Pricing, monetary policy, and aggregate fluctuations.

• Author: Rotemberg, Julio J.

My Ph.D. thesis considered the extent to which firms' reluctance to change their posted prices contributes to business fluctuations and, in particular, makes monetary policy powerful in influencing aggregate output. Ever since, much of my research has been concerned both with the role of monetary policy and with the connection between pricesetting by firms and the evolution of GNP.(1) This article summarizes that research.

Price Rigidity and Monetary Policy

In some ways, the hypothesis that firms are reluctant to change their prices plays the same role as the hypothesis of wage rigidity in traditional Keynesian models. Both imply that expansionary fiscal or monetary policies raise GNP. The difference is that with wage rigidity, firms are only willing to raise their output if their marginal cost of production falls as a result of a decline in real wages. By contrast, firms that were not willing to raise prices will increase their output in response to expansionary monetary and fiscal policies even if doing so raises their marginal cost of production. Thus, the existence of price rigidity can rationalize the observation that real wages tend to increase (albeit slightly) when real GDP rises.(2)

In my early work, I also show that this hypothesis is broadly consistent with the relationship between monetary aggregates and GDP.(3) Another way to gauge whether price rigidity explains the effects of monetary policy is to use the recent explosion of research on how the Federal Reserve System conducts monetary policy.(4) This research allows one to identify monetary policy disturbances through movements in interest rates that are different from those that would represent the "usual" reaction of the Fed to economic circumstances. A good way of ascertaining the empirical relevance of the theory is thus to analyze whether it correctly predicts the way the economy reacts to these monetary disturbances.

It turns out that the theory is remarkably successful in this regard.(5) Because there are good reasons to believe that exchange rate movements are linked to price movements, Alberto Giovannini and I analyzed whether a model of this type could help us to interpret the empirical movements in the dollar/deutschemark exchange rate. While the enormous volatility of exchange rates makes it difficult to explain more than a small fraction of their movements, we find some empirical corroboration for the model in these data.(6) There are thus a number of dimensions in which the hypothesis of price rigidity helps to explain aggregate output fluctuations.(7)

More recently, I have returned to the sort of model I considered in my early work. This was prompted by some research that Michael Woodford and I did on the characteristics of business cycles.(8) We discovered that various definitions of the United States business cycle, including the one that is implicit in the NBER chronology of business cycle peaks and troughs, have a very simple statistical property. Periods in which...