New York Fed President Bill Dudley and I debated The Fed at a crossroads: Where to go next? at Brookings yesterday, moderated by David Wessel.  Bill argued against a more rules-based road ahead for the Fed. I disagreed, but this kind of discussion from the Fed is helpful given recent legislation requiring it to report its rule or strategy.

Surprisingly, most of Bill’s points were directed at the Taylor rule, and here we collided.

First, Bill says that the Taylor rule is not forward-looking because it includes current inflation and output rather than forecasts of those variables.  But the Taylor rule was designed to deal explicitly with forecasts, and it is in fact forward-looking in important ways. Note that when a central bank indicates that it will predictably follow a rule in which the interest rate reacts to the current inflation rate, it automatically says that next period’s interest rate will react to next period’s inflation rate. That’s forward-looking. Moreover, the current level of inflation and output are factors in the forecast of inflation, and the coefficients in the Taylor rule take that into account.

Now, one could replace current inflation with a forecast of inflation in the Taylor rule—as Bill seems to suggest, but the coefficients would most likely have to be different.  And that approach raises the question of whose forecast to use and how to evaluate the rule. Remember that forecasts—including the Fed’s forecasts—are not that good. Also rules with forecasts of inflation and output on the right hand side tend not to be robust.

Second, Bill argues that the equilibrium interest rate in the Taylor rule should not be a constant.  The original rule took the real equilibrium rate to be 2%, meaning that with the 2% target inflation rate in the rule, the equilibrium nominal rate would be 4%.  But there is no reason why a moving rate could not be incorporated into the rule, and many have suggested doing so. In my view debates about the implications of changes in the equilibrium interest rate are more productive if they are conducted within the framework of a policy rule rather than in the abstract. According to the Fed’s dots, the equilibrium rate is slightly lower than FOMC members earlier believed—3.5% rather than 4.0%. That’s no reason to steer away from rules-based policy.

The adjustment that I suggested in 2008 to add a smoothed spread between Libor and the overnight index swap to the rule is an example of how such developments can be incorporated. Bill mentions this adjustment as reason why the Taylor rule should not be used mechanically.  I agree that the Taylor rule should not be used mechanically, and I emphasized that in the original paper.  The adjustment I suggested in 2008 was meant to deal in a systematic way with a particular problem in the money market at that time. The models that were used to find the Taylor rule in the first place implied such an adjustment.

Bill also refers to the zero lower bound as a reason to deviate from the Taylor rule. Well, the zero bound was not a reason for the deviation in 2003-2005 and it is not a reason now, with the zero bound no longer binding. The zero-bound was taken into account in policy rule design research long ago.  My view has been that interest rate rules are best thought of as part of a more encompassing rule in which the instrument becomes money growth in deflationary or hyper inflationary situations.

Third, Bill says the Taylor rule is too simple because it omits certain variables. Well, the Taylor rule is simple, because we made it simple.  At the time people were coming up with all sorts of complex rules that included many of the types of variables that Bill thinks should be there, including asset prices. These rules were too complex to be workable in practice. So we boiled them down. It was amazing that we could. We found that removing certain variables gave just as good a performance in many models and was more robust over all. It certainly was something more practical for policy makers to work with.

In any case Bill seems to want to complicate things again by adding in many more variables.  But saying that you need to take account of everything under the sun is not a good way to make policy. A “careful elucidation,” as he describes it, of key factors sounds better, but not if you don’t have a strategy to react to those factors. That strategy is what is missing from Bill’s favored approach. That is why I disagree with it, and why I favor a rules-based policy for the Fed with the Fed choosing its strategy, which of course does not have to be the Taylor rule.