4.1.1 Out-migration and domestic labour supply
The discussion of the impacts of outmigration begins with the consideration that the latter causes a reduction in the total resident population of the areas of origin, and also tends to bring about a reduction in their labour supply. From a theoretical perspective, however, it has been argued that that the latter is not necessarily the case, not only because some of the people who migrate were not originally a part of the labour force (whether employed or unemployed), but also because the possible wage increases brought about as a consequence of the initial reduction in the labour supply may induce some of the economically inactive to enter the labour force (and some of those already at work to increase the number of hours they work) (Lucas 2005).
The latter are, of course, formalist (and typically neoclassical) arguments, as are those typically used to analyse how deleterious the reduction in the labour supply turns out to be if and when it does occur. Most commonly, the latter is theoretically depicted as depending on the scarcity of labour as a production factor in the areas of origin (i.e. what its ‘marginal productivity’ is). Under the assumption of homogeneous and “unlimited”
labour (as in Lewis’ 1954 model), outmigration is presumed not to have any effect upon
79 domestic production: it merely increases the capital-labour ratio, reduces unemployment and increases income per capita (cf. Massey et al 1998, Abreu 2009). Once the assumptions of homogeneous and unlimited labour are relaxed, however, these conclusions cease to apply. Under homogeneous labour but positive marginal productivity of labour, aggregate output and income are postulated to decrease as a result of emigration even though output and income per capita increase, while distributive effects occur to the benefit of the workers (wages increase) and to the detriment of the holders of capital (the returns to capital are reduced) (Massey et al 1998). Then, once the elementary consideration that labour is not a homogeneous ‘production factor’ is in its turn taken into account, the analysis becomes at once less straightforward, less formalist and more realistic. As highlighted by Lucas (2005), the strategic role played in the production process by some types of workers, the existence of regional bottlenecks in terms of labour availability and the possibility of seasonal labour shortages all mean that even otherwise plentiful labour may in some cases be scarce – and that, in those cases, labour emigration does not have the ‘zero opportunity cost’ assumed in models based on unlimited labour supplies. Much more complex impacts upon production and productivity may therefore occur as a result of outmigration, depending on the geographical and skills’ profiles of the migrants, as well as on the relative scarcity of those skills in the context of the areas from which they migrate (Lucas 2005).
As might be expected, the specific category that has been the object of greatest attention in this context is that of ‘highly-skilled’ workers (i.e. the “brain drain”, arbitrarily but typically operationalised as corresponding to the emigration of those with tertiary education: e.g. Docquier and Rapoport 2004). This focus is accounted for by reference to the assumed importance of the latter in the process of development, their relative scarcity in the context of developing countries and the fact that, due to the selectivity and self- selectivity of migration, skilled migrants tend to be over-represented among the migrant population (Chiswick 2000). A vast ‘pessimistic’ literature on the impact of migration upon development has thus flourished around this particular issue above all others, featuring such notable contributions as Bhagwati’s writings on this subject, which included the suggested levying of a migration tax by the countries of origin in order to create a disincentive to emigration (e.g. Bhagwati 1976). Additionally, insofar as the ‘brain drain’ has been found to disproportionately affect some key sectors, particularly health, skilled migration has also been posited to indirectly affect the ‘quality’ of the labour force, thereby further constraining economic performance (Bhargava and Docquier 2008).
80 After dominating the literature for decades, during the course of which it had a profound impact on both academia and the policy-making world, this pessimistic view of the ‘brain drain’ was eventually challenged by the so-called “optimal brain drain” literature, put forth most prominently by one of the figureheads of the NELM, Oded Stark (2002 and 2005). The key arguments put forth by Stark and other proponents of this view are the following: i) the possibility of migration increases the expected returns to education, thus creating an incentive to the acquisition of knowledge and skills in the areas of origin; ii) for various reasons including immigration restrictions, a significant proportion of those who thus acquire additional knowledge and skills end up remaining in their original areas; and iii) for that reason, the areas in question may end up with higher average educational levels and/or a greater pool of skills than would have been the case in the absence of emigration (Stark 2002 and 2005). This “revisionist” approach to the ‘brain drain’ was met with some (limited) success, but it has also been challenged theoretically in its turn – among others by Schiff (2005), who argued not only that such a combination of incentive structures and actual outcomes is difficult to come by in practice, but also that the “optimal brain drain”
argument fails to take into account the additional public and private expenditures incurred in education as a result.
Thus, we find that the main dividing lines in the theoretical debates around the topic of the impact of outmigration upon the labour supply have largely consisted of: i) the issues of the scarcity and heterogeneity of labour; ii) the possibility of compensating effects occurring in response to the initial impacts upon the labour supply; and iii) the effects of the possibility of migration upon the structure of incentives to skills’ acquisition.
4.1.2 Remittances
The impact of remittances has played a similar role among “migration enthusiasts” to that played by the ‘brain drain’ among “migration pessimists”. This has been especially the case as a result of the massive increase in (and greater formality and visibility of) worldwide remittance flows from the 1990s onwards, but it also reflects the fact that migrant remittances account for a large share of GDP in many developing countries, and that they often exhibit counter-cyclical properties (with migrants sending more when their relatives are most in need) (Carling 2005).
Although most of the enthusiasm around remittances lacks clearly-specified theoretical underpinnings, it has tended to gravitate around two main axes: the effect of
81 remittances in bringing about poverty reduction, hence a direct improvement in social welfare (Acosta et al 2007); and their effect upon investment, serving to alleviate the main constraint to economic development most commonly assumed to characterise underdeveloped regions and countries in general (Bjuggren et al 2010). With such potentially virtuous implications, and as they experienced a massive increase in volume and visibility worldwide, it is of little wonder that remittances should have become, as per Kapur’s (2004) fortunate and oft-quoted expression, a “new development mantra”.
A major weakness of the ‘optimistic’ literature, however, has been the tendency to automatically equate the transfer of private income and wealth with the transfer of capital – which is of course a result of the lack of a theoretically robust concept of capital in the first place. While few non-Marxists have gone as far as criticising the essentialist understanding of capital implicit in this confusion and replacing it with a relational one, many migration scholars have undertaken a similar (if less robust) critique by highlighting that private remittances tend to be used for consumption rather than investment – a consideration which spurred a wave of remittance-use micro-studies (Russell 1992). This latter approach has in its turn been criticised for restricting its attention to the spending behaviour of the immediate recipients (first-round spenders) of remittances, thus neglecting the indirect effect upon investment brought on either as of the subsequent rounds of the multiplier or through mediation by the banking sector; as well as for failing to appreciate the fungible character of remittances vis-à-vis other sources of household income (cf. Massey et al 1998, Abreu 2009). Nevertheless, it has usefully contributed to highlighting the inadequacy of simplistically regarding remittances as ‘capital’. The other main line of argument taken up by the ‘revisionist’ take on remittances has consisted in the idea that the latter breed distortion and dependency: the theoretical argument is that remittances create a disincentive to work amongst recipient households, possibly locking those households in a situation of permanent dependence upon exogenous resources. Thus, especially when migration is sufficiently widespread, remittances themselves are deemed to have a negative effect upon the labour supply, compounding the direct one brought on by outmigration (Chami et al 2003, Ellerman 2003).
In the meantime, while playing a less central a role in the debates, other potential impacts of remittances have been theoretically postulated and empirically explored in the literature. These have included the potential of remittances to bring about ‘Dutch disease’, i.e. an appreciation of the country’s currency compromising its export competitiveness (e.g.
Acosta et al 2007b); the effect of remittances in strengthening the domestic financial sector,
82 by providing liquidity to the latter and constituting a stimulus to branch expansion (e.g.
Gupta et al 2009); the potential inflation-inducing effect of remittances (e.g. Ball et al 2010);
and the impact of remittances upon inequality, under the hypothesis that remittances may exacerbate inequality at the village, regional and/or national levels, given that migrants (especially international ones) are not homogeneously drawn from the population and rarely if ever consist of the poorest (e.g. Docquier et al 2003).
We thus find that, similarly to the case of the effect of migration upon the labour supply, the literature on remittances is largely characterised by isolated emphases on each of the various macroeconomic effects that they bring about, with a ‘standard’ (in this case optimistic) approach being subject to theoretical challenge by a range of ‘revisionist’ (in this case pessimistic) takes.
4.1.3 Other transnational impacts of the diaspora
The increasing attention and emphasis on remittances has been part of a broader shift from regarding migration as a fairly permanent process involving the severing of the ties that link the migrants with their areas of origin to a transnational approach that instead emphasises dual identities and belongings, as well as actions and practices that straddle different areas and countries (Levitt and Nyberg-Sorensen 2004). Money remittances constitute the most visible and economically relevant form of migrant transnationalism, but migration scholars and researchers have also highlighted other forms of transnationalism undertaken “from below” as a ‘by-product’ of migration that have a potential impact upon economic development (Guarnizo and Smith 1998). Most significantly, these have included transfers of technology and know-how, most notably in the case of so-called diaspora knowledge networks (Mayer and Wattiaux 2006); the role of diaspora business networks in fostering foreign investment and business partnerships (Newland and Patrick 2004); and the specific development-inducing effects of the projects and activities undertaken by migrant hometown associations (HTAs) (Mercer et al 2008)15.
15 To these, one might also add the more loosely-defined “social remittances” that consist of the cultural effect exerted by migrants upon the values, beliefs and attitudes of their households and communities of origin (Levitt 1998); or the vast literature on transnational political activities of diasporas (cf. Ostergaard-Nielsen 2003, Mercer et al 2008). Given the less direct relationship between these processes and practices and the process of economic development, however, I shall not address them here beyond this brief mention.
83 Along with remittances, the transfer of knowledge and know-how, albeit defined in vague and hopeful terms, has long been one of the main arguments of migration
“optimists”. Because this transfer usually takes on forms that are intangible and difficult to operationalise, however, the literature has rarely gone beyond enunciating this potential in general terms. The exception to this possibly consists of the focus on the networks of highly-skilled elements of the diaspora who have increasingly come together in order to actively engage in the sharing of know-how and experiences, and to devise ways of contributing to the scientific and economic development of their countries of origin (Mayer and Wattiaux 2006). The emergence of these practices facilitated the shift from a
“traditional emphasis on embedded knowledge of potential returnees in a human capital approach (return option) to a connectionist approach where social capital, including technical and institutional links, is crucial” (id ibid:5). This is hardly the place to undertake a critique of the reliance on the concepts of human and social capital (on which see Fine 2000), so I shall merely highlight the fact that this passage renders clear the theoretical proposition (or hope) that skilled emigrants, both individually and especially in the context of structured networks, actively undertake the transfer of knowledge to their countries of origin and contribute to skills’ acquisition by their compatriots there. They thus constitute a
‘scientific counterpart’ to diaspora business networks, which have also attracted increasing attention by virtue of their capacity to function as chambers of commerce, disseminating knowledge about business and investment opportunities in the countries of origin, facilitating business partnerships and, consequently, attracting foreign direct investment into the country (Newland and Patrick 2004). A similar but certainly more ‘grassroots’ form of migrant network transnationalism is that of hometown associations, which consist of translocal networks bringing together migrants originating in the same community and local members of that community in order to undertake collective initiatives aimed at improving the welfare of the local residents in various ways (Orozco and Welle 2004, Mercer et al 2008). While being often beset by specific difficulties (limited fundraising capacity, difficulties and conflicts regarding the setting of priorities, etc.), these have been argued (and shown) to play an important complementary role vis-à-vis both the provision of public services and the transfer of remittances at the individual and household level, their most successful endeavours typically consisting in the financing and provision of community-level public goods (Mercer et al 2008).
84 4.1.4 Return migration
The last main category of development impacts addressed in the “migration-development”
literature consists of those which are associated with the migrants’ return to their areas of origin. To a large extent, the effects of these are deemed symmetrical with respect to the original outmigration – adding to the labour supply to a varying extent and with varying consequences, depending, among other things, on the age and skills profile of the return migrants. The literature on the economic consequences of return migration has thus typically sought either to empirically identify the specific characteristics of return migration in terms of (self-)selectivity, or to put forth typologies and a priori assertions with regard to those characteristics (Cesare 1974, Cassarino 2004). This has occasionally been complemented by the ‘optimistic’ consideration that many return migrants bring along their savings and know-how (which may be used to start businesses); but also by two
‘pessimistic’ caveats: that many migrants return to their countries with a view to retiring;
and that the skills and know-how that they have acquired abroad are often location-specific and/or associated with specific production processes, and therefore cannot be put to productive use in a different context from that in which they were acquired (O’Conner and Farsakh 1996).
As made apparent in this section, many specific linkages between migration and
“development” have been theoretically posited (and empirically investigated) in the literature. Without having sought to be thorough, and having focused solely on the effects that have a more strictly ‘economic’ character, we have seen that migration has been theoretically postulated:
i) through the initial outmigration, to bring about a decrease in the labour supply (or not), reduce the average skills level of the labour force (or not), decrease total output and income (or not), increase income and output per capita, reduce unemployment, affect the functional distribution of income, and bring about a deterioration in health and education (hence the ‘quality’ of the labour force);
ii) through remittances, to reduce poverty, increase investment, increase output, increase inequality, reduce the labour supply, generate dependence, develop the financial sector, compromise export competitiveness and bring about inflation;
85 iii) through other transnational practices of migrant diasporas, to promote the transfer of skills and knowledge, foster business partnerships and foreign direct investment, and enhance the translocal provision of public goods; and
iv) through return migration, to increase the labour supply (or not) and increase the skills’ levels of the labour force (or not).
Whether or not each of these effects occurs both in general and in specific concrete contexts is a matter for empirical investigation, which has indeed been undertaken in multiple contexts. What this myriad of potential channels renders clear, however, is that it is virtually impossible to settle the “unsettled” migration-development nexus by recourse to deductive methods. Quite simply, the variety, complexity and context-specificity of these channels make it impossible to deduce ultimate consequences from assumptions regarding the intermediate causal linkages. Additionally, I would argue, it is both sterile and misleading to assess each facet of the problem in isolation, without viewing “phenomena, places and problems in terms of their interconnections and relations within a social totality” (Bedford 1981:219).The alternative, of course, consists of case-by-case analyses based on a clear theoretical understanding of what constitutes development as a process, what drives it, and what role migration can be expected to play in that context. Given that, as explained in Chapter 3 (above), the historical materialist conception of development is the one adopted in the context of this thesis, the next section briefly reviews some of the contributions to addressing the migration-development nexus from this theoretical perspective.