Trade Policy
The problems of not being perfect

The effectiveness of trade and industrial policies when markets are imperfectly competitive has been the subject of continuing debate among economists and policy-makers. The policy conclusions which can be drawn from this debate appear very sensitive to assumptions concerning both market structure and the behaviour of firms. Economists have produced contradictory results from somewhat different models, causing in some confusion. The various analyses of export subsidies which have emerged provide a good example of this. If firms sell goods at a price in excess of marginal cost then there is a possibility that welfare may be raised by using export subsidies to expand the output of domestic firms. This argument has been extensively analysed for oligopolistic industries by Krugman, Brander and Spencer, Dixit, Eaton and Grossman. If the assumption that there is a fixed number of firms is replaced by one of free entry, however, then the export subsidy will attract new firms into the industry, and so may not expand the output of individual firms or change their average costs. Recent papers by Horstmann and Markusen and by Venables demonstrate that in this case the effect of export subsidies depends crucially on whether international markets are segmented or integrated. Segmentation means that firms can set sales in domestic and foreign markets independently, so that the price of a product may be different in the two markets. Integrated markets, however, allocate total sales between countries until product prices are equalized in all markets, and firms therefore can control only their total world sales.

This example suggests that policy conclusions are sensitive to assumptions about industry structure, i.e. oligopoly (no entry) versus free entry, and market structure, i.e. segmentation versus integration. In Discussion Paper No. 120, James Markusen and Research Fellow Anthony Venables examine policy effectiveness using a single simplified model which can accommodate a variety of assumptions concerning industry and market structures. The authors consider four permutations; these are generated by oligopoly versus free entry of firms and segmented versus integrated markets. The model permits conflicting results, as well as analysis of such interesting cases as oligopoly with integrated markets and the role of transport costs with assumptions of free entry and segmented markets.

In order to highlight the effects of industry and market structures in a relatively transparent way, Markusen and Venables make some sweeping assumptions. In their model, there are only two economies, with symmetric structures, demand curves are linear, only specific taxes and subsidies are considered, and firms' marginal costs are constant. The authors note that although this sacrifice of generality permits straightforward analysis of the effects of industry and market structures, the policy implications of this simplified model may not be particularly robust.

Markusen and Venables analyse the effects of tariffs and export subsidies for each of the four possible combinations of market and industrial structures, and begin by comparing the segmented and integrated market hypotheses. They find that a small import tariff or export subsidy improves welfare more (or reduces it less) in segmented markets than if markets are integrated, independently of whether free entry or oligopoly is assumed. The intuition behind this result is that with segmented markets the effect of a policy is concentrated on the market in question. A country imposing an import tariff on a good, for example, will force down the price of imports of that good, but if markets are segmented this will have no adverse consequences on its exports of the good. Conversely, with integrated markets the effect of a tariff is spread out over two markets, making it less beneficial.

Under oligopoly, a small import tariff or export subsidy improves welfare more (or reduces it less) than if there is free entry, independently of the assumed market structure. The intuition behind this result is also fairly straightforward. Under oligopoly, such policies tend to expand the size of existing firms. This reduces firms' average costs of production, and these lower costs increase consumer welfare. With free entry, however, part of the industry's expansion takes the form of new firms, with correspondingly smaller changes in the outputs and average costs of existing firms.

Free entry by firms generally implies that an export subsidy reduces welfare. Markusen and Venables find that, whatever the market structure, world welfare under free entry is maximized if there are no taxes or subsidies on trade. Unlike the oligopoly case, free entry implies that reciprocal tariffs or subsidies have no beneficial effects on welfare through the expansion of firms' output. Taxes or subsidies do alter the relative prices faced by consumers, and this distortion reduces welfare. This result must be qualified, however, if there are positive transport costs and if goods are close substitutes. In this situation, with free entry and segmented markets an export subsidy will raise welfare.

Although these results are not robust with respect to key assumptions in the model, they nevertheless help us understand the roles of market structure and entry assumptions in policy analysis. Markusen and Venables conclude that policy is more effective when markets are segmented than when they are integrated, and that, provided transport costs are small, policy is more potent when the number of firms is fixed than when there is free entry. Under free entry, an export subsidy will only raise welfare when markets are segmented, there is a high substitutability between goods, and transport costs are positive. The possibility of economies of scale emerging from a small export subsidy or import tariff means that such policies are far more likely to be beneficial under oligopoly than under free entry.
Other work by Anthony Venables on new approaches to international trade theory incorporating imperfect competition may be found in Discussion Paper No. 74 and in his article in Bulletin No. 12.


Trade Policy with Increasing Returns and Imperfect Competition: Contradictory Results from Competing Assumptions
James Markusen and Anthony Venables

Discussion Paper No. 120, August 1986 (IT)