The debate over the Fed’s dual mandate was reignited recently after Judy Shelton, an economic advisor to President Trump who is being considered for the Fed’s Board of Governors, questioned the desirability of focusing on full employment. “I don’t know that that is really the Fed’s job,” Shelton said. (Shelton was previously the director of the Sound Money Project, prior to it moving to AIER.)
Since the 1946 Employment Act, the Fed has been tasked with the dual mandate of maintaining price stability and full employment. The employment mandate was strengthened in the Humphrey-Hawkins Full Employment Act of 1978.
Shelton is far from alone in criticizing the Fed’s dual mandate. Many economists, including John Taylor and Marc Labonte, have argued that the Fed’s performance would improve with a single mandate on price stability.
Others think Shelton is wrong to question the Fed’s focus on full employment. Claremont McKenna historian David Stein took to Twitter to condemn her remarks. He believes the full-employment mandate “has made the Fed more accountable to working people.”
At first blush, a full-employment mandate might seem harmless — perhaps even wise. We want the Fed to help avoid the sort of large-scale recessions that cause unemployment to rise. It is natural to think a full-employment mandate would help achieve that end. But Shelton is right to question this view.
For starters, it is far from obvious what “full employment” actually means. Economists usually make a distinction between the unemployment rate prevailing at a particular point in time and the rate that would prevail if no one were fooled into over- or under-producing. The latter, known as the natural rate of unemployment, is the best one can hope for given the underlying constraints of the economy and people’s preferences. Some unemployment is desirable, as those between jobs search for a better opportunity or those who are out of work for a longer period, while retraining or relocating in response to structural shocks, continue to look for work.
If the unemployment rate is higher than the natural rate, current output and consumption are too low. If the unemployment rate is lower than the natural rate, current output and consumption are too high. In other words, “full employment” is a good goal only insofar as it is taken to mean the level of employment consistent with the natural rate.
There are three ways the Fed can try to ensure unemployment is in line with the natural rate. It can target a real variable, like output growth or unemployment, at its natural rate and allow nominal variables to adjust accordingly. It can anchor expectations by targeting a nominal variable, like nominal income or inflation, and allow the market to push real variables like output and unemployment toward their natural rates. Or it can target both nominal and real variables, putting some weight on each. The Fed's dual mandate corresponds to the last method.
Those economists who oppose targeting real variables, in whole or in part, usually accept the theoretical basis for targeting a real variable. It is the practical difficulties of targeting a real variable that cause them concern. We do not observe the natural rates of unemployment or output growth. And since technology grows in fits and starts, there is no reason to think these natural rates are stable over time.
Conducting monetary policy to hit a moving target that no one actually observes is a recipe for disaster. If the Fed underestimates the economy’s natural rate of unemployment, as it did throughout the 1960s and ’70s, monetary policy will be too loose. Overproduction and higher prices in the short run and a reduction in potential output in the medium run are likely to result. If it overestimates the natural rate, on the other hand, monetary policy will be too tight and we will produce fewer goods and services than we want.
Targeting nominal variables, in contrast, is much more straightforward. The difference, in terms of production and employment, between a 4 percent and a 5 percent nominal-income target (or a 1 percent and a 2 percent inflation target) is miniscule. What matters most is that the nominal-spending growth and inflation that ultimately result are in line with the nominal-spending growth and inflation expected when production and employment decisions were made. The Fed can anchor expectations along any nominal path by announcing a target. Then, so long as it delivers on its promise to hit that target, no one will be fooled into over- or underproducing. Hence, by targeting a nominal variable, the Fed nonetheless ensures that real variables like unemployment and output growth remain close to their natural rates.
The idea that the Fed can achieve full employment by ignoring real variables altogether and focusing exclusively on a nominal target will no doubt seem strange to many. It is common to hear commentators on financial markets debating whether the Fed should put more weight on price stability or full employment, as if there were a trade-off between the two. But economists who argue that the dual mandate should be scrapped and replaced with a nominal-spending or inflation target do not, by and large, believe that price stability is more important than full employment. Rather, they believe focusing on price stability is a more effective means of achieving full employment. It is a question of means, not ends.
In theory, there is a divine coincidence, where nominal and real (natural rate) targets achieve the same ends. In practice, the difficulties of targeting a real variable suggest a nominal target is more likely to be effective.
More troubling than the debate over the best means to ensure that unemployment corresponds to the natural rate, however, is the disagreement over what the ultimate goal of Fed policy should be. When Stein refers to the importance of “maintaining full employment,” he does not mean keeping unemployment at its natural rate. Instead, he points to the Humphrey-Hawkins Act, which defines full employment as “a situation under which there are useful and rewarding employment opportunities for all adult Americans willing and able to work.” He makes no distinction between benign types of unemployment consistent with the natural rate and the undesirable unemployment most economists agree should be avoided. To Stein, all unemployment is bad unemployment.
In practice, a Fed committed to bringing about full employment as defined by Stein and others on the far left and populist right would be disastrous. To achieve that end, the Fed would have to continuously surprise economic decision makers with more and more money. In the short run, such a policy would result in overproduction and higher prices. In the medium run, our ability to produce would be eroded, as more and more costs would be incurred to deal with higher and higher rates of inflation. It’s a bad policy all around.
Shelton is right to criticize the full-employment aspect of the Fed’s dual mandate. The Fed should do its best to ensure that unemployment is neither too high nor too low. But it does not follow that focusing on full employment is the best way to achieve that end. Indeed, the difficulties of targeting real variables suggest we would be better served by a strict nominal target. At the very least, one should recognize the real constraints of the economy and not seek to push unemployment below — or output above — what is justified given those constraints.