NAIRU and the problem of inverse regression

Abstract

This paper reviews the historical evolution of the concept of the Non-Accelerating Inflation Rate of Unemployment (TDNAI or NAIRU). Some methodological problems faced by those who propose to estimate this variable are reviewed, and it is mainly shown that the calculations of a virtual nairu from the direct sample regression of ?t as a function of ut are always biased due to the fact that the degree of linear association between both variables is not perfect. Additionally, it is shown that, even ignoring the problems examined, the empirical evidence severely questions the relevance of the nairu as a guide for monetary policy or as an instrument for forecasting the inflation rate.