Gross domestic products equals output times the price level (P * Y) which is the same as the money supply times the velocity of money (M * V), ergo recessions (declines in GDP) and inadequate recoveries (failures to catch up to the trend level of GDP) are caused by either declines in the money supply or declines in the velocity of money. What’s the velocity of money? As Bruce Bartlett explains it’s “the rate of turnover of money in the economy—the number of times dollars are spent in the aggregate—which economists call velocity.” And it’s falling:
I should note that there’s another interpretation here. I believe you can’t measure V directly in the MV=PY accounting identity, you just solve for V=PY/M. Which means that your result for V will depend on your measurement of M, and there are different ways of doing this—the M1 measure, the M2 measure, the M3, etc. Bartlett’s math is based on V=PY/M2 which is standard. But for a long time the Fed also calculated a broader metric, M3, composed of M2 + all other CDs (large time deposits, institutional money market mutual fund balances), deposits of eurodollars and repurchase agreements. Starting in March 2006, the Fed stopped collecting this information explaining that M3 “has not played a role in the monetary policy process for many years” because it tracks M2 so closely.
But since 2006 we’ve had a major financial crisis that involved specifically a run in the repo market and a money market fund “breaking the buck” as well as a Fed response that’s involved a sharp increase in the M2 money supply. It’s possible that these two events have created a historically unusual divergence in the M2/M3 trends, and that even though the Fed has been trying to increase the money supply it’s in fact been declining. In my view, the Fed should track M3 again to try to figure out what’s going on. I further note that the ambiguity in what “money” is constitutes a very good reason for monetary policy to target a measurable quantity like P or P*Y rather than money as such.