MANAGED LABOR MIGRATION IN AFGHANISTAN: EXPERIENCE AND EVIDENCE WITH INTERNATIONAL AFGHAN LABOR MOBILITY AT MACRO LEVEL Daniel Garrote Sanchez Background Paper BGP 2a to the World Bank Project on “Afghanistan: Managed International Labor Mobility as Contribution to Economic Development and Growth” STANDARD DISCLAIMER: This volume is a product of the staff of The World Bank. The findings,interpretations, and conclusions expressed in this paper do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. COPYRIGHT STATEMENT: The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, http://www.copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org Managed Labor Migration in Afghanistan: Experience and Evidence with International Afghan Labor Mobility at Macro Level Daniel Garrote Sanchez1 Background Paper BGP 2a to the World Bank Project on “Afghanistan: Managed International Labor Mobility as Contribution to Economic Development and Growth” Washington, D.C and Beirut January 2018 Abstract Migration in Afghanistan has been a relevant phenomenon during the last several decades, driven by a complex combination of protracted conflict, food insecurity, natural disasters, and socioeconomic factors. Around 4.8 million Afghan currently live abroad, most of them in neighboring Iran and Pakistan. While prior migration waves consisted of refugees to a large extent, in the last decade economic migrants have been increasingly prevalent, not only to Iran but also to Gulf Cooperation Council and Organisation for Economic Co-operation and Development countries. Due to the lack of formal mechanisms for migration, however, the vast majority of flows have an irregular nature. As a consequence, official statistics vastly underestimate the value of remittances at 1.7 percent of GDP , while analysis that includes informal channels raise this figure by up to 10 times. Overall, although a relatively small share of families benefits from remittances, they provide a vital source of income and act as a buffer against income shocks. 1 The author is now a senior researcher at the Lebanese Center for Policy Studies. He was a labor economist and migration specialist at the World Bank’s International Labor Mobility Team, Social Protection and Jobs Global Practice, Washington, DC, when the initial draft of the paper was written. Table of Contents Abbreviations and Acronyms.................................................................................................... iii 1 Introduction, Structure, and Main Messages.......................................................................... 7 2. Historical Patterns of Migration and Current Situation............................................................. 9 2.1 Trends in the Stock of Afghan Refugees Overseas........................................................ 9 2.2 The Role of Economic Migration in Afghanistan............................................................ 14 3. Migration and Remittances in Afghanistan............................................................................. 19 3.1 Official Estimates of Remittances............................................................................... 19 3.2 Approximating the True Scope of Remittances in Afghanistan............................................. 22 References............................................................................................................................. 26 List of Figures Figure 1: Trends in Afghan migrants abroad by country of destination, 1990–2015............................... 9 Figure 2: Trends in Afghan migrants abroad by status and gender, 1990–2015.................................... 10 Figure 3: Trends in Afghan refugees by country of asylum, 1990–2015.............................................. 11 Figure 4: Flows of refugee returnees to Afghanistan by country of origin, 1992–2016.......................... 12 Figure 5: Net flows, outflows, and inflows (returnees) of refugees, 2002–2015.................................... 13 Figure 6: Trends in the number of economic migrants from Afghanistan, 1990–2015............................ 15 Figure 7: Share of labor migrants in the total migrant diaspora of Afghanistan, 1990–2015................... 15 Figure 8: Annual flows of deportations of Afghans from Iran, 2002–2010........................................... 17 Figure 9: Flows of remittances into and from Afghanistan, 2008–2015.............................................. 20 Figure 10: Share of households for which remittances are one of the three main sources of income..... 23 List of Tables Table 1: Correlation between inflows of remittances in Afghanistan and key variables, 2008–2015.......... 21 Acronyms and Abbreviations ALCS Afghanistan Living Conditions Survey GCC Gulf Cooperation Council GDP Gross domestic product OECD Organisation for Economic Co-operation and Development UN DESA United Nations Department of Economic and Social Affairs UNHCR United Nations High Commissioner for Refugees iii 6 1 Introduction, Structure, and Main Messages Afghanistan’s economic growth options are limited. Progress in economic reform and macroeconomic stability over the recent decade gave rise to the high growth rate in the country’s gross domestic product (GDP), but this was based to a large extent on high inflows of security- and aid-related resources. The withdrawal of security forces by the US-led coalition and the reduction in aid resources have dimmed Afghanistan’s economic outlook, and its options for autonomous economic growth in a fragile international environment are constrained (IMF 2016). One option under exploration is to use managed economic out-migration of the young Afghan labor force to generate possible growth effects through remittances, skills acquisitions, and the “brain gain” and network effects of return migration. This background paper studies the historical patterns and current situation of international migration flows in Afghanistan from a macro perspective, decomposing its complex nature and drivers. It also aims to shed light on the size, trends, and role of remittances in Afghanistan. Section 2 describes the trends in migration during the last several decades, approximating the different roles of forced and economic migration as well as the size of legal and illegal migration. In turn, section 3 analyzes remittance flows and their drivers during the last decade. Given the lack of reliable figures, it includes different approaches and sources, such as the official statistics of the Central Bank of Afghanistan and surveys of sending and receiving households. The main outcomes are that:  igration in Afghanistan has been a relevant phenomenon during the • M last decades, driven by a complex combination of protracted conflict, food insecurity, natural disasters, and socioeconomic factors.  ince the 1990s the nature of migration has slowly changed, with a • S smaller but still relevant role for forced displacement.  round 4.8 million Afghans resided abroad in 2015, comprising 2.6 • A million refugees and 2.2 million economic migrants, the majority of them in neighboring Iran and Pakistan.  hile prior waves of migration consisted largely of refugees, in the last • W decade economic migrants have seen increasingly important migration flows, not only to Iran but also to Gulf Cooperation Council (GCC) and OECD (Organisation for Economic Co-operation and Development) countries.  he high stock of migrants reflects even larger flows of out- and in- • T migration (returnees). Since the beginning of the 2000s, around 6.5 million migrants returned to Afghanistan, posing additional challenges of reintegration in the country.  etween 2006 and 2015, the average yearly net outflow of economic • B emigrants was around 60,000 people. The accumulated outflow (600,000) represented around 7 percent of the labor force in 2015.  ue to the lack of effective government-managed mechanisms • D for sending labor migrants abroad, most labor migrants travel to neighboring countries irregularly. 7 1.Introduction, Structure, and Main Messages  fficial Afghan statistics report remittance flows of US$342 million in • O 2015, equivalent to 1.7 percent of GDP , but due to the extensive use of informal channels, this figure underestimates the real value.  hile surveys on sending countries suggest remittance levels up to 10 • W times higher, reaching well over 16 percent of GDP , surveys of recipient households yield results similar to official figures. Disparities appear due to the different levels of estimated migration rather than to per migrant remittances. Estimates of remittance levels based on assumptions about the propensity to remit given the length of stay abroad are 3–7 percent of GDP , with a central scenario of around 6 percent of GDP .  s known from other countries, remittance flows depend on the stock • A of migrants, the nature of migration (temporary or permanent), family situation, and economic conditions in both sending and receiving countries. In particular, remittances have a countercyclical nature that serves as a buffer against income shocks.  nly about 6 percent of Afghan households rely on remittances as one • O of their three main sources of income. Remittances are highly relevant for those beneficiaries, though, averaging US$1680 annually (more than half of their income), and are usually consumed in basic needs. 8 2 Historical Patterns of Migration and Current Situation Afghanistan’s long history of migration dates back centuries. In recent history, emigration dramatically increased during the Soviet occupation of 1979–1989, with mass movements of refugees. By its peak in 1990, around 6.7 million Afghans had left the country and were living abroad (Figure 1), representing more than half of the total population living in Afghanistan at that time (around 11.7 million). After a significant drop at the beginning of the 1990s led by inflows of refugee returnees back into the country, the stock of Afghan migrants abroad has again seen an intermittent but upward trend. According to UN DESA (Department of Economic and Social Affairs) statistics, the number of Afghan immigrants was estimated at around 4.8 million in 2015 (Figure 1). Neighboring Iran (2.35 million) and Pakistan (1.6 million) host more than 80 percent of the total Afghan population abroad, while the rest reside in OECD countries (460,000), GCC countries (380,000) and other countries (50,000), mainly India and Central Asia. Figure 1: Trends in Afghan migrants abroad by country of destination, 1990–2015 TOTAL AFGHANS ABROAD BY DESTINATION COUNTRY Thousands 7000 5000 3000 1000 -1000 1990 1995 2000 2005 2010 2015 OECD Saudi Arabia Iran Pakistan Rest Source: UN DESA 2015. Migration patterns in Afghanistan have been shaped by a complex combination of protracted conflict, food insecurity, natural disasters, and socioeconomic factors such as the limited absorption capacity of the domestic labor market, wage differentials (in particular with neighboring Iran), and poverty. As a result, it is hardly a single factor that explains the migration decisions of Afghans, which often blurs the distinction between refugees and economic migrants. Factors motivating migrants to leave their country can be different than those that keep them in the host country and prevent them from returning to Afghanistan. For example, in some cases migrants flee their country to escape violence and insecurity but then stay in the host country, motivated more by economic factors like differences in income-generating opportunities or better access to land. It is important to note that many Afghans increasingly use migration as a strategy to secure their livelihoods, adding a transitory and circular component –temporary and repetitive movement of migrants – to the migration complexity. In the last decades, both inflows and outflows of migrants have been very high, with frequent resettlements, circular migration, and internal migration and displacement (Koser 2009). 9 2.Historical Patterns of Migration and Current Situation In spite of all these caveats and limitations, it is useful to disentangle and approximate the primary purpose of migration. Figure 2 splits the total number of migrants between those who are considered refugees by the United Nations High Commissioner for Refugees (UNHCR) and national authorities and those who are not (economic migrants). This division shows that the majority of Afghan migrants during the last decades were refugees, although economic migrants are increasingly prevalent, representing close to half of the current migrant population. Men represented more than 52 percent of the total migrant population in 2015, down from 55 percent in 2010. The gender disparity differs widely between countries, with a general pattern of higher share of men in countries that host more economic migrants. For example, in Saudi Arabia the share of men among migrants is close to 70 percent. In contrast, there is a more balanced gender composition in countries that predominantly host refugees (as is the case in Pakistan). This illustrates a pattern in which men are overrepresented among Afghans who emigrate in search of job opportunities whereas the gender disparity is less for refugees, who flee with entire families. Figure 2: Trends in Afghan migrants abroad by status and gender, 1990–2015 TOTAL AFGHANS ABROAD BY STATUS Thousands 8000 6000 4000 2000 0 1990 1995 2000 2005 2010 2015 Refugees Economic Migrants SHARE OF MALES IN TOTAL MIGRANT POPULATION % 70 60 50 40 WORLD OECD Saudi Iran Pakistan Rest Arabia 1990 1995 2000 2005 2010 2015 Source: Own calculations based on UN DESA 2015 and UNHCR 2017. 10 2.Historical Patterns of Migration and Current Situation 2.1 Trends in the Stock of Afghan Refugees Overseas The different waves of war and political instability in Afghanistan since 1979 resulted in a majority of international migrants being considered refugees. According to UNHCR data, more than 6 million Afghans fled the country during the Soviet invasion and its withdrawal in 1989 (Figure 3). Most of them arrived in neighboring Pakistan (3.2 million) and Iran (3 million), mainly from rural areas, escaping the conflict (Monsutti 2006). At that time, host neighboring countries welcomed Afghan refugees. Iran provided them with “blue cards,” granting them refugee status as well as access to healthcare, to education, and even to the formal labor market (although only in specific low-skilled jobs where the Iranian labor force was lacking). Refugees mostly resided in urban areas, with only 5–10 percent living in camps (Abbasi-Shavezi et al. 2005). In contrast, Afghan refugees in Pakistan stayed mainly in camps near the border with Afghanistan (IOM 2014a). And while the bulk of refugees in Pakistan were of Pashtun ethnicity, the majority in Iran were Hazaras. Figure 3: Trends in Afghan refugees by country of asylum, 1990–2015 TOTAL AFGHANS ABROAD BY DESTINATION COUNTRY Thousands 7000 6000 5000 4000 3000 2000 1000 0 00 02 04 06 08 80 82 84 86 88 90 92 94 96 98 10 12 14 20 20 20 20 20 20 20 20 19 19 19 19 19 19 19 19 19 19 Iran Pakistan OECD Rest Source: UNHCR 2017. At the beginning of the 1990s, improvement in the conflict situation in Afghanistan and the active repatriation programs of Iranian and Pakistanis authorities both led to a massive flow of refugee returnees to Afghanistan. During that decade, 3.1 million returned, with 1.4 million from Pakistan in 1992 alone. In Iran, the government stopped granting “blue cards,” providing instead temporary registration cards with more limited benefits, which hampered refugees’ living conditions, access to education, and labor market opportunities. In Pakistan, UNHCR changed its policy and stopped providing food aid, leading refugees to move from camps to cities. Overall, a clear change occurred in the refugee policy stance in host neighboring countries, with an aim to reduce the financial burden and promote repatriation despite the still unstable political and security conditions in Afghanistan. A second wave of refugee outflows emerged during the Taliban regime (1994– 2001), escalating around the turn of the century, when the number of Afghan refugees increased from 2.6 to 3.8 million in 2001 (Figure 3). As opposed to the first wave, the vast majority of new refugee flows went to Pakistan, with a smaller share to Iran. 11 2.Historical Patterns of Migration and Current Situation Shortly after the U.S. military intervention and the defeat of the Taliban regime, UNHCR in collaboration with Iranian and Pakistani governments started a voluntary program for refugee return to Afghanistan. Between 2002 and 2015, around 5.8 million refugees returned to their home country (1.5 million from Pakistan in 2002 alone), although repatriation flows decreased significantly over time (Figure 4). Among the almost 6 million returnees, UNHCR directly assisted around 4.8 million (UNHCR 2015). In 2016, however, the number of returnees to Afghanistan increased rapidly again. Around 380,000 registered refugees returned from Pakistan under the UNHCR repatriation initiative (UNHCR 2017), in addition to an estimated 250,000 of undocumented Afghans (IOM 2017). The IOM also estimated 445,000 undocumented returnees from Iran. This would bring the total number of returnees to more than a million in 2016. While the reasons behind this rise are mixed, including economic factors, a main driver mentioned by returnees was the harassment and fear of arrest and deportation. This coincides with the stricter stance of the Pakistani government to accelerate the return to Afghan refugee to their country and to limit the validity of Proof of Registration cards until March 2017. Figure 4: Flows of refugee returnees to Afghanistan by country of origin, 1992–2016 RETURN MIGRATION Thousands 1800 1600 1400 1200 1000 800 600 400 200 0 00 02 03 04 05 06 08 09 07 2 4 5 8 9 01 10 12 13 14 7 11 9 9 9 9 9 9 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 19 19 19 19 19 19 Iran Pakistan Source: UNHCR 2017. During the same period of the U.S. intervention (2002–2015), the stock of Afghan refugees abroad decreased by 1.15 million, from 3.8 million to the current 2.6 million (Figure 5). This implies that more than 4.5 million refugees left the country after 2002 in spite of a general downward trend, and highlights the complex and sizable inflows and outflows of forced migration to neighboring countries. In particular, rapid inflows in 2007 were behind the 0.9 million increase in the stock of Afghan refugees worldwide after increasing security concerns in the country. Turton and Mardsen (2002) highlighted that the provision of assistance to returnees, particularly financial support to cover the cost of repatriation, incentivized many refugees to return to collect grants and then go back to the host country (what has been called the "recyclers"). This phenomenon was particularly relevant among refugees in Pakistan, who had on average shorter distances to travel as many already lived near the border. In addition, the difficulties of resettling in Afghanistan could have been behind the large “backflow” of returnees to Iran and Pakistan as well as the internal movements within Afghanistan (Oeppen and Schlenkhoff 2010). 12 2.Historical Patterns of Migration and Current Situation Figure 5: Net flows, outflows, and inflows (returnees) of refugees, 2002–2015 FLOWS OF AFGHAN REFUGEES Thousands 6000 5000 4000 3000 2000 1000 0 -1000 -2000 Net Flows Inflows Outflows Net Flows Inflows Outflows -3000 1992-2001 2002-2014 Source: Own Iran based calculations World on UNHCR 2017. Pakistan Other 2.1.1 Current Situation of Afghan Refugees In 2015, there were 2.6 million Afghan registered refugees worldwide, representing the second largest international forced-displacement group, after only the Republic of Syria. Among them, the vast majority currently reside in Pakistan (1.55 million) and Iran (close to 1 million). In comparison, around 110,000 Afghan refugees live in OECD countries (mainly Germany (30,000), Austria (15,000), and Sweden (13,000)), while 30,000 live in other countries in Central and South Asia. The Government of Iran regularly registered its refugee population over the last decade to monitor and control migration flows in the "Amayesh" census. Registration is done periodically and provides cards for refugees that give them access to basic services. In 2008, the Ministry of Labor introduced work permits for male refugees aged 16–60. Although 300,000 permits were issued, the short validity period and the high cost of acquisition (US$500) or renewal (US$300) led most Afghan refugees to choose to work in the informal sector without legal working documents. 13 2.Historical Patterns of Migration and Current Situation Most of the 1 million registered refugees arrived in Iran before 2000. Due to a change in policy, those asylum seekers who arrived after the U.S. intervention are assessed on an individual basis, which makes it more difficult to obtain the permits to be registered. This policy, combined with the increasing priority of repatriating a larger flow of Afghan refugees, aims at reducing their number in the country. As a result, most Afghan refugees registered in Iran have lived in the country for an extended period of time (in many cases 20 years or more) and almost half of them were born in exile (Wickramasekara et al. 2006). The fact that most migrants have lived in Iran for a long time or were even born there and thus have acculturated to the Iranian lifestyle makes it more difficult for them to return to Afghanistan (Koepke 2011). Most do not actually intend to return in the future (Wickramasekara et al. 2006). Iran’s more modern, urban lifestyle and better economic opportunities are among the leading factors for this decision. Conditions in Afghanistan are still not very favorable to attract the diaspora, with an unstable security environment in several areas of the country, more limited employment opportunities in non-agriculture sectors, and defective social services and infrastructure (Koepke 2011). In addition, long-term refugees, most of whom fled the country with their entire families, lost many of their social networks and land rights2, and can face discrimination upon arrival (Koepke 2011). These barriers not only deter refugees from returning to Afghanistan but also hinder their reintegration when they decide to do so. In Pakistan, the government and UNHCR conducted a census of Afghans in 2005, providing 1.7 million Proof of Registration cards. However, even registered Afghan refugees do not have the right to work, which forces them to work mostly in the informal sector in low-skilled and low-paid jobs (ICMPD 2013 and US Committee for Refugees and Immigrants 2009). In spite of these conditions, several barriers deter Afghan migrants from returning to their home country. In particular, security issues and the lack of employment and economic opportunities continue to pose a threat to their return and reintegration (Monsutti 2006). ICMPD (2013) also found that a main concern is the lack of land availability. 2.2 The Role of Economic Migration in Afghanistan In contrast with the irregular but downward trend in the number of refugees, the importance of labor migration in Afghanistan significantly increased over the last decades. While in 1990 there were around 380,000 economic migrants, this number rose to close to 1 million in 2000 and to 2.2 million in 2015. This trend represents a net annual flow of 85,000 Afghans over the last 15 years. By country of destination, 1.4 million labor migrants currently reside in Iran, near two-thirds of the total workers who migrated from Afghanistan in search of better economic opportunities. The other most important destinations for economic migrants are GCC countries (with 360,000 workers, mostly in Saudi Arabia)3 and OECD countries (350,000). Among the latter, the countries with the most Afghan economic migrants are Germany (70,000), the United States (62,000), the United Kingdom (59,000), Canada (30,000), and the Netherlands (28,000). In some countries like the United States, these figures only count first-generation Afghan migrants, as those born in the country are considered nationals. This contrasts with countries like those in the GCC, Iran, or Pakistan, where Afghan migrants and their children rarely obtain the host country nationality, and thus the second and third generations are still registered as Afghans. 2 The government has been very slow in allocating land among the returnees who lost their land. 3 Although there is a general lack of information regarding the scope and nature of labor migration of Afghans in GCC countries, it is thought that most of the flows are illegal (only Qatar has signed a bilateral labor agreement with Afghanistan), with some Afghans obtaining illegal Pakistani passports to enter the region (IOM 2014a). 14 2.Historical Patterns of Migration and Current Situation Figure 6: Trends in the number of economic migrants from Afghanistan, 1990–2015 ECONOMIC MIGRANTS BY Thousands DESTINATION COUNTRY 2500 2000 1500 1000 500 0 1990 1995 2000 2005 2010 2015 OECD Saudi Arabia Iran Pakistan Rest Source: Own calculations based on UN DESA 2015 and UNHCR 2017. Overall, the nature of migration in Afghanistan changed drastically in the last decades. Compared to 1990, when migration was mostly related to refugees escaping war and conflict (representing 95 percent of total migration), labor migration is prevalent in the current flows of migrants abroad. As a result, the share of labor migrants in the total stock of Afghan migrants rapidly increased, reaching 45 percent in 2015 (Figure 7). Figure 7: Share of labor migrants in the total migrant diaspora of Afghanistan, 1990–2015 Source: Own calculations based on UN DESA 2015 and UNHCR 2017. SHARE OF ECONOMIC MIGRANTS IN % THE TOTAL MIGRANT POPULATION 100 80 60 40 20 0 WORLD OECD Saudi Iran Pakistan Rest Arabia 1990 1995 2000 2005 2010 2015 15 2.Historical Patterns of Migration and Current Situation The increasing importance of labor migration was driven by the evolution of migration to Iran, where labor migrants already represent 60 percent of the total migrants, compared to the situation in the early 1990s when more than 90 percent were refugees. In developed countries, labor migrants comprise an even higher share, at 75 percent in OECD countries and 100 percent in Saudi Arabia. It is important to note that the quality of the data faces many limitations. For example, no data are available in Pakistan for non-refugees, as they are not counted in the census. As previously mentioned, the Iranian government tightened the procedures to get the Amayesh refugee cards that are now granted on individual basis. As a result, part of the reduction in the share of refugees might be due to the increasing unrecorded status of migrants who left the country for security reasons, although the relevance of recent forms of labor migration has been highlighted in different studies. To date, Afghanistan has no effective mechanism for sending labor migrants abroad. Most labor migrants go to neighboring countries irregularly. A main problem in recording the number of international flows of Afghan labor migrants is the porous nature of the country’s borders, particularly with Pakistan. Many tribes have gone back and forth between the two countries for centuries, such as the nomadic Kuchis. In a study commissioned by UNHCR, Majidi (2008) found that a large number of daily crossings in the Torkham border went vastly unreported. During a week in May 2008, the average crossings were between 10,000 and 20,000 entries and exits per day, in contrast with only a few hundred officially registered. The overall message is that people usually cross the border without documents and registration, mainly for a short duration and a variety of reasons, such as work, visits to relatives, or health checks. In addition to the official number of registered refugees, an estimated 400,000 to 1 million Afghans reside illegally in Pakistan4. Undocumented migrants do not want to report their presence to agencies like UNHCR as they fear being recorded and deported by Pakistan authorities (International Crisis Group 2009). Iran lacks a legal framework under which migrants can work temporarily. As a result, while Afghans with refugee status reside in the country legally, labor migrants have to enter the country illegally. Government authorities estimate that between 1.5–2 million undocumented Afghan labor migrants live in the country (IOM 2014a), a figure similar to that indirectly obtained from UN DESA and UNHCR data (1.4 million). Therefore, labor migrants account for up to two-thirds of all Afghan migrants in Iran. The majority of those undocumented migrants can be categorized as young adult men who come to Iran for a short stay to work as unskilled manual workers and save money to support their families back in Afghanistan. They spend an average of 3.5 years in Iran and most report that they do not intend to stay in the country for more than five years (Majidi 2008). Undocumented migrants rely on smugglers to cross the country, at a cost half that of the lengthy visa process to cross legally (Majidi 2008). To migrate, they also rely on the social networks of relatives and friends in Iran, which help finance the cost of migration and oftentimes determine where new migrants decide to live. 4. Thousands of Afghan refugees in limbo in Pakistan,” AP August 24, 2012, or “Afghanistan taking steps for repatriation of 1.6m refugees,“ Reuters January 25, 2015. 16 2.Historical Patterns of Migration and Current Situation The Iranian government has been increasingly strict with the flows of irregular migration and, as such, has deported a large number of undocumented Afghans. According to UNCHR data, Iranian authorities deported 1.9 million Afghans between 2002 and 2011. The number of deportees rose steadily to a maximum of 400,000 in 2008 (Figure 8). This forced return entails important costs for migrants, who usually leave assets back in Iran that they are unable to claim. However, coercive mechanisms of forced out-migration such as deportations have not succeeded in curving the flows of irregular migration. As Majidi (2008) showed, Afghan deportees often try to smuggle back into Iran, driven by the higher economic opportunities. Since 2010, the Iranian government has tried to regularize undocumented migrants through a process that requires Afghan men without families to return to Afghanistan to apply for visas. However, the high cost of this process, which often entails obtaining an Afghan passport in the first place, deters many migrants from regularizing their status. Figure 8: Annual flows of deportations of Afghans from Iran, 2002–2010 ANNUAL DEPORTATIONS OF AFGHANS IN IRAN 450,000 400,000 350,000 300,000 250,000 200,000 150,000 100,000 50,000 0 2002 2003 2004 2005 2006 2007 2008 2009 2010 Source: UNHCR. 2.2.1 Illegal Migration in Europe In 2015, around 400,000 Afghans were found to be illegally present in Europe, representing the second largest group after Syrians and 20 percent of the total of 2 million illegal migrants in Europe (Eurostat 2016). This represents an eightfold increase with respect to 2014. The vast majority were located in arrival countries like Greece (213,000) and Hungary (98,000), or destination countries such as Germany, Austria, and Sweden. In Germany, arrivals of Afghan migrants rose from 7,000 to 53,000 (half of the estimated Afghan population in the country until then). The agreement between Europe and Turkey on the refugee crisis does not consider Afghans as refugees, and European governments are trying to complete agreements with Afghanistan and third- party countries to return most of the recent influx of irregular migrants.5+6 In December 2015, Germany announced a ban on Afghan asylum seekers on the grounds that they are fleeing poverty and not war, and started to negotiate an agreement with Afghan authorities to return thousands of migrants to their home country, although the Afghan government was reluctant to sign it. In October 2016, the EU signed an agreement with the Afghan government allowing its member states to deport an unlimited number of Afghans that got their asylum applications rejected.7 5. https://www.euractiv.com/section/global-europe/news/commission-defends-plan-to-deport- afghans-to-safe-regions/ 6. http://www.bbc.com/news/world-europe-35649576 7. http://europa.eu/rapid/press-release_MEX-16-3282_en.htm 17 18 3 Migration and Remittances in Afghanistan 3.1 Official Estimates of Remittances Reliable data on remittances in Afghanistan are generally unavailable, which complicates their analysis. The Afghan financial sector includes banks and microfinance institutions, money transfer organizations, and money service providers such as electronic money institutions and “hawala” businesses, as well as foreign exchange dealers (IOM 2014b). The continuous conflict in the country has weakened and limited development of the formal financial sector (Maimbo 2003). As a result, a significant share of remittances in Afghanistan are transferred through a well-developed network of informal brokers, called hawala dealers, that are not monitored and included in official statistics. Another limitation for the analysis is that household surveys might be downwardly biased, as mistrust usually leads people to underreport their money transactions. Therefore, official data understate the real magnitude of remittances to and from the country. As a first approximation, data from the Central Bank of Afghanistan (Da Afghanistan Bank) estimate the inflows of remittances into Afghanistan at US$342 million in 2015, equivalent to 1.7 percent of GDP (Figure 9). Following the standards set by the International Monetary Fund, total remittances include the compensation of employees abroad, workers’ remittances, and migrants’ transfers. Remittance inflows in Afghanistan were volatile in the last decade although with a general upward trend. Since the global financial crisis in 2008, remittances to Afghanistan have increased threefold. With a stock of 4.8 million Afghan migrants abroad, the amount of annual remittances per migrant stands at a low US$70. In turn, remittance outflows are also sizable although declining in recent years, averaging US$150 million in 2015 (0.8 percent of GDP). 19 3. Migration and Remittances in Afghanistan Figure 9: Flows of remittances into and from Afghanistan, 2008–2015 INFLOWS OF REMITTANCES IN AFGHANISTAN % GDP Million US$ 400 3 350 2.5 300 2 250 200 1.5 150 1 100 0.5 50 0 0 2008 2009 2010 2011 2012 2013 2014 2015 Levels As % of GDP (rhs) OUTFLOWS OF REMITTANCES Million US$ IN AFGHANISTAN % GDP 350 3.5 300 3 250 2.5 200 2 150 1.5 100 1 50 0.5 0 0 2008 2009 2010 2011 2012 2013 2014 2015 Levels As % of GDP (rhs) Source: Da Afghanistan Bank 2016 and World Bank 2016. By country of origin, remittances to Afghanistan predominantly come from Iran, in line with the higher stock of Afghan migrants living in the country. In 2015, remittances from Iran accounted for US$144 million, or more than 40 percent of the total amount. The other main sending countries are Pakistan (US$107 million, or 30 percent of the total), Saudi Arabia (US$52 million), Germany (US$13 million), and the United States (US$7 million). On a per capita basis, Afghans in Saudi Arabia send an average of US$142 per year, while in Germany and the United States they send about US$130 and US$110, respectively. This contrasts with much lower per capita figures in Iran and Pakistan (US$61 and US$66, respectively). This phenomenon can be linked to both the higher salaries earned in more developed countries and the higher share of young male labor migrants (versus refugee families) who migrate to those countries with a primary goal of saving money to send home. 20 3. Migration and Remittances in Afghanistan Regarding the outflows of remittances from Afghanistan, the highest recipient country is Pakistan, totaling US$100. This high flow can be related to the presence of a sizable diaspora of poor Afghan refugees in Pakistan (1.7 million), mainly in neighboring areas, who could benefit from the transfers of money sent by relatives in Afghanistan. According to the Collective for Social Science Research (2006), close to 40 percent of Afghan migrants in neighboring areas of Pakistan receive remittances from abroad. IOM (2014b) reports that the high outward remittance flows reported in the last years also likely reflect the compensation of international staff working in Afghanistan. International experience and analysis suggest that remittance flows are shaped by a number of factors in both home and host countries, particularly the number of economic migrants (Clemens and McKenzie 2014), but also their earnings in the host country, the income generated, and the needs of their relatives and friends in the country of origin (Gupta 2005). The level of remittances also depends on personal characteristics of migrants (propensity to remit), among which are social ties and networks back in Afghanistan. In general, those that leave their country with their entire families will be less prone to remit money back home compared to individual migrants with their families in Afghanistan. Remittances are usually countercyclical, which means that they increase when economic conditions in the native country deteriorate. Table 1 shows correlations between the inflows of remittances to Afghanistan and several key variables. First, remittances are highly correlated with the stocks as well as net flows of refugees abroad (+0.6). For example, in 2011 the stock of refugees dropped by almost 400,000, mainly derived from return migration from Iran. During the same year, remittances were reduced by almost 50 percent, contrary to the general upward trend observed during the decade. Remittances are also positively correlated with the GDP growth of most receiving countries, which can be considered a proxy for the evolution of economic conditions and earnings. Interestingly, the highest correlation is with OECD countries (+0.6). In Iran the relation is lower (+0.2) while in Pakistan it is close to zero. With all the limitations of this simple correlation analysis over a very short period of time, the higher co-movements with OECD countries could reflect more propensity to send remittances (and higher quantities) among Afghan migrants in those developed countries. In contrast, Afghans in Pakistan, most of whom are refugees displaced with their entire families, might send lower remittances regardless of the state of the Pakistani economy. Economic growth in GCC countries, which are highly dependent on oil prices, might not be a good proxy for the income of migrants and thus the correlation is not high either. Table 1: Correlation between inflows of remittances in Afghanistan and key variables, 2008–2015 Annual GDP Growth Number of Migrants OECD Gulf Iran Pakistan Afghanistan Stock Net Flows countries countries 0.59 0.14 0.18 0.03 -0.42 0.65 0.60 Source: Own calculations based on UN DESA, UNHCR, and World Bank data. 21 3. Migration and Remittances in Afghanistan During the same period (2008–2015), remittances to Afghanistan were negatively correlated with the evolution of the Afghan economy (-0.4). This pattern highlights the role of remittances as part of the income-generating as well as the risk-diversifying strategy of Afghan families. As in many other countries, remittances are countercyclical and play a shock-mitigating role, supporting the local economy and income of Afghan families more in periods of hardship. This result is in line with other studies like that of Ahmed and Gassmann (2010), who observed that households that rely on remittances are less vulnerable to shocks than those that rely on wage labor, as remittances act as a buffer against shocks. 3.2 Approximating the True Scope of Remittances in Afghanistan To approximate the real magnitude of remittances into Afghanistan, it is useful to enhance the incomplete disposable data with estimates from different sources. The International Fund for Agricultural Development (IFAD) estimated the remittances inflow in 2006 to be around US$2.5 billion, or 29.6 percent of GDP (Orozco 2007). In a similar analysis in 2013, IFAD estimated it to have increased to US$3.2 billion, although the rise was lower than that of nominal GDP growth, which reduced the ratio to 16.3 percent of GDP (IFAD 2013). The study calculated remittances based on three estimates: the total number of migrants living abroad, the percentage of migrants who remit money, and the annual value of remittances sent. Overall, this estimate is 10 times higher than Da Afghanistan Bank’s official figures. Another simple approximation can be made with data on the share of registered financial intermediaries. According to informal communication with representatives of Da Afghanistan Bank, around 200 hawala dealers are registered out of an estimated total of 1900–2500 dealers. Under the strong assumption of similar levels and trends of transactions among the different dealers, this would imply that real remittances are 10 times the official figures, pointing again to around US$3–3.5 billion (or 15–18 percent of the Afghan GDP). Studies on Afghans’ remittances from the main host countries point to a similar degree of underestimation. The Afghanistan Investment Support Agency (2004) estimated the average annual remittance per Afghan migrant in the United States and Canada was US$1500, much higher than the official figures obtained from World Bank bilateral remittance data (US$110). As a result, the total remittances from those countries could reach US$75 million per year, 10 times higher. In a study conducted among deportees in Iran, Majidi (2010) observed that they were mostly single adult men who migrated to Iran alone with the goal of earning higher salaries and sending remittances back to their families in Afghanistan. These remittances often constitute the sole source of income and are a major livelihoods strategy for these families. According to this study, Afghans in Iran send back two-thirds of their monthly wage (which was US$320 on average) in the form of remittances, totaling up to US$2,500 per person annually. This represents a total value of US$500 million among deportees from Iran alone (200,000), and could be even higher if the entire migrant population in Iran was included. 22 3. Migration and Remittances in Afghanistan Information from household surveys in receiving countries yields significantly lower estimates compared to those focused on sending countries, and is more in line with the official figures. Household surveys provide another important source of information regarding the level and impact of remittances entering Afghanistan. According to the IS Academy Migration and Development Survey (UNU-MERIT/IOM), only 7.6 percent of Afghan households received remittances from abroad. This figure is similar to the one obtained in the Afghanistan Living Conditions Survey (ALCS 2013/2014) and the National Risk and Vulnerability Assessment (2007/2008 and 2011/2012), which are nationally representative surveys implemented by the Central Statistics Organization (CSO). However, this low number might reflect the usual problem of underestimation of remittances by household surveys, as they might face selection bias (under- sampling specific groups of the population) and underreporting of monetary information (e.g., remittances) due to mistrust or fear of taxation (Freund and Spatafora 2005). These biases can be particularly relevant in a post-conflict environment like that of Afghanistan. According to the ALCS, around 6 percent of Afghan households rely on remittances as one of their three main sources of income (Figure 10), representing a total of 215,000 households. As expected, the relevance of remittances is much higher among international migrant households (20 percent) than among those that never left the country (3 percent). For around 8 percent of households that returned to Afghanistan (either refugees or economic migrant returnees), remittances represent an important source of income. This suggests that few of them left social networks abroad. On the contrary, it is among households with a member currently living abroad that remittances play the most relevant role: for half of them, remittances represent the most important source of income (around three-fourths of the total), while for another 20 percent remittances are the second source (representing 30 percent of their income) or third source (15 percent of their income). Figure 10: Share of households for which remittances are one of the three main sources of income % HOUSEHOLDS WITH REMITTANCES AS ONE OF THE THREE MAIN SOURCES OF INCOME 80% 60% 40% 20% 0% Total Non-Migrant Int. Refugee Econ. Seasonal Members Migrant returnees returnees workers abroad Main Second Third Source: Afghanistan Living Conditions Survey 2013/2014. 23 3. Migration and Remittances in Afghanistan On average, the amount of remittances among recipient households was about US$1680 per year in 2013–2014 (US$140 monthly), more than half of their total income (ALCS 2013/2014). This shows that for those that benefit from remittances, the magnitude of the transfers received is sizable and can be of key importance. This result was also highlighted by Majidi (2008), who found that for more than two-thirds of recipient households in Afghanistan, remittances from relatives in Iran represented at least 75 percent of their household income. Several studies, such as those of IOM (2008), Vadean (2007), and Siegel et al. (2010), note that remittances are most often used to meet daily needs like food and housing expenditures. Aggregate estimates from the ALCS show that remittances accounted for US$340 million in 2013–2014, or 1.7 percent of GDP . This number is close to the remittance figures provided by Da Afghanistan Bank. However, these estimates might also be downward biased. On one hand, the ALCS only takes into consideration received remittances when they were one of three main sources of income, although comparing the share of households benefiting from remittances with other surveys indicates that this omission could be small (6 percent in the ALCS compared to 7.6 percent in the IS Academy Migration and Development Survey). More problematic could be that the actual number of households in the ALCS that have a member living abroad is 125,000, representing 160,000 current international migrants. Obviating the majority of refugees who usually moved abroad with their entire family (and thus would not be represented in the ALCS but also might not be as prone to sending money abroad), this number is still much smaller than the 2.2 million economic migrants estimated worldwide. Newly migrated Afghans (those that migrated during the year prior to when the ALCS survey was run) represent only 7 percent of the total labor migrants but seem to send the vast majority of remittances according to the ALCS. Therefore the wide divergences between surveys of sending and recipient countries (from 1.7 percent to 16 percent of GDP) are not based on disparities in the average amount sent per migrant but on the size of the migrant population, and in particular those that send money. Annex 1 of a companion policy paper shows different scenarios of the number of migrants who remit. If all labor migrants had the same propensity to remit as the newly migrated, the estimated level of remittances would reach 16.1 percent of GDP , around 10 time the official figures. This estimate would be more in line with other analyses like that of IFAD (2012) and with what can be derived from the share of hawala dealers not registered by Da Afghanistan Bank. However, this figure represents an upper bound and might actually overstate the true value of remittances as migrants are usually less likely to remit the longer they stay in the foreign country. Taking more reasonable assumptions about the propensity to remit given the length of stay abroad, it seems that the current value of remittances is between 3–7 percent of GDP , with a central scenario of 5.9 percent of GDP . Even in the most conservative scenario where remittance flows benefit a relatively small share of the population, they often account for a significant portion of income of those beneficiaries, highlighting their relevance to supplement income and cover basic spending needs. This fact, coupled with the countercyclical nature of remittances, points to a key role of migration as a survival strategy of diversification. Remittances, thus, act as a buffer against shocks that are prevalent in the Afghanistan economy, like the protracted war, conflict and insecurity, environmental catastrophes such as droughts and earthquakes, and economic shocks like unemployment or underemployment 24 25 References Abbasi-Shavazi, M., D. Glazebrook, G. Jamshidiha, H. Mahmoudian and R. Sadeghi. 2005. “Return to Afghanistan? A Study of Afghans Living in Tehran.” Afghanistan Research and Evaluation Unit (AREU) Case Study Series, AREU: Kabul. Afghanistan Investment Support Agency (AISA). 2004. “ Migration and Development. Working with the Diaspora: The Effect of Immigrant Remittances on the Economy of Afghanistan. AISA: Kabul. Ahmed, M. and F. Gassmann. 2010. “Measuring Multidimensional Vulnerability in Afghanistan.” Working Paper MGSoG / 2010 / 004. Maastricht Graduate School of Governance: Maastricht. Clemens, M. and D. McKenzie. 2014. “Why Don’t Remittances Appear to Affect Growth?” Policy Research Working Paper 6856, World Bank: Washington D.C. Collective for Social Science Research. 2006. “Afghans in Peshawar: Migration, Settlements and Social Networks.” Case Studies Series, Karachi, and Afghanistan Research and Evaluation Unit: Kabul. Eurostat. 2016. “Enforcement of Immigration Legislation: Third country Nationals Found to be Illegally Present.” Eurostat: Luxemburg. Freund, C., and N. Spatafora. 2005. “Remittances: Transaction Costs, Determinants, and Informal Flows.” Policy Research Working Paper Series, No. 3704, World Bank, Washington: D.C. Gupta, P. 2005. “Macroeconomic Determinants of Remittances: Evidence from India.” International Monetary Fund Working Paper 05/224. IMF, Washington D.C. IFAD. 2013. “ Sending Money Home to Asia Trends and Opportunities in the World’s Largest Remittance Marketplace.” International Fund for Agricultural Development: Rome. International Centre for Migration Policy Development. 2013. Afghanistan Migration Country Report, ICMPD, Vienna, Austria. International Crisis Group. 2009. “Afghanistan: What Now for Refugees?” Asia Report, No. 175. ICR: Kabul/ Islamabad / Brussels. International Monetary Fund (IMF). 2016. “Islamic Republic of Afghanistan: Second Review of the Staff-Monitored Program,” IMF Country Report No. 16/120. IMF: Washington D.C. International Organization of Migration (IOM). 2008. “Afghan Labour Migration and Human Resource Development.” 3rd Regional Economic Cooperation Conference. IOM: Islamabad. IOM. 2014a. “Afghanistan Migration Profile.” IOM: Kabul. IOM. 2014b. “Afghanistan Remittance Overview and Trends.” IOM: Kabul. 26 References IOM. 2016. “Migration Flows in Europe - Infographics.” IOM: Geneva. IOM. 2017. “Return of Undocumented Afghans.” January. IOM: Kabul. Koepke, B. 2011. “The Situation of Afghans in the Islamic Republic of Iran Nine Years After the Overthrow of the Taliban Regime in Afghanistan.” Middle East Institute, Study Groups: Afghanistan. Koser, K. 2009. “The Migration-Displacement Nexus in Afghanistan.” Brookings Institution, Washington D.C. Maimbo, S. 2003. “The Money Exchange Dealers of Kabul: A Study of the Hawala System in Afghanistan.” Working Paper No. 13, World Bank, Washington, D.C. Majidi, N. 2008. “Research Study on Afghan Deportees from Iran.” Kabul: ILO-UNHCR Cooperation Towards Comprehensive Solutions for Afghan Displacement (August 2008), p. 68. Majidi, N. 2010. “Research Study on Cross Border Population Movements between Afghanistan and Pakistan.” Altai Consulting, commissioned by UNHCR: Kabul. Monsutti, A. 2006. “Afghan Transnational Networks: Looking Beyond Repatriation. Synthesis Paper Series, Afghanistan Research and Evaluation Unit, Kabul. Oeppen, C., and A. Schlenkhoff. 2010. “Beyond the ‘Wild Tribes’: Understanding Modern Afghanistan and Its Diaspora.” Oxford University Press: London. Orozco, M. 2007. “Estimating Global Remittance Flows: A Methodology.” Inter-American Dialogue and International Fund for Agricultural Development: Rome. Siegel, M., M. Vanore, R. Lucas and C. de Neubourg. 2010. “The Netherlands- Afghanistan Remittance Corridor Study.” Report commissioned by the Dutch Ministry of Foreign Affairs, Maastricht Graduate School of Governance, Maastricht. Turton, D., and P. Mardsen. 2002. “Taking Refugees for a Ride? The Politics of Refugee Return to Afghanistan.” The Afghanistan Research and Evaluation Unit: Kabul. UN Department of Economic and Social Affairs (UN DESA). 2015. “Trends in International Migrant Stock: Migrants by Age and Sex.”. http://www.un.org/ en/development/desa/population/migration/data/estimates2/estimates15. shtml United Nations High Commissioner for Refugees (UNHCR). 2013. “Volrep and Border Monitoring Monthly Update.” April 2013. UNHCR: Kabul. 27 References UNHCR. 2015. “Solutions Strategy for Afghan Refugees to Support Voluntary Repatriation, Sustainable Reintegration and Assistance to Host Countries. ” UNHCR: Kabul. UNHCR. 2017. “Tough Choices for Afghan Refugees Returning Home After Years in Exile.” UNHCR, Pakistan. UNHCR. 2017. “Population Statistics Database.” UNHCR: Geneva. United States Committee for Refugees and Immigrants. 2009. “World Refugee Survey 2009 – Iran.” June. USCRI: Arlington, VA. Vadean, F. 2007. “Skills and Remittances: The Case of Afghan, Egyptian and Serbian Immigrants in Germany.” Working Paper, No. 92, Asia Research Institute, Singapore. Wickramasekara, P., J. Sehgal, F. Mehran, L. Noroozi, and S. Eisazadeh. 2006. “Afghan Households and Workers in Iran: Profile and Impact.” ILO-UNHCR Cooperation Towards Comprehensive Solutions for Afghan Displacement. ILO- UNHCR: Geneva. 28 38