| Abstract: | This paper advances a model of multilateral trade negotiations to analyze the effects of the most-favored-nation clause (MFN) on international trade agreements. Negotiations are modeled in a three player, non-cooperative, dynamic bargaining framework that admits the possibility of both bilateral and multilateral agreements. The central result is that bargaining in the presence of MFN results in Pareto efficient, mutually advantageous, multilateral trade agreements. The free-rider problem commonly attributed to the presence of MFN does not arise, and, under a condition of symmetry, each country receives equal gains (or reciprocity) from the agreement. In the absence of MFN, many of these properties may not hold. Examples are given in which at most two of the three countries benefit from agreement. These results suggest that many of the criticisms levied against the MFN clause are misplaced; moreover, attempts to replace unconditional MFN with conditional MFN may sacrifice many of the long-held values of the GATT. |
