Comments on ‘Ranking alternative trade-restricting

This paper discusses a model of duopolistic competition in the context of international trade. There are two countries, a home and a foreign country, but consumption happens only in the home country. There are two differentiated (substitute) products, one produced at home and one abroad. Three policies that influence competition in the home country are considered: tariffs, subsidies and quotas. Assuming that there is a certain target level of imports, the author considers the effects of the various policies listed above on total surplus at home and abroad. There are two models in this paper, one of quantity-setting, and one of price-setting firms. Both models assume a certain rate of conjectured reaction by the opponent, and calculate the effects of policy changes on market equilibrium for the given conjectures. I feel uncomfortable with the use of conjectural variations. First, note that in an oligopoly game, any outcome can be supported by some pair of conjectural variations. Second, there is a strong game-theoretic argument against the use of conjectural variations because they imply inconsistent use of strategies. Further, attempts in the early 80s to justify them either as short-cuts for more complicated muitiperiod games, or by imposing consistency, have failed. Thus, I will discuss only the results of zero conjectural variations, which correspond to the traditional Cournot and Bertrand games (where quantities or prices are used as strategies, respectively). In the Cournot game, it is shown that home tariffs, foreign export taxes, subsidies on home firms or quotas result in identical production levels and prices. In this setting, total world surplus does not differ across regimes. Surpluses at home and abroad do not change across the regimes of tariffs, home industry subsidies, and quotas. Of course, surplus at home is lower and surplus abroad is higher when the regime switches from tariffs to foreign