Three essays on strategic trade policy: precommitment, time consistency, and effects of a ratio quota

Abstract

This dissertation investigates the strategic behavior of firms under imperfect competition in international trade;We utilize the Spencer-Brander multi-stage game in which two firms; one foreign and one domestic, compete as Cournot rivals in a third country. A three-stage game is characterized as: (1) one home government sets subsidies on both export and investment, (2) firms choose the levels of investment, and (3) firms choose output levels. We show that it is not generally optimal to choose investment levels that minimize costs in a Spencer-Brander model. Precommiting both investment and output levels by a home government (as a normal Stackelberg leader) is shown to yield a solution superior to the Spencer-Brander outcome;However, neither of these precommitment models is time consistent because the ex ante subsidy in a precommitment model is larger than that which appears optimal ex post. We show that the time consistent solution can dominate the precommitment solutions, a result that is at odds with the conventional wisdom concerning the inferiority of the time consistent solution. This result emerges because neither precommitment solution is made optimally contingent on foreign investment;We examine the effects of a ratio VER (voluntary export restraints). We claim that under the implementation of a ratio VER, the strategic behavior of firms is most appropriately modeled as simultaneous play, not as Stackelberg leadership for the home firm. Under simultaneous play, we show that (1) there is no equilibrium in pure strategies, and (2) the unique mixed strategy exists with the domestic firm's randomization of its production. In comparing tariffs and quotas, we find that the foreign firm prefers a VER to a tariff system, when the two are chosen such that the home firm's profits under a VER are the same as under a tariff system.