The Economic Impact of Forced Migration

Source: OCP Policy Center

The flow of forced migrants from Syria, Iraq, Afghanistan, Eritrea, Somalia, Mali, and other localities beset by conflicts is a human catastrophe of the first order, the cause of the uprooting of millions of families and of perilous journeys that have led to thousands of deaths. It is also a crisis that has major international implications, from the risk of radicalization and political instability, to the spread of infectious diseases, and to the risk of collapse of the visa-free European Schengen zone under pressure of migrant flows and the inability to agree on an appropriate solution. Since the conflicts that are causing the crisis are endemic and there is little likelihood that the vast majority of migrants can soon return to their homes, the crisis may get worse before it gets better. As policymakers and politicians try to make sense of a complex reality, in a political environment that has become even more explosive in the wake of terrorist attacks, it is important to get the facts straight, and to understand some of the economic regularities associated with mass migration. What are the development and economic implications of large flows of forced migrants? This, after all, is a phenomenon of which the world has many examples and long experience, and which economists have widely studied.

This chapter deals mainly with refugees and asylum seekers, forced migrants who cross international borders, rather than with internally displaced people. It draws on the extensive economic literature on voluntary and forced migration as well on the recent field experience of the World Bank, the UNHCR, and other national and international agencies. It makes the following points:

• Forced migrants tend to arrive in places where there are few job opportunities for them. The problem of finding a livelihood is far more severe in the ten or so poor countries or regions receiving large numbers of forced migrants than in rich countries which, with few exceptions, receive tiny numbers of forced migrants relative to their population. In poor countries or regions receiving large numbers of migrants, the depressing effect on the wages of native unskilled workers can be extremely severe.
• When forced migrants arrive in large numbers in a poor region, they place an enormous strain on public services and infrastructure and on the public purse, and they can also severely fray the social and political fabric, leading to deterioration of the investment climate.
• Concerns that accepting an increased number of forced migrants in advanced countries will cause job losses or falling wages, and place an undue burden on the public purse are largely unjustified. In most instances, in advanced countries, the arrival of young people willing to work is likely to cause a proportionate expansion of investment and output, and may also accelerate the economy’s long-term growth rate.
• The global gains from forced migrants settling in the South -- those accruing to the migrant, the country of origin (through remittances) and the country of destination - are much smaller than from migrants settling in the North. Though most countries in the South receive very few forced migrants, when forced migrants arrive in large numbers in fragile states or regions, net welfare losses may result.
• Forced migration flows which are mismanaged, as at present, create large negative externalities for the surrounding region or even for the world. There is no perfect scheme for allocating the burden, but, absent political solutions to conflicts, any scheme must envisage increased numbers of refugees settling in the North, redoubled efforts to integrate refugees in their country of asylum, and increased development aid for the countries in the South with the largest numbers of refugees. Such a scheme is more likely to materialize if it is based on voluntary – rather than compulsory – targets to welcome refugees and provide aid, and if a new comprehensive framework for dealing with refugees is adopted.

The chapter’s starting point is: what sets forced migrants apart from voluntary ones?

Characteristics of Forced Migrants

The economic literature on migration is principally concerned with the implications of voluntary migration, while the economic literature on forced migration, though growing rapidly, remains quite sparse. The literature on voluntary migration does, nevertheless, help shed some light on the phenomenon of mass forced migration since there is overlap between the two fields of study. For example, analysts of the effects of voluntary migration have often analyzed instances of forced migration to identify the effects of the sudden “exogenous” arrival of large numbers of migrants¹.

However, there are a number of differences between forced and voluntary migrants, which carry important economic and policy implications. Voluntary migrants move for economic gain or to unify their family, and so go at a time and to a place of their choosing, while forced migrants flee, often to the nearest safe haven, to avoid bodily harm. Thus, forced migrants, unlike voluntary migrants, may lose most of their assets, and may end up in a place where job opportunities for them are scarce or completely absent, and may be unable or unwilling to return to their country of origin regardless of how they fare in their adopted place of abode. Compared to voluntary migrants, their adjustment to new conditions may take longer, and may require more support from the host community or from others. These characteristics lead to a number of complex economic challenges which are not present or not present to nearly the same degree in the case of voluntary migration.

Moreover, the legal distinction between refugees, who are defined as forced migrants who seek and are granted asylum, and economic migrants, who are voluntary migrants, has important bearing, since under the 1951 UN Convention member countries are obligated not to return refugees to their country of origin where they may be harmed (“non-refoulement”). Yet, this legal distinction is often hard to draw in practice, and for the purpose of economic analysis the distinction is even fuzzier. In reality, many refugees – while genuinely escaping from danger -­- do not leave at the point of a gun but have some latitude on the timing of their departure and their destination. Confronted with a threat, they will consider the risk and cost of migration, and prospects at different destinations. Like voluntary migrants, a family fearing persecution or war may decide to send first a young, able-bodied and enterprising member to explore. Conversely, some economic migrants may, to all purposes, be forced to migrate by hunger and disease.

From an economic perspective, the passage of time may make the distinction between forced and voluntary migrants, or between refugees and economic migrants, even less meaningful. Since for the vast majority of forced migrants the capacity to work is the only asset, sooner rather than later they must settle and find a job. In many instances, this is sought in a region or country other than that of first arrival, or, less commonly, in the country of origin to which the forced migrant may be willing and able to return. Thus, a migrant who was initially forced to migrate subsequently acts much like a voluntary migrant, and the analysis of the economic impact of voluntary migration can be even more useful in understanding the effect of forced migration.

Still, there are four economic challenges which are uniquely associated with forced migration and which require specific analysis. The short-term challenges include the job market mismatch problem, and the effect on the region of first arrival of large numbers of forced migrants. The long-term challenges relate to the region of settlement of the forced migrant, and specifically the overwhelmingly South-South nature of forced migration, as well as the international coordination problems caused by the phenomenon. These challenges, which are all highly specific to forced migration, and are all related, are treated in turn below.

Job Market Mismatch

The typical forced migrant breadwinner must find a good job, and quickly, to mitigate his or her family’s plight, and to overcome the poverty, uncertainty, and humiliation frequently associated with dependence of humanitarian aid. Finding a job, even a bad job, is also the key to their becoming productive members of their adopted society. Yet, there are many impediments to this happening, including, in many cases, outright refusal of the permission to work. So, the forced migrant’s job prospects are a good place to start in analyzing the economic impact of forced migration. While there is extensive evidence that voluntary migrants respond to the demand for their services, and they go where (and when) they can find jobs (Papademetriou, Sumption & Somerville, 2009), there is also compelling evidence that forced migrants seeking employment will more often than not find themselves in the wrong place, at least initially. If they are not allowed to work, and are not permitted to move on, or are unable to cover the cost of another trip, or are simply unwilling to take the risk of moving, their living standards as well as their ability to contribute to society are severely compromised.

Most countries – whether developed or developing -­- receive only very small numbers of forced migrants relative to their population in a given year and the effect on their job market is minuscule. In those cases, the chances that a forced migrant will eventually find decent work are reasonable even if they may not be as good as those of well-connected natives. The problem arises when the inflow of forced migrants is large relative to the host population. Technically, the job mismatch problem facing forced migrants can be defined as an excess supply of workers which can only be corrected by a fall in wages, or by an increase in investment. Such a decline in wages may not be allowed to occur because it is too large to be politically acceptable (leading to work permit denial) or because of the presence of other labor market restrictions that protect incumbent workers. At the same time, new investments may occur too slowly or be deterred by a fragile investment climate or by political instability, which is sometimes made even worse by a large inflow of migrants. The result is that migrants are not absorbed into the labor force, become dependent on humanitarian aid. The available evidence suggests that the job market mismatch problem facing forced migrants is always severe at first, including for the relatively small number of refugees who find abode in the world’s richest countries, but it can be extreme and persistent for the much larger number of forced migrants that find themselves concentrated in relatively poor regions and in a fragile political environment. Where refugees are not allowed to work, many tend to become part of informal labor markets anyway, in a part of the economy which is largely outside the tax net, work at extremely low wages, and are subject to exploitation and abuse.

The evidence on the job market mismatch problem in both advanced and developing countries is partial and spotty, but, while it is discomforting in developing countries receiving large numbers of forced migrants, it is far from uniformly bleak in advanced countries. For example, Germans and Finns displaced after World War 2 and French and Portuguese displaced in the wake of colonization who returned to their country of origin, were allowed to work and fared fairly well according to the available studies (comp. De Lima & Carrington, 1996; Hunt, 1992). However, these groups are not representative of present day forced migrations of people of diverse ethnic origins. In recent decades, most refugees – which tend to be overwhelmingly unskilled - arrived in advanced countries without knowing the language and, often, without being able to write in the Latin script. Some arrive illegally but even those who arrive legally often have to wait many months or years before they are allowed to work. The refugees that have professional qualifications typically find that they are not recognized and they must accept a lower status or pass new exams which take many years of preparation.

Even so, recent evidence suggests that, despite the increase in work permits granted to refugees in Germany – a country that received large numbers of asylum seekers in 2015 - the national unemployment rate continued to fall in every month compared to the same month in the previous year (IAB 2015, 2). The OECD estimates that the growth of the labor force across the Euro area will barely reach 0.5% per annum in 2016-2017, with a significant part of that growth due to the large net migration to Germany. While the medium and long-term impact will be determined by how well refugees are integrated into the labor market, according to the OECD this inflow “could ease emerging wage pressures in the comparatively-tight German labor market” and “also help to moderate demographic pressures due to population ageing” (OECD 2015c, p.16).

Unfortunately, systematic studies of how modern day refugees have fared in advanced countries over the long run are few and largely confined to the United States. These studies tend to concur that refugees find a job within a few months of arrival, that they earn significantly less than natives with equal qualifications at first, and that knowledge of the English language – which is often scant - is an important determinant of their income. Most refugees arriving in the United States rely initially on government aid, but very few still depend on aid within 2 or 3 years of arrival. A disproportionate number of refugees are small-scale entrepreneurs. While the income of refugees rises quite rapidly over time, there is insufficient evidence to conclude that their incomes converge completely to those of natives. It is worth noting that the more abundant evidence on voluntary immigrants suggests that their incomes do not converge completely (Alcobendas & Rodriguez-Planas 2009; Cortes 2004; Card 2005) even after many years.

A recent longitudinal study of the United States by Cortes (Cortes 2004) finds that refugees adjusted rapidly and – surprisingly - that their income progression significantly outstrips that of economic migrants. While refugees earned 6% less than economic migrants in 1980 and worked 14% fewer hours, by 1990 they earned 20% more, worked 4% more hours, and improved their English skills by 11% more. She shows that refugees invested more in education. Cortes finds that these results cannot be attributed to differences in the ethnic composition of economic migrants and refugees, and speculates that they are due to the fact that refugees are more committed to remaining in the United States in the longer term because they cannot return or do not want to return.

The available studies of the insertion of forced migrants in developing countries are also few, but as mentioned, they offer a much less comforting picture. This is as might be expected, because in the instances studied – unlike in the United States - the refugee inflow is very large relative to the domestic population, and the migrants, which are unskilled, tend to be directly competitive with the local population, which is also predominantly unskilled. Moreover, in the instances studied, wages are low and the job market situation is precarious to start with. According to a recent World Bank study, “The influx of Syrian refugees [in Lebanon] is expected to increase labor supply by between 30 and 50 percent – with the largest impacts on women, youth, and unskilled workers […] The overall unemployment rate and share of informal work in total employment could both increase each by up to 10 percentage points” (Ianchovichina & Ivanic 2014). In a recent welfare assessment of Syrian refugees in Jordan and Lebanon, the World Bank found a high incidence of poverty, and -surprisingly - that access to labor market was not a significant determinant in poverty levels (Verme 2015a), presumably as the only jobs available to refugees are at wage levels which are extremely low.

An ILO study of forced migrants in Lebanon, where they are estimated to account for over 20% of the local population, finds that Syrians (many of whom are allowed to work in Lebanon), have a 30% unemployment rate and their average monthly wage is $277, which compares to the statutory minimum wage of $448. Syrian women earn a little above 1/3 of the minimum wage (Masri & Srour 2014). In comparison, a similar ILO study on Syrian refugees in Jordan indicates “wages for unskilled and semi-skilled wages of JD 4-JD 10 per day, […] insufficient to provide subsistence livelihoods” (Aljuni & Kawar 2014, p. 21). In line with that, a World Bank welfare assessment on Jordan found that 69% of Syrian refugees in Jordan were poor using the UNHCR poverty line (Aljuni & Kawar 2014; Verme 2015a).

A recent study of internally displaced people in Colombia, which are estimated at 4.7 million equivalent to roughly 10% of the country’s population, concludes that localities that saw the largest number of displaced people have seen wages fall very sharply, and, on average across the national territory, displacement has led to a reduction in wages of 28.4 percent in urban areas (Calderón-Mejía & Ibáñez 2009, p. 25). While wages and employment were not much affected in the highly regulated Colombian formal sector, which represents 60 % of the labor market (Calderón-Mejía & Ibáñez 2015, p.6), the effects on the informal sector were especially severe, with wages declining more sharply than they might have done in a more uniform labor market and with informal employment expanding markedly (Calderón-Mejía & Ibáñez 2015).

The job market mismatch problem, which is clearly much more severe in developing countries absorbing large numbers of forced migrants relative to their population than in advanced or developing countries absorbing small numbers, is only one aspect of the impact on the region of first arrival, discussed next.

Impact on the Region of First Arrival

Forced displacement typically takes place in a relatively short time frame and involves large numbers, unlike the more regular inflow of voluntary migrants. As mentioned, the refugee crisis affects specific countries and regions disproportionately, while the vast majority of advanced and developing countries receive very few refugees in a given year; this is a case where initial conditions and the size of the shock matter, and they matter greatly. Over and beyond the labor market impact discussed above, forced displacement can constitute a large demographic shock causing a sudden mismatch of supply and demand of public services and housing in the host community. In countries or regions where the investment climate is inauspicious to start with and where the inflow of foreigners disrupts established social and political equilibria, business expectations can be adversely affected. This can delay or indefinitely impede the private investment response to rising demand for housing and services. Fiscal constraints can limit the required investment in public services, and, in extreme cases, balance of payments constraints can lead to exchange rate devaluation and make the import of food, fuel and other necessities more expensive.

The current media focus on the migrant crisis in European countries, most of which receive very few forced migrants relative to their population, diverts attention from the far more severe challenge in a number of host communities in Africa, the Middle East and South Asia. Over half of the refugees worldwide originate from Syria, Afghanistan and Somalia and over two thirds of displacement takes place within the region of origin (UNHCR 2015), in host communities where social, economic and institutional capacities are low to begin with, and where fiscal space and capacity to borrow is limited. How the arrival problem plays out depends greatly on the specific context of the host community as well as the socioeconomic profile of the displaced population. By way of illustration, the cases of Germany, Italy, Jordan, and Lebanon, countries which have recently attracted large absolute numbers of forced migrants, are compared in Table 1.

Table 1: Ability to Cope: Economy and Refugees 2014

	Germany	Italy	Jordan	Lebanon
Population (#)	80,889,505	61,336,387	6,607,000	4,546,774
Surface area (sq. km)	357,170	301,340	89,320	10,450
GDP (current US$ billion)	3,853	2,144	36	46
Average Annual Growth Rate GDP (2010-2014 %)	1.96	(0.52)	2.70	3.02
GNI per capita, PPP (current int. $)	46,840	34,710	11,910	17,400
Unemployment rate (% of total labor force, modeled ILO estimate 2013)	5.30	12.20	12.60	6.50
Refugee Population (2014, UNHCR)	216,973	93,715	654,141	1,154,040
Refugee Inflow (2014, UNHCR)	40,563	20,582	119,000	403,600
Estimated Net Annual Migration (2011-2015 United Nations Population Division, annual average)	549,998	900,000	400,002	500,001
Refugees to host population ratio (# of refugees per 1,000 inhabitants)	2.63	1.53	87.16	232.39
Source: World Bank, World Development Indicators

The difference in per capita incomes between Germany and Italy, and Jordan and Lebanon, after adjusting for purchasing power, is about 3 or 4 to 1 in the case of Jordan, and 2 or 3 to 1 in the case of Lebanon.

Lebanon and Jordan rank high among countries hosting refugees. However, they are not among the top 10 host countries when all forced migrants, including IDPs, are included. As Chart 1 shows. Syria, Colombia, and Sudan and South Sudan, which are much poorer countries host far larger numbers of displaced people. Lebanon, Jordan, Italy and Germany do not even currently feature among the largest hosts when all forced migrants are considered.

Chart 1: Major Host Countries, UNHCR Database 2015

Germany received close to 1 million asylum seekers in 2015, equivalent to 1.2% of its population, the “largest annual inflow of people seeking asylum of any OECD country ever recorded (OECD 2015). Still, even if most of these people were granted asylum, the stock of refugees in Germany (less than 3 per thousand) would be tiny compared to Lebanon (232 per thousand) and Jordan, (87 per thousand), as Chart 2 shows. Italy’s refugee population currently stands at less than 2 per thousand. In the United States, where Congress has just voted against admitting additional refugees absent a draconian vetting process, and only up to 10,000 Syrian refugees per year, the number of refugees represents less than 1 per thousand of the population. Refugees also historically represent a small share of all migrants, some 7%, compared to voluntary migrants. To put the current flow of refugees in perspective, Spain attracted around 4.3 million new, overwhelmingly voluntary, immigrants over 2000-2008, equivalent to about 9% of its population.  

Chart 2: Incidence of Refugees, UNHCR Global Trends 2014

The flow of large numbers of displaced people in a short time affects the host community profoundly. Below, three dimensions are examined: public services, fiscal balance, and political stability.

Public Services and Infrastructure

In developing countries, low incomes and fiscal constraints mean that the provision of health care, education and social services barely satisfies the needs of the native population to start with, and excludes many, even before the arrival of refugees. Over the period of 2012-2014, a World Bank assessment found that to restore access and quality to these services in Lebanon to the level possible before the arrival of refugees would cost 1.4-1.6 USD billion, (World Bank 2013, p. 3). Similarly restoring the level of Lebanon’s infrastructure services to that prior to the arrival of refugees - defined as water and sanitation services, municipal services, electricity and transport - was estimated to cost 1.1 USD billion. In total the social and impact assessment estimated “that an additional spending of USD 2.5 billion would be required for stabilization, i.e., to reinstate the access to and quality of public services to their pre-Syrian crisis level” (World Bank 2013, p. 1). This sum represents about 5.5% of Lebanon’s GDP.

Because the inflow of migrants is relatively small and public services are better developed and funded to start with, OECD countries have much larger capacity to deal with the immediate demand shock on public services than poorer hosts, even though the concentration of refugees in some regions or cities can cause the shock to be uneven, as in the case of some school districts receiving large numbers of immigrant children. Since the population of both Germany and Italy is projected to decline with the natural increase across the EU turning likely turning negative for the first time in 2015 (OECD 2014, p. 35), migrants can be accommodated without necessarily requiring expansion of the existing infrastructure, or of health services which migrants, who are relatively young, draw on less intensively.

In contrast, the challenge for public services in the small number of developing countries confronted with a surge of forced migrants is compounded by the composition of the migrants. In Jordan and Lebanon, the Syrian refugee population consists of relatively vulnerable groups, placing even greater demand on education and maternal health. Compared to the population of Syria, refugees are disproportionately children and the head of household is disproportionately female (Verme 2015b, p.5), many mothers are under 18, and the level of education of refugees is marginally lower. In contrast, there is anecdotal evidence to suggest that those forced migrants who reach Europe have higher skills and larger assets. For example, an OECD brief finds that “more than 40% of Syrians in the country [Sweden] in 2014 have at least upper secondary education” compared to 15% of those in Lebanon and Jordan (OECD 2015, 8). Still, while refugees in Jordan and Lebanon speak Arabic, those in Italy and Germany will require language training.

Immediate Fiscal Impact

A recent assessment of the fiscal impact of refugees in Lebanon concluded that “public finances that were structurally weak prior to the Syrian shock are now becoming severely strained, with the deficit estimated to widen by USD 2.6 billion [equal to roughly 1.4 % of GDP] over the 2012-2014 time period” (World Bank 2013, p. 2-6). No equivalent assessment is available for Jordan, but the fiscal burden is expected to be less, since policy has been to settle refugees in camps – which are funded by UNHCR – unlike in Lebanon where refugees draw directly on public facilities and resources.

Costs of hosting refugees in OECD countries are higher per refugee than hosting them in developing countries but much smaller in proportion to GDP. Official statistics are difficult to find, but the German magazine Zeit, for example, estimates that hosting a refugee will increase public spending by EUR 10,000-12,000 per refugee per year. Even if all the estimated 1 million asylum seekers in Germany in 2015 are granted refugee status, the costs would then come to EUR 10-12 billion for 2015. In the most extreme scenario, Zeit estimates added total public expenditures of EUR 65 billion for asylum seekers and refugees through 2015-2018, which would be partly offset by about EUR 20 billion in increased government revenues reflecting the expansion of output. The net costs of EUR 45 billion in the five year timespan would represent around less than .3% of GDP and could be covered by drawing on Germany’s projected budget surplus, without raising taxes or raising more debt (Zeit 2015). Italy is expected to receive a far smaller number of migrants but runs a fiscal deficit and, unlike Germany, has a structural unemployment problem. Still, the fiscal impact of hosting more refugees in Italy over the next five years is likely to be marginal.

Whether they are hosted in a developing or advanced country, the fiscal impact of forced migrant inflows depends critically on policies designed to integrate them in the formal labor market, and the tax revenue they generate. Refugees often have some limited assets to draw from on arrival, which tend to boost the consumption of local goods and services; however, these funds run out quickly if refugees cannot replenish them through gainful employment². Settlement regulations also matter. Refugees in camps are less likely to become part of the formal labor market and to contribute through taxes. In Jordan and Lebanon the employment opportunities in the formal sector are scarce. In Europe, asylum seekers have to wait 9 months on average to receive a work permit, though in Germany the delay was recently cut down to under 5 ½ months (OECD 2015c, p.18). One estimate suggests that If they are integrated into labor markets speedily, “the influx of refugees could lift EU growth by 0.2 percentage points, due both to demand-side effects in the short run and to supply-side ones in the longer run” (World Bank MFM GP 2015, p.5). Refugees can also boost tax revenue by facilitating trade and investment flows between the origin and host communities, as appears to be happening in the case of Syrian refugees moving their operations from Syria to Turkey (World Bank MFM GP 2015, p.2) but maintaining business links with Syria.

Social and Political Imbalance

While this analysis has mainly focuses in on the economic effects of the sudden inflow of refugees into a host community, there are also implications for the social and political fabric which can have economic repercussions. Indeed, although real or perceived adverse economic effects can be an important cause of social tensions, concerns about maintaining the political balance, preserving local mores, protecting national security, not to mention the ever-present xenophobia and racism, have a dynamic of their own. In some instances, the inflow of displaced populations reignites preexisting tensions among different ethnic groups over resources, territory, or influence. Examples include tensions caused by a large inflow of Bangladeshis into Northern India (Times of India 2013), Mozambican refugees in Malawi, Eritreans in Eastern Sudan, and displaced Ugandans clashing with native population within Northern Uganda (comp. Refugee Studies Centre 2011, p. 78). The recent terrorist attacks in Paris and Brussels and allegations of sexual harassment by immigrant men in Cologne and Stockholm have ignited fears of communal tensions. These incidents, attitudes and concerns, as much as economic effects, lie at the heart of the widespread resistance to welcoming large numbers of refugees. A demographic shock which affects existing political, ethnic, religious, and social equilibria, creates new winners and losers in the system and can exacerbate existing fragilities. Similar to the economic effects, some of these imbalances might resolve themselves or create new acceptable equilibria, but others might be seen as unacceptably disruptive by specific minorities or even the majority of the population.

Even a cursory examination of countries hosting large numbers of forced migrants reveals that these countries typically exhibit a fragile investment climate. Thus, about half of the top ten hosts of forced migrants do not even feature in the most recent World Economic Forum Competitiveness Rankings, while most of the rest of the group, as well as Lebanon and Jordan, rank very low on the competitiveness list.   

Often concerns hinge on security rather than on material well-being. Studies have shown that it is not just the size of the displaced populations but also the duration of displacement and the pre-existing inter-group  dynamics which determine insecurity as in the case of Afghan refugees in Pakistan (Schmeidl 2002) or refugee-hosting areas in Tanzania (Rutinwa and Kamanga 2003). Encampment of refugees can reinforce an ‘us versus them’ dynamic. Camps or enclosed spaces can strengthen a national narrative, in a positive way, as in the ‘Mayan revival’ of Guatemalan refugees or contribute to radicalization, as in the case of Palestinian refugees in the blockaded Gaza strip. But even refugees who become integrated in the labor market and are self-sufficient, as in the case of urban refugees in Nairobi, still “remain vulnerable to arrest and xenophobic violence, illustrating the significance of contextualizing employment levels with the broader political economy of displacement” (Refugees Studies Centre 2011, p. 65, comp. Campbell 2006).

Still, there is clearly a link between the economic and political impact of mass forced migration host communities. The inflow of refugees “has challenged the already delicate societal and inter-communal balance in Lebanon […]; overcrowding, saturation of basic services and competition for jobs are among the root causes for social tensions between host and refugee communities” (Verme 2015a, p.5). These relationships can work both ways, however. In Germany, “opinion polls suggest that the more favorable economic situation is also linked with a more favorable public opinion regarding receiving refugees, although this has not prevented xenophobic attacks. Public opinion at least appears better prepared for immigration compared with the situation in the early 1990s.” (OECD 2015a, p.11).

The preceding sections have shown that the immediate impact of a large and sudden flows of refugees creates challenges for policy-makers and for forced migrants in any environment, but that the challenge is especially daunting when the inflow of refugees is large relative to the native population, the host country is poor, has little fiscal space, and low administrative capacity. Not surprisingly, while many migrants stay in the country or region of first arrival, many others choose to move again if they can, and to settle permanently in another place, whether in the North or the South, where there is a job for them. The next section examines the long-term impact of migration on the region of settlement.

Settlement, and the Predominantly South-South Nature of Forced Migration

In contrast to the wealth of data-driven research on the economic impact of migration into high income countries, the literature on migration into low-and middle-income countries is sparse³. Yet, the stock of South-South migrants, at 82.3 million, which includes both voluntary and forced migrants, edges that of migrants from the South that reside in the North (UNDESA 2013), at 81.9 million. However, the share of forced migrants that originate in the South and stay in the South is overwhelming. “Refugees account for a relatively small proportion of the global migrant stock. In 2013, the total number of refugees in the world was estimated at 15.7 million, representing about 7% of all international migrants. Nearly nine of every ten refugees in the world had found asylum in developing regions“(OECD, UNDESA 2013). The implication is that about 10% of the South-South migrants are forced, compared to less than 2% of South-North migrants (IOM 2013, 68) (see table 2). Thus, although much of the attention of researchers and journalists is focused on the refugees in the North, their number is tiny, not only in comparison to the size of population but also compared to the number of voluntary migrants in the North. Nevertheless, this section pays much attention to the impact of forced migration in the North as well as to those countries in the South that attract many forced migrants, for two reasons. First, because, as is made clear in the previous discussion, although the North allows entry to only relatively small number of forced migrants, its capacity to absorb larger numbers is considerable. Second, because the rich vein of research on immigration in the North can help shed insights on immigration in the South, albeit with caveats.

Settlement in the North

As discussed, with the passage of time, the economic impact of forced migrants who are allowed to work or who are allowed to move to find work, comes to resemble that of voluntary migrants. Even though the initial decision to migrate is forced, eventually those who are allowed to do so (or do so even if they are not allowed) decide where and when to seek work. For example, the likely economic impact of large numbers of Syrian refugees in Germany where they are being taught German and allowed to work, is similar in the long-term to that of voluntary migrants. For this reason, analyzing the impact of forced migration on the region of settlement can draw usefully on the literature on voluntary migration as well as that on forced migration.

The literature on permanent migration in the North, whether forced or voluntary, is primarily concerned with four issues: the impact on the wages of natives, especially on the wages of native unskilled workers; the impact on unemployment; the long-term fiscal impact of immigration (as distinct from the immediate fiscal cost of supporting refugees); and the impact on aggregate economic growth through investment and increased productivity. Not surprisingly, these issues are closely linked, since, for example, insofar as immigration affects investment and economic growth, it can be expected to affect the wages and employment of natives, and to affect public finances, indirectly as well as directly; and, whether immigrants find jobs affects their ability to pay taxes.

The Effect of Immigration on Economic Growth

Consider the following story. A country, Welcomia, receives a large number of forced (or voluntary) migrants. If migrants are allowed to work, the labor market is flexible, and the investment climate is supportive, one would expect that the influx of migrants will have a similar effect on the economy as an increase in young natives (migrants tend to be young) joining the domestic labor force. The increased inflow of workers raises the demand for capital, including housing, and the return to capital, and stimulates investment. In the longer run, one can expect the economy of Welcomia to return to the same capital-labor ratio and initial wage (Solow 1956). Thus, a large one-time influx of migrants simply scales up the economy proportionately to the increase in the labor force associated with migration. Though the rise of wages of native workers may slow or wages may decline temporarily while the adjustment takes place, increased investment will ensure that wages rise again and are unchanged in the steady state compared to the counterfactual of no new migration. The adjustment to the new equilibrium will occur faster in an open economy which can import capital to accompany the larger labor force and where the expansion of labor-intensive export industries can expand to absorb the additional labor. Flexible and buoyant economies with a conducive investment climate and efficient capital markets will adjust faster to a migration shock – as to any shock - than economies that do not display those characteristics.

The realism of this story hinges on the assumptions that the capital-output ratio is fairly constant, and that the capital stock and the labor-force to grow tend to grow at similar rates over very long periods. It turns out that these are two long-established empirical regularities in economics (Harrod 1926). Moreover, recent research in advanced countries suggests that domestic investment is quite quickly stimulated by a migration surge, so that within a few years the capital-labor ratio tends to return to its prior level in the face of a labor market shock. This conclusion is supported by studies such as Ottaviano and Peri (2008) for the US, Brücker and Jahn (2011) for Germany, Cohen and Hsieh (2000), and Ortega and Peri (2009) in a study of 14 OECD countries. Migration in advanced countries tends to occur at a moderate and quite steady pace (typically adding about 0.5% to the labor force every year), so that its effects are embedded in business plans, making the adjustment smoother. Currently, the inflow of refugees in advanced countries, with few exceptions such as Germany and Sweden, represent a small share of this immigration.

Economic growth is driven not only by the increase in the stock of labor and capital that immigration can trigger, but also – and even more importantly – by the growth in productivity of all factors of production. Unfortunately, few studies have addressed the issue of how immigration affects productivity, and the available evidence is mixed. Orefice (2010) speculates that not only does immigration generate investment opportunities, migrants can affect TFP in host countries in many ways, such as boosting entrepreneurship, providing new opportunities for specialization among workers and to exploit complementarities, increasing flexibility, reinforcing agglomeration economies, and populating remote locations, thus raising the returns to capital as well as wages across the economy. He examines the effect of migration on per capita income using bilateral migrant flows from 86 developing countries to 24 OECD countries from 1998 to 2007 and concludes that, in this process, the skill level of migrants matters a lot. If migrants are high-skilled, they will tend to raise per capita income and prompt a long-term increase in the economy’s capital-labor ratio; if, on the other hand, they are predominantly low-skilled, they will reduce average income per capita in the short run and promote the adoption of less productive, more labor-intensive technologies. In contrast, carrying out an analysis of the effects of migration on US states, Peri (2009) and Peri and Sparber (2008) found that, at the state level, migration is associated with increased output per worker in the long run, a reflection of increased investment and specialization effects as typically natives take on different tasks, leading to efficiency gains through specialization. For example, in states with an influx of unskilled migration, who are typically non-English speakers, natives take on an increased share of communications-intensive jobs. Over the long run (10 years), a net inflow of migrants equal to a 1 percent of employment increases income per worker by 0.6 - 0.9 percent.

Other studies have found that while unskilled immigrants may initially lower average incomes and wages in the host country, in the longer term they and their offspring can help accelerate the growth of income per capita, provided they become integrated. In an attempt to evaluate the long-term impact of unskilled migration, Card (2005) examines the earnings of first generation immigrants in the United States over time and also the educational attainment of their children. He finds that first generation migrants close only about a quarter of the 40 percent-gap in earnings with respect to natives, but their children tend to do slightly better than the children of natives. There is a strong correlation between the education levels of children and those of their parents, but even the children of Mexican immigrants, whose parents have very little education on average, close a very large part of the education gap with respect to the children of natives. A recent cross-country study by the OECD (2009) found a marked difference in the achievement of children of migrants in the United States and other traditional immigration countries, who matched that of the children of natives, and those of children of migrants in Europe where they underperformed by a wide margin, after controlling for socio-economic differences. Results consistent with these findings were reported by Dustmann (Card et al. 2012) and by a recent Johns Hopkins study (Hao and Woo, 2012).

The Wages of Unskilled Workers

The fear that large inflows of unskilled migrants from the South will take jobs away from unskilled natives in the North has motivated a very large literature. Going back to the story of Welcomia in the previous section, consider the different effect of the inflow of large numbers of predominantly unskilled migrants on the native skilled and unskilled workers. Skilled natives, who are complementary to the unskilled immigrants, will tend to end up better off in the new steady state on account of both increased investment and having more unskilled workers to work with. However, the effect on unskilled natives is a priori indeterminate, since, while they, too, benefit from increased investment, they will confront increased competition from the unskilled migrants. Which of these effects dominates will depend critically on the extent to which the unskilled migrants are close substitutes for domestic unskilled workers, a question that has been extensively studied. By estimating a production function including groups of workers classified by education, experience, gender and origin, Ottaviano and Peri (2008) for the U.S., Manacorda et al. (2006) for the UK, and for Brücker and Jahn (2011), D’Amuri and Peri (2011) and Felbermayr et al. (2010) for Germany reach the conclusion that unskilled immigrants and natives are imperfect substitutes. Immigrants who do not speak the language, often cannot read and write in Latin script, have scant networks, and relatively low expectations, tend to do different jobs than unskilled natives, so compete with them only indirectly.

Consistent with these findings, most – though not all - studies of the effect of unskilled migration on the wages of unskilled workers find only small negative effects. The early literature on the subject typically concluded that a 1 percent increase in the immigrant share in the population causes no decline in wages or a decline of 0.1 percent. These “area studies” attempted to exploit the variation in migration incidence across countries, or more typically, across localities in the United States. A partial list of these studies includes Altonji and Card (1991), Butcher and Card (1991), LaLonde and Topel (1991), and Schoeni (1997). However, as Borjas pointed out in a landmark contribution, the area studies often failed to contend with a serious endogeneity problem, since the intensity of migrants in a specific locality is itself influenced by wage and employment opportunities, tending to bias the estimated effect of migration on wages severely downwards (Borjas, 2003). The additional criticism on studies that correlate the variation in immigration intensity across localities in the United States with wages, is that natives (and other migrants) compensate for increased migration by moving elsewhere.

One solution – albeit imperfect - to the endogeneity problem, which also turns out to be highly relevant to the study of forced migration, was to examine instances of mass migration which are prompted by political or other external events and therefore are not plausibly in response to labor market conditions in the receiving country. In 1990, David Card published his classic analysis of the influx of Cuban migrants to Miami during the 1980 Mariel Boatlift (named after the port in Cuba where the boatlift took place) (Card 1990). Card found that, although the Mariel immigrants increased the labor force of the Miami metropolitan area by 7 percent, they had virtually no effect on the wage rates or unemployment levels of less-skilled non-Cuban workers, including blacks and other Hispanics. Effects on previously resident Cuban workers were also very small. Instead, Card’s study finds that the Miami labor force was able to rapidly absorb the flood of predominantly low-skilled, Cuban immigrants. He speculates that two factors may help account for the rapid absorption of the Mariel immigrants: presence of industries in Miami, such as garments, able to expand and employ unskilled migrants, and a compensatory reduction in the rate of migration into Miami from the rest of the United States following the influx of the Mariel immigrants.

Following Card, many other instances of sudden migration surges have been analyzed and yielded similar results (see Hunt 1992; De Lima & Carrington 1996) though some found somewhat greater effects on native wages, in the range of 1-2%.

In a contribution that exemplifies the broader economy-wide approach of the most recent literature, Peri and Ottaviano (2008) refute Borjas’ findings that the effects of migration on the wages of unskilled workers are large. Like Borjas, they examine the effects of migration of different age/education cohorts and their impact on the national labor market in the United States, but, applying a standard production function model, they find instead that immigration over 1990-2006 had small negative effects in the short run on native workers with no high school degree (-0.7 percent) and on average wages (-0.4 percent), while it had small positive effects on native workers with no high school degree (+0.3 percent) and on average native wages (+0.6 percent) in the long run. They also find a wage effect of new migrants on previous migrants on the order of negative 6 percent. Ottaviano and Peri’s conclusions differ so markedly from Borjas’ mainly because they account for the increase of investment over time and its employment-expanding effects, and they account more precisely for the imperfect substitution between migrants and natives. An important benefit that natives – whether they are skilled or unskilled - derive from unskilled migrants, and one that until recently was largely neglected, is that they help reduce the prices of non-traded goods and services that natives use intensively, such as home care, food preparation, gardening, and construction. Cortes (2008) finds that the surge in immigration in the Unites States during 1980-2000 may have reduced the prices of these services by about 10 percent.

Effects on Employment

Many advanced countries, most notably in Europe, experience high and persistent unemployment over many years, even when growth is near or above long-term potential. Especially at a time of prolonged high unemployment like today, the worry that increased immigration will simply make the structural or cyclical unemployment problem worse resonates widely.

Brucker and Jahn (2011) consider an economy where collective bargaining predominates in some “rigid” sectors and where the labor market is very flexible in others. They find that migration can increase unemployment in the rigid sector. The wage in the rigid sectors is determined in a standard wage bargaining model, where unions set the wage in function of the level of unemployment and firms are free to determine employment in function of the prevailing wage. Increased migration with the same education and experience distribution as previous migrants will then contribute to raise unemployment in the rigid sectors (as some of the migrants will demand to work in those sectors) while having a relatively small negative effect on wages in the rigid sector, and reduce the wage in the flexible sector where unemployment remains low by definition. In the long run, these effects will dissipate as investment responds in a way similar to the perfect labor market case. Thus, the model which assumes perfect labor markets tends to overestimate the impact of migration on average wages and to underestimate the effect on unemployment. Brücker (2011) conducted a meta-analysis of studies examining the effect of immigration on unemployment in the U.S. and Europe and found that in general an increase in immigration by 1 percent of the population leads to an increase in unemployment of no more than 0.3 percent.

In contrast, studies of the U.S. labor market, which is among the most flexible, have found no significant effect of immigration on employment opportunities for native workers (Peri 2009), including low-skilled native workers, between 1990 and 2007. Peri and Sparber (2008) report that migrants induce task specialization of natives. They found that, among less-educated workers, those born in the United States tend to have jobs in manufacturing or mining, while migrants tend to have jobs in personal services and agriculture, providing an explanation for why low-skilled migration has a limited impact on employment. In fact, the share of migrants among the less-educated is strongly correlated with the extent of U.S.-born worker specialization in communication tasks. In states with a heavy concentration of less-educated migrants, U.S.-born workers have shifted toward more communication-intensive occupations. Those jobs pay higher wages than manual jobs, and so such a mechanism has stimulated the productivity of workers born in the United States and generated new employment opportunities./

Long-Term Fiscal Impact of Immigration

A commonly held belief is that unskilled immigrants come to “live off” the welfare state in advanced countries. It is true that recently arrived voluntary immigrants generally have a less favorable net fiscal profile than natives, and forced migrants have an even less favorable net fiscal profile as the previous sections suggest. However, this is mainly because they contribute less to taxes and social security than natives, and their labor force participation rates are lower, not because they use social services more intensively. Consequently, the fiscal profile of migrants improves over time if they are allowed to, and find, work, and most studies find that immigration has a minimal fiscal impact on developed countries in the long run.

Three approaches are employed in the literature to measure the fiscal impact of immigration: the accounting approach, which examines the fiscal impact of resident immigrants in any given year; dynamic models, which examine the fiscal impact of additional migration on future public budget balances; and macroeconomic models, which examine the fiscal implications that flow from the overall impact of immigration on the economy. Assumptions regarding the extent to which immigrants should bear the cost of public goods such as defense, infrastructure, and administration significantly affect findings.  Moreover, state and local governments tend to carry the burden of providing public services and welfare support, and can see much larger effects in instances of high immigrant concentration.

A recent OECD Migration Outlook (2013) provides an overview of the literature on the fiscal impact of immigration on OECD countries. Most studies indicate a small fiscal impact (less than +/-1 percent of GDP), subject to methodological approaches and key assumptions. In many European OECD countries, raising immigrants’ employment rate relative to natives would entail significant fiscal gains; in Belgium, France, and Sweden, this would have a budget impact of more than 0.5 percent of GDP. Factors which affect the fiscal impact include characteristics of the immigrant population such as age and reason for migration. The available literature suggests that young unskilled migrants who come to work, and who are, moreover, employed, can be expected to be net fiscal contributors, with the turning point being between the age of 40 and 45. Labor migrants generally have a more positive fiscal impact on their host countries than migrants who emigrate for family or humanitarian reasons.

Employment is the most important factor affecting the net fiscal profile of immigrants, especially in countries with generous welfare states. Raising migrants’ employment rate to those of natives would boost public budget balances significantly, in some countries by as much as 0.5 percent of GDP. Still, as the aforementioned OECD study affirms, “in the long run for most countries, the overall conclusion [is that immigration] is neither a major burden nor a major panacea for the public purse.” (OECD 2013, p.161). From the fiscal perspective, the big difference between immigrants and natives is not in benefits received, but rather in taxes paid. In part because immigrants on average have less education, at each age they earn less and pay substantially lower taxes of all kinds and to all levels of government.

This section has reviewed the evidence on the impact of immigration on growth of investment and output advanced host economies, the wages and employment of unskilled natives, and on public finances, providing a fairly reassuring picture. Yet, this only conveys part of the story. Broad studies conducted using computable general equilibrium models (e.g. World Bank 2006) suggest that not only do the natives of advanced countries derive modest net gains from immigration (with native skilled workers and capital owners clear net gainers, while native unskilled workers are small net losers), but also migrants see a very large increase in real incomes and the country of origin can also derive large benefits from increased remittances. The next section examines how these conclusions hold up in the case of South-South migration.

Settlement in the South

As already mentioned, according to the United Nations, the stock of South-South migrants – comprising both voluntary and forced migrants - is slightly larger than that of South-North migrants. However, the latter has been growing more rapidly, more than doubling since 1990, while the former grew by over 40% (UNDESA Population Division 2013). Chart 3 shows that the major South-South migration corridors are among countries in Asia and among countries in Africa. Thus, 53.8 million migrants have moved from one country of Asia to another country in Asia, and 15.3 million migrants have moved within Africa. The overwhelming majority of migrants from the South either stay in their region of origin or head to Europe or to North-America. Very few migrants from the South head to a developing region outside their own.

Chart 3: International Migrant Population by Major Area of Origin and Destination, UNDESA Population Division, 2013, in millions

 Migration among countries of the South appears to depend greatly on proximity: 80 percent of South-South migration occurs between countries with contiguous borders. These migration corridors are less restricted, entail less cost, pose less risks, and appear to build on networks of families and friends even more than South-North migration, which allows even migrants with limited resources and without travel documents to move across borders (Ratha & Shaw 2007, p.19). Due to these low financial, social and cultural costs of migration, differences in country income appear to play a lesser role in South-South migration than in South-North migration: remarkably “About 38 percent of identified South-South migrants come from countries with higher incomes than their host country”(Ratha & Shaw 2007, p.18). Individual opportunities may of course be greater than these averages indicate, since, even if they are moving to a lower average income location, migrants are still responding to incentives for increased earnings or immediate economic survival needs.

Chart 4: Major refugee-hosting countries end 2014, UNHCR Global Trends 2014

From an intra-household decision perspective, these movements help families diversify their sources of income thereby reducing the risk of a complete loss of income in case of crisis in their home community (Lucas 2005, World Bank 2005). Migration within low- and middle-income countries is also often used for more temporary goals such as seasonal work or as a transit stop on the way to a high income country (Campillo-Carrete 2013; UNDESA, OECD 2013). The different patterns of motivation translate into a different profile of migrant population who is more likely to be poor and less educated than their counterparts who make it to the North. Data is inadequate, but there is a consensus in the existing literature that migration within and between developing countries involves an even higher share of unskilled migrants(Hujo & Piper 2007) than South-North migration. OECD estimates suggest that “in the North, 24 per cent of all migrants have completed tertiary education, while only 15 per cent of migrants in the South have reached this level of education” (IOM 2013, p.68).

The flow of refugees is far more skewed towards the South-South corridor than the flow of migrants. As Table 2 shows, close to 8 million refugees are from the South and find abode in the South, while only 1.8 million refugees find abode in the North. Moreover, refugees represent just 2% of the total flow of South-North migrants, but they represent 11% of the flow of South-South migrants.

Table 2: Number of refugees, share of global refugee stock and share of migrant stock, 2010

Refugees by Migration Pathway, UNHCR Data, 2010
	South-North	North-North	South-South	North-South
Stock (in thousands)	1,756	19	7,939	61
% of global refugee stock	18.0%	0.0%	81.0%	1.0%
% of total migrants (in each pathway)	2.0%	0.0%	11.0%	1.0%

When the migration is forced, the selection of migrants that remain in the South also appears to be skewed towards the most vulnerable segment of the population, those with small or no assets, low education levels, children and young mothers (Verme 2015a; Calderón-Mejía & Ibáñez 2009). This pattern fits with the OECD’s findings on the current inflow of Syrian refugees to Europe with a comparatively high level of education and assets. As already mentioned, only the ones who are the most skilled and have assets to start with can even make it on the arduous journey from their first displacement in Syria to a country in Europe (OECD 2015).

The characteristics of migrants that settle in the South combined with the limited hosting capacity of countries result in a rather different picture of economic impact as compared to migration to the North. The most obvious difference lies in the smaller income gap between the country of origin and the country destination in the case of South-South migration. The South-South migrant moves from a low-productivity environment to another low- or even lower- productivity environment, so the resulting gains for the migrant, and the country of origin (through remittances) are likely to be far smaller than South-North migration. While individual voluntary migrants may, nevertheless, be making significant income gains from moving that are concealed by these averages (just as there are migrants from the North that move to the South to exploit specific opportunities), migrants that are forced to move from one country or region of the South to another may not improve their income or may lose outright.

To estimate the differential impact between South-South to South-North migration, Ratha and Shaw apply an appropriately calibrated global computable general equilibrium model and perform a simulation of an increase in migration to compare movement of 14.2 million workers (equal to 3% of the labor force of advanced countries) from developing countries to advanced countries, and from developing countries to developing countries. The results of the simulations find just a 60 percent wage increase for South-South migration compared to an over 2300 percent increase in wages for South-North migration. The gain for migrants moving from low- to high- income countries in this simulation appears to be inordinately large but may be a reflection of the income gap that persists between the countries of origin with large populations, many of which are very poor, and the advanced countries⁴. Moreover, the authors caveat that this comparatively large real wage increase in high-income countries comes with higher upfront migration costs, and higher taxes. While these factors reduce the overall welfare gains from South-North migration, it is fair to assume that the welfare gain differential would remain extremely large even after adjustment.

Confirming these model projections, the comparatively lower increase in wages for the South-South migrant results in smaller remittances to the country of origin. UN data suggests that in 2010, North-South remittance flows came to $267 billion versus $55 billion in South-South remittances. [Note that these flows include remittances originating in countries such as Russia, the UAE, and Saudi Arabia, which the World Bank has recently classified as high-income; to be checked]. So, while South-South migrants make up almost half of international migration stock, they only remit a less than a quarter of all remittances. Once again the data on remittances flowing to and from low and middle income countries is not very reliable, so an even larger amount of remittances is likely flowing unrecorded through formal and informal channels⁵ than is the case for North-South remittances. No data is available that allows identification of remittances from forced migrants to their country of origin; one can speculate that if, as is often the case, whole families eventually leave (even if the able-bodied leave first), remittances by forced migrants are relatively small and short-lived, and also flow predominantly through informal channels⁶.

The lower wage differential for migrants moving from the South to other countries in the South compared to those moving to the North signals higher competition and less complementarity with native workers. As set out above, South-South migration flows are made up of a very high share of low-skilled, uneducated workers, and that proportion is even higher among forced migrants, and the labor force of developing countries is also predominantly unskilled. Evidence on Syrian migration to Turkey and on regionally displaced Colombians, for example, confirms this (Del Carpio 2015, Ceritoglu 2015, Calderon Ibanez 2009). The increased competition at the low-skill end of the income distribution leads to increased downward pressure on already low wages, with significant and adverse distributional impacts (Maystadt and Verwimp 2009). These trends can be aggravated by the need to find jobs for the rapidly growing young population in poor countries⁷.

Beyond the immediate labor market factors, as already discussed in previous sections, the host countries in the South also have lower capacity to provide for social and public services and to handle additional population pressures: “Western industrialized countries with well-established institutions, powerful economies and stable societies are actually not comparable to developing countries with severe economic, political and bureaucratic deficiencies.” (Czaika 2005, 103) Developing countries also have low capacity to manage the social tensions resulting from large flows of migrants: “policies to manage immigration are lacking while control of the same is failing to contain the inflow of migrants due to scarce resources, weak administrative capacity, and porous borders.” (Ratha, Mohapatra, Scheja 2011, 14; comp. Adepoju 2005; Crush 2000).

In cases where large amounts of humanitarian aid flow into a very poor region hosting refugees, the effects on the native population or at least on the land and capital owners can be positive. A study of the Dadaab camp in Kenya for example finds annual benefits for the host community of “USD 82 million through increased trading and business opportunities, camp-related employment, improved infrastructure” (Zetter 2014, p. 5). These positive effects emanating from humanitarian aid flows will only be sustainable, however, if the region receiving the aid develops a viable source of earnings from exportable products or services or from labor income from a nearby town. 

Returning to the example of Welcomia in the previous section, the country that receives large numbers of migrants and allows them to work, in principle Welcomia could be either a developing or advanced country, and in either case one could see output and investment expand with the arrival of refugees. Indeed, in the many instances where the inflow of refugees is small relative to the size of the population, and the investment climate is favorable, a developing country might derive significant benefits in terms of increased output and investment, even if they might need help to cover the fiscal cost of initially sustaining the refugees. However, these conclusions do not apply in instances where the inflow is very large and causes such large frictions that the investment climate – which may already be fragile to start with – deteriorates. In those instances, while supporting low-capacity hosting countries with aid and technical assistance helps, it is unlikely to represent a durable solution to the forced migration challenge.

Coordination Problem

The rising number of forced migrants and their concentration in a small number of relatively low-income regions creates daunting problems and large costs for those regions, great suffering for the migrants, and a humanitarian challenge of the first order for the international community. But that is only a part of the story. The concentration of “unabsorbed” forced migrants also creates large negative effects on the surrounding region and on the rest of the world. These negative externalities include the risk of political instability and alienation and radicalization of large groups, worsening of fiscal and balance of payment imbalances in host countries which can result in economic shocks and inability to repay debt, increased threat of infectious disease, and large and disorderly movements of refugees to countries where economic opportunities exist. The presence of these externalities as well as the obvious limits in the capacity of countries of first arrival to cope create the need for an internationally coordinated response.

Over and beyond the humanitarian response to the crisis, which will or should involve all countries, as well as civil society and individuals willing and able to help, dealing with the economic and political externalities of the crisis requires spreading its costs in function of countries’ capacity to pay or to carry the burden by hosting the migrants, on the one hand, and the negative impact on them if the crisis is not adequately dealt and mitigated on the other. A number of schemes have been proposed in this regard, and they all envision some combination of hosting forced migrants and covering the costs of other countries that host them but are unable to carry the full cost.

Though some authors have tried (Czaika 2005, 2009; Dennis 1993; Hoertz & GTZ 1995; Rapoport 2014; Schuck 2010), it is difficult, if not impossible, to envision a scheme that allocates these costs “rationally” and precisely, since while the direct costs of hosting refugees may be quantified with some degree of precision (the cost of shelter, food, infrastructure, etc.), the indirect costs (e.g the effect on the social and political fabric) cannot be. As argued above, in advanced countries there can be significant benefits, as well as costs, for the host population from absorbing refugees, and these will vary in function of the investment climate and demographic trends. And the benefits accruing to individual countries from resolving or mitigating the crisis are even more difficult to quantify: what is the value of avoiding radicalization and who gains the most? Moreover, schemes that attempt to “allocate” refugees across countries according to simple formulas such as the share in the population run into a fundamental obstacle, which is that refugees are not objects but people, and they naturally have their own views as to where they will be welcomed and where they want to live and work.

Nevertheless, even the most cursory review of the distribution of forced migrants suggests that the burden is currently distributed extremely unequally, and that a few relatively poor countries are carrying an entirely disproportionate share of the burden. Constructing an index of the forced migration burden that depends on the size of the forced migrant stock, and indicators such as per capita income, and institutional strength in 2003 Czaika found no high-income country among the top 20 most burdened (Czaika 2005, 116).

The preceding discussion suggests some general principles that can help improve the response of the international community.

• Any durable solution to the refugee crisis must provide jobs for the refugees; any other scheme – whether envisaging increased humanitarian aid in the country of first arrival or return to country of origin or transfer to a third country – does not represent a durable solution unless it provides for jobs.
• The developing country of first arrival must prepare for an extended period of adjustment to large flows of forced migrants, and working with international donors, should formulate a development plan capable of absorbing a vastly expanded labor force. Even though it cannot and should not carry the full cost of hosting refugees, the country of first arrival has the greatest interest in helping resolve the crisis on account of its geographic proximity to the crisis.
• Countries which have a strong investment climate, more flexible labor markets, solid macroeconomic fundamentals, and aging populations, need immigration and are especially capable to absorb refugees. Advanced countries and developing countries that fit these characteristics should allow a much larger number of forced migrants to enter and settle.
• A large share of the costs of housing refugees in developing countries of first arrival should be borne by advanced countries. Advanced countries with large current account surpluses and large net foreign assets and that are geographically closest to the crisis are especially well positioned to do more. Countries of first arrival which are host to large numbers of migrants and with large fiscal and balance of payments deficits need the most help.
• The International Community must ensure that the funding of humanitarian and development agencies is adequate for the task of relieving the suffering of refugees and helping countries absorb them. Care must be taken to avoid situations of extended dependency.

Long experience with attempts to deal with international externalities in other areas, such as climate change, customs and trade facilitation, and financial regulation suggests that hard, enforceable targets, are unlikely to be agreed by sovereign nations, or to be applied even if they are agreed. As shown most recently by the Paris Agreement on Climate Change, it is nevertheless possible to make progress with a combination of hard commitments on principles and more flexible, voluntary, arrangements on specific targets. It is, for example, possible to envisage the adoption of general principles on how to deal with refugees which go beyond the 1951 convention, and also to undertake firm commitments to transparency and monitoring of policies regarding refugees. Agreement on this general framework can be combined with voluntary targets on absorption of refugees and on aid flows, and on their periodic revision in light of the evolving situation.

Conclusion

The current refugee crisis is a catastrophe affecting millions of families, endangering the stability of nations that are hosts to large numbers of migrants, and of the region around them. There is little likelihood at present that large numbers of forced migrants will return to their homes. There is no magic bullet. Managing the crisis and mitigating its worst effects requires courage, leadership and political will. A cold-eyed economic assessment such as attempted in this chapter can only help clarify the choices facing political action, not substitute for the willingness to undertake it.

Drawing on the literature on voluntary and forced migration as well on the recent field experience of international agencies, the chapter has identified the centrality of finding jobs for forced migrants, and has highlighted the difficulty of doing this in poor, low-capacity environments, and where public services and the political fabric are fragile to start with. At the same time, the chapter has shown that concerns that accepting forced migrants in advanced countries will cause job losses or falling wages, and place an undue burden on the public purse are largely unjustified, and that the global gains from forced migrants settling in the South are much smaller than migrants settling in the North, and may result in net welfare losses in some instances.

Forced migration flows which are mismanaged, as at present, create large negative political and economic externalities for the world as a whole. There is no perfect scheme for allocating the burden, but, absent political solutions to conflicts, any scheme must envisage increased numbers of refugees settling in the North and increased development aid for the countries in the South with the largest numbers of refugees. Such a scheme is more likely to materialize if it is based on the adoption of basic principles as well as voluntary – rather than compulsory –targets to welcome refugees, combined with firm commitments to transparency and monitoring.

This paper benefited greatly from comments received at two World Bank seminars on a forthcoming flagship report on Forced Migration held in October 2015 and January 2016, and also from numerous comments and discussion with Xavier Devictor, lead advisor in the conflict unit at the World Bank.

Bibliography

Alcobendas, M.A. & Rodriguez-Planas, N., 2009. Immigrants’ Assimilation Process in a Segmented Labor Market. , (4394), p.77. Available at: http://pareto.uab.es/wp/2010/82210.pdf\nhttp://search.ebscohost.com/login.aspx?direct=true&db=ecn&AN=1102872&site=ehost-live.

Aljuni, S. & Kawar, M., 2014. The Impact of the Syrian Refugee Crisis on the Labour Market Jordan: A Preliminary Analysis,

Borjas, G.J., 2003. the Labor Demand Curve Is Downward Sloping; Quarterly Journal of Economics

Borjas, G.J. & Katz, L.F., 2005. The Evolution of the Mexican-born Workforce in the United States,

Brücker, H., 2011. The Labor Market Challenge. Does International Migration Challenge Labor Markets in Host Countries ? A Critical Review of the Recent and Traditional Literature,

Brücker, H. & Jahn, E.J., 2011. Migration and Wage-setting: Reassessing the Labor Market Effects of Migration*. Scandinavian Journal of Economics, 113(2), pp.286–317. Available at: http://doi.wiley.com/10.1111/j.1467-9442.2010.01634.x.

Butcher, K.F. & Card, D., 1991. Immigration and Wages: Evidence from the 1980’s. The American Economic Review, 81(2), pp.292–296.

Calderón-Mejía, V. & Ibáñez, A.M., 2009. Labor Market Effects of Migration-Related Supply Shocks: Evidence from Internally Displaced Populations in Colombia,

Calderón-Mejía, V. & Ibáñez, A.M., 2015. Labour market effects of migration-related supply shocks: evidence from internal refugees in Colombia. Journal of Economic Geography, September.

Campillo-Carrete, B., 2013. South-South Migration : A review of the literature. , (May), pp.1–89.

Card, D., 2005. Is the new immigration really so bad?,

Card, D., 1990. The Impact of the Mariel Boatlift on the Miami Labor Market. Industrial and Labor Relations Review, 43(2), pp.245–257.

Card, D. & Altonji, J.G., 1991. The Effects of Immigration on ther Labor Market Outcomes of Less-skilled Natives. In J. M. Abowd & R. B. Freeman, eds. Immigration, Trade and the Labor Market. Univesity of Chicago Press, pp. 201–234. Available at: http://www.nber.org/chapters/c11773.

Card, D., Dustmann, C. & Preston, I., 2012. Immigration, wages, and compositional amenities. Journal of the European Economic Association, 10(1), pp.78–119.

Card, D. & Lewis, E.G., The Diffusion of Mexican Immigrations During the 1990s: Explanations and Impacts,

Cohen, S. & Hsieh, C.-T., 2000. Macroeconomic and Labor Market Impact of Russian Immigration. Working Papers, (October). Available at: http://ideas.repec.org/p/biu/wpaper/2001-11.html.

Cortes, K.E., 2004. Are Refugees Different from Economic Immigrants? Some Empirical Evidence on the Heterogeneity of Immigrant Groups in the United States. Review of Economics and Statistics, 86(2), pp.465–480.

Cortes, P., 2008. The Effect of Low-Skilled Immigration on U.S. Prices: Evidence from CPI Data. Journal of Political Economy, 116(3), pp.381–422.

D’Amuri, F. & Peri, G., 2011. Immigration, Jobs and Employment Protection: Evidence from Europe,

Friedberg, R.M., 2001. The Impact of Mass Migration on the Labor Market. The Quarterly Journal of Economics, 116(4), pp.1373–1408. Available at: http://www.jstor.org/stable/2696462\nhttp://www.jstor.org/stable/pdfplus/2696462.pdf?acceptTC=true.

Hercowitz, Z. & Yashiv, E., 2000. A Macroeconomic Experiment in Mass Immigration,

Hujo, K. & Piper, N., 2007. South–South Migration: Challenges for development and social policy. Development, 50(4), pp.19–25.

Hunt, J., 1992. The Impact of the 1962 Repatriates from Algeria on the French Labor Market. Industrial and Labor Relations Review, 45(3), pp.556–572.

Ianchovichina, E. & Ivanic, M., 2014. Economic Effects of the Syrian War and the Spread of the Islamic State on the Levant, Available at: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2535615.

IOM, 2013. World Migration Report 2013. Migrant Well-Being and Development,

LaLonde, R.J. & Topel, R.H., 1991. Labor Market Adjustments to Increased Immigration. In J. M. Abowd & R. B. Freeman, eds. Immigration, Trade and the Labor Market. Chicago: University of Chicago Press, pp. 167–200.

De Lima, P.J.F. & Carrington, W.J., 1996. The Impact of 1970s Repatriates from Africa on the Portuguese Labor Market. Industrial and Labor Relations Review, 49(2).

Manacorda, M., Manning, A. & Wadsworth, J., 2006. The Impact of Immigration on the Structure of Male Wages and Employment: Theory and Evidence from Britain, Available at: eprints.lse.ac.uk/.../The_Impact_of_Immigration_on_the_Structure_of_Male_Wages_Theory_and_Evidence_from_Britain.pdf.

Masri, S. & Srour, I., 2014. Assessment of the Impact of Syrian Refugees in Lebanon and Their Employment Profile,

OECD, 2013. International Migration Outlook 2013, Available at: http://www.oecd-ilibrary.org/social-issues-migration-health/international-migration-outlook-2013/chile_migr_outlook-2013-13-en.

OECD, 2015a. International Migration Outlook 2015,

OECD, 2015b. Is this humanitarian migration crisis different ? Migration Policy Debates, (N°7), pp.1–15. Available at: http://www.oecd.org/migration/Is-this-refugee-crisis-different.pdf.

OECD, 2015c. OECD Economic Outlook 2015, Available at: http://discovery.ucl.ac.uk/93372/.

Orefice, G., 2010. Skilled Migration and Economic Performances: Evidence from OECD Countries. Journal of Economic Literature.

Ortega, F. & Peri, G., 2009. The Causes and Effects of International Migrations: Evidence From OECD Countries 1980-2005. NBER Working Paper, (14833).

Ottaviano, G.I.P. & Peri, G., 2008. Immigration and national wages: clarifying the theory and the empirics. NBER Working Paper, p.68. Available at: http://www.nber.org/papers/w14188.

Peri, G. & Sparber, C., 2008. Task Specialization, Immigration, and Wages,

Ratha, D. & Shaw, W., 2007. South-south migration and remittances,

Schoeni, R.F., 1997. The Effect Of Immigrants On The Employment And Wages Of Native Workers: Evidence From The 1970s And 1980s,

Solow, R.M., 1956. A contribution to the theory of economic growth. The Quarterly Journal of Economics, 70(1), pp.65–94.

UNDESA Population Division, 2013. International Migration 2013: Migrants by origin and destination,

UNHCR, 2015. UNHCR Global Trends. Forced Displacement in 2014.,

Verme, P., 2015a. How Poor Are Refugees? A Welfare Assessment of Syrians Living in Jordan and Lebanon.

Verme, P., 2015b. How Poor Are Refugees? Presentation.

World Bank, 2013. Lebanon. Economic and Social Assessment.

World Bank MFM GP, 2015. Briefing for President Jim Kim -Macroeconomic and Fiscal Impacts of the Refugee Crisis.

Notes

¹ These studies are often referred to as Card studies (after David Card,1990; see also section on settlement in advanced countries, below). Econometric studies of the effects of migration often resort to studies of mass forced migration to help get around the problem of causality, and of identifying the independent effect of immigration on the host economy more precisely. It is easy, for example, to mistake migration as the cause of growth of the host economy rather than the result of growth [Borjas,2003].

² Increased consumption can also lead to increasing prices for essential goods, effectively worsening the situation for both displaced and hosts. Alix-Garcia and Saah, for example, show negative real income effects on households in refugee hosting areas of western Tanzania due to  “large increase in the prices of non-aid food items and more modest price effects for aid-related food items” (2010, p. 148).

³ As in Ratha and Shaw, here we classify countries of the South as those the World Bank classifies as low- and middle-income countries; the countries of the North are the high-income countries,

⁴ The results from this general equilibrium exercise need to be taken with a grain of salt because as the authors point out since for one “the parameters driving the results (such as elasticity of wages with respect to changes in the supply of workers) have only a limited empirical foundation” and, second, “the model cannot account for many issues that affect the wage gains from migration (such as the legal status of migrants, effects of migration on workers’ productivity, the potential impact on investment of a greater supply of workers, and adjustment costs faced by migrants)” (Ratha and Shaw 2007, 25). Despite these difficulties, these simulations give us the best available estimate of the relative economic impact of South-South versus South-North migration.

⁵ Ratha and Shaw estimate that the true size of these flows, is believed to be at least 50 percent larger (Ratha and Shaw 2007).

⁶ It is also important to note that even if flows between countries in the South are smaller, they are nevertheless important for development especially in the least developed countries who received two thirds of their remittances from other countries in the South (UNCTAD, 2012). This can be explained by the fact that “migrants from LDCs mainly move to other developing countries and only one out of four migrates to a developed country.” (IOM 2013, 71)

⁷ Some studies, however, find complex distributional effects. Displacement of households in Northern Uganda “… had a positive effect on the value of assets in households that originally had little or no assets but reduced the value of assets of all other households by between 17% and 26%” (Fiala 2009, p. 1). While circumstances of displacement will determine how distributional effects play out, it remains true that the movement will invariably cause ‘winners’ and ‘losers’ (comp. Maystadt & Verwimp 2009; Alix-Garcia & Saah 2010).

The report was originally published by the OCP Policy Center.