This study investigates the stability of long-run log-linear demand functions for narrowly defined monetary aggregates (M1, Monetary Base) in the U.S. during the post World War II period. The hypotheses that the individual time series which appear in such equations (real M1, real Monetary Base, real Personal Income and short-term and long-term nominal interest rates) all have unit roots cannot be rejected. The primary conclusion of this study is that with proper attention to the time series properties of the available data, there exists strong evidence in support of a stable equilibrium demand function for real balances in the post-World War II U.S. economy. The hypothesis of a unitary equilibrium real income elasticity (a velocity function) cannot be rejected. Further, the estimates of equilibrium interest elasticities are approximately -.5 to -.6 for real M1 and -.4 to -.5 for real monetary base. The estimated interest elasticities are significantly different statistically depending on whether long- term or short-term interest rates are used, but the observed differences in these estimates are not of economic significance.

This paper examines the roles played by innovations identified from a simple four-variable VAR characterized by cointegration. Using knowledge of cointegration rank and “textbook” relations that link macroeconomic aggregates, we identify distinct “real” and “nominal” innovations that dictate the long-run behavior of the model. We also examine the explanatory power of transitory innovations that are orthogonal to these permanent shocks. One of the permanent shocks displays all the characteristics of a technology or “supply” innovation, while one of the transitory innovations—identified by imposing short-run price rigid-ity—is interpretable as a “demand” side impulse. The permanent nominal shock bears the imprint of an innovation in aggregate inflation expectations. Historical decomposition and comparison with variables that are external to the model reveals the relative importance of the shocks at various episodes.