Should monetary policy be run by a formula?

When I was studying economics 45 years ago in college, one of the big debates was between Milton Friedman, who argued that the Fed should increase the money supply by a constant percentage each year, and a lot of Keynesians who wanted the Fed to have discretion to vary the money supply and interest rates, depending on the condition of the economy (the level of unemployment and the rate of inflation). Although the Fed, under the leadership of chairmen and members from both political parties, has never adopted the “rules-based” approach advocated by Mr. Friedman, the rules vs. discretion debate lives on.

It’s latest incarnation is captured by the current debate over whether to enshrine the “Taylor rule” into law, as two House members, Rep. Bill Huizenga (R., Mich.) and Rep. Scott Garret (R., N.J.), proposed last year. Briefly stated, the Taylor rule (named after Stanford economist and former Treasury Under-Secretary John Taylor) sets out a formula for determining the federal funds rate – the rate banks charge each other for overnight lending, which the Fed can normally control through its purchases and sales of government securities; the formula factors in the inflation rate and the shortfall between the current level of GDP and “potential GDP,” or what GDP would be if the economy were at full employment.

Mr. Taylor has argued that, by his rule, interest rates were too low in the years before the financial crisis, laying the seeds for the crisis, and since 2010, or so, also have been low. Although there is no chance that legislation embodying the Taylor rule will become law during the next two years, the rule is attractive to those who back some version of “audit the Fed” legislation because it sets a seemingly scientific benchmark for judging what the Fed actually does.

On April 28, former Fed Chairman Ben Bernanke posted a rebuttal to Mr. Taylor and others who want his rule to be adopted into law. You can read it for yourself, but I find it devastating. It shows that once one uses an appropriate measure of consumer price inflation and attaches more importance to slow GDP growth, which is consistent with the Fed’s and public’s concern with high unemployment, a “modified Taylor rule” would have mandated the Fed to be even looser in its monetary policy over the past several years than it actually was. This is not what the Taylor rule advocates have in mind.

More broadly, any kind of rule would put the Fed into a straightjacket and unable to respond to a national financial catastrophe of the kind we saw back in 2008. No one should want that either.

Yes, people are imperfect, and so is discretion. Any rule with no discretion would be worse.