The relationships between European labour costs and European competitiveness were the subject of a joint Zentrum für Europäische Integrationsforschung (ZEI)/CEPR workshop on 'Product Quality, Labour Productivity and Trade' held in Bonn on 23/4 January 1998. The workshop organizer was Stefan Lutz (Union Bank of Switzerland, Zürich, ZEI, Universität Bonn and CEPR).

The workshop was intended to address several questions. Are high labour costs in some European countries, such as Germany, a competitive disadvantage, threatening a flight from investment in those countries and rising unemployment? Or does the high level of skills and consequent higher productivity of workers still give European firms a headstart in the production of high-quality goods? How are European manufacturers coping with the growing competition from low-wage developing countries and what are the likely consequences of this new competition for European wages and employment? And what industrial, trade or labour market policies, if any, should European authorities embrace in dealing with these issues?

Johan Torstensson (Lunds Universitet) presented 'Economic Geography, Comparative Advantage and Trade within Industries: Evidence from the OECD’, which was written with David Greenaway (University of Nottingham). The paper addressed the issue of the determinants of the patterns of specialization in intra-industry trade. Monopolistically competitive models offer economies of scale and horizontal product differentiation as explanations for intra-industry trade. Yet, in their simplest form, they do not help in determining the pattern of specialization within an industry. The authors presented a simple unifying theoretical framework which considered both market access and comparative advantage (due to relative factor endowments) as possible explanations for this pattern of specialization. They then tested the model empirically using data on Swedish imports, on the assumption that these are representative of the broader OECD patterns.

The results support the assumption that both the size of the home market and relative factor endowments, in particular human capital, are important in explaining the choice of quality by producers in a certain industry. The data do not, however, support the presumption that falling trade costs should lead to higher geographical concentration of the production of certain quality spectra over time. Finally, the paper measured the relative importance of vertical, as opposed to horizontal, intra-industry trade. Vertical intra-industry trade – defined as the exchange of low-quality for high-quality versions of the same good – was found to be quantitatively more important than horizontal intra-industry trade, defined as the exchange of different varieties of similar goods. Moreover, while the share of the former seemed to have increased over time, the latter had stagnated in the period from 1981 to 1994.

For Lionel Fontagné (Université Paris I and CEPII, Paris), the distinction between vertical and horizontal product differentiation was important in that the former is likely to imply larger adjustment costs than the latter. Tito Cordella (IMF and CEPR) noted that the theoretical implications of a positive relationship between market size and higher quality depend crucially on the assumption that higher quality implies higher fixed costs rather than higher marginal costs, which may not be the case in many industries.

A quantitative assessment of the importance of intra-industry trade in quality-differentiated products was the central concern of 'Intra-Industry Trade and the Single Market: Quality Matters', presented by Michael Freudenberg (CEPII, Paris) and written with Lionel Fontagné (Université Paris I and CEPII, Paris) and Nicholas Péridy (CEPII, Paris). Using a highly disaggregated dataset of intra-EU trade, the authors were able to examine the changing patterns of specialization and trade within the European market over the period from 1980 to 1994. Trade patterns were identified by breaking up total trade into three types: inter-industry trade; intra-industry trade in horizontally differentiated products; and intra-industry trade in vertically differentiated products. In this case, the distinction between vertical and horizontal intra-industry trade relied on a comparison between the unit value of imports and the unit value of exports for a given good: if these values differed by more than a certain range, trade was classified as vertical, on the assumption that differences in prices reflected differences in quality.

The results showed that overall intra-industry trade in the EU had increased substantially over the period considered, mostly because of a pronounced increase in the exchange of vertically differentiated products. By contrast, exchange of horizontally differentiated varieties did not change significantly. This suggested the existence of a qualitative division of labour in the EU, with adjustments taking place within industries along the quality spectrum, rather than between industries. The second part of the paper used econometric techniques to investigate some of the factors that may have influenced these changes in the pattern of trade. The main conclusions were that vertical intra-industry trade increased with differences in the economic size of member countries, and that the removal of Non-Tariff Barriers did not substantially affect the composition of intra-industry trade.

Carlo Scarpa (Universitá di Bologna) suggested particular care was needed in the use of the dispersion of unit values as a proxy for vertical intra-industry trade, since differences in prices may reflect factors other than differences in costs and quality. Furthermore, the results may be fairly sensitive to the arbitrary choice of a given difference in unit values for classifying goods as vertically rather than as horizontally differentiated, especially if the threshold is assumed to be equal across all products.

Tito Cordella (IMF and CEPR) tackled the issue of the short-term social cost – particularly as perceived by workers – of the relocation of economic activity when barriers to trade, or to foreign investment, fall. His paper, 'Globalization and Relocation in a Vertically Differentiated Industry', written with Isabel Grilo (CORE, Université Catholique de Louvain), presented a partial-equilibrium model of a duopoly with vertically differentiated products, in which firms can relocate their production activity abroad, bearing a fixed cost, and the incentive for relocation comes from a given wage differential. Depending on the relative magnitudes of the fixed cost and the wage differential, four equilibria can arise in the globalization scenario: one in which neither firm relocates; one in which both firms relocate; and two in which either the high-quality or the low-quality firm chooses to produce abroad.

The welfare analyses of these equilibria show that workers are worse off relative to autarky when both firms relocate, but globalization may be welfare-improving when one (and only one) of the two firms relocates. An interesting implication of the model is that, in the latter case, it matters which of the two firms relocates: in an economy which has a strong preference for high quality under marginal-cost pricing, the relocation of the firm producing the high-quality good is preferable in welfare terms. In their conclusions, the authors stressed that large parts of society might bear non-negligible costs in the short-run as a consequence of globalization. Moreover, since the model revealed the possibility of multiple equilibria with quite different welfare implications, there was a role for industrial policy in helping to select the types of firms that relocate their activities abroad.

Stephanie Rosenkranz (Universität Bonn and CEPR) suggested that firms could change the quality of their products in response to the relative preferences of their domestic economies for higher or lower quality. Some participants pointed out that, if a general equilibrium view is taken, the social costs of relocation are likely to be less substantial. The authors replied that, although this was true, their purpose had been to investigate the short-run cost of relocation, as perceived by workers and by public opinion, and as it usually enters the political debate in developed countries.

Offering a theoretical explanation for the recent empirical finding that there exists a negative relationship between wages and unemployment when estimated across regions was the main purpose of 'Regional Unemployment and Specialisation' presented by Uwe Walz (Universität Tübingen and CEPR) and written with Peer Ritter (Universität Tübingen). They considered a region with a finite number of firms and a pool of workers with a given distribution of skills. A larger region implies more specialized firms and therefore a more specialized demand for skills, which drives up the wage rate. Using a model of equilibrium unemployment, based on efficiency wages, they also showed that a larger region will have a lower level of unemployment because of the better matching between firms' demand for, and workers' supply of skills. These two effects yielded the sought-for negative relationship between the level of wages and unemployment across regions.

Stefan Lutz pointed out that one of the main policy implications is the desirability of a more flexible educational system, that is to say, a system that can provide each future worker with very specialized skills, and offering a wide range of skills in the aggregate. This would reduce the skill-mismatch problem, thus increasing wages and reducing unemployment.

Stefan Lutz (Union Bank of Switzerland, Zürich, ZEI, Universität Bonn and CEPR) and Alessandro Turrini (CESPRI – Bocconi, Milano, and CORE, Université Catholique de Louvain) presented 'Skills, Labor Costs and Vertically Differentiated Industries: A General Equilibrium Analysis'. Their paper addressed the much-debated issue of whether high European labour costs (both wage and non-wage) really have a negative bearing on firms' profits. The trade-off between labour costs and industry profits was analyzed within a general-equilibrium model in which one industry is oligopolistic and vertically differentiated. Producing goods of a higher quality requires the employment of a larger amount of skilled labour. Given an underlying skills distribution, the model determines prices, qualities, profits and wages. The non-linearity of the model did not allow for a closed solution, and simulations were performed. The results showed that high wages may be associated with high or low profits, depending on the skills endowment of the economy. The model also shed some light on the effects of the reduction of labour taxation: depending on the skills endowment, it may either lower or increase industry profits.

Isabel Grilo commented on some technical aspects of the model and suggested some extensions; it may be interesting to study what happens when substitutability between different skill levels is reduced and when different degrees of dispersion of skills in the economy are considered. Richard Baldwin (Graduate Institute of International Studies, Geneva, and CEPR) observed that the model could have interesting implications for open economies.

Sébastien Jean (CEPII, Paris) presented 'Does Competition from Emerging Countries Threaten the European Unskilled Labour? An Applied General Equilibrium Approach', written with Olivier Cortes (CEPII, Paris). The authors built a computable general-equilibrium model to assess the effects on European factor rewards of fast growth in large developing countries. They divided the world into three regions (Europe, the rest of the developed world, and emerging countries); each region's production was classified within two perfectly competitive industries and 11 monopolistically competitive industries; and products were differentiated by country of origin, i.e. goods produced in Europe and goods produced in emerging countries were not perfect substitutes in consumers' preferences. Within this framework, they introduced a shock in the form of a doubling in the size of the emerging countries' economies, but which leaves their relative factor endowments unaffected.

The results suggested that the effects of such a shock on European factor prices should be small: the prices of skilled labour and of capital would increase by 0.7% and 0.1% respectively, and the real wage of unskilled workers would decrease by only 0.1%. Michael Pflüger (Universität Freiburg) pointed out that some interesting features of European labour markets had been left out of the model; if wages are rigid, for example, then the effects in terms of unemployment could be non-negligible. Another point, raised also by other participants, was that assuming a process of growth which leaves factor proportions unchanged is quite unrealistic and could strongly influence the results.

'How to Compete in Quality: Factors Explaining Vertical Intra-Industry Trade' by Marion Jansen (Universitat Pompeu Fabra, Barcelona) presented an attempt at an empirical test for the determinants of countries' performance in high-quality exports. The paper proposed a new measure for product quality differences, based on the ranks of prices of imports from different countries, rather than on the usual criterion of dispersion of unit values of imports and exports. Following the suggestions of trade models with vertically differentiated products, R&D, workers' skills, physical capital and country size were suggested as potential determinants of product quality. The results showed that a high level of education, used as a proxy for workers' skills, had a strong and significant effect on the quality of exports, whereas physical capital seemed to play almost no role. The regression exercise also showed a negative effect of income on the quality of exports – a rather unexpected result on theoretical grounds.

Luca Lambertini (Università di Bologna) suggested particular care was required in using differences in prices as a proxy for differences in quality. Differences in prices could reflect strategic behaviour by firms, different firm sizes and technologies, and many other factors that had very little to do with product quality per se. This criticism applied to Jansen's paper as well as to most empirical work on quality and trade. An alternative approach is offered by Bresnahan who measured differences in quality within a certain industry by means of the technical characteristics of products using the hedonic prices approach. Lambertini also pointed out that since every conceivable theoretical framework seemed to yield a strong positive relationship between product quality and income, the empirical finding of a negative relation was indeed puzzling. Some participants observed that the potentially high positive correlation between GDP and education could give rise to problems of collinearity, thus imparting a downward bias to the coefficient on income.

Hylke Vandenbussche (UFSIA, University of Antwerp and CEPR) presented 'Globalisation and the Effects of National versus International Competition on the Labour Market: Theory and Evidence on Belgian Firm Level Data', written with Jozef Konings (CES, Katholieke Universiteit Leuven and CEPR). Their paper developed a theoretical framework showing that important differences exist between national and international competition and the consequent effects on national labour markets and welfare. The effects of an increase in national competition versus an increase in international competition were studied under endogenous wage formation and different types of market structure.

The model predicted that extra foreign competition would have stronger employment and wage effects than extra domestic competition. Furthermore, extra foreign competition would always reduce domestic employment, but the scale of job reduction would be lower in unionized sectors, since some of the competitive pressure would be offset by a reduction in the wage mark-up. The authors' empirical test of their theoretical model, performed on firm-level data for Belgium (augmented by the results of their own survey), confirmed the main predictions. Richard Baldwin appreciated the quality and the interesting conclusions of the empirical work, and suggested some simplifications which might make the thoretical framework more general and easier to work with.

Imperfect information on the part of consumers about product quality was suggested as a possible new reason for foreign direct investment in vertically differentiated industries in 'Intra-Industry Trade and Locational Choice under Incomplete Information: Empirical Evidence from German-Polish Trade and FDI Data'. This paper was presented by Christian Wey and Andrzej Cieslik, and written with Justus Haucap (all Universität des Saarlandes). The authors focused on 'experience goods', i.e. goods whose quality is unknown to the consumer before consumption takes place, and they assumed that firms can relocate, at a cost, between developed and developing countries. Costs of production were assumed to differ in the two locations. If, in forming their beliefs about the quality of a good, consumers took into account the location in which it was produced, then location of production in a high-cost country would be a signal of high quality. In this setting, there exists an equilibrium in which high-quality firms will locate in developed countries and low-quality firms in developing countries, thus increasing the scope for vertical intra-industry trade and reducing the scope for FDI in experience goods sectors.