|
Immigration is returning to the centre of the European policy debate in the 1990s, as global economic, political and demographic trends make it increasingly attractive for citizens of countries in Southern and Eastern Europe, North Africa and Asia to migrate to EC member countries. Papers presented at a London workshop on `The Economics of Migration' on 22/23 November addressed the main issues raised by such large-scale population movements and their key policy implications for the Community. The workshop was organized by Riccardo Faini, a Research Fellow in CEPR's International Trade programme, and Klaus F Zimmermann, Co-Director of the Centre's Human Resources programme. Financial support was provided by the UK Economic and Social Research Council and by the Commission of the European Communities under its ACE programme. Oded Stark (Harvard University) opened the workshop by outlining three alternatives to the orthodox theory that the migration decision is primarily a response to inter-regional wage differentials. First, if future incomes are uncertain and imperfectly positively correlated in a geographically distinct area, migration by a member of an income-pooling family may serve to diversify risk. Second, if individuals' welfare is directly and negatively related to their `relative deprivation', income distributions in their home countries also influence the migration decision. Third, individuals with heterogenous skills may appear equivalent to foreign employers although distinguishable to domestic employers. If a worker with a given skill level prefers to migrate, so will all those with lower skill levels. Migration flows then differ from those in the case where information is symmetric and all employers can distinguish workers observationally. Juan Dolado (Banco de España) questioned the relevance to modern societies of the mental bonding to the home community assumed in the `relative deprivation' model and doubted that modern labour markets are characterized by asymmetric information as the third model suggested. Christopher Pissarides (LSE) asked if there was any empirical support for the `risk diversification' model. Stark cited evidence from India which indicated that women who had grown up in areas particularly afflicted by the risk of drought were more likely to marry into distant villages. Carlo Carraro (Università di Venezia and CEPR) presented a paper on `The Demand for Migration under Perfect Wage Discrimination', written with Antoine Soubeyran, which developed a model in which natives and migrants are observationally distinct and differ in both productivity and aversion to effort. For example, if migrants are both less averse to effort and less productive than natives and the effort effects dominate, their marginal disutility of production will be lower than natives' for equivalent levels of production. At `low' levels of demand, only migrants will be employed; immigration quotas would then limit rates of native unemployment and also reduce firms' profits. More generally, a policy to minimize native unemployment would permit migrants to enter the country only to work in sectors in which existing native labour supply is negligible or natives' disutility of production relatively low. Christopher Pissarides questioned the relevance of the paper's policy conclusions. He argued that these could easily be reversed if the authors relied on a set of different (and at least equally reasonable) assumptions about productivity and effort effects. Similarly, Pedro Telhado Pereira (Universidade Nova de Lisboa) disputed the usefulness of migration models based on the association of disutility of effort with national groupings. In a paper on `The Earnings Dynamics of Immigrant Labour', Christoph Schmidt (Universität München) noted that the empirical finding that migrants' earnings grow faster than those of observationally equivalent natives is conventionally explained by human capital or cohort effects. He then presented a model in which migrants' abilities are known to the employer and to themselves on entry to the host country, but all parties revise their beliefs about workers' abilities over time. Risk-averse migrants accept long-term wage agreements specifying wages below their expected marginal products to protect themselves against the revelation of low ability. If firms agree to such contracts and then increase the wages of migrants whose abilities are revealed to be relatively high, migrant earnings will outgrow native earnings. Provided that re-emigration probabilities are exogenous, multilaterally observed and larger for temporary than for permanent migrants, this model, argued Schmidt, should be observationally distinguishable from the corresponding human capital model. Oded Stark noted that long-term contracts of this type bear little resemblance to those found in the US, where employment turnover is high and labour law gives workers relatively little protection, although they have some intuitive appeal for the modelling of German labour markets. He also doubted whether the probabilities of re-emigration are either observable or exogenous. Presenting his paper on `The Economic Impact of Immigration on the Host Country', Andrea Ichino (Innocenzo Gasparini Institute of Economic Research, Milan) pointed out that immigration need not exert a negative impact on average native earnings of natives if high-skill natives are complements in production with low-skill workers. Indeed, migration may then increase the productivity and hence the earnings of high-skill natives, although low-skill natives' earnings would then fall. In the long run, the distribution of the costs and benefits of such migration among natives will change, as migrants and natives accumulate human capital at different rates. Econometric estimates from cross-country panel data suggested that migration had a significant positive effect on aggregate native earnings. Given the limitations of the data, however, Ichino cautioned that these findings should serve only as a bench-mark for further research. Anton Muscatelli (University of Glasgow) questioned the econometric consistency of Ichino's approach, arguing that regressions of earnings by nationality on national macroeconomic variables including immigration requires single-country rather than cross-country data. In a paper on `Migration and the European Community', written with Thomas Straubhaar, Klaus F Zimmermann (Universität München and CEPR) argued that immigration can compensate for demographic losses arising from the decline and ageing of the European labour force. The economic history of recent decades in Europe, the US and Canada shows that immigration has been economically beneficial for the receiving countries. Economic theory predicts that welfare gains should be reaped from free factor movements as long as social costs and adjustment costs are not prohibitive. He stressed that an efficient policy must stipulate common rules of entry to all EC member countries. Only if immigrants enjoy the same freedom of movement within the Community as its citizens will member countries' marginal productivities of labour converge over time. Zimmermann proposed that entry to the Community be conditional on the holding of a pre-arranged job and the payment of a fixed entrance fee or an immigration tax to meet potential social costs. Most migrants would then be workers with skills of which native supply is scarce. A quota system would only be a second-best solution. Entry by non-economically motivated migrants should be as restricted as their fair treatment under international law permits. Barry McCormick (University of Southampton) argued that implementing such laissez-faire policy recommendations would make immigration a response to excess demands in Community labour markets; the resulting population flows may be inefficient if they dampen the long-term responses of EC labour supplies. Zimmermann's proposed policy also ignores the potential externality generated by the `brain drain' from the sending countries, which could reduce their average incomes and hence increase the pressure of migration into the Community. Richard Freeman (Harvard University) presented an overview of research on migration into the US, whose effects on migrants and natives are well documented, although its effects on sending countries remain relatively neglected. Emigration to the US has led to the movement of up to 10% of the populations of El Salvador and Nicaragua, whose remittances now exceed their home countries' export earnings with significant effects on exchange rates. Freeman stressed the importance of relationships among trade, capital and labour flows. He also contrasted the basically favourable view of immigration in North America with the rather fearful attitude often found in Western Europe; in the US there seems to be more concern to keep out foreign products and capital than people. Closing the workshop, Riccardo Faini (Università di Brescia and CEPR) outlined an agenda for future research on migration and the European Community. He proposed moving to empirical testing of theoretical models of the type presented in the papers and stressed the importance of the interrelationships between capital, labour and commodity flows. Such a programme should add to the understanding of the factors driving the extent and effects of immigration into the European Community in the 1990s. |