Theorising Return Migration: The Conceptual Approach to Return Migrants Revisited

The attention paid by international organisations to the link between migration and development in migrants(cid:146) origin countries has highlighted the need to revisit approaches to return migration. Moreover, the growing diversity of migratory categories (ranging from economic migrants to refugees and asylum seekers) necessitates a distinction between the various types of returnee. We still need to know who returns when, and why; andwhy some returnees appear as actors of change, in specific social and institutional circumstances at home, whereas others do not. The first objective of this paper is to analyse how return has been dealt with by international migration theories, emphasising particularly the assumptions on which they rest. This theoretical overview is necessary to show how return has been defined and located in time and space, andhow the returnee has been depicted. The second objective is to take the various approaches to return migration a step further by elaborating on the theoretical insights that have been extensively proposed. The conceptual approach to returnees is then revisited, taking into account a set of distinguishing criteria,

The at t ent ion paid by int ernat ional organisat ions t o t he link bet ween migrat ion and development in migrant s' origin c ount ries has highlight ed t he need t o revisit approac hes t o ret urn migrat ion.Moreover, t he growing diversit y of migrat ory c at egories ( ranging f rom ec onomic migrant s t o ref ugees and asylum seek ers) nec essit at es a dist inc t ion bet ween t he various t ypes of ret urnee.We st ill need t o k now who ret urns when, and why; andwhy some ret urnees appear as ac t ors of c hange, in spec if ic soc ial and inst it ut ional c irc umst anc es at home, whereas ot hers do not .The f irst obj ec t ive of t his paper is t o analyse how ret urn has been dealt wit h by int ernat ional migrat ion t heories, emphasising part ic ularly t he assumpt ions on whic h t hey rest .This t heoret ic al overview is nec essary t o show how ret urn has been def ined and loc at ed in t ime and spac e, andhow t he ret urnee has been depic t ed.The sec ond obj ec t ive is t o t ak e t he various approac hes t o ret urn migrat ion a st ep f urt her by elaborat ing on t he t heoret ic al insight s t hat have been ex t ensively proposed.The c onc ept ual approac h t o ret urnees is t hen revisit ed, t ak ing int o ac c ount a set of dist inguishing c rit eria, i.e. t he ret urnee' s "preparedness" and "resourc e mobilisat ion".
A s a subprocess of international migration, return migration has been subj ect to various approaches that offer contrasting sets of propositions stemming from neoclassical economics, the new economics of labour migration, structuralism, transnationalism and social network theory.From a qualitative point of view, numerous empirical inquiries have been carried out to better illustrate the multifarious factors that have made return migration a multifaceted and heterogeneous phenomenon.Although return migration has long been subj ect to various interpretations, our understanding of it remains haz y.Not so much because it has been neglected by migration scholars -analyses of return migration have in fact been legion since the 1960s -but rather because its magnitude and configuration are scarcely measurable and comparable, owing to the lack of reliable large-scale quantitative data.

254
Today, the attention paid by international organisations to the link between migration and development has highlighted the need to revisit approaches to return migration.Moreover, the growing diversity of migration categories (ranging from economic migrants to refugees and asylum seekers) necessitates a distinction between the various types of returnee.W e still need to know who returns when, and why; and why some returnees appear as actors of change, in specific social and institutional circumstances at home, whereas others do not.
As a prerequisite to revisiting the conceptual approach to the profiles of returnees, by taking into account a set of distinguishing criteria, the first objective of this paper is to analyse how return has been dealt with by international migration theories, emphasising particularly the assumptions on which they rest.This theoretical overview is necessary to show how return has been defined and located in time and space, and how the returnee has been depicted.
The second objective is to take the various approaches to return migration a step further by using and elaborating on the theoretical insights that have been extensively proposed.The conceptual approach to returnees is then revisited through a set of distinguishing criteria, i.e. the returnee' s " preparedness" and " resource mobilisation" .These criteria are subject to examination in the development of this study.

Theoretical Overview of Return Migration
W hile scholarly approaches related to return migration can be traced back to the 1960s, there is no question that, with hindsight, it was in the 1980s that stimulating scientific debate among scholars took place on the return phenomenon and its impact on origin countries.These debates culminated in the production of several volumes and critical essays, and in the organisation of conferences (Kubat 1984;Council of Europe 1987).Moreover, they contributed intensively to the development of the literature on return migration, together with the growing concern over " co-development" , the " voluntary repatriation of third-country nationals" , the emergence and implementation of bilateral readmission agreements between sending and receiving countries, and the link between international migration and economic development in migrants' origin countries.
It has to be said that the increasing variety of scholarly analyses, together with the resilient politicisation of international migration movements, have been incidental to the ways in which return migration and returnees have been understood and analysed.Oddly enough, just as Mary Kritz noted (1987,948), there exist conceptual problems regarding the definitions of the immigrant -such definitions having a bearing on the formulation of national immigration policies -there also exist several definitional approaches to return migration, and to returnees that are playing a crucial role in orienting, if not shaping, the perceptions, taxonomies and policies adopted by governmental and intergovernmental agencies.
As a prerequisite to exploring how return has been addressed by international migration theorists, it is important to stress that the theoretical insights discussed below have, in various degrees, included return migration as a subcomponent of their analytical approaches.Whereas some of these insights are the outcome of empirical studies, others stem from the collection of fragmented official quantitative data, based on given definitional criteria of the returnee.
This critical review focuses exclusively on theories that have attempted to propose a set of variables aimed at better understanding the magnitude and dynamics of return migration to origin countries.Whatever their views and interpretations, all the theories presented below yield valuable insights.They do so in so far as they differ in terms of level of analysis and with respect to the salience of the issue of return in their respective analytical frameworks.

Neoclassical Economics and the New Economics of Labour Migration
In so far as the neoclassical approach to international migration is based on the notion of wage differentials between receiving and sending areas, as well as on the migrant's expectations for higher earnings in host countries (Todaro 1969, 140), return migration seems to be viewed as the outcome of a failed migration experience which did not yield the expected benefits.In other words, in a neoclassical stance, return migration exclusively involves labour migrants who miscalculated the costs of migration and who did not reap the benefits of higher earnings.Return occurs as a consequence of their failed experiences abroad or because their human capital was not rewarded as expected.Furthermore, unlike the new economics of labour migration (NELM, see below), the neoclassical economics of migration views migrants as individuals who maximise not only their earnings but also the duration of their stay abroad to achieve permanent settlement and family reunification.In this framework of analysis, return cannot but be motivated by a failed migration experience, in terms of expected earnings, employment and duration.
Conversely, while the neoclassical approach to return migration argues that migrants did not successfully maximise their expected earnings, NELM views return migration as the logical outcome of a "calculated strategy", defined at the level of the migrant's household, and resulting from the successful achievement of goals or target.In fact, as Oded Stark's seminal book argues, the NELM approach "shifts the focus of migration theory from individual independence … to mutual interdependence" (Stark 1991, 26), i.e. at the level of the family or the household.Moreover, it views return as the natural outcome of a successful experience abroad during which migrants met their goals (i.e. higher incomes and accumulation of savings) while naturally remitting part of their income to the household.Remittances are part and parcel of a strategy aimed at diversifying the resources of the household with a view to better compensating for the risks, linked to the absence of an efficient insurance market in home countries.They also constitute one explanatory factor in the return decision, together with the attachment to the home country.
With reference to remittances and the likelihood of return, Amelie Constant and Douglas Massey, by analysing data relating to the return migration of guestworkers in Germany from 1984 to 1997, observed that remitters have higher rates of employment in receiving countries and that having a spouse in the home country increases their likelihood of return (Constant and Massey 2002, 27-8).In contrast to the neoclassical model, return migration appears to Oded Stark and his followers (Taylor 1996) as being part of a well-prepared migration project which shapes not only the propensity of migrants to "exert a higher level of work effort [ in receiving countries] than that exerted by native-born workers" (Stark 1991, 392) and to save more money than native-born workers (Stark and Galor 1990), but also their level of socialisation in receiving countries as well as their incentives to gain additional skills or on-the-job training.In other words, skills acquired abroad are viewed as being affected by the probability of return.In fact, according to NELM, migrants go abroad for a limited period of time, until they succeed in providing their households with the liquidity and income they expect to earn.The planning of the migration project has a bearing on the behavioural patterns of the migrant in the host society, as well as on professional advancement.
The neoclassical economics and NELM approaches differ in so far as they posit contrasting sets of interpretations regarding return migration.When neoclassical economists argue that people move permanently to raise and maximise their wages in receiving countries, return migration is viewed as a failure, if not an anomaly.
When NELM contends that people move on a temporary basis to achieve their goals or targets in receiving countries, as a prerequisite to returning home, return migration is viewed as a success story, if not a logical outcome.NELM theorists are adamant about breaking away from the neoclassical image of the failed returnee.The duration of stay abroad is calculated with reference to the needs of the household, in terms of insurance, purchasing power and savings.Once such needs are fulfilled, return migration occurs.In other words, the NELM approach to return migration goes "beyond a response to negative wage differential" (Stark 1996, 11).
There is no question that, despite their contrasting interpretations of return migration, both theoretical schemes give valuable insights regarding the reasons for which people move abroad and return home.Migrants have clearly defined projects or strategies before, during and after their migration experiences.In fact, the above-mentioned "calculated strategy" is for NELM a way of stressing the fact that the migration decision can no longer be viewed "as an act of desperation or boundless optimism" (Stark 1996, 26).Whether they are faced with market failures at home or with the need to compensate for wage differentials between their countries of origin and their areas of destination, migrants plan and try to respond to market uncertainties.
Nonetheless, as far as their analytical frameworks are concerned, both theoretical approaches have several shortcomings.The first relates to the actors themselves and their motivations for return.These seem to be determined by financial or economic factors only, while providing little explanation of how remittances and skills are used in home countries.The second pertains to the fact that returnees are exclusively viewed as foreign-income bearers or "financial intermediaries", as Edward Taylor would put it.Moreover, while neoclassical economics and NELM try to explain when and why the decision to return home takes place, there is virtually no reference to where migrants return.As no reference is made to their social, economic and political environment at home, return experiences seem isolated from each other.Not only do we not understand how the strategies are planned and reshaped when return takes place, but the interaction beyond the returnee's family or household remains unaccounted for.Finally, as shown in the following section, several empirical studies have convincingly demonstrated that the success/ failure paradigm cannot fully explain the return migration phenomenon.This paradigm tends in fact to isolate the decisions and strategies of the returnees from their social and political environment, without correlating them with contextual factors at home.
As the structural approach to return migration contends, return is not only a personal issue, but above all a social and contextual one, affected by situational and structural factors.

The Structural Approach to Return Migration
There is no surprise in mentioning that the empirical findings and theoretical insights produced by anthropologists, sociologists and social geographers have contributed greatly to refining the structural approach to return migration.
This structural approach argues that return is not solely analysed with reference to the individual experience of the migrant, but also with reference to social and institutional factors in countries of origin.In fact, return is also a question of context.
Just like NELM, the structural approach to return migration shows how crucial to the return decision and the reintegration of the migrant are the financial and economic resources brought back to origin countries.Returnees' success or failure is analysed by correlating the "reality" of the home economy and society with the expectations of the returnee.Francesco Cerase's influential article on Italian returnees from the United States provides many emblematic examples of how complex the relationships between the returnee's expectations and the social and economic context (i.e."reality") at home are.Cerase identifies four different types of returnee, emphasising their aspirations, expectations and needs: "Return of failure" pertains to those returnees who could not integrate in their host countries owing to the prejudices and stereotypes they encountered abroad.
Their difficulties in taking an active part in the receiving societies or in adapting themselves to host societies were strong enough to motivate their return.
"Return of conservatism" includes migrants who before emigrating had planned to return home with enough money to buy land with a view to "liberating themselves from loathsome subjection to the landowners" (Cerase 1974, 254).Because of these aspirations and strategies, conservative returnees only tend to satisfy their personal needs, as well as those of their relatives.Conservative returnees do not aim at changing the social context they had left before migrating; rather, they help to preserve it.
"Return of retirement" refers to retired migrants who decide to return to their home countries and to acquire a piece of land and a home where they will spend their old age.
"Return of innovation" is no doubt the most dynamic category of returnees in Cerase's typology.It refers to actors who are "prepared to make use of all the means and new skills they have acquired during their migratory experiences" (Cerase 1974, 251) with a view to achieving their goals in their origin countries, which, according to them, offer greater opportunities to satisfy their expectations.Cerase notes that these returnees view themselves as innovators, for they believe that the skills acquired abroad as well as their savings will have turned them into "carriers of change".Nonetheless, Cerase observes that these returnees are unlikely to be actors of change in their home countries because of the resilience of strong power relations and vested interests which prevent innovators from undertaking any initiatives that could jeopardise the established situation and the traditional power structure.
Cerase's typology of returnees clearly constitutes an attempt to show that situational or contextual factors in origin countries need to be taken into account as a prerequisite to determining whether a return experience is a success or a failure.
There is no question that Cerase's observations have been crucial to subsequent approaches to returnees and return migration issues.In fact, a few years later, a study by George Gmelch elaborated on Cerase's typology while stressing the need to correlate the migrants' intentions to return with their motivations for return.
Intentions to return, whether real or intended, shape the returnees' expectations in origin countries (Rogers 1984;Callea 1986;Richmond 1984).Return appears to be guided by the opportunities that migrants expect to find in their origin countries but also by the opportunities already offered in their respective host countries.As situational and structural factors have a certain bearing on the return decision, according to Gmelch, the return decision cannot be planned properly as these situational factors need to be gauged a posteriori by the migrants.
To the extent that situational factors are gauged a posteriori, migrants are viewed as being "ill prepared for their return" (Gmelch 1980, 143), owing to the fact that it is difficult for them to gather the information needed to secure their return and to gain better awareness of the social, economic and political changes that have, in the meantime, occurred in their origin countries.
Whatever the level of expectations of returnees, the structural approach to return migration contends that these are more often than not readjusted to local realities and that, owing to the strength of traditional vested interests in origin countries, returnees have a limited innovative influence in their origin societies.It also contends that if readjustment does not take place, the returnee may contemplate reemigration.
The structural approach to return migration is essential to show how influential contextual factors may be on the returnees' capacity to innovate and to appear as actors of change.Not only do skills and financial capital shape return experiences, but local power relations, traditions and values in home countries also have a strong bearing on the returnees' capacity to invest their migration experiences in their home countries.
In contrast to the neoclassical economics and the new economics of labour migration theoretical frameworks, the structural approach to return migration focuses on the extent to which returnees may or may not have an impact on their origin societies once return takes place.As explained above, their analytical framework refers to the consequences that return migration may generate in home countries, with reference to two variables: time and space.
Time pertains to the duration of stay abroad and to the change that occurred before and after migration, with reference to the status of returnees and to their origin societies.Social changes in origin societies, as well as professional advancement, are critical to the reintegration process of returnees.As W. Dumon put it, "the returnee can be defined as a person who, in order to be reaccepted, has to readapt to the changed cultural and behavioural patterns of his community of origin and this is resocialization" (Dumon 1986, 122).This process of readjustment takes time, depending on the duration of the migration experience.At the same time, the duration of stay abroad has to be optimised in order to allow migrants to acquire and diversify their skills in the likelihood of investing them once return takes place (Dustmann 2001).As Russell King suggests: If [the duration of stay abroad] is very short, say less than a year or two, the migrant will have gained too little experience to be of any use in promoting modernisation back home.If the period of absence is very long, returnees may be so alienated from their origin society, or they may be so old, that again the influence exerted will be small.Somewhere in between, an optimum length of absence might be found whereby the absence is sufficiently long to have influenced the migrant and allowed him to absorb certain experiences and values, and yet sufficiently short that he still has time and energy upon return to utilise his newly acquired skills and attitudes (King 1986, 19).
As far as space is concerned, structuralists argue that the area of settlement (i.e.rural or urban) determines the reintegration process of returnees and reshapes their expectations.Nora Colton's study on Yemeni returnees from Saudi Arabia is a case in point.The survey she carried out in rural Yemen showed that, despite the fact that returnees had improved their living standards as well as those of their families, they did not "significantly change old values" (Colton 1993, 879); rather they tended to reinforce them.Colton accounts for this phenomenon with reference to the fact that the expectations of returnees are significantly shaped by the high expectations of the return environment (i.e.family and friends who remained at home).
Finally, in the view of structuralists, because returnees adapt their expectations and behaviours to local societies, with a view to becoming "reaccepted", they tend to orient their consumption patterns to unproductive investments and to conspicuous consumption (Byron and Condon 1996, 100).Furthermore, resources tend to be monopolised by the family members who invest savings in the building of big houses and in the purchase of luxury cars, instead of using savings to modernise, for example, agricultural machinery.These consumption patterns reproduce and breed the unequal relationship between the core (receiving countries) and the periphery (sending countries) -a fundamental of the structural approach to international migration, in general, and to return migration, in particular.
Thanks to the structural approach, return is no longer viewed as being exclusively affected by the migration experience of the individual in host countries.As Lewis and Williams highlighted in their article on Portuguese returnees, the "locality" (i.e.local context in migrants' origin countries) has a great "influence on the impact of return migrants" (Lewis and Williams 1986, 125).A business-friendly institutional context, as well as economic progress in origin countries, is crucial to allow productive investments to be made.Existing institutional characteristics in origin countries also affect (negatively or positively) the impact of return migration on development and social progress.These contextual factors are further examined in the second section of this study.
Structuralists have in fact focused more on how returnees' initiatives could favour economic development when faced with local power structures than on the return migration phenomenon per se.They tend to limit the experiences of migration of the returnees to the mere acquisition of skills -which more often than not are wasted owing to the structural constraints inherent in origin economies -and to the use of foreign-earned incomes.In other words, there seems to be no continuum between the returnees' migration experiences in their former receiving countries and their situation in their origin countries.Moreover, the impact of resources, whether financial or human, tangible or intangible, remains extremely limited, owing to the fact that these are embedded in a traditional family context which defines the symbolic and behavioural patterns with which the returnees will need to comply if they want to be reaccepted back home.
Returnees' initiatives are assessed pessimistically by structuralists.The latter also offer a partial vision of return migration whose impact is embedded in a top-down framework of analysis, where the state of the origin country appears as the actor who structures the local power relations and provides more opportunities and upward mobility.Even the innovative returnee depicted by Cerase offers a disillusioned picture of the human and financial potentials of return migrants: Two things account for his [i.e. the innovative returnee] failure: The first is the economy of the village or town of repatriation; the second, the power relations among the various classes which constitute these communities (Cerase 1974, 258).
In the end, returnees fail in pursuing their interests because they have remained for too long outside the "traditional ways of thinking" in their origin societies, at the same time losing their networks of social relationships.In other words, migrants when abroad do not retain links with their countries of origin.
This core/periphery dichotomy draws the line between two separate worlds: the modern countries of immigration and the traditional countries of origin of the returnees.This structural dichotomy, strongly criticised by Rachel Murphy (2002), is based on the assumptions that little information and few exchanges exist between these two worlds and that the returnee will never be in a position to mobilise the adequate resources and skills needed to face the real conditions at home, with a view to facilitating reintegration (Velikonja 1984).As shown in the following sections, these assumptions are strongly questioned by transnationalists and social network theorists.

Transnationalism and Return Migration
This section is not so much an attempt to conceptualise transnationalism as a way of highlighting the assumptions on which it is based when dealing with return migration and returnees.
As of the late 1980s, in an attempt to highlight the dynamic and maintenance of regular migration linkages between sending and receiving countries -a fact often overlooked by the structuralists -and to interpret the back-and-forth movement of people crossing borders, migration scholars from different disciplines started to adopt the transnational terminology initially used by international relations scholars.There is no question that this terminological borrowing has been subject to various interpretations and understandings that generated a great deal of sloppiness in its usage and analytical relevance in the field of migration.
Transnationalism constitutes an attempt to formulate a theoretical and conceptual framework aimed at a better understanding of the strong social and economic links between migrants' host and origin countries.Transnational activities are implemented, according to Alejandro Portes, by "regular and sustained social contacts over time across national borders" (Portes et al. 1999, 219).It also explains how influential such links can be on the identities of migrants.Unlike the structuralists and the advocates of NELM, return does not constitute the end of a migration cycle.In the view of transnationalists, the migration story continues.
Return migration is part and parcel of a circular system of social and economic relationships and exchanges facilitating the reintegration of migrants while conveying knowledge, information and membership.One of the main contrasts between transnationalism and structuralism lies in the fact that, according to transnationalists, returnees prepare their reintegration at home through periodical and regular visits to their home countries.They retain strong links with their home countries and periodically send remittances to their households.
In the field of migration, the conceptual framework is based on two interrelated fields of investigation: transnational identities and transnational mobility.
Transnational identities result from the combination of migrants' origins with the identities they acquire in their host countries.According to transnationalists, this combination leads more to the development of "double identities" than to the emergence of conflicting identities.Migrants are viewed as having the capacity to negotiate their places in society, whether in host or origin countries, with a view to becoming part of it.Unlike the structuralists, who prefer to talk about adjustment, the transnationalists recognise the need for "adaptation" when returning home.The process of adaptation does not entail the abandonment of the identities they acquire abroad.Admittedly, returnees are faced with difficulties of reintegration, at both social and professional levels.However, as mentioned above, the regular contacts they maintain with their households in origin countries, as well as the back-andforth movements which illustrate transnational mobility (Portes 1999), allow their return to be better prepared and organised.While migrants are viewed as being successful in weighing the costs and benefits of return, the actual impact with local realities at home -at social, economic and political levels -may lead to the emergence and consolidation of transnational identities that shape the behaviours and expectations of the returnees.
In the field of transnationalism, the volume edited by Nadje Al-Ali and Khalid Koser presents an enlightening vision of how migrants' conception of "homeland" may rest on various allegiances.Migrants may be attached to their countries of birth, while being at the same time emotionally connected to their places of origin, and vice versa.For transnationalists, the migrants' subjective perceptions of homeland and their self-identification have a bearing on their decision to return and on their process of reintegration, because they provide a meaning which has a social and historical background.Finally, Al-Ali and Koser (2002, 10) argue that "another characteristic of transnational migrants is that they maintain economic, political and social networks that span several societies.What defines membership of these networks is a common country of origin or a shared origin".Common ethnicity, common origin and kinship linkages appear to be the main factors that lubricate transnational activities and define transnational identities.Migrants belong to geographically dispersed groups and "feel linked to one another by their common place-of-origin and their shared religious and social ties" (Levitt 1998, 4).
Transnationalism also aims to illustrate how "the development of new identities among migrants, who are anchored (socially, culturally and physically) neither in their place of origin nor in their place of destination" (Al-Ali and Koser 2002, 4) has been conducive to the gradual deterritorialisation of citizenship.This controversial assertion has been sustained through the recurrent reference to "diasporas" which is still in vogue among the advocates of transnationalism, although some of them are now questioning its analytical relevance, while specifically limiting its use to highly institutionalised transnational communities1 (Schnapper 2001, 31;Faist 1999).There is no doubt that this terminological borrowing is not so much a way of stressing migrants' longing for return to their homeland (as the reference to diaspora would induce us to think) as an attempt to show that "governments of sending countries have moved in recent years to intensify their contacts with their diasporas and involve them in various forms of national life" (Portes 2001, 190).Moreover, when referring to diasporas, the ethnic reference cannot be denied.
Transnationalism also pertains to goal-oriented initiatives that are collectively coordinated and that have been gradually institutionalised as a result of their interaction with sending countries' governments.There exist many emblematic cases in the world showing how governments have been responsive to the political and economic empowering of their migrant communities abroad.More often than not these institutionalised relationships have "reconstructed the ties between the emigrant and the homeland" (Brand 2002, 6;Leichtman 2002) through the creation of state institutions and para-statal bodies aimed above all at responding to the economic, security and political concerns of home countries, more than at promoting return to the homeland.True, this process is not at all new in the history of international migrations (Vertovec 1999, 145).Nonetheless, as the term diaspora has now entered the lexicon of many government officials, it is reasonable to believe that the institutionalisation of transnational activities has made them more manageable, from an economic point of view, and more permeable to political concerns (Al-Ali et al. 2001, 590-1).
In fact, transnationalism refers not only to the maintenance of strong linkages between migrants and their families or households in origin countries, but also to the multifarious ways in which migrants feel linked to one another by their common ethnic origins and in-group solidarity.Their human and financial resources seem to be embedded in an ethnically defined framework of interaction (Hsing 1998).The transnational approach to international migrations tends to view the action of migrants as the direct outcome of their belonging to their own (dispersed) migrant community.In-group solidarity and resources are defined with reference to the transnational community in which their initiatives and expectations are embedded.
Importantly, as mentioned above, transnational practices are viewed as being porous vis-à -vis state interference, as their level of institutionalisation is gaining momentum.Furthermore, in the field of transnational identities, these are viewed as being the direct outcome of practices which evolve in a kind of dual space of identification spanning the nation-states of host and origin countries. 2 While transnationalists seem to agree on the interaction between nation-states and transnational migrant communities, they also admit that further investigations are needed to understand the extent to which this interaction has shaped the magnitude and sphere of influence of both entities.Furthermore, while some of them argue that "immigrant transnationalism is not driven by ideological reasons but by the very logic of global capitalism" (Portes 2001, 187), others, on the contrary, contend that "transnational communities can wield substantial political, economic and social power" (Al-Ali and Koser 2002, 12).
Beyond these divergent interpretations, it is important to mention that, in the field of return migration, transnationalism allows cross-border linkages between the returnees and their migrant communities abroad to be highlighted.Return takes place once enough resources, whether financial or informational, have been gathered and when conditions at home are viewed as being favourable enough.In a transnational stance, return has been dealt with while referring to the ways in which returnees are successful in adapting themselves to their home environment, at all levels.They know how to take advantage of the "identity attributes" they acquired abroad, with a view to distinguishing themselves from the locals.Returnees may be faced with social pressures or feel marginalised by their own origin society, while at the same time trying to negotiate their places in society without denying their own specificities.
Finally, thanks to the transnationalist approach to international migrations, in general, and to return migration, in particular, it is possible to question the binary structuralist vision of cross-border movements, taking into account the circularity of migration movements which facilitates migrants' mobility (Chapman and Prothero 1983-84).The reference to the term diaspora could be said to constitute a way of highlighting the multi-polar mobility of migrants.
2 Luis Eduardo Guarnizo argues that transnational practices and discourses do not necessarily undermine the nation-state as transnational relations are closely interconnected with national state structures, either in the receiving or in the sending country.He adds: "Transnational practices are to nationalism what informal economic practices are to the formal economy.By definition, they are dialectically interrelated: if one disappears, the other will disappear with it.After all, transnational practices are only possible in a global system of nation-states" (Guarnizo 1998).
Despite these valuable insights, it is difficult to understand how the maintenance of strong linkages with their migrant communities abroad allows the returnees to better cope with the traditional vested interests and social pressures that characterise their origin societies.Moreover, in so far as transnationalists focus on the double or hybrid identities of migrants and on their cross-border mobility, the transnational approach to return migration seems to encapsulate their initiatives and projects at home in a fundamental set of mutual obligations, opportunities and expectations stemming from common ethnicity (i.e. the diaspora) and kinship (i.e. the family, the household).
As explained in this study, the theoretical insights stemming from network theories allow the analytical framework of return migration to be better explored, while going beyond the embeddedness of the diaspora and kinship relationships.In fact, whereas transnationalism views returnees as actors who gather the resources needed to secure and prepare their return to the homeland by mobilising resources stemming from the commonality of attributes (e.g.religion and ethnicity), social network theory views them as actors who gather the resources needed to secure and prepare their return to the homeland by mobilising resources stemming from the commonality of interests and available at the level of social and economic crossborder networks.

Social Network Theory and Return Migration
Just like the transnational approach to return migration, social network theory views returnees as being the bearers of tangible and intangible resources.Although the respective impact of tangible and intangible resources on return migrants' initiatives is difficult to evaluate a priori, it seems essential to examine return migration with constant reference to these elements.Just like the transnational approach to return migration, social network theory views returnees as migrants who maintain strong linkages with their former places of settlement in other countries.However, such linkages are not the direct outcome of the abovementioned commonality of attributes.They are not necessarily dependent on diasporas, as defined by transnationalists.Rather, in a network theoretical stance, linkages reflect an experience of migration that may provide a significant adjunct to the returnees' initiatives at home.Resources needed to secure return back home also stem from patterns of interpersonal relationships that may derive from the returnees' past experiences of migration.
Social structures increase the availability of resources and information, while securing the effective initiatives of return migrants.Thus, the composition of networks, which consist of a multiplicity of social structures (Eccles and Nohria 1992), as well as the configuration of linkages, is of paramount importance to examine the fundamentals that define and maintain the cross-border linkages in which return migrants are involved.When analysing cross-border linkages in terms of networks, no pre-established categorical attribute allows the fundamentals of network dynamics to be depicted.
Moreover, social network theorists do not take network membership for granted: first, because networks are selectively organised (Church et al. 2002, 23); second, because membership requires a voluntary act from the actors themselves as well as the consent of other members with a view to guaranteeing the flows of resources as well as the effectiveness and maintenance of cross-border linkages.In fact, crossborder social and economic networks correspond to "a social entity [that] exists as a collectively shared subjective awareness" (Laumann et al. 1983, 21).Furthermore, the formation and maintenance of networks require long-standing interpersonal relationships, as well as the regular exchange of mutually valuable items between actors.This pattern of exchange is maintained thanks to the circularity inherent in these networks.
However, it should be noted that other resources are also important to the success of returnees' initiatives and projects following their return.The availability of these resources would also seem to lie in the social capital from which return migrants benefited before migrating.On the one hand, past migration experience alone does not fully explain the returnees' initiatives.On the other hand, in terms of social capital, return migrants do not represent a homogeneous group.Social capital, which in the words of James Coleman, "inheres in the structure of relations between actors and among actors" (Coleman 1988, S110), has to be viewed as resources provided by the returnees' families or households.In other words, preexisting social and financial resources, which are provided by the family, may shape the performance of return migrants.Social capital pertains to the resources from which the returnees may benefit.It is reasonable to think that social capital and the potential involvement of return migrants in cross-border social networks may be viewed as resources that complement and shape one another.
Fundamentally, whether they are highly skilled or not, economic migrants or refugees, returnees have to be viewed as social actors who may find ways to ensure their return to their homelands, and participate in the dynamics of cross-border networks.The social networks in which returnees are involved constitute systems of social relations that may have a communal or an associative basis.The former refers to long-term relationships between network members whose exchange relations are influenced by their relational contents.The latter refers to a selective group of actors whose relationships are defined in terms of associative membership.
Whether they have a communal or an associative basis, the organisational characteristics of cross-border social and economic networks are responsive to the economic, social and political context in receiving and sending countries.The reference to social network theory allows the gap to be bridged between the organisational structure of networks and the relational content that actors attach to their own involvement and membership in such networks.Network theory articulates two levels of study.
First, return migrants are seen as social actors who are involved in a set of relational ramifications.By analysing the practice of network membership, other elements of analysis may highlight the multiplicity of involvements of these actors, as well as the types of organisation that are influential on their behaviours.Second, different network structures offer different opportunities in a given context, and different orientations and strategies.It may be from this perspective that actors derive their interests and that, at the same time, networks persist.
It now appears clear that cross-border social and economic networks differ from transnational relationships, in terms of organisational patterns, goals and configuration.Networks pertain to "a specific type of relation linking a defined set of persons, objects, or events.... The set of persons, objects, or events on which a network is defined … possess some attribute(s) that identify them as members of the same equivalence class for purposes of determining the network of relations among them" (Knoke and Kuklinski 1982, 12;Thompson et al. 1991).The definition of David Knoke and James Kuklinski emphasises the need to consider the relational content of network ties that underpins the network structure to which returnees may belong.
Furthermore, returnees are viewed as actors who confer a subjective meaning to their embedded actions, in a given context.In fact, network relationships can be based on the principle of "complementarity" (Laumann et al. 1978, 462) which may occur in a situation where actors, who differ in terms of access to resources, personal characteristics and ascribed attributes, decide to enter into a partnership which will be beneficial to both parties.Furthermore, the returnees' awareness of their network involvement must also be taken into account.This can also be defined with reference to the fact that their own vision of the world generates a form of intellectual ambience; a form of distinctiveness that the returnees like cultivating.
This contention echoes the statements made by Phillips and Potter (2003), as well as by Elizabeth Thomas-Hope (1999), when they respectively interviewed returnees to Barbados and Jamaica.Distinctiveness is far from being irrelevant, for it certainly shapes the returnees' feelings of belonging to an "entity" (Weber 1994, 16), which not only generates mutual understanding and conveys referents, but also delimits the boundaries of the social networks in which actors are involved.Moreover, distinctiveness is part and parcel of a process of identification.
Distinctiveness not only illustrates the subjective awareness of the actors involved in cross-border social and economic networks, but also shows this desire to be part of communal social relationships that both delimit the boundaries of the network (between those who are in and those who are out) and may generate mutual understanding.
Clearly, when analysing the configuration of cross-border social networks, attention has to be paid to the meaningfulness for actors of being involved in network structures.In the same vein, their perceived position in the patterns of partnerships seems to have a certain bearing on the extent to which these actors subjectively identify themselves with their networks of social relationships.In other words, there exist as many degrees of network embeddedness as there are various types of relational contents.
The five theoretical approaches that have been critically reviewed in this study and epitomised in Table 1 all contribute to better understanding the return migration phenomenon.Whether these approaches focus primarily on the economic aspects of return migration, at the individual or household levels (i.e.neoclassical economics, NELM) or the micro and macro dimensions of return migration (e.g.structuralism, transnationalism, social network theory), the various ways in which return has been analysed and returnees depicted differ in terms of levels of analysis and research framework.
Despite such differences, they are all illustrative of the various stages of development and maturation that characterise international migration streams (Martin and Widgren 2002).In fact, the foregoing comparative analysis of the theories of return migration sheds light on the economic and non-economic motivations for return, and on the need to contextualise return, in an ad hoc manner.Just as there exist several demand-pull and supply-push factors that account for the dynamics of international migration, there also exist various micro and macro factors that motivate return and shape its configuration, under specific circumstances.It is the observer's task to identify the predominant factors.

The Need to Revisit the Conceptual Approach to the Returnee
Thanks to the insights of transnationalism and social network theory, return is no longer viewed as the end of the migration cycle; rather, it constitutes one stage in the migration process.In fact, while recognising the influence of structural micro and macro factors in origin countries, both theoretical frameworks argue that the maintenance of linkages between receiving and origin countries fosters the ability of migrants to prepare and secure their own return, as opposed to what structuralists contend.Nonetheless, analyses of such linkages differ.
While transnational linkages emerge spontaneously at a cross-border level, on the basis of the commonality of such attributes as ethnicity and kinship, social network theory contends that the emergence of cross-border networks between receiving and sending countries is responsive to contextual and institutional factors.Crossborder social and economic networks are conducive to complementary exchange relations among actors which may go beyond this commonality of attributes.In fact, these exchange relations are viewed as being based on the commonality of interests, and not on attributes.Social network theory constitutes a broader framework of analysis which allows the complexity of return migration issues to be highlighted.
Having explained the analytical fruitfulness of social network theory, four basic reasons for which a revisited conceptual approach to returnees is needed may be identified.
First, the growing diversity inherent in international migration flows (Stalker 2003, 169) suggests that the analytical and interpretative framework of return migration needs to be broadened.This should not only refer to labour migrants, whether skilled or unskilled, but also to migrant students, asylum seekers and refugees.
Second, the emergence and consolidation of regional trading blocks, at a global level, has favoured the liberalisation of markets, as well as the development of the private sector, in many developing economies.Despite the potential resilience of state interference in most developing economies, liberal reforms in many migrant-sending countries have created the basis for increased business activities, not only for non-migrants, but also for migrants in general, and returnees in particular.
Third, cross-border mobility has been sustained by cheaper transport costs.These have made return a multiple-stage process.
Fourth, technological means of communication have favoured the development of flows of information, as well as the strengthening of cross-border linkages, between origin and host countries, while allowing migrants to better prepare their return.
These four reasons account for the need to revisit our analytical variables while recognising that, in terms of migration experiences, length of stay abroad, patterns of resource mobilisation, legal status, motivations and projects, returnees constitute today an extremely heterogeneous group of actors.Similarly, their impact on sending countries and potential for development vary accordingly.
Admittedly, as Rosemarie Rogers (1984) stressed, returnees differ substantially in terms of return motivations.Her seminal paper has in fact demonstrated that not only are reasons to return highly variegated but that they also tend to overlap.Today, return motivations have become diversified as new categories of returnees have been taken into consideration.In fact, scholarly approaches to return motivations do not only concern labour migrants (Kubat, 1984;King 1986), migrant-students (Glaser and Habers 1974), highly skilled migrants (Lowell 2001;McLaughan andSalt 2002, Iredale andGuo 2001;Vertovec 2002;Cervantes and Guellec 2002), entrepreneur-returnees (Cassarino 2000), but also refugees and asylum seekers (Al-Ali et al. 2001;Ammassari and Black 2001;Ghosh 2000).Moreover, the gradual broadening of the return migration spectrum has not only entailed the growing diversity of return motivations, but also the variety of resource mobilisation patterns.These patterns are certainly reflective of the returnees' migration experiences abroad, but not only that.They are also responsive to specific institutional, political and economic conditions at home that need to be considered in order to understand why some returnees may appear as actors of change at home while others do not.

Resource Mobilisation and the Returnee's Preparedness
In the context of this study, it is argued that the propensity of migrants to become actors of change and development at home depends on the extent to which they have provided for the preparation of their return.To be successfully achieved, return preparation requires time, resources and willingness on the part of the migrant.In other words, there exist various degrees of return preparation that differ in terms of resource mobilisation and preparedness.As a prerequisite to introducing the conceptual framework, these criteria need to be further explained and defined.
Resource mobilisation draws on the above-mentioned insights of social network theory and pertains to tangible (i.e.financial capital) and intangible (i.e.contacts, relationships, skills, acquaintances) resources that have been mobilised during the migration experience abroad.Resource mobilisation also includes resources that the migrants had brought with them prior to leaving their origin country (i.e.social capital).In fact, these two subgroups of resources are part and parcel of resource mobilisation.It hardly needs to be stressed that resource mobilisation patterns vary with the experiences of migration of the returnees as well as with their social backgrounds.
Preparedness pertains not only to the willingness of migrants to return home, but also to their readiness to return.In other words, the returnee's preparedness refers to a voluntary act that must be supported by the gathering of sufficient resources and information about post-return conditions at home.Clearly, the returnee's preparedness goes beyond the free-choice basis that has been introduced by the Council of Europe (1987).To strengthen the link between return migration and development at home, return should not simply be viewed as a voluntary act on the part of the migrant but, above all, as a proof of readiness.Figure 1 clarifies the ways in which these concepts interact with each other, while being at the same time reflective of circumstances in host and home countries.

Figure 1. Return Preparation
The returnee's preparednes s

Social capital Intangible resources Tangible resources
Res ourc e M obi l i s ati on The emphasis on the willingness and readiness of the migrant to return (i.e. the returnee's preparedness) yields various analytical benefits: 1.It argues that return is not only a voluntary act.Return also pertains to a process of resource mobilisation that requires time.Moreover, migrants may manifest their wish to return without necessarily being ready to return.
2. With regard to the link between return migration and development, it shows that, irrespective of their legal status in host countries, returnees differ in terms of levels of preparedness and patterns of resource mobilisation.For example, a labour migrant whose experience of migration was optimal (King 1986, 19;Dustmann 2001) (i.e.neither too short nor too long to invest the human and financial capital acquired abroad) will have a higher level of preparedness than the labour migrant whose experience of migration was too short to provide for return readiness.Similarly, a migrant who qualified for refuge or asylum in a host country, and whose length of stay was optimal, will have greater opportunities to mobilise enough resources, whether tangible or intangible, to become prepared for return, than a migrant who did not qualify for asylum or refuge.
3. It regards various types of migrant ranging from economic, skilled and unskilled to refugees.In other words, returnees differ not only in terms of motivations, but also in terms of levels of preparedness and patterns of resource mobilisation.
4. It shows that the returnee's preparedness is not only dependent on the migrant's experience abroad, but also on the perception that significant institutional, economic and political changes have occurred at home.These circumstances have a bearing on how resources are mobilised and used after return.
5. It highlights the fact that the returnee's preparedness is shaped by circumstances in host and home countries, i.e. by pre-and post-return conditions.
6.It takes into account migrants' preparedness to return while arguing that the returnees' impact on development at home is dependent on their levels of preparedness.
Having defined resource mobilisation and the returnee's preparedness, Table 2 has to be viewed as a framework of analysis which is useful in assessing and understanding the extent to which the levels of preparedness, the patterns of resource mobilisation adopted by returnees, as well as pre-and post-return This caveat is exclusively concerned with people who have returned from their host countries.It includes three levels of preparedness which are consequential on how resources, if at all, may be mobilised before and also after return.As mentioned, the positive impact of return migration at home is not only a question of willingness on the part of returnees, but also (if not above all) a question of preparation and resource mobilisation patterns shaped by pre-and post-return conditions.This statement goes beyond the success/failure dichotomy and suggests delving into the micro and macro factors that substantially configure return patterns and turn returnees into potential actors of development.
The first category refers to returnees whose high level of preparedness allows them to organise their own return autonomously while mobilising the resources needed to secure their return.This category pertains to migrants who feel they have gathered enough tangible and intangible resources to carry out their projects in their home countries.They have also developed valuable contacts and acquired skills and knowledge that can constitute a significant adjunct to their initiatives.
They have had time to evaluate the costs and benefits of return, while considering the changes that have occurred in their countries of origin, at institutional, economic and political levels.Some of them may maintain their residential status in their former places of settlement with a view to securing their cross-border mobility.Their high level of preparedness influences their participation in crossborder social and economic networks; these convey informational and financial resources that can foster resource mobilisation not only before return but also afterwards.Some migrants' projects at home may be shaped by public programmes, promoted by origin countries' governments, and aimed at repatriating skilled and business returnees.Although the impact of such return-friendly state-sponsored programmes has still to be better estimated, their implementation may be viewed as a positive change by returnees. 3Often, these programmes are accompanied by the creation of off-shore industrial zones and technological parks in origin countries, aimed at attracting foreign direct investments (FDIs) and business returnees (Cassarino 2000).
The second category includes returnees having a low level of preparedness.This category pertains to migrants whose length of stay abroad was too short to allow tangible and intangible resources to be mobilised, owing to major events which abruptly interrupted their migration experiences, e.g.unexpected family events, ostracism, no real opportunities for social and professional advancement in host countries.These migrants consider that the costs of remaining are higher than those of returning home, even if few resources were mobilised before their return.Hence, resource mobilisation in receiving countries remains extremely limited and the returnee will tend to rely on resources available at home in order to reintegrate.
The third category pertains to returnees whose level of preparedness is non-existent.These returnees neither contemplated return nor did they provide for the preparation of return.Circumstances in host countries prompted them to leave, for example as a result of a rejected application for asylum or following forced repatriation.

Conclusion
This conceptual caveat suggests that, owing to the growing diversity of returnees, we need to approach the return migration phenomenon while taking into account new variables explaining how, and under which circumstances, migrants return.Clearly, as Bimal Ghosh points out, return "is largely influenced by the initial motivations for migration as well as by the duration of the stay abroad and particularly by the conditions under which the return takes place" (Ghosh 2000, 185).The findings presented in Table 2 confirm his argument.At the same time, the reference to the returnee's preparedness (see Figure 1) and patterns of resource mobilisation complements Ghosh's argument.This dual reference takes our understanding of how and why returnees may contribute to development a step further.
This revisited conceptual framework induces us to think that the point is not so much to focus exclusively on the voluntary dimension of return as to apprehend the level of preparedness of the returnee, i.e. willingness and readiness to return.Preparedness is far from being a vague notion; it puts emphasis on the returnees' ability to gather tangible and intangible resources when return takes place autonomously.The higher the level of preparedness, the greater the ability of returnees to mobilise resources autonomously and the stronger their contribution to development.Moreover, the theoretical insights stemming from social network theory are crucial in understanding the ways in which returnees mobilise their resources while at the same time being involved in the dynamic and maintenance of cross-border social and economic networks.These networks do not emerge spontaneously; rather, they are responsive to specific pre-and post-return conditions.They also generate a continuum between the migrants' experiences lived in host countries and their situations in origin countries.This continuum regards exclusively those returnees who benefit from a high level of preparedness.Conversely, it is non-existent for returnees having a low or no level of preparedness.
These remarks are of paramount importance in understanding that the length and type of migration experiences lived abroad have a certain bearing on the various levels of preparedness of returnees and on their potential capacity to contribute to development.Again, the notions pertaining to resource mobilisation and to the returnee's preparedness must be taken into consideration in order to explain why Theorising Return Migration 276 some returnees turn out to be actors of development whereas others do not.Return refers to a preparation process that can be optimally invested in development if it takes place autonomously and if the migration experience is long enough to foster resource mobilisation.How does the above caveat posit itself with regard to these observations?First it recognises that international migration streams have reached a degree of maturation that allows return to be autonomously prepared, if conditions in receiving and sending countries are favourable enough to allow resources to be mobilised.Then it shows that resource mobilisation, which inheres in the preparation process of return and depends on the dynamics of cross-border social and economic networks, is a prerequisite to securing return.Finally it argues that a continuum is needed to allow resources to be mobilised not only before but also after return.

Notes
The author is grateful to Jaap Dronkers, Philippe Fargues, Nicola Hargreaves, Dawn Lyon, Laura Terzera and Nathalie Tocci for their fruitful comments on an earlier draft of this paper.
Paper presented at conference on Integrating Immigrants in Liberal States, Florence: European University Institute, European Forum, 8-9 May 1998).Manuscript.