John Taylor recently had a post on his blog with the accompanying graph showing the actual Fed Funds rate target of the Fed since 2005 and the Fed Funds rate implied by two versions of the Taylor rule, one that he specifically proposed and another used in a study by Janet Yellen that Taylor, in a 1999 paper, had mentioned as a possible alternative version of his rule. Taylor has subsequently tried to put some distance between himself and the alternative version, the alternative version implying a far lower optimal interest-rate target than the version that he now professes to prefer. But while not explicitly endorsing it when first mentioning it as an alternative, neither did Taylor express any reservations about the alternative, providing no hint that he considered it to be inconsistent with the spirit of his rule or to be obviously inferior to his own previous version, for which he now insists he has a preference.
What I find especially noteworthy, aside from the remarkable fact that, as Scott Sumner noted, Taylor’s preferred rule would have called for a rate increase in early 2008, when the economy was already in recession, and on the verge of one of the sharpest one-quarter declines in real GDP on record, in the third quarter of 2008 even before the Lehman panic of September-October, is that both versions of the Taylor rule implied a target interest rate substantially higher than the Fed Funds rate actually in effect for most of 2008. So Taylor is implicitly endorsing a far tighter monetary policy in 2008, after the economy had already entered a recession and started a rapid contraction, than the disastrously tight policy to which the economy was then being subjected by the FOMC.
Now, in fairness to Taylor, he could argue that the difficulties all stemmed from the prolonged period of very low interest rates following the 2001 recession. But that simply underscores the inherent unworkability of a mechanical rule of the type that Taylor is so enamored by. Conditions are rarely ideal, so you can never be sure that the interest rate implied by the Taylor rule (of whichever version) is preferable to the rate chosen at the discretion of the monetary authority. In retrospect, some of the time the FOMC seems to have done better than the Taylor rules, and some of the time one or both of the Taylor rules seem to have done better than the FOMC. Not exactly an overwhelmingly good performance. So why should anyone assume that adopting the Taylor rule would be an improvement, all things considered, over the exercise of discretion?
Taylor wants to argue that the exercise of discretion is bad in and of itself. But which is The Taylor rule? Taylor likes one version of the rule, but he can’t provide any argument that the Taylor rule that he prefers is better than the one that he now says that he doesn’t prefer, though no such preference was expressed when he first mentioned the alternative version. And even now, though he claims to like one version better than the other, he can only conclude his post by saying that more research on the relative merits of the rules is necessary. In other words, adopting the Taylor rule is not sufficient to eliminate policy uncertainty, as the gap in the diagram between the rates implied by the two rules clearly indicates.
The upshot of all this is just that for Taylor to suggest that adopting his rule would somehow reduce policy uncertainty when there is clearly no way to specify the parameters necessary to generate a predictable value for the interest rate target implied by the rule is simply disingenuous. Moreover, to suggest that there is any evidence that following the Taylor rule (whatever such a vague and imprecise concept can possibly mean) would have led to better outcomes than the not very impressive performance of the FOMC is just laughable.
PS This will be my last post until next week after the Jewish New Year. My best wishes go out to all for a happy, healthy, and peaceful New Year.