Berkeley Program in Law and Economics

The conventional wisdom maintains that the emergence of the International Tax Regime since the 1920’s is a miracle. This is so, the argument goes, because taxes are the last topic on which a Hobbesian observer would have predicted sovereign nations to reach a consensus given the zero-sum nature of the game: one’s country gain in revenue is another’s loss. Conversely, this Article predicates that the International Tax Regime is not a miracle, but rather an intelligible (sometimes failed) attempt to solve problems in the strategic interaction among nations in the allocation of international tax bases.

The purpose of this Article is three-fold. First, it elaborates a hypothesis of the pattern of evolution of international tax law since 1913 onwards using game theory as the theoretical framework. Second, it focuses on a structural element of the International Tax Regime which has failed to solve the strategic problem of apportionment of cross-border business income: the arm´s length standard. It shows, using the rule/standard distinction as the theoretical framework, why the arm’s length standard has been unsuccessful in both the developed and developing worlds in solving the said apportionment problem (normally named the Transfer Pricing Problem). Finally, the Article suggests how the arm´s length standard should evolve for increasing the likelihood of solving the Transfer Pricing Problem in both developed and developing countries. The proposal has been written in such a way as to facilitate its addition to Article 9 of the OECD or UN Model Tax Conventions on Income and on Capital.