A Litmus Test of Fed Independence

This commentary was written by Dr. Daniel Thornton of D.L. Thornton Economics. Thornton spent over three decades at the St. Louis Fed as vice president and economic advisor.

I have been reading Ben Bernanke’s new book, 21^st Century Monetary Policy. In Chapter 15 Bernanke writes about Fed Independence and what kinds of things will strengthen the Fed’s ability to maintain it.

I couldn’t help myself. I immediately thought of a litmus test of the Fed’s independence.

Here is my litmus test: At its November meeting, it’s too late for the September meeting, the FOMC should announce that it will reduce its holding of Treasury debt over the next year to no more than 10% of the marketable Treasury debt, which today would be about $2.4 trillion.

The average was 12.8% from 1990 through 2008. This was a period when banks were required to hold reserves. The amount should be lower now that banks are no longer required to hold reserves.

What about the Fed’s paying interest on reserves policy? Many economists believe that this policy prevented banks from increasing the money supply by holding the reserves rather than making loans with them. I don’t worry about this since I have already demonstrated that paying interest on reserves cannot prevent banks from making loans from reserves, here and here.

Well then, what about the effectiveness of QE in reducing long-term rates? I don’t consider this either because there is no compelling evidence that QE significantly reduced long-term rates.

Ben Bernanke often cites the results from event studies, which measure the immediate response of long-term interest rates to QE announcements, as evidence that QE reduced long-term rates. I did a careful analysis of event studies here, which demonstrated that event studies have no evidentiary value. Jim Hamilton’s analysis of Fed announcement effects here came to the same conclusion.

The credit market is international, enormous and very efficient. Hence, basic economic theory suggests that, at best, the Fed’s QE actions could produce only a small effect on the entire structure of interest rates and ever smaller effect on long-term rates.

Consistent with the theory, in a Cato Policy Analysis here, I showed there was no persistent effect of the Fed’s QE announcements on the 10-year Treasury or the Aaa and Baa corporate bond rates. Jim Hamilton here looked at similar data and concluded “at a minimum we are forced to conclude that Fed purchases were only one of many factors influencing bond yields during these episodes, and certainly not the most important factor.” Greenlaw, Hamilton, Harris and West here also found that QE was ineffective in reducing long-term interest rates.

I don’t expect the FOMC will ever undertake my litmus test because most, if not all, of the FOMC participants believe that paying interest on reserves has prevented an explosion of the money supply and that QE significantly reduced long-term rates in spite of the fact that the money supply exploded, as well as compelling theoretical and empirical evidence to the contrary, as Bernanke does in 21^st Century Monetary Policy.

In any event, if the FOMC made such an announcement, I feel certain that Chairman Powell would be called to the White House for a chat and asked to appear before Congress for a closed-door conference.

While transcripts of FOMC meetings show that Janet Yellen was very skeptical of the effectiveness of Fed’s purchases of Treasuries in reducing the 10-year Treasury rate, as the Secretary of the Treasury she would be quite concerned if the Fed started selling Treasuries equal to about 15% of the public debt.

The only thing we are certain QE accomplished is that the Fed financed about 24% of the $21.9 trillion increase in the public debt between October 2008 and December 2021. This is not a record to be proud of. Nor is it monetary policy for the 21^st Century.

It’s the road to perdition.