Working Paper: CEPR ID: DP1042
Authors: Axel A. Weber
Abstract: Modern neo-Keynesian, new classical, and real business cycle models typically differ in the degree to which they incorporate long-run or short-run neutrality propositions. Despite their importance, little firm international evidence on the validity of these neutrality hypotheses is available to date. This paper applies a bivariate VAR approach to test the long-run restrictions implied by a number of neoclassical neutrality propositions. The evidence from the G7 countries appears to be consistent with the long-run neutrality of money and the vertical Phillips curve, but the data largely refute the long-run super-neutrality of money and the `Fisher effect' of inflation on interest rates.
Keywords: unit roots; vector autoregressions; long run neutrality; superneutrality; Phillips curve; Fisher effect; Lucas critique
JEL Codes: E31; E43; E52
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Monetary aggregates (E19) | Real output (E23) |
| Monetary aggregates (E19) | Inflation (E31) |
| Monetary aggregates (E19) | Nominal interest rates (E43) |
| Inflation (E31) | Unemployment (J64) |
| Inflation (E31) | Nominal interest rates (E43) |
| Nominal interest rates (E43) | Real output (E23) |