Quantitative goals for monetary policy: a quantile regression approach

Abstract

A recent paper by Fatas et al. (2006) indicates that formal monetary policy targets–exchange rate, money supply or inflation targets–palpably decrease inflation in a sample of 40 countries. The authors employ various least squares estimations, which pick up the conditional average effect. However, there is wide inflation variability in the authors’ sample. Thus, formal targets could have very different effects in high- and low- inflation countries. Accordingly, we utilize the technique of quantile regression, a method frequently used in labour economics. We find, in a sample of low- and moderate-inflation countries, that formal targets exert no significant impact on low-inflation nations. This result is important for debates over formal targets, such as whether the United States should adopt an inflation target. There are costs and benefits in having formal targets, and finding that targets do not decrease inflation, when it is already moderate, is an important piece of information to consider.