Does money matter? While different schools differ on how to incorporate short-run non-neutralities of money, there is little disagreement over the long-run neutrality hypothesis. The main controversy is empirical, and in Discussion Paper No. 1042, Research Fellow Axel Weber uses explicit tests of coefficient restrictions in bivariate vector autoregressive models to evaluate evidence for the G7 countries, notably Germany. The author carries out neutrality and superneutrality tests, verifying the long-run vertical Phillips curve hypothesis and the Fischer hypothesis, which predicts a unit long-run effect of inflation on nominal interest rates. He finds little evidence against the long-run neutrality of various monetary aggregates in the majority of G7 countries, though the contrary occurs with superneutrality. Evidence from G7 countries does not reject a long-run vertical Phillips curve, while the Fischer effect is rejected by the US and UK data, but not for the remaining G7 countries.

Since, in the case of Germany, tremendous changes have occurred in the policy environment over the past three decades, Weber carries out tests of the invariability of the parameters (the Lucas critique) for this particular country, using a recursive estimation approach. He finds that inference about both the long-run neutrality and superneutrality of money is not invariant with respect to time, with major changes occurring at policy sensitive points in time in the aftermath of major real shocks. No evidence is found of variation of the estimates for the long-run vertical Phillips curve and the Fischer hypothesis.

Testing Long-Run Neutrality: Empirical Evidence for G7 Countries with Special Emphasis on Germany
Axel A Weber

Discussion Paper No. 1042, October 1994 (IM)