New international evidence exists on the Fisher effect in the United S
tates, the United Kingdom, Japan, Germany and Canada over the modern (
post-March 1973) floating exchange rate experience. Some puzzling data
has been observed which suggests that the Fisher effect appears to be
strong only for particular sample periods. It has also been reported
in the literature that the Fisher effect is more likely in the long ru
n, and that it is not common across economic regimes. This study sheds
some light on these controversial issues taking advantage of more pow
erful tests for unit roots and cointegration. The findings for the Fis
her effect contrast with previous results, especially those of Mishkin
. The empirical results appear to accord with historical observations
on interest rate movements in the five countries. Some important resul
ts are as follows: (1) the Fisher effect is not robust to policy chang
es; (2) there is strong evidence of a long-run Fisher effect for the U
nited States, Germany, and Japan, but little evidence for the United K
ingdom and Canada; (3) the short-run Fisher effect is only detected in
Germany.