Fluctuations, Bilateral Trade and the Exchange Rate Regime

In a recent paper, Frankel and Rose (1998) documented endogenous effects of a monetary union, whereby costs and benefits of the union evolve after its implementation. This paper questions their findings on three grounds. First, their main result that trading partners display relatively more synchronized cycles is not robust to the presence of fixed effects, or variables omitted from their estimation liable to generate both intense trade and synchronized cycles. Second, the cost of giving up independent monetary policy is usually evaluated on the basis of the extent of co-fluctuations between business cycles. We bring into focus which measure of the cycle ought to be used for that purpose. In particular, such measure should in our opinion reflect how synchronized cycles would be in the absence of independent monetary policy. Third, documenting the assumption that fixed exchange rate regimes translate into more bilateral trade has proved elusive. We show that using a bilateral rather than cross-country approach brings little improvement on that front.