The perceived odds they go in September are quite low again. But this is not about that. I don’t think there is a percentage in being dovish here. The markets are not spotting you much. Instead, this brief post is a reminder about method.
It is often taken as a given that when the state of the labor market gets back to normal, interest rates should be getting back to normal too, unless inflation is is way off target, which it is not. (That is a separate discussion, and in this note I mostly abstract from inflation to keep it simple.)
That way of thinking embeds the logic of the Taylor Rule. People may not put it in those terms anymore, but the influence here of Taylor is unwitting.
The idea that the funds rate should be normal when the labor market is normal is not written in the stars. Nor does it come out of a private-actor optimization framework, so far as I can tell.
Rather, the Taylor Rule (which I will use as a stand in for all policy rules here) posits that running policy as if the unemployment rate were a reliable proxy of the appropriate interest rate will generate satisfactory macroeconomic outcomes.
The reason is that the Taylor Rule incorporates stabilizing feedback loops that normally allow mistakes generated by applying the rule to be corrected before they do too much damage to the economy.
For example, if the Taylor Rule were wrongly to prescribe a tightening, the Fed followed it, and the economy began to cycle down, then a rising unemployment rate and/or falling inflation rate would normally signal the Fed to reverse course before too much damage to the economy was done.
As I have mentioned before, the Taylor Rule is not precisely “right”, even according to its advocates. Rather, it is normally robust to being somewhat wrong, by virtue of those stabilizing feedback loops embedded in it.
But these are not normal times and the stabilizing loops are largely disabled. The labor market may be almost back to normal but the Fed is still severely constrained by the zero bound. Hence all the discussion of those stretchy remedies for the zero bound, like h money, more QE, raising the inflation target, etc.
When at the zero bound, the stabilizing feedback loops in the Taylor rule are disabled for half the prospective scenarios, the downside ones. Accordingly, it would be reckless for the Fed to follow a Taylor Rule. So it doesn’t.
Rather, in order to view a funds rate increase as a good idea now, the Fed must forecast with some confidence that the economy would not begin to cycle down in response to such a tightening. It needs to be confident because it realizes that reversing course would be very difficult, simply because of the zero bound.
This abandonment of the Taylor Rule and other policy rules like it is difficult to explain, particularly if the Fed is trying to play down the risks associated with the zero bound and avoid the impression that it is relying now heavily — as it must — on discretion.
It does not help that most of the Fed watching community has not internalized this simple point and is still clutching its pearls over the Fed being “behind the curve,” a curve very poorly traced out by the Taylor Rule.
Keynes was right when he said that what people take to be practical common sense is often just some defunct academic idea. For now, the Taylor Rule is defunct, although I could imagine it coming back were the Fed to find itself again safely away from the zero bound.