Intra-National Evidence of the Fisher Effect

There is a large literature devoted to the size of the response of nominal interest rates to changes in expected inflation, the Fisher effect. The interest in the topic stems from the implications of the Fisher effect for both asset pricing models and monetary neutrality propositions. Unfortunately no consensus has been achieved concerning the Fisher effect. Part of the problem for this has been the low power of the tests employed in Fisher effect studies. In this study we use a data set on local city-wide interest rates and local mortgage interest rates to construct a panel of Fisher relationships within the United States. There are well known power increases in both unit root and cointegration tests using panel techniques. Our panel results, however, are not much different from the single group estimates. The size of the Fisher effect depends more on the particular estimator, i.e. OLS, DOLS, MLE, than on whether used in a single group specification or as a panel.