I must say I am a little puzzled by the fascination with NAIRU. I thought work done by Stock and Watson and others suggested that if there is a NAIRU, it lies somewhere between 4 to 7 percent or some range like that, which is quite wide. If that is right, it is not a terribly useful concept for policy. So, I have been reluctant for some time to go down that path very aggressively.

A second example of a seemingly stable indicator that lost some of its reliability is the unemployment rate at which inflation remains stable, also known as the natural rate of unemployment. According to natural-rate theories, inflation will rise when the actual unemployment rate is below the natural rate and inflation will fall when the unemployment rate is greater than the natural rate.

The key to using the natural rate in policy analysis is to identify what the natural rate is, or at a minimum, a fairly narrow range surrounding it. As recently as three years ago, for example, the consensus view was that the natural rate was around 6 percent. As it turned out, the unemployment rate is now 4.7 percent and has been less than 6 percent for 3 ½ years. Over the same period, however, inflation has steadily declined, suggesting that the natural rate is substantially less than 6 percent. Whether the apparent decline in the natural rate is just a temporary change or more permanent is still an open question. In any event, like the monetary aggregates, the natural rate has not proved to be as reliable a gauge to inflationary pressures as was once thought.