This paper argues that Blomberg and Hess's (Journal of International Economics 1997) finding that political variables can be used to predict exchange rate movements better than the random walk model must be seen in the context of the decade and half of previous research which failed to beat this...