Evidence on Policies to Increase the Development Impacts of International Migration David McKenzie and Dean Yang International migration offers individuals and their families the potential to experience im- mediate and large gains in their incomes and offers a number of other positive benefits to the sending communities and countries. However, there are also concerns about the poten- tial costs of migration, including concerns about trafficking and human rights, a desire for remittances to be used more effectively, and concerns about a loss of externalities from skilled workers. As a result, there is increasing interest in policies that can enhance the de- velopment benefits of international migration and mitigate these potential costs. We provide a critical review of recent research on the effectiveness of these policies at three stages of the migration process: pre-departure, during migration, and directed towards pos- sible return. The existing evidence base suggests some areas of policy success: bilateral mi- gration agreements for countries whose workers have few other migration options, developing new savings and remittance products that allow migrants more control over how their money is used, and efforts to provide financial education to migrants and their families. Suggestive evidence, together with theory, offers support for a number of other policies, such as lowering the cost of remittances, reducing passport costs, offering dual cit- izenship, and removing exit barriers to migration. Research offers reasons to be cautious about some policies, including policies enforcing strong rights for migrants, such as high minimum wages. Nevertheless, we find the evidence base to be weak for many policies, with no reliable research on the impact of most return migration programs or whether countries should attempt to induce communal remitting through matching funds. Migration policy, Remittances, Return Migration, Impact Evaluation JEL codes: O15, F22 The World Bank Research Observer # The Author 2015. Published by Oxford University Press on behalf of the International Bank for Reconstruction and Development / THE WORLD BANK. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com doi:10.1093/wbro/lkv001 Advance Access publication January 20, 2015 30:155–192 Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 International migration offers individuals the potential to experience immediate and large gains in their income, with these gains far exceeding those from any other development policy intervention (Clemens, Montenegro, and Pritchett 2009; McKenzie, Gibson, and Stillman 2010). These gains are typically shared with the household members who accompany migrants and, through remittances, with the family members who remain in their home country. Both the number of migrants and the remittances they send have grown rapidly over the past two decades. The number of people living outside their country of birth totaled 232 million in 2013, up from 154 million in 1990,1 whereas recorded remittances sent to developing countries totaled $321 billion in 2010 compared to only $49 billion in 1990.2 In addition to the direct effects of higher incomes, research has found that migration provides a number of other positive benefits to the sending country, including facili- tation of trade, technology transfer, foreign direct investment, incentives for greater investment in globally marketable skills, and the transmission of democratic norms. However, these positive impacts of migration are accompanied by concerns over a number of potential costs. These include concerns about human trafficking and abuse of migrant rights; concerns about the abilities of households that receive large, temporary, flows of remittances to save appropriately; concerns that the bene- fits of migration do not spread far beyond the immediate household; and concerns that developing countries are losing the positive externalities of highly skilled workers. There is increasing interest among development institutions and developing country governments in identifying policies that can enhance the development impacts of international migration by facilitating more of the benefits and mitigat- ing some of the potential costs. This interest has led to the deployment of a wide range of migration policies that occur throughout the different stages of the migra- tion process ( pre-departure, while the migrant is abroad, and upon possible return) and that cover a wide range of different areas (e.g., financial sector, social protec- tion, employment policies). However, the substantial policy interest and growing policy efforts stand in stark contrast to the limited empirical evidence that can help to guide policy. The vast majority of existing research on migration and development is focused on measur- ing and understanding the determinants of migration and the impacts of migration and remittances on development rather than on the effects of policies designed to change these impacts. This situation has begun to change in the last few years, with a number of innovative studies explicitly testing the impacts of migration poli- cies relative to a counterfactual of some alternative policy that could be implement- ed. In this paper, we review this new literature and assess the state of evidence on the impacts of different policies intended to improve the development impacts of in- ternational migration. This means that we do not consider policies whose primary 156 The World Bank Research Observer, vol. 30, no. 2 (August 2015) Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 goal is the well-being of firms and workers in migrant-receiving developed coun- tries. We focus on international migration and do not discuss internal migration policies. Rather than summarizing the results in the literature, we also offer our as- sessment of the rigor and reliability of these results to identify areas where there is considerable policy effort without reliable evidence of its effects. We note that the existing literature has largely addressed the question, “Does a particular policy work as intended?” rather than “Should this particular policy be implemented?” As we will discuss, the evidence base on whether policies work as in- tended is still rather weak in many policy areas, and even fewer studies measure the costs of implementing such policies. In addition, because the first-order effects of migration occur largely for the migrants themselves and their families, some readers may question the rationale for spending public money on increasing these benefits. There are several key reasons for this rationale. First, some migration poli- cies provide public goods, such as the legal infrastructure in which migration takes place and the regulatory policies under which remittances occur. Second, a number of more activist policies are intended to overcome market failures in the informa- tion, credit, insurance, and labor markets. Third, other policies, such as home-town association programs, are intended to overcome collective action problems. Fourth, a set of policies is designed to reduce the perceived negative externalities of highly skilled emigration. Fifth, other policies are designed to reduce inefficiencies associat- ed with behavioral biases and intra-household decision-making under asymmetric information. In the presence of these constraints, the existing levels of migration and remittances are likely to differ from the socially optimal levels. The result is that many governments are implementing policies designed to increase the development impacts of migration, and it is therefore of interest to understand what the effects of such policies have been. We begin with a discussion of policies that occur at the pre-migration stage. These include efforts to facilitate more migration and pre-departure training ses- sions to provide migrants with more information and skills. We then turn to policies directed towards migrants and their families while migrants are abroad. These include rights policies, efforts to expand financial access, policies to make it easier and cheaper for migrants to remit, policies to channel remittances towards purposes with broader development impacts, policies to encourage communal remittances and the development of home-town associations, and integration policies. Finally, we examine policies directed towards returning migrants. These include efforts to remove regulatory, bureaucratic, and informational barriers that inhibit return mi- gration, policies intended to change the incentives to return, and policies designed to make return migrants more productive and reintegration easier. We find the evi- dence base to be weak for many of these policies, especially those directed towards return migration, and we discuss possible reasons for this lack of research before concluding. McKenzie and Yang 157 Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 Pre-Migration Policies Much of international migration takes place without explicit policy actions by devel- oping country governments. Destination countries set their immigration policies, and individuals who are eligible to move under these policies decide whether to do so, whereas others migrate illegally. However, survey results show that many more people around the world express a desire to move to another country than currently do so (Esipova, Ray, and Pugliese 2011), a number of migration quotas, such as the US Diversity Visa program (Green Card lottery) and the H1B visa category, are heavily oversubscribed, and many potential migrants remain misinformed of the potential gains from moving (McKenzie, Gibson, and Stillman 2013). Bureaucratic, financial, informational, and other constraints prevent many individuals who would benefit from moving from doing so. As a result, one key avenue for migration policy to enhance development impacts is by facilitating more people’s migration and ensuring they are well informed when doing so. The main policy levers used to do this are explicit policy efforts to facilitate migration and pre-departure orienta- tions or training programs designed to ensure that migrants are well informed about conditions abroad. Table 1 summarizes these types of policies and the evi- dence for each, which we discuss in detail below. Policy Efforts to Facilitate (or Hinder) Migration There are two categories of policy actions that developing country governments have taken to facilitate international migration. The first consists of bilateral actions, which involve cooperation with governments or employers in destination countries, leading to formal agreements to allow labor migration of specified numbers and types of workers. A leading example of this type of policy action has been the Philippines, which has signed at least 49 bilateral migration agreements with 25 destination countries. Large migration flows have occurred through some of these migration agreements. Given the importance of legal barriers in regulating migration, it seems likely that such agreements have allowed more migration from the signatory countries. However, we are unaware of research that has identified how much of this migration would have happened in the absence of such signed agreements; hence, there is a lack of evidence on the causal effect of such policies. One exception to this lack of research comes from bilateral migration policies ne- gotiated between New Zealand and several Pacific Island nations (including Samoa, Tonga, and Vanuatu) to allow seasonal migration to New Zealand through the Recognized Seasonal Employer (RSE) program. Gibson and McKenzie (2014) examine the impacts of this program on households in Tonga and Vanuatu. They find that individuals who participate in the program lack almost any other opportu- nity to migrate internationally; therefore, the program generates new migration, 158 The World Bank Research Observer, vol. 30, no. 2 (August 2015) Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 McKenzie and Yang Table 1. Examples of Pre-Migration Policies Potential impact on Policy Examples What does it do? development Evidence Pre-departure training Training and orientation Philippines’ Pre- Departure Eases migrant adjustment Increases worker Non-existent, but one study (Omar of workers in legal rights, Orientation Seminars (PDOS); to new country and work productivity and Mahmoud et al) ongoing. customs and language IOM migrant training sessions environment and earnings; allows overseas, productive protects in cases of abuse workers to escape skills, etc. abuse Financial literacy training Included in Philippine PDOS; Improves financial Raise savings and Evidence from RCTs among Indian for workers and their small programs implemented as decision-making among investment; improve workers in Qatar (Seshan and Yang families part of research studies migrants and their ability to cope with 2014), Indonesian migrants and their families left behind risk families (Doi, McKenzie, Zia 2014), and among Pacific Island migrants to Australia and New Zealand (Gibson et al. 2014) Policies to facilitate migration Bilateral migration Philippines with 25 countries Legal agreement to allow Allows more Strong evidence of positive effects from agreements Pacific Islands with labor migration of set individuals to migrate the New Zealand RSE program through New Zealand numbers or types of abroad matched difference-in- differences workers (Gibson and McKenzie 2014) Lower passport costs and Many countries Makes it less costly to Makes it less costly to Suggestive evidence from regressions with procedures migrate migrate other controls (McKenzie, 2007) Remove exit barriers Many countries don’t have these Makes it less costly to Allows more Suggestive evidence from regressions with barriers, but a number restrict migrate individuals to migrate other controls (McKenzie, 2007). women migrating, and others abroad require govt permission for all emigration 159 Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 and this migration results in large increases in income and consumption for the participating households. These authors use a matched difference-in-difference strategy to identify these program impacts, arguing that the newness of the program and the limited availability of places in the program facilitates finding com- parable households to those selected to participate in the program. The second set of actions countries can take are unilateral actions, which do not require the cooperation of other governments. These actions are most evident when used to hinder migration. For example, a number of countries restrict the rights of women to migrate (e.g., Gabon, Libya, Saudi Arabia), and others require all citizens to obtain government permission to travel abroad (e.g., Cuba, Iran, North Korea). McKenzie (2007) shows in a linear regression that countries with such restrictions have 5 to 6 percent fewer migrants per capita than countries with similar income, population, and governance levels that do not have these restrictions. Countries can also affect the ease of migration by imposing high passport fees and cumbersome procedures for obtaining a passport. McKenzie (2007) shows that the cost of a pass- port exceeds 10 percent of per capita income in some countries, and higher pass- port costs are associated with less migration. This association continues to hold after controlling for population, per capita income, and government effectiveness. Governments can also take more positive unilateral actions to facilitate migra- tion. Beam, McKenzie, and Yang (2014) conducted a large-scale randomized experiment in the Philippine province of Sorsogon to test the impact of unilaterally facilitating international labor migration. Households assigned to treatment groups received one or more of several possible interventions intended to facilitate international migration. The interventions targeted the following reasons people might not migrate internationally: 1) information barriers (information about job search, migrating abroad, financing migration, and passport processing); 2) frictions in job search (assistance in enrolling in an online job-finding website to lower search costs and facilitate matching between recruiters and workers); and 3) documentation barriers (assistance and a full subsidy for passport application). Overall, the authors find no evidence that any of the individual interventions or combinations of interventions led to increases in international migration. The most intensive treatment that combined all these interventions led to increases in job search and doubled the rate of international job offers, but it had no identifi- able effect on international labor migration. The authors conclude that unilateral migration facilitation will, at most, induce a trickle, not a flood, of additional emigration. A related study was conducted by Beam (2014), also in Sorsogon province, Philippines. The study randomly assigned participants to a control group or one of a set of treatment groups. The treatments involved offering information on wages and qualifications for typical overseas jobs or the provision of an incentive (a restau- rant voucher) conditional on attending a “job fair” (an event at which job-seekers 160 The World Bank Research Observer, vol. 30, no. 2 (August 2015) Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 connect with placement agencies for overseas jobs). The study found that the infor- mation treatments improved knowledge about overseas wages and job qualifica- tions. Neither information treatments nor the job fair incentive treatment led to increases in search activity for overseas jobs. Along with the previous study, this work suggests limits to the ability of unilateral facilitation policies to foster more migration. Pre-Departure Orientation or Training Programs A number of programs are designed to provide potential migrants with some type of orientation or training prior to departure. These include programs aimed at easing the transition to life abroad, programs that provide technical skills to enhance pro- ductivity abroad, programs designed to prevent human trafficking and abuse, and programs designed to teach financial literacy (which we discuss separately in the next sub-section). Pre-departure orientations have been implemented in a variety of contexts. Perhaps the most prominent is the Philippine government’s Pre-Departure Orientation Seminar (PDOS), a one-day session that has been required since 1983 for labor migrants departing for new jobs overseas and is administered by the government as well as by non-government organizations (NGOs) under contract with the government. The stated aims are to prevent abuse of migrant workers, protect their labor rights, and educate workers on the laws, culture, and customs of destination countries. Additional topics covered include health and safety, financial literacy, and travel procedures and tips. Starting in 2009, the PDOS was expanded to include 4 –6 days of training for migrant household service workers on job skills, first aid, language, culture, and stress management (known as the Comprehensive Pre-Departure Education Program, or CPDEP) (International Labor Organization 2013). Other important examples of pre-departure training programs are the wide range of programs conducted by the International Organization for Migration (IOM). The IOM has been conducting training for migrants since the 1950s, with 352,000 mi- grants involved in their training programs over the 2001 –10 decade (International Organization for Migration 2011). Impact evaluation of pre-departure orientation training has been minimal thus far. A large number of evaluation reports have been conducted, but to our knowl- edge, these are primarily “process” evaluations that involve audits of activities via examination of internal documents, site visits, and in-depth interviews with key internal and external stakeholders. Examples include Norad (2010) and Asis and Aguinas (2012). Of the more limited number of impact evaluations, to our knowl- edge, none use credible research designs that can allow the identification of causal effects of the programs on migrant welfare or other outcomes. For example, McKenzie and Yang 161 Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 IOM (2006) seeks to assess the impacts of an anti-trafficking informational program in Cambodia on knowledge of anti-trafficking messaging among individu- als remaining in Cambodia (non-migrants). It uses survey data on program partici- pants and compares the responses to surveys of non-participants, finding that participants have better knowledge about information relevant for anti-trafficking. The study does not describe how study participants and control group respondents were selected, so it is not possible to assess whether the two populations can be credibly compared to establish the causal impact of the program. Given the popularity of these programs, it is important to conduct prospective, randomized evaluations of pre-departure trainings. One ongoing study (Mahmoud et al. 2013) is doing this by randomly assigning departing labor migrants from the Philippines to different types of modified pre-departure orientation seminars and tracking impacts on migrants over time. Financial Literacy Programs for Migrants and Their Families (Whether Before or After Departure) Due to the large wage gains possible when individuals migrate from developing to developed countries and the large amounts sent home by migrants in the form of re- mittances, migrant-origin households are very often faced with managing amounts of money that are substantially larger than the household budgets of those without migrant members. This situation has raised concerns among policy-makers and non-government organizations that financial decision-making in migrant house- holds may be suboptimal, particularly in households whose members migrated rela- tively recently. Motivated by such concerns, a common type of intervention provides financial literacy or financial education training to migrant workers and their fami- lies who remain behind in the home country. Financial literacy programs are com- monly conducted as part of pre-departure orientation training programs (discussed in the previous section). Important evidence on the impact of providing financial education to migrant workers and their families is provided by Doi, McKenzie, and Zia (2014), who imple- mented a randomized controlled trial among Indonesian women about to depart for overseas work as household servants (maids) and these women’s families. Study participants were randomly assigned to either a control group (that received no training) or to one of three treatment groups in which financial literacy training was provided prior to the migrant’s departure for overseas for either 1) the migrant alone, 2) a left-behind family member alone, or 3) both the migrant and the family member. The training covered financial planning and management, savings, debt management, sending and receiving remittances, and migrant insurance. The study examined impacts on families that remained behind in Indonesia, finding that each type of treatment increased financial knowledge. In addition, training both the 162 The World Bank Research Observer, vol. 30, no. 2 (August 2015) Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 migrant and family led to increases in savings in the origin household, but the other two treatments (migrant only and family only) did not have a similar savings impact. None of the treatments had substantial impacts on remittances received. The study highlights the complementarities of training both migrants and family members. Financial literacy training also occurs frequently at the destination, targeting the migrant. Seshan and Yang (2014) examined such a program for married male migrant workers in Doha, Qatar who were from Kerala, India and whose wives re- mained behind in India. Participants were randomly assigned to either a control group or a treatment group. The treatment group was invited to attend a one-time motivational session on personal financial management that stressed the impor- tance of savings and of making joint financial decisions with spouses who remained behind in India. Impacts were measured via follow-up surveys of both the migrants in Kerala and their wives in India. The treatment led both the migrants and their wives to be more likely to report that they made joint financial decisions with their spouses, and wives became more likely to seek out additional financial education themselves. Among migrants with lower savings at baseline ( prior to treatment), the treatment led to higher total (Qatar plus India) household savings and higher remittances sent by migrants to wives. Gibson et al. (2014) implemented a randomized controlled trial of a financial lit- eracy intervention among migrants in Australia and New Zealand. The study coin- cided with the introduction of a new online tool for comparing remittance transaction fees across providers and the introduction of a new remittance method in New Zealand ( provision to migrants of a second ATM card that could be sent to family members back home to allow withdrawals from the migrant’s New Zealand bank account). The study sample consisted of migrants from Tonga, East Asia, and Sri Lanka. Study participants were randomly assigned either to a control group that received no training or a treatment group that was invited to a financial education session that focused on helping migrants compare among different remittance- sending methods. The authors find that the treatment led to increases in financial knowledge and in seeking information about remittance methods for the Tongan and East Asian participants. The authors also find no effects of the treatment on re- mittance frequency or total amounts of remittances sent. The authors consider this absence of a treatment effect on remittance outcomes to be partly due to barriers to the use of alternative remittance methods for remittance receivers. These studies suggest the diversity of possible impacts of different types of inter- ventions in different populations. A particular financial education intervention could have different impacts in different subpopulations, suggesting that differenti- ated interventions targeted at the particular needs of subpopulations would be worth exploring. Future studies could also examine the impact of financial literacy interventions among family members after the migrants have departed for overseas. McKenzie and Yang 163 Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 Policies while Migrants are Abroad The widest range of policies designed to enhance the development impacts of migra- tion occur once migration has taken place. Table 2 summarizes the range of policy instruments and the evidence for their effectiveness. A first set of policies involves ensuring adequate rights for migrants, with debate as to which rights should be protected. A second broad area of policy concerns financial access and remittances, with policies to ensure that migrants and their families have access to appropriate financial products, efforts to lower the cost of sending remittances, and policies to encourage migrants to contribute more towards activities with broader develop- ment outcomes. A related area attempts to encourage migrants to form home-town associations to build communal public goods. Finally, policies directed towards mi- grants who intend to stay abroad can affect their ability to integrate successfully into their new countries. Rights Policies for Migrants The United Nations International Convention on the Protection of the Rights of All Migrant Workers and Members of Their Families (CMW) of 1990 stipulates a com- prehensive set of civil, political, economic, and social rights for migrants. It is based on the principal of equal treatment of migrants and nationals, including the right not to have identity documents confiscated, the right to equal treatment with nationals before the courts, the right to form associations and trade unions, and the right to equal treatment with regard to remuneration. However, this convention is the least ratified of all the major international human rights treaties; by 2012, no major migrant-receiving country had ratified it (Ruhs 2013). The question that arises is whether it would be good for development for more destination countries to grant all these rights. In almost all cases, we would expect migrants who experience more rights to be directly better off as a result. However, a number of these rights impose costs on the host country government or the host country employers. Consequently, we should expect to see greater rights for migrants resulting in either lower wages or in fewer migrants being hired. Ruhs (2013) documented this pattern across countries, finding that countries that provide migrants with more rights in terms of access to retirement benefits, unemployment benefits, and public education have less open migration programs. McKenzie, Theoharides and Yang (2014) examined a particu- lar type of economic right: the right to a minimum wage or to a wage equal to the prevailing wage for natives. They conduct a difference-in-differences analysis of the effect of a minimum wage change for Filipina domestic workers. They found that doubling the minimum wage that these workers must be paid resulted in higher wages for the workers who migrated but also resulted in a 55 to 57 percent 164 The World Bank Research Observer, vol. 30, no. 2 (August 2015) Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 McKenzie and Yang Table 2. Examples of Policies While Migrants are Abroad Policy Examples What does it do? Potential impact on development Evidence Rights policies for migrant workers Comprehensive Requiring migrants to be paid Sets minimum wages that Can raise the income of the Evidence shows a trade-off economic rights the same wages as nationals migrants must be paid; migrants receiving the policy, between the number of of the destination country potentially also gives migrants but reduce the demand of individuals who get to migrate access to unemployment employers for labor, and and the economic rights they benefits and welfare programs openness of governments receive (Ruhs 2013; McKenzie et al. 2014). Basic rights Freedom to leave, freedom Protects migrants from abuse, Ensures migrants receive No empirical evidence to measure from abuse, ensuring gives them the ability to leave, income promised, and impacts of changes in these migrants receive pay for the and ensures they get paid for protects against physical and rights, but theory strongly jobs they do work done sexual abuse suggests they should be beneficial for development Financial access for migrants and their families Facilitating remittance- El Salvador’s Banco Agricola Allows migrants to set up Raises asset accumulation and Evidence from RCT among linked savings accounts accounts offered to migrants accounts in home country and allows self-insurance via migrants from El Salvador in origin country in US remit funds directly into them buffer stock accumulation (Ashraf et al. forthcoming) that offering migrants accounts in home country can raise savings, but only when migrants have joint or sole ownership of the accounts Facilitating savings Matricula consular (consular Formal identification document Raises asset accumulation and Evidence from RCT among accounts for migrants ID card) issued to Mexican facilitates bank account allows self-insurance via Mexican migrants (Chin et al. in the host country migrants by Mexican opening in US buffer stock accumulation 2011) that issuing matricula consulate consular raises account opening and savings in host country Continued 165 Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 Table 2. Continued 166 Policy Examples What does it do? Potential impact on development Evidence Policies to lower remittance costs and to induce migrants to remit more Lower remittance prices Remittance price comparison Allows migrants to take Fee savings for migrants; could Evidence from RCTs among websites; reductions in advantage of lower remittance raise total money amount of Salvadoran migrants (Aycinena remittance fees fees remittances sent over and et al. 2010) and among above remittance fee savings Guatemalan and Salvadoran migrants (Ambler et al. 2014) that remittance fee discounts lead to large increases in dollar value of remittances sent, over an above remittance fee savings. However, no rigorous evidence on how much policies have The World Bank Research Observer, vol. 30, no. 2 (August 2015) contributed to lowering remittance fees. Improve communication Reductions in Improves communication and Higher remittances; better Evidence from lab-in-field and/or information telecommunications costs reduces information monitoring by migrants of experiment among Salvadoran flows within asymmetries within remittance recipient migrants that migrants send transnational transnational households expenditures more remittances when households recipients are made aware of an income windfall (Ambler 2013); evidence from RCT among immigrants in Ireland (Batista and Narciso 2013) that reducing. Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 McKenzie and Yang Policies to encourage communal remitting and home-town associations Setting up hometown Malian HTAs in France Joins migrants together to Can increase public goods like Promising evidence that HTAs associations potentially send collective water, roads, and electricity increase public good provision remittances from time series, difference-in- differences, Heckman selection (Beauchemin and Schoumaker 2009; Chauvet et al. 2013; Duquette-Rury 2014). No evidence on policies to increase set up of HTAs Matching funds for 3x1 program in Mexico Offers migrant association a Can lead to more remittances No evidence to show that match collective remittances government match for each being sent for community leads to more funding being sent dollar in funding they send needs. May crowd out family than would be sent otherwise; remittances. May crowd out evidence that program can end other uses of public funds up being regressive and politically targeted (Duquette- Rury 2014). Integration policies for migrants Language training Norway, Germany, Sweden Teaches migrants to speak Can increase job prospects, Strong associations between destination country language and hence earnings of language proficiency and migrants; may help in social earnings in many countries, but integration little evidence on impact of training. Continued 167 Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 168 Table 2. Continued Policy Examples What does it do? Potential impact on development Evidence Migrant introduction Norway, Germany, Sweden Language training, job search Can increase job prospects, Mixed. Strong impacts for programs assistance, teaches civics and and hence earnings of unemployed individuals in values of host society migrants; may help in social Finland in a regression- The World Bank Research Observer, vol. 30, no. 2 (August 2015) integration discontinuity (Sarvima ¨ ki and Ha ¨ la ¨ ma ¨ inen 2012); benefits don’t exceed costs in randomized experiment adding job counselling to basic program in Sweden (Joona and Nekby 2012). Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 reduction in the number of workers going to destinations where this new law was binding. An example of a law change that granted more rights to migrant workers comes from Naidu, Nyarko, and Wang (2014), who examined a reform in the United Arab Emirates that gave migrant workers the right to change employers after their con- tract ends without having to receive a letter of no objection from their previous em- ployer. Using variation in the end date of worker contracts around the reform, they find that workers whose contract ended post-reform experienced a 10 percent in- crease in real earnings relative to workers whose contracts ended before the reform. The right to be able to change employers and move from one job to another clearly conveys benefits on migrant workers. However, the authors were not able to examine whether employers adjusted to this reform by changing their recruiting or wage-setting for new workers. It is possible that employers might be less willing to hire new workers or might offer them lower initial wages if there was a greater pos- sibility of them leaving for a different employer. Existing evidence therefore suggests that there is a trade-off between the rights of migrants to be paid equally to native workers in destination countries and the op- portunities for poor people from developing countries to dramatically increase their incomes through migration. Requiring higher minimum wages does limit the number of people who are able to migrate. In contrast, a different set of more basic rights involves freedom from abuse and exploitation and freedom to leave. This right involves migrants’ ability to retain their passports and depart the country at any time they like, to be paid the amount they are contracted to be paid, to be protected from being beaten or sexually abused by their employer, and to report any abuse that does occur and to see employers who violate these rights face the consequences of doing so. Failure to provide migrants with these rights directly reduces the devel- opment impact of their migration (lowering migrants’ incomes and their well- being) and makes it difficult to use revealed preference as an argument for positive migration benefits because violations of these rights typically are violations of the conditions assumed by migrants in making their migration decisions. Although we are unaware of studies that examine the impact of reforms in these more basic rights, there seems little economic rationale not to provide these types of rights to workers. Financial Access for Migrants and their Families Transnational households have special characteristics that have implications for the emphasis and design of financial access interventions. Reliable remittance services, in particular, are crucial for allowing migrants to support family members who are left behind in the origin country. The fact that high proportions of transnational households already use remittance services also means that it can be practically McKenzie and Yang 169 Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 feasible for financial access interventions to be linked in some way to remittance ser- vices (Hall 2010). An example is offers of savings accounts into which migrants can remit in the home country, which help them meet an often-heard policy objec- tive of seeking to raise savings levels in transnational households. Two recent randomized field experiments have provided evidence among migrants in the US of the effectiveness of different approaches to promoting savings in transna- tional households. A randomized controlled trial among migrants from El Salvador by Ashraf et al. (forthcoming) tested ways to stimulate savings in El Salvador. The study was particularly interested in whether migrants sought to control or influence how much was saved by family members in the home country, so the treatments that were offered varied in the degree to which migrants could monitor and control family members’ savings in account at a bank in El Salvador. Migrants were much more likely to open savings accounts at the partner bank in El Salvador and to accumulate more savings at the partner bank if they were assigned to the treatment condition that offered the greatest degree of monitoring and control. Strikingly, the savings impact of offering an account that was only in the name of someone else in El Salvador was much smaller in magnitude and not statistically significantly differ- ent from zero. This result reveals that the frequently made policy recommendation to foster savings in migrants’ home countries by encouraging migrants to remit directly into savings accounts of remittance recipients would be much less effective compared to interventions that also improved and encouraged migrant monitoring and control over home-country savings. A related experiment was conducted among Mexican migrants in Texas by Chin Karkoviata, and Wilcox (2011). That study was interested in understanding the impact of facilitating migrant access to savings accounts in the host country (in this case, the United States) rather than in the origin country. The study randomized Mexican migrants into either a control group or a treatment group that was given assistance in obtaining a matricula consular identity card from the Mexican Consulate. The primary benefit of this card was that it could be used as identifica- tion when opening a bank account in the United States. The study found that mi- grants in the treatment group were more likely to open United States savings accounts, accumulated more savings in the United States, and remitted less to Mexico. The study also found heterogeneity in treatment effects: for migrants who reported at baseline ( prior to treatment) that they had no control over how their re- mittances were used by recipients in Mexico, the treatment had larger positive impacts on United States bank account take-up and savings and led migrants to shift more savings from Mexico to the United States. These existing studies have documented the positive impacts of interventions to facilitate savings access for transnational households with account offers in both the host and origin countries. There is also consistent evidence that migrants have preferences for how remittance recipients in the home country use remittances, 170 The World Bank Research Observer, vol. 30, no. 2 (August 2015) Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 particularly how much of the remittances are saved. A policy lesson is that if an objec- tive is to raise savings levels in the origin country, interventions should give priority to savings services that offer migrants some ability to monitor or control savings. If the policy aim includes increasing savings by migrants in the host country, then facilitating access to host-country savings facilities can help to achieve this objective. Policies to Lower Remittance Costs and to Induce Migrants to Remit More Motivated by the positive development impacts of remittances (see Yang 2011 for an overview), several efforts are underway to seek to reduce barriers to remittance flows. In 2009, the G8 Heads of State Summit agreed to an objective of reducing the average cost of sending remittances from 10 percent to 5 percent in five years (the so-called “5X5 objective”) via policies such as improved information, transpar- ency, and promotion of competition in the money transmission market (G8 2009). An important approach that has been attempted in a variety of contexts is to improve migrants’ ability to compare remittance transaction fees. Migrants who are able to access lower-fee money transmitters may send more remittances. In addition, improved ability to comparison shop may increase competition across money trans- mitters and lead them to lower their prices. One example is Remittance Prices Worldwide, a website (http://remittanceprices.worldbank.org) maintained by the World Bank that provides remittance fee prices across 226 country corridors covering 32 major remittance-sending countries and 89 receiving countries. Governments and institutions such as the World Bank have also set up remittance-price compari- son websites targeted at specific markets, such as Mexico (www.remesamex.gob.mx), the United Kingdom (www.sendmoneyhome.org), and Australia and New Zealand (www.sendmoneyhomepacific.org); the latter were the subjects of the Gibson et al. (2014) remittances financial literacy experiment. Although remittance costs have fallen following these policy efforts, it is unclear how much of the reduction in costs is due explicitly to the policy actions per se rather than to technological advances and competition that might have lowered costs. Several recent randomized field experiments offer insights into how migrants are likely to react to reductions in remittance costs. The existing evidence suggests that response to fee reductions can be quite large. Aycinena, Martinez, and Yang (2010) implemented a randomized field experiment that estimated the causal impact of remittance transaction fees on remittances. In partnership with a money transfer operator, Salvadoran migrants in Washington D.C. were randomly assigned differently sized discounts on remittance transaction fees. The remittance fee reductions led to higher transaction frequency by remitters: each $1 fee reduction led to an additional 0.11 transactions per month, with no change in the dollar amount remitted per transaction. These remittance responses to price reductions are large in magnitude. A $1 reduction in the remittance transaction McKenzie and Yang 171 Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 fee leads to average fee savings per month of only $0.47, but the corresponding in- crease in average remittances sent per month is an order of magnitude larger, $25. In a related study, Ambler et al. (2014) confirm and build on these results. Migrants from Guatemala and El Salvador in Washington D.C. were randomly chosen to be offered $3 discounts on remittance fees (off a base price of $8) for remittances sent through the partner organization. The discounts were valid for 10 weeks for as many transactions as the respondent chose to conduct. These discounts led to large increases in the number of transactions and total amount remitted (measured using administrative data from the partner money transfer company). These effects of the discounts persisted up to 20 weeks after migrants were no longer eligible to receive them. These were real increases in remittances; follow-up survey responses helped rule out the possibility that migrants were sending remittances on behalf of others or shifting from other remittance channels. An important question that arises from the findings of both of these studies is why migrants send so much more in total remittances in response to relatively small changes in remittance fees. Ambler et al. (2014) suggest that this pattern of results could be generated if remittance recipients in the home country have reference- dependent preferences regarding their expected level of remittances, which evolve slowly as remittance levels change, and if migrants are partially naı ¨ve about the extent to which recipients’ preferences exhibit such reference dependency. With this characterization of preferences, migrants might respond to the remittance fee dis- counts by sending more remittance transactions during the discount period, inter- temporally substituting for remittances later (post-discount remittances). However, once migrants increase their remittance frequency during the discount period, recipi- ents in the home country raise their reference point for remittance receipts (now ex- pecting higher total amounts per time period). Migrants therefore do not immediately return to their previous level of remittances, but only do so in a gradual fashion. Other non-experimental work comes to similar conclusions that reductions in re- mittance fees raise remittances. Gibson, McKenzie, and Rohorua (2006) provide ev- idence on remittance responses to fee reductions from survey hypotheticals. Tongan migrants in New Zealand said that they would send substantially higher remittanc- es in response to reductions in the fixed-cost component of the remittance fee. The estimated elasticity of remittances to changes in the fixed-cost component of the re- mittance fee was -0.22. Freund and Spatafora (2005) use cross-country data to show that remittance fees are negatively correlated with total remittance flows at the country level. On the policy front, this evidence suggests that reforms that reduced migrant re- mittance fees can have larger impacts on remittance flows than might have been ex- pected. Such reforms include increases in competition in money transmission markets or improvements in information for migrants on the relative costs of differ- ent money transmission services. 172 The World Bank Research Observer, vol. 30, no. 2 (August 2015) Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 Policies to Channel Remittances towards Purposes with Broader Development Impacts A large variety of initiatives are ongoing to channel migrants’ remittances towards longer-run household investments (enterprises, housing, human capital invest- ments) as well as public goods at the community level. One example is an initiative in collaboration with a Brazilian state-owned bank, Caixa Economica Federal, and the Brazilian Service of Support to the Micro and Small Enterprise (SEBRAE) to cat- alyze small enterprise investment in Brazil by Brazilians in the United States. The project provides web-based entrepreneurship training for migrants and a suite of fi- nancial products (remittance, credit, and savings) to support the establishment and development of the new businesses in Brazil (Hall 2010). Programs that are more oriented towards promoting investments by home-town associations are covered in the next section. Evidence on the demand for and impact of mechanisms that help migrants make such investments in the home country is somewhat limited at present, but two studies provide initial insights. Both studies are focused on the use of remittances for education and seek to enhance migrants’ ability to channel funds towards the education of particular individuals (of their choice) in their home country. Ambler et al. (forthcoming) implemented a randomized controlled trial to test migrants’ demand for a mechanism that allowed them to channel remittances towards educational expenditures for a particular individual of their choice in the home country. Salvadorean migrants in D.C. were offered a new remittance product, named “EduRemesa”, that allowed them to target remittances towards the education of a specific student they selected in El Salvador. The EduRemesa did not actually control the use of the funds for education, but sponsored students were told that the funds were intended to support their education. Three treatment groups differed in the level of subsidy provided for the EduRemesa. The study found no demand at all (exactly zero take-up) for EduRemesa among migrants in the “no match” treatment group, a small amount of take-up (7 percent) in the 1:1 match treatment, and 19 percent take-up in the 3:1 match treatment. The 3:1 match treatment led to statistically significant increases in total educational expenditures on the target student, a reduction in that student’s labor supply, and an increase in the likelihood that the target student attended private school. This study therefore found no evidence that migrants have an unsubsidized or “pure” demand for control over the use of remittances for educational purposes. However, migrants do appear interested in channeling remittances towards the educational expenditures of specific students when given matching funds to do so, and when this occurs, there are positive impacts on the beneficiary students. In a complementary study, De Arcangelis et al. (2014) partnered with a Philippine bank (Bank of the Philippine Islands) with branches in Rome to design McKenzie and Yang 173 Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 and pilot test a new remittance product, called EduPay. This product allowed mi- grants overseas to channel tuition payments for particular students directly to those students’ educational institutions in the Philippines, avoiding the need to send tuition payments via family members or others in the Philippines who might not be completely trusted to make such payments reliably. Proof of concept was demonstrated by successfully implementing a total of 178 EduPay payments for 55 students in the Philippines. A lab-in-the-field experiment indicated that mi- grants are willing to remit more to beneficiaries in the Philippines when their trans- fers can be “labeled” as intended for educational expenses. The impact of allowing labeling was an increase in transfers of 15.3 percent. In addition to this “labeling”, the impact of actually channeling funds to educational institutions was relatively modest (only a 2.2 percent increase on top of the labeling). These results indicate that a remittance product that simply allowed senders to label remittances as in- tended for education could have nearly as much impact on remittance sending as a product that actually channeled payments to schools. Note that this result is incon- sistent with the finding of Ambler et al. (forthcoming), discussed above, that Salvadoran migrants had zero demand for the unsubsidized EduRemesa education- al remittance product. The inconsistency in results across these studies means that it is important to investigate the relative impacts of education-labeled vs. education- channeled remittance products in follow-up work to ascertain whether the experi- mental responses found by De Arcangelis et al. (2014) hold up in a real-world setting. Policies to Encourage Communal Remitting and Home-Town Associations A home-town association (HTA) consists of a group of migrants from the same com- munity in the migrant-origin country. Often, these associations have been sponta- neously created by migrants as a way of socializing with one another, providing services for new migrant arrivals, and potentially working together to implement social projects in their home community. Beauchemin and Schoumaker (2009) use event history models to show that, controlling for other observables, villages in Burkina Faso with a migrant association are four times as likely to have a health center, 2.8 times as likely to have a primary school, and 2.6 times as likely to have a road, suggesting that these HTAs help to build local infrastructure. Chauvet et al. (2013) examine the impact of Malian HTAs in France on the provision of local public goods in Mali using panel data analysis. They use a difference-in-differences strategy to compare changes in infrastructure in villages that have an HTA to changes in villages without HTAs. They find evidence that Malian HTAs have helped to improve schools, health centers, and water amenities in Mali. A challenge facing both studies is the non-random formation of HTAs and their activities, which can only be partially addressed by methods such as difference-in-differences. 174 The World Bank Research Observer, vol. 30, no. 2 (August 2015) Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 Nevertheless, these results suggest that HTAs can have positive effects on local in- frastructure provision, helping to alleviate concerns that collective remittances may simply substitute for public finance or end up being spent on projects that have more limited development impacts for their home communities. This finding raises the question of whether policy should attempt to encourage the formation of such associations and/or to encourage such associations to engage in more activities. One simple thing governments can do is to not stand in the way of such associa- tions. Chauvet et al. (2013) note that French law prohibited foreigners from gather- ing in associations, and the rise in Malian HTAs occurred following a change in this law. However, a number of migrant-sending countries also implement policies to actively encourage HTAs and to attempt to encourage them to send more collective remittances. The most famous example of such an approach is the Mexican tres por uno (3 Â 1) program. This program has a long history; it began in the state of Zacatecas in the 1970s, where some municipalities agreed to match the contributions of migrant as- sociations towards public works (1 for 1). In 1992, the state government added to the match so that each dollar sent by migrants was matched by a dollar each from the municipality and state (2 Â 1). In 1999, federal funding was added to the match to make it three dollars matched for each dollar contributed by the associa- tion (Garcia Zamora 2007). The program became a nationwide program in 2002, and by 2010, it had a program budget of $1.7 billion (Duquette-Rury 2014). The money is used to finance local development projects such as electrification, water, road paving and maintenance, education and health projects, and town beautifica- tion (Aparicio and Mesguer 2012). Despite its long history, until very recently, the program had not been subject to any rigorous evaluation. One key challenge is the non-random selection of munici- palities to participate in the program. Aparicio and Mesguer (2012) examine the correlates of participation and find that participation is, unsurprisingly, higher in high migration communities. However, because very poor municipalities have little migration (McKenzie and Rapoport 2007), the result is that the program can be somewhat regressive, directing state and federal funding towards somewhat wealth- ier communities. Moreover, the program was launched nationwide by President Vicente Fox of the PAN party, and the authors find that municipalities and states with greater PAN electoral support were more likely to participate. Duquette-Rury (2014) provides the first evaluation of the program that attempts to control for selective participation. She estimates the impact of participating in 3 Â 1 over the 2002 –8 period on changes in public goods infrastructure between 2000 and 2010. To attempt to deal with selection, she estimates a Heckman selec- tion equation to model participation in the program with the exclusion restriction that having a PAN incumbent in the 2000 electoral cycle, conditional on vote shares in later elections, predicts participation in the program but has no McKenzie and Yang 175 Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 independent effect on public goods provision. Under this assumption, she finds 3 Â 1 program expenditures to significantly and positive affect household access to sanita- tion, water, and drainage in participating rural villages. However, she also finds that households receive less family remittances as collective remittances to their munici- palities increase. Taken together, this evidence supports the idea that HTAs can increase the provi- sion of local infrastructure. A number of qualitative studies have expressed concerns about the extent to which these projects are sustainable, with funding not often being provided for maintenance (Torres and Kuznetsov 2006). However, the larger concern from the policy perspective is twofold: first, it is empirically unclear to what extent the matching funds lead more such projects to be undertaken versus crowd- ing out funding that the associations would otherwise provide on their own. In par- ticular, evidence from the charitable giving literature has found that 2 for 1 and 3 for 1 matches lead to no increase in giving relative to 1 for 1 matches (Karlan and List 2007). Second, given the sometimes regressive nature of such programs, it is unclear whether public funding devoted to this program is better for development than other uses for this social funding. For example, in Mexico, SEDESOL also runs the Progresa/Oportunidades program, and it is far from clear that additional dollars going towards 3 Â 1 have a greater impact than using this funding for cash trans- fers to poor households. Because a number of other countries, including Haiti, Somalia, the Philippines, Peru, and Colombia (Duquette-Rury 2014), are looking to implement their own co-financing programs with HTAs, additional evaluations to answer these questions are important. In addition to promoting collective remittances, several countries attempt to en- courage their migrants abroad to invest in projects in the origin country. For example, the Philippines Embassy in Qatar brings agricultural projects to pitch as potential investments to migrants there. Most of these programs seem relatively small in scale, and it is unclear whether the government has a comparative advan- tage in identifying investment projects. A notable example is the Mi Comunidad (My community) project that was launched in 1996 by the Mexican state of Guanajuato. The idea was to use investments by migrants to start small maquiladora garment manufacturing plants in migrants’ communities of origin. The state pro- vided technical assistance, worker training, and assistance to get the projects working. Torres and Kuznetsov (2006) examined the program around 2000 and viewed it as highly promising, noting that it had set up 21 plants and generated 500 permanent jobs in the home communities. However, Iskander (2005) noted that the program was a dramatic failure, with only four maquilas surviving after five years, and those that remained were struggling. She notes several issues: the very physical isolation and lack of opportunity that was a spur for migration also isolated plants from production and supply chains, and workers, once trained, often migrat- ed for better prospects in larger cities or abroad. 176 The World Bank Research Observer, vol. 30, no. 2 (August 2015) Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 Integration Policies for Migrants A number of destination countries offer public programs designed to facilitate the eco- nomic and social integration of immigrants. In some cases, these are voluntary, whereas in other cases, they are compulsory for certain groups of migrants, such as asylum-seekers or migrants receiving welfare benefits. These programs have been used for some time in the Nordic countries, but they have also been introduced recent- ly in several other countries, such as Germany (Rinne 2013). Such programs can include language training, assistance finding jobs, and information on the culture and norms of the country. These programs can improve development outcomes to the extent that they allow migrants to access better jobs, earn higher incomes, and/or have better mental health and subjective well-being through easier assimilation. There is a large body of literature that shows associations between earnings and proficiency in the language of the destination country (e.g., Chiswick and Miller 1995), although some evidence suggests that the returns to knowing the native lan- guage can be relatively low for low-skilled occupations (Berman, Lang, and Siniver 2003). Despite this finding, there is very little rigorous evidence that shows that stand-alone language training for migrants is successful in increasing language profi- ciency and raising employment outcomes. A before-after study of migrants in Canada by Weiermair (1976) found gradual improvement in immigrant wages after under- taking a language course, but it is unclear how much of these wage gains would have occurred anyway. Hayfron (2001) uses an instrumental variables approach to measure the impact of language training in Norway and find positive impacts on pro- ficiency but no impact on earnings. However, the instruments used for language pro- ficiency (whether they had a Norwegian wife and their mother tongue) seem likely to also have independent effects on labor market outcomes. More rigorous evidence comes from an evaluation of an immigrant integration program in Finland (Sarvima ¨ ma ¨ ki and Ha ¨ inen 2012). These authors take advan- ¨ la tage of a discontinuity induced by a law change whereby participation in the inte- gration program was only obligatory for those who entered the population register less than two years before the reform was launched. Using regression-discontinuity analysis to compare outcomes for migrants on either side of the reform deadline, they estimate that the integration plan increased employment and annual earnings threefold and halved social benefits. These authors suggest that one of the main impacts of the reform was to provide more resources for language training and to allow immigrants to retain unemployment benefits while taking a language course. The training course is relatively expensive (13,000 euros), but the benefits in terms of higher earnings for those induced to take the courses appear to be larger than this expense. However, the authors note that even with this program, the average person induced to take courses as a result of the reform was still out of work for half of his sixth year in Finland. McKenzie and Yang 177 Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 A related set of policies aims to better connect immigrants to jobs. Rinne (2013) and Butschek and Walter (2013) provide overviews of evaluations of active labor market policies intended to help immigrants, noting that few of these programs are explicitly targeted at immigrants. Joona and Nekby (2012) use a randomized exper- iment in Sweden to examine whether more intensive counselling and coaching by public employment caseworkers improves the employment of immigrants relative to those who only participate in the standard introduction programs for new immi- grants. They find that intensive coaching results in a six percent higher probability of being employed. Based on this finding, however, they calculate that the costs of the program exceed the benefits. Policies Directed towards Returning Migrants Given the large expected income gains most international migrants face when com- paring working abroad to working in their home country, staying abroad may be the privately optimal option for many migrants. Nevertheless, many migrants do return. This situation is most obvious for temporary worker programs, which require workers to return at the end of a specified period, and migrant-receiving countries use a number of policies to attempt to ensure that these workers return. However, return rates can also be quite high for migrants who do not participate in temporary programs. For example, Dumont and Spielvogel (2008) reported exit rates within the first five years of residence ranging between 19 percent in the United States and 60 percent in Ireland, whereas Gibson and McKenzie (2012) showed high return rates for the highest academic achievers from five countries. This return can be motivated by preferences to be closer to family or to other non- income aspects of their home country, by relative changes in the circumstances of the home and destination countries that change the relative attractiveness of the two, by particular migrants not doing as well in finding work as they expected ex ante, and by target savers having achieved some savings goal abroad that enables them to carry out a desired investment in their home country. Many migrants are thus choosing whether to return or stay abroad and presum- ably making what they feel is the best choice for them. There are three broad types of policies that interact with these decisions. The first are policies that seek to remove regulatory, bureaucratic, and informational barriers that prevent individu- als who might want to return from doing so or from being as productive as they could be upon return. The second are policies that seek to encourage people to return who would otherwise choose not to by changing the financial and other in- centives for their return decision. The third are policies that focus on making return migrants more productive and reintegration easier for migrants when they do return. Table 3 summarizes these types of policies, and we discuss each in turn. However, note that, in practice, these categories are not mutually exclusive; a 178 The World Bank Research Observer, vol. 30, no. 2 (August 2015) Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 McKenzie and Yang Table 3. Examples of Types of Return Migration Policies Policy Examples What does it do? Potential impact on development Evidence Policies to remove regulatory, bureaucratic, and informational barriers Dual citizenship 84 þ countries allow Allows migrants to maintain rights Can enable migrants to earn more while Promising. Cross-sectional of a citizen in their home country abroad, and make it easier for them to associations and difference- in- and in destination country return to their home country differences (Le Blang 2011; Mazzolari 2009). Pension portability Over 1,000 bilateral Allows migrants to retain access to Can increase incomes of return Non-existent. agreements retirement benefits accrued even migrants, and removes a disincentive to if they return return Job fairs and Jamaica returning Help migrants find out about job Can remove informational barriers to Non-existent. migrant resident program opportunities in home country migration, allow migrants to find better databases Bulgaria’s job fairs jobs at home Go-and-see visits UNHCR programs IOM Allow refugees to see whether Can remove information barriers Non-existent. programs home communities are now preventing return migration, allow suitable for return refugees to return to home communities Policies changing financial and other incentives for return Assisted Voluntary IRRANA, available to Pay for travel and logistics for Not clear, since alternative is usually Non-existent, although take-up Return Afghans in Norway rejected asylum seekers to return deportation rates suggest relatively little demand from many migrants Pay-to-go schemes Spain’s Voluntary Pays migrants to return to their Lowers cost of return for those intending Non-existent, although take-up Return Program, Aide home countries to return anyway, potentially provides rates suggest relatively little au Retour in France capital that can be used to set up demand from many migrants. livelihoods Continued 179 Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 180 Table 3. Continued Policy Examples What does it do? Potential impact on development Evidence Temporary return TOKTEN TRQN Connects highly-skilled emigrants Potentially helps transfer knowledge, Extremely limited. Qualitative of qualified to home country institutions for ideas, and technology from abroad, evidence reveals some mixed migrants 3 months of consulting/ with spillover benefits to others in effects, scale of programs knowledge-sharing home country typically too low for large effects. Permanent return Return of Qualified Provides funding, and other Potentially helps transfer knowledge, Extremely limited. Qualitative of qualified African Nationals incentives to bring highly-skilled ideas, and technology from abroad, evidence reveals some mixed migrants migrants back to countries of with spillover benefits to others in effects, scale of programs The World Bank Research Observer, vol. 30, no. 2 (August 2015) origin home country typically too low for large effects. Policies to make return migrants more productive and adjust reintegration easier Reintegration The Philippines’ Provides start-up funding and Potentially helps overcome credit Non-existent. support programs reintegration loans for setting up businesses; constraints and skill constraints to program business or vocational training; allow migrants to earn more. May help psycho-social assistance migrants re-adjust to home society. Recognition of Argentina s RAICES Translate and accredit Make it easier for return migrants to be Non-existent skills earned program qualifications earned abroad rewarded for skills learned abroad, abroad increasing their incomes Source: Authors reading of the literature. Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 number of policies have components of each of these types. Furthermore, we note that the evidence base is extremely weak for almost all of these policies. Policies to Remove the Regulatory, Bureaucratic, and Informational Barriers that Inhibit Return Migration There are a number of regulatory and bureaucratic barriers that can make it more difficult and costly for migrants to return to their home countries. One important example is citizenship or residency rights. Migrants from countries without dual cit- izenship may have had to give up their home country citizenship to gain citizenship abroad. If they wish to return later to their home countries, they may face difficul- ties in being allowed to move back permanently as well as the disincentive of having to close off the option of migrating again by potentially giving up their newly ac- quired foreign citizenship. Individuals with only legal residency abroad may be even more reluctant to move back than those with foreign citizenship because returning home may result in violating continuous residence requirements required for them to retain permanent residence status. For example, permanent residents in the United States can lose their permanent residence status if they remain outside the United States for more than one year. Le Blang (2011) use cross-sectional data on migrants in Spain and Germany and finds an association whereby migrants from countries that offer dual citizenship send more remittances and express higher intents to return. Dual citizenship also can confer benefits on migrants if they do not return. Mazzolari (2009) use a differ- ence-in-differences approach to compare immigrants in the United States from five Latin American countries that granted dual citizenship rights in the 1990s to mi- grants from other Latin American countries and finds the migrants who were granted dual citizenship rights increased employment and wage earnings and reduced their use of welfare benefits relative to the comparison group. The number of countries that allow for dual citizenship has grown dramatically, from only 26 in the mid-1970s to 84 countries in 2006 (Le Blang 2011). Dual citizenship makes it easier for migrants to return. However, many migrants marry citizens of other countries, so another constraint they face is a limit to the ability of their non-citizen spouse and children to live, work, and attend schools in the migrant’s country of origin. Malaysia gives permanent residence status to foreign spouses and children as part of its efforts to facilitate return migration (Lowell 2001). It is likely that such efforts help, and they are certainly unlikely to reduce development impact, but we are unaware of any evaluation evidence. An additional bureaucratic constraint to return migration for many migrants concerns the portability of their social benefits, particularly retirement benefits. There are two issues here. The first is that migrants who work in multiple countries may not accrue enough years of work to become fully vested in the pension systems McKenzie and Yang 181 Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 of either their home or their destination countries. The second is whether migrants retain eligibility to receive pension payments if they return to their home countries. OECD countries have many bilateral agreements, but coverage is very limited for mi- grants who move from one developing country to another (Avato et al. 2010). Pension portability can directly improve the well-being of return migrants, but we are unaware of literature that quantifies this finding or that examines the impact of pension portability on the rate of return migration. In addition to regulatory and bureaucratic barriers, informational barriers may prevent some migrants from returning. Information may be lacking about specific job opportunities at home or about changes in conditions in the home country. A number of governments attempt to reduce job search frictions by making it easier for domestic firms to locate emigrant workers who may be interested in returning and vice versa. For example, Jamaica’s returning residents program has a databank of migrants abroad that prospective employers can use (Thomas-Hope 2004), Bulgaria runs an annual job fair to attempt to initiate direct contact between Bulgarian emigrants and leading companies in Bulgaria (TFMI 2012), and Moldova has held job fairs in Italy to provide information about job opportunities in Moldova. It is unclear how successful any of these efforts have been in terms of in- creasing return migration or improving the jobs that return migrants obtain. A more specialized information barrier faces refugees, who may not have migrat- ed for work reasons and who may wish to return to their home communities pro- viding that peace and reconciliation efforts or disaster recovery make their home community safe for return. They may lack information about the conditions on the ground, making them cautious about returning. The UNHCR and IOM programs for the reintegration of refugees organize “go-and-see” visits that are intended to overcome this constraint.3 However, there does not appear to be evidence as to whether this leads to more return and better outcomes for these refugees. Policies Intended to Change Financial and Other Incentives to Return Even in the absence of regulatory, bureaucratic, and informational barriers, many migrants may not choose to return to their home countries. There are two main cases in which policy efforts attempt to change these private decisions. The first case concerns asylum seekers and some refugees and unemployed migrants, with two potential policy motivations – a desire of migrant-receiving countries to avoid the costs and expenses of either forcefully deporting individuals or of having individ- uals depend on welfare benefits and the idea that these are vulnerable groups that may be financially constrained from returning and setting up viable livelihoods. The second, quite distinct case concerns high-skilled emigrants. The idea is that their return will have positive externalities for their home countries that they fail to take into account in their private decisions. We discuss these two cases. 182 The World Bank Research Observer, vol. 30, no. 2 (August 2015) Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 Assisted voluntary return (A VR) programs offer rejected asylum seekers assistance in returning to their country of origin. The typical program involves paid transporta- tion and logistics facilitation to enable travel to the home country, with some pro- grams also offering grants to set up small businesses or access to training programs once they arrive. Destination country governments find these programs to be consid- erably less costly (between one-tenth (Black, Collyer, and Somerville 2011) and one-quarter (Strand et al. 2008) of the cost) than the forced return of rejected asylum applicants (who may otherwise go through appeal processes and involve costly depor- tation procedures). The counterfactual here is not clear because in the absence of the program, migrants may end up being forcefully deported, but they could also poten- tially gain the right to remain in the destination country through an appeal process. The limited available evidence suggests that these programs are often not very popular. Strand et al. (2008) discuss the case of Afghan nationals in Norway, where close to 2000 people had received a final rejection of their asylum application; however, only 69 Afghan adults chose to return through this program over a two- year period, with more than 206 forcibly returned over the same time frame. The situation is a bit different for refugees and unemployed migrants, who have the option of remaining in the destination country. A number of destination countries have voluntary return programs or “pay-to-go” schemes intended to entice refugees to return to their countries of origin after conflict has ended and to entice unemployed migrants to return to their home countries rather than receive welfare assistance. These programs have a long history; France’s Aide au Retour program launched in 1977 to target unemployed migrants in France, offering 10,000 French Francs for mi- grants to return permanently to their country of origin (Plewa 2012). The programs typically provide airfare, some reintegration assistance, and a lump sum resettlement amount, which can be reasonably sizeable (e.g., Denmark offered Iraqi immigrants up to $9,000 per adult and $10,500 per child; Dumont and Spielvogel 2008). Black, Collyer, and Somerville (2011) have identified 128 such programs, noting a resur- gence in interest during the recent global economic crisis. As an example, they note Spain’s Voluntary Return Plan, which was launched in November 2008 and targeted unemployed foreign nationals. However, by April 2010, only 11,400 immigrants had agreed to leave Spain through this program, which was a tiny fraction of the unem- ployed immigrant population and mostly consisted of Latin Americans, although the target of the policy was Moroccans. The low take-up rates for these voluntary return programs suggest that most migrants do not think that participating in these pro- grams will improve their well-being. It remains unclear to what extent such programs largely subsidize the returns of those who would have returned anyway. Concerns about “brain drain” and a desire to benefit from the skills and knowl- edge that migrants have gained abroad motivate a range of policies designed to foster the temporary or permanent return of highly skilled migrants. Temporary return programs typically aim to link highly skilled emigrants with opportunities to McKenzie and Yang 183 Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 help their home countries, relying on a combination of altruism and coverage of the costs of participating. A prominent example is the UNDP’s TOKTEN (Transfer of Knowledge Through Expatriate Nationals) program, launched in 1977.4 The program relies on volunteers who return to their country of origin for a period of between two weeks and three months to share their expertise. They are paid travel costs and a small allowance, but not professional fees. Approximately 5,000 people have participated in this program in nearly 50 developing countries over a 20-year period (Dumont and Spielvogel 2008). The number of consultants involved in any particular country can be small; Sri Lanka received 43 consultants over the eight- year period of 1996 to 2004 who went to universities, government ministries, and NGOs. A qualitative assessment of the performance in Sri Lanka by Wanigaratne (2006) revealed incidences in which universities appear to have benefited from this exchange as well as cases in which consultants made technical recommendations that were not suitable for local conditions. The assessment noted that the sporadic nature of the engagement made sustained impacts doubtful. Such programs are also offered by developing countries themselves rather than through international agencies. Thailand’s Reverse Brain Drain project aims to facilitate technical linkages between Thai institutions and migrants abroad, whereas Argentina’s RAICES program supports short-term returns of scientists to Argentina (Lowell 2001; Jonkers 2008). A related example comes from the Temporary Return of Qualified Nationals to Afghanistan (TRQN) project studied by Kuschminder (2013), which brought 59 highly-skilled Dutch-Afghans to work with a variety of public and private institutions in Afghanistan for three months. The participants are officially volunteers who receive living, travel, accommodation, and travel allow- ances within Afghanistan that total up to $2,000 per month. Kuschminder gives as examples participants who helped teach new computer skills or helped design new curricula for university courses and a participant who was a trained engineer who taught a specialty course on electrical engineering to workers from private firms. The participants and those they worked with complained that the time period was too short in many cases, and it is unclear how much value these migrants were able to add beyond what would have occurred in their absence. This second point is perhaps less of an issue in post-conflict societies with severe skill shortages, but there is little evidence to date to measure actual impacts. Other programs seek to encourage highly skilled migrants to permanently return to their countries of origin or, at least, to return for a period of several years. Countries provide a broader range of incentives in an effort to do this. These incen- tives have included features such as tax exceptions, interest-free or low interest loans, temporary salary supplements to facilitate career entry, and assistance with housing, schooling for children, and employment for spouses (Lowell 2001; Jonkers 2008; TFMI 2012). An example is Malaysia’s Returning Expert Program,5 which provides a flat tax of 15 percent on employment income for five years, the ability to import 184 The World Bank Research Observer, vol. 30, no. 2 (August 2015) Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 two cars tax-free, and permanent residence status to a foreign spouse and children within six months. Perhaps the best known of these programs is the Return of the Qualified African Nationals (RQAN) program managed by the IOM, which has since evolved to become the Migration for Development in Africa (MIDA) program. The RQAN pro- vided airline tickets for return migrants and their dependents, shipment of their possessions, purchase of professional equipment needed for their work, settling-in expenses, and assistance finding positions in their home countries (Lowell 2001). It placed approximately 2,000 highly skilled persons in 41 African countries between 1974 and 1990 (World Bank 2006), so the effective placement in any given African country in a particular year was low on average. Existing “evaluations” of this program appear to consist largely of surveys that directly ask those who partici- pate whether they feel they are contributing to development in their home country and whether they feel the program was important in their decision to return (e.g., Pires, El Nour, and McMahon 1996). The main purpose of such programs is to generate externalities for others in the developing countries of origin. However, the numbers of people involved in these programs have often been relatively small, which, coupled with qualitative evidence on the types of activities in which these participants have been involved, suggests that the aggregate externalities are likely rather limited. In many cases, the pro- grams may end up subsidizing the return of individuals who were likely to return anyway ( potentially speeding up this return), and as Dumont and Spielvogel (2008) note, these policies can have adverse effects in terms of feeding resentment among non-migrants or even potentially encouraging more people to emigrate to obtain these benefits upon return. Policies Intended to Make Return Migrants More Productive and Reintegration Easier A final set of return migration policies is intended to make it easier for returning mi- grants to be productive, earn more, and re-adjust to life in their home countries. One of the most comprehensive of these types of programs is the Overseas Foreign Worker (OFW) reintegration program provided by the Philippines. This program attempts to address both the economic and social needs of returning workers. Training programs are offered for those who would like to start small businesses, and a psycho-social component includes services such as family counselling, stress debriefing, and community organizing programs intended to help the migrant fit back into life in the Philippines (Tornea 2003). As a second example, Poland offers returning migrants dedicated websites and a “return migrant handbook” with in- formation on programs to help them find work and deal with the logistics of resettle- ment (TFMI 2012). McKenzie and Yang 185 Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 Reintegration programs that involve training and/or credit to start small enter- prises are also widely used in programs that help refugees resettle. Strand et al. (2008) provide qualitative evidence on the success of this approach for Afghans re- turning from Norway. They note that in a few cases, moderately successful business- es were established; however, the majority appeared to exist only on paper, with participants using the business as a mechanism to convert start-up business grants into cash and most respondents stating an intention to re-migrate. Although such programs can sound intuitively appealing and potentially over- come financial or skill constraints that limit the ability of migrants to work produc- tively, there is no existing evidence of their success. Moreover, there are at least three concerns with such programs. The first is that not everyone wants to be an en- trepreneur; many return migrants have been working in wage jobs with no experi- ence in running a business. Second, existing evaluations of training programs have had, at best, mixed results, even among those individuals who are interested in starting businesses (McKenzie and Woodruff 2014); there is no reason to expect return migrants to be particularly good at running businesses. Third, it is unclear why such programs should be targeted explicitly at return migrants rather than being part of a portfolio of training and work assistance options offered to all indi- viduals in a given region. Finally, policymakers can help make return migrants more productive by facili- tating recognition of the qualifications and skills gained abroad. For example, Argentina’s RAICES program offers the translation and accreditation of qualifica- tions formally earned abroad (Jonkers 2008). The Bologna Process aims to formal- ize recognition of higher education qualifications within Greater Europe, but many migrants from developing countries still experience difficulties obtaining overseas qualifications that are recognized within their home countries. A second issue is to how to obtain recognition for on-the-job skills learned abroad and for qualified mi- grants to be able to return to employment systems that principally reward years of service. Gibson and McKenzie (2012) find this to be an issue for highly skilled mi- grants returning to public sector jobs in developing countries. However, there does not appear to be research that demonstrates the impact of policy efforts to recognize skills earned abroad. Why is the Evidence Base so Limited? The above discussion has shown wide variation in the extent of evidence for differ- ent policies. The vast majority of evidence is very recent and comes from random- ized controlled trials that explicitly allocate migrants to one policy regime or another. The areas in which these trials have most often occurred are certain finan- cial services interventions and financial education. These interventions are 186 The World Bank Research Observer, vol. 30, no. 2 (August 2015) Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 relatively cheap to implement, do not require a change to the migration decision itself, and can involve large numbers of individuals. In contrast, the evidence base is more limited for other types of migration poli- cies. Table 3 shows that almost all of the policies directed towards returning mi- grants have no rigorous evidence of their effects. The following example from IOM (2005) is representative of the state of most evidence on return migration policies: “This evaluation report summarizes the findings, conclusions and recommenda- tions of a multistakeholder team using participatory approaches to an evaluation of IOM’s reintegration projects . . . It reflects the application of a ‘learning-focused’ process, aimed more at identifying lessons learnt than at assessing actual perfor- mance or impact levels”. That is, the standard approach, if an evaluation is con- ducted at all, is to attempt to examine the process of the policy with no reference to a counter-factual of what would have happened without the policy. We can speculate about several reasons why very little in the way of rigorous evaluations has been done for these types of policies. A first reason is that the size of many of these programs is very small; a program that takes 20 or 30 highly skilled workers over a couple of years will find it very difficult to have sufficient statistical power to detect any impact of the program unless the impacts are massive. A second, related reason is that many of these programs have not been very popular. As a result, there is seldom a natural comparison group of people who wanted to participate in the program but who were not able to be accommodated. A third reason is that some of the intended outcomes (such as spillovers from highly skilled emigrants returning) can be difficult to measure. A final reason may concern the types of organizations that conduct these policies. The IOM and national govern- ment migration agencies are the main implementers of many of these policies, and these agencies have not typically funded or had staff trained in rigorous evalua- tions. However, we believe that none of these obstacles is insurmountable. Especially given the dire state of existing knowledge, there is ample scope for future work to provide more credible evidence on these types of policies. Conclusion Currently, strong evidence is lacking on the impacts of the wide range of policies intended to enhance the development impacts of international migration. Nevertheless, in the last few years, a growing body of literature has begun to dem- onstrate the impacts of some of these programs. The evidence to date largely comes from a few studies of specific cases, and more research is needed to examine the gen- eralizability of these results. Nevertheless, the preliminary evidence suggests some areas of policy success: bilateral migration agreements for seasonal migration with countries whose workers have few other migration options; developing new savings products for migrants that allow them some control over how this money is used; McKenzie and Yang 187 Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 and efforts to provide financial education to migrants and their families. In some other areas, there is suggestive evidence, which combined with theory, offers a rationale for policy action: efforts to lower the cost of remittances, reduce the cost of a passport, provide dual citizenship, and remove exit barriers to migration. Finally, existing research suggests reasons to be cautious about some other types of policies; enforcing strong rights for migrants, such as high minimum wages, does make some migrants better off, but at the cost of reducing the opportunity to migrate for others. Similarly, integration programs that provide language training and job search assistance for migrants can have positive effects, but at relatively high costs. Nevertheless, there are many types of programs for which the only existing evi- dence is largely case study or process evaluation at best, with no consideration of a counterfactual. This is particularly the case for a large range of return migration programs, but it is also true of popular policies such as matching funds to encour- age communal remittances. Moreover, although one can think of reasons to justify these programs (e.g., subsidies to get highly skilled migrants to return might be jus- tified in terms of the externalities they bring), there are also reasons to be apprehen- sive about the true impacts of these programs (e.g., subsidies might end up funding the return of people who would return anyway or could even deter return by signal- ing that the home country is so undesirable that it requires subsidies to induce someone to live there). Therefore, there is a strong need for research to provide better evidence on many of these migration policies to ensure that they can enhance the development impacts of international migration. Notes David McKenzie (corresponding author) is a Lead Economist with the Finance and Private Sector Development Unit of the Development Research Group in the World Bank; Dean Yang is an Associate Professor in the Department of Economics, University of Michigan. We thank three anonymous refer- ees for helpful comments and gratefully acknowledge funding for this paper from the World Bank’s Global Knowledge Partnership on Migration and Development (KNOMAD) program. All views ex- pressed in this paper are those of the authors and do not necessarily represent the views of the World Bank or the University of Michigan. 1. http://esa.un.org/unmigration/documents/The_number_of_international_migrants.pdf [accessed May 5, 2014]. 2. Amounts are expressed in constant 2011 US dollars and sourced from the World Development Indicators, as reported in Clemens and McKenzie (2014). Clemens and McKenzie (2014) note that much of this increase in remittances likely reflects changes in measurement rather than genuine growth. 3. http://www.unhcr.org/pages/49c3646cfe.html [accessed 26 March 2014]. 4. http://www.ilo.org/dyn/migpractice/migmain.showPractice?p_lang=en&p_practice_id=26 [accessed March 26, 2014]. 5. http://www.talentcorp.com.my/malaysians-abroad/returning-expert-programme// [accessed March 26, 2014]. 188 The World Bank Research Observer, vol. 30, no. 2 (August 2015) Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 References Ambler, K. 2013. “Don’t Tell on Me: Experimental Evidence of Asymmetric Information in Transnational Households.” Working Paper, IFPRI. Ambler, K., D. Aycinena, and D. Yang. 2014. “Remittance Responses to Temporary Discounts: A Field Experiment among Central American Migrants.” Working Paper 20522. National Bureau of Economic Research, Cambridge, MA. . Forthcoming. “Channeling Remittances to Education: A Field Experiment among Migrants from El Salvador.” American Economic Journal: Applied Economics. Aparicio, F. J., and C. Meseguer. 2012. “Collective Remittances and the State: The 3 Â 1 Program in Mexican Municipalities." World Development 40 (1) : 206 –22. Asis, M., and D. R. Agunias. 2012. Strengthening Pre-Departure Orientation Programs in Indonesia, Nepal, and the Philippines. Working Paper. International Organization for Migration and Migration Policy Institute. Ashraf, N., D. Aycinena, C. Martinez, and D. Yang. Forthcoming. “Savings in Transnational Households: A Field Experiment among Migrants from El Salvador.” Review of Economics and Statistics. Avato, J., J. Koettl, and R. Sabates-Wheeler. 2010. “Social Security Regimes, Global Estimates, and Good Practices: The Status of Social Protection for International Migrants.” World Development 38 (4) : 455 –66. Aycinena, D., C. Martinez, and D. Yang. 2010. “The Impact of Remittance Fees on Remittance Flows: Evidence from a Field Experiment Among Salvadoran Migrants.” Working Paper. University of Michigan, Ann Arbor, MI. Ambler, K., D. Aycinena, and D. Yang. 2014. “Remittance Responses to Temporary Discounts: A Field Experiment among Central American Migrants.” Working Paper 20522. National Bureau of Economic Research, Cambridge, MA. Batista, C., and G. Narciso. 2013. “Migrant Remittances and Information Flows: Evidence from a Field Experiment”, IZA Discussion Paper 7839. Beam, E. 2014. “Incomplete Information in Job Search: Evidence from a Field Experiment in the Philippines.” Working Paper. National University of Singapore. Beam, E., D. McKenzie, and D. Yang. 2014. “Unilateral Facilitation Does Not Raise International Labor Migration from the Philippines.” Economic Development and Cultural Change, forthcoming. Beauchemin, C., and B. Schoumaker. 2009. “Are migrant associations actors in local development? A national event-history analysis in rural Burkina Faso.” World Development 37 (12) : 1897 – 913. Berman, E., K. Lang, and E. Siniver. 2003. “Language-skill complementarity: returns to immigrant language acquisition.” Labour Economics 10 (3) : 265–90. Black, R., M. Collyer, and W . Somerville. 2011. “Pay-to-go schemes and other noncoercive return pro- grams: is scale possible?” Working Paper. Migration Policy Institute, Washington, DC. Butschek, S., and T. Walter. 2013. “What active labour market programmes work for immigrants in ¨r Europe? A meta-analysis of the evaluation literature.” ZEW Discussion Paper 13– 056. Zentrum fu ¨ ische Wirtschaftsforschung, Mannheim, Germany. Europa ´ -Somps. 2013. “Migrants’ Home Town Associations Chauvet, L., F. Gubert, M. Mercier, and S. Mesple and Local Development in Mali.” Economics Papers from University Paris Dauphine 11715. Department of Economics, Paris University Dauphine. McKenzie and Yang 189 Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 Chin, A., L. Karkoviata, and N. Wilcox. 2011. “Impact of Bank Accounts on Migrant Savings and Remittances: Evidence from a Field Experiment.” Working Paper. University of Houston. Houston, TX. Chiswick, B., and P. Miller. 1995. “The Endogeneity between Language and Earnings: International Analyses.” Journal of Labor Economics 13 (2) : 246 –88. Clemens, M., and D. McKenzie. 2014. “Why don’t remittances appear to affect growth?” Policy Research Working Paper 6856. World Bank, Policy Research Department, Washington, DC. Clemens, M., C. Montenegro, and L. Pritchett. 2009. “The Place Premium: Wage Differences for Identical Workers across the US Border.” HKS Faculty Research Working Paper Series RWP09– 004. John F. Kennedy School of Government, Harvard University, Cambridge, MA. De Arcangelis, G., D. McKenzie, E. Tiongson, and D. Yang . 2014. “Directing Remittances to Education Using Soft and Hard Commitments: Evidence from a Lab-in-the-field Experiment and New Product Take-up among Filipino Migrants in Rome.” Working Paper. Sapienza University of Rome, World Bank, and University of Michigan. Doi, Y., D. McKenzie, and B. Zia. 2014. “Who You Train Matters: Identifying Complementary Effects of Financial Education on Migrant Households.” Journal of Development Economics 109 : 39 –55. Dumont, J.-C., and G. Spielvogel. 2008. “Return migration: a new perspective.” In OECD, ed., International Migration Outlook 2008. OECD, Paris. Duquette-Rury, L. 2014. “Collective remittances and transnational coproduction: the 3 Â 1 program for migrants and household access to public goods in Mexico.” Studies in Comparative International Development 49 (1) : 112–39. Esipova, N., J. Ray, and A. Pugliese. 2011. Gallup World Poll: The Many Faces of Global Migration. Gallup and International Organization for Migration Report. Freund, C., and N. Spatafora. 2005. “Remittances: Transaction Costs, Determinants, and Informal Flows.” Policy Research Working Paper 3704. World Bank, Policy Research Department, Washington, DC. G8. 2009. Final Declaration on Responsible Leadership for a Sustainable Future. Heads of State Summit, L’Aquila, Abruzzo, Italy, July 8–10. Available at: http://www.g8italia2009.it/static/G8_Allegato/ G8_Declaration_08_07_09_final,0.pdf. ´ xico: Lecciones y Garcia Zamora, R. 2007. “El Programa Tres por Uno de remesas colectivas en Me desafı´os.” Migraciones Internacionales 4 (1) : 165– 72. Gibson, J., and D. McKenzie. 2012. “The Economic Consequences of “Brain Drain” of the Best and Brightest: Evidence from Five Countries.” Economic Journal 122 (560) : 339 –75. . 2014. “The Development Impact of a Best Practice Seasonal Worker Policy.” Review of Economics and Statistics 96 (2) : 229 –43. Gibson, J., D. McKenzie, and H. Rohorua. 2006. “How Cost Elastic are Remittances? Evidence from Tongan Migrants in New Zealand.” Pacific Economic Bulletin 21 (1) : 112– 28. Gibson, J., D. McKenzie, and B. Zia. 2014. “The impact of financial literacy training for migrants.” World Bank Economic Review 28 (1) : 130 –61. Hall, J. 2010. “Ten Years of Innovation in Remittances: Lessons Learned and Models for the Future.” Independent Review of the Multilateral Investment Fund Remittance Portfolio, Inter-American Development Bank, Washington, DC. Hayfron, J. 2001. “Language Training, Language Proficiency and Earnings of Immigrants in Norway.” Applied Economics 33 (15) : 1971– 9. International Organization for Migration (IOM). 2005. Evaluation of IOM Reintegration Activities Targeting Vulnerable Returnees and Displaced Populations in Angola. IOM, Geneva. 190 The World Bank Research Observer, vol. 30, no. 2 (August 2015) Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 . 2006. “Mass and Micro-Information Campaign Awareness Impact Assessment - Information Campaign to Combat Trafficking in Women and Children in Cambodia Project.” Joint evaluation MOW A/IOM, March. . 2011. “IOM Migrant Training Programmes Overview, 2010– 2011.” Accessed January 13, 2015. Available at: http://www.iom.int/jahia/webdav/shared/shared/mainsite/activities/ facilitating/IOM_Migrant_Training_Programmes_Overview_2010_2011.pdf. International Labor Organization (ILO). 2013. “Pre-Departure Training Programme, Republic of the Philippines.” Accessed May 1, 2014. Available at: http://www.ilo.org/dyn/migpractice/migmain. showPractice?p_lang=en&p_practice_id=72. Iskander, N. 2005. “Social learning as a productive project: The Tres por Uno (three for one) experience at Zacatecas, Mexico.” In OECD, ed., Migration, Remittances, and Development. OECD, Paris. Joona, P. A., and L. Nekby, 2012. “Intensive Coaching of New Immigrants: An Evaluation Based on Random Program Assignment.” Scandinavian Journal of Economics 114 (2) : 575 –600. Jonkers, K. 2008. “A comparative study of return migration policies targeting the highly skilled in four major sending countries.” MIREM project analytical report mirem-ar 2008/05. Available at: http:// cadmus.eui.eu/bitstream/handle/1814/9454/MIREM%20AR2008_05.pdf ?sequence=1. Karlan, D., and J. List. 2007. “Does Price Matter in Charitable Giving? Evidence from a Large-Scale Natural Field Experiment.” American Economic Review 97 (5): 1774–93. Kuschminder, K. 2013. “Knowledge Transfer and Capacity Building Through the Temporary Return of Qualified Nationals to Afghanistan.” International Migration 52 (5): 191–207. Le Blang, D. 2011. “Harnessing the Diaspora: Dual Citizenship, Migrant Remittances and Return.” Mimeo. University of Virginia, Department of Politics, Charlottesville, VA. Lowell, B. L. 2001. “Policy responses to the international mobility of skilled labour.” International Migration Papers 45. International Labour Organization, Geneva, Switzerland. Available at: http:// www.ilo.org/wcmsp5/groups/public/—ed_protect/—protrav/—migrant/documents/publication/ wcms_201774.pdf. Mazzolari, F. 2009. “Dual citizenship rights: do they make more and richer citizens?” Demography 46 (1) : 169–91. McKenzie, D. 2007. “Paper Walls are Easier to Tear Down: Passport Costs and Legal Barriers to Emigration.” World Development 35 (11) : 2026– 39. McKenzie, D., J. Gibson, and S. Stillman. 2010. “How Important is Selection? Experimental vs. Non- Experimental Measures of the Income Gains from Migration.” Journal of the European Economic Association 8 (4) : 913 –45. . 2013. “A land of milk and honey with streets paved with gold: Do emigrants have over- optimistic expectations about incomes abroad?” Journal of Development Economics 102 : 116 – 27. McKenzie, D., and H. Rapoport. 2007. “Network effects and the dynamics of migration and inequality: theory and evidence from Mexico.” Journal of Development Economics 84 (1) : 1– 24. McKenzie, D., C. Theoharides, and D. Yang. 2014. “Distortions in the International Migrant Labor Market: Evidence from Filipino Migration and Wage Responses to Destination Country Economic Shocks.” American Economic Journal: Applied Economics 6 (2) : 49 –75. McKenzie, D., and C. Woodruff. 2014. “What are we learning from business training evaluations around the developing world?” World Bank Research Observer 29 (1) : 48 –82. Naidu, S., Y. Nyarko, and S.-Y. Wang. 2014. “Worker mobility in a global labor market: Evidence from the United Arab Emirates.” Working Paper. New York University, NY. Norwegian Agency for Development Cooperation (Norad). 2010. “Evaluation of the International Organization for Migration and its Efforts to Combat Human Trafficking.” Oslo, Norway. McKenzie and Yang 191 Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017 Mahmoud, T. O., V. Licuanan, A. Steinmayr, E. Tiongson, and D. Yang. 2013. “Harnessing the Development Benefits of International Migration: A Randomized Evaluation of Enhanced Pre- Departure Orientation Seminars for Migrants from the Philippines.” Ongoing study. Available at: http://www.3ieimpact.org/en/evidence/impact-evaluations/details/720/. Pires, J., A. El Nour, and L. McMahon. 1996. “Survey among qualified African nationals under the bridging fund scheme of the programme for the return and reintegration of qualified African na- tionals (RQAN) – Phase III.” Report, International Organization for Migration, Geneva. Plewa, P. 2012. “The Effects of Voluntary Return Programmes on Migration Flows in the Context of the 1973/74 and 2008/09 Economic Crises.” Comparative Population Studies 37 (1 –2) : 147– 76 . Rinne, U. 2013. “The Evaluation of Immigration Policies.” In A. F. Constant, and K. F. Zimmermann, eds., International Handbook on the Economics of Migration. Cheltenham, UK: Edward Elgar Publishing. Ruhs, M. 2013. The Price of Rights: Regulating International Labor Migration. Princeton, NJ: Princeton University Press. Sarvima ¨ ma ¨ ki, M., and K. Ha ¨ inen. 2012. “Assimilating immigrants. The impact of an integration ¨ la program.” CReAM Discussion Paper Series 1019. Centre for Research and Analysis of Migration (CReAM), Department of Economics, University College London. Available at: http://cream. conference-services.net/resources/952/3365/pdf/MGDNF2013_0192.pdf. Seshan, G., and D. Yang. 2014. “Motivating Migrants: A Field Experiment on Financial Decision- Making in Transnational Households.” Journal of Development Economics 108 : 119 –27. Strand, A., A. Akbari, T. W. Chaudhary, K. B. Harpviken, A. Sarwari, and A. Suhrke. 2008. “Return in Dignity, Return to What? Review of the Voluntary Return Programme to Afghanistan.” Working Paper. Chr. Michelsen Institute, Bergen, Norway. Thomas-Hope, E. 2004. “Migration Situation Analysis, Policy, and Programme Needs for Jamaica.” Report for the United Nations Population Fund. Available at: http://countryoffice.unfpa.org/ caribbean/drive/2004_migration_study_jamaica.pdf. Tornea, V. 2003. “Reintegration program for migrant workers.” Review of Women’s Studies 13 (2) : 203 –20. Torres, F., and Y. Kuznetsov. 2006. “Mexico: Leveraging Migrants’ Capital to Develop Hometown Communities.” In Y. Kuznetsov, ed., Diaspora Networks and the International Migration of Skills. Washington, DC: World Bank. Transatlantic Forum on Migration and Integration (TFMI). 2012. “Welcome home? Challenges and Chances of Return Migration.” Washington, DC. Available at: http://www.gmfus.org/wp-content/ blogs.dir/1/files_mf/1358538301tfmireturnmigration.pdf Wanigaratne, D. G. 2006. “An Evaluation of the UNDP Transfer of Technology through Expatriate National (TOKTEN) Program.” Report. Available at: http://erc.undp.org/evaluationadmin/ downloaddocument.html?docid=1863. Weiermair, K. 1976. “The Economic Effects of Language Training to Immigrants: A Case Study.” International Migration Review 10 (2) : 205 –19. World Bank. 2006. Global Economic Prospects 2006: Economic Implications of Remittances and Migration. Washington, DC: World Bank. Yang, D. 2011. “Migrant Remittances.” Journal of Economic Perspectives 25 (3) : 129– 52. 192 The World Bank Research Observer, vol. 30, no. 2 (August 2015) Downloaded from https://academic.oup.com/wbro/article-abstract/30/2/155/1655090/Evidence-on-Policies-to-Increase-the-Development by Joint Bank-Fund library user on 06 October 2017