Wednesday, 20 September 2017


On interest rate targeting

July 6, 2015
in Blog

Previously I discussed inflation targeting as a popular rule for governing central bank behavior. In this post I will discuss interest rate targeting, another popular recommendation that has its own costs and benefits.

The most prominent interest rate rule is the Taylor rule, devised by John Taylor of Stanford University. Taylor originally proposed the rule as a way of describing past central bank behavior, but it since has been adopted as a prescription for monetary policy. In brief, the Taylor rule says that the policy (short-run) interest rate should be adjusted based on where inflation is relative to some desired level, and where real output is relative to some desired level. All else being equal, too high inflation means the policy rate should increase; too low real output means the policy rate should decrease; and vice versa.

This approach has been modified by introducing expectations, and emphasizing forward-looking behavior. The most popular ‘New Keynesian’ approach to monetary policy, whose most prominent advocate is Michael Woodford of Columbia University, holds that central bankers should provide forward guidance as to the expected future path of interest rates. This should provide market actors some certainty, and help anchor their expectations with respect to the stance of monetary policy, so they can better coordinate their trading behavior to maximize the gains from exchange.

Perhaps the greatest benefit of interest rate targeting is it gives central bankers very precise ‘if x, then y’-style feedback for how to adjust their behavior in light of prevailing macroeconomic conditions. Inflation targeting and NGDP targeting are outcome rules which do very little to tell central bankers how to achieve these outcomes. Interest rate targeting is usually more precise, especially if it follows a modified Taylor rule parameterized to the current state of the macroeconomy. Of course, this precision can be a double-edged sword. If prevailing macroeconomic conditions change—if the parameters that underlie the Taylor rule shift without central bankers being aware of it—conducting monetary policy according to the now incorrectly specified rule can actually cause the economy harm.

In addition, it must be remembered that interest rates are a price of a specific good, namely time. Interest rates coordinate savings and investment decisions over various time horizons. If central bankers intervene such that the current market interest rate deviates from the hypothetical equilibrium interest rate—the rate that coordinates savers’ supply of capital with investors’ demand for capital—the result can be malinvestments and other miscoordination of economic activity. As an example of this, John Taylor has argued that the Federal Reserve kept interest rates ‘too low for too long’ in the early 2000’s, creating artificially cheap capital and helping to fuel the subprime mortgage bubble.

In my opinion, the heavy focus on interest rates for the conduct of monetary policy is misguided. As I argued previously, the stance between monetary policy and interest rates is ambiguous at best, and it is probably a bad idea for central bankers to be meddling with one of the economy’s most important relative prices. Inflation targeting, or better yet NGDP targeting, may be less specific in telling central bankers how to behave, but their other benefits make them worth it. So long as a central bank credibly adopts an inflation or NGDP target, it actually has to do very little work; because market actors find the rule credible, they will only engage in trades predicated on the rule’s functioning. This is what Lars Christensen has dubbed the ‘Chuck Norris effect’—credible central bankers have to do very little, precisely because they are credible.

In closing, I should mention that there is some debate within the macroeconomics and monetary economics literature as to whether interest rates are instruments or targets. When the Federal Reserve conducts monetary policy in an attempt to influence the fed funds rate, which is the overnight loan rate banks charge each other, are they using interest rates as the mechanism by which monetary policy is implemented, or as a signal as to whether a given goal of monetary policy has been achieved? While subtly different, the distinction is important. The view that interest rates are instruments suggests central banks are actively steering the economy, whereas the view that interest rates are targets suggests central banks instead set the background conditions by which the economy steers itself.