by Emi Nakamura and Jon Steinsson

There remains limited empirical evidence on the real consequences of fixed versus floating exchange rate regimes, despite the importance of the question, both in terms of developing open economy macroeconomics models, and in determining appropriate macroeconomic policies. A benchmark reference in this literature is Baxter and Stockman (1989), who conclude that the exchange rate arrangement doesn’t matter. This evidence is deeply unsatisfying. The main empirical exercise is to compare pre-1973 (fixed exchange rate) to 1973-1986 output volatility for a set of countries. However, many other factors changed between the pre and post period aside from the exchange rate regime, making it difficult to know whether the exchange rate regime was the cause of any difference between the two periods.
A more recent and more convincing reference is Broda (2006), who studies the effect of “terms of trade” shocks for countries with fixed versus floating exchange rates. A basic challenge here is that the terms of trade depend mechanically on the exchange rate. Broda focuses on small countries for whom he argues the terms of trade are exogenous. We can show, however, using our identification approach (more on this below) that the assumption of an exogenous terms of trade is violated. Rose (2011) concludes his recent survey on this topic by stating that “similar countries choose radically different exchange rate regimes without substantive consequences for macroeconomic outcomes like output growth and inflation.”
We believe these issues are ripe for re-investigation, given the limitations of the empirical methods that have been used in previous research. propose a new identification approach to study these issues below and present some preliminary results.