Should the Fed have to play by a rule?

What if the U.S. Federal Reserve Board had to implement monetary policy according to a specific rule that would require specific policy actions depending on the circumstances?

That's the intent of a bill Republicans in the House of Representatives recently proposed. The Federal Reserve Accountability and Transparency Act would require the Fed to formulate and make public a monetary policy rule and provide added transparency about the reasons for the Fed's interest rate recommendations.

Economists have long debated whether specific rules are better than giving central bankers the discretion to set monetary policy as they see fit. Here are the arguments for and against policy rules, and a compromise position that many economists advocate.

The case for rules

One reason to adopt policy rules is to avoid what's known as the "time-inconsistency" problem. Here's an example of that problem. Consider a person who follows a rule saying he'll run five miles for four days of every week. This rule is in the individual's long-run health interest.

Suppose that in a given week, he decides to skip the first thre days. According to the rule, he must run on each of the week's remaining days. But when the first day comes, he decides not to run and to stay home and watch TV instead. In this case, the action plan is "time-inconsistent." Plans made in the past aren't necessarily followed when the time comes to execute them. More formally, time inconsistency occurs when deviations occur in optimal long-run plans when making short-run decisions, and those deviations can undermine the ability to achieve long-run goals.

But if the rule has a commitment mechanism of some sort, some way of forcing the individual to follow the rule, time-inconsistency can be avoided. For instance, he could tell his friends he has committed to the rule so that if he doesn't follow it, he'll suffer a loss of reputation. Or he could give a friend $100, and if he doesn't follow the rule, the money is given to charity (and the pot replenished). The more strict the penalty, the more likely it is he'll follow the rule.

When it comes to monetary policy, a rule might say the Fed will stabilize the inflation rate at its target value of 2 percent. But in the short-run, a more expansionary policy might let GDP increase and unemployment go down. If the Fed gives in to this temptation and raises the money growth rate to stimulate the economy, it could also cause higher inflation in the future. So, the long-run outcome is better if the Fed can resist.

How does this relate to the debate about rules versus discretion? If monetary policy is under the control of a politician who has complete discretion, the tendency will be to use monetary policy to juice the economy, enhance growth and lower unemployment to maximize the chance of reelection. But in the long run -- after the election -- that can result in higher-than-desired inflation. The short-run goal of getting reelected undermines the economy's long-run health.

But if the politician is forced to follow the rule, say, he must leave office if he intentionally deviates, this temptation can be avoided. So, rules along with commitment can avoid the temptations that exist under discretionary policy. Another way to do this is to put policy in the hands of an independent body whose members don't care about reelection. But even then, there could be some temptation to deviate from the rule if discretion is allowed.

The other argument for rules is that they provide stability. For example, as reported in The Wall Street Journal, House Financial Services Committee Chair Jeb Hensarling (R-Texas) claims "The overwhelming weight of evidence is that monetary policy is at its best in maintaining stable prices and maximum employment when it follows a clear, predictable monetary policy rule." The basis for this claim is that deviations from expected policy can cause variation in output and prices from their desired values, that is, it can cause instability in these variables. But if a rule is well-known and there's commitment to ensure it's followed, the likelihood of destabilizing policy surprises is much lower.

The case against rules

Following a rigid rule isn't always best. Going back to the jogging example above, suppose the person didn't run because of an ice storm, and running could result in a slip and severe injury, a broken arm perhaps, something certainly not in his long-run health interest. Or perhaps the individual woke up with the flu, and running -- if possible at all -- would risk pneumonia.

Thus, there has to be a way to allow exceptions from the rule when particular contingencies arise. Many of those can be written as part of the rule itself. It could be amended to say if running on a particular day is a risk to health, it can be skipped. But once exceptions are allowed, the door is open for many more -- and the rule can be undermined.

In addition, not all contingencies can be foreseen, and if something unforeseen comes up that could risk the runner's health, but it's not part of the rule, must the rule be followed?

What's needed is a mechanism for allowing some discretion to deviate from the rule. Otherwise, when it comes to the economy, policymakers can end up being forced to follow a prescription that isn't in the economy's best interest.

Many people argue that would have precisely been the case if the Taylor rule existed during the Great Recession. Would it have allowed the Fed to construct the special liquidity facilities that were so important in stopping even worse problems than we had? Could the Fed have engaged in quantitative easing? (Of course, many proponents of the rule want one specifically because they object to the Fed's QE program).

Another problem with a rule is that many important sources of information aren't quantifiable. It's impossible to quantify every key variable in the economy in a way that will allow it be incorporated into a rule.

For example, each of the 12 Federal Reserve district banks spends considerable effort trying to assess business conditions within the district, and a big part of that is talking to business leaders. But how can business sentiment -- what Federal officials are told informally by a wide range of businesses -- be quantified and used in a mathematical rule?

Plus, monetary policy rules are derived from theoretical models of the macroeconomy, but the true model is unknown. A rule that works best in one model may not be the best rule in another. When we discover a model is flawed -- as macroeconomists found out during the Great Recession -- we may also discover the rule derived from it has serious flaws when some types of problems arise. If we're forced to follow the old rule, that could be bad for the economy. So, how and under what circumstances should the rule be allowed to change? How much discretion should be granted?

And even if the correct model is known, its parameters may change over time, and as they change, the coefficients in the optimal policy rule can also change. Structural change can be difficult to detect, and disagreements will arise, so how much discretion should policymakers have to vary a given rule's coefficients? If they have too much discretion, that can result in picking almost any policy rate they want by varying the coefficients of the rule. So, what good is the rule under those circumstances?

The case for a compromise

Perhaps what's needed is "constrained discretion." Former Federal Reserve Governor Frederic Mishkin argues in his book on monetary theory and policy (on which much of the above is based) that the rules-discretion distinction is too stark. He says the best course is a combination of both.

Thus, under constrained discretion, a term invented in a paper by Mishkin and former Fed Chair Ben Bernake, the Fed might be instructed to do its best to hit particular targets for inflation and output, but the precise means of attaining the targets would be up to the policymakers.

Although the general objectives and tactics would be specified, which would constrain behavior, the Fed would still have discretion over the particular actions it takes. It combines the advantages of both rules and discretion, and it's an approach many economists favor. It's also the way the Fed operates currently: It has a mandate to pursue both price stability and full employment, but the precise means of achieving these goals is left up to the Fed.

Many economists oppose forcing the Fed to adopt a rule. For example, the current proposal gets no support at all in this survey of top economists conducted by the University of Chicago Booth School. They believe constrained discretion is a better policy.

Correction: An earlier version of this article incorrectly stated that the Fed would be forced to follow a Taylor rule.