Economic analyses have conventionally concluded that an export subsidy is optimal when firms compete on output, but that when firms compete on price the optimal subsidy is negative: although the home firm will lose profits and market share as a result of an export tax, its losses will be outweighed by greater tax revenue on higher world prices. Unfortunately, however, there is no consensus among economists on whether firms typically compete on output or price. The issue has been complicated further by recent suggestions that instead of firms adapting to pre-set subsidy levels, in the real world they enter into negotiations with prospective foreign buyers in anticipation of a subsidy which is granted subsequently; and that the level of subsidy is generally related to price rather than to the volume of sales. Some studies have shown that an export subsidy may be optimal in these circumstances, even when firms compete on price.
In Discussion Paper No. 327, Research Fellow Peter Neary re-examines whether these arguments do reverse earlier results on optimal policy towards a home exporting firm under price competition. He uses a model in which a home and a foreign firm, facing symmetric demand conditions and identical marginal cost structures, produce differentiated products.
Neary first uses his model to confirm the standard result that, where the government sets the subsidy level before the firms compete, it makes no difference whether the subsidy level is related to the price charged or to the sales achieved by the home firm. He then considers in detail the case where the government sets the subsidy level after the firms compete a policy anticipated by firms in their negotiations with prospective buyers. In this case an output subsidy is redundant as output levels are determined before the government chooses its subsidy, so only a price subsidy is relevant in this ex post situation.
In Neary's model a positive, ex post price subsidy does appear to increase home profits and tax revenue. He argues, however, that in order to determine the optimality of a subsidy it is necessary not only to compare this outcome with the level of welfare in the absence of intervention, but also to examine whether such an ex post subsidy yields a higher level of welfare than a subsidy whose level is chosen by the government before firms take their pricing decisions.
Intuitively it seems highly unlikely that the abandonment of a significant degree of autonomy by the government could increase welfare, and the model suggests that this is indeed the case. The government can always attain a higher level of welfare by committing itself to a particular level of subsidy than by announcing that it will offer a subsidy whose level is related to the prices charged by the home and foreign firms.

Since earlier work has shown that when the government moves first and firms compete on price the optimal subsidy is negative, Neary concludes that his analysis reinforces the view that theories of imperfect competition provide an extremely weak justification for export subsidies. Evidence that in practice subsidies are set ex post is best explained as the result of lobbying by firms to increase their profits rather than government attempts to maximize overall welfare, he suggests.

Export Subsidies and Price Competition
Peter Neary

Discussion Paper No. 327, August 1989 (IT)