile and the impact Ragui Assaad, 2009); (10) one report assess- on households, Marcin Sasin, 2008; Interna- ing good practices in the area of skilled labor tional Migration, Remittances and Household migration management ( The International Poverty Status in Egypt, Rania Roushdy, Ragui Migration Architecture: A Survey with Les- Assaad, Ali Rashed, Dalia Hassanein, 2008); sons for the MENA Region, John McHale, 2008); (4) one technical report on the role of migration (11) one report on social protection arrange- in mitigating demographic and labour market ments for migrant workers and returnees (Social imbalances ( MENA-to-EU migration and Protection for Migrant Workers from North labor market imbalances, Frédéric Docquier Africa to Europe, Johannes Koettl, Marie Alienor and Luca Marchiori, 2008); (5) two technical van den Bosch, Matteo Morgandi, 2009); and, reports on the impact of remittances on work (12) one report on diasporas and returned mi- incentives in domestic labor markets (The Effects grants actions (Why is Diaspora Potential so of Remittances on Labor Markets: Evidence Elusive? Towards a New Diaspora Agenda in from Morocco, Joana Silva, 2008; The Impact North African Economies, Yevgeny Kuznetsov of International Migration and Remittances and Matteo Morgandi, 2008). In addition, this on Domestic Labor Markets: Evidence from report benefited from 3 background papers on Egypt, Christine Binzel and Ragui Assaad, 2009); migration from North African countries: (1) Mi- (6) one report on the impact of highly skilled gration, Diaspora et Développement : Le Cas migration flows on Maghreb sending countries de la Tunisie, A. Ben Haj Zekri, 2008 ; (2) Mi- (Skilled Migration from the MENA Region: gration, Diaspora et Développement: Le Cas Trends, Impacts and Policy Responses, Fré- du Maroc, B. Hamdouch, 2008; and, (3) Migra- déric Docquier, Sara Johansson de Silva and Ab- tion, Diaspora et Développement : Le Cas de deslam Marfouk, 2009); (7) three country studies l’Algérie.Nacer Edine Hammouda, 2008. The on the impact of the institutional framework for background papers for this volume are included migration on the profile of migrants, their inte- in the CD mounted on the inside back cover. gration in host countries and the evolution of legal vs. undocumented migration (Changes in This report benefits extensively from the Immigration in the Netherlands: Trends, Poli- earlier background research underlying the World cies and Incentives, Chris de Neubourg, 2008; Bank’s report: Shaping the Future: A Long- La politique migratoire française dictée par Term Perspective of People and Job Mobility les besoins économiques, Mouna Viprey, 2008; for the Middle East and North Africa, a study Spain, A Migration Boom under a Reactive co-financed by the European Union and the Immigration Policy, Jose Ignacio Conde-Ruiz, World Bank. Additionally, the presentation and Clara I. Gonzalez, and Sara Johansson de Silva, integration of the research findings in this report 2008); (8) two technical papers on the living and have been substantially improved by comments working conditions of returned migrant workers and suggestions offered throughout the process. in home Maghreb countries (Return Migrants Among these, the report has particularly benefit- and Small Enterprise Development in the ed from comments from Farrukh Iqbal, Roberta Maghreb and An Empirical Analysis Using Gatti, Auguste Tano Kouame, Tara Vishwanath, Data on Return Migrants in the Maghreb, Leila Zlaoui, Ruslan Yemtsov, Sonia Plaza, Dilip x Acknowledgements Ratha, Caglar Ozden, and Adriana Jaramillo, from Institute, Oxford University; Maarten Vink, from the World Bank; Ibrahim Awad, from the ILO’s In- the University of Maastricht and the University ternational Migration Program; Zafiris Tzannatos, of Lisbon; Mathias Lucke, from the Kiel Institute from the American University of Beirut; Alexan- for the World Economy; Jean Pierre Cassarino, dre Kolev, from the International Training Center; from the European University Institute; and Mario Hein de Haas, from the International Migration Piacentini, from the University of Geneva. xi Abbreviations and Glossary 2SLS Two Stage Least Squares technique ADR Aged Dependency Ratio AMGE Association des Marocains des Grandes Ecoles ATUGE Association Tunisiens des Grands Ecoles AVR Assisted Voluntary Return Programs CAPMAS Egyptian Central Agency for Public Mobilization and Statistics CEDEFOP European Center for the Development of Vocational Training CGE Computable General Equilibrium DLM Docquier, Lowell, and Marfouk (migration dataset) EAP East Asia and the Pacific Region EC European Commission ECA Europe and Central Asia Region ELMS Egypt Labor Market Survey (1998) ELMP Egypt Labor Market Panel Survey (2006, 1998) ENCDM Moroccan National Household Consumption and Expenditure Survey EU European Union EU15 EU members prior to the accession of 10 candidates on May 1, 2004: comprising Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Lux- embourg, the Netherlands, Portugal, Spain, Sweden, and the United Kingdom. EU27 Full European Union: comprising Austria, Belgium, Bulgaria, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, the Netherlands, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden and the United Kingdom. FDI Foreign Direct Investment GCC Gulf Cooperation Council GDP Gross Domestic Product GNP Gross National Product GLS Generalized Least Squares technique xiii   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options HCM Higher Committee for Migration HIC High Income Countries (World Bank definition) ILM International Labor Migration Database (ILO) ILO International Labor Organization IOM International Organization for Migration IV Instrumental Variables technique LAC Latin America and the Caribbean Region LIC Low Income Countries (World Bank definition) LMIC Lower Middle-Income Countries (World Bank definition) MENA Middle East and North Africa MENA4 North African focus countries: Morocco, Tunisia, Algeria and Egypt MENA9: MENA4 + Djibouti, Iran, Lebanon, Libya and Malta MIREM Program of Collective Action to Support the Reintegration of Return Migrants in their Country of Origin OECD: Organization for Economic Cooperation and Development OTE Office of Tunisians Abroad REAGE Réseau des Algériens Diplômés des Grands Ecoles et Universités Françaises SA South Asia Region SSA Sub-Saharan Africa Region TDR Total Dependency Ratio TIMSS Third International Mathematics and Science Study TFP Total Factor Productivity UMIC Upper Middle-Income Countries (World Bank definition) UN United Nations xiv Overview The subjects of this migration research program instability—are palpable evidence of develop- are four countries in North Africa—Morocco, ment deficiencies at home. Tunisia, Algeria, and Egypt. As it has for the broader Middle East and North Africa region, At the same time, more and more, migration labor migration has played a central role in shap- has come to be concomitantly seen as a powerful ing these four countries’ social and economic mechanism for growth and development. With development. While there are larger migrant migration opening the door to financial resources sending regions in the world, on a per capita (in the form of remittances), investment and basis as well as in share of GDP, worker remit- knowledge transfer (by returned migrants and tances to North Africa1 are among the highest diaspora), skills development, and networks in the world. important for business, the potential channels for migration to aid in development are numerous. Despite this economic importance, there has been surprisingly little attention paid to The information on stocks of North African understanding the migration phenomenon in migrants is quite dated. In 2000, North African the North Africa region. This research program countries had an average emigration rate (to hopes to fill only part of a large knowledge gap, OECD and GCC economies combined) of about focusing on some of the development impacts 5.5 percent of the resident population, and of North African migration, the challenges the reaching as high as 9 percent in Morocco, but region faces with regard to labor migration over the destination and the nature of migration differ the near and long term, and policy options to for Maghreb countries and Egypt. The migrants maximize the potential for migration to be a from the Maghreb countries are often unskilled tool for development. At the onset, it must be workers searching for job opportunities that acknowledged that the study of migration from allow them to settle permanently in Europe. In North Africa is significantly handicapped by data contrast, Egyptian migrants are more likely to limitations, as confirmed by nearly all studies be temporary migrants in other Arab countries. and researchers. These difficulties permeate all aspects of the subject, including migrant stocks The skill levels of migrants also differ, with and characteristics, migrant flows, remittances, a preponderance of low-skilled migration from and others,2 and prevents the answers to many the Maghreb countries (at 70 percent) and a of the questions one would like to ask. 1 Because of data limitations, the analysis of North Africa is In the past, migration was only viewed as a limited to the four countries mentioned. Where possible or ‘failure of development,’ reflecting a clear lack required, data for the broader MENA region is alternatively discussed throughout the report. of opportunities at the country of origin. Indeed, 2 World Bank 2009. Shaping the Future: A Long Term Per- many of the factors that increase the push for spective of People and Job Mobility for the Middle East and migration—unemployment, poverty, and political North Africa. xv   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options greater mix of low and high-skilled migration capped by a lack of appropriate data throughout from Egypt (55 percent, versus 30 percent for the region. As a result, only some few parts of the high skill and 15 percent for medium skill). At potential costs and benefits of migration could be the same time, all countries in the region exhibit addressed in this research program. high rates of high-skilled (or more appropriately, higher-educated) migration, ranging between 8 At the broad level, there is interest in know- percent and 19 percent of the domestic high- ing the impact of migration on growth and pro- skilled workforce. ductivity in North Africa, particularly in light of potential declines in migration and remittances North African migration, particularly from with the global financial crisis. The North African the Maghreb, has also become more feminized, region is estimated to have experienced 10.6 largely in the context of family reunification, percent drop in remittances over 2009. However, but also an increase in female labor market at an economy wide level, the migration related participation. This feminization of migrants has impacts on output from the global financial crisis had important impacts on both North Africa and are likely to have been modest: with remittances Europe. In Spain, for example, the availability of averaging only 4.3 percent of GDP and with re- predominantly female domestic helpers has led gional remittances estimated to have declined by to large increases in native female labor market 10.6 percent over 2009, the net impact on GDP participation (and growth). over 2009 was only about 0.5 percentage points (but as high as 1.3 percentage points in Morocco). And, although for obvious reasons there are no reliable data on undocumented migration, However, cross-country time series analysis with the establishment of migration networks and of remittances and growth also reveals the inter- changes in migration policy, the proportions of play between remittances and the financial sector North African undocumented workers are likely in North Africa; particularly, that remittances to have increased. In the 1960s and 1970s, migra- have a stronger effect on growth if the financial tion generally operated through official channels system is shallower. This is evidence that remit- and had a circular character. However, as soon as tances act as a substitute for the financial system. networks had been established in host countries, Although the growth effect is small—stemming new migrants increasingly circumvented the bu- from the small magnitude of remittances in reaucracy of official recruitment and an autono- North African economies—the results confirm mous “unofficial� channel emerged, parallel to the evidence on the weak financial sector develop- official one. More informal migration came with the ment in MENA countries generally. Investment introduction of new destinations (including Spain climate assessments in the region reveal that and Italy), and the introduction of Schengen visas few private sector firms even use bank credit to (increasing the difficulty of labor entry legally). finance investment, instead relying primarily on retained earnings. And outside of Sub-Saharan The nature of migration from North Africa Africa, no region has a higher percentage of has a large impact on the benefits and costs firms that identify access to finance as a major the region can realize from the migration phe- constraint to doing business. As a result, remit- nomenon. Migration brings different benefits tances exert a stronger influence on growth in and costs to the economy along the different MENA (and North Africa) than in countries with migration stages, but the magnitude of those more well developed financial sectors. Also as a costs and benefits are greatly influenced by the result, MENA countries are more vulnerable to characteristics of migrants and the structural sudden changes in remittances. characteristics and incentives in both home and host countries. Because migration, relative to the popula- tion, remains small (in Morocco, an estimated 11 The ability to analyze the impacts of migra- percent of households have a migrant abroad; in tion in a systematic way is significantly handi- Egypt, 3.5 percent), many impacts of migration xvi Overview cannot be seen at the macroeconomic level, but with crises, and simultaneously start or expand will be evident through the decisions and out- income generating activities, thus diversifying comes of the household. In only two countries in their livelihoods. the region, Morocco and Egypt, do data permit the empirical analysis of the impact of remit- Remittances have also allowed for increased tances, migration, or both. That analysis reveals expenditures on health and education for chil- some important findings: dren and reduced child labor, which over the long run are vital for setting the groundwork for higher First, migration and remittances have re- growth. In both Egypt and Morocco, migration duced the level of poverty in North Africa. While and remittances lead to positive improvements remittances do benefit the better off (with 61 per- with regard to investments in children, and in cent of remittances accruing to the top 2 income Egypt, migration and remittances significantly quintles in Egypt, and 82 percent in Morocco), reduce the probability of child work. Remittances they nonetheless benefit the poor. In Morocco, also trigger a more equal intra-household alloca- for example, remittances account for more than tion of education and food between boys and 50 percent of the household budget of the lowest girls, offsetting a gender biased allocation in edu- income quintile. Moreover, remittances decrease cation and food expenses, and triggering gender the probability of being poor by about 7 percent- parity in children’s human capital accumulation age points in Morocco and by about 9 percentage in rural areas where girls are more marginalized.3 points in Egypt. That is to say, the presence of remittances in the household reduced the prob- Another potential effect from migration ability of being poor in Morocco from around 15 relates to its impacts on labor markets, not only percent to around 8 percent, while in Egypt, it by the outflow of laborers, but also by potential reduced the probability from around 20 percent changes in the labor market decisions of those to around 11 percent. left behind (particularly for households which receive remittances). However, the data from The analysis suggests that while migration Egypt and Morocco suggest that the effects are can play a role in poverty alleviation, it is not a modest, at best. In the case of Egypt, in fact, panacea for poverty reduction at the national the results suggest a positive impact from re- level. In both Morocco and Egypt, the poverty mittances on labor supply, although most of the alleviating impact of migration and remittances at increase is the result of females moving into un- the national level is modest. With only 3.5 percent paid family labor. In both countries, remittances of the Egyptian population receiving remittances also increase the probability for self-employment at the time of the survey, the absence of remit- at the expense of wage employment. Given the tances would have raised the overall poverty nature of the self-employment sector (predomi- headcount by only about 0.4 percentage points. nantly low-skill), this suggests that remittances In Morocco, meanwhile, the absence of remit- may be allotted as capital to more informal activi- tances would have raised the poverty headcount ties that provide greater employment flexibility. by less than a percentage point. Thus, the ability for migration to substantially alleviate poverty on The other evidence on impacts of North a national level is limited. Africa migration can only come from descriptive evidence, which is suggestive (but not definitive) However, for the individual household, of associated costs and benefits of migration. In migration is an important element of the house- terms of costs, the largest potential cost to the hold’s coping strategy. Indeed, remittances region comes from the loss of higher-educated reduce the individual probability for poverty workers needed for growth and development— almost in half in both countries. Moreover, mi- gration plays an important role in providing a wider set of opportunities for both Moroccans 3 Sadiqui (2004) shows that women in rural areas have been and Egyptians, allowing households to save, cope disadvantaged of urban modernization. For instance, their edu- cation and health access has traditionally been limited. xvii   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options often (and inaccurately) referred to as the brain high-skilled migrants emanate from the stock drain. While skilled emigration can increase the of unemployed—or alternatively, have skill sets welfare of the individual, it can, in the general so- that can be more easily replaced from the pool ciety of sending countries, lower the average level of unemployed in the local labor market, the of human capital, thereby draining it of productive true costs of high-skilled migration are lower. capacity. Additionally, since human capital can generate positive externalities through innovation Not only are North Africa’s unemployment and adoption of technology, result in higher fiscal rates high, there is considerable information to revenues, and prompt better governance, it is show that educated workers are not fully de- possible that high-skilled migration may result in ployed to their most productive uses. Productiv- lower economic growth if educated persons leave ity estimates for the region suggest that North a country. Shortages of skilled health personnel, Africa and the broader MENA region suffers from teachers, engineers, and more general “institution persistently low productivity. builders� in developing countries can have impact beyond the migrants’ personal contributions to Several factors underlie the poor use of skills output, leading to major bottlenecks in progress in North Africa. One large factor is the fact that toward economic and social development. the quantity, quality and focus of educational outcomes are not in line with the economic devel- Data on stocks of migrants in OECD coun- opment objectives. MENA countries rank poorly tries and the GCC indicate that North Africa’s in international tests of science and mathematics skilled emigration rates are high (averaging 11 proficiency, and education systems are geared to percent of the domestic stocks of higher edu- prepare workers for civil service, rather than for cated laborers). But it is not possible, given data jobs in a dynamic and competitive private sector. constraints, to evaluate the lost contributions of high-skilled emigrants from North Africa. How- Education reforms have focused on increas- ever, there are at least a few conditions evident in ing access and improving quality, but without North African labor markets that indicate that the enhancing the motivation for the productive use environment for the productive use of education of education in the economy. Among the most in the domestic economy is poor, suggesting that distortive policies inhibiting the productive use the actual ‘lost’ output from skilled migration is of educated labor (and all labor) in the region is lower than might appear at first glance. the legacy of public sector employment. Despite attempts to reduce the size of the public sector, To begin, a significant portion of North Af- MENA (and North African) economies maintain rica’s (and the broader MENA region’s) stock some of the highest levels of public sector em- of higher educated laborers are unemployed. ployment in the world. A body of research on Unemployment rates for university graduates public-sector employment suggests that human are significant and higher for university gradu- capital in the public sector does not contribute ates than for other groups. In Egypt, while those to economic growth, but actually reduces it. with at least secondary education make up only 42 percent of the labor force, they account Ultimately, the cost of skilled migration for 80 percent of the unemployed. In Algeria, needs to be measured against whether the only about 20 percent of the labor force has counterfactual holds: whether lack of skills is completed secondary education or more, but the binding constraint to economic develop- they account for almost twice that proportion ment, and whether the human capital produced of the unemployed, and the same is true for is productively employed in the home country. Morocco. In addition, the skills composition of Both high unemployment and low productivity emigration strongly mirrors the composition of among educated workers point to the possibility unemployment. Returned migrant surveys from that a large portion of the costs of ‘lost’ educated Algeria, Morocco, and Tunisia also point to high workers is attributable not to migration, but levels of unemployment prior to migration. If rather is a feature of a domestic environment xviii Overview in which education is not channeled to its most sector functions in North Africa. Entrepreneur- productive purposes. ship is also higher for those who received voca- tional training while abroad; these returnees have Much attention has been paid to the devel- higher instances of being first time employers or opment benefits that countries can reap from self employed than those who do not migrate. The migrants, through their initiatives as diasporas, results suggest there is potential for small business and perhaps ultimately, from their return. The start-up programs with market studies, micro- review of diaspora networks of migrants from credit, and training components; thereby easing the Maghreb and Egypt reveals they are already some of the constraints on entrepreneurship that well developed in all major destination countries arise from capital market imperfections or other of Europe and North America, including such market failures, such as lack of information. formal systems as associations of university graduates, philanthropic organizations, entrepre- And as important, the ability and incentives neurial associations, and other professional net- for returnees to invest or to have labor market works. However, the ability of diaspora networks success is significantly affected by the conditions to contribute fully to home country development of return. In particular, those migrants who re- has been hindered by the lack of responsiveness, turned for administrative reasons are less likely flexibility and dynamism on the part of govern- to experience upward job mobility, and less likely ment. Diaspora programs and organizations are to invest once home. This suggests a potential too centralized and bureaucratized to provide growing importance for programs to support in effective support to diaspora members and the reintegration of irregular migrants in home their organizations. Thus, there remains large countries. potential for North African economies to better engage the diaspora for growth and development- Increasingly, it is understood that the expe- enhancing investments and to create the virtuous rience of migrants abroad and the welfare gains circles for institutional change. for migration have a critical dependency on the overall policy and institutional environment Return migrants represent another poten- toward migrants in the host country. Much of tially important source of knowledge and invest- the European focus on migration has centered ment to home countries. While lack of appropri- on security, national identity, and vulnerability ate data prevents an empirical investigation of associated with dependence on immigrants for the impact of diaspora or returned migrants on entire productive sectors. Immigration is often growth or development in North Africa, survey blamed for problems in the host country such as data can provide a glimpse into what makes one high unemployment, the draining of government returned migrant contribute more to the home welfare programs, and for increasing inequality. economy than another: that is, what kind of ex- As a result, there is no strong political constitu- perience abroad creates the infrastructure for ency for mass immigration. At the same time, the maximum contributions upon return. From the need for less-skilled immigrants in Europe, analysis of the return migrant survey data, some based on labor demand, is clear. important results emerge. The dichotomy between stated European First, entrepreneurship among return mi- migration policy objectives to control mass im- grants depends strongly on remittances. The migration and the ambivalence toward outcomes results suggest that the higher the remittances, (at least regarding less-skilled illegal immigra- the lesser the budget constraint after return, an tion), has created inherent distortions which important determinant of entrepreneurial behav- have impacted both the incentives for migrants ior. Indeed, for non-investors, access to financing to return, and also the ability for North Africa’s is considered to be the biggest impediment in all migrants—as diaspora and upon return—to con- three countries, especially in Morocco, providing tribute to home country development. Policies further evidence of the weaknesses of financial to limit low-skilled immigration have had little xix   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options impact on the flows, but they certainly have Based upon current education levels, MENA had repercussions on the status of migrants. countries (including North Africa) will provide In the case of Spain, rigid controls on formal a poor match to the projected skill needs. Since labor migration have resulted in a burgeoning most of the additional population will have pri- of irregular workers entering the country, with mary education levels, only a fraction of MENA’s implications on productivity and access to social additional work force would be of interest to services. In the Netherlands, on the other hand, European countries. as opportunities for labor migration waned, fam- ily and asylum migration grew rapidly. In both Moreover, if education levels in the North countries, migration flows have ultimately been Africa labor force remain low, while the demand largely determined by labor demand conditions. for skills increases significantly across the Medi- terranean, there is the considerable risk for the Demographic forces, which make increased region losing scarce talent, expertise, and its edu- North African migration virtually guaranteed, cational investments to Europe—turning skilled increase the need to better understand the po- migration into the so called “brain drain.� While tential impacts of migration. The long term need there are potential positive feedback effects from to replace an aging work force, especially in Eu- skilled migration, the loss of a large portion of ropean countries, will require large scale popu- MENA’s skilled talent is likely to have damaging lation flows, without which European countries impacts on growth, unless accompanying policies will lose hundreds of millions of workers over are put in place to mitigate them. the next four decades and will facing increasing strain on the productive part of the population General equilibrium analysis can only provide to support both the upbringing of the young and scenarios, and results are determined by the the pensions of old. There are direct impacts on assumptions underlying the model. However, financial elements like social security. the scenarios are useful for modeling the types of impacts the region (and Europe) can expect The demographics of North Africa (and the from large-scale increases in migration, whether broader MENA region) suggest that increasing they be low-skilled or high-skilled, and whether migration flows from North Africa would clearly they are accompanied by other feedbacks and attenuate the deterioration of the European policies. These scenarios point to the increasing demographic structure. However, demographic potential costs from large increases in skilled analysis also shows that the ability for MENA migration from North Africa, even in the face of countries to mitigate the labor market imbal- diaspora contributions (although remittances ances is temporary: MENA countries themselves and diaspora contributions mitigate some of will begin to experience serious demographic those effects). However, these scenarios also problems after 2030. Indeed, the MENA aged- suggest the role that accelerated human capital dependency ratio in 2050 will approximately formation can play in the region to mitigate the be equal to the current aged-dependency ratio effects of the brain drain. observed in the EU. To meet the challenge, the MENA region Moreover, while MENA’s demographic profile needs to improve their education system sig- fits the European labor needs, its skills do not. nificantly. This would provide the region with a Projections over the short, medium and long much needed force for economic development, term suggest that labor shortages in Europe will and—in the face of increasing demand for skilled occur across the skills spectrum, with significant migrants—ensure that human capital at home is demand for medium-level skills (such as nurses, not entirely depleted. But the challenge is not intermediate business services) or even relatively only (or even mostly) quantity. It requires seri- low skills (retail salespersons, waiters, etc.). ously addressing the educational quality issues By contrast, jobs at low qualification levels are at all levels of education. But also, education ef- expected to fall. forts will not pay off unless there is an enabling xx Overview environment that allows for a productive use for tion more broadly, the effect of those changes knowledge in the economy. And skills shortages on household welfare (especially the poor), on may not, in fact, be the main problem for poten- inequality, and on households’ livelihood, diver- tial and actual entrepreneurs, whether based sification and risk management strategies. in the North Africa region or diaspora abroad. Political instability, corruption, high informality, The limited knowledge base on the size of and other factors seem to be just as important migration, its determinants and main impacts is bottlenecks in progress toward higher produc- due to two problems, on the input and output tivity and growth. Indeed, these will be the very side. First, on the input side, there is a dearth of same factors that encourage some of the best and data. But also, on the output side, there appears brightest to leave their country in the first place. not to exist clear strategies or approaches on how to turn data into knowledge. Much higher and The agenda for improving the economic and better exploration of existing data will be needed. social impacts of migration is large, and one of the central factors that will determine the abil- For that reason, it is especially important for ity for North African economies to make good data being made widely accessible. Transparency migration policy is the existence of good data. of data and access for research is important both Migration research is hampered by significant to find limitations or deficiencies in the data that data limitations, as confirmed by nearly all stud- can lead to improvements in data collection, ies and researchers. And, while migration is, in thereby leading to improved policies—and that general, data handicapped, North Africa coun- can help citizens and businesses plan for the tries (or the MENA region more generally) are future. considerably more data-challenged—across all areas of economic and social information, both European policy also has an obvious role in in terms of collection and importantly in terms the ability for migration to be a tool for North of public access—than other countries of similar Africa’s development. A central issue is social income levels. protection for migrants, which impacts not only the migrant condition but also the incentives Devising good migration policies and ensur- for return. Social protection for MENA migrants ing well-founded policy options requires both in Europe is largely covered by the so-called internal and external guidance. MENA policy- Euro-MED agreements. In addition to the Euro- makers should continue to review and assess Med agreements, Maghreb countries enjoy more the outcomes of international best practices in advantages through bilateral labor migration and migration management and consider how good social security agreements with several European practice can be adapted to the specific national countries. and regional circumstances. Policymakers must realize, however, that the most critical obstacle Although much has been done and progress to the adaptation of international policies as has been good, there are still loopholes in social practice is a lack of migration data from which to protection systems. One such loophole is related monitor the trends in the size, determinants and to pensions exportability, which requires a ten- impacts of emigration and immigration. year minimum for the social security system, and in some cases, pensions are reduced when Monitoring and evaluation of migration and transferred abroad. Clearly, such provisions dis- other social and economic data is critical in order courage return migration. Another inconsistency to understand the impact of different policies occurs in the interaction between national provi- and adapt and improve the current policy mix. sions for social protection and immigration and Apart from the long run implications of migration alien laws. Most European countries grant full on growth and welfare, it is also essential that social protection coverage to long-term migrants policymakers find a way to gauge the impact of or permanent residents although the interac- the economic crisis on remittances and migra- tion between social and alien law complicates xxi   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options the de facto access. For example, temporary from migration to the host country are precisely migrants have limited access to health care and the elements which increase growth and devel- unemployment insurance, and no access to social opment overall. High-skilled workers with skills assistance or public housing. Yet, continental relevant for a private sector led economy not European countries require up to ten years of only increase the employment potential (and residence before granting permanency, and in the remittance potential) abroad; they increase southern European countries permanent status the employment potential at home. Having an has only very recently become possible. investment climate which provides sufficient job opportunities for these higher educated workers Clearly, a balance needs to be struck between at home not only reduces skilled labor flight, it fiscal sustainability, the risk of migrant attraction also increases the contribution of human capital to generous welfare systems, and ensuring a level to growth and development. An investment cli- playing field on the one hand, and the rights of mate that attracts business development from legal and irregular migrants on the other hand. If returned migrants also attracts the domestic European countries wish to attract more migrant investment for economic growth and develop- workers in the future, their specific vulnerabil- ment. Having the institutions that support dias- ity—when recently arrived, and in old age—must pora engagement will also contribute to a better be taken into consideration, along with the risk climate for investors domestically. of distorting incentives for return migration, cir- cular migration and the labor market flexibility This research program can only fill a part of that migrants potentially can provide. the knowledge gap in understanding the impacts and coming challenges from North Africa migra- Finally, and ultimately, all of the policies tion. It is hoped that the work sets the stage for a which impact the degree to which North Africa sustained commitment from all parties to better can benefit from migration are precisely the poli- analyze and understand the migration phenom- cies that contribute to growth and development enon, and look for possibilities to maximize is at home. The elements of the incentive frame- potential benefits on both sides of the Mediter- work which would increase the positive returns ranean. xxii Vue d’Ensemble du Programme de Recherche Le présent programme de recherche sur la trouver une réponse à de nombreuses questions migration porte sur quatre pays de l’Afrique que l’on souhaiterait poser. du Nord – le Maroc, la Tunisie, l’Algérie et l’Egypte. Comme cela a été le cas pour la ré- Par le passé la migration était seulement gion du Moyen-Orient et de l’Afrique du Nord envisagée comme « un échec du développement dans son ensemble, la migration a joué un rôle », reflet de l’absence d’opportunités dans le pays central pour influer sur le développement social d’origine. En effet, bon nombre des facteurs qui et économique de ces quatre pays. Bien qu’il y incitent à la migration – chômage, pauvreté, et ait dans le monde des régions exportatrices de instabilité politique – sont la preuve palpable main-d’oeuvre plus importantes, sur une base d’une insuffisance de développement dans le par habitant et en part du PIB, les envois de pays. fonds vers l’Afrique du Nord4 sont parmi les plus élevés au monde. Parallèlement, et de plus en plus, la migra- tion est par la même occasion perçue comme En dépit de cette importance économique, il un mécanisme puissant pour la croissance et le est surprenant de constater que peu d’attention développement. La migration ouvrant la porte a été portée à comprendre le phénomène de la aux ressources financières (sous forme d’envois migration dans la région de l’Afrique du Nord. de fonds), à l’investissement et au transfert du Le programme de recherche vise à combler savoir (migrants rentrant au pays et diaspora), certains aspects de l’important fossé au niveau au développement des compétences, et aux ré- du savoir, en mettant l’accent sur les impacts de seaux importants pour les entreprises, les filières développement de la migration nord-africaine, potentielles pour que la migration vienne en aide sur les défis auxquels la région fait face pour ce au développement son nombreuses. qui est de la migration de la main-d’oeuvre dans le proche et le long terme, et sur les options L’information sur les stocks de migrants de politique visant à maximiser le potentiel de nord-africains n’est pas récente. En 2000, le migration afin qu’il constitue un outil du déve- taux moyen d’émigration des pays de l’Afrique loppement. Dès le début il nous faut reconnaître du Nord (par rapport aux économies de l’OCDE que l’étude sur la migration de l’Afrique du Nord et du CCG combinées) se situait aux environs de est fortement contrainte par les limitations des 5,5 % de la population résidente et atteignait 9 données, comme cela est confirmé par pratique- ment toutes les études et tous les chercheurs. Ces difficultés touchent tous les aspects du sujet, 4 En raison des limitations des données, l’analyse pour l’Afrique du Nord est limitée aux quatre pays susmentionnés. Lorsque y compris les stocks et les caractéristiques des cela s’avère possible ou nécessaire, les données pour la région migrants, les flux des migrants, les envois de MENA au sens plus large sont alternativement discutées tout fonds de l’étranger, et autres. Ceci empêche de au long du rapport. xxiii   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options % au Maroc, mais la destination et la nature de cielle. La migration plus informelle a commencé la migration diffèrent pour les pays du Maghreb avec l’introduction de nouvelles destinations et l’Egypte. Les migrants des pays du Maghreb (dont l’Espagne et l’Italie), et l’introduction de sont souvent des travailleurs non qualifiés à la visas Schengen (rendant l’entrée légale de la recherche d’opportunités d’emploi leur permet- main-d’oeuvre plus difficile). tant de s’installer en permanence en Europe. A l’inverse, les migrants égyptiens ont davantage La nature de la migration d’Afrique du Nord tendance à être des migrants temporaires dans a un impact majeur sur les coûts et les avantages d’autres pays arabes. que la région peu tirés du phénomène migratoire. La migration comporte des niveaux variables de Les niveaux des compétences des migrants coûts et avantages pour l’économie tout au long diffèrent aussi, avec une prépondérance de mi- de ses différentes étapes, mais l’ampleur de ces gration peu qualifiée en provenance des pays du coûts et avantages est largement influencée par Maghreb (70 %) et de migration plus diversifiée les caractéristiques des migrants et par les ca- de travailleurs peu qualifiés et très qualifiés en ractéristiques et incitations structurelles du pays provenance d’Egypte (55 % contre 30 %). Pa- d’origine et des pays d’accueil. rallèlement, tous les pays de la région ont fait montre de taux élevés de migrants hautement La capacité à analyser les impacts de la qualifiés (ou plus précisément plus éduqués), se migration d’une manière systématique est for- situant entre 8 % et 19 % de la population active tement entravée par le manque de données nationale hautement qualifiée. appropriées pour l’ensemble de la région. En conséquence, seule une partie des coûts et La migration nord-africaine, en particulier avantages potentiels de la migration a pu être du Maghreb, s’est également féminisée princi- saisie dans le présent programme de recherche. palement dans le contexte de la réunification des familles, mais également en raison d’une Au niveau global, on constate un intérêt à participation croissante des femmes au le marché appréhender l’impact de la migration sur la crois- du travail. Cette féminisation des migrants a eu sance et la productivité en Afrique du Nord, en des impacts importants à la fois pour l’Afrique particulier à la lumière du recul éventuel de la du Nord et l’Europe. En Europe, la disponibilité migration et des envois de fonds dans le contexte prédominante d’employées de maison a contri- de la crise financière mondiale. Il est projeté que bué à une forte hausse de la participation des la Région de l’Afrique du Nord enregistrera une femmes autochtones au marché du travail (et à baisse de 10,6 pour cent des envois de fonds en la croissance). 2009. Toutefois, au niveau de l’économie, les impacts de la migration résultant de la crise Et, quoique pour des raisons évidentes, il n’y financière mondiale seront vraisemblablement a pas de données fiables sur la migration illégale, peu importants : les envois de fonds ne se situant avec l’établissement de réseaux de migration et qu’à environ 4,3 pour cent seulement du PIB et avec les changements intervenus dans la poli- dans le contexte projeté d’une baisse des envois tique migratoire, la proportion de travailleurs régionaux de 10,6 pour cent en 2009, l’impact net nord-africains illégaux a vraisemblablement sur le PIB de 2009 ne sera que de l’ordre de 0,5 augmenté. Dans les années 60 et 70, la migration points de pourcentage (mais pouvant atteindre s’opérait généralement par le biais de filières 1,3 points de pourcentage au Maroc). officielles et revêtait un caractère circulaire. Toutefois, dès que les réseaux ont été établis Cependant, l’analyse de séries chronolo- dans les pays d’accueil, les nouveaux migrants giques transfrontalières des envois de fonds et ont de plus en plus contourné la bureaucratie du de la croissance indique aussi une interaction recrutement officiel et une filière autonome « non entre les envois de fonds et le secteur financier officielle » a émergé en parallèle à la filière offi- en Afrique du Nord : notamment que plus le xxiv Vue d’Ensemble du Programme de Recherche système financier est superficiel, plus fort sera le plus bas. En outre, les envois de fonds contri- l’impact des envois de fonds sur la croissance. buent à minimiser la probabilité de se trouver Ceci confirme que les envois de fonds agissent en situation de pauvreté d’environ 7 points de comme un substitut du système financier. Bien pourcentage au Maroc et d’environ 9 points de que l’impact sur la croissance soit minime – en pourcentage en Egypte. A savoir que la présence raison de la portée restreinte des envois de fonds des envois de fonds dans le ménage réduit la vers les économies nord-africaines – les résultats probabilité d’être pauvre au Maroc d’environ 15 confirment la conclusion d’un faible développe- % à environ 8 % et en Egypte d’environ 20 % à ment général du secteur financier dans les pays environ 11 %. de la région MENA. Les évaluations du climat de l’investissement dans la région montrent que L’analyse indique que si la migration peut rares sont les entreprises privées qui font appel jouer un rôle pour atténuer la pauvreté, ce n’est au crédit bancaire pour financer l’investissement, pas la panacée pour réduire la pauvreté au ni- comptant plutôt et principalement sur les bé- veau national. A la fois au Maroc et en Egypte, néfices non répartis. Et en dehors de l’Afrique l’impact d’atténuation de la pauvreté de la mi- subsaharienne, aucune région n’a un pourcentage gration et des envois de fonds au niveau national plus élevé d’entreprises qui identifient l’accès est modeste. 3,5 % seulement de la population au crédit comme étant une contrainte majeure égyptienne bénéficiaient d’envois de fonds au à la marche des affaires. En conséquence, les moment de l’enquête, l’absence de ces envois envois de fonds ont une influence plus forte sur n’aurait donc relevé l’indice numérique global la croissance dans la région MENA (et l’Afrique de la pauvreté que d’environ 0,4 point de pour- du Nord) que dans les pays dotés de secteurs centage. Au Maroc, l’absence d’envois de fonds financiers mieux développés. Dès lors, les pays aurait contribué à relever l’indice numérique de la région MENA sont plus vulnérables à de de la pauvreté de moins de 1 point de pourcen- soudaines fluctuations des envois de fonds. tage. Ainsi, la capacité de la migration à alléger substantiellement la pauvreté au niveau national Etant donné que la migration, par rapport à est-elle limitée. la population, demeure peu importante (au Ma- roc, environ 11 % des ménages ont un migrant Toutefois, pour le ménage individuel, la mi- à l’étranger ; en Egypte 3,5 %), de nombreux gration est un élément important de la stratégie impacts de la migration ne sont pas visibles au pour faire face aux risques. En effet, les envois niveau macroéconomique mais apparaîtront au de fonds réduisent la probabilité individuelle de niveau des décisions prises par le ménage et des pauvreté de près de la moitié dans les deux pays. résultats. Dans deux pays seulement de la région, En outre, la migration joue un rôle important en le Maroc et l’Egypte, les données permettent de offrant une plus vaste gamme d’opportunités réaliser une analyse empirique de l’impact des à la fois aux Marocains et aux Egyptiens, pour envois de fonds ou de la migration, ou des deux. permettre aux ménages d’épargner, faire face L’analyse fait montre de quelques conclusions aux crises et simultanément pour lancer ou dé- importantes : velopper des activités génératrices de revenus, et donc pour diversifier les moyens d’existence. Premièrement, la migration et les envois de fonds ont réduit la pauvreté en Afrique du Nord. Les envois de fonds ont aussi permis d’ac- Si les envois de fonds bénéficient au segment le croître les dépenses à la santé et à l’éducation des plus aisé (avec 61 % des envois de fonds allant enfants et de réduire le travail des enfants, ce qui aux deux quintiles de revenus supérieurs, au à long terme est vital pour forger les bases d’une Maroc et en Egypte), ils bénéficient néanmoins croissance accrue. En Egypte et au Maroc la mi- aussi aux pauvres. Au Maroc, par exemple, les gration et les envois de fonds ont été à l’origine envois de fonds s’inscrivent pour plus de 50 % d’améliorations au niveau des investissements du budget des ménages du quintile de revenus consentis aux enfants et en Egypte la migration xxv   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options et les envois de fonds ont nettement réduit la comme le capital humain peut générer des exter- probabilité du travail des enfants. Les envois de nalités positives par le biais de l’innovation et de fonds sont également à l’origine d’une allocation l’adoption de la technologie, être à l’origine de intra-ménagère plus équilibrée à l’éducation et recettes fiscales plus importantes, et inciter à une à l’alimentation entre les garçons et les filles, ce meilleure gouvernance, la migration hautement qui a contrebalancé une allocation sexiste des dé- qualifiée peut avoir des impacts substantiels sur penses à l’éducation et à l’alimentation, et incité la croissance si la population éduquée quitte le à une parité entre les sexes au niveau de l’accu- pays. Les pénuries de personnel qualifié de santé, mulation du capital humain des enfants du milieu enseignants, ingénieurs et plus généralement rural où les filles sont davantage marginalisées. de « créateurs d’institutions » dans les pays en développement peuvent avoir un impact allant Un autre impact potentiel de la migration au-delà des contributions personnelles des mi- a trait aux marchés du travail, non seulement grants à la production, et constituer d’importants pour ce qui est des sorties de travailleurs mais goulets d’étranglement pour le développement aussi des changements intervenants dans les économique et social. décisions d’emploi de ceux qui sont restés sur place (notamment pour les ménages bénéfi- Les données relatives aux stocks de migrants ciaires d’envois de fonds). Toutefois, les données dans les pays de l’OCDE et du CCG indiquent concernant l’Egypte et le Maroc indiquent que que les taux de l’émigration qualifiée en Afrique les impacts sont, aux mieux, modestes. Dans du Nord sont élevés (en moyenne 11 % du stock le cas de l’Egypte, les résultats suggèrent un national de travailleurs très éduqués). Mais il impact positif des envois de fonds sur l’offre de n’est pas possible, en raison des contraintes de main-d’oeuvre, même si cette amélioration dérive données, d’évaluer les contributions perdues principalement des femmes qui s’intègrent à la des émigrants qualifiés de l’Afrique du Nord. main-d’oeuvre familiale non rémunérée. Dans Cependant, il y a au moins quelques conditions les deux pays, les envois de fonds augmentent évidentes sur les marchés nord-africains du tra- aussi la probabilité d’un emploi indépendant aux vail qui indiquent que l’environnement propice à dépens d’un emploi salarié. Tenant compte de la une utilisation productive de l’éducation au sein nature du secteur indépendant (principalement de l’économie nationale est faible, ce qui suggère peu qualifié), ceci montre que les envois de fonds que la « perte » productive réelle de la migration peuvent servir de capital pour des activités plus qualifiée est moindre qu’anticipée au départ. informelles garantissant une plus grande flexi- bilité d’emploi. Tout d’abord, une proportion importante du stock des travailleurs qualifiés d’Afrique du Nord Les autres preuves d’impact de la migration (et de la région MENA) est sans emploi. Les taux nord-africaine dérivent seulement d’un témoi- de chômage pour les diplômés universitaires sont gnage descriptif, suggestif (mais non définitif) des significatifs et plus élevés pour ces derniers que coûts et avantages associés à la migration. Le coût pour d’autres groupes. En Egypte, la population éventuel le plus important pour la région est lié à ayant suivi au moins l’enseignement secondaire la perte de travailleurs plus éduqués nécessaires à ne représente que 42 % seulement de la popula- la croissance et au développement – souvent (et tion active mais elle s’inscrit pour 80 % des sans incorrectement) évoquée comme étant l’exode emploi. En Algérie, environ 20 % de la population des cerveaux. Si l’émigration qualifiée peut ac- active ont complété l’enseignement secondaire croître le bien-être de l’individu, pour la société, ou plus, mais ils s’inscrivent pour près de deux elle fait courir le risque de minimiser les perspec- fois la proportion des sans emploi et il en va de tives de croissance, si elle contribue à faire bais- même au Maroc. Par ailleurs, la composition des ser le niveau moyen du capital humain des pays compétences des migrants traduit largement d’origine et donc à les drainer de leur capacité la composition des sans emploi. Les enquêtes productive. Outre la perte de capacité productive, consacrées à la migration de retour en Algérie, xxvi Vue d’Ensemble du Programme de Recherche au Maroc et en Tunisie indiquent aussi de hauts En fin de compte, le coût de la migration qua- niveaux de chômage avant la migration. Si des lifiée doit être mesuré pour établir le bien-fondé migrants qualifiés émanent du stock des sans em- du contrefactuel suivant : le manque de compé- ploi – ou alternativement ont des compétences tences est-il la contrainte la plus importante au qui peuvent pas être facilement remplacées développement économique et le capital humain par le pool des sans emploi du marché local du produit est-il employé d’une manière productive travail, les coûts réels de la migration qualifiée dans le pays d’origine. A la fois le chômage élevé sont moindres. et le faible niveau de productivité des travailleurs éduqués soulignent la possibilité qu’une part Non seulement les taux de chômage de importante des coûts des travailleurs éduqués « l’Afrique du Nord sont élevés, mais une documen- perdus » est une caractéristique de l’environne- tation considérable montre que les travailleurs ment national dans lequel l’éducation n’est pas éduqués ne sont pas déployés de la manière la orientée à ses fins les plus productives (et non plus productive. Les estimations de la producti- pas à la migration). vité de la région indiquent que l’Afrique du Nord et la région MENA au sens large enregistrent en Une attention particulière a été portée aux permanence un faible niveau de productivité. avantages de développement que les pays peu- vent tirés des migrants, par le biais d’initiatives Plusieurs facteurs expliquent le faible niveau telles que les Diasporas, et peut-être en fin de d’utilisation des compétences en Afrique du compte, la migration de retour. L’examen des Nord. Un des facteurs est le fait que la quantité, réseaux de Diaspora de migrants du Maghreb qualité, et l’attention portée aux réalisations pé- et d’Egypte montre qu’ils sont bien développés dagogiques ne sont pas alignés sur les objectifs du dans les principaux pays d’accueil d’Europe et développement économique. Le classement des d’Amérique du Nord, y compris dans le cadre pays de la région MENA dans les tests internatio- d’associations de diplômés universitaires, naux de maîtrise des sciences et mathématiques associations d’entreprises, et autres réseaux est médiocre et les systèmes éducatifs préparent professionnels. Toutefois, leur capacité à les travailleurs pour un emploi dans la fonction contribuer pleinement au développement a été publique plutôt que pour un emploi dans un entravée par le manque de réactivité, flexibilité secteur privé dynamique et compétitif. et dynamisme de la part du Gouvernement. Les programmes et organisations sont trop centra- Les réformes du secteur de l’éducation lisés et bureaucratisés pour assurer un appui ont mis l’accent sur un meilleur accès et sur effectif aux membres de la Diaspora et à leurs l’amélioration de la qualité, mais sans toutefois organisations. Les économies nord-africaines renforcer la motivation pour une utilisation plus ont donc un important potentiel pour mieux productive de l’éducation dans l’économie. Une engager la Diaspora dans des investissements des politiques les plus biaisées qui fait obstacle axés sur le renforcement de la croissance et du à une utilisation productive de la main-d’oeuvre développement et créer des cercles virtuels pour éduquée (et de la main-d’oeuvre en général) le changement institutionnel. de la région est l’héritage de l’emploi public. En dépit des tentatives en vue de réduire la taille du Les migrants qui rentrent au pays constituent secteur public, les économies de la région MENA une autre source potentielle importante de savoir (et nord-africaines) ont toujours les niveaux et d’investissement pour les pays d’origine. Le les plus élevés d’emploi public au monde. Et manque de données adéquates ne permet pas un organe de recherche axé sur l’emploi public de réaliser un examen empirique de l’impact suggère que le capital humain du secteur public de la Diaspora ou des migrants qui rentrent au non seulement ne contribue pas à la croissance pays sur la croissance et le développement en économique, mais qu’en réalité il contribue à la Afrique du Nord, mais les données d’enquête minimiser. peuvent donner un aperçu de ce qui fait qu’un xxvii   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options migrant rentrant au pays peut contribuer plus des programmes d’assistance sociale du gouver- à l’économie de son pays qu’un autre, à savoir : nement et pour l’accroissement des inégalités. quel est le type d’expérience acquise à l’étranger Par conséquent, il n’y a pas de groupe d’intérêt qui contribuer à créer l’infrastructure requise politique fort pour l’immigration de masse. Dans pour une contribution maximum lors du retour le même temps, le besoin d’immigrants moins dans le pays d’origine. Des résultats intéressants qualifiés en Europe, en fonction de la demande émergent de l’analyse des données d’enquête de main-d’œuvre, est clair. auprès des migrants rentrant au pays. La dichotomie entre les objectifs déclarés Tout d’abord, l’esprit d’entreprise des mi- de la politique de migration européenne pour grants qui rentrent au pays dépend largement contrôler l’immigration de masse et l’ambiva- des envois de fonds. Les résultats indiquent lence vis-à-vis des résultats (du moins en ce qui que plus les envois de fonds sont importants, concerne l’immigration illégale moins qualifiée), moindres seront les contraintes budgétaires au a créé des distorsions inhérentes qui ont eu un retour, un déterminant majeur du comportement impact sur les incitations au retour des migrants entrepreneurial. En effet, pour les non inves- mais aussi sur la capacité des migrants d’Afrique tisseurs, l’accès au financement est considéré du Nord –en tant que diaspora et au retour – à comme étant l’obstacle le plus important dans les contribuer au développement du pays d’origine. trois pays, en particulier au Maroc, ce qui ne fait Les politiques qui limitent l’immigration à faibles que confirmer les problèmes du secteur financier qualifications ont eu peu d’impact sur les flux, en Afrique du Nord. mais elles ont eu certainement des répercus- sions sur le statut des migrants. Dans le cas de L’esprit d’entreprise est également plus l’Espagne, les contrôles rigides sur la migration développé pour ceux qui rentrent au pays formelle de la main-d’œuvre ont eu pour résul- avec une expérience antérieure d’employeurs tat un foisonnement de travailleurs irréguliers ou d’indépendants, et pour ceux qui ont suivi entrant dans le pays, avec des implications sur une formation professionnelle au cours de leur la productivité et l’accès aux services sociaux. séjour à l’étranger. Et, tout aussi important, la Aux Pays-Bas d’autre part, alors que les oppor- capacité et les incitations de ceux qui rentrent tunités pour la migration de la main-d’œuvre au pays à investir ou à remporter un succès sur diminuaient, la migration de la famille et l’asile se le marché du travail sont fortement influencées développaient rapidement. Dans les deux pays, par les conditions de leur retour. En particulier, les flux de la migration ont en dernier ressorts les migrants qui sont rentrés pour des raisons été largement déterminés par les conditions de administratives ont vraisemblablement moins la demande de main-d’œuvre. de possibilité de mobilité d’emploi en amont et seront moins susceptibles d’investir à leur retour. Les forces démographiques, qui virtuelle- ment garantissent une migration nord-africaine De plus en plus, il est entendu que l’expé- accrue, ne font que renforcer la nécessité de rience des migrants à l’étranger et les gains de mieux comprendre les impacts potentiels de la bien-être pour la migration ont une dépendance migration. La nécessité à long terme de rempla- critique sur la politique globale et l’environne- cement de la population active vieillissante, en ment institutionnel envers les migrants dans le particulier dans les pays européens, impliquera pays hôte. Une grande parte de l’axer européen des mouvements de population à grande échelle, sur la migration s’est centré sur la sécurité, sans lesquels les pays européens perdront des l’identité nationale et la vulnérabilité apparen- centaines de millions de travailleurs au cours des tée à la dépendance des immigrants pour des quatre prochaines décennies et auront à faire secteurs entiers de la production. L’immigration face à des tensions croissantes sur le segment est souvent blâmée pour les problèmes dans le productif de la population pour venir à l’appui pays hôte tels que le chômage élevé, l’épuisement de l’éducation des jeunes et des retraites des xxviii Vue d’Ensemble du Programme de Recherche personnes âgées. Les impacts sont directs sur des et le peu de talent et d’expertise dont elle dispose éléments financiers tels que la sécurité sociale. et en faveur de l’Europe – à savoir « l’exode des cerveaux ». Les études macroéconomiques an- Les démographiques de l’Afrique du Nord (et térieures ont montré que « l’exode des cerveaux de la région MENA dans son ensemble) montrent » commence à faire montre d’impacts nuisibles que des mouvements croissants de migration en sur le développement lorsqu’il excède un seuil provenance de l’Afrique du Nord atténueraient de 10-15 %. Bien qu’il y ait un retour d’informa- la détérioration de la structure démographique tion d’impacts potentiels positifs de la migration européenne. Toutefois, l’analyse démographique qualifiée, la perte d’une large proportion du talent montre aussi que la capacité des pays de la région de MENA aura vraisemblablement des impacts MENA à mitiger les déséquilibres du marché préjudiciables sur la croissance, sans politiques du travail est temporaire : les pays de la région d’accompagnement pour les atténuer. MENA eux-mêmes commenceront à avoir de sérieux problèmes démographiques après 2030. L’analyse d’équilibre général peut unique- En 2050 le rapport de la population non adulte à ment fournir des scénarios et les résultats sont la population adulte de la région MENA sera ap- déterminés par les hypothèses sous-jacentes proximativement égal à celui observé dans l’UE. au modèle. Toutefois, les scénarios sont utiles pour modéliser les types d’impacts que la région En outre, alors que le profil démographique (et l’Europe) peut anticiper d’une croissance à de MENA s’aligne sur les besoins en main- grande échelle de la migration (qu’elle soit peu d’oeuvre de l’Europe, il n’en va pas de même pour ou très qualifiée, et qu’elle soit accompagnée les compétences. Les projections pour le court, le d’autres retours d’information et de mesures moyen et le long terme montrent que les pénuries de politique). Ces scénarios soulignent la pro- de main-d’oeuvre en Europe se produiront sur babilité d’une hausse des coûts qu’entraînerait l’ensemble du spectre des compétences, avec une une forte croissance de la migration qualifiée importante demande pour les qualifications de en provenance de l’Afrique du Nord, même à la niveau intermédiaire (infirmiers/infirmières, ser- lumière des contributions de la diaspora (même vices commerciaux intermédiaires) voire même si les envois de fonds et les contributions de la pour les qualifications de niveau relativement diaspora atténuent certains de ces impacts). Ce- faible (vendeurs au détail, serveurs/serveuses, pendant, ces scénarios montrent aussi le rôle que etc.). A l’inverse, les emplois à faible niveau de l’accélération de la formation du capital humain qualifications devraient enregistrer du recul. peut jouer dans la région pour mitiger les effets de l’exode des cerveaux. Sur la base des niveaux actuels d’éducation, les pays de la région MENA (et de l’Afrique du Pour relever le défi, la région MENA doit Nord) seront mal positionnés pour faire face aux améliorer son système éducatif. Ceci doterait la besoins projetés en compétences. Etant donné région d’un atout essentiel pour le développe- que la majorité de la population additionnelle ment économique, et – au regard de la demande n’aura qu’un niveau d’enseignement primaire, croissante pour des migrants qualifiés – veillerait seule une fraction de cette population active à ne pas épuiser entièrement le capital humain additionnelle présentera un intérêt pour les pays du pays. Le défi ne se situe pas seulement (ou européens. principalement) au niveau de la quantité. Il im- plique aussi de remédier aux problèmes liés à la Par ailleurs, si les niveaux d’éducation de qualité de l’éducation à tous les niveaux. Mais la population active de l’Afrique du Nord de- aussi, les efforts d’éducation ne seront pas cou- meurent faibles, alors que la demande pour des ronnés de succès sans un environnement propice compétences augmente fortement dans la Mé- permettant une utilisation productive du savoir diterranée, le risque est considérable de voir la dans l’économie. En fait les pénuries de compé- région perdre ses investissements pédagogiques tences peuvent ne pas s’avérer être le principal xxix   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options problème pour les entrepreneurs potentiels et bien-être, il est aussi essentiel que les décideurs établis, basés dans la région de l’Afrique du Nord trouvent un moyen de mesurer l’impact de la ou qu’il s’agisse de la Diaspora à l’étranger. L’ins- crise économique sur les envois de fonds et la tabilité politique, la corruption, le haut niveau migration plus généralement, l’impact de ces d’informalité et d’autres facteurs semblent être changements sur le bien-être des ménages, les des goulets d’étranglement tout aussi importants pauvres en particulier, sur l’inégalité, et sur les pour une productivité et croissance supérieures. conditions de vie des ménages, la diversification Ce sont en fait les mêmes facteurs qui, en premier et les stratégies de gestion du risque. lieu, encouragent certains des meilleurs et des plus brillants à quitter leur pays. La base de connaissances limitée de l’am- pleur de la migration, de ses déterminants et L’agenda pour améliorer les impacts écono- de ses principaux impacts a deux raisons : les miques et sociaux de la migration est vaste, et intrants et les extrants. Premièrement, sur le l’un des facteurs clés qui déterminera la capacité front des intrants, les données font défaut. Mais des économies nord-africaines à adopter une aussi sur le font des extrants, il semble ne pas y politique de migration judicieuse est l’existence avoir de stratégies ou approches claires quant à de données adéquates. Si, en règle générale, les la manière de traduire les données en connais- données sur la migration sont insuffisantes, celles sances. Une analyse plus détaillée et plus précise des pays de l’Afrique du Nord (ou de la région des données existantes est nécessaire. MENA au sens large) le sont encore nettement plus – dans tous les domaines de l’information Pour cette raison, il est particulièrement économique et sociale, à la fois pour ce qui est important que les données soient largement dis- de la collecte et, plus important encore, pour ce ponibles. La transparence et l’accès aux données qui est de l’accès public – que celles d’autres pays pour la recherche sont importants pour déter- à revenus similaires. miner leurs limitations ou lacunes et améliorer leur collecte afin de renforcer les mesures de L’élaboration de saines politiques de migra- politique et d’aider les citoyens et les entreprises tion et la garantie d’options de politique bien à planifier pour l’avenir. fondées impliquent un encadrement interne et externe. Ainsi, bien que les décideurs de MENA La politique européenne a également un rôle devraient poursuivre l’examen et l’évaluation évident à jouer pour que la migration devienne des résultats des pratiques internationales per- un outil du développement de l’Afrique du Nord, formantes et non performantes en matière de et un enjeu central est la protection sociale des gestion de la migration et envisager comment migrants, qui a un impact non seulement sur la adapter les pratiques performantes aux circons- condition du migrant mais aussi sur ses incita- tances nationales et régionales spécifiques, l’obs- tions à rentrer au pays. La protection sociale des tacle le plus critique à l’adaptation des politiques migrants de MENA en Europe est principalement internationales à la pratique est l’absence de couverte par les accords Euro-Med. Outre les données sur la migration sur la base desquelles accords Euro-Med, les pays du Maghreb bénéfi- assurer le suivi des tendances de l’importance, cient de plus d’avantages dans le cadre d’accords des déterminants et des impacts de l’émigration bilatéraux sur la migration de la main-d’oeuvre et et de l’immigration. la sécurité sociale avec plusieurs pays européens. Le suivi et évaluation de la migration et des Bien que les réalisations aient été nom- autres données sociales et économiques sont breuses et les progrès satisfaisants, les systèmes essentiels pour appréhender l’impact de diffé- de protection sociale comportent toujours des rentes politiques et adapter et améliorer leur failles. Un de ces failles consiste de l’exportabilité dosage actuel. Mises à part les implications à des retraites, qui implique une affiliation de 10 long terme de la migration sur la croissance et le ans minimum au système de la sécurité sociale, xxx Vue d’Ensemble du Programme de Recherche et dans certains cas, les retraites sont réduites la migration sont précisément celles qui contri- lorsqu’elles sont transférées à l’étranger. Il est buent à la croissance et au développement du clair que de telles dispositions découragent la pays d’origine. Les éléments du cadre incitatif qui migration de retour. Une autre inconsistance stimulerait une hausse des rendements positifs se situe au niveau de l’interaction entre les dis- de la migration pour le pays d’accueil sont pré- positions nationales de protection sociale et la cisément ceux qui contribuent à une croissance législation sur l’immigration et les étrangers. La accrue et au développement. Les travailleurs plupart des pays européens accorde une couver- hautement qualifiés dotés de compétences perti- ture intégrale de protection sociale aux migrants nentes pour une économie induite par le secteur de long terme et aux résidents permanents mais privé non seulement augmentent le potentiel l’interaction entre la législation sociale et celle d’emploi (et des envois de fonds) à l’étranger régissant les étrangers en complique de facto mais augmentent aussi le potentiel d’emploi dans l’accès. Par exemple, les migrants temporaires le pays d’origine. Un climat d’investissement qui ont un accès limité à l’assurance maladie et garantit suffisamment d’opportunités d’emploi chômage et aucun accès à l’assistance sociale aux travailleurs éduqués dans leur pays d’origine, ou au logement public. Et pourtant, les pays de non seulement réduit le risque de fuite de la l’Europe continentale imposent 10 années de main-d’oeuvre qualifiée mais il augmente aussi la résidence avant d’accorder le statut de résident contribution du capital humain à la croissance et permanent et dans les pays du Sud de l’Europe le au développement. Un climat de l’investissement statut permanent n’est possible que depuis peu. propice au développement entrepreneurial des migrants rentrant au pays stimule aussi l’inves- Il est évident qu’il faut trouver un équilibre tissement national à la croissance économique et entre la pérennité budgétaire, le risque que le au développement. Les institutions qui viennent migrant soit attiré par de généreux systèmes de à l’appui de l’engagement de la diaspora contri- bien-être, et la garantie d’une homogénéisation buent également à améliorer le climat interne des règles d’une part et les droits des migrants pour les investisseurs du pays. officiels et illégaux d’autre part. Si les pays euro- péens souhaitent attirer plus de migrants à l’ave- Le programme de recherche ne peut combler nir, leur vulnérabilité spécifique – arrivés récem- qu’une partie du fossé du savoir pour comprendre ment ou âgés – doit être prise en considération les impacts et défis futurs de la migration nord- ainsi que le risque de distorsions des incitations africaine. Il faut espérer que le programme à la migration de retour, la migration circulaire et prépare la voie d’un engagement durable de la flexibilité des marchés du travail que peuvent toutes les parties concernées à mieux analyser éventuellement garantir les migrants. et comprendre le phénomène de la migration et qu’il se penche sur les possibilités de maximiser Enfin, toutes les politiques qui influent le ses bénéfices potentiels des deux côtés de la degré auquel l’Afrique du Nord peut bénéficier de Méditerranée. xxxi Chapter 1 A Heightened Urgency for Understanding International Migration International migration has occupied an impor- guest-worker programs in Austria, France, and tant space in the social, economic, and cultural Germany. Bilateral migration agreements and development of North Africa. For more than half conventions were signed between the Maghreb a century, the broader Middle East and North and Europe, and large-scale labor migration oc- Africa region5 witnessed some of the largest curred between the mid-1960s and mid-1970s. population movements of any geographic re- gion in the world—a result of extensive transit By the mid-1970s, however, with stagflation migration, labor migration, family reunification, and unemployment growing in Europe, the mi- and large and protracted situations of refugee gration agreements were terminated, and family migration. Labor migration comprises only one, reunification, family formation, and asylum be- but important element of these large people came the only channels for legal migration from movements, developing gradually as a major North Africa. The change in migration policy economic phenomenon in the 1940’s following not only impacted the nature of labor migration, the discovery of oil. By the 1970’s, labor migra- it also impacted the geographic destination. tion from MENA—to destinations both within Because most legal channels for economic mi- the region and externally—was among the most gration to Europe had been closed, other forms extensive in the world. of migration took over. Since the mid-1970s, predominantly Maghreb migrants have instead North Africa’s migration corridors have entered the European Union through family been shaped by geography, cultural, colonial reunification, family formation, and asylum, and and historical ties, trade, conflict, and migra- increasingly as irregular migrants. In France, tion policy. One corridor comprised the draw the average entry of Maghreb migrants fell from of workers from Egypt (along with the Levant, around 70,000 per year in the mid-1970s, to less West Bank and Gaza, and Yemen), who primarily than half that amount in the 1980s. In addition, migrated to oil-rich Gulf countries, as the need Maghreb North African migration has increas- for both skilled and unskilled labor swelled with ing shifted from Northern Europe to Southern oil production. The other major corridor com- European countries. Italy and Spain in particular prised the labor migrants from North Africa’s have become entry countries for North African Maghreb countries—Morocco, Algeria, and Tu- nisia—who overwhelmingly migrated to Western 5 The MENA region, for the purposes of this report, comprises Europe. The main European receiving countries, Bahrain, Djibouti, Egypt, Iran, Iraq, Jordan, Kuwait, Lebanon, France, Germany, Belgium, and the Netherlands, Libya, Malta, Morocco, Oman, Qatar, Saudi Arabia, Syria, Tunisia, the United Arab Emirates, West Bank and Gaza, and Yemen. needed low-skilled labor during their post-war Because of data limitations, the study of North Africa, for the reconstruction industrial boom. Many migrants purposes of this report, is generally limited to Morocco, Tunisia, were recruited from outside Europe through Algeria, and Egypt. 1   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options migrants, as their demand for low-skilled labor tries in the world in 2005, three were North has increased with a sustained economic boom. African6 (Morocco, with 2.7 million emigrants, Their long coastlines along the Mediterranean amounting for about 9 percent of its total popula- have also provided easier access to Europe for tion; Egypt, with 2.4 million, accounting for more undocumented migrants from North Africa and than 3 percent of its population; and Algeria, with Sub-Saharan Africa (SSA), as well as for Asian 1.8 million, and accounting for 5.4 percent of its migrants who use the North African countries as population.) Even Tunisia’s migration is higher transit ports on their way to Europe. than world averages, with more than 620,000 migrants, accounting for more than 6 percent of Changing migration policies and circum- the population (Figure 1.1). At the same time, stances also impacted Egyptian migration. Migra- with many countries of the world experiencing tion from Egypt also started to slow by the early out-migration rates far in excess of those of North 1980s, with the Iran-Iraq war causing oil revenues Africa, it is also clear that the migration is not a to decrease and forcing many Egyptian migrants phenomenon unique to this region. in Iraq to return home. In addition, continuous declines in oil prices starting from the late 1970s Perhaps most importantly, migration has along with new policies towards hiring nation- served as a critical outlet to the region’s disap- als to substitute for foreign workers in the Gulf pointing labor market conditions at home. For reduced Egyptian migration flows. more than two decades, the region has struggled with visibly poor labor market outcomes, with Yet, despite more restrictive migration policy both declining real wages and high unemploy- over the past two decades, the scope of migra- tion from the combined North African countries (Morocco, Tunisia, Algeria, and Egypt) remains 6 http://siteresources.worldbank.org/INTPROSPECTS/Re- impressive. Of the top-thirty emigration coun- sources/334934–1199807908806/Top10.pdf. Figure 1.1: Middle East and North Africa Region Emigration 2005 Number of Emigrants As percent of population 3,000 30 2,500 25 As percent of population Emigration (thousands) 2,000 20 1,500 15 1,000 10 500 5 0 0 Algeria Egypt, Arab Rep. Morocco Tunisia Iran Lebanon Libya Iraq Jordan Syria WBG Yemen Saudi Arabia Bahrain Kuwait UAE Source: World Bank, Development Economics Group data. 2 Chapter 1: A Heightened Urgency for Understanding International Migration ment, which first emerged following the oil price Three Forces Heightening the collapses in the late 1970s and early-1980s. North Interest in North African Migration African countries, along with the broader MENA region, pursued models of development that were The urgency for better understanding the dy- heavily state-led, with nationalization of private namics of migration—its determinants, its im- assets, state planning of economic activities, and pacts, and the effect that policies have on migra- industrial development through protected local tion—has heightened. While migration research markets. Wealth was redistributed through pub- is not new, it was relegated to a relatively minor lic expenditures directed at social development role in both research and in country policy, even and large-scale public sector employment. With while it was a sizable and growing phenomenon significant oil-related wealth and revenues, this (not only for the North African region, but world- development model was able to ensure near full wide). But a greater pressure has developed in employment from the 1960s through the early recent years to better understand how migration 1980s, with the public sector acting as the em- fits into the broader development context. As a ployer of last resort for the labor force. result, migration is capturing an increasing focus from development practitioners and research- When oil prices collapsed, and with a rapidly ers, including the United Nations Development expanding labor force, the ability for the pub- Programme (UNDP), which made migration the lic sector to absorb laborers was significantly theme of its Human Development Report in 2009. diminished, and unemployment began to rise to unprecedented levels. Over the 1990s, labor This increased focus reflects, in part, the force growth in North Africa averaged close to 3 growing the internationalization of the produc- percent a year.7 By the year 2000, unemployment tion of goods and services, reduced commu- in North Africa averaged more than 17 percent nications and transportation costs, and rising of the labor force, accentuating a growing differ- demographic imbalances worldwide, which ence in economic opportunities between North Africa and migration destinations, and increasing the push for work abroad. 7 Includes Morocco, Tunisia, Algeria, Egypt, and Libya. Figure 1.2: Rise in unemployment in North Africa from 1990 to 2000 1990 2000 35 30 25 20 Percent 15 10 5 0 Morocco* Tunisia Algeria Egypt North Africa Source: World Bank data from country sources. * Urban areas only. 3   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options together imply increased immigration flows over likewise expected to have seen a downturn in the coming decades. Also reflected by this surge remittances, estimated at over 10 percent be- in research is a growing awareness that misman- tween 2008–2009. aged migration can be costly to both sending and receiving countries across a multitude of The impact of this decline in remittances in policy areas, including education, integration, North Africa may be large, due not only to the size social protection and financial services. More of the decline, but also to the magnitude of the immediately, an important factor intensifying region’s dependence on them for the economy. the research on migration is the recognition that Not only have North African economies been the global financial crisis will inevitably create hard hit in terms of the projected percent decline significant impacts on migrant labor flows and in remittances (particularly for Morocco), the remittances. region’s dependence on remittances is greater than many other regions, which heightens the A global economic downturn with potential impact of this shock on the regional strong effects on migration economy. On a per capita basis, as well as a share of GDP, the dependence on remittances in North A major impetus for the imperative to better Africa is greater than in any other region of the understand migration is the global financial world. As a result, the interest in better under- crisis, which has spotlighted the dependence standing what this will mean for North African of many economies on migration and worker economies has risen. remittances, and the vulnerability to a sudden decline in both. Worldwide, migration and remit- How North African countries are impacted tances are estimated to have declined sharply by this remittance downturn will depend upon over 2009 and 2010 as a result of deterioration how remittances work through the economy. in the economic and employment situation in Sending regions experience diverse outcomes the migrant-destination countries worldwide. from the impacts of migration and remittances World Bank projections suggest that remittances on macroeconomic balances, on poverty, inequal- worldwide declined by 6 percent over 2009, with ity, investments (both physical and human), regions like Europe and Central Asia particularly institution building, labor market outcomes, affected (Table 1). North African economies are and overall development. The migration shock Table 1.1: Projected workers’ remittances, 2008–2011 (credit, $U.S., billions) Country/Region 2008 2009 2008–2009 change Algeria 2.20 2.19 –0.4% Egypt 8.7 7.8 –10.3% Morocco 6.9 5.7 –17.0% Tunisia 1.9 1.9 –0.5% East Asia Pacific 86.1 84.8 –1.5% Europe Central Asia 57.8 49.3 –14.7% Latin America Caribbean 64.7 58.5 –9.6% Middle East North Africa 34.7 32.2 –7.2% North Africa 4 19.7 17.6 –10.6% South Asia 73.3 72.0 –1.8% Sub-Saharan Africa 21.1 20.5 –1.8% World 337.8 317.2 –6.1% Source: Dilip Ratha et al. 2009. Migration and Development: Brief 11, November 3. 4 Chapter 1: A Heightened Urgency for Understanding International Migration Table 1.2: Remittances per capita and as a share of GDP, 2007 Remittances Country/region ($U.S. billions) 2007 As percent of GDP Per capita North Africa 18.2 4.8 121 Algeria 2.1 1.6 63 Egypt 7.7 5.9 101 Morocco 6.7 9.0 218 Tunisia 1.7 4.9 168 East Asia Pacific 65.3 1.5 34 Europe Central Asia 50.4 1.6 113 Latin America Caribbean 63.1 1.7 113 Middle East North Africa 31.7 3.7 101 High Income OECD 85.7 0.2 89 South Asia 52.1 3.6 34 Sub-Saharan Africa 18.6 2.2 23 Source: World Development Indicators, 2009. emanating from the global economic downturn world, and even some developing regions such has highlighted the lack of understanding of how as East Asian (driven by China) expect to see these migration decisions are made, the impacts strong increases in the number of old, but with they have on sending country development, and fewer working age persons to support them. how policy actions could influence migration In regions like MENA and Sub-Saharan Africa, decisions outcomes. meanwhile, the demographic transition will imply working age persons will comprise a larger share Demographic forces increasing of the population over the next 30 years, and imperative for more well-managed significantly larger than that which will exist in migration industrialized regions. Dramatic demographic forces also heighten the The economic effects of this demographic need to better understand the complex picture discrepancy are enormous. In regions with rapid- of North African migration. In the course of the ly-aging populations, including Europe, changes next 30–40 years, the world population will expe- in the demographic structure will strongly af- rience major changes. Europe will age consider- fect the population’s consumption patterns, ably, while in low-income countries the largest with significantly higher expenditures devoted cohort of young people of all time will enter the to health, long-term care, housing and energy, labor market. Although the world population as and with considerably lower overall savings and a whole will only gradually start to age, consider- investment. As a result, capital per worker will able regional discrepancies in fertility will result decline (with induced effects on interest rates in relatively older regions co-existing next to and wages), resulting in lower GDP per capita strongly expanding regions. growth, unless declining labor supply is compen- sated with higher productivity growth—unlikely The countries of North Africa will experi- with an older, less entrepreneurial population. ence a strong rise in the working age population. Table 1.3 shows the expected change in the world In addition, there will be strong impact on population between 2010–2040, by age category. public expenditures. For governments that have Europe and other industrialized regions of the a role in providing health care and social security, 5   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Table 1.3: Expected changes in world population by age group, 2010–2040 Population in 2010, millions Population in 2040, millions (and share of total population) (and share of total population) Regions Under 15 Age 15–65 Over 65 Under 15 Age 15–65 Over 65 World Western Europe 29,684 124,194 34,709 27,414 107,269 53,940 (15.7) (65.9) (18.4) (14.5) (56.9) (28.6) USA and Canada 69,756 235,835 45,941 73,760 265,610 91,995 (19.8) (67.1) (13.1) (17.1) (61.6) (21.3) East Asia and Pacific 309,350 1,137,704 152,736 255,651 1,056,722 384,933 (19.3) (71.1) (9.5) (15.1) (62.3) (22.7) Latin America 162,764 385,130 40,755 132,549 472,049 113,436 Caribbean (27.7) (65.4) (6.9) (18.0) (64.0) (18.1) Sub-Saharan Africa 365,589 470,544 27,182 489,361 976,649 70,453 (42.3) (54.5) (3.1) (31.8) (63.6) (4.6) MENA (all) 115,708 245,232 15,924 119,989 373,270 55,393 (30.7) (65.1) (4.2) (21.9) (68.0) (10.1) Morocco /Algeria / 48,157 106,494 8,001 47,077 151,055 24,907 Tunisia / Egypt (29.6) (65.5) (4.9) (21.1) (67.7) (11.2) Source: Docquier and Marchiori. 2009, an output of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. the demographic transition will present them will migration for the benefit of both the sending and hard choices between higher taxes (including the receiving region: the receiving region benefits a possible reweighting of taxes from earnings from easing the negative consequences of aging to consumption), increasing debt (which could by keeping its labor force younger and improving jeopardize the welfare of future generations) its old-age dependency ratios; and the sending or reducing the government’s role in providing region benefits by decreasing its current youth health care and social security. bulge, easing the pressure on the labor market, and reducing the potential for a future aging The challenge for high population growth crisis which is certain to come once the current countries will be job creation. High working-age large youth cohorts retire in 50 years from now. population growth regions such as in North Af- rica, meanwhile, face the challenges of providing Recognition that the costs of not sufficient employment opportunities for a labor understanding migration are high force bulge driven by both demographics as well as increasing participation of females in the labor Avoiding adverse impacts from migration is also force. These new workers will have significantly an impetus for change. But perhaps the greatest higher levels of education than previous gen- impetus for better understanding the dynamics erations, with increasing expectations for good of North African to Europe migration is the in- employment.. creasing realization that migration—unmanaged, misunderstood, and unengaged as an important Demographic arbitrage, in the form of migra- element of development—can have significant tion, has the capacity to support both rapidly ag- costs for both sending and receiving regions. ing and high-population-growth regions. The un- Developing countries, in particular, have a lot to avoidable and impending demographic imbalanc- gain from migration, in terms of growth, poverty es facing both regions also creates opportunities reduction, human capital accumulation, and for demographic arbitrage through inter-regional investment, but achieving these gains requires 6 Chapter 1: A Heightened Urgency for Understanding International Migration coherent and conducive migration and develop- permanent immigration, especially of the un- ment policies. skilled. Poor and inefficient education policies, combined with lack of pre-departure training “Brain drain� is one potential result of badly programs and job counseling for potential mi- managed migration. Without an appropriate grants, may accentuate skill mismatch and make policy framework, skilled migration can result the integration into foreign labor markets difficult in brain drain and a related loss of public invest- ments in education and training of skilled labor, This report can only fill a part of the knowl- with no positive feedback for the sending econo- edge gap underlying these three imperatives. mies. Previous macroeconomic studies suggest The global financial crisis is a major factor that skilled migration rates exceeding a threshold spurring migrant-sending regions (and migrant- of 10–15 percent of the skilled labor force begin receiving regions) to better understand how to exert harmful impacts on development (Doc- implicitly dependent they are upon migration. quier, 2009). High-skilled migration rates from But understanding how they may be impacted North African economies are already at those lev- by migration-related shocks, and how to re- els, ranging from 8–19 percent of the high-skilled spond to them, requires the kind of knowledge population. With demographic forces creating stock that is only initiated in this volume (and increasing pull factors for migration from North whose continuation and extension is required Africa, and without accompanying policies to over the long term). The demographic realities boost higher education, the continued exodus worldwide make intensified migration over the of the region’s best and brightest laborers could next 10–30 years axiomatic, but understanding have critical implications on long-term growth. how to best prepare for it requires the type of migration modeling only touched upon in this The potential costs—both to MENA and to volume (and whose sophisticated continuation is Europe—for not getting the migration policy required over the long term). Migration impacts framework right are myriad. Restrictive labor and determinants that are not understood could migration policies may increase the scope for il- put both North Africa and Europe at risk to lose legal migration, creating a fiscal burden on social out on the “win-win� scenarios that are sought services, with migrants accessing some social after by all involved. This volume does not come protection services (for example, health care or close to laying out a blueprint for better migration education services) without contributing to in- outcomes, but it provides a small foundation of come taxes.8 Lack of portability of social security the work that is needed to create and improve services can reduce circular migration, which that blueprint. can significantly reduce the overall development benefits of migration for sending countries—and also reduces the benefits to receiving countries, who stand to gain with circular migration through 8 In addition, irregular migration is often seen as a risk factor for criminal activities, labor exploitation, and human rights abuse, both meeting their labor market needs but also while contributing to xenophobia and social instability in the avoiding the political controversies surrounding receiving countries. 7 Chapter 2 Trends and Characteristics about North African Migration Data Limitations for Understanding only account for stocks of migrants in receiving North African Migration countries. In addition to overlapping and chang- ing categories of migrants, few statistics measure At the onset, it must be acknowledged that migration flows, and circular migrants are not the study of migration from North Africa is sig- necessarily accounted for. Data are even scarcer nificantly handicapped by data limitations, as when it comes to transit or irregular migration. confirmed by nearly all studies and researchers. These difficulties permeate all aspects of the sub- While North Africa countries (or the MENA ject, including migrant stocks and characteristics, region more generally) are not unique in regard migrant flows, remittances, and others.9 to limited data and study of migration, the region is considerably more data-challenged, with both The absence of migration data is not limited data gaps across almost all areas of economic to North Africa. Worldwide, there is a large gap and social information, and importantly, limited between the growing incidence of migration and public access for research and analysis. the availability of reliable, comparable data from which to conduct empirical analysis. A major con- In part because of data limitations, this report straint has been the lack of harmonized interna- has made use of a wide array of available data tional data on migration stocks and flows by coun- sources and methodologies to better understand try of origin and education level. At the most basic the nature of migration, its impacts on the North level, there is no such thing as an agreed-upon African region (as well as on host countries), and definition of international migration. International potential issues going forward. None are perfect, migrants are alternatively defined according to but combined, all allow for a better understand- two main criteria: the country of birth criterion, ing of North Africa’s migration challenges. used by the United Nations population database,10 and according to which a migrant is a resident Data on stocks of migrants worldwide (from living in a country different from his country of 1990 and 2000) has provided a baseline picture birth; and the country of citizenship criterion, of the extent of international migration from which defines a migrant as a resident that does the North Africa region and trends, as well as not have the citizenship of the country he lives in. important information about the education levels Both definitions are flawed: the country of birth and gender of migrant stocks. Our baseline pic- definition implies that there is no additivity of ture of migration is supplemented with country migrants, and the country of citizenship criterion information from the balance of payments on can be misleading when considering the impact of migration on a receiving country. 9 World Bank. 2009. Shaping the Future: A Long Term Per- spective of People and Job Mobility for the Middle East and But equally important, there are no reliable North Africa.. and systematically computed data, and most data 10 http://esa.un.org/migration 9   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options remittances, although (as noted later in this North African Migrant Destinations chapter) remittance data also suffer from limi- and Nature of Migration tations. Household survey data in Morocco and Egypt permits empirical analysis of the impact The most complete picture of North African mi- of remittances on various aspects of welfare gration comes from 2000, relying on a data set and economic outcomes including poverty, described in Parsons et al. (2007), consisting of labor supply decisions, and decisions affecting a 226x226 matrix of origin-destination stocks by children (including health, education, and child country and territories. The data are generated work). Those results are highlighted in Chapter by disaggregating the information on migrant 3. Specialized surveys of return migrants in three stock in each destination country or economy, as Mahgreb countries allow for a glimpse into the given in its census. The reference period is the conditions that have facilitated the reintegration 2000 round of population censuses, so the data of migrants upon return. Several sources of host- do not refer to precisely the same time period, country data (including immigration and social and makes our picture of North African migra- protection laws and arrangements, labor force tion quite dated. Moreover, the data set provides surveys, and social security records) allow for a stocks of migrants, rather than flows, a serious better understanding of the institutional frame- limitation for migration analysis. Thus, one of the work for migration, as well as the experience of most important areas of interest to researchers: North African migrants abroad (including their “What has happened to migration over the past access to labor markets, education, housing, decade?� is not possible with the current data. and the like). And finally, data on stocks of mi- grants worldwide (by age and education), along With a resident population amounting to 135 with demographic projections prepared by the million, and 7.4 million emigrants living abroad Population Division of the Department of Eco- in 2000, the emigration “rate� for North Africa nomic and Social Affairs of the United Nations is about 5.5 percent, almost double the world Secretariat, and information on tax-transfer pro- emigration rate. Morocco’s migration rate is the files of migrants and natives by education level highest, at more than 9 percent of the population, provide the foundation for examining potential but all four countries have higher than average consequences of migration over time, using a rates of emigration. computable general equilibrium model. Both the destination of migrants and the It is important to note, however, that this nature of migration differ for Maghreb countries report has also drawn upon the vast empirical and Egypt. The migrants from the Maghreb work from other regions, as well as migration countries are often unskilled workers searching theory, to frame the issues of North African for job opportunities that allow them to settle migration where direct empirical analysis is im- permanently in Europe. Algerian and Tunisian possible. While the migration experience differs migrants are highly concentrated into a few des- from country to country, important lessons can tination countries (with more than 60 percent of be drawn through the many commonalities that their migrants located in France), but emigrants do exist. from Morocco are more geographically dispersed (Table 2.1). But all of the Maghreb migrants are Only parts of the potential costs and benefits concentrated in Europe, with France the main of migration could be addressed in this research destination of emigrants from Algeria, Morocco program. The analysis represents only a first step and Tunisia. in a long road required for the North Africa and broader MENA region in elevating migration data In contrast, Egyptian migrants (and migrants collection and analysis—and indeed all statistics from other Mashreq countries, including Lebanon, collection, analysis, access and public dissemina- Syria, and Jordan) are often of skilled and are tion—to the position it needs to occupy for good more likely to be temporary migrants in other Arab economic policy making. countries. Though this group of countries primar- 10 Chapter 2: Trends and Characteristics about North African Migration Table 2.1: Emigration from North African countries in 2000   Total migration Main destination Emigration Herfindahl   Stock Rate index Country Stock Share North Africa 4 7,441,150 5.5% Algeria 2,070,840 6.8% 0.428 France 1,333,587 64.4% Egypt 2,173,711 3.2% 0.232 Saudi Arabia 1,015,124 46.7% Morocco 2,589,108 9.3% 0.131 France 759,011 29.3% Tunisia 607,491 6.4% 0.373 France 364,498 60.0% Source: Parsons et al (2007) ily sends migrants to other Arab countries, it also Maghreb North Africa increased in absolute terms, exports workers to Western destinations (such as while migrants, as a proportion of the domestic the United States, Europe, and Australia), but its labor force, remained relatively constant at ap- export of workers to Western destinations tends to proximately 8.9 percent.11 While a comprehensive be permanent in nature. The primary destination analysis of migrant stocks is not available past of Egyptian migrants in 2000 was Saudi Arabia. 2000, it is likely that the flow of emigrants from North African countries has increased further over More than half of North Africa’s migrants the intervening 10-year period. On the push side, are in Europe. The regional destination of North the North Africa region is facing an unprecedented Africa’s emigrants is provided in Table 2.2. Emi- challenge in labor markets. High unemployment gration destination is distinguished by members persists in North Africa, despite significant reduc- of the OECD, including the EU15, North America, tions in unemployment, which have taken place Oceania (i.e. New Zealand and Australia), the in the context of rising oil prices and growth EU27 and other MENA countries (including GCC over the 2000s.12 High population growth over members). Migration to the OECD (South-North the 1970s and 1980s, combined with increasing migration) is dominant in Algeria, Morocco and participation rates (particularly among women), Tunisia. In all these cases, the share of EU27 host countries is important. Although Egypt’s migra- 11 Maghreb migrants in OECD economies as a percent of the tion rate to the EU27 is low, it is high in absolute Maghreb labor force. terms, with about 200,000 migrants in the EU27. 12 In Algeria, for example, the official unemployment rate was halved from 2000 to 2005 (from close to 30 percent to 15 per- cent), and was in large part supported through oil revenues Migration flows have increased since 1990. (public works and construction associated with Economic Over the period 1990–2000, migration from Recovery Program). Table 2.2: Location of North African emigrants in 2000 OECD EU15 NAM PAC EU27 MENA GCC North Africa 4 Algeria 81.0% 79.0% 1.8% 0.1% 79.1% 9.2% 0.9% Egypt 17.8% 8.7% 7.4% 1.6% 8.9% 75.8% 51.6% Morocco 74.9% 71.9% 2.8% 0.1% 71.9% 16.5% 1.7% Tunisia 77.7% 75.0% 2.3% 0.1% 75.1% 12.8% 2.6% Source: Parsons et al (2007) Legend: NAM = U.S.+Canada; PAC=Australia+New Zealand 11   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options are now contributing to a labor force growing by Remittances some 2.8 percent a year.13 The largest increases in the labor force have come from those with ter- North Africa’s large population movements have tiary educations, who face higher unemployment in turn generated sizeable transfers of money. As rates than average, and who are younger, with a region, MENA receives about 10 percent of the significantly greater access to information and world’s remittance flows, with North Africa ac- higher expectations for employment (in addition counting for a large portion. MENA’s remittances to better potential for migration, both financially surpass other external flows such as ODA or FDI, and legally). In addition, environmental factors making remittances the most important source of may also contribute to increased pressures to external financing within the MENA region. This emigrate in North Africa, as all four countries are is particularly true in North African economies, already worse off than the severe stress threshold where remittances far surpass other financial flows. in terms of renewable freshwater resources per capita. This can only get worse with future, albeit Official figures may underestimate the level modest, population growth. of remittance monies returning to North Africa. Official remittances only include those remit- Estimates of the stock of Egyptian workers tances that come back through official banking abroad confirm an increasing migration trend channels. Many migrants, who do not trust banks, over this decade. According to CAPMAS there remit their money back home through informal were approximately four million Egyptians living and unrecorded means. abroad, which represents about four percent of the total population of Egypt. According to the CAPMAS figures, there has been an almost 79 13 Includes Morocco, Tunisia, Libya, Algeria and Egypt. percent increase in the number of Egyptians liv- 14 Different sources often provide quite distinct migration figures. According to the 1996 Census the number of Egyptian ing abroad in 2006 compared to the 1996 Census migrants abroad was about 2.18 million, while the ILM provides figures of 2.2 million (Figure 2.1).14 an estimate of 2.715 million for the same year. Figure 2.1: Estimates of the Stock of Egyptian migrants abroad, 1970–2006 CAPMAS ILM Data Base 5 4 Millions of Migrants 3 2 1 0 1960 1970 1980 1990 2000 2010 Year Sources: CAPMAS Population Censuses and ILO, International Labor Migration Database (ILM). From Source: Roushdy, Assaad, and Rashed 2009, an output of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. 12 Chapter 2: Trends and Characteristics about North African Migration Table 2.3: Inward Remittance Flows, Per Capita Remittances, National ODA and FDI, 2007 Inward Remittance Per Capita Remittances as Flows (U.S.$ Remittances a Percentage of ODA (U.S.$ FDI (U.S.$ million, 2007) (U.S.$, 2007) GDP in 2006 million, 2006) million, 2007) Algeria 2,906 85.9 2.2 209 1,664 Egypt, Arab Rep. 5,865 77.7 5.0 873 7,620 Morocco 5,700 185.0 9.5 1,046 3,800 Tunisia 1,669 163.6 5.0 432 1,620 Source: World Bank, Migration and Remittances Factbook 2008, 2008; World Bank Data and Statistics Website. The importance of remittances in North international competitiveness, which in turn can Africa is highest in Morocco, where inward re- lead to a decline in production and manufacture mittance flows were five times higher than ODA, of other tradeable goods. Large remittance in- and more than one and half times the size of FDI flows can cause an increase in inflation, which in 2007. Algeria’s remittances constitute a very can result in an increase in poverty. Remittances small part of GDP, yet were still higher than ODA may slow down growth by reducing the work ef- and FDI combined. The scope and long-lasting forts of remittance recipients. The impact may history of out-migration and remittances have be particularly negative if the lost labor is con- made them a structural feature of many econo- centrated in particularly important or productive mies in the region. sectors to the economy. Moreover, remittances may not generate those often-referred-to multi- At the macroeconomic level, there are nu- plier effects in terms of income and employment merous benefits that North Africa may realize creation if they are used primarily for family from this important source of external finance. consumption and investments in real estate. Re- By insuring against income shocks, remittances mittances may also lead to increased inequality, may allow families, especially poor ones, to depending upon who remits. smooth consumption and better meet basic needs. Remittances may open up opportunities The Skills Composition of North for families to invest in their children’s education African Migrants and healthcare, as well as to invest in small busi- nesses or to build savings (savings which may A newly developed dataset allows for an analysis provide a resource base for retirement income). of migrants worldwide by their skill level. Until Remittances can lead to poverty alleviation recently, such a dataset did not exist. While the through their direct impact on the incomes of Parsons et al. database can provide a good picture the poor. Remittances may generate multiplier of overall migration flows, it does not distinguish effects—especially in poor countries with high between high-skill and low-skill migration or the unemployment—through expenditures linkages, gender dimension.15 A recently developed dataset if some of these funds are spent on domesti- by Docquier, Lowell and Marfouk (DLM), is based cally produced goods or services. There are also on the aggregation of harmonized immigration potential dynamic impacts of remittances. Over data collected in OECD host countries, where time, they may improve a country’s creditworthi- information about the birth country, gender, age ness and thereby enhance its access to interna- tional capital markets. 15 The need for better migration data has been strongly em- At the same time, the impact of remittances phasized by international agencies and researchers. A regional conference, held in Addis Ababa in 2000, at the initiative of on North Africa’s development is not guaranteed. the UNECA, IOM, and IDRC, highlighted the lack of adequate The large inflows of remittances can cause real data as a major problem for monitoring the scope and impact exchange rate appreciation as well as loss of of brain drain in Africa. 13   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options and educational attainment of immigrants is avail- North African migrants are predominantly able (see Box 2.1). Nonetheless, for the purpose of low-skilled. However, the skill composition of understanding skilled emigration (particularly from North Africa’s migrants differs considerably for North Africa) OECD-based databases have some the Maghreb countries (Morocco, Tunisia, and shortcomings. First, as shown in tables 2.1 and Algeria) and for Egypt. Maghreb migrants are 2.2, most of Egypt’s emigration is directed not to predominantly low-skilled workers, with more the OECD but to the GCC. Thus, the DLM data are than 70 percent of migrants having primary likely to provide a reasonable picture of skilled emi- education or less. About 15 percent of migrants gration from the Maghreb, but there is much more have completed secondary education, and only uncertainty as to Egypt. To mitigate this limitation, about 15 percent of migrants are what would be the DLM data was extended to include estimates of termed high-skilled, with tertiary education or the immigration structure for the GCC.16 more. In contrast, while the majority of Egyptian migrants [55 percent] are also low-skilled, almost Still, there exist some limitations with the 30 percent have higher education (Figure 2.2). extended DLM dataset. Because census data is only collected with low frequency, the latest At the same time, North Africa exhibits a high point in time is 2000. Thus, very recent changes, degree of high-skilled emigration. When North for example, in reaction to policy changes in re- Africa’s skilled migrant stocks are contrasted ceiving countries, cannot be traced. Third, it is with domestic stocks of higher educated labor- not able to distinguish between those who were ers, all four countries demonstrate relatively educated abroad and those who received their high-skilled emigration rates (often alternatively, education at home—an issue with implications for public spending on education and public re- 16 For Saudi Arabia, the DLM database was extended through the turns to investment. Finally, when considering use of labor force survey data on the age and education level of the emigration of skills, it must be remembered guest workers in 1990 and 2000. In the remaining GCC countries that “education� and “years of schooling� is an (Bahrain, Kuwait, Oman, Qatar and United Arab Emirates), the imperfect measure of skills and ability. This is, of DLM data starts from Parsons et al. bilateral stocks and applies the age and educational structure observed in Saudi Arabia, course, a generic problem in studies looking at giving a reasonable estimates of the educational composition the role of human capital in the economy. of migrants to GCC nations. Box 2.1: Notes on the DLM database on migration by education level The database includes data, typically drawn from cen- of people born in the home country and belonging to suses from the 30 OECD member states for two periods the same educational category. Calculating the skilled (1990 and 2000), with the highest level of detail on coun- emigration rate as a proportion of the total educated try of birth and level of educational attainment, and age labor force is more appropriate to evaluate the pres- and gender. The database concentrates on individuals sure imposed on the local labor market. aged 25+ (when education is assumed to be completed). Selection rates, in contrast, represent the share Three levels of schooling are identified: low-skilled: of skilled migrants in total emigration stocks. It is including lower-secondary, primary and no school; quite possible for a country to have comparatively medium skilled for those with high school diplomas high-skilled emigration rates but low selection rates— and high-skilled for those with tertiary education. Morocco is an example—if the human capital base is Higher skilled emigration rates are obtained by weak. comparing the emigration stocks to the total number Source: The World Bank, 2009. “Skilled Migration from the MENA Region: Trends, Impacts and Policy Responses.� An output of the EC-Funded World Bank program of International Migration from the Middle East and North Africa and Poverty Reduction Strategies. 14 Chapter 2: Trends and Characteristics about North African Migration Figure 2.2: Skills composition of North African migrants, 2000 Up to Primary Secondary Tertiary + Egypt Algeria Tunisia Morocco 0% 20% 40% 60% 80% 100% Sources: Calculations from DLM database 2007. and incorrectly, referred to as “brain drain�). Yet, from the sending country perspective, Table 2.4 highlights the skilled emigration rates emigration still involves a significant outflow of (the percent of the domestic stock of high- human capital because of a weak skill base in the skilled labor who emigrate) for the period 2000. sending country. Relatively high in all four countries (with skilled emigration rates significantly exceeding overall The reasons for departure of the highly emigration rates, from Table 2.1), the issue of educated are not only high wage differentials skilled emigration is of particular relevance to between Maghreb and Europe, but also general Morocco, with almost 20 percent of its skilled labor market and social conditions, such as high workforce living abroad. This can be contrasted unemployment rates among the highly skilled, to a level of 9 percent for the entire MENA region, lack of career opportunity, etc. 11 percent for the Latin American and Caribbean countries, and 12 percent for Sub-Saharan Africa Skilled emigration from North Africa has as a whole. Thus, from the receiving country grown significantly between 1990 and 2000—by perspective, Maghreb countries (and to a lesser 90 percent for the North African region overall, extent, Egypt) offer largely low/medium skills. and by 45 percent for Egypt, 64 percent for Table 2.4: Skilled emigration rates of North African countries in 2000 Total to EU27 to NA to GCC to OECD EU27% North Africa 4     Algeria 9.6% 7.1% 2.1% 0.2% 9.4% 73.7% Egypt 8.3% 0.9% 3.5% 3.8% 4.5% 11.3% Morocco 18.5% 13.3% 4.3% 0.6% 17.9% 72.0% Tunisia 12.9% 9.6% 2.3% 0.6% 12.3% 74.6% Source: Source: Calculations from DLM, 2007 15   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Tunisia, 78 percent for Morocco, and 205 per- and Heering, 2002).17 A survey of returned mi- cent for Algeria. This compares with low-skilled grants in Morocco, Algeria and Tunisia suggests emigration growth of only 21 percent for the that as late as 2007, between 10–17 percent of North African region as a whole, 10 percent for returned migrants were unemployed at the time Egypt, 4 percent for Algeria and Tunisia, and 45 of migration.18 percent for Morocco. Although skilled emigration has grown faster than low-skilled emigration over Migration has become more feminized the 1990s, skilled emigration rates have fallen throughout North Africa (save Algeria), reflect- The gender composition of North Africa migrants ing an overall increase in educational attainment has also changed. In the early times, North Af- in the individual countries. rican migrants were male and single. With the introduction of family reunification programs Recent Changes in North African (and later ongoing family formation), however, Migration the gender balance of migration has shifted, with increasing numbers of women migrating to The motivation behind migration join their spouses. As a result, the proportion of decisions has evolved over time females in the migrant stocks has steadily risen. Among OECD economies, the share of female North Africa’s first migrants, who originated migrants from Morocco increased in relation to mostly from poor rural areas, emigrated for eco- total migrants from 40 percent to 42 percent nomic reasons, seeking to improve the standard between 1990 and 2000, while in Algeria, the of living of their families (World Bank, 2006). female share in total migrants increased from 38 Emigrants included even (or mostly) those percent to 41 percent. people who were employed (Collyer, 2004). In Egypt, most of the migrants to Arab countries Women have generally migrated in the were farmers along with skilled and unskilled context of family reunification. For example, 80 workers. Evidence suggests that even migra- percent of women interviewed in a study of Mo- tion through family regroupings hides a certain roccan migrants (van der Erf and Heering, 2002) element of economic rationale, whereas mar- were married at time of migration. Although the rying into a migrant family provides a way for primary purpose of migration may have been more economic and social opportunities (World family reunification, it has served as a side door Bank, 2006). The strong correlation between for labor migration. In particular, the opening of economic growth and migration flows (officially opportunities in both the domestic help sector for family reunification/formation reasons) in and in certain lighter agriculture activities has the Netherlands over the 1980s and 1990s also attracted many female migrants, particularly to suggests that there are economically motivated Spain.19 reasons. However, although economic reasons continue to be main motivation, the current situation is 17 As regards women, 77% of them were doing household work more nuanced. For example migration is less in- before migration, compared with 66% after migration. Employ- ment rate among them grew from 2% (pre-migration) to 28% discriminate with respect to labor market status. after migration (van der Erf and Heering, 2002). Currently, much more so than in the past, it is 18 Gubert and Nordman 2008a. “Who Benefits Most from Migra- the unemployed and the inactive that migrate, tion? An Empirical Analysis Using Data on Return Migrants in the Maghreb,� an output of the EC-Funded World Bank Program thus helping to ease labor market pressures in of International Migration from Middle East and North Africa the home country. In the survey of Eurostat and Poverty Reduction Strategies. and NIDI (Netherlands Interdisciplinary Demo- 19 Moroccan female migration as well as its impact on gender roles, intra-household allocation and children’s welfare receive graphic Institute, 2000), just over a half of men more attention in a companion paper: Herrera, Dudwick and were non-employed at time of migration, and Murrugarra 2008 “Remittances, gender roles and children wel- almost all had a job after migration (van der Erf fare outcomes in Morocco.� 16 Chapter 2: Trends and Characteristics about North African Migration Figure 2.3: Growth in migrant population in OECD countries, 1990–2000 (Total migrant stock, thousands) Male Female 1,200 1,000 800 600 400 200 0 Algeria Algeria Morocco Morocco Tunisia Tunisia 1990 2000 1990 2000 1990 2000 Sources: Calculations from DLM database. Surveys in host countries find an increasing North Africa has increasingly become labor market participation rate among migrant a destination region for migrants (both women. Between 30 percent and 50 percent of permanent and transit) active Moroccan migrants in European countries are females (a 45 percent increase over the last North Africa is not only a sending region for two decades (World Bank, 2006)). migrants, it is also a receiving region, and in- migration has become increasingly important over The feminization of international migration the past decade. Since the mid 1990s, Morocco raises specific economic issues related to the has developed into a transit migration country for gendered determinants and consequences of thousands of sub-Saharan Africans. Although most migration. In particular, the migration of higher consider Morocco a country of transit on their way educated women is likely to affect sending to Spain and other destinations, many who fail to countries in a very peculiar way. Many studies enter Europe settle in Morocco on a longer-term have emphasized the role of female education in basis, rather than return to their more unstable raising labor productivity and economic growth, and substantially poorer home countries. 21 In suggesting that educational gender gaps are an 2007, the Ministry of the Interior estimated the impediment to economic development.20 Klasen flow of illegal migrants in Morocco at 15,000 (this, (1999) and Dollar and Gatti (1999) demonstrated in addition to the 60,000 legal foreign residents), that gender inequality acts as a significant con- but that flow has been as high as 36,000 in 2003.22 straint on growth in cross-country regressions, a result confirmed by Blackden et al. (2006) in the Morocco is not the only transit country case of sub-Saharan Africa. As a result, societies for migrants. An estimated 65,000–120,000 that have a preference for not investing in girls or that lose a high proportion of skilled women through emigration may experience slower 20 See Coulombe and Tremblay 2006. 21 de Haas 2005a. growth and reduced income. 22 Di Bartolomeo et al. 2009. 17   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options sub-Saharan Africans enter the Maghreb yearly, recruitment and an autonomous “unofficial� of which 70–80 percent are estimated to migrate channel emerged, parallel to the official one (de through Libya.23 Libya’s current migrant popu- Haas 2007c). The surge in informal migration lation has been estimated as high as a million came in the 1990s with the introduction of new persons, of whom some half may be illegal. Thus, destinations (Spain and Italy). The convenience North Africa itself increasingly faces the same of migrants’ networks, the sustained high de- challenges as traditional North African migration mand for flexible, cheap labor, and a large degree destination countries. of informality in sectors where North African (in particular Moroccan) migrants work make it easy Managed and official migration has to find job without formal intermediation. The become increasingly undocumented introduction of Schengen visas and the increasing restrictiveness of immigration policies in Europe Finally, the official nature of North Africa’s migra- seem to have further contributed to making mi- tion has also changed with the establishment of gration “irregular� as both demand and supply migration networks and with changes in migra- of migrant work continued to be strong, but no tion policy. Although for obvious reasons there options existed for labor migration. are no reliable data on undocumented migration, the proportions of undocumented workers are Currently the most common form of Moroc- likely to have increased, largely due to a change can migration to southern Europe is overstaying in destination. In Morocco, for example, during a tourist visa (58 percent of surveyed Moroccan the period of bilateral labor agreements of the migrants in Eurostat and NIDI, 2000), followed by 1960–70s, migration generally operated through entering a country clandestinely, without proper official channels and had a circular character. authorization or documents (13 percent). The However, as soon as networks had been estab- stock of undocumented migrants has on several lished in host countries, new migrants increas- ingly preferred to circumvent the hassle of official 23 de Haas 2006. Figure 2.4: Moroccan immigration in Spain, 2000–2007 (Total migrant stock, thousands) 800 700 600 500 400 300 200 100 0 2000 2001 2002 2003 2004 2005 2006 2007 Year Sources: Conde-Ruiz, Gonzalez, and de Silva 2008, an output of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. 18 Chapter 2: Trends and Characteristics about North African Migration occasions reached level at which host coun- importantly, there is virtually no information to try governments felt compelled to embark on link the various characteristics of the migration regularization campaigns, through which many experience, be it employment, education, wages, Moroccan migrants gained residency rights. training received, legality, entrepreneurship abroad, family, and so on, with the experience of Much about North African migration is families back home. Almost all our information unknown on the development impact of migration is based on the characteristics of the migrant family left While the statistics provide a baseline picture of behind, not the actual migrant. North African migration, the picture is far from comprehensive. There is limited information Nonetheless, with household level data avail- about the conditions of migrants prior to migra- able in two countries, Morocco and Egypt, there tion. There is no information on migrant flows. is the potential to examine several dimensions Thus, the very basic question about how North of migration and remittances, in terms of their African migrants has changed over the last de- impact on the economy. In the following chapter, cade, in terms of their education, in terms of their some of these effects that migration has had on gender, and so forth cannot be answered. And development in North Africa are explored. 19 Chapter 3 The Development Impact of North African Migration In the past, migration was only viewed as a “fail- in health and education, into businesses, or into ure of development,� reflecting a clear lack of savings which may reduce vulnerability to eco- opportunities at the country of origin. Indeed, nomic shocks. They can smooth consumption and many of the factors that increase the push for help families, especially poor ones, meet basic migration—unemployment, poverty, and political needs. However, they can also cause real ex- instability—are palpable evidence of develop- change rate appreciation and loss of international ment deficiencies at home. competitiveness, and they may result in reduced work efforts of remittance recipients. Return Today, migration has come to be concomi- migrants may make important contributions to tantly seen as a powerful mechanism for growth development through knowledge-transfer and and development. With migration opening the investment. Or migrants may return at the end of door to financial resources (in the form of remit- their working life cycle to retire, and the costs of tances), investment and knowledge transfer (by their return (economically) may be higher than returned migrants and diaspora), skills develop- their contributions. ment, and networks important for business, the potential channels for migration to aid in devel- This section presents a shutter view into opment are numerous. what is known about North African migration, and the many ways in which migration has The development impact of migration varies changed the economic and social landscape in worldwide. Understanding the potential for mi- North Africa. However, as already mentioned, gration to aid in development, a pressing question the ability to analyze the impacts of migration among development practitioners has been why in a systematic way are handicapped by a lack migration has seemingly promoted economic of appropriate data throughout the region. As development in some cases but failed to do so a result, only parts of the potential costs and in others. One of the difficulties in arriving at benefits of migration can be addressed herein. an answer to that question is the sheer number of complex and sometimes divergent channels In two countries, Morocco and Egypt, data through which migration contributes to (or de- permits the empirical analysis of the impact of tracts from) development. Migration entails the remittances or migration, or both. The analysis loss of workers and skills, at least temporarily, reveals some important findings: who might otherwise contribute to growth and development. However, if those workers were First, migration and remittances have re- unemployed or severely underemployed, their duced the level of poverty in North Africa. While economic contributions might be limited. Remit- remittances do benefit the better off (with 61 tances sent home by migrants can provide fami- percent of remittances accruing to the top 2 in- lies left behind with opportunities which would come quintiles both in Morocco and Egypt), they otherwise be unavailable, including investments nonetheless do benefit the poor. In Morocco, for 21   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options example, remittances account for more than 50 and benefits of migration. In terms of costs, the percent of the household budget of the lowest largest potential cost to the region comes from income quintile. Moreover, remittances decrease the loss of higher-educated workers needed the probability of being poor by about 7 percent- for growth and development—often (and inac- age points in Morocco and by about 9 percentage curately) referred to as the brain drain. North points in Egypt. That is to say, the presence of African countries have high rates of high-skilled remittances in the household reduced the prob- migration, but over the 1990–2000 period the ability of being poor in Morocco from around 15 rates of skilled migration dropped, not because percent to around 8 percent, while in Egypt, it of drops in the number of skilled workers go- reduced the probability from around 20 percent ing abroad, but because of a general increase to around 11 percent. in education levels. Though lack of data makes estimations of the costs of brain drain in North Remittances have also allowed for increased Africa impossible, what can be said is that the expenditures on health and education for chil- environment for the productive use of educa- dren and reduced child labor, which over the long tion in the economy is poor. That could suggest run are vital for setting the groundwork for higher that the lost earnings from the act of migration growth. In both Egypt and Morocco, migration are somewhat lower than might be suggested, and remittances lead to positive improvements if those high-skilled migrants were from the with regard to investments in children, and in ranks of the unemployed or easily replaceable Egypt, migration and remittances significantly in the domestic economy by the unemployed. reduce the probability of child work. Remit- But ascertaining this would, in the end, require tances also trigger a more equal intra household information that is not available. What is certain allocation on education and food between boys is that the region is already bearing large costs and girls, offsetting a gendered bias allocation of lost education domestically. in education and food expenses, and triggering gender parity in children’s human capital ac- A Framework for Understanding the cumulation in rural areas where girls are more Development Impact of Migration marginalized.24 At the household level, the benefits from migra- Another potential impact from migration tion are revealed through preference. Migration relates to its effects on labor markets, not only is a decision that is made at the household level, by the outflow of laborers, but also by potential made by households evaluating their individual changes in the labor market decisions of those family benefits and family costs. The benefits left behind (particularly for households which are often economic, but there are often non- receive remittances). However, the data from economic benefits as well, including family and Egypt and Morocco suggest that the effects are security. Revealed preference confirms that for modest, at best. In the case of Egypt, in fact, the those who have migrated, the expected benefits results suggest a positive impact by remittances exceed the expected costs. on labor supply, although most of the increase is the result of females moving into unpaid family The impacts on the broader economy are labor. In both countries, remittances also in- more difficult to predict. While the migration crease the probability for self-employment at the decision may consistently generate positive out- expense of wage employment. Given the nature comes at the household level, there is significant of the self-employment sector (predominantly heterogeneity across the world in terms of the low-skill), this suggests that remittances may be way migration (and remittances) impact growth, allotted as capital to more informal activities that poverty, and development for the broader provide greater employment flexibility. 24 Sadiqui (2004) shows that women in rural areas have been There is also descriptive evidence that is disadvantaged of urban modernization, For instance, their edu- suggestive (but not definitive) of associated costs cation and health access has traditionally been limited. 22 Chapter 3: The Development Impact of North African Migration economy. With the myriad transmission channels the actions or policies taken (either by origin through which migration and remittances can or destination countries) which may impact contribute to development, including through migrant decisions (Figure 3.1). income, consumption, investment, government policies, potential parental absence, and removal Migration brings different benefits and costs of potential entrepreneurial individuals from the to the economy along the different migration community,25 countries can realize substantially stages. The migrant lifecycle represents the divergent benefits (and costs) from the migration various stages of the life of the migrant, each phenomenon. of which brings different economic value to the sending country, or could result in different The economy-wide impacts of migration economic or societal costs. The first stage in can be conceptualized as the interaction of the lifecycle is at the point of migration, where three key factors: the stage of migration, the migration itself has a cost (benefit) to the characteristics of migration, and the character- economy, but it has potential benefits, with the istics of both the sending and host country. To ultimate sign based on the discounted value of understand both the ways in which migration the person’s contribution to development had can generate gains to sending countries, as he/she stayed (earnings and otherwise), less well as how countries differ in terms of these the discounted value of the societal resources gains, it is important to center the discussion the migrant would have absorbed had he/she around a basic framework of the transmission stayed, plus the discounted value of any effects channels of migration to the country of origin. that the act of migration has on the people who How migration impacts sending countries can be conceptualized as the interaction of three basic factors: the lifecycle of the migrant, the Sasin and McKenzie. 2007. “Migration Operational Vehicle: 25 individual characteristics of the migrant, and Operational Note 1; Migration, Poverty and Human Capital.� Figure 3.1: Migration’s Development Impact Prism Factors Affecting the Economy-wide Impacts of Migration Education Migration Employment status Time Abroad ics Mi Marital status ist gr Return Migration characteristics, er an ct Family characteristics, tL ara etc. ife C Migrant Ch ycle Development Impact Ec s ono m i c A c ti o n Economic and social policies, home and host countries 23   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options remain (positive or negative).26 To be more Migrant and structural characteristics deter- precise, the impacts from the first stage in the mine the magnitude of those costs and benefits. migrant life cycle can be considered the pure ef- Table 3.1 elaborates the positive and negative po- fects of migration as a result of time of absence, tential impacts that migration can have on send- and not considering subsequent contributions ing countries throughout the migration process. from the migrant, but allowing for impacts on However, these impacts are purely hypothetical. the actions of those remaining. For example, The magnitudes of these positive and negative the migration act may induce non-migrants impacts are undetermined, conditional upon to invest in education and skills accumulation both the characteristics of the migrants (and the in the expectation of future prospects abroad migration decision), as well as structural charac- (whether or not they actually migrate), which teristics in both sending and receiving countries, contributes to growth over the long run. During which both impact the effectiveness of these the second stage in the migrant lifecycle, the channels and the responses of migrants to these period of residing abroad, the migrant can add incentives. The characteristics of migrants—all economic value to the home country, primar- the characteristics which have bearing on the ily through international remittances and the contributions the migrant and/or migrant family myriad actions which have come to be grouped could make to development, including aspects of as “diaspora initiatives,� but migrants may also education, employment, the nature of migration incur costs to the home country. Some high (permanent or temporary; single or family, etc)— migrant-sending countries, for example, devote will obviously have an impact on the ultimate ben- extensive resources toward the protection efits/costs from migration to the home country. and management of their migrant community abroad. Also, there are the potential social costs of family dismemberment. Finally, there is the 26 Although the impacts on decisions made by family members or others remaining in the home country are not occurring pre- stage of return, where the migrant’s net contri- cisely at the same time of migrant exit (and may not occur for butions to society can be evaluated relative to years after), this is a convenient manipulation to separate out the net contributions had the migrant stayed. the effects that can be considered mainly the result of migration In addition, there may be costs associated with (for example, effects on the labor market conditions, effects on changed gender roles in the household) and not the result of re-integrating the migrant back into the home subsequent actions taken by migrants while abroad (such as country. those made possible through the receipt of remittances). Table 3.1: Channels of Development Impact over the Migration Lifecycle Migrant characteristics/Economic Migration Time abroad Return actions ê Channels ê Migrant characteristics: [Education Positive: Societal Positive: Remittances; Positive: Increased level, Occupation, Employment resources not consumed investment home while earnings capacity upon status and prospects]; migration while abroad; positive abroad; tourism revenues; return; knowledge characteristics [single or family, impacts on decisions knowledge transfer while transfer; investment upon expected temporary or permanent]; made by those abroad; positive impacts return; effect on growth- Family characteristics [Ages, income, remaining on decisions made by enhancing institutional schooling, etc] those remaining/receiving reforms upon return. remittances. Structural characteristics in home/ Negative: Lost economic Negative: Societal resources Negative: Societal host countries which impact present contributions while consumed while abroad resources consumed upon and future earnings of migrants (job abroad; negative (reverse remittances); return; negative impacts prospects, wages, training) and may impacts on decisions negative impacts on on decisions made by impact the decisions of migrants/ made by those decisions made by those non-migrants potential migrants remaining. remaining/receiving remittances 24 Chapter 3: The Development Impact of North African Migration An unemployed worker with poor employment worldwide). Besides a general lack of detail on prospects at home, for example, will entail fewer migrants, existing national statistical systems lost economic contributions than an employed are not designed to collect information on the worker with continued strong employment pros- situation of migrants prior to migration. This pects. A high-skilled migrant, other things equal, is a fundamental requirement for investigating will entail higher lost economic contributions to the consequences of migration for international the sending country than a lower skilled migrant. migrants and their households, communities of But also, structural characteristics in home and origin or destination, and countries of origin or host countries will impact the ultimate develop- destination. Population censuses record data on ment impact of migration. For example, the cost persons in enumerated households, which should of transferring remittances in host countries will include international migrants present in the impact the value of remittances sent home by household (however defined), at the time of the migrants, while a strong investment climate will census only. Since there is no data on their living impact the potential for investment by returned conditions prior to migrating, it is impossible to migrants. measure the consequences for the migrant or his/ her household resulting from migration. Therefore, the degree to which sending countries can benefit from migration will depend For most countries, the household survey implicitly on the migrants themselves and the is the foundation of information for monitoring incentive framework within the home country.27 and analyzing migration. Good household surveys And many of the elements of the incentive frame- have the capacity to collect extensive informa- work which would increase the positive returns tion about migrants, but they are only able to do from migration to the host country are precisely so from households in which someone remains the elements which increase growth and devel- behind who can provide information about the opment overall. High-skilled workers with skills emigrant. Thus when the entire household relevant for a private-sector-led economy not has emigrated, there is no one left to provide only increase the employment potential (and data about them (neighbors can rarely provide remittance potential) abroad; they increase the detailed, reliable data). This is a significant employment potential at home. Having job op- shortcoming, to the degree that most emigrants portunities for these higher educated workers from the country migrate as households rather at home not only reduces skilled labor flight, it than individuals (leaving a household behind), also increases the contribution of human capital and the characteristics of those who emigrate to growth and development. An investment as households are generally not similar to those climate that attracts investment from returned who emigrate as individuals (those who depart migrants also attracts the domestic investment as individuals are more likely to be young, single for economic growth and development. adults, and perhaps mostly males, as is the case of individual emigrants from the countries of The Evidence of the Development North Africa, in contrast to a much broader age- Impact of Migration sex distribution of persons in families leaving). While it is analytically convenient to view mi- The analysis of migration impact in North gration impact over various stages and through Africa is limited to Morocco and Egypt, where various channels, a host of analytical challenges nationally representative surveys permit the make measuring the true impact of migration and analysis of conditions of families with migrants remittances impossible. versus families without migrants. But data sources have limitations, discussed later in this The largest impediments to measuring impact center on the endogeneity of migration 27 The incentive framework in the home country, in addition to decisions (discussed in the following section) the incentive framework in the host country, when combined and data constraints (not only in the region, but determines the differential in opportunities. 25   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options report, which limit their ability to underpin policy net effect on GDP from the remittance decline recommendations. was about 0.5 percentage points (0.433 * 0.106) [alternatively, simply subtracting the envisaged The Impact of Migration on Growth remittance decline would reduce GDP by 0.5 percent], with the largest estimated decline in At the broadest level, there is interest in know- Morocco (1.3 percentage points of GDP). ing the impact of migration on growth and productivity in North Africa, particularly in But the other way in which to look at remit- light of declines in migration and remittances tances is not through their contribution to GDP, with the global financial crisis. With regard to but through their role in growth. If remittances migration, and with an absence of information are spent differently than other sources of fi- on migration stocks, this central question cannot nance—for example: if they are channeled into be answered. What can, however, be examined expenditures that fuel long-term growth (such is the role of remittances in output and growth as education or productive investments); if they in the region. improve country creditworthiness (increasing access to international capital); if they generate The North African region is estimated to have multiplier effects through expenditure link- experienced a 10.6 percent drop in remittances ages—then remittances may additionally raise over 2009. What effect did this have on output economic growth, although some growth effects and overall growth in North Africa? There are may not be visible for some time. two ways to evaluate the decline in remittances to the region: first to simply to look at remit- However, the overall impact of remittances tances through its role in the national accounts on growth is not a priori positive. The impact of as a component of GDP, and see what the drop remittances on an economy depends largely on in remittances meant in terms of total economic the structure of the receiving economy and its output. Table 3.2 presents the size of remittances capacity to absorb large financial inflows. And the in all four North African countries as a share academic literature on growth and remittances of GDP in 2008. Remittances accounted for an has found mixed results, with most finding that average of 4.33 percent of GDP in North Africa remittances have benefited at the microeconomic in 2008. The World Bank estimated remittances level, but have not or only marginally contributed in the region declined by 10.6 percent, with the to larger economic growth.28 largest remittance declines in Egypt and Mo- rocco (with estimated declines of 10.3 percent and 17 percent, respectively). As a result, the 28 Pradhan et al. 2004. Table 3.2: National Accounting: Impacts of Remittance Decline on GDP Expected Projected decline in GDP GDP growth Remittances as a result of Projected without Remittances as a share of remittance GDP growth remittance Country 2008 GDP 2008 GDP 2008 decline 2009 decline Algeria 2,202 178,882 1.2 0.0 2.2 2.2 Egypt 9,476 162,818 5.8 0.6 3.8 4.4 Morocco 6,730 86,329 7.8 1.3 5.0 6.3 Tunisia 1,870 40,180 4.7 0.0 3.0 3.0 North Africa 20,278 468,209 4.3 0.5 3.4 3.9 Source: World Bank, 2009, “Remittances and Growth in the Middle East and North Africa,� an output of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. 26 Chapter 3: The Development Impact of North African Migration Three reasons may explain the inconclusive remittances have a greater (positive) impact on findings in the academic literature. First, there growth in less financially developed countries, are fundamental measurement problems associ- with remittances acting as a substitute for credit ated with evaluating macroeconomic impacts of from the financial sector. In financially developed remittances. Not only is the causality difficult to economies, on the other hand, because credit is disentangle (with remittances both impacting more easily available and people need not wait for growth and also being impacted by growth), but remittances for investment purposes, the impact the actual direction of the relationship fluctu- on growth is smaller.30 ates. At the aggregate level, however, it would be difficult to disentangle these effects. More Empirical estimation of impact of generally, measurement is hindered by the lack remittances on growth of a suitable model for dealing with the complex and simultaneous channels through which remit- An analysis of the impact of remittances on tances can impact growth. growth in North Africa was conducted, using the framework developed by Ramirez and Sharma Second, large sets of accurate data on the (2008) in their study of remittances and growth allocation of remittances are scarce, making it in Latin America. In that paper, the authors es- difficult to analyze their broad impact on con- timated the impact of remittances on economic sumption, savings, and investments. Not only are growth in selected upper income and lower in- the data scarce, they are also suspect. In fact, the come Latin American and Caribbean countries, way remittances are reported to be spent may not using a recently developed panel unit root and reveal the likely difference in patterns of expen- panel cointegration tests and the Fully-Modified diture allocation with and without remittances. OLS (FMOLS) methodology. Third, the impact of remittances on a country’s The North African analysis focuses on four economy depends largely on the structure of the North African countries (Egypt, Morocco, Tuni- receiving-economy and its capacity to absorb large sia, and Algeria) with annual data for the period financial inflows. As noted by Kireyev, “There is 1980–2007. To explore the relationship between no obvious blueprint for establishing unam- remittances, financial development and growth, biguously the direction of the macroeconomic two models were estimated. First the model was impact of remittances. Their ultimate macro- estimated without the interaction of financial economic impact will depend on the structural development variables. However in the second characteristics of a particular economy, the model, remittances were allowed to interact country-specific transmission mechanisms, with one of the financial development variables. and elasticities, primarily the marginal pro- This enables the determination of the impact pensities to import, consume and invest.�29 of remittances on growth through the financial development variables. The method used for One of the more important characteristics estimation was the Panel system Generalized that can determine the ultimate impact of re- Methods of Moments regressions (GMM) to mittances is the development of the financial control for endogeneity and serial correlation, sector. A body of emerging research has evolved following Arellano and Bover (1995) and taking that looks at the role remittances play in easing into account time specific effects. A description domestic credit constraints. Given the difficulties of the methodology is included in Appendix 8. associated with borrowing and getting insurance in developing countries, particularly in rural Table 3.3 presents the estimation results. areas, migrant remittances can substitute for a The results suggest that remittance flows have lack of financial development and hence promote a positive and significant impact on economic economic growth. One recent study examining the impact of remittances in the presence of dif- 29 Kirevev 2006. fering levels of financial development found that 30 Giuliano et al. 2006. 27   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Table 3.3: Estimation results: Impact of remittances on per capita growth in North Africa Dependent variable: per Without financial development variable With financial development variable capita growth Coefficient t-stat Coefficient t-stat Lag growth (–1) –0.40 –3.19 –0.48 –2.93 Lag remittances (–1) 0.01 2.46 0.05 2.01 Capital 0.06 2.96 0.08 2.38 Labor 0.07 8.24 0.27 1.68 Credit 0.07 2.41 Remittances*credit –0.06 –2.31 Source: Adams, Keller, Mottaghi, and van den Bosch 2009, an output of of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. Note: All variables are represented in natural logs. The model includes common time dummies. growth in both models, but the introduction helping firms and individuals circumvent these of the financial sector variable significantly credit constraints. While the impact on growth in increases the impact that remittances has on North Africa is small, the degree to which remit- growth, similar to the results from Latin America. tances act as a substitute for a well-developed In particular, a one percentage point increase in financial sector highlights the fact is that the remittances as a percent of GDP, all other things lack of such well developed financial sector is being equal, increases GDP growth per capita by one of the region’s vulnerabilities, Because the 0.05 percentage points. financial sector cannot provide diversified fund- ing sources for both individuals and enterprises The sign of the coefficient on the interaction for mitigating risk and insuring vulnerability, nor term (remittances*domestic credit) is negative, can households and firms can borrow from the implying that remittances act as a substitute for financial sector to smooth the consequences of an credit in North Africa, offering response to the economic downturn, therefore remittances are needs for credit and insurance that the market having to fill a central role—that of coping with has failed to provide. And indeed, the weaker and managing shocks. In the face of shocks to is the financial sector, the greater is the growth remittances themselves, the region is particularly impact from remittances. at risk in the absence of other mechanisms for smoothing consumption and investment. As in many developing regions, North Africa’s financial sectors are generally underdeveloped, Moreover, and as will be further discussed with limited access by households and firms to in Chapter 6, growing demographic imbalances financial services. A recent World Bank report worldwide create the potential for significantly highlights the financial sector challenge:31 Few higher migration flows from the North African private sector firms in MENA even use bank region over the coming decades. Under these credit to finance investment, relying instead circumstances, the vulnerability to remittance primarily on retained earnings (Figure 3.2). shocks is likely to increase in the absence of More than 38 percent of firms in the North Africa better functioning financial markets. identify access to finance as a major constraint to business, higher than for all other regions Measuring Migration’s Impact at the but Sub-Saharan Africa. There are fewer bank Household Level branches in North Africa than in other emerging economies (per 1000 people). Because migration, relative to the population, re- mains small (in Morocco, an estimated 11 percent As a result, remittances play a relatively large role (relative to developed economies) in 31 World Bank, MEDP 2009. 28 Chapter 3: The Development Impact of North African Migration Figure 3.2: Use of Bank Credit by North African Firms 25 Proportion of Firms Using Banks to 20 Finance Investment (%) 15 10 5 0 Middle East East Asia and Europe and Latin America South Asia Sub-Saharan and North Africa the Pacific Central Asia and Caribbean Africa Source: MENA Economic Developments and Prospects 2009. Note: Regional averages unweighted of households have a migrant abroad; in Egypt, cess to basic services and the neighborhood 3.5 percent), many effects of migration cannot infrastructure. It also contains a great deal of be seen at the macroeconomic level, but will be information on the household members’ educa- evident through the decisions and outcomes of tion, employment status, time allocation, job the household. In the following section, the analy- mobility, earnings, migration and household sis from the North African region is discussed. enterprises. With regard to migration questions, each round of the Egypt Labor Market Surveys Data sources for analyzing remittance- (ELMSs) contains information on internal and receiving households in North Africa international migration history (e.g., place of birth, year leaving place of birth, and the place Household surveys with specific questions on and date of the previous two moves if different remittances are only available for two of the from the current place of residence). ELMS 98 in- North African countries: Morocco and Egypt.32 cludes only one (yes/no) question on whether the Both surveys, while very useful for understand- household receives remittances from relative(s) ing some of the characteristics of migration or living abroad. However, in ELMPS 06, a new receiving remittances in North Africa, also have module on current migrants and remittances was important limitations. added and it includes questions on whether the household receives remittances from household The nationwide Egypt Labor Market Survey (ELMS) data is one of the first true nationwide 32 Tunisia carried out a Household Consumption Survey in longitudinal surveys to be carried out in Egypt. 2005, on a sample of 13,000 households. On the household roster It attempted to track households and individu- is a question identifying the residence status of all persons, in- cluding both permanent residents and temporarily absent mem- als first interviewed in 1998 as part of the Egypt bers, but there is no specification of where the absent person is. Labor Market Survey of 1998 (ELMS 98) and The only other information related to international migration is re-interview them in 2006. The Egypt Labor a question on whether the person providing the main source of Market Panel Survey (ELMPS) 06 and ELMS economic support for the household lives with the household or elsewhere. The number of households identified as including 98 provide detailed information on household an emigrant was considered too low to be statistically useful, so housing conditions, ownership of durables, ac- the data from these questions were not analyzed or published. 29   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options members living abroad, the amount and type of and poverty is multifaceted and indeterminate. these remittances, and which household member Poverty can both create increased incentives for receives the remittances. ELMPS 06 also includes migration, but it also can hinder the ability to mi- information on the place and reason of migra- grate. Likewise, while voluntary migration almost tion for individuals who were in the household always lowers poverty at the individual level, the in 1998 but were not found in 2006 because impact of migration on families left behind can they migrated between the 1998 and 2006. be either positive or negative, with factors such Although a rich source of information on labor as lost earnings balanced against remittances the market dynamics and individual and household migrant may be able to send home. characteristics, the ELMPS was not designed to measure migration. Migration is largely network- Worldwide, most studies suggest that inter- driven, with significant clustering of migrants national remittances reduce poverty. The reason among certain communities. While the ELMS is for this finding is simple: in most developing nationally representative, the sample frame was countries the amount of remittances received not designed to obtain accurate trends of migra- from a migrant working abroad is between two tion and remittances flows that would coincide and five times the amount of money that worker with official estimates. However, to the best of could receive from wages earned working at our knowledge, the ELMPS 06 is the only recent home. Since international remittances are so national household survey that collects informa- much larger than “home earnings,� the receipt tion on incidence of international migration and of remittances serves to increase the incomes remittances. of remittance-receiving households. Since some of these remittance-receiving households Morocco’s 2000/01 ENCDM survey is a multi- are “poor� before the receipt of remittances, purpose household survey, covering 0.3 percent the receipt of remittances helps to reduce the of the population. The module on household number of households living in poverty. But it is transfers contains a question about the source of also important to note that not all migration— transfers, for which “family living abroad� is an perhaps very little of it—results in remittances. option. For many purposes, the ENDCM survey Migrants who settle abroad with their families is relatively good. For instance, total household are less likely to contribute to development consumption recorded in the survey is 253 bil- through remittances. lion dirham, compared with 299 billion dirham recorded by the National Account Statistics for There are alternative ways of looking at the the period of the survey. However, remittances in impact of remittances on poverty. Within the the survey add up to only 2.3 percent of total con- lifecycle of migration, understanding the impact sumption (or 2.8 percent, when including foreign of remittances can be looked at two different pensions) against the Balance of Payments figure ways. On the one hand, it is possible to treat of 12 percent. Even assuming that not all private remittances as a simple exogenous transfer of in- transfers from abroad “pass� through households come by migrants. When treated as an exogenous budgets (e.g., some can be directly invested in transfer, the research question becomes: How real estate or stock exchange and some may be do remittances, in total or at the margin, affect money of migrants currently residing abroad, the observed level of poverty and inequality in a who are not included in the sample) it seems that specific country? remittances in the survey are underestimated. The analysis of remittance impact in North Impacts on Poverty and Inequality Africa treats remittances as a substitute for domestic (home) earnings. When treated as One of the most important ways in which mi- a potential substitute for home earnings, the gration can impact development in sending question becomes: How does the observed level countries is through its effect on poverty alle- of poverty and inequality in a country compare viation. But the relationship between migration to a “counterfactual scenario� without migration 30 Chapter 3: The Development Impact of North African Migration and remittances but including an imputation for A profile of migrant-sending, the home earnings of migrants had those people remittance-receiving households in stayed and worked at home? This is the more North Africa33 interesting, but also more challenging economic question, and requires econometric techniques According to the surveys, 11 percent of Moroccan to compare the level of poverty in a country households received remittances in 2001, while with and without remittances. That analysis is 3.5 percent of Egyptian households received complicated by the problems of selection and remittances in 2006. endogeneity. The problem of selection concerns the issue that migrants may differ in unobserved While Europe has absorbed most of Moroc- characteristics from non-migrants, and attempts co’s migrants, the Gulf region has always been the to estimate the impacts of migration on the main destination of Egyptian migrants, followed basis of the observed outcomes of non-migrant by other Arab countries. Overall, through 2006, households will be subject to bias. Endogeneity, 46 percent of Egyptian migrants have been to meanwhile, reflects the issue that decisions on Gulf Countries, 41 percent to other Arab coun- migration, remittances, labor supply, expendi- tries and only 3 percent have been to the rest of ture allocations, school attendance, and other household choices are made simultaneously, and hence the characteristics which explain migra- 33 The following two sections are drawn from Sasin 2008, and Roushdy, Assaad and Rashed 2008, outputs of the EC-Funded tion may also shape other household decisions World Bank Program of International Migration from Middle East (see Box 3.1). and North Africa and Poverty Reduction Strategies. Box 3.1: A note on measuring migration/remittance impact Measuring the impact of migration and remittances and remittances variables correlated with unobserved is hindered by methodological problems. First, a key determinants of household poverty status. Reverse empirical problem is self-selection, which is due to the causality is a big challenge when investigating the ef- fact that the pool of either migrants or remittances- fect of migration on poverty status: for instance, people recipient households is not a random sample of the falling into poverty can systematically change their population. Migrants’ families can systematically differ migration behavior, thus causing a statistically signifi- from those of non-migrants along several observ- cant correlation to be observed between poverty status able (e.g., age, education, marital status) but also and migration, but which should not be interpreted as unobservable preferences and characteristics (e.g., a causal impact of migration on poverty. income shocks, ability, talent, risk aversion). As a To address this problem, many studies use some result, evaluating the impact of migration on develop- kind of two-stage Heckman selection model to test for ment outcomes based simply on association would selection bias in the household receipt of remittances. yield erroneous results: for example, migrant families There are econometric techniques, such as instrumen- systematically save and invest more than non-migrant tal variables (IV), which are often used as a remedy families, thus from the start they are more able to exit for both the endogeneity and selection bias problems. from poverty or to develop risk management strategies However, the IV technique requires a set of instrumen- to avoid falling into poverty. tal variables that are correlated with migration and Second, other methodological problems are fre- remittance decisions but uncorrelated with household quent, including endogeneity, reverse causality bias, living standards. Such variables are often difficult to measurement errors, and omitted variable bias. These find. In the literature, historical migration rates have factors challenge the consistency of models to evalu- been occasionally used to instrument for migration and ate the impact of migration by rendering the migration receipt of remittances, with longer lags. 31   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Figure 3.3: Distribution of Egyptian Migrants by Year of Departure and Country of Destination Saudi Arabia Kuwait Emirates Others Iraq Libya Jordan Other Gulf Countries Other Arab Countries 60 60 50 50 Parent of Egyptian Migrant Parent of Egyptian Migrant 40 40 30 30 20 20 10 10 0 0 Pre-1991 1991 1996 2001 Total Pre-1991 1991 1996 2001 Total –1995 –2000 –2005 –1995 –2000 –2005 Source: Roushdy, Assaad, and Rashed 2009, an output of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. the world. Prior to the first Gulf War, Iraq was the experiencing a small but steady increase in their main destination of Egyptian migrants followed share of Egyptian migrants. by Saudi Arabia. However, the war caused Egyp- tian migrants to return from Iraq. According to Moroccan and Egyptian remittance-receiving the ELMPS 06 sample, by 2006, there were only households differ according to many charac- about 0.7 percent of Egyptian migrants in Iraq. teristics. While remittances disproportionately In contrast, Kuwait saw an increase in its share accrue to urban households in Morocco (73 of Egyptian migrants directly after its liberation, percent of remittance-receiving households but this was followed by a steady decline thereaf- are urban, compared with 56 percent of non- ter. Recent trends show that, in 2006, more than receiving households), remittances in Egypt are 10 percent of Egyptians migrants were living in disproportionately directed to rural families (69 Kuwait. Saudi Arabia continued to be the main percent of remittance receiving families live in destination for both current and return migrants rural areas, versus 51 percent of non-receiving since the 1990s, although it experienced a decline households). Remittance-receiving households in its share during the mid 1990s, which could in both Morocco and Egypt are more likely to be due to the fall in oil prices during that period. be headed by a woman, but while in Morocco The United Arab Emirates started with a very the difference is relatively modest (28 percent small share of Egyptian migrants, but its share of remittance-receiving households are female has increased significantly over the years (from headed, versus 16 percent of non-receiving), the around 1 percent before the 1990s to more than difference in Egypt is considerable: 59 percent 8 percent in 2006). The share of the other Gulf of remittance-receiving households in Egypt are countries has always been very small (less than female headed, versus only 16 percent of non- 1 percent) throughout the period. After the Iraq receiving households. In both countries, receiv- war, Jordan and Libya continued to be the two ing households are more than usually likely to be main destinations for migrants among the Arab headed by a woman, and employment rates are non-Gulf countries. Both countries have been lower, which reflects the absence of a migrant. 32 Chapter 3: The Development Impact of North African Migration In Morocco, the household head is more likely Importantly, in both countries, remittances to be better educated, while in Egypt there is no accrue disproportionately to richer households. substantial discrepancy between the education But the magnitude of disproportion is sig- composition of household heads of remittances- nificantly higher in Morocco, where the highest receiving versus non-receiving households. quintile accounts for 61 percent of all receipts, (Table 3.4). compared with only 3 percent for the bottom Table 3.4: Characteristics of remittance-receiving households in Morocco and Egypt Egypt   Morocco Household does Household Household does Household not receive receives not receive receives Characteristics of Household remittances remittances remittances remittances Total 96.7 3.3   89 11 Household characteristics           Location           Urban 50 31   56 73 Rural 51 69   44 27 Average household size 4.5 4.5   5.9 5.5 Household wealth           Lowest Quintile — 8.1   — 3 Second Quintile — 14   — 5 Third Quintile — 16.6   — 10 Forth Quintile — 26.2   — 21 Highest Quintile — 35.2   — 61 Household head characteristics         Male 84 41   84 72 Female 16 59   16 28 Education           Illiterate 34 37   51 40 literate without any school 9 8       elementary school 11 6       middle school 5 4       Secondary & post-secondary 26 36   13 20 university & above 15 9       No. of Employed       2.1 1.6 Employment Status           Inactive 22 47   Wage worker 48 18       Employer 18 17       Self Employed 9 5       Unpaid family worker 3 13       Sources: Roushdy, Assaad, and Rashed [utilizing ELMPS 2006 (Egypt)];Sasin 2008 [utilizing ENCDM 2001 (Morocco)], outputs of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. 33   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options one. In Egypt, 35 percent of remittances accrue tances, since a potentially large portion of the to the highest quintile, versus 8 percent for the migration story will not be found in remittance- bottom. This is not an unusual characteristic of receiving households. remittances—worldwide, remittance transfers accruing disproportionately to the richer strata A full description of the methodology for of the population is a common finding. measuring the impact of migration/remittances on household poverty is included in Appendix While remittances benefit the better off, they 2 (Egypt) and Appendix 3 (Morocco). Instru- nonetheless benefit the poor. In Morocco, when mental variables techniques are used to correct evaluating households according to pre-remit- for the fact that remittances are (i) not the only tance income (since households may progress outcome of migration, and (ii) endogenous. A on the income distribution precisely as a result regression framework is utilized to explain the of remittances and migration), the percent of consumption level and poverty status of Moroc- households in the bottom quintile benefiting from can households by their characteristics, includ- remittances rises to seven percent, with remit- ing the migration status (proxied by remittance tances accounting for more than 50 percent of receipt), and controlling for other factors, such the household budget (Table 3.5).34 as health, gender, or family situation. Addition- ally, to control for endogeneity, an instrumental It does not appear that Moroccan families variable technique was used to correct for the passively depend on remittances. Only for less possible bias, utilizing the share of return mi- than a quarter of them are transfers the main grants in the population at the province level in source of income, while two-third tend to rely the 1994 Census as the instrument.35 In Egypt, on labor earnings instead, continuing to contrib- since both poverty and migration were binary ute economically to their communities. Other variables (poverty identified as belonging to the studies, such as de Haas (2006) or World Bank bottom quintile of the wealth distribution, using (2008), reach the same conclusions. In any case, the information on household assets), the model remittances provide an opportunity to diversify estimation strategy was not a trivial choice. Both households’ livelihoods, which is important in a two-stage least square estimation technique reducing vulnerability. In Egypt, where the data permits separate 34 The ELMSs questionnaires do not include a household con- analysis of migrant-sending households and re- sumption and income module, and thus cannot provide direct measures of the share of remittances in the household budget, mittance receiving families, 86 percent of families while the wealth status (both poverty and quintiles) was prox- receiving remittances have a household member ied following Filmer and Pritchett (2001), using factor analysis abroad, while only 66 percent of migrant-sending based on household asset ownership and housing characteristics. families receive remittances. This exemplifies 35 For full details of the empirical analysis in Morocco, refer to Sasin, 2008: an output of the EC-Funded World Bank Program some of the problems with attempting to measure of International Migration from Middle East and North Africa the impact of migration through data on remit- and Poverty Reduction Strategies. Table 3.5: Distribution of remittances in Morocco, “adjusted� quintiles “Adjusted� quintiles All Q1 Q2 Q3 Q4 Q5 Incidence (of remit. receipt) 11% 7% 8% 9% 12% 19% Remittance share in budget 13% 52% 24% 18% 15% 8% Per capita (in dirham per month) 140 190 110 108 126 156 Progressivity (share in total remit.) 100% 18% 11% 13% 22% 36% Source: Sasin 2008 [Calculations utilizing ENCDM (2001)], an output of the EC-Funded World Bank Program of International Migra- tion from Middle East and North Africa and Poverty Reduction Strategies. 34 Chapter 3: The Development Impact of North African Migration was employed (following upon Angrist (1991), These findings are similar to those reached which provided certain conditions under which in other studies analyzing the impact of remit- a two-stage linear model can perform well with tances on poverty. In their study of the broader binary endogenous variables models (Acosta MENA region, Adams and Page (2003, 2005) find 2006)) as well as a generalized least squares that, in the broader MENA region, a 10-percent- model. As a robustness check, a bivariate probit age point increase in the share of remittances (two equation probit) model was also estimated, in GDP reduces the poverty headcount by 5.7 using the biprobit command in STATA but imple- percent and the severity of poverty by 2 percent menting it as an IV estimation. This specification on average, while a study of 10 Latin American allowed for the binary nature of poverty and countries (Acosta, Fajnzylber and Lopez 2007) migration and, at the same time, dealt with self- found that poverty headcounts fall by no more selection and endogeneity of migration (remit- than 5 percent with the inclusion of international tances) by allowing the error terms in both the remittances in household income. However, the poverty and migration (remittances) equations results vary greatly depending upon the distribu- to be correlated.36 tion of remittance-receiving households across income levels. In Mexico and El Salvador, for ex- The empirical results on poverty ample, where most migrants are from low-income alleviation households, remittance flows reduce moderate poverty ($2 a day) by 15 percent. Empirical analysis provides supporting evidence of the poverty alleviating impact of remittances The analysis suggests that while migration in North Africa. Remittances decrease the can play a role in poverty alleviation, it is not a likelihood of household poverty by about 8.8 panacea for poverty reduction at the national percentage points in Egypt, while they reduce level. In both Morocco and Egypt, the poverty the likelihood of household poverty by about alleviating impact of migration and remittances at 7.1 percentage points in Morocco (Table 3.6). the national level is modest. With only 3.5 percent That is, households receiving remittances have of the Egyptian population receiving remittances an 8.8 percentage point lower probability of be- at the time of the survey, the absence of remit- ing poor than a households that do not receive tances would have raised the overall poverty remittances in Egypt, while households receiving headcount by only about 0.4 percentage points. remittances have a 7.1 percentage point lower In Morocco, meanwhile, the absence of remit- probability of being poor than households that do tances would have raised the poverty headcount not receive remittances in Morocco. With poverty rates in Egypt and Morocco of 20 percent and 15.3 percent, respectively, remittance-receiving 36 For further detail on the empirical analysis in Egypt, refer to Roushdy, Assaad, and Rashed 2009,, an output of the EC-Funded households have reduced their probability of World Bank Program of International Migration from Middle East poverty by close to half. and North Africa and Poverty Reduction Strategies. Table 3.6: Impact of Remittances on Poverty North Africa and Comparator Countries Reduction in probability of Reduction in poverty with 1 percentage point Country/region poverty with remittances increase in remittances/GDP share Morocco 7.1 Egypt 8.8 MENA (Adams and Page, 2003) 0.6 Latin America (Acosta, 2007) 0.4 Sources: Sasin 2008 [Calculations utilizing ENCDM (2001)]; Roushdy, Assaad, and Rashed 2009 (Using ELMPS 06), outputs of the EC- Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. Acosta (2007); Adams and Page (2003). 35   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options by less than a percentage point. Thus, the ability have been particularly affected. Indeed, data for migration to substantially alleviate poverty on from ENCDM 2000/01 confirms that migration is a national level is limited. more significant for some areas than for others. For example, 12 provinces (out of 60+) account However, for the individual household, for a half of remittances and in as much as 10 migration is an important element of the house- provinces more than 20 percent of the popu- hold’s coping strategy. Indeed, remittances re- lation receives remittances. In the traditional duce the probability for poverty by almost half emigration region of Oriental, and particularly in both countries. These findings support the in provinces of Nador, Jerada, or Taourirt, but qualitative evidence on migration as well, which also in places like Al Hoceima, Tanger, Tetouan, highlights the importance of economic (income) or Tiznit, migration and remittances are an diversification, and the exceptional importance extremely important element of population’s of migration and remittances as a method for livelihoods (see Table 3.9). Similarly, as de Haas moving out and remaining out of poverty. For (2006) reports, in certain valleys in the Middle example, “Moving out of Poverty,� (World Bank, Atlas (Todgha) international remittances are 2008), a participatory qualitative study of poverty received by as much as 40 percent of all house- in Morocco based on interviews with focus groups holds, for which these transfers constitute over within communities, reports that receiving for- a half of the total income. It is often from these eign remittances, salaries, or pensions (especially migration belts that the anecdotal evidence on from employment in Europe) allowed households the omnipresence and importance of migration to save, cope with crises, and simultaneously is gathered. start or expand income generating activities, thus diversifying their livelihoods. In particular, Moreover, migration plays an important role the report finds that migration of one or more in providing a wider set of opportunities for both family members, especially abroad, appears as a Moroccans and Egyptians to accumulate savings key avenue of upward social mobility. Community and to gain skills that could be useful when re- participants mentioned migration as one of the turning home. These benefits may well provide constituting characteristics of a socioeconomic additional welfare effects that might not be cap- status equivalent to the highest quintile. tured by the short-term impact of migration on household poverty status. Indeed, during the past And within communities, the ability of migra- decade, households with migration experience tion to lower poverty and raise living standards in Morocco clearly fared better than average. may be very high. Migration in Morocco is geo- ENCDM 2000/01 has asked respondents about graphically concentrated. Although migration has whether their living conditions had improved become an integral part of the Moroccan socio- compared to ten years before the survey, and economic landscape in general, certain regions the answers show an interesting pattern. Overall, Table 3.7: Regional differences in the impact of remittances in Morocco Poverty rate Region/province Remittance Receipt Current- Less remittances Change Country-wide 11% 15.3% 16.1% –0.8% o/w region: Oriental 16% 18.2% 22.2% –4.0% o/w province: Jerada 19% 21.9% 28.4% –6.5% Nador 19% 15.3% 20.3% –5.0% Taourirt 25% 19.7% 31.1% –11.5% Source: Sasin 2008, an output of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. 36 Chapter 3: The Development Impact of North African Migration according to this subjective measure, there has But the changes in labor supply flow through been a general deterioration in living standards more nuanced channels as well. Remittances during the decade preceding the survey. Almost a may contribute to increased entrepreneurship half of the population claimed that their situation as they may alleviate credit constraints that had worsened, compared with only 30 percent block investment. These may lead to more in- for whom it had improved. However, households vestment by households in their own businesses with access to international migration have fared and increasing engagement in self-employment much better than the rest. Migration thus has activities. A related issue is whether this will be proven to be an important livelihood strategy in accompanied by a reallocation of labor supply uncertain times. from other forms of employment. In order to satisfy operational needs of the new businesses Impacts on Household Labor Market or the need for absorption of the absentee mem- Decisions ber’s role, some household members may move from formal employment to self-employment. Labor markets are not only affected by the out- Empirically, this is often the case for females, as flow of migrant labor, they may also be affected women leave formal sector employment to work by changes in labor market decisions of those either in the informal or the unpaid family sector. left behind. Migration decisions are made at the But this may also occur for non-migrant men. household level, and both labor outflows and the receipt of remittances can lead to reallocation Some descriptive statistics of labor of the labor supplied by remaining household supply and employment among workers. migrant and non-migrant families Remittances are expected, other things being Before analyzing how remittances have impacted equal, to exert a negative influence on the labor labor supply decisions in North Africa, it is use- supply of the remittance-receiving households, ful to provide a picture in both countries of the insofar as it increases the reservation wage. In labor conditions that spark these household deci- this way, remittances can have negative impact sions. As with the poverty section, the analysis on output and productivity. Much of the recent on remittance-receiving households relies on the research on remittances has found that migra- two household surveys from the region in which tion and remittances tend to reduce household such analysis is possible: for Egypt, the Egypt labor supply, as non-migrant household members Labor Market Panel of 2006 (ELMS 06), with substitute increased income for more leisure, but the analysis focused on young men and women there are various ways in which remittances (and age 16–34 who had left school by the time of the migration) influence labor market decisions by survey in 2006/06. As Egypt’s population is very family members left behind. young, with a median age of 23, these youth make up a relatively high share of 46 percent of the There are two main reasons why changes can total workforce. In the case of Morocco, the data occur in the participation in the labor supply of comes from the 2000/01 ENCDM, with the analy- individuals living in households that receive remit- sis focused on men and women aged 15 to 64. tances. First, remittances constitute a new flow of non-labor income that may lift budget constraints The survey data confirms the high levels of and, therefore, raise reservation wages. Second, unemployment that have beleaguered the North remittances may signal the absence of a household African (and broader MENA) region for two member (the migrant). His/her absence generally decade (Figure 3.4). Morocco’s unemployment, entails forgone income, and there may further- while almost non-existent in rural areas, aver- more be significant migration-related expenses. aged some 23 percent of the labor force in urban These two income flows—remittances versus lost areas in 2001. Egypt’s unemployment is also high income—have expected opposite effects on labor for young workers (15–34), averaging 15 percent force participation of non-migrant relatives. of the young labor force in 2006. As is the case 37   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Figure 3.4: Unemployment in Morocco (2001) and Egypt (2006) Morocco Rural Morocco Urban Egypt (young adults) 35 30 Share of Total Labor Force (percent) 25 20 15 10 5 0 Male Female Total Source: Author’s calculations from Binzel and Assaad (2009) and Silva (2008), outputs of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. throughout the region, female unemployment with the problem of unemployment particularly exceeds male unemployment by large margins. acute for females. This reflects the tendency of educated women in Egypt to work for the govern- The survey data also confirms a recognized ment / in the public sector (Assaad 2007), since characteristic of low labor force participation it is associated with short working hours, social among women in the region. While overall labor security, and other benefits that allow women force participation averages about 60 percent to combine market work with domestic work of the working age (in the case of Egypt, young including child care. Yet, job opportunities in the working age) population, female labor force public sector have decreased over the last decade participation is significantly lower than male. In without an increase in similar jobs in the private the case of Morocco, male rates of participation sector, which has left many young and educated are about double those of females, while in the women unemployed (Assaad 2007). case of Egypt, male rates of participation are nearly triple those of like-age females, with only Education also has a strong influence over 31 percent of young women participating in some the labor market choices. As Table 3.8 shows, aspect of the work force (paid or unpaid). self-employment in Morocco is significantly more likely for lower levels of education, both for men For both men and women, higher education and women, suggesting that the self-employment is associated with worsening unemployment, sector consists more heavily of ‘coping’ activi- but the situation is considerably more severe ties (trade and services which don’t require any for women. As Table 3.8 shows, in Morocco, skills, etc.) than entrepreneurial endeavors. It almost a quarter of women with higher educa- is likely that most farm work is classified under tion in Morocco are unemployed (relative to self-employment as well, as the probability of only 11 percent unemployment for those with self-employment is significantly more likely in primary schooling). Prior results from Egypt lead to similar findings,37 although in Egypt the great- est concentration of unemployment occurs for For example, see Galal 2002. ECES WP67: The Paradox of 37 those with intermediate (secondary education), Unemployment and Education in Egypt. 38 Chapter 3: The Development Impact of North African Migration Figure 3.5: Labor force participation rates in Morocco and Egypt by Gender, 2000 (Morocco) and 2006 (Egypt) Morocco Rural Morocco urban Egypt (young adults) 100 90 80 70 60 50 40 30 20 10 0 Male Female Total Sources: Silva 2009 (Morocco); Binzel and Assaad, 2009 (Egypt),, outputs of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. rural areas (Table 3.9). The share of women Finally, in both countries, there are strong engaged in unpaid work decreases considerably gender differences with regard to the allocation with the level of education, while the share of of workers across activities, with women working women engaged in wage work increases, con- in predominantly unpaid subsistence activities sistent with the increased earning potential with and men working predominantly in wage or self- higher levels of schooling. employed activities (Figure 3.6). Table 3.8: Employment status of Moroccan labor force age population, by education Inactive Wage earner Self-employed Unpaid work Unemployed F None 60% 7% 9% 22% 2% Primary 64% 13% 6% 6% 11% Secondary 62% 20% 4% 1% 14% Tertiary 46% 27% 4% 0% 23% All ages 61% 10% 7% 15% 6% M None 9% 34% 40% 14% 4% Primary 16% 36% 20% 12% 15% Secondary 36% 32% 12% 5% 15% Tertiary 27% 42% 10% 3% 17% All ages 17% 35% 26% 11% 12% MF 40% 22% 16% 13% 9% Source: Silva 2009, using 2000/01 ENCDM, an output of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. 39   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Table 3.9: Employment patterns of Moroccan non-migrants, by location Inactive Wage earner Self-employed Unpaid work Unemployed R 30% 15% 21% 31% 3% U 46% 26% 13% 2% 12% UR 40% 22% 16% 13% 9% Source: Silva 2009, using 2000/01 ENCDM, an output of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. Figure 3.6: Male and female employment by sector, Morocco and Egypt Morocco M Egypt M Morocco F Egypt F 80% 70% Share of Total Employment 60% 50% 40% 30% 20% 10% 0 Wage and Work Salary Unpaid Family Self-employed Source: Binzel and Assaad 2009 (using ELMPS 2006); Silva 2009 (Using ENCDM 2001), outputs of the EC-Funded World Bank Pro- gram of International Migration from Middle East and North Africa and Poverty Reduction Strategies. With regard to migrant versus non-migrant tance receiving or migrant sending households households, detailed summary statistics of male versus those without migrants abroad or receiv- and female labor force participation (differenti- ing remittances. Women in remittance-receiving ated by remittance-receiving households versus households (RRHH) are more likely to engage non-receiving households; migrant-sending in subsistence work or unpaid family work, and households versus non-sending households) can less likely to hold a salaried job. The results are be found in Appendices 3–4. A few observations similar for households with at least one migrant emerge from these statistics. abroad, and suggest that women may be driven out of wage and salary work into other types of In both Morocco and Egypt, women in remit- work, such as unpaid family work and subsistence tance-receiving/migrant sending households are work. This might suggest that young women help less likely to engage in market work. In the case replace the labor of the migrant. Labor force par- of Egypt, while labor force participation among ticipation of young women increases slightly, so young females is about 31 percent of the age does, unemployment. For young women who are group (16–34), whether or not the family receives currently employed, migration and remittances remittances, the type of employment in which seem to lead to a decrease in the hours worked, young women engage differs strongly for remit- especially in wage and salary work (Figure 3.7). 40 Chapter 3: The Development Impact of North African Migration Figure 3.7: Employment of women in Egypt, by household remittance status Unemployed Wage/salary work Self-employed Unpaid/subsistence 50 45 40 Percent of Female Labor Force 35 30 25 20 15 10 5 0 Non-receiving Households Receiving Households Source: Binzel and Assaad 2009, an output of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. The empirical results on labor supply for young men (15–24) and older men (45–64), decisions which may suggest that remittances allow for extended schooling than otherwise (in the case The empirical analysis of remittance-receiving of younger men) and earlier retirement. households in Egypt and Morocco present a mixed picture of how migration and remittances Remittances increase the probability of self- impact labor supply. employment. Urban females from remittance- receiving households are 6.6 percent more likely In Morocco, remittances are associated with to be self-employed then counterparts from non- reduced labor supply, for both men and women, recipient households, while among rural females in both rural and urban areas, and increased receiving remittances is associated with an in- engagement in more informal activities. The crease in the propensity for self-employment of magnitude of the labor reduction is small, with 2.3 percent. Males have lower incentive effects remittances reducing the likelihood of labor sup- compared to females, ranging from 1.9–5.6 per- ply only by about 2.7 percent. The labor market cent, depending on the area of residence. In both effects are greater for females, a finding consis- cases, the increased self-employment occurs tent with most empirical studies worldwide. In along with a reduced probability for working in Morocco, women in remittance-receiving house- wage employment. Given the nature of the self- holds are almost 3 percent more likely to be out employment sector (predominantly low-skill), of the labor force than women in non-receiving this suggests that remittances may be allotted as households, with the impact particularly strong capital to more informal activities that provide for women in rural areas. The magnitude of this more flexibility. labor-effect rises with age for women, with re- mittances having the strongest impact on labor The receipt of remittances also increases supply among older women, while for men the unpaid work and unemployment, although the labor-effect exhibits a U-shape in relation to results differ by gender and location. Remit- age. In particular, the receipt of remittances has tances increase the probability of being engaged the greatest impact on labor force withdrawal in unpaid work (for women in urban areas and 41   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Table 3.10: Impacts of remittances on informality in Morocco: Multinomial Logit analysis Inactive Wage earner Self-employed Unpaid work Unemployed Nr Obs. 20,0775 11,223 8,062 8,083 4,439 Share 40% 23% 15% 13% 9% Urban Male 0.030*** –0.057*** 0.010*** –0.005*** 0.022*** (0.0005) (0.0007) (0.0006) (0.0001) (0.0005) Female 0.019*** –0.032*** 0.014*** 0.001*** –0.002*** (0.0005) (0.0004) (0.0003) (0.0001) (0.0001) Rural Male 0.015*** –0.093*** 0.057*** 0.008*** 0.013*** (0.0005) (0.0013) (0.0015) (0.0011) (0.0005) Female 0.041*** –0.013*** 0.057*** –0.003*** 0.0003*** (0.0011) (0.0002) (0.00149) (0.0006) (0.0001) ***Significant at 1 percent level. From Silva, 2009, an output of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. men in rural areas) or in becoming unemployed specialization following migration and/or remit- (for women in rural areas and men in urban ar- tances. Similarly, time allocated to subsistence eas). In all cases, wage employment is reduced work increases for women living in a migrant with the receipt of remittances. household while remittance income does not play any significant role. If, however, a young woman In Egypt, overall participation is positively in a migrant household works for a wage, she is impacted by migration and remittances, although able to reduce her weekly workload. More specifi- for men the impact on participation is not cally, how many hours leisure she can substitute significant. Labor choices are, however, signifi- for market work positively depends on the value cantly impacted by both the migration changes in of remittances received per month. For every 100 household composition, as well as by the receipt LE received, a young woman works 6 hours per of remittances. Young women react to changes week less in the market. Whereas young women in the household composition brought about by who work unpaid for the family significantly migration and help replace the labor of the mi- increase the number of hours worked per week, grant in the family enterprise and in subsistence young men significantly and very strongly reduce work. Although it is possible that the increase their unpaid family work in response to having in unpaid market work for women is due to an an international migrant in the household and/or expansion of family enterprise activity due to the to receiving remittances. Similar to their female savings made possible by migration, it would be counterparts, however, they generally work less reasonable to expect, in that case, to observe an in wage and salary work; although the impact on increase in male unpaid labor as well, which is men’s time allocation appears to be stronger than not the case. on women’s time allocation. In terms of hours worked, migration and The fact that women replace the labor of remittances create a situation where young the migrant without experiencing a reduction in women shift leisure time to unpaid family work. other chores, such as domestic work, suggests Since time spent on domestic work, including that those who leave Egypt and work abroad childcare, is not reduced, it suggests that young have been underemployed beforehand. Job op- women’s total work burden significantly increases portunities especially for (male) youth on the in migrant and remittance-receiving households Egyptian labor market are very limited (Assaad provided that they work unpaid for the fam- 2007). Hence, the fact that they work in the ily. There is no significant increase in domestic family business may be less a free choice and work, which is evidence against intra-household more a result of not finding a (better) job on the 42 Chapter 3: The Development Impact of North African Migration Table 3.11: The association between receiving remittances/having a migrant and the probability of participating in labor markets and self-employment El Salvador Mexico Rural Urban Nicaragua Rural Morocco Egypt Male Female Males Female Males Female Males Females Males Females Males Females Labor force –4.6% –9.9% –5.7% –10.7% –2.4% –10.4% –11% –2% –2.2% –3.1% 2.0% 3.6% participation Self- 0.3% –3.3% 10.0% –0.4% 4.3% 2.90%     3.8% 4.01% 29% 5.1% employment Sources: Silva 2009; Binzel and Assaad 2009, outputs of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. market. And as men are required to contribute Egypt. In the case of Egypt, the results suggest to the household income, they are likely to end a positive impact from remittances on labor sup- up working in a family enterprise rather than ply, although most of the increase is the result of becoming unemployed. The increase in young females moving into unpaid family labor; while women’s labor force participation in response to in Morocco, the negative labor effects are on migration, which is mainly driven by the increase the order of 2–4 percent. Similar studies of the in unpaid family work, also suggests that young labor market impacts in El Salvador, Nicaragua women would take up a job provided that working and Mexico found remittances reduced the conditions are perceived as appropriate. Women probability of employment by as much as 11 in the Middle East and North Africa region percent, with an average labor supply effect of face many restrictions when it comes to labor 7 percent. The relationship between remittances market participation, for instance, job mobility, and self-employment is also generally positive, commuting time and means, work time, and the as in Morocco and Egypt. type of job (World Bank 2004). Moreover, it is usually important for women that their job still Impacts on Decisions Affecting gives them enough time to fulfill their domestic Children chores. As a result of these restrictions, women’s labor force participation is very low in the region, Unlike foreign aid, most international remit- limiting the region’s economic development. In tances go directly to households in developing Egpyt, for instance, labor force participation only countries. For this reason, one of the most reached 26.9 percent in 2006 (Assaad 2007). important channels for remittances’ impacts on Hence, policies that help improve working con- development is at the level of household. From ditions and public transportation and that help the standpoint of economic development, the narrow the wage gap between men and women basic question is: How do these households spend will enable and encourage women to take up a their remittance earnings? Do households spend job and perhaps remain employed even after mar- their remittances on newly desired consumer riage. At the household level, migration (through goods for the family, or do they channel them potential temporary female headship) may also into human and physical investments? help women to gain bargaining power, which has proven to have a positive effect on household There are at least three views on how house- expenditures as women tend to spend more on holds spend or use their remittances and the schooling and health (Quisumbing and Maluccio impact of these monies on economic develop- 2000, Thomas 1997). ment. The first, and probably most widespread view within the literature is that remittances are Remittances do not appear, overall, to signifi- fungible and are spent at the margin like income cantly reduce labor supply in either Morocco or from any other source. In other words, a household 43   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options treats a dollar of remittance income just like a female household headship, household decisions dollar of wage income, and the contribution of on how remittances are allocated may be affected, remittances to development is the same as that which may have significant implications over the from any other source of income. The second view long term. As a result, remittances may facilitate takes a more pessimistic position, arguing that human capital formation, contributing to higher the receipt of remittances can cause behavioral long-term growth. Indeed, when worldwide data changes at the household level that may lower are examined, children in remittance-receiving their development impact relative to the receipt of households have a lower school dropout ratio.38 income from other sources. For example, a recent review of the literature by Chami, Fullenkamp, and Decisions about how to spend the money from Jahjah (2003:10–11) reports that: (a) a “significant remittances, who will benefit from the money, and proportion, and often the majority,� of remittances the short and long term effects that remittances are spent on “status-oriented� consumption; and might have on family structure are not gender (b) the ways in which remittances are typically neutral (Ramirez et al. 2005). In particular, the invested—in housing, land and jewelry—are “not decision over how remittances are allocated in the necessarily productive� to the economy as a whole. household budget will be determined by: i) mi- grants and their spouses’ allocation preferences, But a third, and more recent, view of remit- which are directly related to the individual’s sex tances is more positive, arguing that remittances and gender preferences; and ii) the bargaining can actually increase investments in human and power between the migrant and their spouses physical capital. One way remittances can result over the household decision-making process, in in increased human capital investments is because particular, the budget control (Figure 3.8). migration and remittances are not gender-neutral. Female headship is significantly more common among remittance-receiving households in North 38 Ratha and Mohapatra, Increasing the Macroeconomic Im- Africa. Partly because migration often results in pact of Remittances on Development, p. 4. Figure 3.8: Conceptual framework for gendered preferences over remittances Gendered Channels of Impacts on Transmission in sending households sending households Intra-household Remittances Expenditures allocation Female Children’s Education Enrollment Gendered Preferences Children’s Nutritional Status Source: Herrera, Dudwick, and Murrugarra, 2008, an output of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. 44 Chapter 3: The Development Impact of North African Migration But the effect of migration and remittances The evidence from North Africa again relies on schooling is known to be ambiguous, par- on the household surveys for Morocco and Egypt, ticularly as children supplement for missing the only countries with a nationally representa- labor in the migration process. Remittances can tive sample including migration information. As relax budget constraints that would otherwise part of this migration research program, both limit spending on schooling particularly for girls survey data were explored for the relationship and older children, but the absence of a parent between migration/remittances and decisions on may have a damaging effect on the family that child investment. In Egypt, the analysis focused ultimately negatively impacts school enrollment. on the impacts of migration and remittances on Migration may also positively affect schooling school attendance and child work.39 In Morocco, through its effect on the returns to education: on the other hand, the analysis focused on the if there is a large migration prospect for un- correlation between remittances and household skilled labor, returns to education become lower spending on health and education.40 thereby potentially reducing school attainment. Alternatively, if there is a large migration pros- The empirical results on investments pect for highly skilled labor, school attainment in children may be affected positively (the central argument of the potential for skilled migration to lead to In both countries, migration and remittances “brain gain�). lead to positive improvements with regard to investments in children. In the case of Egypt, mi- Similarly, the final direction of the relation- gration and remittances significantly reduce the ship between migration and child work is not probability of child work for boys in the 6–14 age clear. The income effect from remittances can range. Migration results in a moderate increase reduce the need for the income generated from in domestic work for boys age 15–17, however, child work. Income from remittances can also which may indicate that boys are substituting for raise the reservation wage of left-behind house- their father or absent migrant household workers hold members including children and possibly in domestic chores. Based on these results, the reduce their labor supply. In contrast, the need remittance income effect appears to be dominant for children to work in a family business may for market work for young boys, while the sub- increase in order to replace absent migrant stitution effect is dominant for older boys with household member(s). Migration can generate regard to domestic work. remittance money that can be used to start a family business or buy a farm, which may result For girls, the results are less robust, but in a need for household members, including when looking at work inclusive of all types, remit- children, to work. If children are not needed to tances exert a negative impact on long-duration do market work themselves, they may be needed work, which appears to stem from a reduction to do more household chores to replace migrant in domestic work. Households using remittance adults or adults that work on the family business money to purchase timesaving devices may fa- as a result of migration. cilitate this impact. In addition to the mechanisms above, mi- With regard to schooling, the findings for boys gration can affect child work through wages. If suggest that remittances have a substantial effect migration is occurring at a large-scale, this may on school attendance for the university-aged. This increase wages due to the reduction in labor makes sense given that pre-university education supply because of out-migration, thereby mak- ing child work more rewarding. Based on the 39 Elbadawy and Assaad 2009, an output of the EC-Funded World above discussion, the effect of migration and Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. remittances on children’s schooling is country- 40 Herrera, Dudwick, and Murrugarra 2008, an output of the specific (or even province-specific) and can vary EC-Funded World Bank Program of International Migration from according to each country’s peculiarities. Middle East and North Africa and Poverty Reduction Strategies. 45   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options for boys is becoming a necessity in Egypt and location resources induced by male heads, which hence is less income elastic compared to uni- is documented in MOP (2005)“Fathers prefer versity education. As for girls’ schooling, there to give education and money to sons over is a positive impact on school attendance (ever- daughters because the first ones will have a enrolling) and some positive effect on conditional higher return, girls will marry at any point.� attendance for girls 15–17, and a mild effect on school attendance of university-aged girls. The Similar evidence is found for the education impacts of remittances/migration on tutoring, share. Among non remittance-receiving house- which is nearly a necessity in Egypt in preparation holds, the boys’ demographic share of education for higher education, were insignificant. household expenses is larger than girls, resulting in a gender gap for this allocation; however, this In Morocco, remittances trigger a more equal difference is smaller among receiving house- intra-household allocation for education and food holds. These findings indicate that remittances between boys and girls, closing the existing gen- offset the gender bias in allocation of education der gap in rural areas. The econometric results and food expenses, triggering gender parity in for the food share expenses in rural areas indicate children’s human capital accumulation in rural that boys’ demographic share have an increased areas where girls are more marginalized.41 allocation of food while the girls’ do not. However, in presence of remittances this effect in favor of The Potential Effects of High-skilled boys is offset, closing the gender gap for food al- Migration from North Africa location. Since female headship affects the food share among remittance-receiving households High-skilled migration is often referred to as in rural areas, these results could suggest that the greatest ‘cost’ of migration. High-skilled women lead a more even distribution of remit- tances on food expenses between girls and boys, 41 Sadiqui (2004) shows that women in rural areas have been not revealing gender preferences. Therefore, disadvantaged of the urban modernization, for instance, their women attempt to deal with the inequality in al- education and health access has traditionally been limited. Figure 3.9: Impact of Girls and Boys’ Demographic Share on Education Boys Girls 2.5 2.0 Percent of Total HH Expenses 1.5 1.0 0.5 0 Non-receiving Households Receiving Households Source: Herrera, Dudwick, and Murrugarra 2008, an output of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. 46 Chapter 3: The Development Impact of North African Migration emigration can pose significant costs to sending developed Diaspora which is plugged into both countries. Skilled emigration can increase the the receiving and sending country economies, welfare of the individual, but for the society, spurring the diffusion of know-how, encourag- skilled emigration carries the risk of lowering ing innovation and growth, or strengthening growth prospects if it lowers the average level of cross-border business networks. Both skilled human capital in sending countries, thus drain- and low-skilled migration could also result in a ing them of capacity. The new growth literature transfer of norms and values (e.g., importance provides ample evidence of the role of human of education, lower fertility) that have a positive capital in, all other things being equal, foster- impact on growth over time. ing higher steady state economic growth (e.g., Lucas 1988, Azariadis and Drazen 1990, Romer As important, the cost of skilled migration 1990, de la Fuente and Ciccone 2003). The needs also to be measured against whether the loss of high-skilled labor is not only the loss of counterfactual holds: whether lack of skills is productive capacity by the workers themselves. the binding constraint to economic develop- Human capital can generate positive externalities ment, and whether the human capital produced through innovation and adoption of technology, is productively employed in the home country. result in higher fiscal revenues, and prompt If and when educated persons end up unem- better governance—benefits that are lost if the ployed, inactive, or are not utilized effectively educated persons leave. Shortages of skilled in production (either because the education health personnel, teachers, engineers and more system does not produce the skills needed for the generally “institution builders� in developing economy, or because productive workers are not countries can have impact beyond the migrants’ allocated efficiently to production), the foregone personal contributions to output, contributing to investments in education will have already been major bottlenecks in progress toward economic lost before migration even took place. The op- and social development. The loss of educated portunity cost of migration would then be much females in the home country may be associated lower. The cost of high-skilled migration, then, with declines in child and maternal health. depends very much on the country’s incentive framework, which determines their productive But high-skilled emigration can also bring employment at home. important benefits. Experience shows that skilled emigration may also involve positive externali- The impacts of skilled migration will differ ties, quite separate from remittances. Emigra- from country to country—depending on how ef- tion prospects could lead to a brain gain: if the fectively education creates productive skills, and foreign skill price is higher than the domestic how skills are effectively utilized in production. skill price, a higher emigration rate will raise the However, evaluating those costs is not possible return to human capital investment and increase given the current data. Accurately analyzing the level of investment. For a strong enough even the lost labor contributions of high-skilled induced investment effect, it is possible that emigrants would require information on the mi- a higher emigration rate will actually raise the grant prior to migration, including employment steady-state domestic human capital stock, if not and wages, or output (which would allow for all of the educationally-enhanced people end up estimates of the individual projected earnings leaving the country. Kapur and McHale argue that lost or projected output lost). while brain drain may signify a flight from weak institutions by those who have the possibility to It is possible to provide descriptive evidence leave, the ones who were perhaps best able to of some of the costs On the one hand, public change those institutions are also likely to be the education absorbs considerable resources in the first to leave, perpetuating the flows. Moreover, MENA region, and MENA countries also spend and as will be discussed more at the end of this comparatively more on tertiary and secondary chapter, high-skilled emigration can enhance education than on primary education. In absolute growth in later stages of migration through a well- and PPP (purchasing power parity) adjusted 47   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Table 3.12: Public education spending is comparatively high in MENA 2000 or closest 1995–2003 Ratio of spending 2002 Education/ Primary/ Tertiary/ Tertiary/ Tertiary spending/student GDP, % Secondary Secondary Primary PPP constant U.S.$ Algeria 6.1 1.52 Egypt, Arab Rep. of 5.6 Morocco 5.9 2.56 1.88 4.81 3442 Tunisia 6.8 1.53 2.66 4.07 4065 MENA Total 5.3 1.46 2.48 3.97 4222 Asia 3.6 1.27 2.20 2.79 2606 LAC 3.9 1.15 2.15 2.47 2172 Source: Estimates based on World Bank, 2008. Asian comparators: Republic of Korea, Indonesia, Malaysia, Thailand, Philippine, China; LAC comparators: Argentina, Brazil, Chile, Mexico, Peru. terms, spending on tertiary education in Morocco Such “brain waste� will have incurred the same and Tunisia is thirty and fifty five percent higher education cost as skilled emigration (lost public than in Asian countries. Importantly, education investment, lost social returns to human capi- spending is high in some countries with high lev- tal)—but carries less of the potential benefits els of higher skilled emigration: Algeria, Morocco (networking, knowledge generation, high private and Tunisia all spend above the regional average. returns to education, and remittances). For ex- In Morocco and Tunisia, tertiary level spending ample, Mattoo, Neagu, and Ozden (2008), found is especially high in absolute and relative terms. that after controlling for age, experience in the These investments are thus lost as students go U.S., and education level, highly skilled immi- abroad to work. grants from certain Middle East and North Afri- can countries were less likely to work in skilled Moreover, MENA countries generally are occupations. A large proportion of the variation not fast-growing, dynamic, or knowledge-based can be explained by the country of origin charac- economies, suggesting that there is a problem teristics that affect the quality of human capital in human capital formation. Economic growth of the immigrants such as investment in tertiary has been modest over the past four decades at education and the use of English in instruction. 4 percent, compared to 8 percent for East Asia; Another study (Batalova et al. 2008) finds that high technology exports make up only 4 percent brain waste generally diminished over time, and of manufactured exports compared to 31 percent that whether they recuperate their occupational in East Asia, and KBE indicators like patents or status, and how fast, depends on factors like scientific articles by residents are several times language skills. higher in East Asia.42 A numerical partial equilibrium simula- Finally, skilled migrants do not always end up tion exercise combines potential positive and in skilled jobs abroad. Skilled emigrants, tapping negative theoretical effects of increasing skilled into host country knowledge, can help transfer emigration rates significantly (for MENA sending such knowledge back home, upgrade their own countries), highlighting the links between tech- skills, and get a higher private return on their nology transfer, human capital formation, and investment in education than they would have growth. The simulation contrasts a worst-case in the home country, part of which is sent home as remittances. However, some highly educated migrants do end up in low-skilled occupations. 42 World Development Indicators 2009. 48 Chapter 3: The Development Impact of North African Migration Figure 3.10: Probability that a foreign-educated bachelor degree holder enters a skilled occupation (2000 census) S. Africa India Canada UK Hong Kong Australia Kenya Ireland Japan Taiwan China Hungary Germany Italy Russia Ghana Lebanon Nigeria Vietnam Brazil Jamaica Romania Iran Philippines Ukraine Pakistan South Colombia Poland Iraq Egypt Haiti Morocco Ethiopia Cuba Guatemala Dominican El Salvador Mexico Ecuador 0% 10% 20% 30% 40% 50% 60% 70% 80% Source: Mattoo, Neagu and Ozden 2008 scenario and a best-case scenario. The worst case improved institutional quality, and a lower risk scenario is one in which highly skilled migrants premium for foreign investors. do not remit, a large diaspora does not improve institutional quality at home or attractiveness for The simulations suggest that an increase in investors, and skilled emigration prospects do skilled emigration rates might affect income per not favor human capital formation. In the best- capita negatively in MENA countries, especially case scenario highly skilled migrants remit the for countries with already high emigration rates same fraction of their income as the low-skilled (around ten percent). Even considering such and there are important positive externalities positive effects as brain gain, high remittances, in terms of human capital building at home, and positive feedback effects on technology and 49   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Figure 3.11: Change in income per capita: worst-case and best-case scenarios. Worst-case Best-case 1% 0% –1% –2% –3% –4% –5% –6% –7% Algeria Bahrain Egypt Iran Iraq Jordan Lebanon Libya Morocco Oman S.Arabia Tunisia Yemen Source: World Bank 2009 High-skilled Migration in the Middle East and North Africa: Trends, Impacts and Policy Implications, an output of the EC-Funded World Bank program of International Migration from the Middle East and North Africa and Poverty Reduc- tion Strategies. institutional quality from the Diaspora, the cost their own economies. Very high unemployment is high, especially for countries like the North for the young and educated—a typical migrant, African countries where skilled emigration rates hence—is a feature of North Africa’s (and the are high. These results are in line with Docquier broader MENA region’s) labor markets. For ex- and Marchiori (2009). ample, in Morocco, one in four university gradu- ates active in the labor market is unemployed; Conditions that suggest a lower true in Algeria, one in five, and in Egypt, one in six. cost of high-skilled migration Unemployment rates of graduate women are also higher than those of men—twice and three times However, the above relatively pessimistic view in Algeria and Egypt, respectively. In Egypt, as of skilled emigration as a “brain drain� hinges on seen below, while those with at least secondary crucial assumptions: that skills can be effectively education make up only 42 percent of the labor produced and used at home and that skills short- force, they account for 80 percent of the unem- ages are a binding constraint to economic growth. ployed. In Algeria, only about 20 percent of the This assumption may need to be qualified in the labor force has completed secondary education case of North African countries. Table 3.14 out- or more, but they account for almost twice that lines many of the conditions that influence the proportion of the unemployed, and the same is ultimate impact of skilled migration on sending true for Morocco. (Figure 3.12). economies. From the table, there are at least a few conditions evident in North African labor markets The skills composition of emigration has which would suggest that at the actual lost out- tended to mirror the unemployment situation in put—the opportunity cost—from skilled migration domestic labor markets. Figure 3.13 shows the is lower than might appear at first glance. composition of both domestic unemployment and migration by skill level. The composition of To begin with, North African countries clear- unemployment reflects more recent data (gener- ly suffer from a form of “brain waste� already in ally, 2005), while the composition of emigration 50 Chapter 3: The Development Impact of North African Migration Table 3.14: Conditions that influence the impact of skilled emigration on sending countries Conditions which heighten costs of skilled emigration Conditions with lower costs of skilled emigration Education costs • If migration prospects encourage human capital • If migration prospects encourage human capital formation formation at home (public education) abroad • Private education Human capital • If skilled emigration reduces the supply or pro- • If skilled emigration increases human capital formation, portion of skilled labor and a part of that increase remains at home (brain gain) • If skilled emigration induces occupational short- ages in strategic sectors Economic effects • If skilled workers tend to remit less • Skilled emigration induces a large amount of remittances • Skilled migration generates Diaspora or network externalities (trade, FDI, technology adoption) • Tourism revenues Transfer of norms • If the integration of the world labor market in- • If emigration improves institutions at origin through diffu- duces wage pressure at origin sion of values • If migrants transfer other norms e.g., fertility, the impor- tance of health and education investments, etc. Employment opportunities at home • Migrants would have been productively employed • High unemployment and inefficient high-skilled labor at home and contribute to growth and well being allocation in sending country (“brain in the drain�) (“brain drain�) • Accompanying institutions lacking (e.g., in health sector): binding constraint to growth not emigrating labor Employment opportunities abroad • Skilled emigrants not productively employed in • Skilled emigrants productively employed in receiving receiving country (“brain waste�) countries (high private returns to education) Source: World Bank 2009. High-skilled Migration in the Middle East and North Africa: Trends, Impacts and Policy Implications, an output of the EC-Funded World Bank program of International Migration from the Middle East and North Africa and Poverty Reduction Strategies. reflects 2000, the latest year for which data is turned migrants to the Maghreb is highlighted. available. In Egypt, emigration tilts heavily to- From those surveys, between 10–17 percent of ward medium and higher skilled migration (more returned migrants were unemployed at the time than two thirds of migrants have more than of migration. Moreover, the data suggest that a secondary education), which parallels significant large proportion of returned migrants were previ- stocks of unemployed from that education pool. ously working in categories for which they may In Morocco, as well, the composition is largely re- have been relatively under-employed (including flective of the unemployment situation at home. seasonal and family work), which would also tend to suggest a lower economic contribution. Returned migrants’ surveys in Morocco, Tunisia, and Algeria, analyzed for the purpose Not only are unemployment rates high, of this migration research program43 provide there is considerable information to show that further suggestive evidence that some portion of the migration pool may come from the ranks 43 Gubert and Nordman 2008a and 2008b, outputs of the EC- of unemployed. In Table 3.15, the employment Funded World Bank program of International Migration from the status prior to migration for a sample of re- Middle East and North Africa and Poverty Reduction Strategies. 51   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options educated workers are not fully deployed to their return to higher education have been estimated most productive uses. MENA’s private rates of in a number of country studies, and they indicate Figure 3.12: MENA laborers with secondary and tertiary education, as share of labor force and unemployed Share of labor force Share of Unemployment 90 80 70 60 50 40 30 20 10 0 Morocco Algeria Iran Oman Tunisia Jordan Bahrain Egypt Source: World Bank 2009 High-skilled Migration in the Middle East and North Africa: Trends, Impacts and Policy Implications, an output of the EC-Funded World Bank program of International Migration from the Middle East and North Africa and Poverty Reduc- tion Strategies. Figure 3.13: Unemployment and migration from North Africa, 2000 Unemployed Emigration 18 16 14 Percent of labor force 12 10 8 6 4 2 0 Algeria Egypt, Arab Rep. Morocco Tunisia Source: World Bank 2009a (emigration rates; skilled emigration rates); World Bank 2008 (unemployment rates of secondary edu- cated workers); World Bank 2007 (unemploymemt rates). Note: Darker segment of bars reflect unemployed or migrants with secondary education of higher. 52 Chapter 3: The Development Impact of North African Migration Table 3.15: Employment Status of Returned Migrants from the Maghreb, prior to Migration Employment status Percent of total respondents prior to migration Algeria Tunisia Morocco Total Waged 37.5 36.6 19.0 31.3 Employer 1.8 1.2 0.7 1.3 Self-employed 15.1 14.6 15.1 14.9 Seasonal worker 12.4 15.8 9.8 12.7 Family worker 2.1 3.4 5.6 3.7 Unemployed 17.2 9.9 9.8 12.4 Retired 0.3 0.0 0.3 0.2 Student 10.3 12.7 28.9 17.0 Inactive 3.3 4.3 1.0 2.9 Other 0.0 1.2 9.8 3.5 Source: Gubert and Nordman 2008b, an output of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. Total reflects unweighted average of countries. poor private and social returns to higher school- not in line with economic development objec- ing, relative to comparator countries worldwide.44 tives. The education systems in MENA coun- Total Factor Productivity (TFP) estimates for tries differ from other regions in terms of the the region also suggest that North Africa and the specializations they produce. MENA countries broader MENA region suffer from persistently overall have a lower share of university students low productivity. While TFP is a measure of pro- in science, engineering and technology than Asia ductivity across all factors of production (not and LAC comparators, and a considerably higher just laborers, and not just educated workers) the share in education, humanities, and/or social sci- degree to which MENA’s productivity falls short ences. This structure is arguably reflecting a sys- of other regions suggests systemic problems with tem preparing for civil service employment rather the use of human capital in production. than for dynamic private sector led growth. Table 3.16 presents TFP growth estimates Moreover, MENA countries perform quite for the North Africa (and MENA) region from poorly in international tests of science and math- 1970–2005 (along with the growth of other ematics proficiency, reinforcing the notion that factors of production), compared with the pro- education in the region suffers from important ductivity growth experienced in other regions. quality deficiencies, impacting the true contribu- The methodology behind TFP estimates can be tions to output at home. Figure 3.14 shows the found in Appendix 1. Although North Africa’s percentage of MENA eighth graders able to meet productivity improved significantly over the early high benchmarks in mathematics and science, ac- 2000s, it remains below that of all other regions cording to internationally administered tests. As but Latin America, suggesting a lower overall late as 2007, the percent of MENA eighth graders economic contribution of educated workers from able to meet high benchmarks in science was only the region. 8 percent, and less than 3 percent were able to meet high benchmarks in math. That compares Much empirical research has been devoted to with a world median of 34 percent in science and better explaining the poor returns to education 26 percent in math. And indeed, some of the qual- in MENA and North Africa (both privately and ity issues, together with language barriers, may economy-wide). One important factor underlying the poor outcomes is the fact that the quantity, quality, and focus of educational outcomes are 44 World Bank 2008. 53   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Table 3.16: GDP per laborer growth and growth of accumulation and productivity by region, 1970–2005 (% per year) Region/ Average annual GDP Average annual growth of Average annual growth of Average annual Country per laborer growth human capital per laborer fixed capital per laborer TFP growth Sub-Saharan Africa 1970s 1.1 0.5 2.9 –0.3 1980s 0.2 0.6 1.3 –0.7 1990s 0.5 0.5 0.8 –0.1 2000–2005 2.4 0.5 1.2 1.7 East Asia and the Pacific 1970s 3.8 0.9 5.2 1.2 1980s 6.1 1.0 6.0 3.1 1990s 7.2 0.7 8.5 3.3 2000–2005 6.4 0.7 8.7 2.5 Latin America Caribbean 1970s 2.9 0.7 3.9 0.9 1980s –1.7 0.9 0.2 –2.3 1990s 0.8 0.9 0.7 0.0 2000–2005 0.0 0.9 0.3 –0.7 Middle East North Africa 1970s 2.6 1.1 7.2 –1.0 1980s 0.2 1.4 1.9 –1.5 1990s 0.6 1.3 –0.5 0.0 2000–2005 1.2 1.3 –0.1 0.5 North Africa 1970s 3.4 1.0 5.1 0.7 1980s 1.5 1.5 3.2 –0.7 1990s 0.3 1.3 –0.4 –0.3 2000–2005 1.3 1.3 0.4 0.4 South Asia 1960s 2.3 0.6 4.1 0.3 1970s 0.7 0.9 1.9 –0.7 1980s 3.4 0.9 3.2 1.6 1990s 3.6 0.9 3.8 1.5 World 1970s 2.4 1.0 4.0 0.2 1980s 3.3 0.8 3.6 1.3 1990s 3.8 0.8 4.5 1.5 2000–2005 3.7 0.8 4.7 1.4 Note: Regional estimates weighted by population. Source: Keller and Nabli 2007 (updated for the purpose of this report). 54 Chapter 3: The Development Impact of North African Migration Figure 3.14: Percent of MENA eighth graders reaching high benchmarks in mathematics and science, 2007 Science Math 40 35 30 Percent od Students 25 20 15 10 5 0 Iran Kuwait Tunisia Morocco Algeria Yemen Median Source: TIMSS 2007. explain some of the relatively lower success of exceptionally high public sector employment some highly educated MENA migrants in foreign has lowered MENA’s growth is not entirely clear, labor markets, especially the U.S. but a past study estimated that the loss of GDP growth in the MENA region between 1985—1995 Past education reforms in the region have strictly due to public administration employment focused on increasing access and attempted to was some 8.4 percent—or close to 1 percentage improve quality but did not emphasize enhancing point per year.46 the motivation for the productive use of educa- tion in the economy. Among the most distortive And although there may be shortages of policies inhibiting the productive use of educated adequate skills in the North African countries, labor (and all labor) in the region is the legacy of lack of education does not appear to be the key public sector employment. Despite attempts to binding constraint to private sector activity and reduce the size of the public sector, MENA (and job creation. According to the World Bank’s en- North African) economies maintain some of the terprise surveys, the percentage of firms listing highest levels of public sector employment in the skills shortages as the main business impediment world, averaging 29 percent of all employment in is comparatively low (in fact, lower than in ECA MENA in 2000 (and 26 percent in North Africa). countries or in LAC), much below the importance Most of these public sector workers come from given to, for example, access to finance, informal the stock of educated laborers. sector practice, basic infrastructure or corruption. Indeed, many of the weaknesses in the invest- A body of research on public sector employ- ment climate may be contributing to the flight of ment argues that human capital in the public educated persons looking for better opportunities. sector, especially within the administrative civil service, may not significantly contribute The much-debated case of emigration of health to economic growth, and in fact, may actually staff supports some of these counterarguments. In reduce economic growth if government workers use some of their powers to generate rents for 45 Pritchett, 1999. themselves.45 The degree to which the region’s 46 Pissarides, 2000. 55   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Table 3.17: Skills shortages and the business environment (% of firms listing the respective area as the major obstacle to business) Egypt Algeria Morocco ECA LAC Political 29 Access to 25 Tax Rates 35 Tax Rates 18 Practices 18 instability Finance Informal Sector Practices 17 Practices 20 Access to 14 Access to 15 Political 12 Informal Sector Informal Sector Finance Finance instability Tax Rates 12 Access to Land 11 Practices 11 Practices 12 Access to 11 Informal Sector Informal Sector Finance Corruption 9 Corruption 11 Access to Land 11 Political 10 Crime, Theft & 11 instability Disorder Inadequately 6 Inadequately 5 Inadequately 4 Inadequately 9 Inadequately 7 educated educated educated educated educated workforce workforce workforce workforce workforce Source: World Bank (2009) High-skilled Migration in the Middle East and North Africa: Trends, Impacts and Policy Implications, an output of the EC-Funded World Bank program of International Migration from the Middle East and North Africa and Poverty Reduc- tion Strategies Egypt, infant mortality is six times higher than in system seems to be a more important factor than high-income countries and only seventy percent the lack of doctors, per se. of pregnant women receive prenatal care. At the same time, physician emigration rates are low, at In the end, both factors—high unemploy- some 4 percent (much lower than in, for example, ment and low productivity among educated Lebanon or Iran) and the physician density in workers—point to lower than expected contri- Egypt more than fulfills the WHO’s recommen- butions by educated workers in the domestic dation of 20 doctors per 10,000 inhabitants. A economy, which in some sense also means that review by the WHO suggests that in Egypt, the the true costs from skilled migration might be specialization of physicians is poorly matched lower than that suggested by high-skilled mi- with needs on the ground (rural facilities, family gration rates alone. At the same time, and as medicine, primary care) and with some comple- importantly, there are significant costs to lost mentary health staff such as midwives and nurses education. The MENA region is bearing those (for whom emigration rates, by the way, have been costs, at least in large part through a domestic low). Thus, the lack of an enabling health service environment in which education is not channeled to its most productive purposes. 56 Chapter 4 Further Contributions of Migrants to Host-Country Development, While Abroad and At Return While remittances provide a major vehicle for diaspora for growth and development-enhancing development impact of migration while abroad, investments and to create the virtuous circles remittances are not the only channel. Dias- for institutional change. While the foundation pora—the migrant community abroad—can be a for successful diaspora contribution to develop- potentially valuable ‘second’ human capital stock ment exist throughout North Africa, the lack for the home country. This stock is valuable (rela- of responsiveness, flexibility, and dynamism on tive to truly foreign human capital) to the extent the part of the government hiders the ability of that diaspora members retain disproportionate diaspora networks of migrants from the Maghreb connections to the home economy. These con- and from Egypt to contribute fully to develop- nections support economic transactions such as: ment. Diaspora programs and organizations are trade, direct investment, financial investments, too centralized and bureaucratized to provide remittances, tourism, and knowledge exchange.47 effective support to diaspora members and their They can also be used as bridges for transac- organizations. tions with foreign nationals, bridges that may be especially important in international trading This section also reviews the experiences environments fraught with search, monitoring, of return migrants to North Africa, using survey and other transaction costs.48 data, which allows for a view into what makes one returned migrant contribute more to the Return migrants, like the diaspora, are also home economy than another: that is, what kind increasingly recognized for their potential to of experience abroad creates the infrastructure provide technological, managerial and entrepre- for the maximum contributions upon return. neurial know-how back at home. In addition to The analysis suggests several key features of the potential financial resources, the work experi- migration experience are integral to successful ence abroad and connections to both the host integration upon return (both in terms of labor and home economy make established migrants market success and business investment). First, (either from the diaspora or upon return) po- the conditions of return play a significant role in tential engines of innovation, employment, and the ability for return migrants to experience ei- economic growth at home. ther occupational improvements back home or to become entrepreneurs. Forced return of irregular This section reviews the diaspora networks migrants makes the potential for their successful of migrants from North Africa, evaluates their po- tential for contributing to development, and dis- 47 Gould (1994) and Head and Ries (1998) use a gravity equa- cusses some of the factors that hinder their full tion framework to show that larger diasporas are associated contribution. It also recommends policy actions with greater trade. for North African countries to better engage the 48 See Kapur and McHale (2005a). 57   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options contribution—either through occupational mobil- literature confirms that diasporas can have a ity or through entrepreneurship—to the home positive influence on trade and investment, there economy significantly less likely. Many receiving is surprisingly little empirical investigation into countries have instituted assisted voluntary return what makes one country’s diaspora more effec- programs (AVR) to increase the return rates for tive in supporting the sending country trade and irregular migrants and provide (financial) sup- investment than another’s. port for their return, but to date, the evidence has not confirmed their success. Second, training Diasporas worldwide have at times made star- received in the host country significantly improves tling contributions to home countries, with some the probability of entrepreneurship upon return. common elements. The often-mentioned case of And third, savings brought with the migrant upon Taiwan is useful for understanding the potentially return significantly improve the potential for mi- large contribution diaspora can make (and the grants to experience either upward job mobility potential for North African diaspora to do the upon return and to be able to become entrepre- same). When the Taiwanese government decided neurs. The results suggest the potential for sev- to promote the venture capital (VC) industry in the eral policies in both home and host countries to beginning of the 80s, it had neither the capabili- improve the contributions of migrants at return. ties, nor a blueprint to do that. However, through a process of intense interactions with the Taiwanese The Role (or Potential Role) of diaspora in Silicon Valley, new institutions such as North African Diaspora the Pilot Fund provided matching capital contribu- tions to private venture capital (VC) funds. Once The traditional fear of skilled migration is that the first venture funds proved successful, domestic the loss of society’s best and brightest, hurts IT firms created their own VC funds. economic growth and development at home. Yet, evidence is emerging that skilled diaspora can The first element of this successful engage- be a potentially valuable “second� human capital ment was a search network, consisting initially stock for the home economy, promoting develop- of key dynamic and forward-looking members ment at home. With connections to both home of the Taiwanese government, as well as leading and sending countries, diasporas can support overseas Chinese engineers in Silicon Valley. This many important economic transactions, including network did not have a blueprint, yet it did have trade, direct investment, financial investments, a role model (Silicon Valley) and a clear idea of remittances, tourism and knowledge exchange. “what to do next.� By defining each subsequent They can also be used as bridges for transactions step along the road, the network became wider with foreign nationals, bridges that may be espe- and eventually incorporated skeptics and op- cially important in international trading environ- ponents. ments fraught with search, monitoring and other transaction costs. And, through their continued The extension of projects of diaspora en- interest and involvement in the home country, trepreneurs from co-founding joint firms in the diaspora can also help to effect institutional home countries to co-creating institutional changes important for growth. infrastructure so that these firms can flourish is natural, and almost matter-of-fact. The initial Although diasporas can have a major role to objectives in the Taiwanese case were clear: to play in development, empirical investigations into advance their professional interests by setting their impact on sending countries are limited. up technology firms at their home countries. Yet Most studies have concentrated on the diasporas’ as the constraints of the home country institu- impacts on trade flows and investment. Using tional environment became apparent to them, gravity models to analyze the determinants of they engaged in advancing institutional reform trade flows, most studies have found that bilateral to remedy some of the constraints. Successful trade flows are positively related to the number growth of knowledge-based firms and creation of overseas nationals. But while the empirical of appropriate institutional environment became 58 Chapter 4: Further Contributions of Migrants to Host-Country Development, While Abroad and At Return two sides of the same coin. Innovation entrepre- valuable to gauge the absolute and relative size neurship has blossomed into institutional and of skilled diasporas, and the economic sectors in policy entrepreneurship. which change and interaction with home coun- tries is most likely to take place. But diasporas are not always successful as agents for change. For every success story of In 2000 France was the main receiving country diaspora achievements, such as from Taiwan, of first-generation North African migrants, fol- there are many more cases where diasporas lowed by Spain and Italy (Figure 4.1). However have tried but failed to make a contribution to in terms of more highly educated migrants stocks, their home country. Ultimately, the ability for France is the largest recipient, but the United diasporas to effectively contribute to develop- States and Canada follow (Figure 4.2). Finally, ment depends on several key factors, including among the largest receivers of North African mi- the size of the diaspora group, their motivation, grants (Table 4.1) the United States, Canada, and the government’s view of the political and and (to a lesser extent) the UK, Switzerland and economic relevance of the diaspora community, Australia host the diasporas with highest education and its willingness (and capacity) to tap into level relative to their number. These are the coun- their resources.49 tries where the probability would be higher to find some of the kinds of over-achievers that produced Description of North Africa’s diaspora lasting institutional changes in other regions. Diaspora contributions to development can come Considering professional occupations, 50 from many segments of the diaspora population, North Africa countries display concentrated and the types of interventions they can make can groups of skilled diaspora in some of the key also vary. But for practical purposes, the analysis sectors that have hosted change-makers from of diaspora groups with potential to make impor- other regions of the world. For instance, in tant contributions to home development generally year 2000 around 9,600 Algerians, 7,000 Moroc- concentrates on first-generation skilled migrants. cans and 3,100 Tunisians with tertiary degrees Skilled migrants are significantly more likely to worked in science and engineering profession have both the financial resources and knowledge in France alone, with Canada at the second to develop business ventures in the home country. place (Table 4.2). Egyptians highly skilled in sci- First-generation migrants, meanwhile, are the ences were especially concentrated in the United most likely to remain connected to their home States, again with Canada at the second place. To country, sharing the same social networks and place these figures into perspective, in the same idiosyncrasies associated with shared experiences period about 12,400 Argentineans and 6,700 of going to the same school and university. Chileans with tertiary degrees worked in the United States in science, technology, engineer- From a purely quantitative perspective, ing, business, finance, IT and health combined North African diasporas are not particularly (Table 4.3). Therefore, on a purely quantitative different from those of other mid-size middle- perspective, diasporas from North Africa—even income countries. Stocks of tertiary educated if not as large as Indians or Chinese—are not expatriates from North Africa’s economies are particularly different from other diasporas from not insignificant, and even if they do not reach mid-size middle-income countries that have been the mass of historically successful diasporas such shown in the literature to be capable of having as Taiwan, China and India, there are sufficient a remarkable impact on their country of origin. examples worldwide of smaller diasporas of skilled migrants who are just as productive for home country development as large ones. 49 Owusu 2006. 50 In order to partially control for “brain-waste� and inactivity, we consider in this section high-skilled foreign-born (e.g. first The OECD database on migrants stocks, generation) migrants who were actively employed in selected though referring to the year 2000, is still very professional occupations at time of the census. 59   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Qualitative evidence of North countries of Europe and North America. These African diaspora achievements groups are the basis of the “search networks� described before. North Africa also demonstrates some success sto- ries with diaspora initiatives. Diaspora networks One example within North Africa includes of migrants from the Maghreb and from Egypt are the associations of university graduates, stem- already well developed in all major destination ming from the fact that French leading universi- Figure 4.1: Stock of North African migrants in select OECD economies Algeria Egypt Morocco Tunisia 2500 2000 1500 1000 500 0 Belgium Canada France Greece Italy Spain Switzerland UK USA Source: Kuznetsov, Yevgeny and Matteo Morgani. 2008. Figure 4.2: Stock of North African Tertiary-Educated Migrants in Select OECD economies Algeria Egypt Morocco Tunisia 400 350 300 250 200 150 100 50 0 Belgium Canada France Greece Italy Spain Switzerland UK USA Source: Kuznetsov, Yevgeny and Matteo Morgani. 2008. 60 Chapter 4: Further Contributions of Migrants to Host-Country Development, While Abroad and At Return Table 4.1: Share of migrants with tertiary education in the top 7 destination countries of the OECD Country of residence Algeria Egypt Morocco Tunisia Australia 33% 31% 27% 14% Belgium 14% 41% 10% 19% Canada 62% 65% 52% 63% France 15% 43% 17% 16% Greece 31% 26% 26% 22% Italy 12% 22% 5% 5% Spain 12% 33% 10% 24% Switzerland 31% 51% 25% 24% UK 37% 44% 24% 26% USA 53% 63% 43% 49% OECD - Total 16% 47% 14% 16% Source: Kuznetsov, Yevgeny and Matteo Morgani. 2008. “Why is Diaspora Potential so Elusive? Towards a New Diaspora Agenda in North African Economies.� An output of the EC-funded World Bank program of International Migration from the Middle East and North Africa and Poverty Reduction Strategies. Note: Table includes the top 7 OECD destination countries of Egypt and the Maghreb, except for Germany (6ht destination country for Moroccans) for which data was not available in the same database. Table 4.2: Active foreign-born tertiary educated migrants in selected OECD countries, selected occupations (2000) Algeria Egypt Morocco Tunisia Algeria Egypt Morocco Tunisia   12 Corporate managers 22 Life science and health professionals Australia 4 773 30 4 10 1003 5 0 Belgium 0 0 Canada 0 0 0 0 190 1575 270 35 Switzerland 104 142 174 98 94 85 56 24 Spain 520 60 4840 60 40 60 1300 0 France 21681 582 12828 5778 15052 458 7828 3581 UK 565 1455 640 234 45 1496 32 24 Greece 2 753 13 4 0 356 5 2 Total 22888 3991 18563 6213 15444 5180 9504 3671   21 Physical, mathematical and engineering 23 Teaching professionals science professionals Australia 14 646 16 0 34 755 22 4 Belgium 682 0 Canada 1305 2160 1055 465 755 1385 1040 285 Switzerland 166 147 234 123 94 71 88 61 Spain 120 40 920 20 240 40 2700 20 France 13428 429 8986 4219 17062 235 11560 2707 UK 300 433 170 39 460 799 288 104 Greece 4 495 14 3 3 624 10 3 Total 15368 4473 12111 4889 18671 4029 15769 3203 Source: Kuznetsov, Yevgeny and Matteo Morgani. 2008. “Why is Diaspora Potential so Elusive? Towards a New Diaspora Agenda in North African Economies.� An output of the EC-funded World Bank program of International Migration from the Middle East and North Africa and Poverty Reduction Strategies. 61   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Table 4.3: Active foreign-born tertiary educated migrants in United States, selected occupations (2000) Occupations: Algeria Egypt Morocco Tunisia Argent Chile China UK India Iran Mex Philp Management 500 6490 1960 345 5430 2850 40145 47690 55460 19570 28665 32440 Business 209 3475 704 139 2140 1165 31890 17465 31475 7085 15700 42185 & financial operations Computer & 355 2070 904 204 1639 825 63670 16705 128535 7570 8230 22405 mathematical science Architecture 365 3095 564 165 1690 990 35830 13990 38365 11410 11700 20205 &engineering Life, physical, 230 1430 360 130 1590 925 37605 9630 20720 3095 3930 7860 & social science Total 1659 16560 4492 983 12489 6755 209140 105480 274555 48730 68225 125095 Source: Kuznetsov, Yevgeny and Matteo Morgani. 2008. “Why is Diaspora Potential so Elusive? Towards a New Diaspora Agenda in North African Economies.� An output of the EC-funded World Bank program of International Migration from the Middle East and North Africa and Poverty Reduction Strategies. ties (Grandes Ecoles) still remain the preferred the university and techno-hub of Toulouse, the option for tertiary education of Moroccan elites. headquarters of Airbus, the European Aerospace Three associations51 perform similar functions consortium (see Box 4.1). The organization for each one of the Maghreb diasporas: they link opened its network of professionals to Moroc- together perspective students with French uni- can companies searching for expertise in R&D. versities, enhance alumni networks with private This exemplifies the potential of a specialized companies interested specifically in diaspora network of professionals to provide avenues to graduates, and they offer cultural and social solve complex problems across boarders by re- events. ducing the search and information costs and by leveraging the inner desire to contribute to the Based on a similar concept (but with a more development of the home country. specialized focus) the NGO Maroc Entrepreneurs was founded in France by seven self-starting stu- Under certain conditions, these search net- dents of Business Administration, with the pur- works allow talented individuals to locate and pose of fostering entrepreneurship, particularly interact with the dynamic segments of home back in Morocco. Nine years after its foundation, country institutions, and produce transforma- the organization is still run by volunteers, but tional impacts and reforms now counts a network of about 3,000 students and 7,000 Moroccan graduates in Economics and However, while these social entrepreneurs Business,52 with offices in London and Paris. The exist, it is a challenge to find a suitable partner organization offers business training, and every among home country institutions (win-win year has a business plan competition for graduat- ing students. The authors of winning projects are matched with experienced Moroccan business- 51 The three associations are: ATUGE (Association Tunisiens men as mentors. des Grands Ecoles), AMGE (Association des Marocains des Grandes Ecoles) and REAGE (Réseau des Algériens Diplômés A similar specialized professional network des Grands Ecoles et Universités Françaises). 52 Source: interview by the authors with the current chief rep- is the organization Savoir et Developpement, resentatives of Maroc Entrepreneurs in the Paris and London created by the Moroccan scientific diaspora in Offices. 62 Chapter 4: Further Contributions of Migrants to Host-Country Development, While Abroad and At Return Box 4.1: Savoir et Développement: a European “Silicon Valley� Diaspora? An interesting example that shares similar traits to take structured and large research and development some of the early stage initiatives that took off around is limited, this organization is able to fill the gap by scientific hubs in North America is the spontaneous providing R & D services to the market of small and network of Moroccan scientists “Savoir & Développe- medium firms, as well as what we would classify as ment�1 grown in 1999 in Southern France’s scientific “consulting.� In terms of institutional cooperation, the pole of Toulouse (the headquarter of Airbus, the organization is collaborating with a local university to European aerospace consortium). In the words of the propose a nanotechnology research initiative. The or- author who documents this case, “the projects and ganization has also branched out to public institutions the subjects discussed by the group are extremely such as the Organization for Moroccan Abroad (with practical and concrete.�2 The organization serves as a which they organize annual meetings for Moroccan search network to match Moroccan scientists abroad students), and local municipalities to work on treat- with Moroccan small and medium enterprises willing ment facilities of urban waste. Time will tell whether to undertake innovation at the firm level. The member this search network will be able to translate such of the organization work either on R&D projects, or to promising steps into even more concrete and effective help firms finding partners for their R&D needs. In a actions at the institutional level, within the scientific business environment where local capacity to under- and business communities. Source: Kuznetsoy, Yevgeny and Matteo Morgani.2008 1 http://www.savdev.org/ 2 This draws from Boujour 2005 matches). This hinders the potential for these offering a clear mission that is recognized both groups to make altruistic, positive contributions by diasporas and home institutions. to the development of their place of origin. In their absence, the diaspora network in North Af- A third, yet again very unique, case of in- rica serves the more limited purposes of diaspora stitutional development led by diasporas is the members alone, such as career advancement, project of the International University of Rabat.53 business contacts, and socialization. This private university plans to provide advanced undergraduate- and masters-level courses in sci- The case of the philanthropic organization ence, architecture, social sciences and languages Coptic Orphans (see Box 4.3) is an example of through an intercollegiate faculty drawing from a how transaction and information costs can be network of international universities that employ reduced through the intermediation of a com- Moroccan academics. Not surprisingly, the proj- mon institution that ensures logistical support, ect originated with the initiative of a Moroccan entry points and the organizational structure professor working at the Polytechnic University to scale up—in this case through the Egyptian of Nantes, France, who desired to return to Coptic Church. The story begins with a highly Morocco and to contribute to the moderniza- entrepreneurial diaspora member, who gave rise tion of the country. A core point of strength of to a small private philanthropic initiative with the project design is its reliance on a network funding from her Coptic Church community in of well-positioned members of the Moroccan the United States. The enterprise evolved into a diaspora in other academic circles in the world. structured and professional NGO that reached Domestically, the endorsement of a powerful seg- with scale the poorest strata of the population ment of a home country government (the King in Egypt. The use of the Church as intermediary provided a shared set of beliefs that could tran- 53 Sources: Interview with a member of the Moroccan profes- scend cultural difference and could appeal to the sional diaspora and former government employee; Afrik.com, second generation of migrants in North America, Attalib.org, 63   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Box 4.2: Unlikely reform initiated and sustained by diaspora members: electricity in rural area of Morocco When a French steel plant in the city of Argentière-la- of their community of origin, to the European Union Bessée was closed down in the wave of heavy industry and the French Cooperation. The NGO also created retrenchments of the 1980s, laid-off workers received a platform where specialized associations could trans- seed funding to start local enterprises. Defying the fer knowledge and know-how: for instance, the case French authorities’ plan to invest funding locally, 54 of the association of volunteers from Electricité de migrants—under the leadership of a Moroccan worker France and the “Hydraulics without borders.� Equally and active member in local trade union—decided to important was the participatory, staged, and dialogic return to their region of origin, the Souss valley, and to method \the group used to identify priorities, receive invest in the development of electrification. The lack feedback, and establish local village institutions that of electricity was symbolic of the divide between the would plan for the future interventions and sustain the modern Morocco and the politically neglected territo- existing infrastructure. These innovations at the local ries such as the Taroudannt province, where govern- level did not come unnoticed by some government ment intervention had been scarce since colonial institutions, by then interested in servicing also these times. Turning into social entrepreneurs, these former regions (such as the Foundation Mohammed V), which insiders with a new modernizing vision engaged with co-funded initiatives and began including the local their communities of origin in an iterative process of communities in their infrastructure expansion plans. technology adaptation and institutional development. Following the hierarchy of diaspora effects illus- A key to their success was the ability to become inter- trated in this paper, over its 20 years of existence this faces between local needs and external resources, and organization of migrants also generated a remarkable a basic understanding of the technology that they were degree of institutional change at the national level. attempting to transfer. Through the establishment Perhaps the most telling evidence is that the public of a formal NGO and of a search network—which in electricity company eventually adopted the associa- 1997 reached about 60 organizations and established tion’s technical standards and participatory processes offices in Rabat, the Taroudannt region, and five cities in a massive rural electrification plan in 1996. More in France—Migration et Développement was able to broadly, the experience reversed, through demonstra- tap into a wealth of private and public investors on tion, a long-lasting bureaucratic paradigm that used to both sides of the Mediterranean; from urban migrants map this region within the boundaries of the “useless in Morocco interested in investing in the development Morocco.� Source:Kuznetsov and Morgandi 2009, an output of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. himself, and a number of former ministers), who human capital development, or it will become an donated public land for the construction of the additional springboard for brain drain. private institution in the Technopole of Rabat, shielded the project from possible interference Initiatives by national governments by competing institutions. With an initial in- vestment of 50M€, and planned revenue from The number and potential of skilled and unskilled substantial (10,000€) but still internationally migrants from North Africa has not remained competitive university fees, the school aspires unnoticed by policymakers. A large number of to serve the local as well as the regional public. initiatives have been implemented by both send- As the project is still at the preparatory stage, it ing and receiving countries to engage diasporas is too early to assess whether it will function as in collective remittance projects, encourage planned, and whether this intended local center return or circular migration, and invest in their of excellence will have positive spillovers on local country of origin. 64 Chapter 4: Further Contributions of Migrants to Host-Country Development, While Abroad and At Return Box 4.3: Diaspora-led philanthropy transforms the way traditional charities operate in the Egyptian Coptic Church Descending from Christian converts of the first century, percent drop-out rate among girl beneficiaries. As the Coptic Egyptians have a long tradition of cohesion as a communities reached by the NGO became hundreds, minority in their home and host countries alike, mak- procedures became standardized, and today donors ing this group somehow singular with respect to other receive transparent accounts on the use of their dona- North African diasporas. When a Coptic-American vis- tions (a remarkable improvement in light of a tradition ited an orphanage run by Coptic Nuns in Cairo and wit- that would discourage the giver from inquiring about nessed the situation of resident children from poverty- the result of his charity). stricken families, she began fundraising within the Cop- By exposing the direct link between diasporas and tic Churches of the United States with unexpected suc- communities of origin, Coptic orphans awakened not cess. Coptic communities abroad had a long tradition of only compassion but also a sense of capacity among charitable giving through bilateral exchanges with twin diaspora members to share some of the “fortune� that parishes in Egypt, but their effects—comparable to the made their destiny so different from that of their fathers. typical collective remittances—only contributed to the In exchange for the contributions, the NGO offered sustainment of the existing charitable system. Under- identity and opportunities for fulfilling engagement: standing the potential to leverage diaspora resources to through summer English teaching programs, networks produce a more lasting impact on the children’s lives, of young Egyptian-Americans were able to re-establish Nermien Riad’s next step was the creation of a program the cultural link with Coptic Egypt. A further success to counter early school dropout and girls’ high illiteracy factor was the NGO’s ability to build on the logistical rates, by pairing adult women—volunteers from the strengths of the Coptic Church, an institution which is local community—with younger girls from low-income both local and transnational, and whose beliefs and families in various church districts. The girls would re- cultural norms are shared among sending and receiving ceive both academic support and a role model through communities and conducive to charity itself. The orga- their educated “big-sister� role models. Local churches nization also strives to stick to the intrinsic motivation played an intermediation role by providing the com- of Coptic charity: for instance professional staff is kept munication infrastructure, volunteers, and the outreach to a minimum, and limited to the coordinating offices in capacity in communities. Cairo, Australia, Canada and the United States. A second program directly targeted children who The relationship between the State and NGOs suffered from the death of their father, granting their continues to be complex and ambivalent in Egypt. On families an alternative to traditional orphanages to raise one hand the role of NGOs in improving the welfare of their children. In its 20 years of operation, these initia- citizens is recognized, on the other all civic activities tives that started in few communities around Cairo now are closely controlled and their legal status can be reach the entire country, and distribute $2.3 M a year challenged by the state. This ambivalence may be all (2007). Complementary workshops targeting children the more true for a NGO which represents diaspora and parents (for instance on rights, nutrition, study hab- Copts, who provide remittances and are well connected its) support change processes at the community level. with eminent people abroad, but nonetheless work to Through these actions, the diaspora NGO was serve a local minority who is discriminated against able to channel part of the Copts’ remittances into a and at times persecuted. Especially for a diaspora professionally-managed development program, and organization that embodies, in its way of proceeding, to become a change factor across a large network a “change� agenda, the willingness and ability to col- of traditional and hierarchical charitable institu- laborate with the state—at the state’s pace—without tions. Introducing individual mentoring and mentors’ loosing its focus on results may depend on a future training enhanced local volunteerism. New account- accidental encounter with an enlightened internal ability mechanisms allowed monitoring the behavior champion. This is yet to occur, but the change brought of beneficiary families towards the welfare of their within a hierarchical and complex organization such as children, and to record the remarkable improvement in the Coptic Church is, per se, a remarkable product of literacy, academic performance and achievement of a 0 diaspora social entrepreneurship. Source: Brinkerhoff (2007) and interview with C.O. Executive Director Nermien Riad. 65   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Initiatives driven by host country gov- tate return and investment by diaspora members. ernments (bilateral aid agencies or foreign The international experience suggests that such ministries) in receiving countries in Europe tried bodies serve as initial gateways to home country to engage non-specialized diaspora organizations institutions, but they rarely offer specialized in development projects (co-development). De services that a potential “overachiever� would Haas (2005) provides a very thorough review of be able to take advantage of, and direct contact such programs in the EU—many of which are with home country institutions is usually more intended for Maghreb migrants—and, his general relevant. However, they play an important con- findings and recommendations are very relevant vening function—as they can organize diaspora to our discussion and worth summarizing here. congresses, which draw together more or less specialized segments of migrants—and link them Since the 1970s, some of the most popular with country institutions. and best financed programs have focused on co-development or “brain circulation� initiatives Thus, though North African diasporas have with a more or less hidden agenda of return, for already made significant contributions to institu- instance by requiring participants that benefit tional development and promising and significant from seed funding for their investment projects examples of diaspora contributions to home to leave the country for a number of years. Evalu- country development exist, the promise of di- ations of experiences in the Netherlands and in aspora remains immature. The challenge for the France54 showed that such initiatives did not government is scaling up and institutionalization engage the brightest and most dynamic diaspora (i.e., rendering some governance and structure) individuals—let alone groups—because such to many promising diaspora networks, which are conditions would be appealing only to “failed� often informal. For North African diasporas to migration experiences. In terms of impact, it make strong contributions, they require a higher appears that investment projects by return mi- level of support to diaspora members and its asso- grants under such circumstances did not prove ciation in a less centralized and bureaucratic way. sustainable. Strengthening ties with North Africa’s A number of governments have attempted diaspora to involve diasporas as potential partners in local development cooperation, especially by North Africa demonstrates segments of home giving grants to specific projects or by match- institutions and foreign champions that manage ing collective remittances. A key challenge of to advance a shared agenda. However, North such programs has been to increase the capac- Africa’s public institutions that have the role of ity of migrants’ spontaneous organizations to liaising with migrants have a limited capacity to take advantage of these funds, without losing, work as interlocutors in very specialized fields in the process, the ingenuity of the effort and such as technology. These kinds of matches re- without fostering an “over-professionalization� quire special skills, and understandably a public of migrants. It appears that the most promising agency cannot host them all in house. results in terms of quality of projects have been obtained where migrants’ organizations were The analysis of more mature diaspora experi- able to partner with professional NGOs to build ences than those of North Africa, already docu- capacity.55 mented by the literature, is that this process of All North African countries have established offices for migrants abroad in the main desti- 54 REMPLOD program in 1970s in the Netherlands, Aide au Re- nation countries, but there has been no clear tour program 1980s, and Programme Developpement Locale Migration, 1995, in France. evaluation so far of the quality of their functions. 55 De Haas cites as positively evaluated examples the small grant These offices are also in charge of disseminating program for non-professional development organization Linkis information on incentives and initiatives to facili- in the Netherlands. 66 Chapter 4: Further Contributions of Migrants to Host-Country Development, While Abroad and At Return Table 4.4: Institutionalization of diaspora networks as search networks Characterization of better performing segments Examples Informal networks Individual champions, usually • Ireland in the 70’s, ‘overachievers,’ from government, • India in the 70’s and 80’s diaspora, and private sector sides • Networks of scientists, business school graduates from Maghreb • Most middle-income and many low income countries now, Some The champions (personalities) • Taiwan experience with early stage venture capital institutionalization create institutional platforms to • Mexico’ Red de Talentos institutionalize interactions • Diaspora initiatives promoted by private sector associations such as TiE • Associations of Maghreb business school graduates Institutionalized A process of matching of Diaspora • GlobalScot networks members and institutions in home • ChileGlobal countries to generate and support joint projects emergence and institutionalization goes through How can government provide coherent a three stages. centralized framework to assure diversity of bottom-up initiatives to fit specific local circum- In the first stage, informal networks between stances? The evidence from successful experi- individual champions, typically overachievers, ences worldwide suggests a two-prong approach: emerge. The ability to act innovatively and think both a centralized framework (which is up to ‘out of the box’ stems both from personal repu- government to establish) which makes diaspora tation and credibility (not necessarily linked to members feel welcome and assures basic rules of an official position) and the ability to leverage their engagement, and an institutional space for resources provided by an official position. This bottom-up creativity and initiative of all agents duality allows flexibility and opens the door to involved. institutionalization of personal and informal net- works. The second stage of evolution is their par- North Africa countries have developed a tial institutionalization, where diaspora search number of initiatives to engage with diasporas networks can help to formalize other networks such as building databases of diasporas, websites while making them more effective as a means to promote diaspora dialogues, and subsidies for incubating new programs as governance for short scientific collaborations. Ministries of structures as well as new projects. Finally, the Foreign Relations develop outreach programs to third stage is the fully institutionalized search link to its people abroad. This effort should be network. continued. This is a framework for the diaspora engagement to establish a dialogue to make The full formalization of diaspora search diaspora feel welcome and needed at home—a networks is typically not desirable, particularly first prong of the strategy we propose. However, in a context of a developing economy. An insti- the framework cannot substitute for decentral- tutional home is desirable, yet many informal ized and largely bottom-up creativity—diaspora features (characteristic of Stage 2 of the evolu- members and organizations working with home tion) should remain. Full institutionalization can country institutions to change things for the easily result in stifling of creativity and capture of better. a heretofore vibrant network by vested interests. In short, interests of powerful organizations may Striking such a balance is far from trivial and overtake dynamic searches of diaspora members. the centralized ministries of diasporas (as, say, in 67   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options India), proved to be at best, of very limited value. ogy and technology-driven SMEs as a means of The relevant analogy is promotion of innovation. employment creation. One needs a tolerant innovation climate, yet it is private agents, not the governments which make Furthermore, many efforts are being carried innovation happen (with Ministries of Sciences out by bilateral development agencies as well as and technology more often part of the problem, international organizations to mobilize diaspora rather than solution). engagement in their home countries. However, rather than providing financial incentive to con- One blending of bottom-up creativity with an vince diasporas to invest in development proj- explicit national framework of ‘rules of the game’ ects in their home countries, where success has is a national contest between diaspora initiatives, proved questionable, these organizations could which could provide matching funds to organiza- focus on three lines of assistance: tions in a country interested in articulating and running diaspora initiatives that would advance • Evaluation and documentation of successful their own missions and objectives (for instance, approaches, which will build critical knowl- an organization interested in using diaspora edge of the advantage of the diaspora social professionals to test new methods of delivery of entrepreneurs who are already engaged in preventive medicine to low-income population). the cause. These types of diaspora contests can help to pilot • Sponsoring of gatherings of diaspora orga- promising activities. nizations from different countries to favor knowledge exchange and provide informal Attempts to foster diaspora talent in a de- opportunities for mutual assistance and centralized way already exist in the Maghreb and learning. there is a diaspora agenda, at least in the narrow • Assistance to home country governments sense: diaspora ministries, Ministries of Foreign in establishing enabling policy frameworks, Relations, and so forth. These are entry points including the removal of bureaucratic hurdles of diaspora engagement which play also a coor- that prevent diaspora-led NGOs to act freely dination role: advocating reasonable institutional in the home country as entities. environment for diaspora engagement and main- taining dialogue with diasporas. These agencies North African countries already recognize are Diaspora “embassies� in the home countries. the potential for diasporas to be economic part- But just like embassies are just entry points to ners and global bridge builders. And the evidence the governments, one needs “diaspora agenda in worldwide suggests that diasporas do not need to the broad sense�—a process of engagement with be large and voluminous to produce an impact. specialized government agencies (Ministries of However, North African countries do need to Health, Education, Science and Technology) and ensure that the diasporas are better engaged agents to elicit credible commitments between if they are to be the valuable source of tangible the agents with resources and expertise at home help and resources (remittances, philanthropic and relevant diaspora members. contributions, investments, and technological and organizational knowledge) to home country As a short-term dimension of their diaspora development. strategy, North African countries should empha- size decentralized ‘light touch’ agendas allowing a Return Migration diversity of NGOs and research efforts to flourish. Once a diversity of diaspora engagement efforts Migration can lead to important forms of benefi- is apparent, one can, over the medium-term, cial transfers back to home countries, in the form move to strengthen some of them by establishing of technological, managerial, and entrepreneurial a diaspora contest between local organizations know-how. Some migrants who return home may in specific areas (such as in Russia and Mexico), have acquired the financial resources, and the including, but not limited to, science, technol- work experience, while abroad to provide an 68 Chapter 4: Further Contributions of Migrants to Host-Country Development, While Abroad and At Return impetus to the local economy and become en- With regard to labor market performance, gines of innovation, employment, and economic empirical studies have focused on whether growth. The development impact of return mi- returnees are able to apply at home what they gration, however, as with all stages of migration, learned abroad through a comparison of the varies strongly across countries. Understanding wages of return migrants to the wages of those the factors that influence the ultimate contribu- who stayed in the home country.58 Contrasting tion—be they characteristics of the migrant and results emerge from this literature, with some their return, characteristics of the home country, finding that experience is neither penalized or characteristics of the time abroad—are impor- nor rewarded,59 and some finding that there is tant information for guiding policy-makers both a wage premium for having gone abroad,60 with in North Africa and in destination countries. large differences in results across gender and host country (foreign experience was found to It is generally acknowledged that while the strongly matter for women but not for men). number of North Africans returning home was sizable up to the mid-1970s, returns since then Another stream of empirical studies has have been limited in size. This tendency of some examined the impact of return migration on the migrant workers to settle for good in immigration development of small businesses in the home countries (or, at least, to stay longer) is due to country.61 There are two ways through which several factors, among which are poor economic experience abroad might enable migrants to prospects in the home countries, high income contribute to small business development: first, differentials between home and host countries, accumulated savings abroad might contribute to and the closure of European frontiers, making alleviate domestic capital market imperfections; legal circular migration impossible. Lower-bound secondly, overseas work experience might gener- estimates of the number of return migrants in ate new skills and new ideas. From this literature, Morocco, Algeria and Tunisia computed from there is information on North Africa. McCormick census data are provided by the website of the and Wahba (2001) explore the extent to which MIREM project.56 In the case of Morocco, about returnees to Egypt become entrepreneurs and 68,000 international migrants returned between the way this process is influenced by overseas 1975 and 1982—almost 10,000 per year—and savings, overseas work experience, and pre- 117,132 returnees were recorded in the 1994 migration formal education. Using data drawn census. In the case of Algeria, 29,863 individuals from the ELMS 88, which included a return mi- interviewed in 1998 were abroad ten years ear- gration module, they estimated a simple model lier, suggesting a return migration flow of about of the probability that a return migrant is an 2,600 individuals per year over the 1987–1998 pe- entrepreneur. Their findings suggest that total riod (RGPH 1998). Turning to Tunisia, the return savings accumulated overseas and the length of migration flow is estimated at 5,931 individuals overseas employment positively and significantly per year over the 1982–1984 period, and at 3,553 affect the probability of becoming an entrepre- individuals per year over the 1999–2004 period.57 neur among literate returnees. By contrast, Development impact of return migration See www.mirem.eu/donnes/statistiques/statistiques for statis- 56 tics based on census data. 57 Gubert and Nordman 2008b, an output of the EC-Funded Data on return migration are scarce, and there World Bank Program of International Migration from Middle East is only a small amount of empirical literature and North Africa and Poverty Reduction Strategies. 58 See Kiker and Traynham 1977; Enchautegui 1993; Co, Gang that has examined the impact of return migra- and Myesong-Su 2000; de Coulon and Piracha 2005; Rooth and tion from the sending country perspective. Two Saarela 2007 issues in particular have received research at- 59 See Enchautegui 1993. 60 See Co et al. 2000. tention: the labor market performance of return 61 See for example Ilahi 1999; McCormick and Wahba 2001; migrants, and the characteristics of businesses Ammassari 2003; Black, King and Tiemoko 2003; Wahba 2003; created by returnees. Mesnard 2004; Nicholson 2004. 69   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options longer periods overseas have no influence on the By contrast, it is important to clarify that likelihood of becoming an entrepreneur among other questions cannot be addressed: First, illiterate returnees. since there are no non-migrant individuals in the sample, the questions of whether the Information from returned migrants’ entrepreneurial behavior of return migrants studies of North Africa differ from that of non-migrants or whether experience abroad affects the characteristics of Lack of appropriate data prevents an empirical businesses established by the returnees cannot investigation of the impact of returned migrants be explored. in North Africa. However, three recent surveys on returned migrants simultaneously conducted in Second, the data set focuses on returnees Algeria, Morocco, and Tunisia in 2006 as part of and, as such, is not a representative sample the MIREM project62 provide a window into the of migrants in general. Since migrants from factors which motivated and shaped the migra- Maghreb countries are not mandated to return tory stages; how the human, social and financial (even though some of them are sometimes “en- capital of the interviewees changed over time; couraged� to do so), returnees are unlikely to and how patterns of reintegration and entrepre- constitute a random sample of the migrant popu- neurship differed among returnees and countries. lation. It may be the case that those who have failed economically or socially in host countries About 330 returned migrants were inter- or those who are retired are overrepresented in viewed in each country using a common ques- the return migrant population. Controlling for tionnaire. In each country, the sampling proce- this would require having data on migrants who dure was based on a geographical stratification still reside in immigration countries. Since such process. A few specific regions were selected data could not be collected for obvious logistic using official statistics on return flows, so the and financial reasons, the conclusions that are survey data should not be viewed as reflecting na- derived from the analyses that follow only apply tional trends (Table 4.5). For the MIREM project, to the surveyed returnees, and they cannot be a returnee is defined as “any person returning generalized to the whole population of migrants. to his/her country of origin, in the course of the last ten years, after having been an international Nevertheless, while the studies are limiting migrant (whether short-term or long-term) in in this sense, they provide a valuable blueprint another country. Return may be permanent or into what makes one returned migrant contribute temporary. It may be independently decided by more to the home economy than another: that is, the migrant or forced by unexpected circum- what kind of experience abroad creates the infra- stances.�63 This definition partially draws on structure for the maximum contributions upon the one recommended by the United Nations. It return. In this case, they are indeed valuable. refers specifically to migrants who returned to Moreover, perhaps no migration policy is more their country of origin in the course of the last heavily touted than circular migration. The Eu- ten years, as this time limit allows for assessment ropean Union, in particular, has invited to Com- of the impact of the experience of migration on mission to propose ways and means to facilitate the interviewee’s pattern of reintegration. It also circular and temporary migration between third allows the respondents to recount their migratory countries and the EU. Since the surveys provide experiences more precisely. information on the difficulties returned migrants have faced in reintegrating into the home coun- The questionnaire is structured around try, including re-entry into the labor market, three modules relating to the different migratory they are very useful for helping policymakers to stages: the returnees’ conditions before they left for abroad; the returnees’ experience of migra- 62 see www.mirem.eu/mirem?set_language=en, for further de- tion lived abroad; and the returnees’ post-return tails on the whole project. conditions in the country of origin. 63 Gubert, Flore and Christophe Nordman. 2008b 70 Chapter 4: Further Contributions of Migrants to Host-Country Development, While Abroad and At Return Table 4.5: Composition of national samples for return migrants Algeria Morocco Tunisia N % N % N % Wilayas Regions Governorates Algiers 104 31.3 Tadla-Azilal 111 33.6 Tunis 122 37.0 Setif 82 24.7 Casablanca 99 40.0 Ariana 40 12.1 Bejaia 75 22.6 Chaouia-Ourdigha 57 17.3 Sfax 40 12.1 Tlemcen 71 21.4 Rabat-Salé- 50 15.2 Sousse 40 12.1 Zemmour-Zaër Other Regions 13 3.9 Nabeul 28 8.5 Medenine 25 7.6 Mahdia 20 6.1 La Manouba 15 4.5 Total 332 100.0 Total 330 100.0 Total 330 100.0 Source: Gubert and Nordman, 2008b, an output of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. improve the ease with which migrants to return The return migrant survey data provides and contribute to home country development. an opportunity to examine the conditions un- der which migrants experienced job mobility. Return migration, reintegration, and Not all mobility is good, however, and thus a entrepreneurship methodology for establishing upward mobility versus downward mobility was needed. The full The return migrant survey data was analyzed for methodology behind occupational classifications, several purposes: first to better understand what along with the econometric model to test for oc- factors have been most important in helping sup- cupational mobility, is presented in Appendix 9. port occupational upward mobility for returned The occupations from the North African sample migrants, and second, to better understand were classified into 5 mutually exclusive groups returnees’ entrepreneurial behavior. or clusters, which can be ranked from 1–5, with 1 being the highest occupational status and 5 Determinants of return migration being the lowest: occupation mobility Group 1: Employers, or employees with Occupation mobility is important in two ways in secured jobs relation to reintegration. First, for the sending Group 2: Employees with secured part- country, upward occupational mobility ensures time/short-term jobs that return migrants will be able to make greater Group 3: Small self-employed economic contributions upon their return. With- Group 4: Employees with unsecured jobs or out greater success in the labor market upon unemployed return, the gains from migration decline signifi- Group 5: Family workers cantly. But secondly, for the individual, occupa- tional status is a good indicator for overall social Using this ranking of occupations, the no- status, as it encompasses many other dimensions tion of occupation mobility (either upward or including living standards, prestige, and the like. downward) could be examined. Occupation As such, upward movement within occupations mobility is unevenly distributed in the Maghreb. is a reasonable proxy for changes in individual Algerian migrants are less likely than both Tu- economic well-being. nisian and Moroccan migrants to experience 71   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options upward mobility (17 percent, versus 30 percent Surprisingly, the migrants’ receiving country for Tunisians and 26 percent for Moroccans). has little effect on the probability of having up- They are also more likely to withdraw from ward/downward mobility, once their demograph- the labor market post-migration (36 percent, ics, education, professional status, overseas stay versus 22 percent for Tunisians and 21 percent and return characteristics are accounted for. for Moroccans). Characteristics and Determinants of To understand the forces behind occupa- Entrepreneurship tional mobility in the Maghreb, a probit model was used to disentangle possible determinants Migration and return can be an important mecha- of occupational mobility between pre- and post- nism for entrepreneurship. With (often) financial migration periods. The estimation results for capital, business knowledge, and social networks upward mobility are presented in Table 4.6 (the created abroad, return migrants possess many of results for downward mobility can be found in the key ingredients to small business creation, Appendix 9). which can be a vital source of both growth and employment creation in the home country. The estimation of occupational mobility re- Indeed, given many of the problems with the veals some important results. availability of credit in North African countries, the return of migrants with savings might be a First, and most significant, mobility is in large particularly important means for small business part conditioned by the migrants’ initial positions investment (Figure 3.4). in the distribution of occupations. Those return- ees who were family workers prior to migration More generally, understanding what fa- or small, self-employed workers were much more cilitates self-employment and small business likely to experience upward mobility than those formation among return migrants is important who had secured part-time or short-term jobs to governments and international organizations, (Group 2). Thus, and not entirely surprisingly, which aim to target schemes to assist return the greatest potential payoff to migration oc- migrants in business development. curs for those at the bottom of the occupational spectrum. As with occupational mobility, to understand the determinants of business creation among Secondly, a high amount of remittances is sig- return migrants, a probit version of a discrete nificantly and positively associated with upward choice econometric model was utilized to esti- mobility for North African return migrants. The mate the probability of entrepreneurship since result suggests that the higher the remittances, return. The methodology and full results can be the lesser the budget constraint after return, found in Appendix 10. From the econometric an important determinant of entrepreneurship analysis, and also from descriptive analysis of behavior. the surveys, certain salient features about North African entrepreneurship emerge (Table 4.7).64 Third, the conditions of return play an impor- tant role in the North African migrants’ profes- First, one third of returnees did invest in sional trajectories. In particular, those migrants projects and businesses after return although who returned for administrative reasons are less likely to experience an upward mobility. The ef- 64 Two definitions for “entrepreneur� are alternatively used. fect is particularly strong in the case of Tunisian In the restricted definition, an entrepreneur is defined as any returnees. “Forced� returns are indeed likely to individual who is either: an employer, a regular self-employed, be unprepared returns and to negatively affect or an irregular self-employed with at least one employee. In the extended definition, an entrepreneur is defined as any individual the migrants’ professional reintegration in their who is either an employer, a regular self-employed, an irregular home country. A symmetric result is found in the self-employed with at least one employee, or anyone who in- model of downward mobility. vested in a project hiring at least one employee.. 72 Chapter 4: Further Contributions of Migrants to Host-Country Development, While Abroad and At Return this share strongly varies between countries. positions that did not allow them to acquire any Algeria clearly stands apart, with both a lower entrepreneurial skill. In addition, Algerian re- share of returnees being either employers or self- turnees are older on average and most of them employed and a lower share of returnees being are now retired. investors. This lower propensity to invest partly results from the fact that a significant share of Second, entrepreneurs among returnees Algerians within the sample went to France as are on average different in some ways from early as in the 1960s and occupied low-qualified non-entrepreneurs: they are more likely to be Table 4.6: Probit model of upward occupational mobility after migration with selection Pooled sample Algeria Morocco Tunisia Individual characteristics Female 0.139 0.062 –0.153 –0.589 Age at time of migration 0.016 –0.003 0.004 0.015 Born in urban area 0.048 –0.069 0.101 0.240 Bi-national 0.144 –0.425 0.157 0.947* Family status prior to migration Married before migration –0.233 0.099 –0.725 –0.906** Household size before migration 0.005 0.006 0.018 –0.001 Occupational status prior to migration Was in Group 1 –0.440* –0.531 –1.233* –0.352 Was in Group 2 0.529** 0.608* 0.328 0.741* Was in Group 3 2.705*** — 3.038*** — Characteristics of overseas stay Worked during last migration –0.044 –0.211 –0.195 –0.477 Trained during migration 0.158 0.287 0.026 –0.068 Diploma recognized (1=yes) 0.229 0.186 0.210 0.255 Migration duration (years) 0.016 –0.045*** –0.055** 0.017 Past remittance behavior Sent less than €500 a year –0.057 –0.377 –0.476 –0.693* Sent between €501 and €1000 0.111 0.532 –0.249 –0.100 Sent more than €1000 a year 0.393** 0.726 0.744 0.346 Return conditions Time elapsed since return 0.048*** –0.024 –0.005 0.056 Back to birth place –0.002 0.376 –0.229 –0.263 Returned for administrative reasons –0.719*** –0583 –0.774 –1.043*** Location after return is capital 0.133 –0.950* 1.558** –0.711* Location after return is secondary city 0.042 –0.339 1.458** –0.514 Algerian returnees –0.073 Moroccan returnees –0.032 Tunisian returnees Source: From Gubert and Nordman, 2008a, an output of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. 73   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options male, are younger, and have medium educa- for returnees with experience as employers or tion levels. self-employed, for those who received vocational training while abroad, and for those who inde- In addition, the probability of becoming an pendently and freely chose to return. Again, the entrepreneur after return is significantly higher conditions of return play a significant role in the Table 4.7: Probit model of becoming an entrepreneur after migration (marginal effects) Algeria Morocco Tunisia Demographic characteristics Female –0.138*** –0.254*** –0.341*** Age –0.002 –0.003 –0.032*** Region of origin: urban 0.050 0.035 0.182** Bi-national 0.271** 0.116 0.160 Education (reference is none) Primary 0.203* 0.225** 0.445*** Preparatory 0.291** –0.170 0.519*** Secondary 0.298*** 0.186* 0.370*** University 0.468*** 0.252** 0.297** Higher diplomas 0.255** –0.034 –0.049 Was an entrepreneur before leaving 0.266*** 0.186* 0.176* Characteristics of overseas stay Trained during migration 0.087 0.209** 0.179** Duration of last migration (years) –0.001 0.008 0.019*** Sent less than €500 a year 0.108 0.048 –0.0.25 Sent between €501 and €1000 0.071 0.006 0.153 Sent more than €1000 a year 0.065 0.266*** 0.253*** Conditions of return Time elapsed since return 0.021*** 0.016* 0.053*** Forced return –0.077 –0.195*** –0.168* Plans to re-migrate –0.108** –0.102* –0.113* Back to birth place 0.017 0.116* 0.056 Back to capital city (ref. is small city) –0.132** 0.234*** –0.171** Back to secondary city (ref. is small city) –0.071 0.189* –0.016 Destination country (ref. is France) Germany 0.131 0.032 0.317** North America 0.081 0.194 –0.228 Other Europe 0.108 0.032 –0.064 Spain –0.014 0.063 Italy 0.164 0.229** –0.080 MENA 0.011 Unknown 0.323** –0.251 Source: From Gubert and Nordman, 2008b, an ouput of the EC-Funded World Bank Program of International 74 Chapter 4: Further Contributions of Migrants to Host-Country Development, While Abroad and At Return future contributions return migrants can make form a homogenous group, and some factors may at home. have particularly important influence over how well returnees do upon return. Surprisingly enough, there is no clear corre- lation between migration duration and entrepre- Both job mobility and entrepreneurship are neurship even after controlling for the potential associated with the conditions of return, with endogeneity of migration duration. forced return significantly lowering the prob- ability for returnees to experience occupational Entrepreneurs do not form a homogenous mobility or to start businesses. While the return group, though, and sharp differences emerge of migrants is a cornerstone of European states’ when employers and self-employed are consid- rights to determine who shall enter and remain ered separately. Compared to their self-employed on their territory, the conditions of return may counterparts, employers appear much more have a strong influence over the ability for return educated on average, are more likely to reside migrants to find economic success. Assisted vol- in urban areas after return, have received more untary return programs (IVR), which currently training during their migration stay, and the vast operate out of some 18 countries in Europe, have majority have deliberately chosen to return. been designed to increase return for irregular mi- grants, but to date, the success of these programs Third, projects or businesses owned by is doubtful. The OECD (2008b, 194) concluded: returnees appear to be concentrated in a few sectors. Overall, the wholesale and retail trade Despite the many initiatives and sums sector ranks first, followed by hotels and restau- spent by host countries, assisted vol- rants, agriculture, and the manufacturing and untary return programs are of limited construction sectors. This hierarchy is roughly impact, at least when assessed in the the same in all countries. Most enterprises owned light of the numbers of people involved by return migrants are rather small, with less and in comparison with return flows than 10 employees, and were created using the as a whole. This no doubt reflects the returnees’ own savings. Whatever the country, fact that return is only an option if administrative constraints rank first among the the political, economic and social difficulties faced by investors, followed by exces- situation in the home country is re- sive competition and lack of capital. Turning to stored and stabilized. Yet even in this non-investors, access to financing is considered case, AVR programmes will not make to be the biggest impediment in all three coun- much difference for migrants unless tries, especially in Morocco. financial constraints are the primary barrier to their return. Policy alternatives for North Africa and Europe The evidence suggests entrepreneurship is influenced both by training received abroad and The empirical evidence on return migration is financial resources brought home. This suggests thin, and much greater knowledge will be needed potential interventions by host countries. Small to be able to formulate policies which can ensure business start-up programs with market studies, greater contributions of return migrants to devel- micro-credit, and training components could be opment (as well as facilitate their re-integration tested to ascertain if they could ease constraints into labor markets). But there is initial evidence on entrepreneurship that arise from capital mar- to suggest that North African returnees have an ket imperfections or other market failures, such ability to both experience upward job mobility as lack of information. The French Development and to create small or medium businesses and Agency is presently preparing to launch an ex- generate jobs, factors which clearly impact the perimental program aimed at helping migrants ultimate benefits a country can expect to receive from the Maghreb to set up their business. For from its returned migrants. But returnees do not these programs to be effective, however, local 75   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options governments should also create better conditions the evidence suggests that remittances (or ac- for returnees to integrate and stay in their home cumulated savings abroad) are important for countries. Moreover, training upon return may be future contributions of migrants, in part because equally important (and beneficial). financial markets in the region are weak. Under- standing that, programs and arrangements to Finally, savings—or start-up capital—are es- enhance access to finance among return migrants sential for both entrepreneurship and for upward (but more generally, improving access to finance job mobility upon return. Like the analysis of the in the economies broadly) need to be included relationship between remittances and growth, as key among migrant reintegration objectives. 76 Chapter 5 European Migration Policy, Labor Market Conditions, and Migration Outcomes The development impact of migration is directly an impact on both the incentives for migrants influenced by the experience of migrants abroad. to return, and also the ability for North Africa’s The contributions of North African migrants to migrants—as diaspora and upon return—to con- home-country development over the migration tribute to home country development. life cycle—through remittances, knowledge trans- fer, and investments while abroad, and through Immigrant labor in Spain, one of the coun- occupational mobility, knowledge transfer, in- tries examined more closely under this research vestment, and institution building upon return, program, has contributed positively to growth will be largely determined by host country labor through both increases in the working age market conditions for migrants. These conditions population and increased employment. Although will not only determine the amount migrants can migration has been associated with lower pro- remit, they will also affect the contributions of ductivity (with immigrants primarily working migrants upon return, through both the quality in low-productivity sectors), migration has also of job experience and training received. allowed for strong increases in native female labor force participation. Thus, by taking on low Increasingly, it is understood that the experi- productivity jobs, migrants have opened up the ence of migrants abroad critically depends on the possibility for women to take up work, which in overall policy and institutional environment for turn may enhance productivity in the long term. migrants. Much of the European focus on migra- tion has centered on security, national identity, In both Spain and the Netherlands, policies and vulnerability associated with dependence on to limit less skilled migration have had little immigrants for entire productive sectors. Immi- impact on the flows, but they certainly have had gration is often blamed for high unemployment, repercussions on the status of migrants. In the for draining government welfare programs, and case of Spain, rigid controls on formal labor mi- for increasing inequality in host countries. As a gration have resulted in a burgeoning of irregular result, there is no strong political constituency workers entering the country. In the Netherlands, for mass immigration. At the same time, the need on the other hand, as opportunities for labor for less-skilled immigrants in Europe, based on migration waned, family and asylum migration labor demand, is clear. grew rapidly. In both countries, migration flows have ultimately been largely determined by The dichotomy between stated European labor demand conditions. While policy has not migration policy objectives to control mass immi- achieved its ultimate objectives regarding low- gration and the ambivalence toward outcomes (at skilled migration flows, it has resulted in grow- least regarding less-skilled illegal immigration), ing informality, lower productivity, and reduced has created inherent distortions which have had access to social insurance programs. 77   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Interactions Between Migration the overall migration architecture. Each ele- Policy, Migration Flows and ment can be affected by the unilateral polices of Development Outcomes receiving-country governments, the unilateral policies of sending-country governments, or by Migration flows and the situation of migrants are the negotiated bilateral/multilateral policies of affected by a range of policies, both in sending senders and receivers. and receiving countries. Perhaps most obviously, migrants are affected by migration policies in Controls are the immigration and emigration host countries which dictate who can come and policies that directly determine who is allowed to who cannot, by integration policies, and by bor- move. Receiving country immigration policies are der control and protection policies that affect the the dominant feature of the architecture, but oth- scope for undocumented migration. But migrants er policies matter as well. Policies that affect the are also impacted by a host of policies outside of formation of diaspora networks or their links to actual migration policy, including those relating the home economy (connections); policies that to host countries’ labor markets, social welfare, provide assistance for return or help reintegrate financial sector, international trade, and invest- returnees into the home economy (circulation); ment and development cooperation policies, to policies to invest in human capital in sending and name a few. Combined, these policies—mixed receiving countries (creation); policies to com- with underlying labor market conditions—have pensate sending countries for losses or provide bearing on the types of migration flows, the abil- a share of gains from international migration ity for migrants to participate and benefit from (compensation); and policies to attract or retain growth in host countries, and the contribution international talent (competition), all have a role migrants are able to make to sending country to play in the migration experience. development, both when abroad and at return. Rich-country migration policy features Migration policies in host countries reflect, at least in part, concerns over migrant effects on A review of labor receiving OECD countries the host country. Migration flows can impact host for this migration program66 reveals four basic countries through a variety of channels, including features characterizing migration policy: first, economic output, poverty, labor conditions of na- it is restrictive—only a relatively small fraction tive workers, social protection systems, security, of poor-country residents have the option of be- and cultural and national identity, among others. coming a rich-country resident. It is selective— Many of the effects on host countries are positive, routes for legal access are heavily tilted towards but there are sufficient examples of deleterious the highly skilled. It is rigid—governments effects, as well. The nature of this report pre- generally emphasize permanent movement cludes a thorough review of the complex and over circulation. And it is ambivalent—many changing factors that enter into host country governments combine restrictive polices (espe- migration policy formation. A wide theoretical cially towards the unskilled) with a measure of and empirical literature on both how migration forbearance towards illegal migrants. policy is formulated as well as the impacts of migration on host countries can be consulted The restrictiveness of receiving country mi- for a better understanding of the challenges host gration policy may not be apparent when one countries face in formulating migration policy.65 considers that OECD economies on average have a 12 percent foreign-born population, The architecture of migration The “migration architecture� includes a broad 65 See, for example, Zimmermann 2005; Tranaes and Zimmer- mann 2004. range of policies and arrangements that have 66 McHale, 2009, an output of the of the EC-Funded World Bank impact on the migration experience. Table 5.1 Program of International Migration from Middle East and North details some of the key elements that comprise Africa and Poverty Reduction Strategies. 78 Chapter 5: European Migration Policy, Labor Market Conditions, and Migration Outcomes Table 5. 1: The International Migration Architecture Examples Unilateral Receiving- Unilateral Sending- Bilateral/Multilateral Country Policies Country Policies Policies Controls Immigration policies (e.g., se- Emigration policies (e.g., refusal to Free movement agreements lective issuance of permanent issue passports) (e.g., access policies for EU and temporary visas) accession countries or NAFTA- related TN visas) Connections Cooperation with diaspora Government coordination of diaspo- Inter-government cooperation in organizations in allocating ras; matching subsidies for diaspora coordinating diaspora organiza- development aid investments tions; multilateral lender sup- port of diaspora organizations Circulation Assisted return programs; Targeted return inducements; pro- Free movement agreements; ensuring reasonable probability grams to leverage the human, financial inter-government cooperation in of re-entry; tax withholding and social capital of returnees returning illegal migrants conditional on return Creation Forward-looking education Education investment responses in International student exchange policies to avoid (reoccurring) anticipation of emigration (either more programs “crisis� foreign recruitment; public investment to compensate for policies towards foreign stu- foreseeable losses, or less investment dent recruitment due to lowered social return) Compensation Payroll tax sharing for tempo- A Bhagwati tax on overseas income; Bilateral cooperation in migrant rary migrants; migration-relat- conditional tertiary education funding taxation schemes; migration-re- ed development assistance that becomes repayable on emigration lated multilateral development assistance Competition Tax and benefit policies to pro- Public sector working conditions to Inter-government agreements vide better “value propositions� retain health and education profes- to refrain from active recruit- for attracting international sionals; broader institutional reform to ment of critical personnel such talent; immigrant integration stem talent exodus as health care professionals policies hardly suggestive of closed borders. But legal seen as part of the broader process of attracting access channels to rich-country labor markets and selecting highly skilled permanent movers. do not come close to meeting the demand. The overwhelming majority of the world’s citizens And finally, to some degree, rich-country do not have the option of arbitraging the large migration policies are marked by a certain degree income gaps that exist between poor and rich of ambivalence toward low-skilled, irregular countries—opportunities that many would take migrants. On the one hand, migration policy is if they could.67 In terms of selectivity, English- aimed toward attracting higher skilled individu- speaking, post-graduate degree holders with a als, but on the other hand, countries take few handful of years of relevant professional experi- efforts to make illegal residence less attractive— ence are significantly more likely to have access for example using harsh employer sanctions. to one or more rich-country labor markets. This The ability for low-skilled labor to continue to access is most explicit in the case of countries live and work in rich countries illegally, with with formal points systems, but is also present in migrants living alternatively on the margins of (skilled) employment-based systems. In contrast, society—reflects a general ambivalence toward there is a dearth of access routes for the lesser the tension between the clear demand for low- skilled. Rich-country systems exhibit rigidity skilled laborers in their economies and their because they are primarily designed to facilitate permanent settlement. While temporary access routes are growing in importance, they are often 67 Illegal flows provide one indicator of this unmet demand. 79   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options policy stance designed to exclude them. It is dealing with the actual phenomenon of immi- that ambivalence in receiving country migration gration, especially after the year 2000. This has policy that has perhaps the greatest distortive resulted in the cyclical creation of large pockets effects on achieving better migration outcomes of irregular immigrants, which, from time to time, (both in receiving and sending countries). have had to be incorporated into the system by means of extraordinary regularization processes, Case studies of the migration architecture in as there was no other policy mechanism to deal the Netherlands and Spain reveal a few insights with them on an individual basis. into the dichotomy between de jure migration policy objectives and de facto migration out- Every time in the past 30 years that new im- comes, which have bearing on North African migrant legislation had been passed, there has migration and development. been a corresponding process of regularization of immigrants. From the approval of the National The Migration Architecture and Law of Foreigners in 1985 up until now, there Labor Market Conditions in Spain have been six different episodes of extraordinary regularizations, taking place in 1986, 1991, 1996, During most of the 1990s, Spain was a major 2000, 2001, and 2005 as shown in Table 5.2. All sender of migrants, who headed for Latin Amer- these processes established certain prerequisites ica and Northern Europe. However, since the and set limited application periods for all irregu- turn of the century, the inflow of migrants has lar immigrants. From the year 2000 onwards, been so intense that Spain has evolved into one this process of extraordinary regularization has of Europe’s key immigrant countries. After 2000, become more relevant in terms of the number of the inflows have reached an average of 600,000 applicants and concessions. For example, from new arrivals per year and the number of immi- the year 2000 to the year 2002, when the fourth grants increased from 0.9 million (2.2 percent of and fifth of these extraordinary regularizations the population) in the year 2000 to 4.7 millions took place, the number of residents increased by (10.5 percent of the total population) in the year almost 500,000. In the last of this extraordinary 2007. Such growth rates—a five-fold increase regularization in 2005, four years later than the over seven years—has no parallel in any other previous one, the number of successful appli- of the OECD countries for the last decades. In cants was nearly of 550,000 immigrants. Since fact, 10 percent of all the immigrants to OECD 2000, hence, over one million migrants have been countries for the period 2000–2005 chose Spain regularized in Spain. as their destination country. The—by far— most important category of immigrants is Latin The Spanish experience points to repeated Americans followed by other EU members and failures of successive governments since 1985 to Northern Africa. adjust to the realities of large-scale immigration; large pockets of irregular immigrants continued This large and relatively sudden influx of to exist. Different legal frameworks proved to be migrants has, in turn, meant that the population too rigid and ineffective to deal with the large growth rate in Spain has been higher than at any number of immigrants coming to Spain. This has point in time for the last hundred years. In fact, resulted in large inflows of regular and irregular the Spanish population increased more in the migrants. period 2000–2007 (by 4.6 million) than in the previous two full decades (by 3.4 million people). There are three main reasons why so many immigrants have decided to come to Spain with- As in most European countries, immigration out appropriate residence permits in the past: policy in Spain has mainly been concerned with controlling, rather than managing, the inflow of immigrants.68 The series of different legislatives 68 Appendix 10 has a review of the key phases of migration policy measures have proved relatively ineffective when formation in Spain over time. 80 Chapter 5: European Migration Policy, Labor Market Conditions, and Migration Outcomes Table 5.2: Legislative development and regularization processes in Spain No. Legislative regularized GDP Unemployment Year reform Focus immigrants growth* rate 1986 LOE 1985 • Tight control of inflows 38,181 2.4% 20.6% • Speedy deportation 1991 — • Extraordinary regularization 109,135 3.7% 16.3% 1996 Reglamento • Scale of offenses 21,283 2.5% 21.7% 1996 • Introduction of work and residence permits 2000 LODYLE 2004/4 • Introduction of family reunification 169,157 4.8% 13.5% • Acess to health care and education extended to irregular immigrants 2001 LOE 2000/8 Same as before 310,327 4.5% 10.3% 2005 Reglamento • New legal idea of arraigo laboral (2 years 577,923 3.3% 8.4% 2004 residence and 1 year work) • New legal idea of arraigo social (3 years of residence) • More flexible and realistic recruiting policy at the home countries Source: Conde-Ruiz et al 2008, an output of the of the EC-Funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. Note: See Appendix 11 for review of key Spanish immigration policy phases. *average of the previous 5 years. i) only option—no legal labor migration options legal migration for low-skilled workers, combined existed to respond to labor demand in Spain; with the ambivalence toward their existence in ii) access to the labor markets—immigrants the economy (demonstrated through repeated can find work in the informal sector in Spain; regularizations) and continuing demand for less- (iii) expectations effects due to previous regu- skilled workers, have contributed to the steady larizations—migrants come with the expectation growth in irregular migrants in the Spanish that they will to be able to regulate their situation economy. Without access to formal labor markets, in the next extraordinary regularization process. this has led to increased informalization of the migrant labor market. With regard to integration of migrants into the Spanish economy and society, the legal Estimates of informality, based on employ- framework has been progressively adapted to ment surveys and social security records, sug- protect the rights of both legal immigrants and gest that migrant workers are more prone to irregular immigrants, including through laws be working in the informal sector than natives. penalizing discrimination against immigrants, Latin Americans who reside illegally can clearly as well as ensuring irregular migrants the right not hold legal jobs. Rumanians and Bulgarians to medical assistance, access to education, and have legal access to living in Spain but cannot legal counsel when dealing with the authorities. work in the country due to the moratorium im- posed on recent EU members’ access to labor The nexus between policy, markets, markets. The inherent tension creates pressures and migration outcomes for informal work. It is not possible to empirically link North African But in addition, and surprisingly, many migration outcomes to Spain’s migration policy. North Africans (predominantly Moroccans, but However, some consequences of Spain’s migra- also Algerians), who generally reside legally in tion architecture are evident. Rigid controls on Spain and so unlike the former group would have 81   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Figure 5.1: Difference between the proportion of “irregular residents� and the proportion of “irregular workers� in Spain by nationality, 2007 Proportion of irregular residents minus proportion of irregular workers Morocco Algeria Peru Ecuador Colombia Brazil Bolivia Argentina Romania and Bulgaria –40 –30 –20 –10 0 10 20 Source: Conde-Ruiz, et al. 2008. the right to work, have a relatively high share of the removal of some of the supply constraints informal sector work. Only about 5.5 percent of has had a positive impact on Spanish economic Moroccans were in an irregular situation in Spain growth in recent years. in 2007 (compared with 74 percent of Boliv- ians, 63 percent of Brazilians, and 51 percent of Immigration can impact growth through Argentineans The explanation for this puzzling three major channels: productivity growth, in- result is that North African (mostly Moroccan) creases in the employment rate, and increases migrants, because of their skills profile, tend to in the working age population. Previous studies work in a limited number of sectors where, by of growth in Spain have concluded that immigra- consequence, informality is high. Competition tion has had a positive impact on growth in Spain for jobs therefore forces North African’s legal through the effect on the working age population migrants to accept poorer working conditions, (they are younger) and on the employment rate. including informal jobs. Thus, the inability for The impact on productivity growth (i.e., GDP per Spanish migration policy to stem the large inflows worker) has been negative, however. This is not of irregular migrants impacts the wellbeing of surprising, as immigrants are disproportionately North African immigrants, because when working employed in low-productivity sectors. However, in the informal sector they cannot have access and additionally, immigrant workers have “freed to social insurance (pensions, unemployment up� native workers in Spain to move to more benefit) and active labor market programs. productive sectors, improving native productivity (see Box 5.1 for a summary of the main findings The impact of migration on economic on immigration and growth in Spain). growth The improvements in the demographic factor Migration policy has in contributed to larger (the proportion of the population that is of work- number of irregular migrants than intended in ing age) can be seen in Figure 5.2. In 2007, more Spain, but policy has not been able to stem the than 86 percent of immigrants were between 16 underlying demand for low-skilled migrants. And, and 64 years old, compared to 66 percent of na- with the availability of low-cost, less skilled labor, tives. In the absence of immigration inflows, the 82 Chapter 5: European Migration Policy, Labor Market Conditions, and Migration Outcomes Box 5.1: Migration and Economic Growth in Spain The simplest way to assess the direct impact of im- production per employment unit); ii) employment rate migration on regional and national economic growth -ER- (i.e. ratio between the employed and the working is to examine the development of the three factors age population from 16 to 64 years old); iii) the demo- that completely determine its development. By means graphic factor -DF- (the ratio between the working age of an accounting identity, the per capita GDP can be population and the total population). broken down as the outcome of: i) productivity -Pr- (i.e. GDP PIB L Pop16–64 = . . Pop L Pop16–64 Pop GDP per capita Productivity (Pr) Emplotment Rate (TE) Demographic Factor (DF) The impact of immigration on the per capita income growth rate: different studies Immigration % GDP per Capita Total Demographic F. Employment R. Productivity OEP (2006): 2001–2005 1.60 0.40 0.40 0.20 –0.20 Bank of Spain (2006): 2000–2005 1.70 0.40 0.31 0.27 –0.18 Conde-Ruiz et al. (2008): 2000–20006 1.78 0.05 0.35 0.21 –0.51 The impact of immigration on GDP growth rate: different studies % GDP Total GDP pc Population OEP (2006): 2001–2005 3.10 1.60 0.40 1.20 Bank of Spain (2006): 2000–2005 3.30 1.60 0.40 1.20 Conde-Ruiz et al. (2008): 2000–20006 3.34 1.3 0.05 1.25 The differing results between these studies can be methodology employed to calculate the impact on attributed to differences in the period under examina- productivity. tion (our analysis includes a longer period) and the Source: Conde-Ruiz et al (2008) number of individuals aged between 16 and 24 market occurred in parallel with: i) a decline in would have decreased by as much as 1.5 million unemployment among Spaniards; ii) an spec- in the last seven years. tacular increase in the female activity rate; iii) a stable level of fixed-term contract rate at around But immigrants have also increased growth 30 percent. In fact, unemployment rates fell in Spain through the employment rate. First, between 2000 and 2006 in virtually all Spanish they have increased the total Spanish employ- regions, while at the same time female activity ment rate because they have higher employ- rates increased universally—especially in high ment rate than natives (68.6 percent versus immigration regions. 66.2 percent). This is mainly because almost all immigrants come to Spain to work—the activ- There are several reasons to believe that im- ity rate of immigrants is almost 10 percentage migrants have contributed to the increased em- points higher than natives. But additionally, the ployment among natives. First, immigrants have intensive immigrant inflow to the Spanish labor increased the availability of domestic helpers and 83   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Figure 5.2: Evolution of the Working Age Population in Spain (Natives vs. Immigrants, in millions) Population from 16 to 64 years old Natives from 16 to 64 years old 30 29 28 27 26 1998 1999 2000 2001 2002 2003 2004 2005 2006 Source: Conde-Ruiz, et al. 2008. reduced the cost of this service, thus facilitat- Immigration has a negative impact on la- ing the increase of female labor participation. bor productivity (GDP per worker) in Spain. OEP (2006) has estimated that for every point Conde-Ruiz et al. (2008) show that in regions increase in the immigration rate, the female ac- where growth rate of employment is high (i.e. tivity rises by 0.6 points, and that over one-third regions that receive a lot of immigrants), the of the 12 percentage points by which the female productivity growth rate is low and vice versa. activity rate has increased in the last decade is The elasticity of employment to productivity attributable to immigration. Thus, by taking on growth is –0.68. low productivity jobs, migrants have opened up the possibility for women to take up work, which The accounting explanation is easy; regions in turn may be productivity enhancing. that have received more immigrants (i.e. the average growth rate of employment is high) are Second, immigrants have reduced struc- the same that have experienced a lower produc- tural unemployment (that is, non-accelerating tivity growth. However, the economic interpreta- inflation rate of unemployment (NAIRU)) by tion is more complex. There are two potential reducing the mismatch between labor supply explanations: i) regions that attract relatively and demand. In the context of a reduction in the more immigrants experience falling labor costs; unemployment rate like that observed since the the economy becomes more labor intensive, and mid 1990s, the OEP report (2006) suggests that therefore productivity goes down; or ii) regions the effect of immigration on the NAIRU is large which have adopted technological changes in- and negative, amounting for nearly 2 percentage crease productivity but do not generate labor points’ reduction in the last ten years. demand for immigrants (because immigrants usually do not have sufficient skills or enough Third, immigrants have probably helped human capital to work there). sectors with demand restrictions. Compared to natives, immigrants are more concentrated in The characteristics of labor immigrants differ construction, a sector that was growing faster from those of natives in at least three dimensions than the total economy over the year 2000–2006. that indicate that they are working in lower 84 Chapter 5: European Migration Policy, Labor Market Conditions, and Migration Outcomes productivity sectors and occupations: type of fied. Fernandez and Ortega (2007) similarly contract, wage levels, and job qualification. show that immigrants are much more likely to be overeducated at arrival than comparable na- First, their type of contract indicates they tives. They also find considerable differences in are in lower productivity occupations. The inci- the magnitudes of the gap across nationalities: dence of fixed-term contracts is higher among i) the native immigrants differences are largest immigrants than among natives. According to for East European immigrants, who are also the the LFS, 55 percent of immigrants have a tem- immigrants with the highest educational levels; porary contract compared to 28 percent for na- ii) the over-educational gap is smaller, although tives. Fernandez and Ortega (2007) show that still sizable, for Latin American and African im- the incidence of temporary contracts is much migrants. It is important to point out that by higher for newly arrived male immigrants than education level, the distribution of immigrants for comparable natives, while it is very similar for does not differ too much with respect to na- newly arrived female immigrants and their native tives. Concretely, immigrants from outside the counterparts. The number of immigrant workers European Union (but including Rumania and with temporary contracts has increased by more Bulgaria) are proportionally more likely to have than 1.1 million in the period 2000–2007, but at primary and secondary education and less likely the same time the temporary contract rate of to have tertiary levels of education, but the dif- the economy has remained constant. In all, this ferences are quite small. suggests that immigrants have been substituted for natives in terms of temporary contracts. Thus, while Spanish productivity has declined with immigration, the decline can Secondly, different researchers have also explained by labor immigrants’ characteristics shown that immigrants receive, on average, in the Spanish labor market (low productivity, a wage that is 25 percent lower than natives. precarious jobs, and over-qualification). How- When adjusting for training and experience, the ever, this negative impact may be a short-run differences are not as large, but are still positive. phenomenon. There are, in fact, two channels Simón, Ramos y Sanromá (2007) find that legal through which the immigrant’s productivity immigrants from developing countries exhibit could increase in the future. First, the literature lower mean wages and a more compressed wage on the “assimilation of immigrants� which structure than native-born workers in Spain, and analyzes the speed of “catch up� of immigrants’ that disparities can be explained by their dif- characteristics to those of natives (Adsera and ferent observed characteristics (mostly due to Chiswick 2007; Peracchi and Depalo 2006), has occupational and work-place segregation). More pointed out that the longer that immigrants re- specifically, the Bank of Spain (2005) estimated side in the country, the smaller the differences the unexplained differences to be just 7 percent; between natives and immigrants become, and Simón, et al. (2007) obtained similar results. Ca- that differences disappear altogether after 15 nal-Dominguez and Rodriguez-Guitiérrez (2007) years. Fernandez and Ortega (2007) find that, find that the unexplained component of the wage compared to natives, immigrants have initially difference between native and immigrant work- higher participation rates, but also higher un- ers decreases along the wage distribution, even employment rates, and higher incidence of becoming negative at the end. They also detect over-qualification and of temporary contracts. that the wage gaps between immigrants and na- However, five years after their arrival, participa- tives are largest for the low skill workers. tion rates start to converge to native rates, and unemployment rates decrease to levels even low- Finally, most immigrants are over-educated er than those of natives. Second, the intense im- for the job they are doing. According to OECD migrant inflow experienced in such a short time data, 42.9 percent of immigrants have higher period and fully absorbed by the labor market levels of education than required for their jobs, has reduced the growth rate of the capital/labor while 24.2 percent of the natives are overquali- ratio and consequently the productivity growth 85   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options rate. In the future, as long as the investment in system in a way that labor intensive industries productive capital increases, the productivity are placed in the South instead of in the North. (that is, the wages) of all workers (included the To accomplish that aim, the government would immigrants) will increase as well. strive to liberalize trade policies of developed countries and selectively stimulate industrializa- The Migration Architecture of tion in developing countries. Thus, the Dutch the Netherlands policy in the 1970s was aimed at reducing the push factors for migration in poor countries by The Netherlands immigration experience is simi- improving the work—and income—options of lar to that of some other Western European coun- the potential migrants in their home countries, tries, marked by large scale labor immigration and thereby discouraging immigration in the in the 1960s, temporary workers who became Netherlands. permanent, family reunification beyond 1970s, and subsequent increasing efforts to reduce ir- In the attempt to minimize the number of regular migration. foreign workers, and due to the view that labor migration would be of a temporary character, During the 1960s, the Netherlands concluded more effort was placed on the actual realiza- labor migration agreements with several labor- tion of returning migrants back to their home abundant countries around the Mediterranean: countries. The government proposed to give a Spain, Portugal, Turkey, Greece, Morocco, Yugo- bonus for each worker that would voluntarily slavia, and Tunisia. This resulted in increasing and permanently return. After a strong criti- inflows of both regular and irregular migrants, cism against such a premium, the government and with the acceptance throughout society that dropped the proposal. Nevertheless, attempts growth required foreign labor. to encourage return migration took hold in a number of projects. The 1973 oil crisis had a very strong negative effect on the Dutch economy, especially in sectors From temporary to permanent like manufacturing and construction, both with a residents strong presence of foreign workers. Many guest workers left at that time. However, those from By the end of the 1970s, however, it became poorer countries, including Morocco, had less clear that many immigrants would stay in the incentive to do so, as the economic and political country despite government policy in the in that situation in their home countries was unstable. decade to facilitate return. As a consequence, Although the recruitment policies assumed that the policy at the time could best be classified these workers would leave the Netherlands once as categorical welfare policy, with the aim to their services were not needed anymore, this did provide the necessary services such as housing not happen to the extent predicted. Additionally, and cultural programs to facilitate temporary like any other unemployed person, guest workers stay and later return. As the preservation of were entitled to income-replacing social security immigrants’ own ethnic identity was given benefits, despite being unemployed, providing an special emphasis, the government supported additional incentive to stay. ethnic cultural education and mother-tongue teaching, to facilitate reintegration upon the During this time, migration policy centered migrants’ (expected) return to their home on three themes: limiting labor migration, focus- countries. Also, given the lack of coordinated ing on migrating jobs through trade and devel- central government intervention, many private opment policies, and an idea that policies were institutions were set up, and partially subsidized needed to move the Dutch economy away from by the government, to provide welfare services any dependence on low qualified foreign workers. to the various immigrant groups. Also, employ- The Den Uyl government, for example, set itself ers, in many cases, provided housing for guest an ambitious objective of changing the economic workers and their families. 86 Chapter 5: European Migration Policy, Labor Market Conditions, and Migration Outcomes However, with recognition of the position A number of notable projects were setup of immigrants in the country, debates began in the 1980s and 1990s toward improving mi- about the need to acknowledge permanent grant civic and economic integration, including residence for some immigrant groups and the projects to enhance literacy and educational ac- need to introduce integration policies for them tivities for ethnic minorities and Dutch language (rather than reintegration policies). As many training, and civic orientation classes. ethnic minority groups were facing a number of problems such as a weak legal status, high Since the mid 1980s, the government also unemployment rates, low education, and dis- undertook initiatives to improve the weak labor crimination, they became the target population market position of ethnic minorities. And finally, of an extensive catalogues of policies aimed at during the same decade, the government decided ethnic emancipation. By the early 1980s, the to delegate the task of disbursing residential con- government had redefined its perspective on im- centrations of disadvantaged minorities to local migration, acknowledging that the Netherlands authorities, in the light of the ineffectiveness of had indeed become an immigration country and central housing policies. This has led to the im- there was need for an effective policy providing plementation of housing dispersal policies to fight a regulatory framework for the large populations housing deprivation of among ethnic minorities of foreigners residing non-temporarily in the in numerous municipalities. Despite the general country and facing persistent socio-economic ineffectiveness of central housing policy, a fun- disadvantages. damental change was introduce by the central government through a legislative revision in 1981 The main policy objective was to reduce providing legally residing foreigners full access socio-economic disadvantages of the members to social housing. As social housing accounts for of minority groups and to “emancipate� minor- the larger share of all housing in big Dutch cities, ity groups through the creation of new ethnic this measure improved the housing situation of communities (pillars) in the Dutch society. minorities considerably (Penninx 2006). This involved large-scale government funding of minority organizations, ethnic language and Doors are opened, and difficult to close culture education institutes, and ethnic radio and television broadcasting channels, all with Part of the new integration policy included en- the goal of fostering emancipation. During the hanced opportunities for family unification. Im- 1980s numerous projects on the national as well migration for family reunification and formation as on the local level were started to provide reasons was considered functional to the integra- better educational opportunities, improved tion of the earlier labor migrant. It was thought chances on the labor market and access to social that families were considered to integrate more housing (previously impossible). The over- easily in the local ethnic community than single arching aim was to provide ethnic minorities, men and that these ethnic communities would as separate groups, with an equivalent status become an integral part of the Dutch society. and equal developmental opportunities as the Dutch. The policy rested on the improvement At the same time, the Netherlands also intro- in their legal status and their socio-economic duced much stricter control on labor migration, positions as well as on the preservation of their and later on immigration for family and asylum own cultural and religious practices. It was reasons. Visa requirements were introduced for expected that this positive identification would countries with high immigration potential, such contribute to a better integration of the ethnic as Morocco, Turkey, and Suriname. In the 1990s, groups (and, by implication, of the individuals) family unity was no longer considered as an asset into Dutch society. In return, the immigrants to migrant integration, but more as a curse. The were expected to acquire the basic skill (lan- Law on Integration Abroad (Wet Inburgering in guage) that would allow them to function well het Buitenland) from 2005 clearly sees family in Dutch society. migration as a potential threat to integration. 87   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options The declared goal of the “civic integration test plication for a work permit relates to a specific job abroad,� which is required for migrants entering and requires an individualized labor market test. the country on the grounds of family reunification or family formation, is to improve the integration With regard to irregular migrants, meanwhile, conditions for new migrants. Though the reduc- the growing concerns in the Netherlands about tion of the number of immigrants allegedly is not undocumented migration materialized in “the the primary goal of the policy, the government linkage act� in 1998 which completely excludes still expects a drop in family migration. That, in irregular migrants from public services includ- fact, happened when the number of applicants ing housing, secondary or higher education, and for family formation dropped significantly in health care, by preventing them from obtaining 2006. a social security number (which is a prerequisite for regular employment, social security benefits, As for labor migration, the basic policy of and subsidized housing). the Dutch government (since the turn in 1973), is to not allow permanent settlement unless How migration has adapted to changing there are compelling reasons to do the opposite policies (Doomernik 2005). Different from France and Germany where a formal recruitment stop was Like Spain, the Netherlands presents an example introduced, labor migration continued to be pos- of a country struggling to adapt to migration sible even after the Netherlands proclaimed itself flows. Migration policies have been largely reac- closed to labor migration. First, immigration was tive, with economic and political realities driving regulated by the Foreign Employees Act from migration flows, and migration and integration 1979 to 1995 (Wet Buitelandse Werknemers) policies readjusting to these realities with a and from 1995 onwards through the Foreign somewhat trial-and-error approach. It also points Workers Employment Act (Wet Arbeid Vreem- to the problems associated with developing the delingen) (Bruquetas-Callejo et al. 2007). The right mix policies for circular migration, balanc- admission was a temporary matter and only the ing the desire for circularity with the desire for workers that were already in Europe were ad- integration. mitted. Employers were only allowed to employ workers from outside the European Economic On the migration flows front, the most ob- Area (EEA) if they could prove that suitable vious feature of migration policy has been its employees could not be found within this area inability to reduce migration from its growth- (Roodenburg et al. 2003). Although the end of determined trajectory. Looking at immigration labor immigration was proclaimed, the above- and GDP growth in Netherlands, changes in the mentioned policies continued to channel the number of immigrants69 closely follows the busi- entrance of those workers that were considered ness cycle (see Figures 5.3 and 5.4). When the beneficial to the Dutch labor market. economy is growing, the number of immigrants increase, and the decreases when economic The main policy instrument that currently growth is depressed. That indicates that labor regulates employment of “third country nation- demand remains an important factor driving als� is the Foreign Workers Employment Act the number of immigrants in the country. The (WAV). The Act is entirely demand driven, mean- institutional environment may change the way ing that employers are responsible for obtaining in which migrants enter, but it has far less abil- the work permit. Preferential treatment is given ity to stem the number of immigrants entering to “priority labor supply.� which means that a the country. This casts a peculiar light on the third country national can only be hired when intended government policy to make migration the employer can demonstrate that every pos- more selective. It may be that the government sible attempt has been undertaken to employ a person of Dutch nationality or the nationality of one of the countries within the EEA. Every ap- 69 lagged two years. 88 Chapter 5: European Migration Policy, Labor Market Conditions, and Migration Outcomes here overestimates its influence to limit immi- On the integration front, the objectives of gration under the condition of an excess labor migration policy in the Netherlands have been demand. in many ways successful. The socio-economic Figure 5.3: Immigration in the Netherlands (per 1.000 inhabitants)* and GDP growth 1980–2007 GDP Growth Immigration .009 5 4 .008 Immigration (share population) 3 (two years lagged) GDP growth (%) .007 2 1 .006 0 .005 –1 .004 –2 80 82 84 86 88 90 92 94 96 98 00 02 04 06 Source: de Neubourg, Chris. 2008. “Changes in Immigration in the Netherlands: Trends, Policies and Incentive.� An output of the EC-funded World Bank program of International Migration from the Middle East and North Africa and Poverty Reduction Strategies. Figure 5.4: Immigration in the Netherlands (per 1.000 inhabitants)* and GDP growth 1960–1980 GDP Growth Immigration .008 9 8 Immigration (share population) 7 .007 6 (two years lagged) GDP growth (%) 5 .006 4 3 .005 2 1 .004 0 60 62 64 66 68 70 72 74 76 78 80 Source: de Neubourg, Chris. 2008. “Changes in Immigration in the Netherlands: Trends, Policies and Incentive.� An output of the EC-funded World Bank program of International Migration from the Middle East and North Africa and Poverty Reduction Strategies. 89   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options position of the non-OECD migrant population is cent less than among the indigenous population, generally a lot worse than that of the indigenous but well above the earlier rates. While educational population on whatever indicator is chosen to disparities between second-generation migrants describe the differences (as is illustrated for and Dutch natives remain, the convergence has low-income incidence in Figure 5.5). Since the been impressive (Figure 5.6). recognition, however, that migrants are becoming permanent residents in the country in the mid- The policies aimed toward socio-economic 1970’s, and with increasing policies to reduce equity, however, have had less success on the socio-economic disadvantages, the situation has employment front. Employment rates of migrants gradually been improved. Given the permanently were generally 13 percent below the overall em- growing stream of additional immigrants to ployment rates during the entire period between absorb, this in itself signals that the integration 1996 and 2006. Unemployment rates are very policies that have been followed since the mid- high and in the order of magnitude of 15–16 per- 1980’s had some effect. The situation has also cent which is well above the overall employment improved considerably with regard to housing. of 5–7 percent for the same period. Educational disparities have also been greatly What Spain and the Netherlands reduced. Only 8 percent of the Dutch population Tell Us about Migration Policy finished primary education as their only degree, compared with 31 and 34 percent of Turkish and Spain and the Netherlands are comparable Moroccan migrants, respectively (in 2005). This studies in the efforts to manage migration flows is down from more than 50 percent only 10 years through migration policy. In neither country have earlier. Significant progress is seen among the sec- flows varied far from the trajectory of labor de- ond generation, of whom almost 20 percent held mand. Migration policies have been largely reac- a higher education degree in 2006; still seven per- tive: economic and political realities have driven Figure 5.5: Share households with low income (households disposing up to and including 100% of the low-income threshold), per ethnic group, the Netherlands, 2000–2005 (in percentages) 2000 2001 2002 2003 2004 2005 40 35 30 25 20 15 10 5 0 Turkish Moroccan Surinamese Antillean Other Non-western Natives non-western total Source: CBS Statline, 2007; SCP, 2007 Note: Years displayed in chronologic order from left to right. 90 Chapter 5: European Migration Policy, Labor Market Conditions, and Migration Outcomes Figure 5.6: Educational Catch-Up Among Migrant Communities in the Netherlands Higher Secondary II Secondary II Basic 100 90 80 70 60 50 40 30 20 10 0 1sr gen 2nd gen 1sr gen 2nd gen 1sr gen 2nd gen 1sr gen 2nd gen Dutch Turkish Moroccan Surinamese Antillean Source: de Neubourg, et al. 2008, an output of the EC-funded World Bank Program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. migration flows, and migration and integration on the status of migrants. In the case of Spain, policies have readjusted to these realities with rigid controls on formal labor migration have a trial-and-error approach. Both countries have resulted in a burgeoning of irregular workers also been marked by a certain ambivalence to the entering the country. In the Netherlands, on the issue of low-skilled migration (from North Africa other hand, as opportunities for labor migra- and elsewhere): while both economies have need tion waned, family and asylum migration grew unskilled workers, the political realities of mass rapidly. In both countries, the rigid migration immigration of low-skilled workers has resulted architecture has resulted in increases in irregular in repeated efforts to contain their flow. migrants, working in lower productivity, informal jobs, with poorer working conditions. This has In the end, policies had little impact on the in all probability reduced gains in the welfare flows, but they certainly have had repercussions of migrants. 91 Chapter 6 Demographic Forces and the Implications for North Africa While Chapter 3 outlined the impact of migration The direct impact of demographic change oc- on North African economies over the last decade, curs through the so-called dependency ratio. The the migration landscape is being rapidly changed total dependency ratio (TDR) is the ratio of the by global demographics, the result of which makes economically dependent part of the population increased migration axiomatic. How North Africa to the productive part. The economically depen- (and MENA more broadly) prepares for these large dent part is recognized to be children that are demographic forces will strongly impact the ulti- too young to work, and individuals that are too mate costs and benefits of migration for the region. old; that is, generally, individuals under the age of 15 and over the age of 65. The total dependency Demographic Changes and the ratio is important because as it increases, there Opportunities for Arbitrage is increased strain on the productive part of the population to support both the upbringing of the The world population is going through significant young and the pensions of old. There are direct shifts that will affect labor demand and, ulti- impacts on financial elements like social security. mately, the growth and welfare prospects of all re- gions. Some regions and countries, most notably Figure 6.1 shows that the total dependency Europe, but also North America, China, and other ratio has been higher in the MENA region than in East and Central Asian countries, will experience the EU since 1950 and is still expected to remain significant reductions in their available labor higher until 2020. After 2020, the rise in life ex- force in the coming decades. A general decrease pectancy will push the EU ratio upward while the in fertility along with reduced mortality rates is drop in fertility rates will push the MENA ratio causing a progressive and inescapable ageing, or downwards. The European results are similar graying of European nations. Without increases whether one focuses on the EU15 EU1571 or the in migration, these countries will lose hundreds of EU27.72 In the same vein, the demographic evolu- millions workers between now and 2050. tions observed in North Africa and the broader MENA region are almost identical. On the other hand, the countries of North Africa and the broader MENA70 region share a 70 For the purpose of the demographic analysis which follows, the relatively young population. Working-age popu- MENA region is comprised of the focus North African countries lation growth is high, putting strong pressures (Morocco, Algeria, Tunisia and Egypt) as well as Djibouti, Iran, on labor markets, with rapid growth of new Lebanon, Libya and Malta. 71 EU15 comprises Austria, Belgium, Denmark, Finland, France, entrants to the work force. However, as fertility Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, rates continue to fall, this phenomenon will pass. Portugal, Spain, Sweden, and the United Kingdom. After 2030, these young entrants will be retiring, 72 EU27 comprises Austria, Belgium, Bulgaria, Cyprus, Czech and MENA emigration countries will begin to Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, confront the same difficulties that Europe faces the Netherlands, Poland, Portugal, Romania, Slovakia, Slovenia, today, with rapidly aging populations. Spain, Sweden and the United Kingdom. 93   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Figure 6.1: Total dependency ratio (TDR) by region (1950–2050) EU15 EU27 MENA4 MENA9 100% 90% 80% 70% 60% 50% 40% 30% 10% 10% 0% 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 Source: Docquier and Marchiori 2008, an output of the EC-funded World Bank program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. All of the countries in Europe will experi- be more severely impacted. In the Magrheb ence a similar demographic transition, although and Egypt, meanwhile (MENA4, below), the some countries, including Italy, Spain, Austria, graying of the population will occur after 2030 Germany and Portugal (in the EU15), will (Figure 6.2). Figure 6.2: Aged dependency ratio (ADR) in North African and other MENA countries (1950–2050) Algeria Egypt Morocco Tunisia Others 80% 70% 60% 50% 40% 30% 10% 10% 0% 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 Source: Docquier and Marchiori 2008, an output of the EC-funded World Bank program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. 94 Chapter 6: Demographic Forces and the Implications for North Africa Figure 6.3: Aged dependency ratio (ADR) by consolidated region EU15 EU27 MENA4 MENA9 60% 50% 40% 30% 10% 10% 0% 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040 2050 Source: Docquier and Marchiori 2008, an output of the EC-funded World Bank program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. It is useful to also consider the evolution social protection systems, in particular, health of total and aged dependency ratios for con- care and pensions.75 Governments then face solidated regions, i.e., aggregating EU and North hard choices between higher taxes (including Africa (MENA4)73 or the broader MENA region a possible reweighting of taxes from earnings (MENA9).74 In a sense, this consolidation pro- to consumption), increasing debt (which could vides the full picture of dependency ratios if there jeopardize the welfare of future generations) and were full labor mobility. Clearly, this consolidation a reduced government role in providing health smoothes the trends and attenuates the rise in care and social security. It is also clear that the dependency; however it is also clear North African demographic opportunity for MENA to help ease countries could not hope to fill the labor market the labor shortages in Europe is limited: MENA gaps in Europe. With a total population of around countries themselves will begin to experience 137 million, North African countries have only about serious demographic problems after 2030. In- 27 percent of the EU27 population (500 million). deed, the MENA aged-dependency ratio in 2050 Thus, there would remain a large increase in the will be approximately equal to the current aged- dependency ratio even with full labor mobility with dependency ratio observed in the EU. North African countries, and even with full mobility between the broader MENA region and Europe. While demographics fit, the skills do not The figures confirm that increasing migration However, a recent World Bank study points to flows from North Africa or the broader MENA significant disparities between projected Eu- region to EU27 countries would clearly attenuate ropean demand and MENA skills.76 Projections the deterioration of the European demographic structure. The magnitude of the imbalances 73 MENA4 includes Morocco, Algeria, Tunisia and Egypt. is such that migration into Europe and other 74 MENA9 includes Morocco, Algeria, Tunisia, Egypt, Djibouti, deficit regions is almost certain to intensify into Iran, Lebanon, Libya, and Malta. the future. In the absence of immigration, the 75 Holzmann 2005. 76 World Bank, 2009. Shaping the Future: A Long-Term Per- European region’s shrinking and aging popula- spective of People and Job Mobility for the Middle East and tion will impose new challenges on Europe’s North Africa. 95   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options over the short-, medium-, and long-term suggest board and cover workforce replacement and that labor shortages in Europe will occur across expansion (Figure 6.5). the skills spectrum, with significant demand for medium-level skills (such as nurses or inter- Based upon current education levels, how- mediate business services) or even relatively ever, MENA countries (including North Africa) low skills (retail sales persons, waiters, etc.). will provide a poor match to the projected skill A study by CEDEFOP “Future Skills Needs needs. Since most of the additional population in Europe—Medium-Term Forecast�77 for the will have primary education levels, only a frac- period 2006—15, projects a drastic increase in tion of MENA’s additional work force would be qualification requirements across most jobs in of interest to European countries. Europe (including the EU-25, Norway, and Swit- zerland) with more than 12.5 million additional Moreover, if education levels in the North jobs required at the highest qualification level Africa labor force remain low, while the demand across all occupational groups (see Box 6.1 for for skills increases significantly across the Medi- definitions) and an additional increase of about terranean, there is the considerable risk that 9.5 million jobs required at the medium level of the region will lose scarce talent, expertise, and qualifications. its educational investments to Europe—the so called “brain drain.� By contrast, jobs at low qualification levels are expected to fall by 8.5 million across all oc- Although projections suggest that medium- cupational groups (see Figure 6.4). level skills will also experience significant increases in demand, the EU has agreed to The projections take into account demand union-wide Blue Card for highly skilled non-EU expansion as well as replacement demand (from nationals, and a number of European countries attrition). In all categories, replacement demand have recently introduced programs aimed at exceeds demand from expansion, a clear reflec- attracting a more qualified labor force. A more tion of population aging and attrition of the selective European immigration policy may labor force. In terms of occupation, because of over time increase skilled emigration from the expanding demand for higher skills, demand will MENA region, especially from North Africa. And be high for technical and professional occupation although, as noted in Chapter 3, there are poten- groups. But because of attrition, there will also be tial positive externalities on the effects of skilled significant demand for service workers and shop, migration (through remittances, brain gain, and market, and sales workers as well as elementary occupations; however, the demand for formal See http://www.trainingvillage.gr/etv/Information_resources/ 77 qualifications is expected to increase across the Bookshop/publication_details.asp?pub_id=488. Box 6.1: CEDEFOP Definition of Occupational Groups Highly skilled, nonmanual occupations: Managers, workers, plant and machine operators, and assem- professionals, technicians, and associate professionals blers Skilled nonmanual occupations: Clerical support Elementary occupations: Cleaners and helpers, workers, service, and sales workers agricultural workers, laborers in mining, construction, manufacturing and transports, food preparation as- Skilled manual occupations: Skilled agricultural, sistants, street and related sales and services workers, forestry, and fishery workers, craft and related trade refuse workers, and others Source: World Bank 2009. Shaping the Future: A Long-Term Perspective of People and Job-Mobility for the Middle East and North Africa. 96 Chapter 6: Demographic Forces and the Implications for North Africa Figure 6.4: Expected Employment Changes in Europe 2006–2015 by Occupational Groups Net Employment Change (in millions) and Highest Level of Qualification, EU-25, Norway, and Switzerland High Medium Low High-skilled non-manual occupations Low-skilled non-manual occupations Skilled manual occupations Elementary occupations –5 0 5 10 Source: World Bank 2009, Shaping the Future: A Long-Term Perspective of People and Job Mobility for the Middle East and North Africa. Note: Projections for EU24, Norway and Switzerland. Figure 6.5: Increase in Occupational Demand in Europe Broad Groups (change in millions), 2006–2015 Expansion Demand Replacement Demand Total Requirement Legislators, senior officials and managers Professionals Technicians and associate professionals Clerks Service workers, shop/market sales workers Skilled agricultural and fishery workers Craft and related trades workers Plant and machine operators and assemblers Elementary occupations –5 0 5 10 15 Source: World Bank, 2009; Shaping the Future: A Long-Term Perspective of People and Job Mobility in the Middle East and North Africa. Note: Projections for EU24, Norway and Switzerland. 97   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Figure 6.6: Educational attainment of adult population in North Africa, 1960–2000 Upto post-secondary Upto secondary Upto primary No education 100% 90% Proportion of total population age 15+ 80% 70% 60% 50% 40% 30% 20% 10% 0% Algeria Algeria Algeria Egypt Egypt Egypt Tunisia Tunisia Tunisia 1960 1980 2000 1960 1980 2000 1960 1980 2000 Source: Barro-Lee 2000. contributions through return migration), sizeable and fiscal balances in both regions—a computable increases in skilled migration could overwhelm general equilibrium (CGE) model with overlapping the potential for positive feedback. generations of individuals who are heterogeneous in skills, date and place of birth provides some Scenarios of How Rising Migration insight. General equilibrium analysis can only may Impact North Africa provide scenarios, and results are determined by the assumptions underlying the model. However, The impacts of increasing migration from North the scenarios are useful for modeling the types of Africa to Europe over the coming decades could impacts the region (and Europe) can expect from be potentially large, both for sending countries large-scale increases in migration (whether they and for host countries. Because the contributions be low-skilled or high-skilled, and whether they of workers (both fiscally and in terms of output) are accompanied by other feedbacks and policies). are heterogeneous, the results will depend on skill level of migrants and the way in which they The model employed here allows for hetero- are employed in Europe (particularly whether geneous use of skills abroad, changes in educa- the skills of higher-educated workers are utilized tional attainment emerging from the prospect or wasted in low-skilled jobs), with potentially of migration (brain gain), the remittance of a significant impacts in terms of both output and fraction of consumption by migrants, potential fiscal balances. Immigrants induce many effects diaspora and returned migrant contributions, on the welfare of the population in the host coun- and fiscal changes determined by taxes on labor try, through the capital market, the markets for earnings and consumption of pension benefits, skilled and unskilled labor, and through public welfare transfers, and general consumption in- finance. These direct effects can be reinforced cluding schooling, both in Europe and in MENA.78 by host country responses. To better understand what this demographic 78 With North Africa accounting for about 80 percent of MENA- to-EU migration, multiplying the MENA deviations from the transition might mean for MENA and North Africa baseline by 1.25 can approximate the impact on each North as well as Europe—in terms of output, productivity, African country. 98 Chapter 6: Demographic Forces and the Implications for North Africa The full details of the model and methodology be- impacts in ways that need to be considered hind the scenarios are included in Appendix 12. through a dynamic and endogenous process. Obviously, the additional migration that will be The driving forces behind potential outcomes drawn from MENA (and North Africa) will to come from (1) who migrates (high-skilled or low- some degree involve bits and pieces of all sce- skilled), (2) how those skills are utilized abroad, narios. The ultimate impact will depend on which (3) whether there are any positive feedback of these outcomes dominate in the MENA case. channels from migration (such as remittances), and (4) whether policies are employed to change The results the underlying assumptions on education. As a result, five basic scenarios were considered: Effects on taxes • The “Skilled� scenario, assuming the ad- The models collectively suggest how increased ditional migrants arriving from MENA to the MENA-to-EU migration would generate signifi- EU between 2010 and 2050 are skilled and cant changes in EU tax rates and GNI per capita, are employed as skilled workers in the EU significantly smoothing the fiscal and economic (i.e., perfect assimilation). burdens of aging. • The “Unskilled� scenario assumes that all additional migrants are unskilled and em- The analysis of skilled versus unskilled ployed as unskilled workers in the EU. migration (both utilized effectively in the host country) reveals that selecting immigrants has • The “Brain Waste� scenario assumes that only a moderate impact on host country tax rates. all additional migrants are skilled and em- Figures 6.7 and 6.8 present the resulting tax rates ployed as unskilled workers in the EU. compared to the baseline, where it is clear that • The “Brain Gain� scenario, where skilled immigration shocks reduce the tax rate in the EU workers are employed as skilled workers in and raise it in MENA. At destination, the “Skilled� the EU, and where there are also ex-ante scenario is unsurprisingly the most effective one effects on the human capital formation at to relieve the pressure on the pension systems home (as well as various diaspora exter- since skilled individuals contribute more to, and nalities). benefit less from, welfare programs. MENA on • and finally, the “Expansionary� scenario,� the other hand would have to cope with a higher where the additional migrants arriving from tax rate if it loses part of its labor force, and the MENA are skilled and employed as skilled “Skilled� scenario will have the worst fiscal impli- workers, and where simultaneously the cations. However, Figures 6.7 and 6.8 also show MENA region achieves substantial increases that the difference between these two scenarios in the level of education at home over the is relatively small, explained by induced effects period (while in the baseline, the proportion on wage rates, skill premiums and interest rates. of skilled among youngest cohort is constant When the additional immigrants are skilled, the at 15 percent from 2000 onwards, and thus skill premium decreases, and this reduces the the proportion of skilled in the workforce and average contribution of a skilled worker in fi- in the population reaches 15 percent in the nancing pension systems. Conversely, when the long run, in the expansionary scenario, the additional migrants are unskilled, a skilled worker share of skilled among the youngest cohort will pay more taxes since the skill premium is continues to rise after 2000 to reach 20 per- enhanced. Moreover, when additional migrants cent in 2030, with the consequence that 20 come to Europe (or leave MENA), it leads to an percent of the population will be educated increase (decrease) in GDP and thus in the debt. in the long run). Since GDP rises more rapidly under the “Skilled� scenario than under the “Unskilled� scenario, the These scenarios are useful only in the sense public deficit to be financed will be larger when that they can provide insight into the potential additional migrants are skilled. 99   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Figure 6.7: Tax rate in EU15 – “Skilled� vs “Unskilled� (in percent) Baseline Skilled Unskilled 60% 55% 50% 45% 40% 2000 2020 2040 2060 2080 2100 2120 2140 Source: Docquier and Marchiori 2009, an ouput of the EC-Funded World Bank program of International Migration from the Middle East and North Africa and Poverty Reduction Strategies. Figure 6.8: Tax rate in MENA – “Skilled� vs “Unskilled� (in percent) Baseline Skilled Unskilled 35% 30% 25% 20% 15% 2000 2020 2040 2060 2080 2100 2120 2140 Source: Docquier and Marchiori 2009, an ouput of the EC-Funded World Bank program of International Migration from the Middle East and North Africa and Poverty Reduction Strategies. Effects on GNI per capita EU to MENA leads to higher returns to capital in Europe. Since Europe is technologically more While tax rates in Europe are only marginally advanced, labor outflows from MENA induce a affected by the selection of migrants, selection more than proportional reallocation of physical has a strong and positive impact on Europe’s capital to Europe. Thus, increased migration GDP per capita. The reallocation of labor from to the EU acts to enhance GDP per capita (see 100 Chapter 6: Demographic Forces and the Implications for North Africa Figure 6.9). This effect is particularly strong if amounts is equivalent across scenarios. As a re- the migrants are skilled and employed as such at sult, however, increasing unskilled emigration in destination (“Skilled� scenario). This is obviously MENA leads to improvements in GDP per capita, due to the fact that skilled workers are more pro- which increasing skilled emigration continues to ductive. In addition, the agglomeration of skilled induce detrimental consequences. labor in EU increases the marginal productivity of capital in this region. Variants of the skilled migration scenario alter the results. A “Brain Waste� variant (in In MENA on the other hand, migration which skilled migrants are employed as unskilled reduces GDP per capita, particularly when mi- at destination) combines the worst effects of the grants are skilled, in the absence of any feedback selective and non-selective shocks, since MENA effects (Figure 6.10, top). However, when remit- loses its most productive workers, who are em- tances are included in the analysis, the negative ployed as unskilled in Europe and thus contribute impact is mitigated, with almost equivalent poorly to the EU’s economy. In a second variant impacts whether migrants are high-skilled or accounting for positive externalities of the brain low-skilled—the explanation hinging on general drain, the negative impact of skilled migration equilibrium effects. The selection of immigrants on sending countries is mitigated (in terms of affects the skill premium at destination. Under effects on income per capita and inequality [not a “Skilled� scenario the skill premium drops and shown here]). Still, MENA would suffer from an skilled individuals earn less while unskilled earn enlarged brain drain. more than in the baseline. Thus, the amount of remittances sent back by a skilled migrant Finally, there is another possibility in which is lower than in the baseline and the one of an MENA countries prepare for greater skilled mi- unskilled migrant is higher than in the baseline. gration to Europe by accelerating their human The reverse is true under an `unskilled’ migration capital formation. In this case, a skilled emigra- shock. Consequently, the increase in remittance tion shock could go along with a rise in education Figure 6.9: GDP per capita in EU (percent deviation from the baseline) – “Skilled� vs “Unskilled� Skilled Unskilled 6% 5% 4% 3% 2% 1% 0% –1% 2000 2020 2040 2060 2080 2100 Source: Docquier and Marchiori 2008, an output of the EC-funded World Bank program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. 101   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Figure 6.10: GDP per capita in MENA (percent deviation from the baseline) – “Skilled� vs “Unskilled�: with and without diaspora effect Skilled Unskilled Without remittance effects 1% 0% –1% –2% –3% –4% 2000 2020 2040 2060 2080 2100 With remittance effects 1% 0% –1% –2% –3% 2000 2020 2040 2060 2080 2100 Source: Docquier and Marchiori 2008, an output of the EC-funded World Bank program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. levels of the MENA population and income per the Mediterranean, but at an accelerating speed. capita (Figure 6.11 and 6.12). However, if North African countries are to pro- vide a rapidly increasing skilled work force, the Policy implications for North Africa current situation poses a significant threat to these countries. The demographics and migration ties that exist between North Africa and Europe suggest that The threat is that the skills pools in send- it is likely that many of the migrants Europe will ing countries becomes entirely depleted, with need will continue to come from the South of implications for growth, tax income, an so forth. 102 Chapter 6: Demographic Forces and the Implications for North Africa Figure 6.11: GNI per cap. in Europe – Brain drain variants (in % dev. from baseline) Brain Waste Skilled = Brain Gain = Expansionary 5% 4% 3% 2% 1% 0% –1% 2000 2020 2040 2060 2080 2100 Figure 6.12: GNI per cap. in MENA – Brain drain variants (in % dev. from baseline) Brain Gain Skilled = Brain Waste Expansionary 6% 5% 4% 3% 2% 1% 0% –1% –2% –3% –4% 2000 2020 2040 2060 2080 2100 Source: Docquier and Marchiori 2008, an output of the EC-funded World Bank program of International Migration from Middle East and North Africa and Poverty Reduction Strategies. Although skilled emigration, as discussed, can in general have not delivered a highly educated carry many benefits, a scenario where all or a and employable labor force so far. The problem majority of persons with education leave their relates to both quality and to quantity. Although country must have severe and negative implica- the region has made impressive efforts at ex- tions on the economy. panding access—the average age of schooling has increased rapidly, and alphabetization efforts are However, as discussed in a recent report paying off—the MENA region generally lags be- (World Bank 2008), MENA’s education systems hind other regions, especially for young females 103   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options (by 2005, two in five Moroccan young women, in demand. Such efforts are also vital to ensure that and one in five Egyptian young women were still the input into higher levels of education through illiterate). As discussed earlier, the prospects of basic and secondary levels is adequate. government employment and the lack of a dynamic private sector have introduced a disproportionate Moving ahead on this ambitious agenda preference for humanities and softer skills at the by addressing the weaknesses outlined above, expense of technology and management, and the will be essential for improving outcomes in the quality problems in education are witnessed in high region, but also for providing prospective mi- unemployment rates for graduates, “brain waste� grants with the necessary skills to ensure higher abroad, falling returns to education, and poor per- productivity and higher rates of employment formance of students in international comparisons. abroad. However, there will also be a need for specific partnerships in order to provide a better Currently, education systems do not neces- match between skills demanded abroad and skills sarily provide incentives for improving perfor- provided by the education systems. mance. Monitoring mechanisms are lacking, whether at the level of institution or instructor. European countries stand to win much from The public sector concentrates on service de- improved education systems in the North African livery rather than quality control, there is little countries and in MENA overall. The reform agenda autonomy, private-public partnerships remain is ambitious, however, and it seems only reasonable rare, and the general public and more specifically that all countries that wish to draw on the MENA the consumers of education services, do not have work force intensify collaboration in this area. the possibility to influence the education agenda. Skills that would normally increase employabil- The impetus for addressing skills shortages ity—communication skills, project management, is strong, but it is also urgent. As skills gaps risk analysis—are not emphasized. growing in coming years, action must be taken now. It takes several years to train a skilled At the same time, the rapid growth in the worker, from basic education through specialized youth population is also putting much higher pres- skills. Specialization may be possible through sure on education systems than in the past. Higher short-term training. However, such training will demand for education and weak outcomes are all still require a basis of more basic skills and a the more a cause for concern given that MENA’s broader capacity for absorbing new knowledge, public education systems in general are quite which in turn must generally be acquired through costly, and provide comparatively more funding the childhood and youth education system. Since for higher levels of education. All of this suggests these general foundations cannot be built over a that more resources, by themselves, can hardly be few years, the efforts at improving and expanding the antidote. Efficiency gains through a thorough education must start now. reform of education systems will be necessary. Finally, education efforts will not pay off As discussed in World Bank (2008) and World unless there is an enabling environment that Bank (2009), North African countries can and must allows for a productive use for knowledge in the prepare for the future by improving their education economy (discussed further in the final chapter). system significantly. This would provide the region And skills shortages may not, in fact, be the main with a much needed force for economic develop- problem for potential and actual entrepreneurs, ment, and—in the face of increasing demand for whether based in the North Africa region or di- skilled migrants—ensure that human capital at aspora abroad. Political instability, corruption, home is not entirely depleted. This requires ad- high informality, and other factors seem to be just dressing serious quality issues at all levels of edu- as important bottlenecks to higher productivity cation, with respect to curriculum, focus, as well as and growth. Indeed, these will be the very same teaching methods, to make sure that qualifications factors that encourage some of the best and reflect knowledge and competence in skills that are brightest to leave their country in the first place. 104 Chapter 7 Further Agenda for Improving the Development Impact of North African Migration The previous chapters have highlighted some of high risk period for the migrant. Another is the the challenges the North African region faces in possible adverse effects of social security systems both improving the potential for positive impacts which require direct continued contributions from migration and preparing for the large de- from benefactors—even if they do not live in mographic changes that will be occurring over the country. the coming decades. Devising good migration policies and en- With increased migration to Europe foresee- suring well-founded policy options requires able, particularly for the region’s higher-educated both internal and external guidance. Given the workers, the largest challenge facing North Africa importance of international migration to their will occur on the education front, and involves economies, the data that North African coun- both the expansion of education and significant tries collect and make available on international improvement in education quality. This would migration needs to be significantly improved. provide the region with a much needed force How they go about it, and the benefits they can for economic development, and—in the face expect, depend on each country’s existing level of increasing demand for skilled migrants—en- of knowledge on migration. But all countries sure that human capital at home is not entirely can make small adjustments to current data depleted. collection instruments to significantly enhance their knowledge of both the determinants and But there remain other large issues to ad- consequences of migration Moreover, data are dress, both by the North African region and of limited value if they are not made widely ac- within Europe, if they are to mutually contribute cessible to citizens of the country: for research, to the region’s sustainable development. Social including finding limitations or deficiencies in protection regimes obviously have important the data that can lead to improvements in data implications for migrants themselves, and their collection; for research that can lead to improved capacity to cope with various risks—health, old policies; and for the private sector to plan its own age, unemployment, and others. Access and investment, sales, and advertising strategies. It portability have important repercussions, not is therefore crucial for all four countries to move only on migrant welfare, but also on incentives towards mechanisms to make data available to for return. MENA country nationals in Europe the public. are generally well covered by social protection through the Euro-MED agreements. However, Finally, many of the elements of the incentive some unresolved issues remain for migrants with framework which would increase the positive residence permits. One is the social protection returns from migration to the host country are of migrants who are newly arrived—normally a precisely the elements which increase growth 105   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options and development overall. Thus, as North Af- risk of attracting migrants to generous welfare rican economies aim to enhance the develop- systems is widely discussed. Howwever, stud- ment potential from migration, they will need ies suggest that the attractiveness of generous to strengthen these foundations for economic welfare systems is a much smaller factor than growth and development. economic opportunities or network effects for mi- gration (Gubert and Nordman 2007). In contrast, The Protection of North Africa’s incomplete social protection systems may intro- Workers Abroad, and its Impacts on duce perverse incentives. For example, if pension Migration Flows benefits are not exportable across countries, migrants are discouraged to return to the sending Improving the impacts from migration also means countries as they approach or enter retirement improving the conditions of migrants abroad, and (see Box 7.1 for explanation of exportability and improving the conditions for migrants to return. portability). Reallocation of skilled workers across Migration is part of a household risk diversifica- immigration countries is not possible in the face tion strategy, driven by lack of credit and insur- of an asymmetric demand shock. ance opportunities in sending countries.79 Thus, it happens partly in response to inadequate social Similarly, workers with no access to social protection at home. Migrants are often more insurance and contribution based benefits such vulnerable to income risks as they lack extensive as health and pensions, are discouraged from networks and are perhaps unacquainted with the participating in the formal economy since this society where they have arrived. De jure access would involve payments without benefits. does not guarantee de facto access to social ser- vices: for example, many benefits (for example, The fiscal impact is often ambiguous but unemployment) require a certain time of contri- it is clear that return migration may bring fis- bution before they can be paid out. Although this cal benefits, as the costs associated with social is the case for non-migrants as well as migrants, assistance of the elderly are reduced. It would recently arrived migrants are more vulnerable. seem particularly unfair to impose contributions on migrants who then cannot claim benefits; yet, More broadly, however, social protection this is sometimes the case. systems have much wider implications for migra- tion flows, fiscal costs, and economic structure (namely the size of the formal economy). The 79 See Bloom and Stark 1985; Stark 1991. Box 7.1: Exportability and Portability explained Exportability of pension benefits means that national transfer acquired social security rights independent social law allows for pensions being paid extrater- of nationality and country of residence� (Cruz, 2004). It ritorially. That is, benefits are paid independent of the ensures that not only acquired entitlements—that is, country of residence of the pensioner, as long as the when the individual fulfils the qualifying conditions for individual qualifies for a pension according to national a pension—can be exported, but also entitlements in law. It is therefore a unilateral measure, independent the process of being acquired. That is, short periods of of the legislation of the country of residence. contribution in one country, which are not sufficient to Portability, on the other hand, refers to the (partial qualify for a pension there, can be carried along to the or complete) totalization of the period of contributions other country, and vice versa. Portability is therefore between the host and the home country. Portability is a bilateral measure that requires the coordination of the migrant worker’s ability to “preserve, maintain and social security institutions in two countries. Source: World Bank 2009. Shaping the Future: A Long-Term Perspective of People and Job-Mobility for the Middle East and North Africa. 106 Chapter 7: Further Agenda for Improving the Development Impact of North African Migration Figure 7.1: Social services for migrant Sending/home Country Receiving/host Country  Social services for remaining  Portability of acquired  Social services for migrants and family members social rights between their families  Social services for returning receiving and sending  Labor market conditions for migrants country migrants. Source: Koettl, et al. 2009, an output of the EC-funded World Bank program of International Migration from the Middle East and North Africa and Poverty Reduction Strategies. The social protection system for interna- by bilateral agreements between tional migrants is illustrated in Figure 7.1 above the migrant-sending and receiving and include (i) access to social services for mi- country, which in principle ensures grants and their families in receiving countries; transferability of acquired rights (ii) access to social services for remaining family and no discrimination with regards members in sending countries; (iii) access to to social security benefits; social services in sending countries for return- ing migrants; (iv) portability of acquired social Regime II: Access to social security benefits in rights between receiving and sending country the absence of bilateral agreements, which includes both the totalization of years of which means that national laws of contributions irrespective of where they were the migrant-receiving country alone made, and the right to receive the benefits in determine the migrant’s access to country of choice; (v) labor market conditions benefits; for migrants, including hiring procedures, visa regulations, and workers’ rights. Regime III: No access to portable social security benefits, not even on a voluntary Social protection for MENA migrants in basis; and Europe is largely covered by the so-called Euro- MED agreements. In addition to the Euro-Med Regime IV: Undocumented but also legal mi- agreements, Maghreb countries enjoy more ad- grants who participate in the infor- vantages through bilateral labor migration and mal sector of the host country. social security agreements with several European countries. The Euro-MED agreements govern the As seen in Figure 7.2, migrants from Maghreb, social protection framework for migrant workers especially, enjoy quite high levels of social pro- from Algeria, Tunisia and Morocco, for countries tection compared to the rest of the world and for which the North African countries do not have indeed compared to MENA migrants whose main more favorably bilateral social security agree- destination is not Europe. ments.) As a result of these agreements, MENA migrants overall compare favorably within the However, although much has been done and global pool of migrants. Holzmann et al. (2005) progress has been good, there are still loopholes describe four main regimes of social protection in the social protection systems. One such loop- for labor migrants: hole is related to pensions exportability, which requires a ten-year minimum participation in Regime I: Access to social security benefits the social security system, and in some cases and advanced portability regulated (for example, Germany), pensions are reduced 107   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Figure 7.2: Social protection regimes for migrants I II III IV 100 90 Percentage of migrants covered by 80 different types of regimes 70 60 50 40 30 20 10 0 Global MENA, all Maghreb Source: Koettl 2006. when transferred abroad. Clearly, such provisions Clearly, a balance needs to be struck be- discourage return migration. tween, on the one hand, fiscal sustainability, the risk of migrant attraction to generous welfare Another inconsistency occurs in the inter- systems, and ensuring a level playing field, and action between national provisions for social on the other hand the rights of legal and irregular protection and immigration and alien laws. Most migrants. If European countries wish to attract European countries grant full social protection more migrant workers in the future, their specific coverage to long-term migrants or permanent vulnerability—when recently arrived, and in old residents although the interaction between social age—must be taken into consideration, as well and alien law complicates the de facto access. For as the risk of distorting incentives for return mi- example, temporary migrants have limited access gration, circular migration and the labor market to health care and unemployment insurance, and flexibility that migrants potentially can provide. no access to social assistance or public housing. Yet, continental European countries require up Social protection plans to manage to ten years of residence before granting perma- flows and protect workers nency, and in the southern European countries permanent status has only very recently become Part of that strategy will need to come through possible. the design of European social protection regimes, including both provisions for access and portabil- Moreover, the interaction between social and ity, which impact both migrants’ welfare, migrant alien law sometimes increases migrant vulner- flows, and social security institutions. In particu- ability. For example, several governments have lar, circular or return migration may be discour- introduced provisions in their alien law that en- aged if benefits earned through working years’ danger the migrant’s residence status (whether contributions are not portable across borders. permanent or temporary) when claiming welfare benefits. Hence, although legally residing mi- The social protection for MENA-born mi- grants contribute—by law—to welfare programs, grant workers, with an emphasis on workers, they are discouraged from claiming benefits. is adequate if there is a bilateral social security 108 Chapter 7: Further Agenda for Improving the Development Impact of North African Migration agreement in place between the migrant’s host tries in Europe for work, and thus less likely to and home country. Such agreements effectively reach vesting. govern the access to and portability of social security benefits for migrants from the signatory Also, without totalization, replacement states in their host country. In the case of MENA- rates are lower—this is the case for example born migrants, the Euro-Mediterranean Agree- for German migrants. With regard to healthcare ments only exist between the EU on the one benefits, migrants who return to their home hand, and Algeria, Morocco, and Tunisia on the country for retirement might not be able to find other. Since these Maghreb countries supply the adequate private healthcare insurance at home vast majority of MENA-born migrants to Europe, given their age. the social protection of MENA-born migrants in the EU is for the most part satisfactory. The social security institutions of migrant- sending and receiving countries are affected Access to social protection for legal migrants whenever benefits with a pre-saving element in the EU is thus adequate, with the exception are not fully portable. Hence, the lack of porta- of the problematic interaction of social law with bility of old age and healthcare benefits creates alien law. The latter deters migrants from claim- distortions in the system, which affects the ing benefits that they are entitled to for fear concerned social security institutions in vari- of losing their temporary work and residence ous ways. The pension institutions of the host permission. There is, moreover, an inverse rela- country are negatively affected since migrants tion between the vulnerability of migrants and want to avoid paying contributions and par- their access to social protection. Shortly after ticipate in the informal sector, but are positively arrival, when migrants are most vulnerable, ac- affected by receiving contributions without cess to social protection is low, but increases paying benefits and paying lower benefits. The over time until migrants are granted full access pension institutions in the home country also when receiving a permanent work and residence gain from not paying benefits or paying lower permission. To some extent, this is not unlike the benefits to migrants who decide to stay in the accumulated rights to some benefits for native host country, but at the same time lose from populations (such as waiting periods before un- having to provide for returning migrants who employment insurance can be accessed). In the might receive a means-based pension or social case of migrants, however, the trade-off here is assistance without having paid contributions or one of risk of abuse of generous welfare systems taxes to the home country. In the case of health (which applies to all populations) and the specific care, MENA’s healthcare usually has a pre-saving risks of migrant abuse (which apply to migrant element: young and healthy workers pay more circumstances). into the system than the old and sick. If migrants work abroad while young and healthy but return With regard to portability, bilateral social when old, the health systems of MENA countries security agreements are crucial. Although Euro- may suffer the financial consequences. Health- pean countries usually allow for the exportability care institutions in both migrant-sending and of pension benefits, only bilateral agreements can receiving countries would benefit from increased assure the totalization of periods of contribution. portability since it would remove obstacles in the In the absence of bilateral agreements, returning trade of healthcare services. migrants who do not fulfill the qualifying condi- tions for an old-age pension (usually at least 10 While these difficulties may cancel out be- years of contribution) lose all their contributions tween developed countries between which flows and the benefits associated with them. While na- are more symmetrical, the situation is clearly tionals also have required vesting periods, they more acute in the case of flows from a poorer do not usually move out of country as often, so region (MENA) to a richer, labor importing one their benefits collect throughout multiple jobs. (Europe), where migrants spend their productive Migrants are more likely to move between coun- life abroad and return for retirement. 109   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Policy implications for improving the social to its émigrés abroad, deserves more attention in protection of MENA-born migrants in Europe and order to evaluate if such a policy can serve as a the Gulf are possible on three levels: unilaterally, model to unilaterally improve the social protec- multilaterally, and bilaterally. tion of migrants. Unilaterally, countries should align the con- A multilateral approach to address issues of ditions for exportability of old-age benefits for access to and portability of social security ben- migrants with those that apply to nationals. Most efits for international migrants is the first-best countries allow that pension benefits are paid solution. Such an approach creates common to every country in the world, but some coun- standards and removes discrimination among tries apply reductions. Totalization of periods migrants that is created through the current of contribution cannot be achieved unilaterally bilateral policies. The ILO Convention 157 on but only in cooperation with the migrant’s home the maintenance of social security rights is the country. Yet, in the absence of totalization, if a best example of what such an approach could migrant does not meet the qualifying conditions look like. Also, the various regional efforts to for an old-age pension, she should at least be able cooperate in the field of social security—through to receive a lump sum payment of her and her the AMU, the GCC, and the EU—can serve as employer’s contributions when returning to the platforms to harmonize social security systems, home country. This is currently rarely the case. deal with access to social protection for migrants, and enhance portability of benefits. The Euro- In the case of healthcare benefits, the health- Mediterranean Agreements for the Maghreb care institutions of migrant-receiving countries countries show how such an approach might should reimburse migrants for healthcare ser- work and lead to common standards. Within the vices they consume in the home country. This MENA Region, initiatives are not moving ahead, would not only benefit migrants who return to however. their home country—either temporarily or per- manently—but also the healthcare insurance Hence, for the time being, bilateral agree- institutions in both migrant sending and receiving ments as a second-best solution are more re- countries because they could realize compara- alistic. Some form of multilateral cooperation tive cost advantages through increased trade in within the MENA Region is indispensable as healthcare services. long as access to social protection for migrants is not granted and as long as the disparities In this regard it would also make sense to al- among the systems prevail. Yet, for the sake low migrants who have contributed long enough of fast and efficient results, bilateral initia- to the healthcare system of the host country to tives—in particular between the EU and MENA stay voluntarily insured, even after return, as it countries—are preferable. Although Maghreb might be prohibitively expensive for them to find countries are mostly well covered, other MENA private insurance after return because of their migrant groups do not enjoy the same access age. This can only make sense, though, if these to social protection. migrants are allowed to consume healthcare services in their home country at the expense Improving social protection systems would of the healthcare system of the host country. If not only benefit migrants and their family mem- not, this would lead to the situation of French bers by guaranteeing non-discriminatory access retirees who returned to Morocco: they have to to, and enhanced portability of, social security continue to contribute to the French healthcare benefits. It would also affect the concerned system without being reimbursed for healthcare social security institutions by removing dis- services they receive in Morocco. tortions within the systems and may facilitate the circular migration movements currently Finally, the experience of Tunisia, which ex- favored as the most desirable forms of people tends the coverage of the social security system mobility. 110 Chapter 7: Further Agenda for Improving the Development Impact of North African Migration North Africa’s Migration Information The North African countries should be mo- and Capacity Gap tivated to also monitor immigration. Different estimates suggest that each year, large numbers Devising good migration policies and ensuring (some 30–40,000) of mostly sub-Saharan Afri- well-founded policy options requires both inter- cans enter Algeria and Morocco, some of whom nal and external guidance. Thus, while MENA stay and some of whom go on to Europe (de Haas policymakers should continue to review and 2006). Thus, understanding the size and impact assess the outcomes of international best and of immigration is also becoming increasingly worst practices in migration management and important. While this is an important premise, consider how good practice can be adapted to this discussion focuses on emigration, which is specific national and regional circumstances, the less straightforward to capture. most critical obstacle to adaptation of interna- tional policies as practice is a lack of migration The limited knowledge base on the size of data from which to monitor the trends in the migration, its determinants and main impacts is size, determinants, and impacts of emigration due to two problems, on the input and output and immigration. side. On the input side, there is a dearth of data. Expansion of the statistical instruments used to Monitoring and evaluation of migration and collect data on migration, as well as improve- other social and economic data is critical in order ments in existing instruments are sorely needed. to understand the impact of different policies On the output side, clear strategies or approaches and adapt and improve the current policy mix. on how to turn data into knowledge appear not Apart from the long run implications of migration to exist.. Much higher and better exploration of on growth and welfare, it is also essential that existing data will be needed. policymakers find a way to gauge the impact of the economic crisis on remittances and migra- Data: Statistical capacity and tion more broadly, the effect of those changes instruments on household welfare, especially the poor, on inequality, and on households’ livelihood, diver- The data that countries collect on international sification and risk management strategies. migration should reflect its importance to the country, and to the welfare of its citizens, wheth- Paradoxically, while emigration and immigra- er at home or abroad. Since all four countries tion phenomena is central to MENA countries, have indicated a significant increase in interest and receives increasing attention in the policy in international migration the question is how forum, the basis on which to inform policymak- they should go about it, and what benefits can ing and monitor policies remains weak. There they expect. There are three possible levels of are large information gaps, especially in the knowledge about international migration that understanding of migration determinants and are possible: outcomes in the four focus countries. The statisti- cal capacity and interest in monitoring migration A. Gauging the size of the emigration phe- and analyzing its impacts is still limited, although nomenon. This involves basic estimates of Morocco especially, but also Egypt appear to be the stock of immigrants, and, if possible, more advanced than Algeria and Tunisia in terms outflows, taking into account transit and of initiatives. For example, Algeria did not until undocumented migration; basic estimates 2008 have a way of counting emigrants or im- of households with (recent) emigrants, and migrants. None of the countries has a consistent outflows if available. These data are available database tracking migration stocks (or flows) from censuses in all countries. over time, and so snap shots through different B. Characterizing migrants compared to surveys are not comparable. The work on under- non-migrants. This requires data on cer- standing the impact and drivers of migration are tain key characteristics of the immigrants moreover only at its beginning. and emigrants and their households, such as 111   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options age, sex, education, marital status, employ- from each of these four focus countries (and ment status, etc., and how these compare most other developing countries in the world) with the characteristics of non-migrants; is a rare event, a large sample size is needed this information would help identify, among to capture a sufficient number of households other things, the extent of skilled migration present in one round who are not there at the vs. unskilled migration. Exhaustive ques- time of the second round, which is commonly tions are costly for census and these data in practice one year later. Among the focus are more likely to come from household countries, Tunisia and Egypt have labor force surveys. At issue is whether sample sizes are surveys with fairly large samples that have a large enough to capture sufficient migrant rotating panel, so that for their panels it is households (this is unfortunately not the known if the household has migrated away or case in Tunisia). not. The new National Demographic Survey C. Understanding what drives migration, in Morocco also has several rounds, appar- and the consequences of migration. ently for only one year. This requires data on the characteristics of migrants and their households at the time of Migration data are collected to inform migration compared to non-migrant house- policymakers of the extent, causes, and effects holds; data on extent and use of remittances, of migration, and guide policy interventions working conditions, living conditions and to maximize benefits and minimize costs. To integration in host country. This generally re- understand this, it is necessary to consider the quires specialized surveys. Only Morocco and general relevance to policies in what we learn Egypt have specialized surveys that collect from studies of the determinants of interna- data on international migration, but most of tional migration and of the consequences of in- these do not have adequate sample designs to ternational migration. Results from determinant ensure adequate numbers of recent migrants, studies tell us what it is that affects, or causes, and none has an adequate questionnaire de- international migration, and thereby can identify sign. The latter requires, for the investigation factors that policymakers may be able to influ- of the determinants of (e.g., emigration) ence in order to alter migration levels, types, or decisions, data on the situation of the migrant directions. But whether policymakers should prior to or at the time of emigration, not data do that cannot be inferred from studies of the at the time of interview. Moreover, these data determinants of migration. Instead it requires should refer to not only the migrant but also knowing what the consequences are of migra- to the household composition (which may tion. If the consequences of a particular type have to be reconstituted from data on other of international migration are deemed positive household members being born, dying, or (or negative), then the policy implication would migrating into or out of the household since be to encourage (or discourage) that type of the time the migrant left. migration. But a study of consequences tells us nothing about how to alter migration. That only Added to these considerations is the type of comes from studies of determinants. It is clear, emigrants one can hope to capture: then, that policymakers need to have results on both the determinants and consequences of D. Capturing migrating households rather each type of international migration of interest than individuals. For identifying the emi- if they are to develop or improve policies on gration of entire households, a longitudinal or international migration (Bilsborrow et al. 1984, panel survey is needed: a single round survey 1997). Specialized surveys of international migra- (or population census) cannot obtain data on tion offer the best (but not inexpensive) means whole households departing since there is for collecting the data necessary to carry out the no one there to provide reliable data on the analytical studies needed to understand both the household that left before the time of the determinants and consequences of international second round. Since international migration migration. 112 Chapter 7: Further Agenda for Improving the Development Impact of North African Migration Table 7.1 provides an overview of the exis- A logical approach for the countries to pur- tence of some emigration data in different census sue is to add modules on emigration to existing and surveys in the four focus countries. It is surveys (see Box 7.2), but also investigate the important to note that the table informs us pri- possibilities of specialized migration surveys that marily on whether the questionnaire captures can adequately capture the broader policy ques- any information on emigration. It does not detail tions surrounding migration determinants and issues in the broader statistical system, such as impacts. Given where countries are now in terms whether sample sizes are generally large enough of previous data collection with specialized sur- to permit such data to be representative, if data veys, both Morocco and Egypt, having carried out are timely, or if it is available for exploitation. their own specialized surveys already on one or It is clear, however, that data are still scarce, more types of migrants (immigrants, emigrants), especially for comparisons over time, and that and having strong statistical offices, could be especially Algeria, especially, has a long way to thinking of how to improve and expand those go to provide a good understanding of migration surveys to collect better data, or design a new stocks, let alone determinants or impacts. It is better specialized survey. Tunisia and Algeria, on also clear, however, that statistical systems are the other hand, are ready to undertake surveys increasingly cognizant of the importance of filling of the types Morocco and Egypt have already the information gaps and that new initiatives are carried out, while learning from the mistakes being launched. and difficulties encountered in the latter. Later Table 7.1: Data on emigration in statistical instruments Algeria Egypt Morocco Tunisia Census 2008 population 2006 census 2004 census census Household surveys Consumption/ • 2000/2001: remittances from 2005 some expenditure surveys abroad information, but • 2006/2007: yes, both emigration, very small sample immigration and return migration (but not available) Specialized surveys on 2008: International migration migration survey Labor force surveys 2007 onwards 2005 and onwards, some limited information Demographic and health surveys Living standard 1991,1998/1999 measurement surveys Other Post censal surveys Border surveys • Survey of socio-economic inte- gration into receiving countries • 2000 survey on return migrants Other ELMPS, 2006, • Upcoming: National Demographic Survey 2009–2010, panel (two years) • MIREM return migration 113   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Box 7.2: Modules on emigration to add to existing surveys A non exhaustive list of possible questions to add to cruitment if relevant, cost of job brokerage, cost labor force or other household surveys: of travel) • Work/activity in destination country Identifying migrants and their characteristics (Sex, age, • Adaptation/assimilation in destination country age at departure, education level, marital status, family (language, education achieved abroad, citizen- left behind; year of departure, country of destination) ship, future plans) • Migration intentions among other family mem- • Work/activity of emigrant prior to departure bers • Reasons of emigration (work, studies, family, • Remittances (frequency, amount, use) network) • Similarly, modules can be added on immigration, • Policy relevant questions (terms of employment and on return migration. in destination country e.g. contract, form of re- Source: Bilsborrow 2009, an output of the EC-funded World Bank program of International Migration from the Middle East and North Africa and Poverty Reduction Strategies. on they can move on to develop more complex government and statistical institutes has many specialized surveys of international migration. practical implications. For example, it means that data can be explored by a much wider group of Turning data into information and practitioners, especially researchers knowledge However, this involves appropriate sharing Data are of course nothing but numbers, and are of tasks—as discussed in this report, migration of limited value if they are not explored in an ex- research is an area fraught with many method- haustive manner. A country’s ability to carry out ological pitfalls and researchers may be better this exhaustive research is partly related to sta- tuned in to both issues and methods necessary tistical capacity. According to the World Bank’s for the exploitation rather than collection of statistical capacity rankings, Egypt especially, but statistical information. Much more useful and also Morocco, score above developing countries policy relevant information and knowledge can on average, while Tunisia is slightly above, and be derived from the data, if it is shared widely. Algeria slightly below the average. Sharing data also has practical implications However, this is referring to in-house capac- for the quality of the statistical system, as it is also ity only, and is not taking into account data trans- a means of encouraging user-driven innovation. parency and access. Yet, data sharing outside of Researchers can find limitations or deficiencies Table 7.2: 2008 statistical capacity indicator (out of 100) Indicator Algeria Egypt Morocco Tunisia All countries* Overall 62 88 74 67 65 Statistical Practice 50 70 90 70 56 Data Collection 50 100 60 60 62 Indicator availability 87 93 72 72 77 Source: World Bank. ** All countries include low- and middle-income IDA/IBRD countries with a population of over 1 million. 114 Chapter 7: Further Agenda for Improving the Development Impact of North African Migration in data that can lead to improvements in data col- migration, and also to better prepare workers lection, which in turn will also benefit “in house� for the type of work demanded by competitive exploitation of these data within the statistical firms in a global economy. Not surprisingly, the agencies. Wider distribution encourages much development of more private-sector relevant more intense testing of the existing statistical skills not only increases the employment poten- approaches. tial (and remittance-sending potential) for mi- grants abroad, it also increases the employment It is therefore crucial for all four countries to potential for the population at home. move towards mechanisms to make micro-data files available while, of course, ensuring confiden- Having an investment climate which provides tiality. These data need to be made available to sufficient job opportunities for these higher edu- the public in a timely fashion, together with clear cated workers at home not only reduces skilled documentation of the sample size, the sampling labor flight, it also increases the contribution of frame used, the sampling methods used, and human capital to growth and development. An sampling weights. This may require additional investment climate that attracts business devel- resources. As statistical capacity building is part opment from returned migrants also attracts the of donor assistance to MENA countries, the domestic investment for economic growth and emphasis on access could be factored into such development, and a financial sector that provides programs and appropriately budgeted for. capital for return migrant investors also provides capital for domestic investors. Institutional capac- Improving North Africa’s ity that can support the successful engagement Environment for Migration of diasporas for development is the same institu- Contributions to Development tional capacity that can support successful private sector growth and job creation. In short, the pri- Finally, and ultimately, the degree to which North orities for enhancing the returns from migration Africa can benefit from emigration will depend as are, not surprisingly, the priorities for enhancing much on its own education, investment climate, growth and development more broadly. and institutional reforms as on the policies in Europe. And it is perhaps not surprising that It is not the purpose of this research pro- many of the elements of the incentive framework gram to review the social and structural reform which would increase the positive returns from agenda remaining for North African countries: migration to the host country are precisely the the agenda is distinct for each country, and each elements which increase growth and develop- is at a different place in their own transition. ment overall. However, it is upon this broader economic and institutional foundation that the migration and This research program has emphasized the development nexus is built. Inevitably, for North need for both quantity and quality enhancements African economies to enhance the development in education, both to avoid the depletion of im- potential from migration, they will need to portant skills with the likelihood of increasing strengthen these foundations as well. 115 References Acosta, P. 2006. “Labor Supply, School Atten- ment: Evidence from Egypt.� World Devel- dance, and Remittances from International opment 33(3): 431–454. Migration: The Case of El Salvador.� World Azariadis, C. and A. Drazen. 1990. “Threshold Bank Policy Research Working Paper 3903, Externalities in Economic Development,� The World Bank, Washington, DC. Quarterly Journal of Economics, 501–526 Acosta, P. 2007. “Entrepreneuship, Labor Mar- Barham, B. and S. Boucher. 1998. “Migration, kets, and International Remittances: Evi- Remittances and Inequality: estimating the dence from El Salvador.� In International net effects of migration in income distribu- Migration and Economic Development, tion,� Journal of Development Economics, Remittances, and Brain Drain, edited by vol. 55, 307–331. Caglar Ozden and Maurice Schiff, Chapter 4. Barro, R.J. and J.W. Lee. 2000. “International World Bank publications. Data on Educational Attainment: Updates Adams, Richard, Jennifer Keller, Lili Mottaghi, and Implications,� Oxford Economic Papers and Marie Alienor van den Bosch. 2009. 53, 541–563. “Remittances and Growth in North Africa,� Barsoum, Ghada. 2007. “Egypt Labor Market an output of the EC-funded World Bank Panel Survey 2006: Report on Methodology program of International Migration from the and Data Collection.� Working Paper No. Middle East and North Africa and Poverty 0704. Economic Research Forum, Cairo. Reduction Strategies. Batalova J., P. A. Creticos and M. Fix. 2008. “Un- Aghion P., C. Meghir, and J. Vandenbussche. even Progress: The Employment Pathways 2006. “Distance to Frontier, Growth, and the of Skilled Immigrants in the United States,� Composition of Human Capital.� Journal of October.Migration Policy Institute (MPI), Economic Growth. Washington, DC, United States. Al Ali, Nadje. 2006. “Relationship between migra- Beine, M., F. Docquier and H. Rapoport. 2001. tion within and from the MENA and pro-poor “Brain Drain and Economic Growth: Theory policies.� Discussion Paper, University of and Evidence.� Journal of Development Exeter. Institute of Arab and Islamic Studies. Economics 64 (1), 275–89. Amuedo-Dorantes, C. and S. Pozo. 2006. “Migra- Beine, M., F. Docquier and H. Rapoport. 2008. tion, Remittances, and Male and Female Em- “Brain Drain and Human Capital Formation ployment Patterns,� American Economic in Developing Countries: Winners and Los- Review, 96(2), 222–226. ers.� Economic Journal 118, 631–652. Assaad, R. 2007. “Labor Supply, Employment and Benhabib, J. and M.M. Spiegel. 2005. “Human Unemployment in the Egyptian Economy, Capital and Technology Diffusion.� In Hand- 1988–2006.� Economic Research Forum book of Economic Growth, edited by Ph. Working Paper, 200701. Aghion and S. Durlauf, Vol 1, Chapter 13, Assaad, R. and M. Arntz. 2005. “Constrianed 935–966, Elsevier. Geographical Mobility and Gendered Labor Bhargava, A. and F. Docquier. 2008. “HIV Pan- Market Outcomes Under Structural Adjust- demic, Medical Brain Drain and Economic 117   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Development in Sub-Saharan Africa.� World lence of Rare Diseases: Three Unbiased Bank Economic Review, 22(2), 345–366. Estimates.� Vital and Health Statistics Bilsborrow, Richard E. and Ingrid Ivins. 2009. 2(11). “An Assessment of Sources of Data on In- Boujour, S. 2005. “The Politics of Migration and ternational Migration in the Maghreb and Development: The migration-development Egypt, Focusing on Population Censuses and connection in Dutch political discourse and Household Surveys.� An output of the EC- policy since 1970.� Final paper for Florence funded World Bank program of International School on Euro-Mediterranean Migration and Migration from the Middle East and North Development. September. Africa and Poverty Reduction Strategies. Bustamente, Jorge A., Rodolfo Corona, and Bilsborrow, Richard E., and Centro de Estudios Jorge Santibáñez. 1994. Encuesta sobre sobre Población y Desarrollo Social (CEPAR). Migración en la Frontera Norte de México: 2007. The Living Conditions of Refugees, Síntesis Ejecutiva. Tijuana, Mexico: El Co- Asylum-seekers and Other Colombians legio de la Frontera Norte, Consejo Nacional in Ecuador: Millennium Development de Población, and Secretaria del Trabajo y Indicators and Coping Behaviour. Geneva, Previsión Social. Switzerland: United Nations High Commis- Cantisani, Giambattista, Samir Farid, David sioner for Refugees (unhcr.org/statistics) Pearce, and Nicolas Perrin. 2009. “Guide Bilsborrow, Richard E., Graeme Hugo, Amarjit on the Compilation of Statistics on Interna- S. Oberai, and Hania Zlotnik. 1997. Inter- tional Migration in the Euro-Mediterranean national Migration Statistics: Guidelines Region.� MEDSTAT II, Euro-Mediterranean for Improving Data Collection Systems. Statistical Cooperation. February 11 draft. Geneva: International Labour Office. Carletto, Gero, Benjamin Davis, Marco Stampini, Bilsborrow, Richard E., A. S. Oberai, and Guy and Alberto Zezzo. 2004. “A Country on the Standing. 1984. Migration Surveys in Low- Move: Internal Mobiility and International Mi- Income Countries: Guidelines for Survey gration in Post-Communist Albania.� Rome: and Questionnaire Design. London and FAO. Dover, NH: Croom Helm, for the International Carrington, W. J. and E. Detragiache. 1998. Labour Office. “How Big is the Brain Drain?� IMF Working Bilsborrow, Richard E. and Hania Zlotnik. 1995. Paper. WP/98/102. International Monetary “The Systems Approach and the Measure- Fund,Washington, DC. ment of the Determinants of International CELADE. 1977. “Investigación de la Migración Migration.� In Causes of International Internacional en Latinoamérica (IMILA).� Migration, edited by R. F. van der Erf and Boletín Demográfico 10. L. Heering. Luxembourg: Office for Official CEPAR. 2005. Encuesta Demográfica y de Publications of the European Communities, Salud Materna e Infantil: ENDEMAIN IV. 61–76. Quito, Ecuador and Atlanta, Ga.: Centro de Binzel, Christine, and Ragui Assaad. 2009. “The Estudios sobre Población y Desarrollo Social; Impact of International Migration and Remit- Centers for Disease Control. tances on Domestic Labor Markets: Evidence Chiswick, B.R. and P.W. Miller. 1992. “Language in from Egypt.� An output of the EC-funded the Labor Market: the Immigrant Experience World Bank program of International Migra- in Canada and the United States,� In Im- tion from the Middle East and North Africa migration, Language and Ethnic Issues: and Poverty Reduction Strategies. Canada and the United States, edited by B. Binzel, C. and R. Assaad. 2008. “Pathways to Mar- R. Chiswick. American Enterprise Institute, riage in Egypt: How the Timing of Marriage Washington. for Young Men is Affected by their Labor Choi, Inbom. 2003. “Korean Diaspora in the Market Trajectory.� Unpublished mimeo. Making: Its Current Status and Impact Birnbaum, Z.W. and M. Sirken. 1965. “Design on the Korean Economy.� In The Korean of Sample Surveys to Estimate the Preva- Diaspora in the World Economy, edited 118 References by C. Bergsten and Fred and Inbom Choi, gration Institute, James Martin 21st Century Institute for International Economics. School, WP3, University of Oxford Washington, DC. de La Fuente, A. and A. Ciccone. 2003. “Human Clemens, M.A. 2007. “Do Visas Kill? Health Ef- Capital and Growth in a Global and Knowl- fects of African Health Professional Emigra- edge-based Economy.� Office for Official tion.� Working Paper 114. Center For Global Publications of the European Communities, Development, Washington D.C. Luxembourg. Clemens, M.A. and G. Pettersson. 2006. “A New de Neubourg, Chris. 2008. “Changes in Immi- Database of Health Professional Emigration gration in the Netherlands: Trends, Policies from Africa.� Working Paper 95. Center for and Incentive.� An output of the EC-funded Global Development. World Bank program of International Migra- Commander, S., M. Kangasniemi, L.A. Winters. tion from the Middle East and North Africa 2004. “The Brain Drain: Curse or Boon? A and Poverty Reduction Strategies. Survey of the Literature,� In Challenges to Di Bartolomeo, Anna, Tamarace Fakhoury, and Globalization: Analyzing the Economics, Delphine Perrine. 2009. “Morocco: The edited by R. E. Baldwin and L. A. Winters, Demographic Economic Framework of Mi- Chicago University Press. gration, The Legal Framework of Migration, Conde-Ruiz, J-I, C. Gonzalez, and S. Johansson the Socio-Political Framework of Migration.� de Silva, 2008, “An Immigration Boom under CARIM-Migration Profile. European Univer- a Reactive Immigration policy: the case of sity Institute, Robert Schuman Centre for Spain.� An output of the EC-funded World Advanced Studies. Bank program of International Migration Docquier, F., O. Faye and P. Pestieau. 2008. “Is from the Middle East and North Africa and Migration a Good Substitute for Subsidies?� Poverty Reduction Strategies. Journal of Development Economics 86, Coulombe, S. and J-F. Tremblay. 2006. “Literacy 263–276. and Growth.� Topics in Macro-economics Docquier, F. and E. Lodigiani. 2009. “Skilled 6(2), Article 4. Migration and Business Networks.� Open de Haas, Hein. 2005a. “Morocco: From Emigra- Economies Review. Forthcoming. tion Country to Africa’s Migration Passage to Docquier, Frédéric, Sara Johansson de Silva, and Europe.� Mimeo, Migration Policy Institute. Abdeslam Marfouk. 2009. “Skilled Migration de Haas, Hein. 2005b. “Morocco’s migration from the MENA Region: Trends, Impacts transition: trends, determinants and future and Policy Responses.� An output of the EC- scenarios.� Global Migration Perspectives funded World Bank program of International No. 28. Global Commission on International Migration from the Middle East and North Migration. Africa and Poverty Reduction Strategies. de Haas, Hein. 2006. “Migration, Remittances Docquier F., B. Lindsay Lowell, and A. Marfouk. and Regional Development in Southern Mo- 2007. “A gendered Assessment of the Brain rocco,� Geoforum, 37(4), 565–580. drain,� IZA discussion papers, Institute for de Haas, Hein. 2006. “Trans-Saharan Migration the Study of Labor (IZA), No. 3232, Decem- to North Africa and the EU: Historical Roots ber. and Current Trends.� Mimeo, Migration Docquier, Frédéric and Luca Marchiori. 2008. Policy Institute. “MENA-to-EU migration and labor market de Haas, Hein. 2007. “Between courting and imbalances.� An output of the EC-funded controlling: The Moroccan state and ‘its’ World Bank program of International Migra- emigrants, Centre on Migration.� Policy and tion from the Middle East and North Africa Society, WP-07–54, University of Oxford and Poverty Reduction Strategies. de Haas, Hein. 2007. “The impact of international Docquier F. and A. Marfouk. 2006. “Interna- migration on social and economic develop- tional Migration by Educational Attainment ment in Moroccan sending regions: a review (1990–2000).� In International Migration, of the empirical literature.� International Mi- Remittances and Development, edited by 119   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options C. Ozden and M. Schiff. Parlgrave, Macmil- Funkhauser, E. 1992. “Migration from Nicaragua: lan, New York. Some Recent Evidence.� World Develop- World Bank. 2009. Skilled Migration from the ment, 20(8), 1209–18. MENA region: trends and impacts, by Gallina, Andrea. 2006. “The Impact of Interna- F. Docquier, S. Johansson de Silva, and A. tional Migration on the Economic Develop- Marfouk ment of Countries in the Mediterranean Egypt, Ministry of Manpower and Emigration. Basin.� Paper for the Expert Group Meeting 2003. International Migration in Egypt. on International Migration and Development Cairo: 2003. in the Arab Region: Challenges and Oppor- Elbadawy, Asmaa and Ragui Assaad. 2009. “Im- tunities, Beirut, May. pact of International Migration and Remit- Goodman, Leo A. 1961. “Snowball Sampling.� tances on Child Schooling and Child Work: Annals of Mathematical Statistics, Vol. The Case of Egypt.� An output of the EC- 32(1): 148–170. funded World Bank program of International Goodman, R. and Leslie Kish. 1950. “Controlled Migration from the Middle East and North Selection: A Technique in Probability Sam- Africa and Poverty Reduction Strategies. pling.� Journal of the American Statistical ENCDM. 2001. Enquête nationale sur la consuma- Association 45:350–72. tion et les dépenses des ménages 2000/2001. Gould, David. 1994. “Immigrant Links to the Haut commissariat au plan, Rabat. Home Country: Empirical Implications for ELMPS. 2006. Egypt Labor Market Panel Survey U.S. Bilateral Trade Flows.� Review of Eco- of 2006. Publicly Accessible Database, Eco- nomics and Statistics, 76. nomic Research Forum, Cairo, Egypt (www. Groenewold, George and Richard E. Bilsborrow. erf.org.eg). 2008. “Design of Samples for International Fadloullah, Abdellatif, A. Berrada, and M. Migration Surveys: Methodological Consid- Khachani. 2000. Facteurs D’Attraction et erations and Lessons Learned from a Multi- de Repulsion a L’Origine des Flux Migra- Country Study in Africa and Europe.� In In- toires Internationnaux: Rapport national ternational Migration in Europe: Current Le Maroc. European Commission: Eurostat. Trends and Issues, edited by C. Bonifazi, M. Fargues, F. 2006. “International Migration in the Okolski, J. J. Schoorl, and P. Simon. Rome: Arab Region: Trends and Policies,� Popula- Universiti di Roma, 293–312.. tion Division, United Nations Expert Group Gubert, Flore and Christophe Nordman. 2008a. Meeting on International Migration and De- “Return Migrants and Small Enterprise Devel- velopment in the Arab Region, Department of opment in the Maghreb.� An output of the EC- Economic and Social Affairs United Nations funded World Bank program of International Secretariat, UN/POP/EGM/2006/09, Beirut, Migration from the Middle East and North 15–17 May. Africa and Poverty Reduction Strategies. Fargues, Philippe. 2008. “Emerging Demographic Gubert, Flore and Christophe Nordman. 2008b. Patterns across the Mediterranean and their “An Empirical Analysis Using Data on Re- Implications for Migration through 2030.� turn Migrants in the Maghreb.� An output Washington, DC: Migration Policy Institute, of the EC-funded World Bank program of November. International Migration from the Middle East Fargues, Philippe. 2009. “Irregularity as Normal- and North Africa and Poverty Reduction ity among Immigrants South and East of Strategies. the Mediterranean.� Analytic and Synthesis Hamdouch, B. 2009. “Migration, Diaspora et Notes. CARIM-AS-2009/02. Florence, Italy: Dévéloppement: Le cas du Maroc.� Mimeo. CARIM. Middle East and North Africa Department, Funkhauser, E. 1995. “Remittances from Interna- the World Bank tional Migration: A Comparison of El Salvador Hammouda, Nacer-Edine. 2008. Migration, and Nicaragua.� The Review of Economics Diaspora et Développement  : Le Cas de and Statistics, 77(1), 137–46. l’Algérie. unpublished background paper. 120 References Hanson, G. 2007. “Emigration, Remittances, and in Pakistan: Preliminary Results from the Labor Force Participation in Mexico.� Work- PLM Survey, 1979.’ Studies in Population, ing paper 28. Institute for the integration of Labour Force and Migration Project Report Latin America and the Caribbean (INTAL). No. 6.� Islamabad: Pakistan Institute of De- Head, Keith and John Ries. 1998. “Immigration velopment Economics. and Trade Creation: Econometric Evidence Kalsbeek, W.D. 2003. “Sampling Minority Groups from Canada.� Canadian Journal of Eco- in Health Surveys.� Statistics in Medicine, nomics 31, no 1. 22(9): 1527–49. Hiebert D. 2002. “Canadian Immigration and Kalton, Graham and D. W. Anderson. 1986. the Selection-settlement Service Trade-off: “Sampling Rare Populations.� Journal of the Exploring Immigrant Economic Participation Royal Statistical Society Series A: Statis- in British Columbia.� Working Paper Series, tics in Society 149:65–82. No. 02–25. Research on Immigration and In- Kangasniemi, M., S. Commander and L.A. tegration in the Metropolis (RIIM). October. Winters. 2007. “Is the Medical Brain Drain Herrera, Catalina, Nora Dudwick, and Edmundo Beneficial? Evidence from Overseas Doctors Murrugarra. 2008. Remittances, Gender in the UK.� Social Science & Medicine 65: Roles and Children’s Welfare Outcomes in 915–23. Morocco, an output of the EC-funded World Kapur, D. and J. McHale. 2005. “The Global Bank program of International Migration Migration of Talent: What Does it Mean for from the Middle East and North Africa and Developing Countries?� Center For Global Poverty Reduction Strategies. Development, Washington D.C. INDEC. n.d. Complementary Survey on Inter- Kirevev, Alexei. 2006. “The Macroeconomics of national Migrants (Encuesta Complemen- Remittances: The Case of Tajikistan�. IMF taria sobre Migrantes Iinternacionales). Working Paper WP/06/2; International Mon- Buenos Aires: Instituto Nacional de Estadis- etary Fund. tica y Census. (www.indec.gov.ar/webcenso/ Kish, Leslie. 1965. Survey Sampling. John Wiley ecmi/index_ecmi.asp). & Sons. INEC (Instituto Nacional de Estadísticas y Cen- Kugler M. and H. Rapoport (2007), “International sos). 2001. VI Censo de Población y V de Labor and Capital Flows: Complements or Vivienda. Quito, Ecuador: INEC. Substitutes?� Economics Letters, 94 (2), Isaoglu, Aysen. 2007. “A Cautionary Note 155–162. about the Discrepancies among the UN, Koettl, Johannes, Marie Alienor van den Bosch, OECD and CARIM Databases on Interna- Matteo Morgandi. 2008. “Social Protection tional Migration.� In Mediterranean Mi- for Migrant Workers from North Africa to gration—2006–2007 Report, edited by Europe.� An output of the EC-funded World Philippe Fargues. Florence, Italy: CARIM. Bank program of International Migration International Labour Office. 2009. “Information from the Middle East and North Africa and Note: ILO Labour Migration Module. Geneva: Poverty Reduction Strategies. ILO Bureau of Statistics, March. Kuznetsov, Yevgeny and Matteo Morgani. 2008. International Organization for Migration. 2005. “Why is Diaspora Potential so Elusive? World Migration 2005: Costs and Benefits Towards a New Diaspora Agenda in North of International Migration. Geneva, Swit- African Economies.� An output of the EC- zerland.: IOM. funded World Bank program of International Irfan, Mohammad. 1981. “An Introduction to Migration from the Middle East and North ‘Studies in Population, Labour Force and Africa and Poverty Reduction Strategies. Migration’, a PIDE/ILO-UNFPA Project.� Lodigiani E. 2008. “Diaspora Externalities and Research Report Series No. 118. Islamabad: Technology Diffusion.� IRES Discussion Pakistan Institute of Development Economics. paper 2008–8. Irfan, Mohammad, Lionel Demery, and Arif Gh- Lohr, Sharon L. 1999. Sampling: Design and ulam Mohammad. 1984. ‘“Migration Patterns Analysis. Duxbury Press. 121   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Lucas, R. E., Jr. 1988. “On the Mechanics of Eco- Musette, Mohamed Saib, B. Saad, B. Hassan, nomic Development.� Journal of Monetary and H. Eddine. 2006. “Systems of Statistical Economics 22:3–42. Information on Migrant Workers in Central Makhlouf, Hesham H., et al. 1999. Push and Maghreb.� International Migration Papers Pull Factors of International Migration, 76E. Geneva, Switzerland: International Country Report, Egypt. European Commis- Labour Office. sion: Eurostat. Nassar, H. 2005. “Migration, Transfers and De- Massey, D. S., J. Arango, G. Hugo, A. Kouaouci, A. velopment in Egypt.� Euro-Mediterranean Pellegrino, and J. E. Taylor. 1993. “Theories Consortium for Applied Research on Inter- of International Migration: A Review and national Migration Research Report 2005/01, Appraisal.� Population and Development European University Institute, Robert Review 19:431–466. Schuman Centre for Advanced Studies. Massey, D. and Parrado, E. 1998. “International Niimi, Y., C. Ozden and M. Schiff. 2008. “Migration, Migration and Business Formation in Mexi- Remittances and the Brain Drain: Causes and co.� Social Science Quarterly, 79:1–20. Linkages.� Manuscript: World Bank. Mattoo A., I. C. Neagu and C. Özden. 2008. “Brain OECD. 2008a. “Cross-Border Higher Education waste? Educated Immigrants in the U.S. and Development.� Policy Brief, OECD Ob- Labor Market.� Journal of Development server, January. Economics 87:255–269 OECD. 2008b. International Migration Out- McHale, John. 2008. “The International Migration look. OECD: Paris. Architecture: A Survey with Lessons for the Ojeda, Gabriel, Myriam Ordoñez, and Hernando MENA Region.� An output of the EC-funded Ochoa. 2005. Salud Sexual y Reproductiva World Bank program of International Migra- en Colombia: Encuesta Nacional de De- tion from the Middle East and North Africa mografía y Salud 2005. Bogotá, Colombia: and Poverty Reduction Strategies. Profamilia, and Columbia, MD. Macro Inter- McKenzie, D. J. 2005. “Beyond Remittances: national Inc. The Effects of Migration on Mexican Okolski, Marek and et al. 1995. “Trends and Pat- Households.� In International Migration, terns of Emigration: Poland (Preliminary Remittances and Brain Drain. edited by Report).� Paper presented at the Second C. Ozden and M. Schiff. The World Bank, Training Workshop on International Migra- Washington D.C. tion Surveys in Central and Eastern Europe McKenzie, David and Hillel Rapoport. 2005. “Net- (United National Economic Commission work Effects and the Dynamics of Migration for Europe and United Nations Population and Inequality: Theory and Evidence from Fund), Warsaw, Poland, 5–7 June. Mexico.� Bureau for Research in Economic O’Rourke, Michael. 2006. “Migration flows—A Analysis of Development (BREAD) Working Summary of the UK’s Recent Experience and Paper No. 063, April. the Immediate Challenges Ahead.� Presented McKenzie, David. and Hillel Rapoport. 2006. “Can at UN Expert Group Meeting on Measuring migration reduce educational attainment? International Migration, 4–7 December, New Evidence from Mexico.� Policy Research York. Working Paper Series 3952, The World Bank. Organisation for Economic Co-operation and MEDSTAT. 2008. “Morocco – Country Statistical Development and SOPEMI. 2005. Trends Situation Report.� December. in International Migration: Continuous Morrison, A. R., M. Schiff, and M. Sjöblom. 2007. Reporting System on Migration, Annual The International Migration of Women. Report, 2004 edition. Paris, France: Or- Palgrave McMillan. ganisation for Economic Co-operation and Mountford, A. 1997. “Can a Brain Drain Be Development. Good for Growth in the Source Economy?� Owasu, Thomas.2006. “Transnationalism Among Journal of Development Economics 53 African Immigrants in North America: The (2), 287–303. Case of Ghanaians in Canada.� In The New 122 References African Diaspora in North America: Trends, tion, Remittances and Household Poverty Community Building, and Adaptation, Status in Egypt.� An output of the EC-funded edited by Kwado Konadu-Agyemang, Baffour World Bank program of International Migra- K. Takyi, and John A. Arthurs. Lexicon Books. tion from the Middle East and North Africa Özden, C. (2006), “Brain Drain in Middle East and Poverty Reduction Strategies. & North Africa—The Patterns Under the Saleh, Madiha, and Ahmed Morsy. 2009a. “Sourc- Surface.� United Nations Expert Group es for International Migration Data in Egypt.� Meeting on International Migration and De- Presented at MEDSTAT II Regional Work- velopment in the Arab Region, Population shop on Developing Model Questionnaires in Division, Department of Economic and Social Coordinated Household International Migra- Affairs United Nations Secretariat, UN/POP/ tion Surveys in the Mediterranean countries EGM/2006/10, Beirut, 15–17 May. (MED-HIMS). Brussels, February 16–19. Parsons C. R., R. Skeldon, T. L. Walmsley, and L. Saleh, Madiha, and Ahmed Morsy. 2009b. “Egyp- A. Winters. 2007. “Quantifying International tian Labor Migration Project: survey design Migration: a Database of Bilateral Migrant and main findings.� Presented at MEDSTAT Stocks.� Policy Research Working Paper II Regional Workshop on Developing Model 4165, World Bank. Questionnaires in Coordinated Household Pearce, S.C. 2006. “Immigrants Women in the International Migration Surveys in the Medi- United States: A Demographic Portrait.� terranean countries (MED-HIMS). Brussels, Special report, Immigration Policy Center, February 16–19. Summer 2006. Sasin, Marcin. 2008. “Morocco’s migration: the Petroni, Rita and Danny Childers. 2003. “Cover- profile and the impact on households.� An age Measurement from the Perspective of output of the EC-funded World Bank pro- March 2001 A.C.E.� U.S. Census Bureau, gram of International Migration from the Washington, D. C. Middle East and North Africa and Poverty Pissarides, Chrisopher. 2000. “Human capital and Reduction Strategies. growth: a synthesis report.� OECD Develop- Sasin, Marcin and David McKenzie. 2007. “Migra- ment Center Technical Paper No. 168. tion, remittances, poverty, and human capi- Poutvaara, P. 2004. “Public Education in an tal: conceptual and empirical challenges.� Integrated Europe: Studying to Migrate and Policy Research Working Paper 4272. World Teaching to Stay.� CESifo Working Paper, Bank, Washington DC No 1369. Schiff, M. (2005), “Brain Gain: Claims About its Pradhan G., M. Upadhyay, and K. Upadhyaya. Size and Impact on Welfare and Growth are 2004. “Remittances and Economic Growth Greatly Exaggerated.� IZA discussion paper in Developing Countries.� Conference paper No 1599. for Midwest Economics Association. Schoorl, Jeannette J., Liesbeth Heering, Ingrid Pritchett, Lant. 1999. “Has Education Had a Esveldt, George Groenewold, Rob F. van der Growth Payoff in the MENA Region?� World Erf, Alinda Bosch, Helga de Valk, Bart J. de Bank Middle East and North Africa Work- Bruijn, European Commission, and Statistical ing Paper 18. December. Office of the European Communities. 2000. Rodriguez, E. R. and E. R. Tiongson. 2001. “Tem- Push and Pull Factors of International porary Migration Overseas and Household Migration: A Comparative Report. Luxem- Labor Supply: Evidence from Urban Philip- bourg: Office for Official Publications of the pines.� International Migration Review European Communities. 35(3), 709–25. Silva, Joana. 2008. “The Effects of Remittances Romer, P. 1990. “Endogenous Technological on Labor Markets: Evidence from Morocco.� Change.� Journal of Political Economy 98 An output of the EC-funded World Bank (5), part 2, 71–102. program of International Migration from the Roushdy, Rania, Ragui ASsaad, Ali Rashed, and Middle East and North Africa and Poverty Dalia Hassanein. 2008. “International Migra- Reduction Strategies. 123   Labor Migration from North Africa – Development Impact, Challenges, and Policy Options Sirken, Monroe G. 1970. “Household Surveys Odense, University Press of Southern Den- with Multiplicity.� Journal of the American mark. Statistical Association 65 (239): 257–266. Unesco. 2006. Global Education Digest 2006: Sirken, Monroe G. 1998. “A Short History of Comparing Education Statistics Across Network Sampling.� Proceedings of the the World. Institute for Statistics, Montreal, American Statistical Association: 1–6. Canada. Som, Ranjan Kumar. 1973. Recall Lapse in United Nations. 2004. International Migration Demographic Enquiries. New York: Asia and Development. New York: United Na- Publishing House. tions, Department of Economic and Social Somoza, Jorge L. 1977. “Una Idea para Estimar Affairs. la Población Emigrante por Sexo y Edad en United Nations. 2005. World Economic and el Censo de un País.� Notas de Población Social Survey, 2004: International Migra- 89: 106 (CELADE). tion (Sales No. E/2004/75/Rev.1/Add.1). Somoza, Jorge L. 1981a. “Indirect Estimates of New York, NY: United Nations. Emigration: Application of Two Procedures United Nations. 2006a. International Migra- Using Information on Residence of Children tion and Development: Report of the and Siblings.� In Indirect Procedures for Secretary-General (A/60/871). New York: Estimating Emigration, International United Nations. Union for the Scientific Study of Popula- United Nations. 2006b. “Trends in Total Migrant tion (IUSSP) Papers No. 18: 35–60. Liege, Stock: The 2005 Revision.� POP / DB / MIG / Belgium: IUSSP. Rev. 2005. Database in digital form, and Wall Somoza, Jorge L.1981b. “A Proposal for Estimat- Chart International Migration 2006. ing the Emigrant Population by Sex and Age United Nations Development Programme. from Special Census Questions.� In Indirect 2009. Human Development Report 2009: Procedures for Estimating Emigration, Overcoming Barriers: Human Mobility International Union for the Scientific and Development. New York: United Na- Study of Population (IUSSP) Papers No. tions. 18: 3–18. Liege, Belgium: IUSSP. van Dalen, Hendrik P., George Groenewold, and Stark, O. and Y. Wang. 2002. “Inducing Human Tineke Fokkema. 2005. “The Effect of Remit- Capital Formation: Migration as a Substitute tances on Emigration Intentions in Egypt, for Subsidies.� Journal Public Economics Morocco, and Turkey.� Population Studies, 86(1), 29–46. 59(1): 375–392. Stark, O., C. Helmenstein and A. Prskawetz. 1997. van der Erf, Rob; Liesbeth Heering. 2002. “Moroc- “A Brain Gain with a Brain Drain.� Econom- can Migration Dynamics.� IOM ics Letters 55, 227: 34. Vidal, J.P. 1998. “The effect of Emigration on Hu- Stark, O., C. Helmenstein and A. Prskawetz. 1998. man Capital Formation.� Journal of Popula- “Human Capital Depletion, Human Capital tion Economics 11(4): 589–600. Formation, and Migration: A Blessing or a Viprey, Mouna. 2008. “La politique migra- ’Curse’?� Economics Letters 60 (3), 363: 7. toire française dictée par les besoins Sudman, Seymour, M. Sirken, and C. Cowan. économiques.� An output of the EC-funded 1988. “Sampling Rare and Elusive Popula- World Bank program of International Migra- tions.� Science, 240 (20 May): 991–996. tion from the Middle East and North Africa Thailand, National Statistical Office (NSO). 2007. and Poverty Reduction Strategies. The Labor Force Survey: Whole Kingdom, Wahba, J. 2007. “An Overview of International Quarter 4: October – December 2006. Re- & International Migration in Egypt.� ERF port in Thai and English. Bangkok: NSO, Working Paper, 0703. Ministry of Information and Communication Wahba, J. 2004. “Does International Migration Technology. Matter? A Study of Egyptian Return Mi- Tranaes, T. and K. F. Zimmermann. 2004. grants,� Arab Migration in a Globalized Migrants, Work, and the Welfare State, World, Genova, IOM. 124 References Wood, Charles. 1982. “Equilibrium and historical- from Philippine Migrants’ Exchange Rate structural perspectives on migration.� Inter- Shocks’, Economic Journal, forthcoming. national Migration Review 16(2): 298–319. Zaba, Basia. 1986. Measurement of Emigra- Woodruff, C. and Zenteno, R. 2007. “\Migration tion Using Indirect Techniques: Manual Networks and Microenterprises in Mexico.� for the Collection and Analysis of Data Journal of Development Economics 82: on Residence of Relatives. Liège, Belgium: 509–28. Ordina Editions. World Bank. 2005. A better Investment Climate Zaba, Basia. 1987. “The Indirect Estimation of for Everyone, World Development Report Migration: A Critical Review.� International 2005, World Bank, Washington D.C. Migration Review 21:1395–1445. World Bank. 2006. “Fostering Higher Growth and Zekri, A. Ben Haj. 2008. “Migration, Diaspora Employment with Productive Diversification et Développement  : Le Cas de la Tunisie.� and Competitiveness.� Morocco, Country unpublished background paper. Economic Memorandum, Report 32948- Zimmermann, K. 2005. European Migration: MOR, vol. 2, Washington DC. What Do We Know? Oxford: Oxford Uni- World Bank. 2006. Global Economic Prospects versity Press. 2006. World Bank, Washington D.C. Zlotnik, H. 1990. “International Migration Policies World Bank. 2007. Job creation in a High and the Status of Female Migrants.� Interna- Growth Era: MENA Economic Develop- tional Migration Review 24 (2): 372–381. ment and Prospects 2007. World Bank, Zlotnik, H. 1992. “Empirical Identification of Washington D.C. International Migration Systems.� In Inter- World Bank. 2008. The Road Not Traveled: national Migration Systems: A Global Ap- Education Reform in the Middle East and proach, edited by M. M. Kritz, L. L. Lim, and North Africa. MENA Development Report, H. Zlotnik, 19–40. Oxford, UK, and New York: World Bank, Washington D.C. Clarendon Press, Oxford University Press. World Bank. 2009. Shaping the Future: A long Zlotnik, H. 1995. “The South-to-North Migra- term perspective of people and job mobility tion of Women.� International Migration for the Middle East and North Africa. The Review 29 (1), 229–254. World Bank, Washington D.C. Zohry, Ayman. 2003. Contemporary Egyptian Yang, D. (2008). ‘International Migration, Remit- Migration, 2003. Cairo: Ministry of Man- tances, and Household Investment: Evidence power and Emigration. 125 THE WORLD BANK GROUP 1818 H Street, N.W. Washington, D.C. 20433 USA www/worldbank.org