The Federal Reserve wants to get monetary policy back to normal without scaring or surprising the financial markets. Now, try defining "normal," and you can see it's going to be difficult.

A vital instrument of abnormal monetary policy has been the promise to keep interest rates at (roughly) zero for an extended period. Once rates have been raised off that floor, this kind of time-based commitment no longer works.


Ordinarily, incoming data would (or should) dictate how quickly interest rates go up. Quite possibly, if inflation fails to rise to its 2 percent target, the data will call for more rate cuts. The point is, nobody knows what the data will say. That's why normal policy is inherently more confusing than policy at the zero lower bound -- hence more capable of springing surprises.

In speeches during the past week, both Fed Vice Chairman Stanley Fischer and Bank of England Chief Economist Andrew Haldane delved into this.

Fischer said: "With respect to forward guidance: its role has been and continues to be important in the long period in which eventual liftoff has been the key interest rate decision confronting the FOMC and the focus of market expectations. However, as monetary policy is normalized, interest rates will sometimes have to be increased, and sometimes decreased."

Haldane, discussing Britain, where headline inflation fell to zero in February, noted that professional forecasters unanimously expect the next change in interest rates to be upward.

"I do not currently see an immediate case for a policy change in either direction. If one were required, given the asymmetry of inflation risks, I think the chances of a rate rise or cut are broadly evenly balanced. In other words, my view would be that policy may need to move off either foot in the immediate period ahead, depending on which way risks break," Haldane said.

The challenge, it seems, is to persuade financial markets that policy really will be data-driven. In an ideal world, the central bank would react predictably to data: Doubts about what it might do would arise solely from unavoidable uncertainty about what the data will be.

The world isn't ideal, but the point is still worth bearing in mind. This week the Financial Times quoted William O'Donnell, a strategist at RBS Securities, expressing a widely held view: "Data dependency and psychoanalysis of the Fed will continue to hold the reins of U.S. rates." As long as forecasters think the Fed needs psychoanalyzing, there's a problem with the way it's communicating.

A "Taylor rule" for monetary policy could help in presenting Fed decisions, even if it wasn't used to dictate them. Taylor-type rules, as suggested by Stanford University's John Taylor, explicitly link interest rates to inflation and the amount of slack in the economy. Fischer and Haldane both touched on the idea.

Fischer said that a Taylor-type rule would ignore many factors that ought to influence interest rates, and that the Fed's policy makers have to use their judgment in reacting to special circumstances -- but he also said that it "can provide the starting-point" for decisions. If the Fed leaned more openly on a data-based rule in explaining itself, it would lighten the burden on the markets' stressed psychoanalysts.

Haldane described the results of so-called optimal-control simulations, rather than a Taylor-type rule. Fed Chair Janet Yellen has often expressed interest in this approach. Optimal control is a rule-based method too, but much more complex than a Taylor rule. It's forward-looking and involves minimizing the economic losses predicted by a specific economic model. Everything therefore depends on whether the model in question is any good.

Nonetheless, the two approaches have important things in common: They put structure on one's thinking and move data center-stage. As Haldane says:

"Of course, there are good reasons why monetary policy is not set by algorithm. These simulations ignore a number of important practical uncertainties. For example, they assume that this policy path is fully credible and effective in stimulating demand and inflation expectations. In practice, the effects of a policy easing cannot be known with certainty and policy credibility cannot be guaranteed. Moreover, there are upside as well as downside risks to the inflation outlook. Nonetheless, these experiments are in my view the right baseline when assessing the appropriate policy stance today."

Starting-point, baseline, whatever. Policy rules shouldn't be followed slavishly. Nonetheless, taking them more seriously - - and being seen to do so -- would help to make markets more comfortable with data-driven policy.

Clive Crook is a Bloomberg View columnist and a member of the Bloomberg View editorial board.