Both the comments and the paper were presented at The First European Central Banking Conference, held in Frankfurt, Germany, November 2-5, 2000.
A little inflation may be a good thing in a modern economy. In particular, it may be possible to maintain lower rates of unemployment with low inflation than with zero or very low inflation. In two recent articles, George Akerlof , George Perry and I (1996,2000) (hereafter, ADP) have presented evidence for two specific mechanisms by which inflation may affect the equilibrium level of unemployment. We modeled those mechanisms, and argued the case that large permanent reductions in unemployment may be obtained by moving from either a high or very low rate of inflation to a moderate rate (2-4% in the United States). The two mechanisms we examined were nominal rigidity in wage setting, and near rationality in the use of inflationary expectations in price and wage setting.
In these papers we estimated Phillips Curve relations from which we deduced the magnitude of the effects of nominal rigidity and near rationality on the long-run relationship between inflation and unemployment. But these empirical exercises do little to verify the nature of this relationship because that was simply assumed when we accepted the dictates of our theory in setting up the specification we estimated. In this paper Charles Wyplosz takes a very different approach to roughly the same empirical problem. He estimates models of unemployment in which he allows the NAIRU, or natural rate of unemployment, to vary with the rate of inflation in a very general way. Such an approach has advantages and disadvantages relative to the approach my colleagues at The Brookings Institution and I have taken. Below I discuss Wyplosz’s results and present some of my own. Taken together, our results suggest the possibility that very low rates of inflation may cause unemployment to be higher than it would be at moderate rates of inflation in the Euro zone.
View Wyplosz’s paper from the ECB website » (PDF)