The lack of a relationship between the size of the coefficients in a Taylor rule for monetary policy conduct and the underlying preferences for stabilization of macroeconomic goals is well known. I often have it as a check subject in my exams in monetary economics. When I present the result to students first time—it is fleshed out in Lars Svensson’s “Inflation Forecast Targeting: Implementing and Monitoring Inflation Targets” (European Economic Review 41, 1997, 1111-1141) for a simple backward-looking IS/AS model—I often state that many tend to overlook this, and that it is a common misconception that, e.g., a relatively high coefficient on the output gap in the rule indicates a relatively high preference for output gap stabilization.
I sometimes fear that I thereby make the classic mistake of putting up a straw man in lack of better motivation for why this is an important result to emphasize. But I just became aware of a new example of the peculiar endurance of this misconception. In John Cochrane’s positive and very appetizing review of John Taylor’s new book, “First Principles” (W. W. Norton & Company, Inc., 2012), Cochrane writes:
The Taylor rule actually stands quite a bit to the left of the “inflation targeting” tradition that says central banks should only respond to inflation, ditching the whole GDP response — because, in John’s words (p. 127)
‘Some Federal Reserve officials worry that a focus on the goal of price stability would lead to more unemployment. But history shows just the opposite.’
John answers that the “dual response” really is a “single mandate.” It is a a worthy effort, but one I find strained. The reason for the GDP response is, explicitly in the models, to accomplish a tradeoff between inflation and output volatility.
There are two intertwined mistakes here. As Svensson’s model clearly shows, even a strict inflation-targeting central bank, i.e., one that only cares about inflation stability, would respond optimally to the output gap. So following Taylor’s rule would be a good idea (given that the “magic numbers” 1.5 and 0.5 somehow were appropriate for all countries over the globe). Why is that? Well, output may be a good predictor for future inflation. So it serves as an intermediate target worth responding to—even for a completely right-wing “inflation nutter”. It is not a goal variable per se.
Therefore, it is also false to claim that the reason for the GDP response is “explicitly in the models, to accomplish a tradeoff between inflation and output volatility”. To emphasize this point with a different model example, note that in the simple New Keynesian model, a Taylor rule with only a response to inflation will actually be one that may secure the optimal tradeoff between inflation and output volatility. This has been known at least since Clarida, Galí and Gertler “The Science of Monetary Policy: A New Keynesian Perspective” (Journal of Economic Literature 1999, 1662-1007).
All of these matters are simple facts which are completely orthogonal to your own policy preferences, to whether you favor the Taylor-rule approach to monetary policymaking or not, and so on. So one more time for the world: There is no simple relationship (if any) between Taylor-rule coefficients (their size or existence) and policy preferences.
But at least I apparently don’t put up straw men during teaching on this point. Hopefully the point will sink in over time, however.
Apologies up front in advance if I have misinterpreted Cochrane’s post in this dimension.
