Forget the Money Supply, It's the Interest Rate, Stupid

The only sensible view on monetary policy is to be a hawk when the situation calls for it and a dove when the situation calls for that. Anyone who is always a hawk or always a dove is off-base. However, there might well be political benefits to being always a hawk or always a dove. And there can be a pecuniary benefit to being always a hawk: it can be a good way to gin up gold sales by scaring everyone about inflation.

Lately inflation has jumped way up. There is a debate about whether that rise is mostly temporary or has a big permanent component. And this is in a context where the Fed has decided it wants somewhat higher inflation than it had before the pandemic. The Fed’s current monetary policy framework is meant to be ill-defined enough to not tie the Fed’s hands—contrary to the desire of some economists, including me, for more of a monetary policy rule. (See “Next Generation Monetary Policy.”) But I interpret the Fed as shifting from an implicit target of about 1.5% per year up to an implicit target of 2.5% per year: going from not being fully willing to do what it would take to get inflation back up to their nominal 2% per year target to being willing to exceed their 2% per year target somewhat for quite some time.

Whatever one’s take on how tight or how stimulative monetary policy should be, the money supply is no longer very relevant to monetary policy on the margin. The Fed now has what is sometimes called a “floor system.” The Fed tries to keep the monetary base—paper currency plus reserves held in electronic accounts at the Fed—large enough that banks have nothing better to do with the last dollar of reserves than leave it earning interest as excess reserves in a reserve account—or earning interest in the Fed’s overnight facility based on Treasury bill repurchase agreements. Those reserves held as excess reserves or funds earning interest from repurchase agreements are economically idle and don’t stimulate the economy. They are part of the money supply, but one can say they aren’t part of the active money supply. Being inactive matters: for example, idle reserves only add 1-for-1 to other money supply measures. I explain all of this in greater detail, with graphs, in “Supply and Demand for the Monetary Base: How the Fed Currently Determines Interest Rates.”

In their July 20, 2021 op-ed “Too Much Money Portends High Inflation,” John Greenwood and Steve Hanke claim the size of the money supply matters, contrary to what Jerome Powell says:

In his Feb. 23 testimony to Congress, Fed Chairman Jerome Powell said that the growth in the money supply, specifically M2, “doesn’t really have important implications.” The experts, the press and the bond vigilantes were as quick to unlearn monetarism, if they ever had learned it, as Mr. Powell. Reporting about U.S. inflation rarely contains the words “money supply.” …

Wrong. The inflation upticks aren’t temporary and were predictable, driven by an extraordinary explosion in the money supply. Since March 2020, the M2 has been growing at an average annualized rate of 23.9%—the fastest since World War II. There is so much money out there that banks don’t know what to do with it. Via reverse repurchase agreements, banks and money-market funds are lending money to the Fed to the tune of $860 billion. That’s unprecedented.

Notice how they say there is so much money sloshing around that a lot of it sits idle in the overnight facility based on repos that I mentioned above. But they don’t seem to understand that money idled like that doesn’t add to economic stimulus. It doesn’t lower the interest rate because the interest rate on reserves has put a floor under interest rates. (For technical reasons, the floor on interest rates is a tad lower than the interest rate on reserves itself, but not much.)

To see that the money supply is no longer crucial in a floor system where the interest rate has been disconnected from marginal changes in the monetary base, consider the following policy shift:

  • The Fed raises its target rate and the interest rate on reserves by 500 basis points (= 5 percentage points, where a percentage point is a per annum rate).

  • The Fed doubles the monetary base.

I assure you, this would be quite contractionary. The high interest rate on reserves would reduce the active part of the monetary base even though the entire monetary base doubled.

Money doesn’t matter any more on the margin. But that isn’t to say money doesn’t matter. A floor system only works when the total size of the monetary base is kept quite large. So a large monetary base is a precondition for the kind of logic I laid out above. Friedman’s dictum “Inflation is always and everywhere a monetary phenomenon” is not lost, as John Greenwood and Steve Hanke claim; rather, money—and in particular, “Supply and Demand for the Monetary Base,” is what is enabling the Fed to set interest rates.

It is certainly reasonable for John Greenwood and Steve Hanke to be arguing that the Fed should be tightening sooner than it is saying it will (that is well within the range of reasonable disagreement), but they should leave their money supply rhetoric behind and argue forthrightly for the Fed to raise its target rate and to reduce long-term asset purchases in order to put other rates on a higher track.