International Policy Coordination in Dynamic Macroeconomic Models

Recent analyses of the gains to policy coordination have focussed on the strategic aspects of macroeconomic policy making in a static setting. A major theme is that noncooperative policy making is likely to be Pareto inefficient because of the presence of beggar-thy-neighbor policies. This paper extends the analysis to a dynamic setting, thereby introducing three important points of realism to the static game. First, the payoffs to beggar-thy-neighbor policies look very different in one-period and multiperiod games, and thus so do the gains to coordination. Second, we show that policy coordination may reduce economic welfare if governments are nyopic in their policy making, as is sometimes claimed. Third, governments act under a fundamental constraint that they cannot bind the actions of later governments, and we investigate how this constraint alters the gains to policy coordination.