Did monetary forces cause the Great Depression? A Bayesian VAR analysis for the U.S

This paper recasts Temin's (1976) question of whether monetary forces caused the Great Depression in a modern time series framework. We adopt a Bayesian estimation and forecasting algorithm to evaluate the eects of monetary policy against nonmon-etary alternatives, allowing for time-varying parameters and coecient updating. We nd that the predictive power of monetary policy is very small for the early phase of the depression and breaks down almost entirely after 1931. During the propagation phase of 1930-31, monetary policy is able to forecast correctly at short time horizons put invariably predicts recovery at longer horizons. Conrming Temin (1976), we nd that nonmonetary leading indicators, particularly on residential construction and equipment investment, have impressive predictive power. Already in September 1929, they forecast about two thirds of downturn correctly. Our time varying framework also permits us to examine the stability of the dynamic parameter structure of our estimates. We nd that the monetary impulse responses exhibit remarkable struc-tural instability and react clearly to changes in the monetary regime that occurred during the depression. We nd this phenomenon to be discomforting in the light of the Lucas (1976) critique, as it suggests that the money/income relationship may itself have been endogenous to policy and was not in the set of deep parameters of the U.S. economy. Given the instability and poor predictive power of monetary instruments and the strong showing of leading indicators on real activity, we remain skeptical with regard to a monetary interpretation of the Great Depression in the U.S