The role monetary aggregates should play, if any, in the conduct of monetary policy remains a controversial issue. While no major central bank relies on monetary targeting, both the European Central Bank and the Swiss National Bank attach considerable weight to monetary variables, in particular broad aggregates, in analysing and forecasting inflation. Of course, this practice reflects the fact that in both economies shocks to velocity have historically tended to be small.
Indeed, research on the relationship between money growth and inflation in these economies, in the form of estimated money-demand relationships or models for forecasting inflation, typically have not found much evidence of in-stability over time. By contrast, in the United States, where shifts to velocity have been large and unpredictable, money plays no role in the framework of monetary policy. Indeed, the Federal Reserve recently decided to cease compiling M3 data.
In this paper we study the relationship between money growth and inflation in Japan. Doing so is interesting for several reasons. First, Japan has experienced large fluctuations in the rate of inflation in the post-1970 data we consider. Measured in terms of changes over four quarters, inflation peaked in 1974 in Japan at about 25% and reached a trough in 2002 of -1.4%. Bordo and Filardo (2005 and 2007) study the inflation experiences of a large number of countries since the late 19th century and conclude that while money growth is generally not a useful information variable for inflation in episodes in which it is low and stable, it does contain useful information in periods of high inflation and deflation.
One would for this reason expect to find quite a strong relationship between money growth and inflation in Japan. Indeed, as we discuss below, there is a large empirical literature finding strong linkages between money and prices in Japan, although the stability of these relationships remains an unsettled issue.
