Migration, Development and Conflict: State-of-the-Art Overview
Ninna Nyberg S0rensen, Nicholas Van Hear, and Poul Engberg-Pedersen
ABSTRACT
This paper provides a state of the art overview. of current thinking and available evidence on the relations between migration, development and conflict, including the role of aid in migrant- and refugee-producing areas. It offers evidence and conclusions related to four critical issues:
Poverty and migration. People in developing countries require resources and connections to engage in international migration. There is no direct link between poverty, economic development, population growth, social and political change on one hand, and international migration on the other. Therefore poverty reduction is not in itself a migration-reducing strategy.
Conflicts, refugees, and migration. Violent conflicts produce displaced persons, migrants and refugees. People on the move may contribute both to conflict prevention and reconciliation and to sustained conflict. Most refugees do not have the resources to move beyond neighbouring areas: they remain internally displaced or move across borders to first countries of asylum within their region. Aid to developing countries receiving large inflows of refugees is poverty-oriented to the extent that these are poor countries, but it is uncertain what effect such aid has in terms of reducing the number of people seeking asylum in developed countries. Furthermore, aid to neighbouring countries may attract refugees from countries in war and crisis.
Migrants as a development resource. International liberalization has gone far with respect to movement of capital, goods and services, but not to labour mobility. Current international institutions provide little space or initiatives for negotiations on labour mobility and the flow of remittances. There is a
6
Nyberg SQrensen, Van Hear, and Engberg-Pedersen
pressing need to reinforce the view of migrants as a development resource. Remittances are double the size of aid and at least as well targeted at the poor. Migrant diasporas are engaged in transnational practices with direct effects on aid and development; developed countries recognize their dependence on immigrant labour; and policies on development aid, humanitarian relief, migration, and refugee protection are often internally inconsistent and sometimes mutually contradictory.
Aid and migration. Aid policies face a critical challenge to balance a focus on poverty reduction with mitigating the conditions that produce refugees, while also interacting constructively with migrant diasporas and their transnational practices. The current emphasis on aid selectivity tends to allocate development aid to the well performing countries and humanitarian assistance to the crisis countries and trouble spots. However, development aid is more effective than humanitarian assistance in preventing violent conflicts, promoting reconciliation and democratization, and encouraging poverty-reducing development investments by migrant diasporas.
The paper synthesizes current knowledge of migration-development-conflict dynamics, including an assessment of the intended and unintended consequences of development and humanitarian policy interventions. The first section examines whether recent developments in the sphere of international migration provide evidence of a "crisis", as well as the connections between migration, globalization and the changing nature of conflict. Section two summarizes current thinking on the main issues at stake in the migration-development nexus. Section three examines available evidence on the relations between migration and development. Section four discusses the consequent challenges to the aid community, including the current debates about coherence and selectivity in aid and relief. The final section elaborates on the four conclusions of this summary.
THE SIGNIFICANCE OF MIGRATION
Throughout history, migration has been intimately related to economic and social development: it is often seen as the result of imbalances in development, but also as influencing development. Assessments of the influence of migration on development have varied over time: sometimes migration has been seen as beneficial and at others detrimental to development, depending on the historical moment and circumstances. With the variation in perspectives has come variation in migration and development policies. In the latter quarter of the twentieth century, the view in Europe shifted from seeing migration as a factor contributing to economic growth in the receiving states and to development in the sending states, to the prevailing view that immigration pressures have reached intolerable levels. More restrictive legislation has been accompanied by tendencies to confuse the status of refugees and illegal migrants and to lump together concerns about security with the problem'of asylum seekers. Development and
State-of-the-art overview
conflict prevention are seen as needed in the migrant-sending countries to curtail unwanted migration. However, while there has been much talk of improving economic and security conditions in source countries - assumed to alleviate migration pressures - so far the emphasis has been on policies aimed at curbing immigration at the destination end - a trend that has gained momentum in the wake of the events of 11 September 2001.
The prevailing sense of an "international migration crisis" has profoundly influenced the formation of policy. In the latter part of the 1990s, perceived immigration pressure ascended to the status of a worldwide security issue (Weiner, 1995) and to a priority policy concern in the European Community. On the initiative of the Netherlands, the Council of the European Union set up a High Level Working Group (HLWG) on Asylum and Migration in December 1998 charged with preparing action plans encompassing concerns about border controls, coordination of development aid and reallocation of aid to six migrant-producing countries and regions: Afghanistan and neighbouring regions, Morocco, Somalia, Sri Lanka, Iraq, and Albania and neighbouring regions.
The HLWG action plans contain proposals to coordinate action within three areas: foreign policy, development policy and migration/asylum policy. The basic instruments and components are: protection of human rights; support for democratization and measures for the promotion of constitutional governance; social and economic development; combating poverty; support for conflict prevention and reconciliation; cooperation with the Office of the UN High Commissioner for Refugees (UNHCR), the International Organization for Migration (IOM) and human rights organizations; respecting refugees' and asylum seekers' right to protection; and measures to combat illegal migration. Implementation of HLWG action plans has come up against certain difficulties, not least the perception among several of the six countries' negotiators that the security of developed countries (DCs) and not development in less developed countries (LDCs) is the major concern of the European Community.
This section gives some of the background to the policy arguments advanced. First, it explores the extent to which the perception of a "migration crisis" rests on plausible grounds. Then the changing dynamics of mass migration in the current era are explored, focusing on the effects of globalization, new forms of conflict and other imperatives to migrate.
Is there a crisis of mobility?
It has been estimated that some 150 million people currently live outside the country of their birth, a reflection of the acceleration of migration worldwide in recent decades. But at about 2.5 per cent of the world's population, this proportion is not that much different from parts of the last century, or indeed
8                      Nyberg Sflrensen, Van Hear, and Engberg-Pedersen
earlier eras when population movements peaked. The significance of changes during the post-colonial era lies not in the fact of global migration - which has existed for centuries - but rather in the great increase in the magnitude, density, velocity, and diversity of global connections, in the growing awareness of these global relationships, and in the growing recognition of the possibilities for activities that transcend state boundaries.
Defining international migrants as those who reside in countries other than those of their birth for more than one year, the number of such persons has doubled from 75 million in 1965 to an estimated 150 million in 2000 (IOM, 2000a). Of these about 80 to 97 million were migrant workers and members of their families (ILO, 2001),andbetween 12.1 million (UNHCR, 2001) and 14.5 million (USCR, 2001) were refugees. In addition to the refugees outside their countries of origin, there were some 20-25 million internally displaced persons forced to move within their states.
Zolberg (2001) has traced the evolution of alarmist popular social science commentary on migration, paralleled in more moderate form in the academic literature. In different ways, Kennedy (1993), Kaplan (1994) and Brimelow (1995) provide apocalyptic visions of a western world beset by massive migration pressures from "barbarous", "degenerating" regions of the developing world, coupled with overwrought anxieties about growing "imbalances" between the native population and other racial categories. In the mid-1990s such visions caught the imaginations of policy-makers, particularly in North America. The academic literature has developed in a similar way. Thus "crisis" is a much-used term in the context of migration, no less than in other arenas. The title of an influential book by Myron Weiner (1995), The global migration crisis, referred to what he and others see as a diffuse phenomenon widely felt and experienced throughout the world. However, careful scrutiny of today's migration reveals less a global migration crisis than a series of migration crises (often serious) around the globe (Van Hear 1998).
Among the factors contributing to an increase in the volume and velocity of migration in the last 50 years are the liberalization of exit, first from the post-colonial world (the "South"), as imperial restrictions on movement of colonial subjects fell away, and later from former communist countries (the "East"), after the collapse of communism. Increased possibilities of out-migration have been coupled with greater awareness of growing disparities in life chances between rich and poor countries, and the spread of violent conflicts often in the same poor regions (Zolberg, 2001). Nevertheless, the majority of refugees stay within their region in the developing world, or in the post communist world. The number of refugees has moreover fallen in recent years, from a peak of 17.6-18.2 million in 1992 (UNHCR, 1993; USCR, 2001) to 12-14.5 millionin2000 (UNHCR, 2001; USCR, 2001), although the number of internally displaced people has concomitantly risen, reflecting increasing pressure to contain forced migrants in their
State-of-the-art overview
9
countries or regions of origin, a trend partly a result of unwarranted anxiety about migration in western countries (Shacknove, 1993; Chimni, 1999). Looking at international migration more generally, the proportion of people living in countries other than those in which they were born has stayed more or less constant over the last three decades (Zlotnik, 1998). Thus while the current era has been presented as "The age of migration" (Castles and Miller 1993), the volume of migration has historical precedents - indeed the proportion of people on the move was probably greater in the decades straddling 1900 than it is 100 years later.
The changing dynamics of migration
Past and present migration may be seen as both a manifestation and a consequence of globalization. Globalization involves a number of related processes, among the most important being the steadily increased circulation of capital, production and goods; the global penetration of new technologies in the form of means of transport, communication and media; and the elaboration of regional, international or transnational political developments and alliances such as the European Union, the North American Free Trade Agreement (NAFTA), and numerous grassroots social organizations and movements. With globalization, mobility has increased and the chains of interaction have been lengthened and spread considerably. However, the movement of capital, goods and information has been liberalized to a much larger extent than the movement of people, whose mobility continues to be heavily regulated. While globalization and liberalization have involved freeing up international trade and capital flows, the international movement of labour, another essential factor of production, has if anything become more restricted (Rodrik, 2001). While national borders are being constantly criss-crossed by processes of communication and exchange, the actual bodily movement of people remains constrained (Smith and Guarnizo, 1998).
These processes have exacerbated imbalances among regions, countries and communities, giving further impetus to migration. A related consequence of globalization has been further differentiation of migrants in terms of ethnic and class backgrounds, as well as an increased feminization of migration.
The feminization of migration
This is seen in the emergence of new groups and types of migrants, including young single women or female family breadwinners who move independently rather than under the authority of older relatives and men. Female migration is on the increase within as well as from many parts of the developing world. Current migration to Europe is increasingly female, and typically male dominated migration streams towards Europe - for example, from Morocco - have throughout the 1990s changed towards including more and more autonomous female migrants.
O                  Nyberg S0rensen, Van Hear, and Engberg-Pedersen
Female migrants from LDCs differ in terms of background, including women from rural backgrounds migrating autonomously or through family reunification programmes, low-skilled women from urban backgrounds increasingly migrating autonomously because of divorce/repudiation and poverty, and women with secondary or higher education involved in autonomous migration because they could not obtain jobs in accordance with their qualifications at home. A fourth and increasing group is women fleeing civil unrest. While female migration may form part of an integrated family strategy, it may also take place within female networks, separate from those of men. There can thus be significant differences of opinion within the family, and wives and daughters may migrate as a consequence of the wishes of husbands and parents as well as despite such wishes.
Female migration is linked to new global economic transformations and the resulting restructuring of the labour force. In Europe many women find employment as domestic workers or the broader service sector. Some enter the sex industry, at times involuntarily through trafficking in prostitdtion networks. While some observers posit that female migrants in Europe resemble a slave labour force, existing on the margins and "fenced in" by society (Anthias and Lazaridis, 2000), others point to the relative autonomy of women even among sex workers (Lisborg, 2001). Despite such differences, most migrant women share the experience of deskilling. However, the sale of domestic services on the global market reveals that the tasks housewives usually perform for free in fact hold the potential for making significant contributions to both household finances and the national economy through remittances. Women, to a larger extent than men, are subject to social pressure to look after their relatives back home. Female migrants not only tend to be better remitters, they also tend to organize around important development issues of family welfare, health, schooling and the local environment. Upon return, women seem to have made some progress in the household and kinship sphere, in some instances leading to larger equity between partners in household decisions and reduced domestic violence.
Migration and development policies often ignore migrants' gendered identities and practices. When women are targeted as a special group, their transnational engagement in both sending and receiving societies is often overlooked. It is therefore important that policies are designed according to the opportunities and constraints specific to different groups (for example women and men, younger and older women, autonomous or dependent female migrants), as well as according to specific groups' transnational spheres of action. Migrants not only contribute remittances while abroad. They also contribute new skills and life views whether they return or not. Their abilities to do so depend on whether they have equitable access to services and training. International agencies should therefore approach migrants' gender-specific concerns and make sure to follow up effectively on gender awareness campaigns and programmes when women return. Unless properly assisted, women may lose newly gained gender rights to men, who seem to regain their traditional gender privileges upon return (Pessar, 2001).
State-of-the-art overview
The changing dynamics of conflict
While the factors impelling people to move to better their lives have changed in intensity rather than substance in recent years, changes in the nature of conflict since the demise of the Cold War have been accompanied by changes in the nature of displacement both within and among countries in the developing world and beyond (Duffield, 2001; Anderson, 1999; Kaldor, 2001; Collier, 2000; Keen, 1998; Reno, 1998; Richards, 1996; Gurr, 2000; UNHCR. 2001; Global IDP project, 1998; Schmeidl, 2001; McGregor, 1993). As new forms of conflict and upheaval have engulfed many parts of the world, the view is growing that the very nature of the political economy in such turbulent regions is an adaptation to globalization (Duffield, 2001). In a world where some countries and regions have ceased to hold the significance they had during the Cold War, conflict, and the migration associated with it, may be seen not as aberrations, but as novel forms of response.
There is a general consensus that, since the end of the Cold War, conflicts have taken new forms, in which civilians are seen as important components of warfare rather than simply incidental to it. Displacement has become a tactic or object of warfare rather than being an unintended outcome of it (UN High Commissioner for Refugees S. Ogata, cited in UNHCR, 1997). Such trends have rendered some types of intervention by the relief and aid community outmoded (Duffield, 2001). A recent analysis has put this shift in context succinctly:
The fact that the great majority of armed conflicts now are internal conflicts reflects major structural changes in global politics. Geopolitics is not what it used to be. In an era of "de-territorialisation" of economic activities, territorial gains are no longer as important to states.... Weapons capabilities are now such that war between major powers has become virtually impossible, while technological change has put arms in the hands of warlords and militias which previously only states could afford, thereby changing the political and security landscape (Pieterse, 1998:7).
Much has been made of the harm relief assistance can do in terms of exacerbating conflict and its consequences, including internal displacement and the flight of refugees (Anderson, 1999; Duffield, 2001). Similar conclusions are reached by those looking at development assistance. For example, a recent study concludes that
Rapid economic change in either ppsitive or negative direction involves redistribution of opportunity, status, and deprivation in ways that are often inconsistent with deeply held notions of what is fairand what is acceptable. Reciprocally, ethnic politics intrudes on the apparent technical rationality of development policy; rules are bent, locations skewed, privatisations distorted (Herring and Esman, 2001: 1, our emphasis).
Therefore, "Some [aid] interventions may be conducive to peaceful coexistence and equity; others may aggravate tensions and precipitate conflict" (Ibid: 3).
12                    Nyberg SQrensen, Van Hear, and Engberg-Pedersen
Thus not only does conflict generate displacement, but development itself as well as interventions designed to mitigate conflict may have similar effects. Some of the implications of this are discussed further below in the section on "Challenges to Aid".
MIGRATION AND DEVELOPMENT - IS THERE A LINK?
The current relationship between migration and development has rightly been characterized as an "unsettled" one (Papademetriou and Martin, 1991). Since the publication of the influential Ascencio report, a new consensus has arisen that rather than stemming or containing migration pressure, development can stimulate migration in the short term by raising people's expectations and by enhancing the resources that are needed to move (Ascencio, 1990; see also OECD, 1992; IOM, 1996). Some of the work known as the "new economics of migration" suggests that the demand for remittances from migrants, for example, increases as development proceeds and both investment opportunities and returns on investment increase: by enhancing development, remittances may therefore propel or perpetuate migration. Put another way, there is a "migration hump" that has to be overcome before people are encouraged to stay put by the development of their homelands and migration begins to decline (Martin, 1997; Martin and Taylor, 2001). Accompanying this view, models of migration based on economic forces such as pull and push factors have been supplemented by approaches recognizing mediating factors such as social networks, improved communication and transportation linkages, trade competition between countries, government migration policies, and violent conflicts within countries, yielding a more dynamic analysis of how migrations begin, how and why they stop or continue, and the extent to which migration can be controlled.
As migration has steadily climbed up the list of public and policy concerns, it has become increasingly recognized that migration can be affected - intentionally or not-by interventions in the kindred arenas of development policy and assistance, as well as by wider policies and practice in the foreign and domestic spheres. However, the precise links among these arenas of policy and practice - not least in terms of cause and effect - are imperfectly understood by analysts as much as by policy-makers.
This section first summarizes migration-development links in the literature on economically motivated migration. It then summarizes the links between development/relief and forced migration motivated by conflict, human rights abuse or other political dimensions. Subsequently arguments and evidence from the new literature on migrant diasporas and transnational migration are discussed. After a short discussion of the gap between migration policy outputs and outcomes, the challenges posed to the international aid system are outlined.
State-of-the-art overview
Migration-development links
Conventionally, international migration is understood to occur as a consequence of imbalances in development between sending and receiving societies. Tht most basic assumption is that if growth in material resources fails to keep up wit! demographic growth, strong migration pressures from LDCs to DCs will evolve In classical theory, migration occurs due to a combination of supply-push anc demand-pull factors. Diminishing migration pressures are thus dependent or eliminating levels of overpopulation and poverty in LDCs. Leaving aside tht question of whether there is any empirical basis for this assumption, the migration-development link is often understood to revolve around "the three R's' of Recruitment, Remittances and Return (Papademetriou and Martin, 1991).
Recruitment in a broad sense covers the conditions producing emigration. Such conditions include both migrant motivation (why people migrate) and facilitating factors/agents (what/who makes movement possible). Negative or low economic growth, population growth, high under- and unemployment rates, combined with unequal income distribution, and high pressures on land and urbar environments drive people to seek employment abroad due to a lack of alternatives back home. Poor governance is another major factor for emigration, especially among the highly skilled. Recruitment mechanisms range from individual to collective, from official to unofficial, and from government-led tc employment-led. There is no consensus on the optimal recruitment mechanism, but evidence suggests that worker recruitment eventually creates network linking particular rural or urban communities in the sending countries with specific labour markets in the receiving countries (Gamburd, 2000). When such networks are established, they become valuable assets for those who have access to them. Moreover, they represent the means by which migration becomes a self-perpetuating, semi-autonomous process.
Depending on their income in the migration destinations, migrants' contribution to local development in the sending countries can be significant. Remittances from migrants benefit local households in LDCs by sustaining daily living and debt repayment (Athukorala, 1993). Over time, remittances may be invested in consumer durables and better housing, education and the purchase of land or small businesses. At the national level, remittances contribute positively to the balance of payments by providing much-needed foreign exchange. The remittance-development link is highly debated (Massey et al., 1998; Taylor, 1999). Evidence suggests that remittances affect LDCs first by being spent on family maintenance and improvement of housing; in a second stage, spending tends to be on "conspicuous" consumption (often resulting in tensions, inflation and worsening of the position of the poorest); in a third phase, however, remittances are invested in productive activities, including improvement of land productivity. Analysis of the developmental impact of remittances needs to consider the initial
14                  Nyberg S0rensen, Van Hear, and Engberg-Pedersen
conditions under which people go abroad. Poor families obviously need more time than the better off to gain from migration (Gamburd, 2000).
Return is generally seen as the natural "end product" of the migration cycle. Ideally, migrants are expected to have saved capital and acquired skills abroad that can be productively invested in the sending country. Evidence nevertheless suggests that migrants, unless highly skilled, often do not acquire skills abroad that are useful at home. If skills are acquired, returning migrants often prefer to work in another, generally private, sector back home (Martin, 1991). Return is not necessarily promoted by home governments who may have a more direct interest in continuing flows of remittances than in incorporating returnees in the local labour market. Incentives to return have therefore primarily been initiated by receiving countries (Collinson, 1996). A study of Jamaican return migration suggests that return programmes attract only few migrants and generally only those who were planning to return in any event (Thomas-Hope, 2002). To the extent that highly skilled migration is determined by poor governance in the country of origin, return of skilled migrants can only be expected when local governance radically improves.
Assessing migration-development links through the three R's tends to reduce migration to an economic act and to view migrants in their role as labourers only. As this section has suggested, there are other dimensions - social, cultural and political - which also have to be taken into account. In the following section, the links between development and forced migration are discussed.
Refugee-development links
Conflict and human rights abuse associated with poor governance have become among the key factors impelling much current migration. It is no coincidence that conflict-ridden countries are often those with severe economic difficulties. Such combinations of motivations create difficulties in maintaining a clear distinction between voluntary and forced migration, as has been recognized for some time (Richmond, 1994). What begins as economic migration may transmute into internal displacement or international refugee movements, and conversely, what are originally refugee movements may over time develop into other forms of movement (Van Hear, 1998; Stepputat and S0rensen, 2001). When migrants from developing countries arrive in the developed world, refugees may live alongside co-nationals who are not necessarily refugees but rather part of broader communities of newcomers (Steen, 1993; Crisp, 1999); and refugees who remain within their region of origin may also enter prior currents of labour migration. Nevertheless, refugees are distinct from other kinds of migrant in international law.
Each stage of forced displacement has development implications. As with "economic" migration, refugee flight involves the loss of labour, skilled workers and
State-of-the-art overview
capital for the country of origin. Mass arrivals of asylum seekers - usually in countries neighbouring those from which refugees have fled, but also in more far-flung states - have short-term damaging effects, particularly in terms of strains on resources hosts must provide; however in the longer term the impacts of such mass arrivals may be more beneficial, particularly in terms of the economic, human and social capital newcomers bring with them (Van Hear, 1998).
Beyond flight and reception, these medium and long-term outcomes also have profound development implications. Conventionally there are three such outcomes, known as the three "durable solutions" (Chimni, 1999; Kibreab, 1999): repatriation; local integration, usually in the country of first asylum; and resettlement in a third country. The feasibility and attractiveness of these "durable solutions" have varied over time, partly determined by geo-political considerations: during the Cold War, resettlement or local integration were more the norm, because this suited the purposes of the West, while since the end of the Cold War, new imperatives have prevailed and repatriation has come to be seen as the most desirable durable solution (Chimni, 1999). Coupled with repatriation have been efforts to deter out-migration and to contain would-be migrants in their countries or regions of origin. Containment has been attempted by physically preventing people from leaving, or by emphasizing the "internal flight alternative", that is internal displacement. Countries in refugee-generating zones that have in the past been generous in hosting refugees are increasingly reluctant to continue to do so, taking their cue from western countries' restrictive polices and practices towards asylum seekers. The durable solution of "local integration", and the potential it offers for developing strife-torn regions, has therefore fallen by the wayside - at least this is the case as an official policy, while informal integration is widespread (Jacobsen, 2001).
Conventionally, displacement is represented as a temporary phenomenon, manifested in the form of temporary residence in refugee camps, more often in neighbouring countries than in countries further afield. Only if asylum becomes permanent may we speak of local integration or resettlement: the refugee may become an established resident, and eventually a citizen of the country of asylum. Temporary status is not supposed to last long: either the conditions that forced flight should be resolved and the displaced can go home, or the displaced should be incorporated permanently into their place of refuge. Such at least is the assumption of the "refugee regime" - the body of institutions, law, policy and practice, national and international that exists to deal with forced migration (Van Hear, 1998a).
Needless to say, the real world is messier than in this ideal scheme. "Resolution" of displacement often takes a long time, which the original architects of the refugee regime did not anticipate. The displaced often find themselves in a state of protracted limbo. Nationality or citizenship may not be easily acquired or reacquired, and is often disputed or problematic. People in such circumstances
j 6                  Nyberg S0rensen, Van Hear, and Engberg-Pedersen
develop ambiguous relationships towards the places in which they find themselves, and this may seriously constrain the influence they can have on the development of such places of residence.
Nevertheless the presence of refugees in the places of settlement does have impacts, during local integration in the first asylum country, during and after resettlement in a third country, and during and after repatriation to their country of origin (Harrell-Bond, 1986;Kuhlman, 1994;Kibreab, 1996;Black, 1998; Van Damme, 1999; Jacobsen, 1997; 2001; Bakewell, 2000). Among these effects, positive and negative, are:
-    changes in local markets for food, housing, land, transport, and other goods, services and resources;
-    changes in local labour markets;
-    changes in the local economy and society wrought by the introduction of humanitarian assistance;
-    demands on health care, education and other services;
-    demographic changes, and related influences on health, mortality and morbidity;
-    influences on infrastructure;
-    ecological and environmental changes.
There are also development implications for countries of origin. While refugee flight deprives their homelands of labour and skills, it also opens the possibility of remittances from refugees who manage to find employment sufficiently remunerative to allow surpluses to be sent home. The impact of remittances from both refugees and economic migrants is considered in more detail below.
Migrant diasporas and transnational practices
An important result of the extraordinary new focus on migration is a much greater awareness of the significance of migration, including the factors motivating migration, the factors attracting migrants to particular destination areas, the social networks linking areas of origin with areas of destination, and the improved communication and transportation networks enabling longdistance ties across geo-political divides. Over the past ten years, academic and other literature has stressed the importance of locating migration within transnational processes in terms of global economic connections and the formation of transnational migratory groups. The literature on transnational migration provides essential new insights into contemporary forms of migration and also raises general conceptual issues about ways of understanding migration in a global context.
Contrary to conventional migration theory's binary focus on the process of emigration from and immigration to 'particular nation states, transnational
State-of-the-art overview
17
approaches suggest that migration should be understood as social processes linking together countries of origin and destination. Contemporary migrants are designated "transmigrants" in as far as they develop and maintain multiple relations- familial, social, economic, political, organizational and religious ties -that span borders (Glick Schiller et al., 1992). Contrary to prevailing interpretations that portray migrant settlement as a process involving a break with home, transnational approaches suggest that the struggle for incorporation and adaptation in migrant destinations takes place within a framework of interests and obligations that result from migrants' simultaneous engagement in countries of origin and destination. Thus contemporary migration can only be understood by studying socio-economic, political and other relations spanning sending and receiving societies (Levitt, 2001).
But transnationalism is not limited to migrants' activities and networks. Migrants have become increasingly important, not only as a source of remittances, investments, and political contributions, but also as potential "ambassadors" or lobbyists in defence of national interests abroad. Many migrant-sending states recognize that although many migrants are unlikely to return, they can still advance state consolidation and national development from abroad (Levitt, 2001). Migrants have the potential to be organized into strong lobbies that advocate for sending country interests. In response, sending states may endow migrants with special rights, protections, and recognitions, in the hope of ensuring their long-term support (Basch et al., 1994; R.C. Smith, 1998; Guarnizo, 1997; Roberts et al., 1999). The interplay between "transnationalism from above" (by sending states) and "transnationalism from below" (by migrant groups) is evident in the practices of numerous "home-state" and "home-town" associations connecting migrants and their resources to their homelands often by promoting community development projects (Goldring, 1998; M.P. Smith, 2001); it is also seen in governments offering bonds at high state-guaranteed rates of interest to undertake major national development projects by mobilizing worldwide diasporic loyalties (Rayaprol, 1997;Sengupta, 1998).
Recently, international development agencies like the World Bank and the Inter-American Development Bank have acknowledged the development potential of migrant diasporas. Initiatives to leverage the impact of migrant remittances, such as by supporting regulatory reforms that will enable popular savings and microcredit institutions to become formal, regulated institutions, are currently being discussed (Martin, 2001).
Mobility and migration policy
If mobile populations have proven to be beneficial to local development, highly restrictive entry policies may interfere. While some analysts have expressed considerable scepticism regarding the effectiveness of migration policies and regulations in determining the level and composition for current immigration to
8                  Nyberg S0rensen, Van Hear, and Engberg-Pedersen
DCs (e.g. Massey, 1995), others hold that vast uncontrolled flows through transnational networks are unlikely to occur because immigration is in fact severely controlled by the countries to which people want to go (Sen, 1994). Yet others argue that measures to control immigration cannot be said to have failed because they have not seriously been tried (R.C. Smith, 2001).
A country' s right to determine who enters its borders, and under what conditions, is regarded as the essence of state sovereignty (Haus, 2001). With the increase in immigration to DCs in the post-war period (OECD, 1992;IOM, 1996), many states began to search for ways to stop or slow the influx. Since the early 1970s, almost all receiving countries have been trying to reassert control over migration flows, often using similar policies and in response to public opinion, which increasingly became hostile to high levels of immigration. To the extent that immigration persisted, the gap between the goals of immigration policies and the result of these policies grew and has since come to be known as the "gap hypothesis" (Cornelius et al., 1994; Hollifield, 2001).
One reason for the gap between policy goals and results undoubtedly has to do with the dominant approach among DCs to view migrants only in their role as labourers and economic actors and to ignore or overlook the prospective incorporation of migrants into society and polity. Other gap-facilitating factors, such as the presence of employers who have an interest in recruiting labour from LDCs; foreign policy considerations, or occasionally historic ties of obligation towards particular migrant groups; and a positive stance toward family reunification initiatives, have also played an important role. South-North movement induced by persecution and violent conflicts has likewise been important (Zolberg,2001).
Contrary to the goal of curbing international migration, increasingly stringent policies may benefit human smugglers and employers who hire undocumented migrants to avoid complying with existing pay and working conditions regulations, rather than stemming migration (Tacoli and Okali, 2001). Thus one unintended consequence of tightened migration controls - with measures directed against economic migrants even affecting genuine asylum seekers - is the growth in trafficking and human smuggling organizations (IOM, 2000b; Guiraudon and Joppke,2001).
Aid policy challenges
In addition to migration and refugee policies, aid policies - including development cooperation, humanitarian interventions and humanitarian assistance - have a bearing on migration-development links. Managing migration requires an understanding of why people migrate and the solutions to migration pressures lie mostly within emigration countries (Martin and Taylor, 2001). It is therefore argued that
State-of-the-art overview
19
policies should concentrate not on the migrants themselves (e.g. by limiting their mobility), but rather on ensuring that migration is a choice and not the only option.
Aligning migration and development policies is a complex affair that at times encompasses conflicting objectives. For example, skills acquisition/preferential quota systems in Europe may lead to brain-drain from LDCs, while skills retention and return may lead to capacity-building. Another potentially conflicting outcome is that while return/repatriation generally is seen as the successful end product of the migration/refugee cycle, the return of migrants and refugees means a decrease in remittances and foreign exchange for LDCs (Ferris, 2001).
Recent evidence suggests that policies that restrict migration are costly. In LDCs, they hurt the poor more than the rich (de Haan, 1999). In the long run, other policy instruments may prove more effective in reducing unwanted migration. Such instruments should be directed towards reducing the demographic and economic differences that promote economic migration, and increasing respect for democracy and human rights to minimize the number of asylum seekers and refugees. The impact of democratization or the promotion of "good governance" will take time to take effect - indeed such interventions may stimulate more upheaval and refugee migration in the short run. Therefore "democratization" may produce a "refugee hump" which somewhat parallels the "migration hump" induced by free trade orby "development" (Schmeidl, 2001).
There is some evidence to show that the provision of humanitarian aid in neighbouring countries can stimulate further refugee immigration from countries suffering conflict or widespread human rights abuse: flows of relief to Afghan refugees in Pakistan, to refugees in southern Africa (e.g. Mozambicans in Malawi), to refugees in the Horn of Africa (e.g. Ethiopians, Somalis), and to refugees in South-East Asia (Burma/Thailand) provide examples of this. But there is little evidence to show that refugee flows to further destinations, such as the west, are affected one way or another by such humanitarian intervention. Arguably, however, such intervention may have indirect influence, as "complex humanitarian emergencies" involving cocktails of conflict, humanitarian aid and refugee flows spill over borders to create "bad neighbourhoods" from which people may be impelled to escape altogether to safety further afield.
As recently argued by Zolberg (2001), it is important for DCs to avoid focusing exclusively on what they perceive as security-threatening refugee situations and limit their assistance to only such countries. Other situations, such as declining access to arable land, decreasing farm productivity, less liveable urban environments, recurring "natural" and "man-made" disaster, degraded natural resources, weak off-farm employment prospects, and increasingly restricted international migration may be acutely threatening for the populations involved.
20                   Vyber g S&rensen, Van Hear, and Engberg-Pedersen
This section has reviewed some of the current thinking on the relations between migration and development. To understand better those relations and the prospects for linking the policies and practices that address them, the following section explores in more detail existing assumptions and evidence on the migration-development nexus.
ASSUMPTIONS AND EVIDENCE ON MIGRATION-DEVELOPMENT RELATIONS
Consideration of existing and potential migration-development links involves posing fundamental questions about the migrants, the nature of their movement, and the effects of migration on the socio-economic and political structures of source areas and destinations. In recent years, most concern has tended to revolve around the positive and negative impact of foreigners on the receiving societies. To the extent that a sending country perspective has been included, migrants from LDCs have often been viewed through the prism of concern about the migration problems they pose for the Western world due to rapid population growth, poverty and conflict in the source countries (for an elaboration of the latter relationship, see Collier, 2000). In the following we turn our gaze towards LDCs to review findings from conventional as well as transnational literature and indicate various migration-development prospects of migrant diasporas. Subsequently we review the literature linking remittances, development and relief.
Migration-development links
We define migrant diasporas as being constituted by people dispersed among diverse destinations outside their home country; transnational practices are defined as activities carried out by migrant diasporas over several locations. Such practices may enhance the life chances of migrants' family members in developing countries as well as having wider developmental impacts. The ways in which migrants distribute their resources and loyalties between sending and receiving countries is, in part, determined by the kinds of institutional opportunities and government policies they encounter.
The impact of local development conditions on migration
How does development in areas of origin affect migration? It is generally believed that a lack of local development options is the root cause of economic migration. Much evidence suggests that increased local development prospects may increase migration in the short term, but ultimately make migration less necessary and attractive (OECD, 1992; Martin and Taylor, 2001). Much of the literature focuses on population movement as a result of locally determined crisis - demographic, economic or environmental - whereas the literature on structural adjustment and other external factors determining local development has only to a limited extent paid attention to the effects on migration (de Haan, 1999).
State-of-the-art overview
21
Many studies have paid attention to the characteristics of the migrants, underlining that migration is a selective process. Such studies have pointed out that migrants are usually not the poorest in the areas of origin, and that young adult men (often slightly better educated than the national average) tend to constitute the bulk of migrants from LDCs, especially in Africa (Mitchell, 1960; Chant and Radcliffe, 1992). Finally there is evidence that the local level of economic development influences the type and duration of migration. Duration of migration may be longer among migrants from economically dynamic regions than from more stagnant regions (Lindström, 1996).
The transnational literature has pointed to an interesting dynamic linking local development conditions in countries of out-migration to wider global processes (Portes, 2001):
-    The globalization of production means that multinational activities of large corporations operating in LDCs introduce new consumption aspirations and new sources of information about life elsewhere, thereby reinforcing popular incentives for out-migration.
-    Once mass migration has taken off and migrant diasporas are established abroad, a flow of transnational economic and information resources starts, ranging from occasional remittances to the emergence of a class of full-time transnational entrepreneurs.
-    The cumulative effects of these dynamics come to the attention of national governments that reorient their international activities through embassies, consulates, and missions to recapture the loyalty of their expatriates and guide their investments and political mobilizations.
-    The increased demand created by migrant remittances and investments in their home countries support, in turn, the further expansion of the market for their multinationals and encourage local firms to go abroad themselves, establishing branches in areas of diaspora concentration.
The existence of such dynamics makes evident how difficult it is to delimit the study of local migration-development causal effects from wider global and transnational processes.
The impact of migration on local development
How does migration affect development in areas of origin? A negative impact of migration is reported by various conventional analyses that point to the selective nature of migration, the lowering of local labour intensity when the most productive household members go abroad (Lipton, 1980), the tendency of remittances to be insignificant among the poorest (Massey et al., 1998), and that return migration is likely to be by old and less successful migrants whereby skills transfers are unlikely to have much developmental effect (Collinson, 1996). In addition migration may have an inflationary effect on the local economy and increase local income disparities. Other analyses suggest that migration helps to
22                  Nyberg S0rensen, Van Hear, and Engberg-Pedersen
alleviate local unemployment and infuses local economies with remittances and acquired skills (Ghosh, 1992), thereby promoting development.
The transnational literature generally shows a positive effect of migration on development in the countries of origin (for US-bound migration, see Glick Schiller et al., 1992; Rouse, 1992; Lessinger, 1992; Basch et al., 1994; S0rensen, 1994; Smith and Guarnizo, 1998; Portes et al., 1999; van der Veer, 2000; for Europe-bound migration see Soysal, 1994; Anthias and Lazaridis, 2000; S0rensen, 1995, 1999; 0stergaard-Nielsen, 2001). Although pointing to a variety of migration-development dynamics, many studies suggest that an important resource for the development of LDCs is people connected by transnational networks. Despite recent findings that point to limited numerical involvement of migrants in transnational activities, these activities remain significant because of their prospective growth and their impact on development projects in LDCs (Portes, 2001).
If transnational activities are important for national development, they are even more vital at the local level. Hometown associations have served as platforms and vehicles for matching fund schemes that pool remittances with government funds and expertise, often resulting in significant improvements in local health, education, and sanitation conditions, benefiting migrant- and non-migrant households alike (M.P. Smith, 2001). Towns and rural villages that are connected to home-town associations abroad tend to be better off in terms of infrastructure and access to services (Landolt, 2001). To enhance the positive impact, however, support in the form of services, training and infrastructure must be provided. Only then is migration likely to contribute to sustainable local development (Tacoli and Okali, 2001).
More individualized efforts have served the purpose of preventing the decline of rural communities. Many migrants do not leave in order to start a new life elsewhere but rather to better the one they already have back home (Kyle, 2000). Those who remain abroad for extended periods or eventually settle there may continue to remit sums to family members back home. Even if the immediate family resettle abroad, more distant family members may be able to count on remittances in times of acute crisis (Gardner, 1995).
If transnational literature generally shows a positive effect of transnational migration on development, it also suggests that the institutional bridges linking migrants with their home countries do not appear overnight. As the economies of sending countries come to depend increasingly on migrant remittances, their governments must contend with the transnational concerns of a growing proportion of their citizens. And one can find less positive sides to the story. For example, in some sending regions, migrants' transfer of resources has resulted in inflation of real estate prices, concentration of land tenure in the hands of families connected to migration, and increased unemployment (Fletcher, 1999). In other regions, local political feaders have been the first to depart,
State-of-the-art overview
23
depriving local communities of valuable social and political capital (S0rensen, 1999).
Developmental impact of voluntary and forced migration
Evidence suggests that the links between economic and refugee related migration are greater than hitherto expected. With the difficulties in maintaining a clear distinction between voluntary and forced migration in mind, it is still relevant - especially in relation to policy - to pose the following questions: is the relationship between voluntary and forced migration and development of the same nature and do migrants and refugees have the same interests in contributing to local development? This is perhaps the least studied but most relevant area within the field. Evidence suggests that, in the long run, development will alleviate economically driven migrations while democratization is likely to reduce forced movements (Zolberg, 2001). Less is known about the developmental impact of refugee related migrations.
Studies of refugee migration show that refugees generally come from poor countries. Hence economic factors such as lack of development options could be considered the root cause of refugee migration as well. This view is nevertheless countered by the fact that not all poor countries send refugees, leading to speculation that poverty causes political violence leading to conflict, or that poverty interacts with political violence as a root cause. Evidence suggests that poverty may provide the final "push" forpeople to leave politically unsatisfying environments (Schmeidl, 2001).
An inclusion of the role played by transnational social networks in prompting, facilitating and redirecting the movement of asylum seekers and other immigrants into Western Europe may nevertheless raise questions about the motivations for flight. For this reason, Crisp (1999) advocates that the issues of means and motivations remain rigorously separated. That does not alter the general evidence, however, that migrants and refugees alike continue to send substantial remittances to their countries of origin.
Some analysts assume that refugees wish to leave political activism behind, while economic migrants can be politicized from afar (Basch et al., 1994). Other studies have concluded that refugees who fled their countries of origin on a collective basis take a more political stand towards their homelands than economic migrants who often left on a more individualized basis (Pessar, 2001).
Many studies have documented how migrants have contributed to economic and social development in their countries of origin. Whether the developmental impact is regarded as having positive or negative consequences may depend on the extent to which the countries or areas of origin are internally differentiated. When new opportunities due to migration are introduced to more differentiated
24                    Nyberg S0rensen, Van Hear, and Engberg-Pedersen
societies, increased polarization is often the result (Gardner, 1995). But even in such societies, migration may have an equalizing effect at the household level (Gamburd,2000).
Returnlrepatriation  and development
The literature seems to suggest that return or repatriation are a prerequisite for migrants' continued engagement with local development. Yet, inadequate attention has been given to selectivity in terms of returnees' personal characteristics, duration of stay abroad, and the motivations underlying different types of return (Ghosh, 2000).
Some attention has been given to return migration in the 1970s, following the dismantling of Western Europe's guest worker programme (Collinson, 1996), and to more recent return migration of workers from the Gulf States (Gamburd, 2000). Evidence suggests that return after a relatively short period abroad, especially among low skilled migrants and if caused by an inability to adapt to the foreign environment or due to unforeseen and adverse family circumstances, is unlikely to contribute to development. Return following a longer stay abroad when the migrant has saved money to meet specific development purposes back home - such as building a house or investing in business related activities - has far better developmental prospects. Whether return will benefit local development will vary and is primarily determined by two factors: the aptitude and preparation of the return migrant, and whether or not the country of origin provides a propitious social, economic and institutional environment for the migrant to use their economic and human capital productively (Ghosh, 2000).
While some LDCs may be anxious for their migrant populations abroad to return, there may be less interest in refugee repatriation. For repatriation to be successful, a political climate facilitating former adversaries to begin to work together is needed. Recent evidence suggests a valuable consequence of hiring local professional people to take part in relief operations. Such people can be critical elements for rehabilitation. Yet they are usually the first to leave, not only because of their greater mobility but also because the risks they face. Employment of local professionals can help to create a critical mass that keeps the skilled and educated within their own country (Bissell and Natsios, 2001).
If return is seen as the end product of the migration cycle in the conventional literature, evidence from transnational studies suggests that return is not a prerequisite for continued engagement with local development. Governments of migrant sending countries have increasingly moved to intensify their contacts with their diasporas and involve them in various forms in national life. Concrete actions include the granting of dual citizenship rights, rights to vote in national elections, representation in national legislatures, cultural and religious pro-
State-of-the-art overview
25
grammes abroad targeting emigrants, and even in some cases providing services abroad for undocumented migrants that wish to legalize their undocumented status (Mahler, 1998). Portes (2001) attributes the new extra-territorial ambitions of developing country governments to the aggregate volume of remittances, migrants' actual or potential investment in the home economy, and their political influence in terms of both contributions to parties and candidates in national elections and organized mobilizations abroad.
Though only a limited number of systematic studies and comparative case assessments have been carried out so far, evidence suggests that sending governments increasingly promote transnational participation (Levitt, 2001; Portes, 2001). Evidence also suggests that states with a history of violent conflict may be more eager to capture the resources of refugees abroad than to encourage their return and participation in the post-conflict nation-state building process (Koser, 2002).
Incentives to contribute to local development
What determines migrants' incentives to contribute to development in their countries of origin? Migration, and the form it takes, is usually consistent with populations' social and cultural values (Sowell, 1996), and these values structure the patterns of migration (de Haan, 1999). Evidence suggests that population mobility often is a central element in the livelihoods of many households in LDCs (Stepputat and S0rensen, 2001; S0rensen and Ol wig, 2002). However, most development policies target sedentary populations or may even have sedent-arism as their goal. It is therefore argued that the global policy environment works to the detriment of migration benefiting local development. For example, policymakers often ignore the fact that mobility is an important part of people's livelihood diversification strategies and assume that land redistribution schemes and credit initiatives can be based on household members living together in a single place (Tacoli and Okali, 2001).
But the context of reception may have even more important implications for the directions of migrants' social and economic investments. Some analysts suggest that the greater the gap between the human capital that migrants bring along and the (lack of) opportunities they encounter in the countries of destination, the greater their motivation to engage in developmental activities towards their countries of origin (Grasmuck and Pessar, 1991; Pessar, 2001).
Contrary to these assumptions, evidence from recent transnational studies suggests that economic and political practices of migrant populations should not be reduced to a function of the opportunity structures in migrant receiving states. While more inclusive structures, which in principle allow for migrant incorporation, may exclude dialogue on homeland politics, they may also facilitate mobilization around homeland political concerns (0stergaard-Nielsen, 2001).
26                  Nyberg S0rensen, Van Hear, and Engberg-Pedersen
Remittances, development and relief
There is increasing evidence that remittances from abroad are crucial to the survival of communities in many developing countries, including many which produce refugees. Estimated at about US$75 billion a year in the early 1990s (Russell, 1993) and at US$100 billion in 2000 (Martin, 2001), migrants' remittances represent a large proportion of world financial flows and amount to substantially more than global overseas development assistance. To underline their importance for the developing world, 60 per cent of remittances were thought to go to developing countries in the year 2000 (Martin, 2001).
There has been a considerable amount of research about the effects of remittances sent by economic migrants to relatively stable low-income and middle-income countries. The overall finding is that remittances to developing countries go first and foremost to lower-middle income and low-income countries. Lower middle income countries receive the largest amount, but remittances constitute a much higher share of total international flows to low income countries. In the second half of the 1990s, foreign aid and remittances to low-income countries were of almost equal size, each constituting a third of international flows. Taking remittances through unofficial channels into account, remittances are surely greater than foreign aid, and constitute a more constant source of income to developing countries than other private flows and foreign direct investments (Gammeltoft, 2002).
Refugees also remit a share of their resources but less is known about remittances and other transfers by and to refugees. Are these essential for the survival of those left at home or languishing in refugee camps? Are the transfers used in ways similar to those remitted by economic migrants - for survival or daily essentials, or as investment in productive activity - or are they channelled as support for armed groups that may prolong conflict and retard peace-building? How might money and other transfers be encouraged to assist in post-conflict reconstruction?
Exploring such questions has only just begun (de Montclos, 2001; Van Hear, 2000), and requires examining the flow and role of remittances to and from refugees and displaced people during and after conflicts or refugee crises. It also involves tracking the flow of resources among three different categories of people: refugees in far-flung diasporas, refugees in countries neighbouring their homelands, and those left behind in the country of origin, including the internally displaced. The limited evidence so far points'to the ambivalent nature of remittal .ces for both refugees and their families at home.
While considerable work has been accomplished on the developmental impact of remittance flows to LDCs, two scenarios can be identified in which the flow and role of remittances differ in their impacts on relief and development. These
State-of-the-art overview
two scenarios involve transfer from (and to) refugees and displaced people during conflicts and after conflicts. Comparison is needed of the impact of transfers by "economic" migrants to (low- and middle-income) homelands enjoying relative peace and stability on one hand, and on the other the role of remittances during conflicts or refugee-producing crises, as well as the role or potential of remittances after conflicts, in the period of repatriation, reintegration and post-conflict reconstruction.
Remittances during conflicts!refugee crises
Remittances from abroad help families to survive during conflict and to sustain communities in crisis - both in countries of origin and in countries of first asylum. In conflict-torn societies and regions, the scope for investment in "productive" enterprises may be limitedin conditions of great insecurity. However, investment of remittances in social activities may be seen as reconstruction of the social fabric, in which "productive" activities are embedded. By facilitating the accumulation or repair of social capital, such investment may lay the foundation for later reconstruction and development (Goodhand, Hulme and Lewer, 2000; Van Hear, 2001).
At the same time, remittances and other transfers, as well as international lobbying by diasporas, may help perpetuate the conflicts or crises that beset such families and communities, by providing support for armed conflict. For Collier (2000) the existence of a large diaspora is a powerful risk factor predisposing a country to civil war, or its resumption. Anderson (1999), another influential writer on conflict, holds a similar view. Duffield (2001) notes that many contemporary wars are sustained by regional and global linkages through which local resources are sold and arms and other essential supplies are bought. Diasporas are crucial in the flow of money and resources on which warring parties depend, helping to market commodities extracted from war-torn areas, or more indirectly supplying the finance or lubricating the connections needed to effect such transactions.
The balance between these positive and negative influences of migrants, diasporas and their transfers varies from case to case. More thought needs to be given to the extent to which policy interventions can encourage the deployment of transnational activities in positive direction, such as towards conflict resolution or post-conflict reconstruction, considered below.
Remittances after conflicts
Remittances are potentially a powerful resource for post-conflict reconstruction (Koser, 2002). Much depends on the extent to which remittances are actually used for reconstruction, and the means and policies that can be deployed to encourage that outcome. There is also the conundrum that if the resolution of conflict or crisis is accompanied by large-scale repatriation, the source of remittances will obviously diminish, raising potential perhaps for instability and
28                  Nyberg SQrensen, Van Hear, and Engberg-Pedersen
further conflict. There may even be an argument against repatriation on these grounds. Such was the thrust of an appeal in 1995 by the president of El Salvador for the US authorities to refrain from repatriating Salvadorans whose temporary protection in the US was imminently expiring (Mahler, 2001).
Repatriation or restrictions on immigration may have far-reaching consequences for development. The consequences include the possibility that a diminution of remittances may lead to hardship, instability, socio-economic or political upheaval, and even the resumption or provocation of conflict - and then quite likely renewed out-migration. The curtailment of immigration and the implementation of repatriation may therefore imperil the very economic and political security -in broader terms the human security - that the international community claims to want to foster. Moreover, the trend towards containment of forced migrants in countries or regions of origin will mean that those remaining in such places may have less in the way of earning and therefore potential remittance power than those in more prosperous asylum countries. In the longer term, as already indicated, remittances have the potential to be harnessed for the reconstruction and development of societies recovering from the distress of war or economic collapse; diminution of such transfers through repatriation will likely undermine such potential. It follows that migration policies that purport to be oriented to the country of origin of migrants cannot afford to leave those abroad out of consideration, especially those hosted by relatively affluent countries.
CHALLENGES TO AID: COHERENCE AND SELECTIVITY
Underlying international thinking on aid and migration has been the question of the effectiveness of aid in reducing migration and refugee flows, by generating local development; preventing and resolving local conflicts; and retaining refugees in neighbouring areas/first countries of asylum. The migration-development nexus poses three challenges to the present international aid system:
1)  Can and should development cooperation and humanitarian assistance be linked or even merged, in order to maximize the potentially mitigating impact of aid on migration-producing circumstances, including poverty and conflicts?
2)  Can and should aid (development and humanitarian) be allocated more selectively among developing countries, to maximize the potential impact of aid on migration, and if so, should the selectivity be based on development, humanitarian or other criteria?
3)  Can and should aid, which tends ,to be space-bound and state-centred, interact more directly with mobile populations, diasporas and transnational
State-of-the-art overview
29
communities? Can and should aid extend its support of national and local civil society to the international level as well?
The first question has been addressed by research and aid agencies, though mainly from the point of increasing the effectiveness of humanitarian assistance in general. The second question about selectivity has been addressed out of concern for the local development effect in developing countries, but only recently in a migration context. The third question represents a quite new challenge to the aid community. A positive answer to these three sets of questions would produce two new dilemmas:
Integration of development and humanitarian assistance has proved to be complex and costly, and makes limited sense at a time of declining aid budgets. Greater selectivity in aid allocations would pull aid in two different directions: development aid would go to the countries performing well in economic and political terms, whereas humanitarian assistance would go to failed and conflict-producing states. The first dilemma is that development aid has a greater potential than humanitarian aid in terms of preventing violent conflict and the migration it generates. If approaches and instruments of development aid are used to address migration-producing factors, the selectivity criteria emphasized so far for development aid would need to be reversed - and then the potential of aid for reducing poverty in "good performance" countries would be lost. Conversely, if only humanitarian assistance is used in crisis countries, aid would have a very limited role in conflict prevention, because humanitarian assistance tends to be delivered after it might have had a conflict-preventing role.
The second dilemma is also related to the aid allocation issue. Comparing the motivations of aid donors and migrants, it seems that migrants' remittances and investments in their countries of origin may follow paths that replace, supplement or even undermine aid. Remittances sent by low-skilled migrants to poorer areas of origin are likely to be for social and livelihood purposes, and their allocation is unlikely to follow the geopolitical and/or commercial political and economic objectives of aid donors. The dilemma is that allocation decisions are taken in different spheres and that the decision-makers have little experience with collaboration and coordination.
This section first summarizes the logic and achievements of integrating development and humanitarian assistance - what has been called linking relief and development. Subsequently, arguments are discussed for and against greater selectivity in the allocation of development and humanitarian assistance, respectively. Finally, the challenges posed to aid by mobility, remittances and transnational communities are discussed, and ways to address the two dilemmas are explored.
30                    Nyberg S0rensen, Van Hear, and Engberg-Pedersen
Linking relief, recovery, development and conflict prevention
The rationale for linking relief and development assistance is that people and societies in need of relief usually are likely to be the most vulnerable and hence likely to be exposed to new disasters (with natural and/or human causes). To reduce their vulnerability, relief and recovery should be development-oriented. In a good summary of current thinking, the International Federation of Red Cross and Red Crescent Societies (IFRCRCS, 2001: 12-33) discusses four ways for aid to secure and strengthen recovery.
First, the delivery of relief should support, not undermine, recovery. This includes making use of existing institutions in the disaster area, which contributes to long-term capacity building, and providing relief plus assets (for example, tools, seeds and other agricultural inputs) for the recovery of the disaster victims' livelihood. Relief can provide breathing space and a minimum of resources from which to move towards sustainable recovery. However, it can only lead to recovery if it is followed up by development interventions.
The second main approach suggested by IFRCRCS concerns the inclusion of risk-awareness in development promotion. In conflicts and disasters, speed is indeed decisive for the saving of lives, but even the most efficient international NGOs and UN relief agencies are likely to be much too late in arriving at disaster sites. This means that community preparedness is the only practical solution for poor, high-risk countries. The capacity of communities and local institutions to respond to conflicts and recurrent disasters must be strengthened.
The third approach concerns the funding gap between relief and recovery. Donor agencies are fully aware of the persistent rigidity of their budget lines, despite a decade of discussions of the relief-development continuum and development-oriented relief. Some relief donors operate with a six-month spending window on emergency funds. The funding gap between relief and recovery programmes remains a real challenge to both humanitarian and development agencies.
The fourth approach of the IFRCRCS to addressing recovery concerns a linking of aid and advocacy of structural changes at political and economic levels. The realization is simple: "Aid alone will never be able to combat root causes and break the cycle of disasters" (IFRCRCS, 2001: 28). Aid cannot address all aspects of conflict, climate change, under-development, structural poverty, and uneven globalization and marginalization. But relief provides an opportunity and entry-point for political advocacy by international NGOs, UN agencies and their local partners. Complex emergencies demand concurrent action within relief, rehabilitation, recovery, development (including risk awareness and local disaster preparedness), and advocacy. Concurrent actions differ from continuum and
State-of-the-art overview
31
integrated approaches, since they can be taken ad hoc by the best capable local, national and international agencies within an overall strategic framework.
The attempts during the 1990s to link relief, rehabilitation and development within the operational aid activities of individual donor agencies faced numerous problems. Such efforts were revived in slightly more coordinated form in the late 1990s in the shape of the "Brookings Process", under which UNHCR, UNDP and the World Bank undertook to bridge "the gap between humanitarian assistance and long term development" (Crisp, 2001; World Bank, 1998a). However, a countervailing trend is also apparent, given the tendencies for multilateral agencies such as the World Bank and UNHCR to concentrate more on their "core" activities - development and refugee protection, respectively -mainly as a consequence of declining aid resources. Similarly, the recent management and organizational reforms in the European Commission kept its Humanitarian Office (ECHO) and humanitarian relief outside the integration of all the Commission's external relations, which is aimed at establishing an integrated EU response to global crises. Thus, there have been both political and operational reasons for limiting the integration of humanitarian assistance with development cooperation and political-economic relations.
In 2001, the UN Secretary-General published a report on the prevention of armed conflict (see also Brahimi, 2000). He reiterated his pledge to move the UN from a culture of reaction to a culture of prevention; the report's underlying message seemed to be that the UN and its member states and partners should do more in all fields. The report was based on the premise that conflict prevention and sustainable and equitable development are mutually reinforcing activities. It quotes studies by the Carnegie Commission showing that the international community spent about US$200 billion on the seven major interventions of the 1990s (Bosnia, Somalia, Rwanda, Haiti, the Persian Gulf, Cambodia and El Salvador, exclusive of Kosovo and East Timor), whereas a preventive approach might have cost only some US$70 billion. These figures are highly uncertain, but they undoubtedly reflect the cost-effectiveness of prevention over cure in conflict management. Kofi Annan's recommended solutions are more resources, earlier interventions, coordinated planning and coordinated implementation, all within a framework of respect for the national sovereignty of individual states.
There is consensus that the prevention ■ of violent conflict must be multidimensional, and it should be part of all development and humanitarian aid -that is, "mainstreamed" in the aid vocabulary. This is a challenge to the international community, but it does not mean that all these measures have to be delivered as integrated packages by individual agencies. The international community needs conflict- and country-specific strategic frameworks for conflict prevention, reconciliation and resolution, but the implementation of the
32                  Nyberg S0rensen, Van Hear, and Engberg-Pedersen
individual measures may be handled by the agencies in accordance with their respective capacities. Much improved conflict analysis is seen as a major requirement for the international community and for individual agencies (Addison, 2000; Royal Danish Ministry of Foreign Affairs, 2000; Ramsbotham and Woodhouse, 1996).
Arguably such analysis has already being integrated into development discourse. A fundamental policy shift in this respect can be traced in the 1990s (Macrae, 2000; Duffield, 2001; Crisp, 2001). In the first half of that decade, the focus was on developing institutional arrangements that allowed aid agencies to work in conflict zones, and to help civilians caught up in them. Such initiatives had limited success and from the mid-1990s the focus shifted towards conflict resolution and post-war reconstruction (Duffield, 2001). Such shifts were manifested in debate over what came to be called "complex emergencies" or "complex political emergencies", that is conflict-related humanitarian disasters involving much social dislocation and often forced displacement, and requiring system-wide responses from the international community. This debate reflects what Duffield (and others) describe as "the blurring and convergence during the 1990s of development and security" (ibid: 15); in a shift in the politics of development, the object now is to resolve conflicts and to reconstruct societies so as to avoid future wars. "Through a circular form of reinforcement and mutuality", the perception now is that "development is ultimately impossible without stability and, at the same time, security is not sustainable without development" (ibid: 16). This reinforces the integrated understanding of relief, conflict resolution, migration, development and security, which is a major institutional challenge to the international community.
Selectivity in development and humanitarian assistance
Two partly opposing arguments are currently being made for the introduction of greater selectivity in the allocation of aid. On the one hand, the case is made that development aid should be given primarily to the good performers among developing countries, because this will enhance effectiveness and pull more people out of extreme poverty: development assistance is effective in developing countries with good governance, sound economic policies and capable institutions. On the other hand, the case is made (reinforced after 11 September 2001) that aid should be given primarily to the present and future trouble spots and crisis countries in order to diffuse or control conflicts, reduce the appeal of fundamentalism, and contain refugees in neighbouring areas (first countries of asylum); this argument is made for both development and humanitarian aid.
Because these two arguments point towards different allocation patterns for aid, there is pressure in OECD countries and multilateral institutions for a clearer distinction between development aid for good performers (in economic and political terms) and humanitarian aid for crisis countries. Since the political
State-of-the-art overview
33
interest in crisis prevention and control is increasing in connection with the fight against international terrorism, there is pressure for an increase in humanitarian interventions, including aid, possibly to be financed through a reduction in allocations for long-term development cooperation. The international debate has only recently started to examine how these policy prescriptions fit with attempts to address migration-development links in a coherent manner.
Performance-based development aid
The World Bank's Assessing Aid: What Works, What Doesn't, and Why (1998) was a milestone in the move from needs- towards performance-based aid. According to Beynon (2001), it prompted a vigorous, healthy but at times hotly contested debate. "Two opposing viewpoints have emerged. According to one, aid only really works when government policies are good, and a more selective allocation of aid to "good policy/high poverty" countries will lead to larger reductions in poverty. According to the other, aid effectiveness is not conditional on policy and the implications of the former for more selective aid allocations are treated with concern" (Beynon, 2001: 1). The debate is ongoing, but some elements of agreement are emerging (Degnbol-Martinussen and Engberg-Pedersen, 2003), including:
1)  Aid has contributed significantly to a reduction of poverty in recipient countries, through economic growth, income redistribution, improved health and education, that is, a combination of resource transfer, societal change, capacity building and human resource development.
2)  Aid is, of course, most effective in developing countries with macro-economic stability, coherent policies, and good institutional capacities. It is essential that improved policy-making and capacity building are aid objectives in all developing countries, in particular in the weakest and least developed countries.
3)  The greatest global effect in terms of income poverty reduction will be achieved through (re-) allocating aid to the low-income countries with the largest number of poor people. This is more effective than a re-allocation of aid on the basis only of the "quality" of policies and institutions.
4)  Humanitarian assistance, technical cooperation, and development projects and programmes should be allocated on the basis of poverty, social and political needs, and institutional opportunities. With very few exceptions where only humanitarian assistance is possible (examples in 2002 were the Democratic Republic of Congo, Somalia, Liberia and Sierra Leone), all of these aid forms are needed in all low-income countries.
This emerging consensus would help resolve the dilemma of development aid going to the strong countries and humanitarian assistance to the crisis countries.
34                  Nyberg S0rensen, Van Hear, and Engberg-Pedersen
Recognition that the tools and approaches of development aid are needed in the poorest trouble spots (that is, present and potential crisis countries) would also give development cooperation a renewed role in migration-producing circumstances.
Selectivity and the effectiveness of humanitarian assistance
Resources for humanitarian assistance have increased, but not commensurate with the increasingly diverse use of this type of aid. For some donors, Eastern Europe and Central Asia have been added to the target areas for conflict and disaster relief; the closer links between relief and both development assistance and broader humanitarian interventions, and the use of aid for asylum seekers in donor countries have reduced the resources available for "pure" relief; and national and international mobility is adding new target groups (internally displaced persons and other types of migrants) to the recipients of humanitarian assistance.
The terror attacks in New York and Washington in September 2001 have revived the justification of aid as a potentially powerful tool of conflict prevention, reconciliation and resolution, although there is no immediate and direct link between poverty and terrorism. Humanitarian assistance can only help to diffuse tensions and the risk of fundamentalist attacks if aid agencies become better at engaging constructively with the adversaries of the present world order. This requires that aid agencies have a significant presence in the world's actual or potential trouble spots, reflecting a development-oriented, long-term commitment.
These are requirements that contradict the emphasis on speed and top-down service delivery in the operations of relief agencies. Research on aid as an arena of conflict over knowledge and resources (Long, 1992; Olivier de Sardan, 1995) suggests new demands on aid agency staff, who must be personally involved in local dialogues in ways that are new and alien to the operations and roles of both relief and development agencies in the past. Contrary to the ideological emphasis on aid to the good performers, this new aid rationale calls for strategic use of aid in the countries that are affected by conflicts and are most likely to produce migration and/or attacks (political and/or terror) on international institutions and donor countries. These forms of selective humanitarianism may become a threat to fundamental principles of need and neutrality in humanitarian assistance.
Aid, remittances and diasporas: New policy fields and options
There is not much research on the strengths and weaknesses of aid seen from a migration perspective. The new challenga is for policy and research to explore ways in which aid can influence migration-producing factors (both conflict- and
State-of-the-art overview
35
development-related); and interact with migrants and diasporas to reduce violent conflicts and poverty and promote development in developing countries. The evidence on the first issue may be summarized as follows:
-    Development aid has the instruments and the approach to influence migration-producing factors; but there is a tendency for aid increasingly to go to the relatively well-performing developing countries that do not produce (forced) migration.
-    Humanitarian aid goes to migration-producing circumstances (such as violent conflicts), and it is likely to do so even more in future; but it tends to arrive after the event and it lacks the instruments and the agency presence to prevent violent conflicts and reduce forced migration. It may, however, reduce migration from LDCs to DCs to the extent that it helps keeping refugees in neighbouring areas, that is, in first countries of asylum.
-    The attempts to integrate humanitarian and development aid as a conflict-preventive and migration-controlling measure have suffered from high institutional complexities and financial costs. This applies also to the attempts to link aid (mainly humanitarian) with peace building and peacekeeping through humanitarian interventions in complex political emergencies. If aid is used mainly for peacemaking and containment, there is a risk of under-utilizing its potential for prevention of violent conflicts.
There is almost no evidence on the second question, the links between aid and diasporas. While recognizing that the motivations behind migrants' remittances are likely to combine economic, political and social dimensions, and that their effects are likely to be a mixture of survival, consumption and development, it is necessary to work with rather simplistic scenarios about the possible links between aid donors and migrant communities. Four situations may be distinguished:
1)  Aid and remittances to relatively peaceful, low-income countries (LICs). Remittances provide income, foreign exchange and ideas for both human- and private-sector development. In addition to concentrating on policy-making, institutional capacity building, democratization and so on, aid donors could facilitate the involvement of diasporas in private and community sector development at all levels of society and in support to "poor people on the move".
2)  Aid and remittances to relatively peaceful middle-income countries (MICs). Again, remittances provide livelihood support, but they are also likely to be part of overall foreign direct investments to MICs. The evidence that global poverty is best reduced though re-allocation of aid to the poorest countries would suggest a reduction in official aid to these countries, particularly if remittances and investments do seek out economic  opportunities  in  them.  Still,  aid  agencies could  facilitate
36                  Nyberg S0rensen, Van Hear, and Engberg-Pedersen
collaboration between the state, national and international civil society, that is, the diaspora, aimed at nation and state building.
3)  Aid and remittances to present and potential trouble spots, including conflict-affected countries and failed states. This is where the greatest new challenges lie. Bilateral and multilateral aid agencies could seek ways to combine their different aid types with the activities of transnational communities, aimed at ensuring that the remittances and other activities provide resources, security and political space to the poor and other conflict victims, rather than fuelling violent conflicts. This calls for a case-specific approach, where donors - possibly under UN leadership - invite international NGOs and diaspora organizations for transparent dialogues on the overall resource flows to the country, including to possible conflicting parties.
4)  Aid and remittances to "post-conflict" countries and regions. While in some ways a sub-set of the previous section, the possibilities and techniques are somewhat different in post-conflict states and regions attempting the three "re-s": repatriation, reintegration and reconstruction. Here the focus should be on mobilizing diaspora resources for reconstruction as part of wider international peace-building, reconciliation and reconstruction efforts, with special emphasis on avoiding the generation of new tensions that might lead to new rounds of conflict and displacement.
CONCLUSIONS
This state-of-the-art overview has shown that current thinking is still tentative and available evidence is sketchy in many areas regarding the links between migration and development. As a starting-point for the exploration of policy options (in the policy paper at the end of this volume), the study offers four conclusions.
Poverty and migration. People in developing countries require resources and connections to engage in international migration. There is little evidence of a direct link between poverty, economic development, population growth, social and political change on the one hand and international migration on the other. The "migration hump" suggests that some economic development generates both the resources and the incentives for people to migrate. By implication, poverty reduction is not in itself a migration-reducing strategy. As long as poverty reduction is the overriding goal of aid and development cooperation, there should be no direct link between aid and migration control.
Conflicts, refugees, and migration. Violent conflicts produce displaced persons, migrants and refugees. People on the move may both contribute to
State-of-the-art overview
37
conflict prevention and reconciliation and to renewed and sustained conflict. The impact of migrant and refugee diasporas on conflict or stability in their country and region of origin varies between situations and over time. The large majority of refugees do not have the resources or the opportunities to move beyond neighbouring areas; they remain internally displaced or move across borders to first countries of asylum in the region. By implication, aid to developing countries receiving large inflows of refugees is poverty-oriented to the extent that these are poor countries in need of international support for local settlement of refugees, but it is uncertain whether such aid has much effect in terms of reducing the number of people seeking asylum in developed countries. There is some evidence that aid to neighbouring countries has served as a pull-factor attracting refugees from countries in war, anarchy or even economic and environmental crisis. The alternative of providing the aid directly to the refugee producing countries faces the challenge of controlling the intended and unintended effects of aid on conflict, where aid, aid agencies and their staff are drawn into local political processes and violent conflicts.
Migrants as a development resource. In the nineteenth and early twentieth century, migrants were viewed as a resource contributing to the development of both sending and receiving countries. In the last quarter of the twentieth century, migrants from developing to developed countries were increasingly seen by the latter as a problem in need of regulation. The dominant mode of regulation has been stricter immigration controls. Liberalization has been deep and global with respect to capital, goods and services, but not to labour mobility. Current international political-economic institutions and regimes provide neither space nor initiatives for negotiations on labour mobility and the flow of remittances. Based on both evidence and political interests, there is a pressing need to reinforce the view of migrants as a development resource, for at least four reasons. First, the remittances by migrants and refugees are likely to be double the size of aid and may be at least as well targeted at the poor in both conflict-ridden and stable developing countries. Second, migrant diasporas are engaged in a variety of transnational practices (such as relief, investments, cultural exchange and political advocacy) with direct effects on international development cooperation. Third, both private and public sectors in developed countries recognize their immediate and long-term dependence on immigrant labour with an ever more complex skills mixture. Fourth, policies on development cooperation, humanitarian relief, migration, and refugee protection are internally inconsistent and occasionally mutually contradictory. Viewing migrant diasporas as a development resource and seeking links between aid and migrants' transnational practices could address some of these trends and concerns.
Aid and migration. Aid policies face a critical challenge to balance a focus on poverty reduction with mitigating the conditions that produce refugees, while at the same time interacting more constructively with migrant diasporas and their transnational practices. Donors must revisit their current approach to selectivity
38                    Nyberg S0rensen, Van Hear, and Engberg-Pedersen
in aid, which tends to allocate development aid to the well performing countries and humanitarian assistance to the crisis countries and trouble spots. The long-term approaches and instruments of development aid are more effective than (shorter-term) humanitarian assistance in preventing violent conflicts, promoting reconciliation and democratization, and encouraging poverty-reducing development investments by migrant diasporas.
These interconnected issues are pursued in the final policy analysis chapter of this volume, which explores ways in which migration and development can mutually reinforce one another.
REFERENCES
Addison, T. 2000      "Aid and conflict", in F. Tarp (Ed.), Foreign Aid and Development: Lessons Learnt and Directions for the Future, Routledge, London. Anderson, M.B. J999      Do No Harm: How Aid Can Support Peace - or War, Lynne Rienner, London. Anthias, F. 2000      "Metaphors of home: gendering new migrations to southern Europe", in F. Anthias, and G. Lazaridis (Eds.), Gender and Migration in Southern Europe: Women on the move, Berg Publishers, Oxford. Anthias, F., and G. Lazaridis (Eds.) 2000      Gender and Migration in Southern Europe: Women on the Move, Berg Publishers, Oxford. Ascencio, D.
1990      Unauthorised Migration: An Economic Development Response, Report of the US Commission for the study of international migration and cooperative economic development, US government printing office, Washington, DC. Athukorala, P.
1993       Enhancing Developmental Impact of Migrant Remittances: A Review of Asian Experiences, Asian Regional Programme on International Labour Migration, ILO-UNDP Project, New Delhi.
Bakewell, O.
2000       "Repatriation and self-settled refugees in Zambia: bringing solutions to the wrong problems", Journal of Refugee Studies, 13(4): 356-373.
Basch, L., N. Glide Schiller and C. Szanton Blanc
1994       Nations Unbound: Transnational Projects, Postcolonial Predicaments, and Deterritorialized Nation States, Gordon & Breach Publishers, Pennsylvania.
Beynon, J.
2001       Policy Implications for Aid Allocations of Recent Research on Aid Effectiveness and Selectivity, paper presented at the Joint Development Centre/ DAC Experts Seminar on Aid Effectiveness, Selectivity and Poor Performers, 17 January 2001, OECD, Paris.   "
State-of-the-art overview
39
Bissei, R., and A. Natsios 2001      "Development assistance and international migration", in A.R. Zolberg, and P.M. Benda (Eds.), Global Migrants, Global Refugees: Problems and Solutions, Berghahn Books, New York.
Black, R.
1998       Refugees, Environment and Development, Longman, Harlow. Brahimi, L.
2000      Report of the Panel on United Nations Peace Operations, United Nations, New York.
Brimelow, P.
1995         Alien Nation: Common Sense about America's Immigration Disaster, Ran-
dom House, New York. Castles, S., and M.J. Miller
1993       The Age of Migration: International Population Movements in the Modern World, Guilford Press, New York.
Chant, S., and S.A. Radcliffe 1992      "Migration and development: the importance of gender", in S. Chant (Ed.), Gender and Migration in Developing Countries, Belhaven Press, London. Chimni,B.S.
1999      From Resettlement to Involuntary Repatriation: Towards a Critical History of Durable Solutions to Refugee Problems, Evaluation and Policy Analysis Unit, New Issues in Refugee Research, Working Paper No2, UNHCR, Geneva.
Collier, P.
2000      Economic Causes of Civil Conflict and their Implications for Policy, World Bank, Washington, DC.
Collinson, S. 1996      Shore to Shore. The Politics of Migration in Euro-Maghreb Relations, The Royal Institute of International Affairs, Middle East Programme, London. Cornelius, W.A., P. Martin and J. Hollifield
1994       Controlling Immigration: A Global Perspective, Stanford University Press, Stanford.
Crisp, J.
1999 Policy Challenges of the New Diasporas: Migrant Networks and their Impact on Asylum Flows and Regimes, Evaluation and Policy Analysis Unit, New Issues in Refugee Research, Working Paper no.7, UNHCR, Geneva.
2001       "Mind the gap! UNHCR, humanitarian assistance and the development process", International Migration Review, 35( 133): 168-191.
Degnbol-Martinussen, J., and P. Engberg-Pedersen 2003      Aid: Understanding International Development Cooperation, Zed Press, London. De Haan, A. 1999      "Livelihood and poverty: the role of migration - a critical review of the migration literature", Development Studies 36(2):3-47. De Montclos, M.
2001       "Réseaux financiers et hawilad: le role de la diaspora somalienne dans la reconstruction de leur pays", in L. Cambresy and V. Lassailly-Jacob (Eds.), Populations refugiées: de ľ exil et retour, Institut de recherche pour le developpement, Paris.