The completion of the European Community's single-market programme is intended to stimulate the European economy by boosting both intra-EC trade and the Community's trade with the rest of the world. CEPR's research programme, The Consequences of `1992' for International Trade, has produced a number of thorough academic analyses of the implications of `1992' for external trade, which the European Commission's own 1988 Cecchini Report largely neglected. Many of these were presented at the programme's final conference, `Trade Flows and Trade Policies After 1992', held in Paris on 16/18 January, which was organized by L Alan Winters, Co-Director of CEPR's International Trade programme, and hosted by the Institut National de la Statistique et des Etudes Economiques. Financial support for this research programme has been provided by the Commission of the European Communities under its SPES programme and by the UK Department of Trade and Industry and Foreign and Commonwealth Office.

General Equilibrium Models
Michael Gasiorek (University of Sussex) and Anthony Venables (University of Southampton and CEPR) presented their paper, `1992: Trade and Welfare; A General Equilibrium Model', written with Alasdair Smith. This developed a computable general equilibrium model of EC market integration with seven EC country blocs and the rest of the world, 15 industries (of which 14 are imperfectly competitive), and five factors of production (four types of labour differentiated by skill level and capital). They used this model to simulate the effects of `1992' in two ways: first assuming a 2.5% reduction in intra-EC trade costs, and second that this cost reduction is accompanied by the abolition of price discrimination across national markets (full integration).

Gasiorek reported that EC output for all the imperfectly competitive industries expands in both cases, which raises factor demands and hence wages in all countries, with skilled workers usually gaining more than unskilled. Because real resources are saved in intra-EC trade, and firms' access to a larger market enables them to reap increased scale economies, integration also brings welfare gains to all economies. These are largest where labour is cheap and initial industrial concentration high, particularly in Iberia and `rest of EC South' (Greece and Ireland), which display output growth of about 5% and reap welfare gains of around 2% of GDP.

In the short run, these welfare gains stem mainly from the direct reductions in trade costs; in the longer run, however, increasing output in imperfectly competitive industries assumes greater importance, although it outweighs the trade cost effect only in the case of full integration. Firm entry and exit have relatively small welfare effects, except in food products, and the loss of tariff revenue as EC imports from the rest of the world fall is also small, with the notable exceptions of food products and textiles and clothing. For the short-run, integrated market simulation, such trade diversion accounts for some 80% of the welfare loss in textiles and clothing.

Gasiorek found that `1992' has the greatest effects for industries with the greatest output growth: EC exports rise significantly (and imports fall) for metal products, non-metallic mineral products, transport, electrical goods, food products and textiles. For the long-run, integrated market simulation, sectoral exports rise by between zero and 35%, while imports fall by between 4% and 40%.

Jan Haaland (Norwegian School of Economics and Business Administration, Bergen) presented his joint paper with Victor Norman, `Global Production Effects of European Integration'. This used a similar model to assess the effects of integration in the Community and the European Economic Area (EEA) which essentially extends the single market to include EFTA on global production and trade patterns. For a model with the same broad theoretical structure and with five main world regions (EC, EFTA, US, Japan and rest of the world), 13 sectors and three (non-tradable) primary factors, capital and two types of labour Haaland reported the results of simulating the same integration experiments as Gasiorek, Smith and Venables. These indicated that the effects of integration on EEA countries' trade are proportionally some ten times greater than for the rest of the world. EFTA stands to gain the equivalent of 3.7% of its traded goods output from participation in the EEA programme almost twice the expected gain to the Community. Such participation also has major sectoral effects: EFTA countries exhibit very strong growth in skill-intensive and engineering industries as full participants in the `1992' programme; but these sectors suffer heavily from the Community's increased competitiveness if EFTA remains outside the single market. Their results support Gasiorek, Smith and Venables's prediction that skilled labour's real wages will rise by proportionally more than those of other factors.

Larry Karp (University of Southampton and CEPR) suggested that Gasiorek, Smith and Venables extend their model to consider the effects of vertical integration and monopsony power. Harry Flam (Institute for International Economic Studies, Stockholm) noted that integration may increase consumer choice: it may reduce the total number of types of goods produced in the EC, but those that remain may be more widely available. Richard Baldwin (Institut Universitaire des Hautes Etudes Internationales, Genève, and CEPR) noted that in Haaland and Norman's model integration affects the return to capital and suggested that they extend it to consider the implications of integration for the capital stock and hence for economic growth.

Case-Studies
Three papers focused in greater detail on specific industries. In `Pharmaceuticals Who's Afraid of 1992?', Gernot Klepper (Institut für Weltwirtschaft, Kiel, and CEPR) noted that the pharmaceuticals market is one of the most segmented and regulated in the Community. National certification procedures, patents and (in Southern countries) price controls all reduce competition and encourage price discrimination. The Commission's directives on transparency of policies and on the authorization of medicinal products should make price discrimination more difficult and also allow the introduction of Community-wide certification: they therefore represent a step towards an integrated market.

Klepper developed models of price-discriminating monopoly and duopoly in which suppliers sell in both price-controlled and uncontrolled markets; he explicitly considered the activities of firms that arbitrage drugs between high- and low-price markets. According to the conventional wisdom, relaxing price controls in the controlled market after `1992' will raise prices in both markets as arbitrage becomes less profitable. If arbitrageurs' cost structures are very convex and the level of their activity is very sensitive to the size of the price differentials, however, a producer in the uncontrolled market may find it more profitable to reduce its price and increase sales. Harmonization of certification procedures across EC markets should also reduce price discrimination, with similarly ambiguous effects on prices in the uncontrolled market.

Klepper suggested that the European Commission's `1992'-related directives on pharmaceuticals are more likely to induce governments to reduce discriminatory price controls than to reduce impediments to arbitrage. From the casual information available on arbitrageurs' cost structures, he forecast that prices in southern Europe will rise in response to these directives, so producer profits should rise at the expense of consumer surplus; but there will be little change in the uncontrolled markets of the North.

Damien Neven (Université de Liège, INSEAD and CEPR) argued that price controls like the length of patents are instruments that governments use to balance the incentive for research and development against the cost of monopoly rents during the period of protection. Models of integration's welfare effects should therefore make price controls endogenous and take account of the effective life of a drug, the prevailing market conditions after patents have expired and the design of suitable incentives for research in the Community as a whole given that governments imposing price controls can `free ride' on research conducted elsewhere.

L Alan Winters (University of Birmingham and CEPR) presented his paper, `Integration, Trade Policy and European Footwear Trade', which investigated the effects of various EC trade policies on footwear. First, Spain and Portugal's accession to the Community led to trade creation and hence enhanced consumer welfare in both the original and new members. This was offset in the original members, however, by trade diversion as Spain's new preferential access reduced imports from more efficient producers in the rest of the world. Spanish and Portuguese producers gained strongly, while their counterparts in the original members suffered some loss of profits.

Second, the welfare losses arising from France and Italy's imposition of bilateral quotas and voluntary export restraints (VERs) on South Korea and Taiwan in 1988 were greatest in the restricted markets themselves, but they also spilled over into other EC markets as reductions in the competitive pressures on French and Italian producers enabled them to increase their prices elsewhere. The EC-wide VERs that replaced them in 1990 were no more restrictive in total, but they significantly redistributed the costs of this protection towards consumers elsewhere in the Community. Italian producers suffered virtually no loss: their gains in the newly protected UK and German markets offset their losses from liberalization in Italy. Since French producers' sales were concentrated in the home market, however, the switch to an EC-wide quota reduced their protection. Taiwanese and Korean producers also gained, since they now earned scarcity rents on all sales to the Community.

Third, footwear imports from East European countries could more than double if the quantitative restrictions they currently face are removed, provided they can overcome their supply constraints. This would confer considerable benefits on EC consumers worth up to 200 million ecu per annum with relatively mild losses for producers.

Kym Anderson (GATT) noted that Winters's partial equilibrium analysis of such a simple market as footwear highlighted the possible complexity of integration's effects. Quantitative modelling of trade policy is required to determine not just the magnitude but even the sign of the relevant changes in some cases.

In his paper, `The Integration of Financial Services and Economic Welfare', Cillian Ryan (University of Birmingham) focused on the intertemporal component of financial services. Because the value of information in the financial services sector is high, intermediaries jealously guard many of the data that are commonly available for other industries; the Cecchini Report therefore used relatively crude methods to assess financial integration's effects.

Ryan developed a general equilibrium model of short- and long-run financial markets in six EC countries to derive alternative measures with which to forecast the effects of the convergence of production efficiencies and the integration of financial markets. His results suggested that the French and Italian financial sectors and to a lesser extent those of the UK and the Netherlands were much less efficient in providing intertemporal financial services than those of Belgium or Germany. The former should therefore enjoy the bulk of the expected gains, which he found to be considerably greater than those predicted by Cecchini. Ryan attributed this discrepancy to his model's broader framework, in which gains accrue to depositors as well as borrowers, and factor reallocations between the goods and services sectors may yield further gains. As countries' financial sectors are forced into line with best- practice techniques, the demand for their services increases as their prices fall; but this does not fully compensate for the job losses arising from their increased efficiency.

Jacob Kol (Erasmus Universiteit Rotterdam) suggested considering intra- and extra-EC trade flows in financial services separately, which may yield insights into the global implications of European integration. Ryan agreed that this would be highly desirable, but the necessary data are not available.

Theoretical Advances
Larry Karp (University of Southampton and CEPR) and Jeffrey Perloff (University of California at Berkeley) presented the first of two papers on new theoretical aspects of market integration, `The Long-Run Value of Inflexibility'. This employed a dynamic duopoly model of domestic adjustment policies to compare the effects on firms' market power of non-linear adjustment subsidies or taxes, which depend on values of inputs or outputs, with those of proportional or linear subsidies, which depend on their quantities. A linear adjustment subsidy enhances a domestic firm's strategic power by raising the level of domestic investment, which precommits it to a higher level of output and thus discourages pre-emptive investment by foreign rivals. In contrast, ad valorem taxation raises both the level and rate of growth of adjustment costs as investment increases; it may therefore achieve the same effect as a linear policy with a lower level of government transfers. Far from establishing a new case for strategic trade policy, however, these results indicate the fragile theoretical basis for any such policy and the difficulty of satisfying the informational requirements to apply it successfully.

Harry Flam stressed that the linear tax or subsidy in models of this type applied to adjustment costs rather than investment: in the linear case, a subsidy is always optimal because it represents a tax on disinvestment, which encourages aggressive behaviour and also discourages defensive behaviour. If a non-linear policy with the same effect could be found, the main result would apply with greater force. In practice, however, policies are usually unidirectional: they either encourage expansion or ease contraction, but not both.

Ian Wooton (University of Western Ontario) presented his joint paper with Jan Haaland, `Market Integration, Competition and Welfare', in which they disputed the conventional wisdom that abolishing price discrimination between imperfectly competitive markets will reduce prices in the domestic market so that consumers gain and producers lose. Price-cost margins in standard models are directly related to market shares, and economic integration is represented as a move from segmented to integrated markets. Before integration, firms can price discriminate between markets; domestic firms generally capture larger shares of their home markets, where they can charge higher prices. After `1992', however, their shares of the enlarged Community market will determine the extent of their monopoly power; as firms lose power in domestic markets, their price-cost margins decline.

Wooton used their simple model of international trade in an oligopolistic industry to show that market integration may reduce competition and lead to consumer losses if significant trade costs remain or consumer preferences are biased towards home-produced goods. This is because the initial `segmented markets' case involves `dumping', which becomes impossible once integration is complete. If firms find it more expensive to maintain foreign market shares once the lower prices required in export markets also apply in the home market and therefore reduce their exports, competitive pressure on domestic firms will decline, and prices will rise in all markets. Their simulations indicated that the likelihood of overall losses following integration increases with the substitutability of products; even in the `normal' case where the domestic price falls, integration's welfare consequences for consumers are ambiguous, since import prices also increase.

Konstantine Gatsios (Fitzwilliam College, Cambridge, and CEPR) was not surprised that consumers can lose from integration, but he welcomed Haaland and Wooton's identification of trade costs and consumers' home market bias as possible sources of this result, although he expressed reservations about the value to policy-makers of the distinction between them. Richard Baldwin suggested that the most heavily protected markets are those with the least efficient firms, so there is likely to be a bias in barriers to competition as well as in consumption.

Econometric Investigations
The final two papers concentrated on econometric estimations of the parameters that characterize individuals' responses to the single-market programme. In `The Effects of 1992: Macro-Economic Import Functions with Imperfect Competition', written jointly with Joël Toujas-Bernate, Joaquim Oliveira- Martins (OECD) considered the effects of including product differentiation at the firm level into aggregate import demand equations, in addition to the conventional differentiation of products by location of production. The absence of long-run relationships between relative prices and market shares highlights the weakness of trade models based solely on price effects.

Oliveira-Martins noted that models of international trade with imperfect competition and product differentiation suggest that the number of firms (or varieties) in an individual country also influences its aggregate market share. Using an index of industrial activity as a proxy for the number of firms, he found that a composite price index including firm numbers rather than relative prices alone improved the explanation of market shares in a demand system measuring the shares of domestic, EC and foreign products in the main European markets. The parameter estimates suggested that changes in non-price factors have quite large effects on market shares, while price changes have only a moderate impact.

Lars-Hendrik Röller (INSEAD) questioned the estimators' asymptotic properties since the number of degrees of freedom was small, and he suggested conducting some further tests of the results' robustness. Riccardo Faini (Università degli Studi di Brescia and CEPR) suggested the authors' assumption of homothetic demand may account for their unsatisfactory results for long-run movements in market shares using relative prices alone.

Presenting the final paper of the conference, `Bilateral Trade Elasticities for Exploring the Effects of 1992', Paul Brenton (University of Birmingham) and L Alan Winters (University of Birmingham and CEPR) considered the modelling of `1992' on the basis of price elasticities estimated from disaggregated bilateral trade functions. In their earlier model of allocation of expenditure among domestic and foreign suppliers using an Almost Ideal Demand System on German data for 70 manufacturing industries they found that price elasticities were rather low, with the majority close to unity, while the expenditure elasticity for domestic supplies exceeded unity and that for imports was less than one. In the present paper they first confirmed their basic results by applying such a system to Italian data for 15 manufacturing industries and the (more restrictive) constant elasticity of substitution (CES) demand system to German and Italian data for the same sub-set of industries. They then tested the sensitivity of the price elasticities to changes in functional form by using alternative estimations with instrumental estimates for the domestic price and a first-differenced transformation of the CES model. Their results did not lead to any significant change in the magnitude of the elasticities.

Brenton and Winters concluded that the effects of removing trade barriers as part of the `1992' programme will be relatively small, if the sensitivity of trade flows to price changes is as low as these estimates suggest. Where such barriers constrain quantities, however, there may be very large effects on prices and rents after liberalization, in which case more attention must be directed towards issues such as the distribution of the rents arising from such quantitative restrictions.

Riccardo Faini cautioned that these low price elasticities may simply reflect measurement errors because import prices are proxied by unit values; it is also difficult to find a convincing reason for the expenditure elasticities of domestic goods to be larger than those of imports. Patrick Rey (ENSAE, Paris, and CEPR) noted that Brenton and Winters had estimated the demands of wholesalers rather than final purchasers, and the nature of the distribution sector may have significant implications for its sensitivity to price and income changes.

There was broad agreement at the conference that `1992' will produce significant gains for the EC and for EFTA if the EEA is formed at virtually no cost to the rest of the world. Nevertheless, considerable uncertainty remains about the economic effects of completing the single market: critical parameters are not well estimated, subtle and unmeasured issues such as the extent of consumer bias towards home goods could influence integration's overall effects, inflexibility may confer welfare gains, and general equilibrium feedbacks may be important. These results do not necessarily undermine previous estimates of the effects of `1992', but they do suggest a need for rigorous work of a sophistication not often found in this debate before firm conclusions can be drawn.


'Trade Flows and Trade Policy' edited by L Alan Winters
is available from: Centre for Economic Policy Research, 90-98 Goswell Road, London, EC1V 7RR
ISBN (hardback) 0 521 44020 3