Do temporal causality tests provide information on policy dominance?

Abstract

A number of recent papers have raised serious questions about the validity of the German dominance hypothesis, using Granger (temporal) causality tests. If Germany dominates within the European Monetary System, then German monetary policy, measured by either money stocks or interest rates should Granger (temporally) cause other EMS countries’ monetary policies, but not vice versa. Empirical evidence leads analysts to conclude that the German dominance hypothesis is invalid, or at a minimum, in need of significant reformulation. Explores similar Granger causality tests, using the recent cointegration and error‐correction modelling strategy, for the US and a group of developing countries during the Bretton Woods period, where conventional wisdom suggests that US policy dominated. Finds significant evidence of two‐way causality between the US money stock and the money stocks of a large number of developing countries. These findings raise a serious questions about the interpretation and/or appropriateness of the Granger causality test for investigating policy dominance hypotheses.