This paper investigates the consequences of the completion of the internal market in the EC using a computable general equilibrium model of trade under imperfect competition. The focus of the paper is the welfare consequences of reducing trade barriers and the changes in production and trade flows with the rest of the world. Welfare changes by country are reported and these are decomposed by source of gain. Two sets of results are reported: a `segmented market' experiment where trade costs are reduced by an amount equal to 2.5% of the value of trade, and an `integrated market' experiment in which there is the same trade cost reduction plus a switch from a segmented to an integrated market equilibrium. In both cases we find large welfare effects arising from imperfect competition. Intra-EC trade liberalization has pro-competitive effects which make a substantial contribution to the welfare change in the first set of experiments and are the most important component of the welfare change in the second set.

This paper investigates the consequences of the completion of the internal market in the EC using a computable general equilibrium model of trade under imperfect competition. The focus of the paper is on the welfare consequences of a reduction in trade barriers and on the changes in production and trade flows with the rest of the world. We report two sets of results: a `segmented market' experiment in which there is a reduction in the costs of trade by an amount equal to 2.5% of the value of trade; and an `integrated market' experiment in which there is the same trade cost reduction plus a switch to integrated from segmented markets. This means that firms can no longer price discriminate between markets but compete on an EC-wide basis, setting the same producer price in all markets. This experiment implies a much greater pro-competitive effect of trade liberalization as each firm's effective market shares are reduced. Within each of these experiments we consider the short run, where the numbers of firms are held constant, and the long run where there is free entry and exit of firms from any of the EC countries. In both sets of experiments we find large welfare effects from factors associated with imperfect competition. Intra-EC trade liberalization has pro-competitive effects which make a substantial contribution to the welfare change in the first set of experiments and are by far the most important component of the welfare change in the second set of experiments.

The model contains a perfectly competitive composite sector and a number of imperfectly competitive industries. Production uses intermediate goods and five primary factors of production. The imperfectly competitive industries include 13 manufacturing industries, as well as banking and finance. These industries operate under increasing returns to scale and support an equilibrium with intra-industry trade. The changes in trade costs and market structure occur only in the 14 imperfectly competitive industries; the perfectly competitive sector changes in response to changes in other sectors. The model is disaggregated into seven `countries': France, Germany, Italy, the United Kingdom, `Other EC North' (Benelux & Denmark), Iberia (Spain and Portugal) and Greece/Ireland.

In the experiments EC output in all imperfectly competitive industries rises. The distribution of the output changes across industries and depends on a number of interdependent factors. These include the relative share of intra-EC trade in production of each industry, the degree of economies of scale, the degree of concentration, the elasticity of demand for the individual product varieties and cost changes due to general equilibrium changes in input prices. The size of the external trade effects are largest in industries where the production increase is large, and where initial imports are relatively small. In most industries quantity changes are larger in the long run. This occurs due to the exit of firms remaining firms operate on a larger scale, thus giving lower marginal costs and the consequent increase in production and exports. Moving to an integrated market implies a much more pro-competitive policy and so also magnifies the changes. EC exporters benefit substantially from the integrated market scenario. Some industries greatly increase their sales to the rest of the world and particularly in the long run. It is in those industries which benefit the most from the pro-competitive effect of the trade liberalization that the change in EC exports to the rest of the world is the greatest. There is also a very substantial decline in imports of some goods.

We report welfare changes by country of the various experiments which are decomposed by source of gain. Under segmented markets, welfare gains are comparatively small for most of the EC although rising slightly in the long run. Significantly larger gains are reported in the integrated market experiment in the long run. Except for France and `Other EC North', all countries experience welfare gains in excess of 1% of GDP; and for Greece/Ireland and Iberia the gains exceed 2% of GDP. It is worth noting that the policy experiment directly affects only the 14 industries, which account for a little less than 40% of GDP.

The decomposition of the welfare gains by industry and by type of gain indicate that the gains largely follow the pattern of the output changes and that the greatest gains arise from the change in consumer and intermediate surpluses arising from the changes in prices. The paper also reports on the decomposition of the welfare gains by looking at the direct cost saving of the policy, at `distortions' in each industry and at changes in external terms of trade. The pro-competitive effect of the policy accounts for a significant share of the gains in most industries in the long run; whereas in the short run a higher proportion of the welfare gain is generated by the direct effect of the reduction in trade costs.

We undertake two types of sensitivity analysis, rather than assuming a uniform cost reduction of 2.5% of trade costs across all industries, we assess the consequences of a differential experiment that maintains the same overall size of the policy experiment. The aggregate welfare effects are now considerably larger, because the industries in which the trade cost reduction is now greater include some of those with the highest degrees of concentration and economies of scale. Moreover, the changes in the policy effects are not directly proportional to the changes in policy as a consequence of the interplay of general equilibrium effects. The second kind of sensitivity analysis reported is with respect to market structure, where we change the level of disaggregation within each industry i.e. a change in the number of sub-industries.

By modelling intra-EC trade liberalization in a general equilibrium model, we can address several important questions that cannot be treated in a partial equilibrium approach. We find modest effects on factor markets as trade liberalization has differential effects across industries with different factor intensities. General equilibrium effects enter into the accounting of the welfare effects of the policy change, but not with sufficient force to make the order of magnitude of welfare changes different from those that would be derived from a partial equilibrium approach. We also find quite large effects on the EC's external trade, as intra-EC trade liberalization reduces demand for imports from outside the EC. The analysis suggests that there are potentially large welfare gains from factors associated with imperfect competition and hence confirms the importance of the effects of `1992' on market structure and competition. It should also be stressed, however, that much scope remains for improving our understanding of how to model the interaction between trade policy and industrial organization.

1992: Trade and Welfare; A General Equilibrium Model
Michael Gasiorek, Alasdair Smith and Anthony J Venables

Discussion Paper No. 672, March 1992 (IT)