Adam Check (University of Oregon) presented preliminary work which attempts to relate the observed reduction in correlation between the money supply and inflation in the USA after the mid-80s to changes in Fed policy. The quantity theory of money states that changes in the money supply should cause one-for-one changes in the price level. Data on the money supply and aggregate price levels support this theory only until the mid 1980s, after which the correlation between the variables appears to be much lower (possibly zero). There are two leading hypothesis for this change. First, the way in which money is measured may provide an inadequate representation of the true money supply. Second, changes in Fed policy e.g. changing the relative weighting of fulfilling the Dual Mandate could have affected the observed correlation.

Sargent and Surico (2011) showed that changes in the monetary policy rule in a New Keynesian model could change the parameter estimates in a regression of money on inflation. The model is estimated using Bayesian techniques, and contour plots showing ranges for the regression coefficient mapped in policy parameter space reveals that as policy makers responded more aggressively to inflation, the correlation between money and inflation fell.

Slides from the presentation can be found here.