Restoring the reputation of monetarism

Today’s post is brought to you by my sponsor, Mechanize. They’re hiring junior software engineers at $300K/year base salary. Apply now!

* * *

In one sense, monetarism is always true. By controlling the monetary base we can control the path of NGDP. That is even true in a world where the central bank pays interest on bank reserves (IOR). Unfortunately, with the advent of IOR, monetarism no longer appears to be true, at least to most informed observers. In this post I’ll explain how we can move from a world where monetarism is true but counterintuitive, to a world where monetarism is obviously true.

Back in the 20th century, you would occasionally hear economists refer to the concept of “high-powered money”. At the time, this term was attached to the monetary base, which is composed of currency in circulation and bank reserves. Both of those assets earned a zero nominal rate of return and hence were costly to hold in a world where risk-free interest rates of safe Treasury securities were well above zero.

During this period, banks didn’t want to hold large reserves. By the early 2000s the monetary base was about 98% currency and roughly 2% commercial bank deposits at the Fed. This meant that an injection of new base money through open market operations was a sort of “hot potato”, which the public tried to get rid of each time the new injection exceeded current base demand. Attempts to get rid of excess cash balances led to more spending, pushing up nominal GDP. Eventually, aggregate national income rose high enough so that people were willing to voluntarily hold this excess cash, and equilibrium was restored. Base money was “high-powered” because it had a big impact on aggregates like the price level and NGDP.

Unfortunately, this simple monetarist system ended in 2008 when the Fed began paying interest on bank reserves. One component of the monetary base became a close substitute for Treasury bills and no longer represented high-powered money. Only the currency part of the base is now high-powered. The Fed lost tight control of the quantity of high-powered money (currency), as although a $1 million open market purchase continued to increase the monetary base by $1 million, it no longer had a predictable impact on the currency stock. Banks might choose to simply sit on the newly injected base money with an increasingly bloated reserve deposit account at the Fed.

My preferred solution to this problem is to return to the pre-2008 system of scarce reserves and no IOR. Unfortunately, a recent post by David Beckworth suggests, that this option is a political non-starter:

Don [Kohn] and I discussed this counterfactual and agree it is one reason why we can never go back to the pre-2008 operating system. Even if the Fed did return to a scarce reserve operating system, it would do so with IOR. There have been calls to change the Federal Reserve Act and eliminate IOR, but that would effectively be imposing a big, distortionary tax on banks.I believe this is why Fed Chair Kevin Warsh said in his House testimony in July that “I’m not of the mistaken view we can go back to where we were when I arrived at the Fed in 2006…”. The other reason we cannot go back is the post-GFC liquidity regulations which have increased the structural demand for reserves. There are ways to tweak these regulations to reduce the demand for reserves they create, but there would still probably be marginally more demand even after making these changes. Throw in the IOR and there is no going back to the pre-2008 system. So it is time to put to bed the idea that we can go back.

To be clear, I’d still favor going back. The tax on reserves was not very important in a world where reserves are a small fraction of the monetary base. And liquidity regulations can be changed—why not go back to the pre-2008 regulatory system for reserves (while continuing to have stricter capital requirements to reduce bank default risk)? But it’s a moot point, as David is likely correct that my proposal is not within the Overton Window.