Venice played host to a CEPR conference on ‘Trade and Factor
Mobility’ on 24/25 January 1997. The meeting, held under the auspices
of the Centre’s research programme on ‘European Migration: From
Economic Analysis to Policy Response’, received financial support from
Centro Studi Luca d’Agliano, with additional funding provided by the
Dipartmento di Scienze Economiche, Università degli Studi di Venezia.
The organizers were Riccardo Faini (Università degli Studi di
Brescia and CEPR), Jaime de Melo (Université de Genève and CEPR)
and Klaus Zimmermann (Universität München and
Anthony Venables (LSE and CEPR) presented ‘Trade Liberalization and Factor Mobility: An Overview’, which illustrated how liberalization can alter factor prices and thereby change the incentives for factor mobility. Venables started with the familiar Heckscher-Ohlin result that trade liberalization reduces the incentives for factor mobility, but went on to show that this need not be the case in a specific-factors model (which allows for a wider range of factor substitution and complementarity relationships), or in an imperfectly-competitive model that allows for increasing returns to scale.
André Sapir (Université Libre de Bruxelles and CEPR) suggested that separating the issue of substitutability and complementarity of factors from the experiment of trade liberalization might help enhance the clarity of exposition. He also noted that the choice among models should depend on the identities of the countries involved: for example, the Heckscher-Ohlin model (in which differences in endowments drive trade) makes more sense for thinking about trade between rich and poor countries than between similarly wealthy nations. Jeffrey Williamson (Harvard University) elaborated on Sapir’s suggestion by recommending a distinction between models which are more relevant for understanding transitional events (e.g. German re-unification) and those best suited for understanding non-transitional states. James Markusen (University of Colorado) pointed out the limitations of the Helpman-Krugman framework of imperfect competition (in particular, the underlying assumption of symmetric substitutability) as the basis for analysing the complementarity or subsitutability of trade and factor flows.
In ‘The King Never Emigrates’, Arye Hillman (Bar-Ilan University), Gil Epstein (Bar-Ilan University), and Heinrich Ursprung (Universität Konstanz) offered a provocative theoretical contribution on the importance of the political-economic institutions in countries of potential emigration. Government policies can change the economic incentive to migrate by affecting the income distribution, and if highly productive workers are sufficiently disadvantaged by ‘the king’s’ redistributive system, they will emigrate (to the disadvantage of those left behind). In general, citizens can be ranked according to their ‘closeness’ to the king, those furthest away being taxed the most. When citizens optimize over productive and privilege-seeking activities, the most productive end up furthest from the king, are heavily taxed, and might have a strong incentive to emigrate. An alternative model allows privilege-seeking to be a ‘difficult activity’, with the result that the least productive workers end up furthest from the king.
Panos Hatzipanayatou (University of Thessaloniki) presented ‘Trade Liberalization and Public Good Provision: Migration Promoting or Migration Demoting?’, written jointly with Konstantine Gatsios (Athens University of Economics and Business and CEPR) and Michael Michael (University of Cyprus). The authors constructed a general equilibrium trade model of a small country producing two private tradable goods and one public consumption good. The model allowed perfect international labour mobility, but assumed that trade is restricted. The authors identified conditions under which trade liberalization leads to a reduction in the supply of the public good and promotes emigration. Ignazio Muso (Università Ca’ Foscari di Venezia) noted that the results of the paper depend on assumptions concerning a number of partial cross-derivatives. He also suggested that an assessment of the welfare impact of trade liberalization and its consequences would be useful. Riccardo Faini noted an odd feature of the model: the temporary nature of the migration implies that citizens abroad continue to consume domestic public goods. Anthony Venables suggested using a more sophisticated supply-side analysis of public goods provision in which the government is allowed to alter tax rates rather than simply adjust expenditure to reach the budget constraint.
The paper given by Ian Wooton (University of Glasgow and CEPR), entitled ‘Regional Integration, Trade and Migration: Are Demand Linkages Relevant in Europe?’, was co-authored with Rodney Ludema (Georgetown University). The authors employed Krugman’s economic-geography model, in which demand linkages can promote the geographic agglomeration of manufacturing activity, and they allowed for imperfectly mobile manufacturing labour to make the model more representative of the European economy. They showed that, if some workers are reluctant to migrate, then the forces of agglomeration may be constrained and complete agglomeration may be avoided. Another interesting conclusion concerned the timing of different elements of liberalization: it may be advisable to complete trade liberalization before lifting restrictions on international labour movements.
Giorgio Basevi suggested the authors consider an alternative shape to the labour supply curve which could yield stable equilibria with complete agglomeration. James Markusen remarked that more fully developed, micro-foundations to the model’s labour supply side would be useful. André Sapir argued the need for comparisons of the US and European potential for agglomeration and interregional migration to confront the issue of interregional differences in social security systems. Riccardo Faini inquired about the determinants of preferences for mobility: for example, will labourers’ reluctance to move away from home increase or decrease as income rises?
In the first of a series of more empirical papers, Jeffrey Williamson (Harvard University) presented his joint work with William Collins (Harvard University) and Kevin O’Rourke (University College Dublin), entitled ‘Were Trade and Factor Mobility Substitutes in History?’. After a brief review of the theoretical ambiguities concerning the relationship between trade and factor mobility, the paper attempted to resolve the issue empirically, using data for the Atlantic Economies from 1870 to 1940. The authors examined the data in annual time-series form within each country, and in decade-averaged panel data form by pooling all of the available evidence. They concluded that the notion that trade and factor mobility substitute for one another definitely was not supported by historical experience, but they were unable to make any definitive claims concerning potential complementarity. They found no evidence of a ‘regime switch’ from complementarity to substitutability as New World frontiers closed, and concluded that New World policy-makers did not behave as if they viewed trade and migration as substitutes for one another.
Gianni Toniolo (Il Università di Roma and CEPR) recommended that the authors attempt to discuss and explain individual country experiences rather than generalize on the basis of the overall regression results. Anthony Venables suggested paying more attention to the commodity composition, as opposed to just the volume, of trade flows. Riccardo Faini suggested incorporating a measure of transport costs. Arye Hillman referred the authors to Australia’s Bridgen Report for more insight as to how policy-makers thought about trade and migration earlier this century.
James Markusen (University of Colorado) presented a paper jointly authored with Stephen Zahniser (University of Colorado) entitled ‘Liberalization and Incentives for Labour Migration: Theory with Applications to NAFTA’. The authors suggested that one of the motivations for NAFTA (from the US perspective) was to reduce incentives for Mexicans to migrate to the United States. Basic Heckscher-Ohlin trade theory suggested that Mexico’s abundant factor (unskilled labour) should be made better off by liberalization, thus reducing the incentive to migrate. However, by drawing on the Feenstra-Hanson and Markusen-Venables models (in both of which the demand for skilled relative to unskilled labour rises in both Mexico and the United States), and by examining the nature of Mexican maize production, Markusen and Zahniser showed that liberalization need not reduce the incentive.
Pasquale Sgro noted that, since NAFTA has only been in place for a few years (during which Mexico suffered a major macroeconomic crisis), it is difficult to make medium- or long-term predictions about its consequences for labour mobility. Alessandra Venturini (Università degli Studi di Bergamo) inquired about the propensity of Mexicans to migrate after acquiring experience in border-area manufacturing establishments. Jeffrey Williamson suggested confronting the possibility that migration had been income-constrained and that rising Mexican incomes could lead to more migration, or, alternatively, focusing exclusively on wage patterns and leaving aside explicit discussion of migration. Riccardo Faini noted that if wages rise for the rich (whose migration decisions are not income-constrained), then they may be less likely to migrate; and, if wages simultaneously fall for the poor (who are income-constrained), then they too may be less likely to migrate. Thus, migration can fall for both groups as wage inequality widens.
‘Trade and Migration: A Production-Theory Approach’, in which imports and foreign labour services are both treated as inputs to the production function, was delivered by Ulrich Kohli (Université de Genève). By estimating a four-factor cost function (imports, non-resident labour, resident labour and capital) using Swiss national accounts data, Kohli tried to identify the substitutability and complementarity relationships among the various factors. He found that, in Swiss experience, imports and non-resident labour were complements from 1950 to 1986.
Marzio Galeotti (Università degli Studi di Bergamo) praised the flexibility of the production theory approach, but suggested some alternative means of implementing it. These included: distinguishing labour by education and experience; estimating a GNP – as opposed to a cost – function for the sake of identifying the relevant elasticities; and, in contrast to the cost-function approach, recognizing that the supply of inputs (especially capital and labour) may not be perfectly elastic. Galeotti further recommended using instrumental variables to circumvent the econometric problems posed by endogeneity of the right-hand side variables. André Sapir suggested an interpretation of the complementarity result based on immigrant employment in a non-traded sector (e.g. services). Klaus Zimmermann questioned the modeling of technological advance, suggesting that it is endogenous, and wondered if capacity utilization data could be used to refine the story.
The impact of the opening of the eastern borders on the Austrian and German labour markets was investigated by Rudolph Winter-Ebmer (Universität Linz) and Klaus Zimmermann (Universität München) in ‘East-West Trade and Migration: The Austro-German Case’. Owing to their geographical proximity, Austria and Germany are heavily affected by increased immigration and trade with Eastern Europe. After detailing the impact in specific regions and industries, the paper analysed the employment and wage effects of immigration econometrically using industrial panel data. Overall, the results were mixed: more negative employment and wage effects were found in Austria than in Germany, but rising trade relations with Eastern Europe do not seem to have substantially influenced employment and wages in either case.
Several issues were raised by Maria Schenkel (Università degli Studi di Venezia) for consideration: whether instrumental variables should be used to deal with endogeneity of the ‘independent’ variables; whether the period of dramatic macroeconomic change (in Germany in particular) makes it difficult to identify the effects of immigration and trade; whether it would be useful to disaggregate labour by skill level; and whether internal German migration (from east to west) serves as a substitute for immigration from other countries. Riccardo Faini expressed concern about the specification of the regression equations, in particular about the endogeneity of output growth (a right-hand side variable). Jaime de Melo noted that the different relative sizes of Germany and Austria could have important implications for the observed responses to rising trade and migration.
The final paper, by Jean-Marie Grether (Université de Genève), Jaime de Melo (Université de Genève) and Riccardo Faini (Università degli Studi di Brescia), was entitled ‘Globalization and Migratory Pressures from Developing Countries: A Simulation Analysis’. The authors used a Ricardo-Viner model to study the determinants of the supply of emigrants from developing countries. Their model incorporated migration costs and risk-spreading behaviour by households in the sending countries. The paper showed that, if exports respond strongly to trade liberalization, then a real appreciation may be accompanied by lower migration levels. If exports are slow to respond to liberalization, however, then a depreciation may occur along with more migration. The composition of consumption may also be an important consideration: if exports are a large part of household consumption, then liberalization might lower real wages and promote migration. The paper emphasized that trade and migration policies should not be assessed separately.
Alessandra Venturini emphasized the significance of the authors’ sensitivity analysis in which additional (and, for some circumstances, perhaps more appropriate) hypotheses were adopted, providing a wider range of results. Jeffrey Williamson questioned the time dimension of the model and suggested allowing for capital accumulation as a response to liberalization and migration. He also stressed the potential importance of demographic structures in the trade/migration relationship. Arye Hillman asked about the receiving country government’s objective function. He also invoked the Coase theorem in suggesting that, in theory, potential migrants could either be paid by the receiving country to stay at home, or be forced to pay to immigrate. Ulrich Kohli pointed out, however, that it is unlikely that the Coase theorem can operate in practice when countries cannot effectively enforce property rights regarding who crosses the border. James Markusen argued that, in the US/Mexico case, it is clear why the United States wants to limit Mexican immigration: Mexicans consume more public goods than they contribute in taxes.
The conference succeeded in its objective of bringing together the theory and empirics of the ‘trade and factor mobility’ issue, with the papers approaching the topic from a variety of viewpoints. It is clear that theory remains somewhat ambivalent about the nature of the relationship and that there is a wide variety of models from which to choose. Care must be taken, therefore, in selecting a framework appropriate to the setting in mind. Furthermore, empirical work on the topic is still quite young, with the result that clear conclusions concerning the trade and factor mobility nexus must await additional research.
CEPR, 77 Bastwick St, London EC1V 3PZ
Tel: +44 (0)20 7183 8801 Fax: +44 (0)20 7183 8820
Email: email@example.com Webmaster: firstname.lastname@example.org