On the Distribution of Exchange Rate Regime Treatment Effects on International Trade

This paper provides evidence of heterogeneous treatment effects on trade from switching among three types of de-facto exchange rate regimes: freely floating, currency bands, and pegs or currency unions. A cottage literature at the interface of macroeconomics and international economics focuses on the consequences of exchange rate regimes for economic outcome such as trade. The majority of contributions points to trade-stimulating average effects of tighter exchange rate tying in general and of currency unions in specific. While there is great variability of the estimated quantitative effects across studies, all of the associated work adopted at least two and most of it all of the following three assumptions: assignment of countries to exchange rate regimes is random, the treatment effect of adopting a currency union is independent of the underlying regime transition, and it is homogeneous and hence fully captured by the average. This paper allows for self-selection into exchange rate regimes conditional on observable characteristics and a given regime state prior to a transition and provides evidence of strong impact heterogeneity on bilateral trade among otherwise observationally equivalent country-pairs.