David R. Henderson  

Contra Garett Jones: Interest Rates are NOT an Instrument

The Federal Pay Scale: Inflate... Public vs. Private Sector Comp...

Garett Jones argues that the growth rate of the money supply is a target--and targets are often hard to hit, while short-term interest rates are an instrument. He writes:

This is known as the "instruments versus targets" distinction in macroeconomics, and it pays to make the distinction clear. One reason John Taylor's rule for setting short-term interest rates swept the field of macroeconomics is because it told central bankers exactly what to do with the instruments that central bankers actually use and understand. Taylor didn't say "do good things, don't do bad things," he said, "If inflation falls 1% cut the short term rate by 1.5%". Practical advice, not a noble goal.

Garett is right that the growth of the money supply is a target and not an instrument. But he's wrong in claiming that short-term interest rates are an instrument. They are not. They too are just a target.

Consider the three actual instruments the Federal Reserve has for conducting monetary policy:
1. Setting the discount rate, that is, the interest rate at which the Fed lends to member banks.
2. Changing reserve requirements.
3. Buying and selling bonds with federal open market purchases and sales.

#1 is rarely used. #2 is even more rarely used and most monetary scholars and the Fed officials themselves regard it as too blunt an instrument.

That leaves #3. This is the one the Fed uses most. But the Fed doesn't set interest rates. They try to affect interest rates. Moreover, their target interest rate--yes, target--is the Fed funds rate. The Fed funds rate is the rate at which banks lend money to each other overnight. A bank will typically borrow overnight when it would otherwise have too few reserves to meet reserve requirements. How involved is the Fed in the Fed funds market? Ironically, given the name, not involved at all. The Fed neither borrows nor lends in that market. Instead, the Fed tries to affect overall liquidity by buying or selling bonds, and hopes that this affects the Fed funds rate.

In short, the Fed has a target--the Fed funds rate. That rate is not an instrument. The relevant instrument is open-market purchases and sales of bonds.

Comments and Sharing

CATEGORIES: Monetary Policy , Money

COMMENTS (11 to date)
Daniel Kuehn writes:


I made this point adamantly in my Macro Political Economy class, but the professor did not come around to seeing things my way.

To a large extent it's semantics - as are a lot of the "endogenous money supply" questions that these semantics are pertinent to. But I definitely think yours is the more natural way to talk about it.

David R. Henderson writes:

@Daniel Kuehn,
Thanks. Oh and, by the way, congrats on your paper on immigration. Haven't looked at it yet but I look forward to doing so

John Hall writes:

I feel like the basic textbook treatment of the Fed under-emphasizes the importance of repos in the practical management of U.S. monetary policy.

Daniel writes:

The interest rate on reserves is now also an instrument.

John writes:

#3 and interest rates on reserves are the same thing.

No, John the FFR is not the same as IOR. The first is a market rate between banks, the second is the standing offer the Fed makes to sterilize reserves of banks.

Correct the above to IOER (E=excess). which is the new wrinkle since the financial crisis began.

E. Barandiaran writes:

In a model you can assume any variable to be an instrument (an exogenous variable strictly controlled by government) or a target (an endogenous variable of special interest to government). Now, if you are a serious scholar, you try to find evidence to support your choice of instruments and targets, even if you play games like choosing some variables as intermediate targets or whatever.

It's difficult to provide reliable evidence to support that a particular variable is an instrument. In addition to being controlled by government, one must show that an instrument has a reliable and stable relationship with other variables and eventually, through a mechanism of interactions and causation, with a target. This is why Milton Friedman's monetary analysis was wrong and why Charles Goodhart's Law is right.

Note 1: Both the penalty rate on banks' required reserves and the rate paid on excess reserves have always been set by the Fed. But are they instruments? I don't think so because despite huge amounts of money spent on building models the Fed has no clear idea of what effect they may have on banks' portfolio and funding.

Note 2: central banks can influence banks' behavior in ways that are not amenable to analysis. It's called moral suasion, and in times of crisis it amounts to moral coercion.

Greg Jaxon writes:

I guess we simply assume without proof that centrally planning all this monetary policy is Good, Legal, Wise, ...

1. Setting the discount rate, that is, the interest rate at which the Fed lends to member banks.

What a bogus definition! Let's recollect a definition that preceded the existence of "The Fed". Discounting of bills in trade had a long history of stability without the rate ever being "set".

3. Buying and selling bonds with federal open market purchases and sales.

The whole premise of "open market" action is that it ensures FMV pricing for the assets - sort of a circular definition. But what if the policy is structurally guaranteed to Buy more than it Sells? In what sense is a market with risk-free profit built-in assigning a fair price? The new practice of bond speculation can eternally bet that the Fed must continue to Buy (and on average over the long term it has - Boy Has It Ever!)

This central plan is a house of cards.

Justin Rietz writes:

This post on the New York Fed's blog has a good description of how the Fed targets the Fed Funds rate, whether with open market operations or a corridor/floor using the discount rate and IOR:


Jia writes:

Do you think china's PBC has money supply target?

Comments for this entry have been closed
Return to top