ment of an unexpectedly large growth in the money supply will lead to an immediate increase in interest rates. This is often taken as confirmation of the monetarist view that the way to bring down interest rates, even in the short run, is to reduce the rate of money growth. In this paper we provide an alternative explanation of the announcement effect that is consistent with the Keynesian view that, in the short run, the way to bring down interest rates is to expand the money supply. We do not address the issue of inflationary expectations in a formal way, though, in the long run, we would expect the well-known Fisher effect to hold. Looking closely at the announcement effect, it is clear that whatever may be the true short-run relationship between money growth and interest rates, it cannot be inferred from a simple correlation of announcements of changes in the money stock with the resulting changes in interest rates. This is because of several properties of the announcement phenomenon that we capture in the model

A long-run link between money and prices is evident for the United States since the Korean War if the M2 measure of money is used and the velocity of M2 (V2) is modeled as a mean-reverting series. This link between M2 and prices is the basis for a dynamic model of inflation that compares favorably in forecasting exercises with Phillips-curve and more typical monetarist approaches. The behavior of V2 is examined from 1870 to the present, providing a basis for reconsidering previous findings that V2 follows a random walk. Copyright 1991 by American Economic Association.