By Dotsey, Michael; Reid, Max
Economic Review , Vol. 78, No. 4
The U.S. economy has experienced nine recessions over the post-World War II period. Whether the causes of these recessions are primarily real or monetary has been widely debated. In this paper we examine two seemingly conflicting results regarding the primary causes of contractions in U.S. economic activity since the end of World War II. One set of results obtained by Hamilton (1983) shows that major downturns in U.S. economic activity are associated with prior exogenous increases in oil prices, while another set of results established by Romer and Romer (1989) indicate that exogenous tightening in monetary policy is the major cause of declines in industrial production and increases in unemployment.
We note that while Hamilton is careful not to rule out the role policy may play in determining economic activity, he does place heavy emphasis on the effects of oil prices. Romer and Romer are more emphatic in their belief that they have uncovered exogenous monetary policy and that this policy is solely responsible for the events they study. We wish to examine their contention by testing whether real disturbances could simultaneously be influencing Federal Reserve policy and downturns in economic activity. Given Hamilton's work and the fact that four of the six episodes that the Romers associate with exogenous monetary policy are very close to oil price shocks, we check to see if these shocks are responsible for their results. We find that including oil prices in their analysis makes monetary policy as specified by the Romers insignificant.
Negating the results of Romer and Romer does not imply that monetary policy plays no role in determining economic activity. Following McCallum's (1983) suggestion, which is also implemented by Sims (1991), we use interest rates as a proxy for monetary policy in Hamilton's model. Specifically, we use the federal funds rate and the spread between the ten-year Treasury bill rate and the funds rate as depicting the relative tightness of monetary policy. In this setting we find that both oil price increases and movements in interest rates are significant in our statistical analysis of real GNP and employment. Further, an analysis of impulse response functions and variance decompositions indicates that innovations in both oil price increases and interest rates are associated with subsequent movements in real economic activity.
II. LITERATURE REVIEW
Here we review the analysis presented in the papers of Romer and Romer (1989) and Hamilton (1983) that are of primary interest to the subject of this paper. More broadly, these two papers represent contributions to the ongoing debate in macroeconomics concerning the primary source of economic fluctuations. Are these sources primarily real or monetary?
Romer and Romer (1989) adopt the perspective of the seminal work of Friedman and Schwartz (1963) that monetary policy explains much of the variation in economic activity. In performing their investigation of the relationship between monetary policy and movements in U.S. economic activity over the post-World War II period, they use Friedman and Schartz's methodology, which they term the "narrative approach." This approach attempts to isolate historically exogenous monetary policy and then analyze the effects of such policy on economic activity. Whether or not they have accurately identified exogenous monetary shocks is the basis of our critical evaluation of their work.
The Romers' (1989) conclusion is that six of the eight postwar recessions in their data set were caused by contractionary monetary shocks. The identification of these monetary shocks is based on examinations of the "Record of Policy Actions" of the Board of Governors and the Federal Open Market Committee (FOMC), as well as the minutes of the FOMC prior to their discontinuance in 1976. The Romers identify as shocks, "only episodes in which the Federal Reserve attempted to exert a contractionary influence on the economy in order to reduce inflation" (p. …