Migration and Development: A Theoretical Perspective.

The debate on migration and development has swung back and forth like a pendulum, from developmentalist optimism in the 1950s and 1960s, to neo-Marxist pessimism over the 1970s and 1980s, towards more optimistic views in the 1990s and 2000s. This paper argues how such discursive shifts in the migration and development debate should be primarily seen as part of more general paradigm shifts in social and development theory. However, the classical opposition between pessimistic and optimistic views is challenged by empirical evidence pointing to the heterogeneity of migration impacts. By integrating and amending insights from the new economics of labor migration, livelihood perspectives in development studies and transnational perspectives in migration studies - which share several though as yet unobserved conceptual parallels - this paper elaborates the contours of a conceptual framework that simultaneously integrates agency and structure perspectives and is therefore able to account for the heterogeneous nature of migration-development interactions. The resulting perspective reveals the naivety of recent views celebrating migration as self-help development "from below". These views are largely ideologically driven and shift the attention away from structural constraints and the vital role of states in shaping favorable conditions for positive development impacts of migration to occur.

Introduction

In the past few years there has been a remarkable renaissance in optimism and the overall interest in the issue of migration and development by policy makers and scholars. After decades of pessimism and concerns on brain drain, governments of migrant sending countries have put renewed hopes on transnationally oriented migrants and “Diasporas” as potential investors and actors of development. Surging remittances, in particular, are often believed to be a more effective instrument for income redistribution, poverty reduction and economic growth than large, bureaucratic development programs or development aid (Jones, 1998; Kapur, 2003; Ratha, 2003).

However, the recent re‐discovery of the migration‐development nexus tends to go along with a certain neglect of the insights that have emerged from decades of prior research and policy experience. Furthermore, there has been a tendency to study causes and impacts of migration separately, which constitute largely separate strands of migration literature. This is unfortunate, since the developmental factors influencing migration decisions are also likely to shape the developmental outcomes in sending countries and communities (Taylor, 1999). Third, and more generally, the scholarly debate on migration has tended to separate the developmental causes (determinants) and effects (impacts) of migration artificially from more general processes of social (including economic ) change.

Rather, we need to see migration as (1) a process which is an integral part of broader transformation processes embodied in the term “development”, but (2) also has its internal, self‐sustaining and self‐undermining dynamics, and (3) impacts on such transformation processes in its own right. This contextuality has important theoretical implications. Because migration is not an exogenous variable, but an integral part of wider social and development processes, the development impacts of migration are also fundamentally heterogeneous.

The specific debate on migration and development has evolved rather separately from general migration theory. Because of their focus on migration processes or their focus on migrant receiving societies, general migration theories do not offer many specific insights into the nature of migration impacts on development in sending societies, let alone the heterogeneity of such impacts. We therefore need to put the specific debate on migration and development in a broader perspective of social and migration theory.

The first aim of this paper is to review how specific theories on migration and development have evolved over the past half century. It shows how discursive shifts in the debate on migration and development reflect more general paradigm shifts in social and development theory. First, we discuss opposing traditional “optimistic” and “pessimistic” views, and analyze their intimate connections with general functionalist and structuralist strands of social theory. Subsequently, we review alternative, more “pluralist” (which simultaneously take into account agency and structure) and refined views on migration and development that have emerged more recently. The second aim of this paper is to elaborate the contours of a conceptual framework for analyzing heterogeneous migration and development interactions within a broader social theory perspective. This is done through integrating and amending insights from recent, pluralist perspectives on migration and development which have evolved largely separately in migration economics, development studies and migrant studies.

Migration and development optimists versus pessimists

Over the past five decades, the impact of migration on development in migrant sending communities and countries has been the subject of continuous and sometimes heated debate, opposing views of the “migration optimists” and “migration pessimists” (cf. Taylor, 1999). This division in views on migration and development reflects deeper paradigmatic divisions in social theory (i.e., functionalist versus structuralist paradigms) and development theory (i.e., balanced growth versus asymmetric development paradigms). To a considerable extent, this also reflects ideological divisions between neoliberal and state‐centrist views. Table 1 summarizes the opposed views of these two schools of thought on migration and development.

Table 1

 
O

Migration optimists		Migration pessimists
Functionalist	↔	Structuralist
Neo‐classical	↔	Neo‐Marxist
Modernization	↔	Disintegration
Net North‐South transfer	↔	Net South‐North transfer
Brain gain	↔	Brain drain
More equality	↔	More inequality
Remittance investment	↔	Consumption
Development	↔	Dependency
Less migration	↔	More migration

The scholarly and policy debates on migration and development have tended to swing back and forth like a pendulum from optimism until the early 1970s to pessimism until the 1990s, and back again to more optimistic views in recent years (see Table 2). As the paper will argue, these shifts reflect more general paradigmatic shifts in social and development theory. The following sections will explore the theoretical roots of these different strands of thinking on migration and development.

Table 2

 
M

Period	Research community	Policy field
Until 1973	Development and migration optimism	Developmentalist views; capital and knowledge transfers by migrants would help developing countries in development take‐off. Development strongly linked to return.
1973–1990	Development and migration pessimism (dependency, brain drain)	Growing skepticism; concerns on brain drain; after experiments with return migration policies focused on integration in receiving countries. Migration largely out of sight in development field, tightening of immigration policies.
1990–2001	Readjustment to more subtle views under influence empirical work (NELM, livelihood approaches, transnationalism)	Persistent skepticism and near‐neglect of the issue; “migration and development, nobody believes that anymore” (Taylor et al., 1996a:401) further tightening of immigration policies.
>2001	Boom in research, in particular on remittances. Generally positive views. De‐linking of development with return.	Resurgence of migration and development optimism under influence of remittance boom, and a sudden turnaround of views: remittances, brain gain, diaspora involvement as vital development tools. Development contribution of migration often framed within renewed hopes put on circular and return migration.

Source: Adapted from De Haas (2007a).

Optimistic Views: Neo‐Classical and Developmentalist Theory

Neo‐classical migration theory perceives migration as a form of optimal allocation of production factors to the benefit of both sending and receiving countries. In this perspective of “balanced growth”, the re‐allocation of labor from rural, agricultural areas to urban, industrial sectors (within or across borders), is considered as a prerequisite for economic growth and, hence, as an constituent component of the entire development process (Todaro, 1969:139). The free movement of labor – in an unconstrained market environment – will eventually lead to the increasing scarcity of labor, coinciding with a higher marginal productivity of labor and increasing wage levels in migrant sending countries. Capital flows are expected to go in exactly the opposite direction, that is, from the labor‐scarce to the capital‐scarce migrant sending countries. Eventually, this process of factor price equalization (the Heckscher‐Ohlin model) predicts that migration ceases once wage levels at the origin and destination converge (Massey ).

In a strictly neo‐classical world, the developmental role of migration is entirely realized through factor price equalization. As Djajic (1986) pointed out, earlier neo‐classical migration theory ruled out the possibility of a gain for non‐migrants. Strictly speaking, neo‐classical migration theory has therefore no place for money remittances flowing to origin countries (Taylor, 1999:65). Neo‐classical migration theory tends to view migrants as atomistic, utility maximizing individuals, and tends to disregard other migration motives as well as migrants’ belonging to social groups such as households, families and communities.

According to dominant views of the 1950s and 1960s in development theory, return migrants were seen as important agents of change and innovation. It was expected that migrants not only bring back money, but also new ideas, knowledge, and entrepreneurial attitudes. In this way, migrants were expected to play positive role in development and contribute to the accelerated spatial diffusion of modernization in developing countries. Also remittances have been attributed an important role in stimulating economic growth.

Such optimistic views were rooted in earlier studies on rural‐to‐urban migration within Europe and the United States and based on the historical experience with emigration from Europe to North America. This also reflected “developmentalist” views which dominated in development theory and theory in the first two decades following the Second World War. Rooted in evolutionary views on development, freshly decolonialized countries were expected to quickly follow the same path of modernization, industrialization, and rapid economic growth as many Western countries had gone through. Assuming that capital constraints formed the major problem these countries faced, the developmentalist model postulated that through large‐scale capital transfer (e.g., through loans, aid, and remittances) poor countries would be able to jump on the bandwagon of rapid economic development and industrialization. Internal and international labor migration was seen as integral parts of this process contributing to a more optimal spatial allocation of production factors and, hence, better aggregate outcomes.

In the same post‐war period, large‐scale labor migration from “developing” to “developed” countries started to gain momentum. Many labor surplus countries became involved in the migration process amidst expectations of the “dawning of a new era” (Papademetriou, 1985:212). Governments of developing countries, for instance in the Mediterranean, started to actively encourage emigration, which they considered as one of the principal instruments to promote development (Heinemeijer ; Adler, 1981; Penninx, 1982).

At the macro level, remittances were considered a vital source of hard currency. At the meso and micro level, migration was expected to lead to the economic improvement of migrant sending regions. Remittances would “improve income distribution and quality of life beyond what other available development approaches could deliver” (Keely and Tran, 1989:500). Moreover, it was expected that labor migrants or “guestworkers” would re‐invest substantially in enterprises in origin countries after their widely expected return. Migrant workers were seen as representing “a hope for the industrial development of their native land” (Beijer, 1970:102) and it was widely thought that “large‐scale emigration can contribute to the best of both worlds: rapid growth in the country of immigration… and rapid growth in the country of origin” (Kindleberger, 1965:253).

Although this optimism would diminish after 1970, several governments, particularly in Asia and the Pacific regions, have continued to see international migration as a major instrument of national economic development (Bertram, 1986, 1999; Fraenkel, 2006). A combination of migration, remittances, aid, and (government) bureaucracy [the so‐called “MIRAB” model (Bertram, 1986, 1999)] was expected to contribute to the economic take‐off of developing countries (Mckee and Tisdell, 1988:418; Hayes, 1991).

Pessimistic Views: Cumulative Causation and the “Migrant Syndrome”

As from the late 1960s, optimistic views were increasingly challenged under the combined influence of a paradigm shift in social and development theory towards historical‐structuralist and dependency (Frank, 1966, 1969) views as well as empirical studies and policy experiences that often did not support optimistic views (De Mas, 1978; Penninx, 1982). In fact, these new views turned the argument of neo‐classical and developmentalist approaches completely upside down: instead of decreasing, migration was now seen as increasing spatial (inter‐region and international) disparities in developmental levels.

The historical‐structuralist paradigm sees migration as a “flight from misery” caused by global capitalist expansion, which is therefore inherently unable to resolve the structural conditions that cause migration. Quite on the contrary, migration is seen as aggravating problems of underdevelopment. As Papademetriou (1985:211–212) argued, in sending countries, migration would contribute to “the evolution into an uncontrolled depletion of their already meager supplies of skilled manpower – and the most healthy, dynamic, and productive members of their populations.”

This coincided with increasing concern about the “brain drain.” Although many sending country governments have been comparatively positive towards the emigration of lower educated citizens, the attitude towards the emigration of skilled people has generally been more negative. It is perceived to deprive poor countries of their scarce skilled and professional labor resources in which states have invested many years of education (Baldwin, 1970). Also views on the development contribution of migration and remittances reversed, with the dominant vision becoming that remittances rather fueled consumption and inflation in origin regions and that migrants rarely invested their money in productive enterprises.

These pessimistic views seemed to fit particularly well into cumulative causation theory elaborated by Myrdal (1957). Cumulative causation theory holds that capitalist development is inevitably marked by deepening spatial welfare inequalities. Once differential growth has occurred, internal and external economies of scale (agglomeration and multiplier effects) perpetuate and deepen the bipolar pattern characterized by the vicious cycle of poverty in the periphery and the accelerated growth of the core region. So, economic activities in areas and countries with an initial advantage drain investment and encourage the out‐migration of the most talented populations from peripheral area and countries. Although positive “spread effects” also occur – such as increased demand for agricultural products and raw materials trade from the periphery (or remittances) – these do not match the negative “backwash effects.” Myrdal therefore argued that, without strong state policy, the capitalist system fosters increasing spatial inequalities.

Cumulative causation theory can be applied on national and international level, and obviously comes close to center‐periphery models and neo‐Marxist development theory. Thus, cumulative causation theory can be well applied to historical‐structuralist views on migration and development. Migration is expected to undermine regional and national economies by depriving them of their valuable human and material capital resources, which are exploited for the benefit of industrialized countries (international migration) and urban‐based capitalist elite groups within developing countries (internal migration) in need of cheap migrant labor. Migration undermines regional and local economies by depriving communities of their most valuable labor force, increasing dependence on core countries (of which remittances are but one manifestation) and stimulating further out‐migration. In this way, the productive structures at the origin would be progressively undermined, contributing to “asymmetric growth”– as opposed to the neo‐classical equilibrium model of factor price equalization – and the increasing underdevelopment and dependency of the underdeveloped on the developed core countries (cf. Almeida, 1973). In its turn, such pauperization is seen as encouraging further out‐migration. This also reveals an implicit assumption that migration is a more or less linear function of spatial opportunity disparities, underdevelopment and poverty.

Empirical studies conducted in migrant sending regions seemed largely to confirm these rather grim predictions of cumulative causation (for review articles, see Lewis, 1986; Lipton, 1980), corroborating the hypothesis that migration contributes to the “development of underdevelopment” instead of the reverse (Almeida, 1973; Rhoades, 1979; Lipton, 1980; Reichert, 1981; Rubenstein, 1992; Binford, 2003, Keely and Tran, 1989:501). Negative perspectives were often amalgamated into what Reichert (1981), called the “migrant syndrome,” or the vicious circle of:

Figure I summarizes the various negative feedback mechanisms through which migration is believed to increase, rather than decrease problems of underdevelopment and, hence, deepen inequalities between sending and receiving countries. Although the brain drain has attracted most attention, perhaps more relevant in the context of low skilled migration is the “brawn drain” (Penninx, 1982:793) – the large‐scale departure of young, able‐bodied men from rural areas (Lewis, 1986). This lost labor effect is typically blamed for causing a shortage of agricultural labor (Taylor, 1984) and decreasing agricultural productivity (Rubenstein, 1992:133). Moreover, migrants are typically believed to be talented young men who are the most significant agricultural innovators (Lipton, 1980:7+11). Likewise, other traditional economic sectors, such as craft industries, are expected to suffer from this lost labor effect.

Figure I

 Conceptual Framework of the “Migrant Syndrome” (Pessimistic Perspectives)

Furthermore, migration is believed to increase inequality with in migrant sending communities. Because migrants tend to be the already employed, more entrepreneurial, open‐minded, and relatively better educated people, remittances and other benefits of migration would also disproportionally accrue to the already better‐off (Lipton, 1980, Zachariah, Mathew, and Rajan, 2001). Therefore, migration will not contribute to poverty alleviation. The gradual undermining of traditional economies is even likely to increase the deprivation of the (non‐migrant) worst‐off.

Another widespread assumption in the migration and development literature is that migrants and their families do not invest their money productively but rather spend their money on “conspicuous consumption” such as imported consumer goods, and on so‐called non‐productive enterprises such as housing (Entzinger, 1985:268; Lewis, 1986). In his seminal review, Lipton (1980:12) concluded that recipients use remittances first to pay off debts incurred in financing migration or for education of their children. According to Lipton, more than 90 percent of remittances are spent on everyday consumption. Most consumption behavior serves to reinforce status, such as high payments for bride prices, feasts, funerals and the construction of pompous, luxurious houses. Remittances may also be directly used to finance the migration of family members (cf. Van Dalen, Groenewold, and Fokkema, 2005).

According to Lipton, investments only come in the fourth place of remittance use. Moreover, these were negatively evaluated as “consumptive investments”– a capital transfer more than capital creation – such as the purchase of land, the use of remittances to hire workers (e.g., for irrigation maintenance) where once family labor was used, or for labor‐replacing mechanization rather than the generation of extra output or the better use of scarce land inputs (for largely similar voices, see Rubenstein, 1992; Lewis, 1986; Zachariah, Mathew, and Rajan, 2001). Other studies mention a lack of creativity and innovation of migrant investors, which would render the establishment of typical labor migrant investments such as grocery shops, small restaurants, and trucks, or “second rank propositions in an overcrowded sector” (Penninx, 1982:802–803).

Within pessimistic frameworks, such “unproductive” expenses are usually thought to weaken local and regional economies and increase dependency. First, increased consumption and land purchase by migrants were reported to provoke inflatory pressures (cf. Russell, 1992) and soaring land prices (Appleyard, 1989; Rubenstein, 1992), from which the already poorer non‐migrants would suffer most – leading to more inequality. Second, many purchased items (e.g., TV sets, household appliances, refrigerators, stylish clothing and fabrics, building materials, ornaments, modern foodstuffs, fertilizers, etc.) would not be locally produced, but have to be imported from urban areas or from abroad. This is assumed to have the double effect of crowding out traditional, local production, and strengthening the economies of core areas, thereby intensifying the process of asymmetric growth and increasing regional disparities between the core and periphery. Third, the scarce productive investments by migrants would be mainly made in urban areas outside the village or region of origin (Lipton, 1980; Lewis, 1986). This leakage of remittance investments out of migrant sending areas further exacerbates regional disparities in wealth. This all corroborates the predictions of cumulative causation theory, according to which migration increases rather than decreases spatial inequalities.

Also the socio‐cultural effects of migration have usually received a bad press. Migration is usually believed to provoke consumerist, non‐productive and remittance‐dependent attitudes in migrant‐sending communities. The exposure to the wealth of (return) migrants and the goods and ideas they bring with them, would contribute to changing rural tastes (Lipton, 1980:12), lowering the demand for locally produced goods, increasing the demands for imported urban or foreign‐produced goods, and thereby increasing the general costs of living in sending communities. Migration is often held responsible for the disruption of traditional kinship systems and care structures (King and Vullnetari, 2006), the loss of community solidarity or the undermining of their “sociocultural integrity” (Hayes, 1991), and the breakdown of traditional institutions regulating village life and agriculture (De Haas, 1998). The exposure of rural youth to the relative wealth and success of migrants, combined with changing “urban” tastes and material aspirations, makes the rural way of life less appealing, discourage local people from working in traditional sectors, and encourage even more out‐migration. This would lead to a “culture of migration” (Massey , Heering, Van Der Erf, and Van Wissen, 2004), in which youth can only imagine a future through migrating, decreasing their willingness to work and build a future locally.

In sum, migratory cumulative causation theory postulates that migration deepens underdevelopment in migrant sending societies through various negative feedback mechanisms (backwash effects), which in its turn fuels further out‐migration, thereby perpetuating the vicious circle of the “migrant syndrome”. Put in Neo‐Marxist terms, migration not only reproduces but also reinforces the capitalist system based on class and spatial inequalities. The main positive effect of migration – the increase in family welfare for migrants themselves – is assumed to be only temporary and therefore artificial or “cosmetic” (Lewis, 1986). One‐sided dependency on migrant remittances is even considered dangerous, based on the assumption that remittances will rapidly decrease after migrants have returned or have settled and start to integrate in receiving societies, which would imply the gradual cutting of social and economic ties with origin societies.

A Critique of Deterministic Theories

In the 1970s and 1980s there was an expansion in the number of empirical micro‐studies on development impacts of migration conducted in various migrant sending countries in Latin America (mainly Mexico) and, to a lesser extent, the Mediterranean. Most studies tended to support pessimistic, historical‐structural views to varying degrees (cf. Almeida, 1973; Rhoades, 1979; Reichert, 1981; Park, 1992; Rubenstein, 1992). The influence of pessimistic views on migration and development has been enormous, and many of its views – in particular on migrants’ supposed inclination to spend money unproductively – have at least until very recently pervaded scholarly and, particularly, policy views on migration and development. Still, the idea of migration as a developing‐undermining, destabilizing product of poverty, as a problem which should be “solved” through restrictive immigration policies or aid and development programs, retain currency among academics, politicians, and the media.

However, the validity of these pessimistic views can be questioned because of their deterministic and circular nature and a logical inconsistency in their central arguments. First, the deterministic and self‐affirming nature of these theories does not leave any room for heterogeneity with regards to specific, localized migration impacts. For instance, they do not make plausible for what precise reasons would positive spread effects (e.g., remittances) not match negative backwash effects under certain conditions. They predict this outcome, but do not give a plausible explanation, and ignore empirical evidence that positive development impacts are possible at least under certain circumstances.

Second, there is an uncomfortable circularity in the idea that the vicious cycle of impoverishment in the periphery and growth at the core seems to go on ad infinitum. It seems unrealistic that there are no counter‐mechanisms which level‐off or change the nature of this process over time. In particular, how far can impoverishment go on without decreasing migration? At some point, it must lead to less migration because impoverishment will also decrease the proportion of people who are able to assume the costs and risks of migrating.

This brings us to the following criticism, which is the implicit but empirically unsubstantiated assumption that the relationship between development and levels of out‐migration is linear and inversely proportional. Empirical evidence rather suggests that this relationship is curvilinear and that development at least initially tends to coincide with rapid increases in migration rates because social and economic development enables and inspires people to migrate (De Haas, 2007b). In line with the mobility transition theory (Zelinsky, 1971) and the “migration hump” (Martin and Taylor, 1996), the relation between economic development and net emigration is J‐ or inverted U‐curve like, rather than linear and inversely proportional. In general, more developed societies tend to be more, not less, mobile (Skeldon, 1997).

The above critique enables us to identify an inherent logical contradiction between the two central arguments that migration pessimists make: on the one hand, migration is assumed to breed inequality because migrants come from better‐off groups within society. This is broadly consistent with empirical evidence. On the other hand, further impoverishment of the region of origin is expected to lead to more migration. This is logically inconsistent, as the first argument correctly supposes that a certain threshold of wealth needs to precede migration and the second argument supposes a negative‐linear relationship between wealth and migration. We have also to observe that neoclassical, “push–pull” and other place‐utility migration theories (erroneously) assume a negative linear relationship between sending country development and emigration, but at least apply this assumption consistently.

The fourth and final critique is empirical. An increasing body of empirical research has appeared in the 1980s and 1990s indicating that the development impacts of migration are fundamentally heterogeneous, and that, under favorable economic and political conditions, migration has played a positive role in the development of regions and countries of origin. For instance, in south‐European countries such as Spain and Italy and East Asian countries such as Malaysia and South‐Korea, migrants have often played a positive role in national development through remittances, investments, entrepreneurship as well as contributions to public debate and social change (cf. Massey ; Agunias, 2006; De Haas, 2007a).

So, the self‐reinforcing cyclical mechanisms of asymmetrical, polarizing development embodied in cumulative causation theory cannot be taken as axiomatic. So, looking back, does this all mean that the migration optimists were right after all? The likely answer is that neither the pessimists nor the optimists were right, as the heterogeneity of real‐life migration‐development interactions is too high to fit them into deterministic theoretical schemes predicting the development outcome of migration.

Papademetriou and Martin (1991) already argued that there is no automatic mechanism by which international migration leads to development. Although few would contest this observation, it adds little to our understanding of the factors explaining the heterogeneity of migration‐development interactions. To achieve this, the real challenge is to elaborate an appropriate theoretical framework that is refined enough to deal with the heterogeneity and complexities of migration‐development interactions, but that does not restrict itself to empiricism and “all is local and singular” relativism. This can only be done via systematic theoretical and empirical research which should “help us make sense of social structures and processes that never recur in the same form, yet express common principles of causality” (Tilly, 1984; in Skeldon, 1997:13). Unravelling principles determining the spatial and intertemporal heterogeneity of migration and development interactions should therefore be the core aim of analysis.

Findings from empirical work studies are clearly contradictory. In some cases, migration has a positive effect on the different dimensions of social and economic development, in other cases it seems to have no effect or even negative effects (De Haas, 2009). This cannot just pertain to differences in paradigmatic orientation – leading to different interpretations of similar empirical data – political ideology or methodology, but also relates to real, existing differences. Empirical research has highlighted that the spiraling down mechanisms of cumulative causation do not always hold true, but that the perfect neo‐classical world does not exist in reality either. Structural constraints such as highly unequal access to employment, markets, education and power do matter in the daily lives of many people in poor countries, and do severely limit their capability to overcome poverty and general underdevelopment. It would be unrealistic that migration alone would enable people to profoundly change structures.

So, discarding the rigidity of structuralist and neo‐Marxist approaches is not to say that structural constraints do not matter. While neo‐classical and developmentalist perspectives on migration and development tend to underestimate, structuralist perspectives tend to overestimate the importance of structural constraints and thereby also largely rule out agency. Hence, an improved theoretical perspective on migration and development has to be able to account for the role of structure – the constraining or enabling general political, institutional, economic social, and cultural context in which migration takes place – as well as agency – the limited but real capacity of individuals to overcome constraints and potentially reshape structure.

Pluralist views on migration and development interactions

Most empirical work from the late 1980s and 1990s increasingly acknowledged the heterogeneous, non‐deterministic nature of migration impacts on development. This corresponded with a general paradigm shift in contemporary social theory, away from grand theories towards more pluralist, hybrid approaches, which simultaneously take into account agency and structure. Social scientists, influenced by post‐modernist thinking and Giddens (1984) structuration theory, sought to harmonize agency and structure‐oriented approaches. Recognizing the relevance of both structure and agency is essential, as this enables us to better deal with the heterogeneity of migration‐development interactions. In such a “pluralist” approach, the results of the structure‐actor interactions allow for a greater variety of outcome than would have been allowed from either the single aggregation of individual decision making (Skeldon, 1997:18) or from the unidirectional imperatives of structures.

This general paradigm shift in social theory has also deeply affected the scholarly debate on migration and development. Over the 1980s and 1990s, the most crucial innovation to the debate came from the new economics of labor migration (NELM). Due to disciplinary divisions, it has remained unobserved that NELM has strong conceptual parallels with other “pluralist” strands in development thinking – the so‐called livelihood approaches – and sociological and anthropological research on migrants’ transnationalism. The following sections will review these three strands of literature and show how these can be integrated to provide a more nuanced perspective on reciprocal migration and development interactions, which integrates structure and agency perspectives, and gives sufficient analytical room for explaining the heterogeneous relationship between migration and wider development processes.

New Economics of Labor Migration (NELM)

In the 1980s and 1990s, the new economics of labor migration emerged as a critical response to neo‐classical migration theory (Massey ). The NELM theory rejects neo‐classical models, which largely ignore constraints and were evaluated as too rigid to deal with the diverse realities of the migration and development interactions. It was particularly Stark (1978, 1991) who revitalized thinking on migration in and from the developing world by placing the behavior of individual migrants in a wider societal context and by considering not the individual, but the family or the household as the most appropriate decision‐making unit. This new approach also increases the scope for integrating factors other than individual utility maximization as affecting migration decision‐making.

The new economics of labor migration models migration as risk‐sharing behavior of families or households. Better than individuals, households seem able to diversify their resources, such as labor, in order to minimize income risks (Stark and Levhari, 1982). The assumption is that people, households and families act not only to maximize income but also to minimize and spread risks. Internal and international migration can then be perceived as a household response to income risk, as migrant remittances provide income insurance for households of origin. This risk‐spreading motive can even explain the occurrence of migration in the absence of (expected) wage differentials. The idea is that for households as a whole it may be a Pareto‐superior strategy to have members migrate elsewhere, either as a means of risk sharing and/or as an investment in access to higher earnings streams (Lucas and Stark, 1985:902).

Migration is not only perceived as household risk spreading strategy, but also as a way to overcome various market constraints. The new economics of labor migration places the household in imperfect credit (capital) and risk (insurance) markets that prevail in most developing countries (Stark and Levhari, 1982; Stark and Bloom, 1985, Taylor, 1986; Taylor and Wyatt, 1996; Taylor, 1999). Such markets are often weakly developed or difficult to access for non‐elite groups. In particular through international remittances, migration can be a household strategy to overcome such market constraints by enabling households to invest in productive activities and to improve their welfare (Stark, 1980). While remittances are ignored in neo‐classical migration theory, within NELM they are perceived as one of the most essential motives for migrating. Conceptually, this also implies that the development contribution of migrants is not necessarily linked to return migration. Still‐abroad migrants, permanent settlers and their descendants can contribute to development by remitting money.

Besides providing a radically different conceptualization of migration as a household strategy to diversify risk and overcome market constraints, NELM also criticized the design of most prior empirical work. According to Taylor ,

prior work has been unduly pessimistic about the prospects for development as a result of international migration, largely because it has failed to take into account the complex, often indirect ways that migration and remittances influence the economic status of households and the communities that contain them.

Such criticism focused on the lack of analytical rigor, the prevalence of deductive reasoning over empirical testing, as well as the important methodological deficiencies of much prior empirical work. Many studies of migration impacts consist of non‐comparative remittance‐use studies that disregard income fungibility and the indirect, community‐wide impacts of migration (Taylor, 1999).

Migration as a Household Livelihood Strategy

New economics of labor migration has striking – though as yet unobserved – conceptual parallels with livelihood approaches. These have evolved as of the late 1970s among geographers, anthropologists, and sociologists conducting micro‐research in developing countries, who observed that the diverse and contradictory findings from their empirical work did not fit into rather rigid neo‐Marxist schemes. This made them argue that the poor cannot only be reduced to passive victims of global capitalist forces but try to actively improve their livelihoods within the constraining conditions they live in. This points to the fundamental role of human agency (Lieten and Nieuwenhuys, 1989).

A livelihood comprises the capabilities, assets (including both material and social resources), and activities required for a means of living (Carney, 1998). A livelihood encompasses not only the households’ income generating activities, but also the social institutions, intra‐household relations, and mechanisms of access to resources through the life cycle (Ellis, 1998). A livelihood strategy can then be defined as a strategic or deliberate choice of a combination of activities by households and their individual members to maintain, secure, and improve their livelihoods. This particular choice is based on (selective) access to assets, perceptions of opportunities, as well as aspirations of actors. Since these differ from household to household and from individual to individual, livelihood strategies tend to be so heterogeneous.

The emergence of the livelihood concept has meant a departure from rather rigid and theoretically deductive historical‐structuralist views towards more empirical approaches. This went along with the insight that people – generally, but all the more in the prevailing circumstances of economic, political and environmental uncertainty and hardship – organize their livelihoods not individually but within wider social contexts, such as households, village communities, and ethnic groups. For many social settings, the household was recognized as the most appropriate unit of analysis (McDowell and De Haan, 1997:3).

In this context, migration has been increasingly recognized as one of the main elements of the strategies households employ to diversify, secure, and, potentially, durably improve, their livelihoods. This is often combined with other strategies, such as agricultural intensification and local non‐farm activities (McDowell and De Haan, 1997; Bebbington, 1999; Ellis, 2000). It has increasingly been recognized that migration is often more than a short‐term survival strategy by rural populations, who were uprooted by global capitalist forces and more or less forced to join the ranks of a new international proletariat. Rather, empirical work suggested that migration is often a deliberate decision to improve livelihoods, enable investments (Bebbington, 1999:2027), and help to reduce fluctuations in the family income that often used to be largely dependent on climatic vagaries (McDowell and De Haan, 1997:18; De Haan ). Migration can then be seen as a means to acquire a wider range of assets which insure against future shocks and stresses (De Haan ). Although this has been mainly applied for rural‐urban internal migration in poor countries, there is no a priori reason why this diversification‐through‐migration argument cannot also be extended to international migration and urban households.

This comes strikingly close to the premises of NELM. Both approaches can be easily integrated if we see internal as well as international migration as part of a broader household livelihood strategy to diversify income and overcome development constraints in the place of origin. There is also a striking similarity in how over the 1970s and 1980s structuralist and functionalist views of migration converged towards more pluralist views recognizing the relevance of both agency and structural constraints. While livelihood approaches meant for many sociologists and anthropologists a departure from rather rigid structuralism towards an increasing recognition of the role of agency, economists and other scholars reasoning from neo‐classical approaches moved in the opposite direction. In particular, the emergence of NELM marked a departure from neo‐classical and actor‐oriented approaches towards a household‐level based perspective which explains migration from the structural constraints and imperfect markets within which migration decisions are made.

New economics of labor migration adopted a household‐oriented approach that was already common in other fields of social science, a fact that was explicitly acknowledged by Lucas and Stark (1985:901), who stated that economists have begun to address questions of household composition more traditionally posed by anthropologists and sociologists. They therefore proposed to

extend the recent intergenerational view of the household to a spatial dimension.... and dualistic theories of development must be revised: Instead of an urban sector and a rural sector, each with its own populace benefiting from the sectoral‐specific speeds of development, the family straddles the two. Classes cease to be only peasants and workers, and a hybrid peasants‐worker group emerges. This perception is not new to anthropologists but has not previously been integrated with the economics of the household (Lucas and Stark, 1985:915).

So, within a household perspective it is not either migration or activities at the origin, but often both. This also indicates that the impact of a migration strategy cannot be properly evaluated outside its relationship with other multi‐sectoral and multi‐local livelihood strategies, that is, the entire “portfolio” of household activities (Stark, 1991). Research attempting to isolate migration and migrants from their wider social and economic context is therefore not able to assess the relation between migration and broader transformation processes embodied in the term development.

Internal and international migrants tend to maintain close links with their communities of origin over much longer periods than has previously been assumed (McDowell and De Haan, 1997:1). This also exemplifies that the development contribution of migration is not necessarily linked to the return of migrants. Migration and economic activities at the origin are not mutually exclusive, but are in fact often combined. Without a household approach, such multiple strategies cannot be captured. This view, which is shared by both NELM and livelihoods approaches seems to better reflect the realities of daily life for millions of migrants in developing countries than neo‐classical or structuralist approaches.

The choice of the household as the primary unit of analysis can be seen as a kind of optimum strategy or a compromise between agency and structure approaches, acknowledging that the forms of households vary across time, space, and social groups. In perceiving migration as a household livelihood strategy, we acknowledge that structural forces leave at least some room for agency, although at highly varying degrees. Household approaches seem particularly applicable in developing countries where for many people it is not possible to secure the family income through private insurance markets or government programs (Bauer and Zimmermann, 1998), increasing the importance of implicit contracts within families and communities.

A Transnational Perspective on Migration and Development

The rise of new economics and livelihood perspectives on migration and development have coincided with a third trend in migration studies, that is, the “transnational turn” in the study of the settlement and integration of migrant communities in receiving countries (Glick Schiller, Basch, and Blanc‐Szanton, 1991, Castles and Miller, 2009; Faist, 2004). There has been growing recognition of the increased possibilities for migrants and their families to live transnationally and to adopt transnational identities (cf. Vertovec, 1999; Guarnizo, Portes, and Haller, 2003). This relates to the improved technical possibilities for migrants to foster links with their societies of origin through the (mobile) telephone, fax, (satellite) television and the internet, and to remit money through globalized formal or informal banking systems. This increasingly enables migrants and their families to foster double loyalties, to travel back and forth, to relate to people, and to work and to do business simultaneously in distant places. It is true that also 19th and early 20th century migrants kept intensive transnational ties, but it is likely that technological revolutions have substantially increased the scope for migrants and their families to pursue transnational livelihoods on a more constant, day‐to‐day basis.

This transnationalization of migrants’ lives has challenged assimilationist models of migrant integration, as well as the modernist political construct of the nation‐state and citizenship. The implication is that clear‐cut dichotomies of “origin” or “destination” and categories such as “permanent,”“temporary,” and “return” migration are increasingly difficult to sustain in a world in which the lives of migrants are characterized by circulation and simultaneous commitment to two or more societies or communities (De Haas, 2005).

This has fundamental implications for the study of migration and development, because this implies that integration in receiving societies and commitment to origin societies are not necessarily substitutes, but can be complements. It has long been assumed that migrants’ integration would coincide with a gradual loosening of ties with societies of origin and that “permanent” migration would therefore inevitably represent a “loss” or “drain”. This assumption explains much of the prior pessimism on the sustainability of remittances and the idea that migrants’ contribution to development in origin countries is strongly linked to return migration. However, empirical studies have indicated that migrants may maintain strong transnational ties over sustained periods and that these ties can even become trans‐generational. They also show that migrants’ engagement with origin countries is not conditional on their return, but can be maintained through remitting money and ideas, telecommunications, holiday visits and pendular migration patterns.

The sustainability of transnational ties is exemplified by persistent remittances, transnational marriages and the involvement of migrants in social, cultural and political affairs of their origin countries. It seems therefore incorrect to automatically interpret migrants’ commitment towards their countries of origin as a manifestation of failed integration. Conversely, migrants’ deeper involvement in their receiving societies does not necessarily lead to less significant commitment to their countries of origin. The reverse is also possible (Snel, Engbersen, and Leerkes, 2006). After all, successfully “integrated” migrants also have increased financial and human resources that potentially enable them to set up enterprises or participate in public debate in origin countries.

It is evident that insights from transnationalism studies have many parallels with and complement NELM and livelihood approaches. They can be combined if we conceptualize international migration as an integral part of transnational livelihood strategies pursued by households and other social groups. Return visits and return migration, remittances, transnational business activities as well as investments and political involvement in origin countries are all expressions of the transnational character of a migrant’s life. The fact that migrants often maintain long‐term ties with origin countries and that integration does not necessarily preclude or can even encourage such transnational engagement, casts doubt on the assumption that the departure of migrants would automatically represent a loss in the form of a brain or brawn drain.

Empirical evidence

Over the previous decades, a growing number of empirical studies have countered pessimistic views on migration and development. Earlier (1996a, 1996b) and more recent (cf. Agunias, 2006; Katseli, Lucas, and Xenogiani, 2006, Rapoport and Docquier, 2005; Özden and Schiff, 2005; De Haas, 2007a) reviews of the literature have pointed to the potentially positive role of migrants and remittances in social, economic and political transformation processes in societies and communities of origin. Largely in line with NELM and livelihood approaches, the evidence reviewed in the above publications supports the view that migration is rather a deliberate attempt by social groups (typically, but not exclusively, households) to spread income risks, to improve their social and economic status and, hence, to overcome local development constraints. Particularly, remittances are an expression of strong transnational social bonds and of the wish to improve the lives of those left behind. However, the accumulated evidence also exemplified the fact that migration and remittances are no panacea to overcome structural development constraints (Taylor, 1999).

International remittances generally help to diversify and also to substantially increase household income. They have a crucial insurance function in protecting people from the destabilizing effects of absent or ill functioning markets, failing state policies and a lack of state‐provided social security. On the national level, there is substantial evidence that remittances have proved to be an increasingly important, less volatile, less pro‐cyclical, and therefore a more reliable source of foreign currency than other capital flows to developing countries. However, this does not necessarily imply that they contribute to poverty alleviation. As migration is a selective process, most international remittances do not tend to flow to the poorest members of communities nor to the poorest countries. However, poor non‐migrant families are often affected indirectly through the economy‐wide effects of remittance expenditure on wages, prices and employment in migrant sending communities. Therefore, most studies conclude that remittances reduce poverty, albeit to a limited extent.

The effect of migration and remittances on income inequality in migrant sending communities is more ambiguous, because this depends fundamentally on the varying and changing selectivity of migration. Pioneer migrants tend to be from relatively wealthy households, and migration and remittances therefore often initially reinforces inequality. However, in later stages selectivity can decrease, primarily due to the establishment of migrant networks, which, ceteris paribus, reduce the costs and risks of migration. As a consequence of this diffusion process, the initially negative effect of remittances on income equality might therefore be dampened or even reversed.

Several studies also indicated that remittance receiving households often have a higher propensity to invest than non‐migrant households. Moreover, they indicate that consumptive expenses and so‐called “non‐productive investments” such as on housing can have highly positive multiplier effects in local and regional economies, which generate employment and income for non‐migrants and can contribute to poverty reduction. This coincided with criticism on arbitrary definitions of what actually constitutes “productive investments”, which in its turn reflects rather narrow views on what actually constitutes development. If we adopt a broader, capabilities‐focused perspective on human development as proposed by Sen (1999)– who defined development as the process of expanding the substantive freedoms that people enjoy – expenditure in areas such as education, health, food, medicines and housing, as well as community projects in education, health and recreational facilities should be seen as developmental as long as they enhance people’s wellbeing and capabilities. However, the extent to which migrants invest crucially depends on the selectivity of migration as well as the more general development conditions in regions of origin. These ultimately determine the extent to which migrants are compelled to invest in, to continue or rather to withdraw from social and economic activities in origin countries.

Also the universal validity of the brain drain hypothesis has been increasingly questioned, making room for a much more nuanced picture. Not all migrants are highly skilled and the brain drain seems to be truly massive only in a minority of, generally small and/or very poor, countries. Furthermore, a brain drain can be accompanied by a significant brain gain, because the prospect of moving abroad may stimulate the incentive to study among stay‐behinds (Worldbank, 2005; Stark, Helmenstein, and Prskawetz, 1997; Lowell and Findlay, 2002). However, this only seems to occur if the opportunity to migrate increases the economic returns to education. Therefore, migration can also create negative incentive structures for education in cases of low skilled, often irregular migration, where few if any positive returns on education can be expected (Mckenzie, 2006). Although migrants often play an important and positive role in the civil society in countries of origin, they may also contribute to sustained conflicts, for instance by providing support for warring parties (Nyberg‐Sorensen, Van Hear, and Engberg‐Pedersen, 2002; Van Hear, 2004).

Figure II summarizes the accumulated insights into the various mechanisms through which migration can affect development in migrant regions areas in the short to medium term. This conceptual framework combines insights from the NELM livelihood approaches and transnationalism studies while casting the concept of development within a capability framework as developed by Sen. This conceptual framework is pluralist because it emphasizes the contextual conditionalities of migration impacts on development, exemplifying that the degree to which the development potential of migration is exploited fundamentally depends on the more general investment environment. Although migration is often a strategy to overcome local development constraints, it is unlikely that migration alone can solve more general constraints such as endemic corruption, misguided macro‐economic policies, credit and insurance market failure and insecure property rights.

Figure II

 Conceptual Framework of Pluralist Heterogeneous Migration and Development Interactions (Community Level)

Amendments to migration as a transnational household strategy

Although the presented “pluralist” perspectives seem more refined and realistic than the rather deterministic neo‐classical and structuralist views, they can be criticized for their focus on households as well as labor migration, plus a certain bias towards transnationally active migrants. This critique can be used to amend the conceptual framework elaborated above in order to make it less rigid and also applicable for other forms of “non‐labor” migration.

Firstly, although household approaches seem the best compromise to harmonize agency and structure approaches, this involves the risk of reifying the household, when it comes to be seen as a unit with a clear will, plans, strategy, and aims (Lieten and Nieuwenhuys, 1989:8). Criticism on household approaches has focused on the underlying assumption of households as monolithic, altruistic units taking unanimous decisions to the advantage of the whole group (Rodenburg, 1997; Carling, 2005). This may mask intra‐household age, gender and other inequalities, and can also disguise the importance of migration‐relevant social bonds with non‐household family, community members and friends. It also rules out agency of individual household members and, hence, their potential ability to revolt against the will of powerful household members by, for instance, migrating without consent.

Second, there is a problematic circularity in the way in which NELM and livelihood approaches tend to link initial migration motives and strategies to consequences of migration. The direct link that, in particular, NELM draws between motives of migration and the act of remitting, is often more unsettled in practice (Lindley, 2007). For instance, a person migrating abroad with the intention to earn money to allow her household to invest in a private enterprise might end up not doing so because of political or economic crises in origin countries or because her transnational family bonds weaken more rapidly than anticipated.

Through their common bias towards transnationally active migrants, case‐study based empirical work on transnationalism does often not pay sufficient attention to counterfactual cases of migrants following a more classical path of assimilation and fading of transnational ties (Guarnizo, Portes, and Haller, 2003). The other way around, a refugee who primarily migrates to escape life‐threatening circumstances, may end up remitting substantial amount of money or become a transnational entrepreneur (Lindley, 2007). The same can be applied to student migrants who might intend to return after graduating, but who often end up working and remitting money.

This reflects the more fundamental problem that conventional categories used to classify migrants (e.g., economic, refugee, asylum, family, student) primarily reflect bureaucratic and legal categories and conceal the often complex, mixed and shifting motivations of migrants. For all these reasons, it would be preferable to remove the “L” from “NELM” and to extend this theory to almost all forms of migratory mobility. This would also acknowledge the fact that migration is not necessarily a preconceived “strategy” to improve livelihoods through investing. Finally, it de‐links initial migration intention from eventual development consequence. This is another reason to broaden our concept of development towards Sen’s capabilities perspective discussed above. This enables us to go beyond economic interpretations or a narrow focus on labor migration and to perceive migration within a broader framework of (economic, social or political) opportunity rather than income differentials.

Contextualizing migration‐development interactions

The new economics of labor migration and livelihood as well as transnational approaches towards migration can all be situated within a broader paradigm shift in social theory towards approaches attempting to harmonize agency and structure. This leads to a more optimistic assessment of the development potential of migration and points to the ability of individuals and households to overcome structural development constraints through migrating in a deliberate attempt to diversify, secure and improve their livelihoods.

However, the significant empirical and theoretical advances that have been made over the past decades highlight the fundamentally heterogeneous nature of migration‐development interactions as well as their contingency on spatial and temporal scales of analysis, which should forestall any blanket assertions on this issue. To understand this heterogeneity, we need to study these migration‐development interactions in the development context of which they are an intrinsic part. Migration is not an independent variable “causing” development (or the reverse), but is an endogenous variable, an integral part of change itself and a factor that may enable further change. This is why it is more correct to refer to the reciprocal relationship between migration and broader development processes instead of the – one‐way –“impact” of migration on development.

Figure III depicts this reciprocal nature of migration and development interactions. When analysing the factors which underlie the differentiation in migration and development relationships, a distinction can be made between (I) the development context at the general, macro (national, international) level; (II) the development context at the local or regional level; and (III) the factors related to the migrant and her direct social and economic environment on the household, family and community level. These three sets of variables are mutually linked through various direct functional relations and feedback mechanisms.

Figure III

 General Conceptual Framework for Analyzing Migration‐development Interactions

The macro‐level development context – the above‐regional (national, international) whole of political, social, and economic structures – partly determines the local development context, for instance through public infrastructure, policies, social facilities, legislature, taxation, market access or regional development programs.

The macro‐context also largely determines the extent to which there are opportunities to migrate, either internally or abroad, for instance through immigration policies, labor demand and, income levels. Such opportunity structures affect the magnitude, nature (undocumented, legal, labor, political, family), and the (initial) selectivity of migration.

The local development context determines to what extent people are able to lead lives they have reason to value and to enhance the substantive choices they have (following Sen’s definition) through local livelihood activities. The extent to which they perceive this is possible affects their (1) aspiration to migrate. A second way through which the local development context affects the propensity to migrate is the influence of development on the (2) capability to migrate through drawing on (a) financial/material, (b) social and (c) human capital. Thus, people’s propensity to migrate is seen as a function of their aspirations and capabilities to do so; and migration may therefore increase as long aspirations increase faster than local livelihood opportunities.

In their turn, migration processes affect the local development context through their effects (system feedbacks) on labor supply, consumption, investments, inequality, social stratification, relative deprivation, local culture and aspirations (for more specification, see Figure II). As the above review has shown, the nature of these impacts is spatially heterogeneous, and is contingent on the characteristics of the local development context as set by the behavior of previous actors. In their turn, such migration‐induced processes of social and economic change affect people’s (1) aspirations (for instance through increasing inequality and relative deprivation) and (2) capabilities to migrate (arrow c), while the creation of social capital through the formation of migrant networks tends to facilitate additional movement.

Changes in the local development context – for instance as the result of migration – may eventually affect the macro‐level development context, albeit to a limited extent, because of the limited magnitude of migration and remittances and the predominantly individual, family and community character of migration.

The conceptual embedding of the specific analysis of localized migration impacts into the broader development context at the macro‐level helps to understand the heterogeneity of migration impacts. The extent to which migration can contribute to regional, and even national development, fundamentally depends on the more general macro‐level development context. Micro‐empirical evidence highlighting the often positive role of migration and remittances in households’ livelihoods is often inaccurately taken as evidence that migration does stimulate development in more general terms and on the macro‐level. However, to argue from “migration and remittances durably improve households’ living standards” to “migration stimulates national development” is to commit a classical ecological fallacy by transferring inferences made on a micro‐level scale of analysis to a macro‐level scale of analysis.

General development is a complex and multifaceted process, involving and requiring structural social, political and institutional reform, which cannot realistically be achieved by individual migrants or remittances alone, and requires active state intervention. Notwithstanding their often considerable blessings for individuals, households and communities, migration and remittances are no panacea to solve more structural development problems. If states fail to implement general social and economic reform, migration and remittances are unlikely to contribute in nation‐wide sustainable development (Gammage, 2006, Taylor, Moran‐Taylor, and Ruiz, 2006). Therefore, migrants and remittances can neither be blamed for a lack of development, nor be expected to trigger take‐off development in generally unattractive investment environments.

As Heinemeijer already observed, development in migrant‐sending regions is therefore a prerequisite for investment by migrants rather than a consequence of migration. While stressing the developmental potential of migration, the now substantial body of empirical evidence also highlights the complexity, heterogeneity and socially differentiated nature of migration‐development interactions. This provides a warning against recent optimistic views on migration and development by pointing at the real but fundamentally limited ability of individual migrants to overcome structural constraints and, hence, the paramount importance of the more general development context in determining the extent to which the development potential of migration can be realized.

Depending on this broader context, migration may enable people to retreat from, just as much as to engage and invest in, social, political and economic activities in origin countries. It is the very capabilities‐enhancing potential of migration that also increases the freedom of migrants and their families to effectively withdraw from such activities. However, if development in origin countries takes a positive turn, if trust in governments increases and economic growth starts to take off, migrants are likely to be among the first to join in and recognize such new opportunities, reinforcing these positive trends through investing, circulating and returning to their origin countries. Such mutually reinforcing migration‐development processes seem to have occurred in several former emigration countries as diverse as Spain, Taiwan, South Korea, and, recently, Turkey.

Conclusion

The preceding analysis has exemplified that discursive shifts in the scholarly and policy debate on migration and development reflect more general paradigm shifts in social and development theory. It has frequently been argued that it is possible to combine and integrate different theoretical perspectives on migration (1993, 1998). However, attempts to combine different theoretical perspectives are more problematic than sometimes suggested. As Kuhn (1962) argued, proponents of different paradigms live in different worlds, use different vocabularies, and use different criteria determining the legitimacy both of problems and of proposed solutions in terms of methodology and analysis. Each paradigm therefore has the tendency to satisfy the criteria it sets for itself and to reject the very problem definition as well as evaluation criteria used by other paradigms (Kuhn, 1962:109).

Because of this circularity, there are no objective, “scientific” criteria against which to externally examine the superiority of competing paradigms, the evaluation of which therefore inevitably involves non‐scientific values. This is for instance evident in the diametrically opposed analysis of “dependency”, in neo‐Marxist and NELM approaches. Ultimately, differentiating valuations of migration in its reciprocal relation to development are strongly related to differentiating a priori assumptions about what actually constitutes development. The above study of evolution of migration and development theory does corroborate Kuhn’s position, in the sense that progress has not been gradual and cumulative but rather a “revolutionary” process in which this field of study has been re‐conceived three times based on new theoretical and methodological fundamentals.

This raises the more fundamental question whether the recent shift towards highly optimistic views in policy but also academic circles reflects a veritable change in migration‐development interactions, the use of other methodological and analytical tools, or is rather the deductive echo of a general paradigm shift in research and policy away from dependency and state‐centrist to neo‐classical and neo‐liberal views in general. In social science, structuralist theory has become increasingly discredited. This has led to less negative interpretations of dependency and a more positive value being attributed to the global incorporation of regions and countries in the developing world, a process of which migration is an integral part. However, a bias towards migration and development success stories might obscure situations in which migration did contribute to worsening underdevelopment.

It is important to note that current optimism on the development potential of migration and development also has a strong ideological dimension, as it fits very well into (neo) liberal political philosophies. On a critical note, Kapur (2003) has pointed to the ideological roots of recent remittance euphoria. He argued that remittances strike the right cognitive chords, and fit in with a communitarian, “third way” approach, exemplifying the principle of self‐help: “Immigrants, rather than governments, then become the biggest provider of “foreign aid” (Kapur, 2003:10).

This shows a real danger that ignorance or neglect of previous empirical and theoretical work leads to uninformed and, hence, naïve optimism somehow reminiscent of earlier developmentalist beliefs in migration and development. If anything, the accumulated empirical and theoretical evidence stresses the fundamentally heterogeneous nature of migration‐development interactions and in particular their contingency on more general development conditions. In other words, structure matters.

Public policies which improve the functioning of social, legal, economic and political institutions, the access of ordinary people to basic amenities and markets and which restore trust in governments, are crucial not only for creating a fertile ground for development in general, but also for compelling more migrants to invest and/or return in origin countries. Policy and scholarly discourses celebrating migration, remittances and transnational engagement as self‐help development “from below” shift attention away from structural constraints and the real but limited ability of individuals to overcome these. This exemplifies the crucial role states continue to play in shaping favorable general conditions for human development to occur.