The economic case for trade policy has traditionally depended on the ability of governments to change the prices of imports and exports on world markets to their national advantage. Under imperfect competition, trade policy may also lead to potential gains through changes in the number of firms or the degree of competition in an industry. It may increase the scale at which firms operate and/or strengthen the competitive position of domestic vis-à-vis foreign firms. Although a large number of papers have been written on `strategic trade policy' in recent years, considerable uncertainty remains concerning the likely effects of employing trade policy in a particular industry.

In Discussion Paper No. 412, Research Fellow Anthony Venables uses a computable model of trade under imperfect competition to simulate the effects of an import tariff and an export subsidy for a wide range of industries, under eight different assumptions about the nature of market structure and firms' conduct. The use of such a large number of simulations casts light on the influence of different assumptions about market structure and conduct, and different industry characteristics, on the impact of policy. He seeks to assess the sensitivity of policy to the choice of equilibrium concept employed, the characteristics of industries in which policy is most successfully used, and the size of potential gains.

Venables applies the various theoretical models to a world of six `countries' (France, West Germany, Italy, the UK, the `rest of the EC' and the `rest of the world'), and he calibrates them to each of the industries under study. Whereas for given parameters each theory predicts different levels of output and volumes of trade, as well as different effects of policy change, the calibration procedure reverses this process. Data are available on levels of output and trade, for which different theories then imply different values of unobserved parameters consistent with the industry being in equilibrium.

Venables's results suggest three main conclusions. First, there are gains from the unilateral use of import tariffs (without retaliation) for all industries and for all types of market structure and conduct. These welfare gains are rather small, however, and in very few cases exceed 2% of the value of consumption. Second, there are gains from use of export subsidies in many cases, which are less sensitive to choice of equilibrium concept than a reading of the theoretical literature would suggest. In almost all cases, however, the gains are very small (rarely reaching as much as 1% of the value of consumption), because they represent the net effects of a welfare loss (through the terms-of-trade effect of the policy), and a welfare gain (through the expansion of domestic firms). Also, export subsidies will be lower for an industry with a large volume of exports since a heavy revenue cost will be incurred in subsidizing existing exports, in order to achieve the marginal expansion in exports. Third, gains from either policy will increase with the concentration of the industry.
Venables concludes that models of this type provide a weak case for policy intervention, since even if government gets the policy right, the maximum gains it can expect from it are rather small.

Trade Policy Under Imperfect Competition: A Numerical Assessment
Anthony J Venables

Discussion Paper No. 412, April 1990 (IT)