Strategic Trade Policy with Incompletely Informed Policymakers

Abstract

Ever since the inception of research on strategic trade policy, economists have warned that the informational requirements are high, and unlikely to be met in practice. This paper investigates the implications of incomplete information for a simple, rent-shifting trade policy of the type proposed in Brander-Spencer (1985). We find that asymmetric information undermines the precommitrnent effect of unilateral government intervention. This "screening" effect induces a downward distortion in the optimal subsidy, and it may be so great as to require a tax rather than a subsidy for high levels of uncertainty, given a zero-profit participation constraint. Second, in contrast to the full-information case with strategic substitutes, the introduction of a rival interventionist government reinforces rather than countervails the precommitment effect, by reducing the incentive for the domestic firm to misrepresent its private information. Finally, when a nonintervention-profit participation constraint is substituted for the conventional zero-profit participation constraint to take into account the special relationship between firms and policymakers in trade, the government eschews intervention altogether for high levels of uncertainty.