The closer economic integration of Western Europe, its opening to the East and demographic pressures in the Third World have all brought migration to the forefront of the European policy debate. Researchers in the Centre's network on `Economics of European Migration' met at a workshop in Paris on 26/27 November, to discuss empirical findings on competing explanations of migration and its impacts on host countries and income convergence across countries. Papers presented focused on internal migration within and international migration from South European countries, effects of East-West migration, and the impact of migration on labour markets and government budgets in host countries. The workshop was organized by François Bourguignon, Professor of Economics at the Département et Laboratoire d'Economie Théorique et Appliquée (DELTA), Paris, Riccardo Faini, Professor of Economics at the Università degli Studi di Brescia, Research Fellow at the Innocenzo Gasparini Institute for Economic Research, Milan, and in CEPR's International Trade programme, Fiorella Padoa Schioppa Kostoris, Professor of Economics at the Università degli Studi di Roma, `La Sapienza', and Research Fellow in CEPR's Human Resources programme, and Klaus F Zimmermann, Director of the Seminar for Labor and Population Economics and Dean of the Faculty of Economics at the Universität München and Co-Director of CEPR's Human Resources programme. The financial support of a SPES grant from the Commission of the European Communities is gratefully acknowledged.

Southern Europe
Alessandra Goria (Fondazione Eni Enrico Mattei, Milano) and Andrea Ichino (Innocenzo Gasparini Institute for Economic Research, Milan) presented `Migration and Convergence Among Italian Regions'. In contrast to previous analyses of the relationship between migration and growth, their migration-augmented Solow growth model allowed for differential effects of immigration and emigration. They found this distinction important and concluded that migration has helped the convergence of per capita GDP between Italian regions as unproductive emigrants left their regions of origin to find productive employment elsewhere.

Jörn-Steffen Pischke (MIT) suggested extending the model to relax the assumption that migrants' human capital endowment is fixed, since immigrants typically acquire additional human capital in the host region. Christian Dustmann (Universität Bielefeld and CEPR) added that migration is likely to be selective on unobserved ability if there is unemployment in the receiving region. François Bourguignon suggested incorporating capital mobility and the endogeneity of government capital investments. Klaus F Zimmermann cautioned that the results depend on the stability of the relative size of the shadow economy across regions.

In `Migration and Growth: The Experience of Southern Europe', Riccardo Faini and Alessandra Venturini (Università degli Studi di Bergamo) investigated the determination of gross migration flows from Greece, Portugal, Spain and Turkey during 1960-90. For a model in which the balance of location preferences and income levels in the home country determine migrant supply, they concluded that the decline in the magnitude of migration flows from Southern Europe reflects increases in the income levels of these sending regions.

Andrea Ichino and Christoph M Schmidt (Universität München and CEPR) questioned the emphasis on migrant supply in accounting for observed population flows, since the receiving countries of Northern Europe placed restrictions on immigration following the oil crisis. Rudolf Winter-Ebmer (Universität Linz and CEPR) suggested using the estimates to calculate predicted dates of peak emigration. Jörn-Steffen Pischke cautioned that employment growth in host countries may be endogenous.

Yannis Tsamourgelis (University of Athens) presented the final paper on Southern Europe, `The Impact of Migration on Employment, Wages and Unemployment of the Host Country: The Case of Southern Europe'. He developed a model in which migration is a shock to labour supply analogous to a rise in the native population. In the host country's labour market, a matching technology links unemployment and vacancies while labour unions act in a right-to-manage framework. Tsamourgelis found that immigration facilitates hiring and thereby influences labour demand positively, which partially alleviates its negative effects on employment.

Johannes Velling (Zentrum für Europäische Wirtschaftsforschung, Mannheim) stressed the severity of recent critiques of the Beveridge curve. François Bourguignon suggested using gross rather than net migration flows, which would match the disaggregation of determinants of net employment flows into vacancies and unemployed workers. Christoph M Schmidt suggested that the steady increase in measured vacancies over time may reflect increased sampling effort rather than genuine growth of vacancies, while Christian Dustmann cautioned that it may reflect institutional changes not modelled in the paper. Velling proposed comparing measured vacancies and inflows into employment to test for the temporal consistency of vacancy measurements.

East-West Migration Flows
Silvia Weyerbrock (University of California, Berkeley) then presented her paper, `Can the EC Absorb More Immigrants? A General Equilibrium Analysis of the Labour Market and Macroeconomic Effects of East-West Migration in Europe'. Her simulations of various cases of migration from Eastern Europe and the former Soviet Union to the European Community, under both flexible and fixed wage regimes, indicated that immigration's adverse effects on unemployment and per capita income will be minor, even with no growth in the EC capital stock. Capital growth and wage flexibility will alleviate its effects on labour markets.

Klaus F Zimmermann questioned whether the assumption of perfectly competitive factor markets is justified during a period of transition in Europe. Christian Dustmann suggested using the model to calculate an optimal level of immigration and called for more detailed sensitivity analysis regarding the choice of parameters. Richard Portes (CEPR and Birkbeck College, London) suggested aggregating the rural sectors and incorporating labour heterogeneity and a more detailed regional structure into the model. Jörn-Steffen Pischke proposed splitting the Community into industrialized and non-industrialized regions and using the model to analyse the impact of East-West migration on sending as well as receiving countries.

In `The Assimilation of East European Immigrants in the German Labour Market', Christoph M Schmidt recalled the period after World War II when the German economy last faced a massive integration problem as a result of large-scale East-West migration flows. Schmidt argued that analysis of these post-war migrant cohorts does not pose the self-selection problems encountered in studying the population flows of guest-workers in the 1960s. He used two German micro surveys from 1982 and 1990 to demonstrate that there are important differences in the distribution of labour market outcomes between native Germans and immigrants. His results describing educational choices, participation behaviour, employment and wages stressed the importance of cohort effects and age at time of immigration as decisive factors for success in the labour market.

Klaus F Zimmermann argued that the long time-span between immigration and sampling that results from focusing on post-war entry cohorts will make identification of assimilation patterns difficult, and he suggested concentrating the comparison on native contemporaries of the post-war immigrant cohort. Andrea Ichino and Christian Dustmann suggested conducting separate analyses for male and female immigrants, while Johannes Velling proposed distinguishing immigrants from East Germany before and after the construction of the Berlin Wall.
Two papers, `Labor Market Effects of Immigration to Germany', by Jörn-Steffen Pischke and Johannes Velling, and `Immigration and the Earnings of Young Native Workers', by Rudolf Winter-Ebmer and Josef Zweimüller (Universität Linz), analysed the impact of differences in the regional or industry share of foreign workers on labour market outcomes of native workers. Pischke and Velling aggregated German county (Kreis) data from the late 1980s to produce data on 167 larger labour market regions. They found evidence to suggest that foreigners concentrate in areas that do well in terms of employment but the foreign share had no impact on native unemployment. These findings were confirmed by data on the immigrant influx into German counties after 1989.

Klaus F Zimmermann questioned the quality of the regional data, since any definition of local labour market areas for sets of contiguous regions must follow arbitrary rules. Several participants discussed the use of instrumental variable techniques to account for the endogeneity of the share of foreigners.

Winter-Ebmer and Zweimüller used comprehensive data on all employees from a random sample of 5,000 Austrian firms based on social security records. They adopted an insider-outsider framework to argue that immigrant workers will be close substitutes for young male native workers. Their results demonstrated, however, that native workers' earnings are higher in firms with high shares of foreign workers, even with the use of instrumental variable techniques.

Christoph M Schmidt argued that most Austrian wage bargaining takes place at the regional level, while these results might be driven at the firm level by unobserved productivity differences rather than the proposed insider-outsider distinction. He also suggested using the earnings of foreigners in these firms to control for firm-specific effects, while Klaus F Zimmermann proposed controlling for differences in demand across product markets.

In `Budget Incidence of Immigration into Switzerland: A Cross-Sectional Analysis of the Public Transfer System', joint with Thomas Straubhaar, René Weber (Universität Basel and Bundesamt für Aussenwirtschaft, Bern) compared the impact of resident foreign and native Swiss households on the government budget. He used the 1990 Swiss Consumer Survey to show that the contribution of the average foreign household to net public revenue exceeded that of the average native household and concluded that, from a cross-sectional perspective, immigrants benefit the native population though their impact on the public transfer system.

Christoph M Schmidt suggested that the observed difference in average net budget contributions results mainly from the different age compositions of native and foreign populations, since foreigners tend to be more concentrated at their prime working age. Klaus F Zimmermann suggested that these differences in age structures undermine the ability of such a cross-sectional approach to yield useful insights into the life-cycle impact of foreigners on the government budget. François Bourguignon proposed including military service by Swiss men as a positive contribution to the budget.

In `Differences in Labor Market Behavior between Temporary and Permanent Migrant Women', Christian Dustmann developed a two-period neoclassical model of married migrant women's intertemporal labour supply decisions to show that temporary migrants work more hours than permanent migrants in the initial period. In an empirical implementation on German data, he distinguished migrant women on the basis of data from interviews concerning their intentions to return migrate; his estimation results supported the model's predictions concerning participation rates but not those concerning hours worked.

Several participants discussed the problems of comparing schooling received in the sending and host countries. Riccardo Faini suggested using the model's intuition to consider behavioural differences between temporary and permanent migrants' savings behaviour. Andrea Ichino stressed the potential endogeneity of return intentions to the current state of the labour market and suggested using instrumental variable techniques. In contrast, Roxane Silberman (Laboratoire d'Analyse Secondaire et Méthodes Appliquées en Sociologie, Paris) stressed that sociological studies had consistently found migrant women's participation rates to depend on family formation at the time of migration to the host country.