Real exchange rates and inflation in exchange-rate based stabilizations: an empirical examination

Considerable research has focused on explaining why currencies appreciate in real terms after the nominal exchange rate is stabilized, but this research generally has taken a theoretical approach, and rarely has tested its hypotheses empirically. In this paper I estimate a simple error-correction model for Mexico, based on the Salter-Swan framework, in which inflation is determined by (1) the gap between the actual real exchange rate and the exchange rate that clears the market for non-traded goods, and (2) persistence effects of past inflation. Using this model, I decompose the excess of Mexican inflation in 1988-94 over peso-adjusted international inflation rates--that is, the real appreciation of the peso--into that part attributable to the initial undervaluation of the peso, that part explained by the subsequent expansion of domestic demand, and that part attributable to inertial inflation. The results indicate that the effects of inertial inflation in appreciating the real exchange rate were quite temporary, lasting only about a year after the stabilization of the peso in 1988. Of the real appreciation that took place between 1988 and 1994, about half was attributable to the expansion of domestic demand-which appreciated the equilibrium real exchange rate in the non-tradeables sector--and about half reflected the correction of the initial undervaluation of the real exchange rate relative to its equilibrium level in the non-tradeables sector. Finally, the paper uses the model to illustrate the impact of various prospective exchange rate policies on inflation and the real exchange rate in Mexico.