On the non-inflationary effects of long-term unemployment reductions

The paper critically examines the New Keynesian explanation of hysteresis based on the role of long-term unemployment. We first examine its analytical foundations, according to which rehiring long-term unemployed individuals would not be possible without accelerating inflation. Then we empirically assess its validity along two lines of inquiry. First, we investigate the reversibility of long-term unemployment. Then we focus on episodes of sustained long-term unemployment reductions to check for inflationary effects. Specifically, in a panel of 25 OECD countries (from 1983 to 2016), we verify by means of local projections whether they are associated with inflationary pressures in a subsequent five-year window. Two main results emerge: i) the evolution of the long-term unemployment rate is almost completely synchronous with the dynamics of the total unemployment rate, both during downswings and upswings; ii) we do not find indications of accelerating or persistently higher inflation during and after episodes of strong declines in the long- term unemployment rate, even when they occur in country-years in which the actual unemployment rate was estimated to be below a conventionally estimated Non-Accelerating Inflation Rate of Unemployment (NAIRU). Our results call into question the role of long-term unemployment in causing hysteresis and provide support to policy implications that are at variance with the conventional wisdom that regards the NAIRU as an inflationary barrier.