Theory and empirical evidence
Tyler Cowen directed me to a fascinating Thomas Sargent essay on Bob Lucas. This paragraph caught my eye:
Bob invited me to dinner at his house the night before the seminar. At that time, I was watching Neil Wallace rework monetary theory from the ground up and remarked to Bob that I had reservations about working with Cagan’s model, even under rational expectations, because the heart of the model was an ad hoc demand for real balances understood as an inverse function of the public’s anticipated rate of inflation. Neil had convinced me that empirical work really should wait until the foundations of monetary theory had been properly set forth and provided a deep enough theory of valued fiat money. Bob shot back immediately that ‘if theorizing to build deep foundations do not imply a demand function for money that looked much like Cagan’s, then it should be ignored for empirical work’.
I often see people attempt to derive monetary economics from first principles. They ask questions like: “Why should open market operations matter? After all, OMOs merely swap one government liability (base money) for another (T-bills)?” In fact, there is a mountain of empirical evidence that OMOs matter a great deal.
I’ve devoted much of my life to a very close examination of monetary data, during all sorts of historical periods and under a wide variety of different monetary regimes. Thus I have a pretty good sense of which monetary theories are plausible and which are not. Phillip Cagan’s empirical work on money demand during periods of high inflation is one important piece of the puzzle, but it’s merely one of hundreds of pieces of evidence that monetary economists need to be aware of.
I’m a bit dismayed when I see conventional macroeconomists argue that the recent high inflation shows that the MMT model is wrong. It does nothing of the sort. The MMT model allows for the possibility that excessively expansionary fiscal policy might lead to high inflation. Instead, I’m dismayed that it took these recent events to make people realize that MMT was a worthless model. The MMT model has always been wildly inconsistent with hundreds of years of empirical evidence on the effects of monetary policy. It would be like someone saying that the recent invasion of Ukraine convinced them that Putin is an evil dictator. What? The previous 20 years didn’t already convince you of that fact?