Chapter 13 The ADA and the DSU of the WTO

Beginning with the assumption that antidumping laws are used to address adverse shocks in import-competing industries, this chapter provides an explanation for the infrequent utilization of the Dispute Settlement Understanding under the Antidumping Agreement. It does so with a very simple model that represents the shock by a one-dimensional random variable. This is found on an interpretation of the ADA as a de facto escape clause. ADA signatories are homogeneous, which enables the representation of the expected frequency of shocks over each member's import-competing sectors by the binomial distribution with identical parameters. The explanation for the infrequency of utilization of the DSU invokes a repeated Prisoners’ Dilemma with two levels of cooperation in an infinite horizon game. The high level is free trade in all sectors. The low level is the application of ADA duties in sectors incurring the shock in a manner that is consistent with the ADA. The high level of cooperation in all sectors in every period is not sustainable for any degree of patience. A convex combination of the high and low levels of cooperation is sustainable for some degrees of patience under the folk theorem. However, this combination of cooperation is attainable only with the support of the DSU. The extent of importance of the DSU depends on the completeness of information with which signatories are endowed. With complete information, dispute resolution does not occur in equilibrium. However, its presence supports cooperation through its mandate to sanction retaliation. If filing were prohibitively costly, disputes would never arise, and cooperation would be expected to evaporate. In the instance of incomplete information with costless filing, disputes would occur in equilibrium whenever an AD action was taken. In the most realistic circumstance, that of incomplete information and nonprohibitive filing costs, disputes would arise only when the number of AD actions exceeded their (common) expectation. This provides a conceptual explanation for the observations of Tarullo (2002) and Bown (2005) that ADA disputes are infrequent.