The Longer, the Weaker? Considering the Role of Long-Term Unemployment in an ‘Original’ Phillips Curve

A recent explanation of hysteresis suggests that downward wage rigidity may be caused by long-term unemployment as individuals who are detached from the labor market are not good inflation fighters. Based on this hypothesis, some scholars have attempted to estimate the Phillips curve and the Non-Accelerating Inflation Rate of Unemployment (NAIRU) either by including only short-term unemployment or by taking into account weighted measures of unemployment duration. This essay investigates whether the conclusion that long-term unemployment is less relevant to the inflation path than short-term unemployment is supported by data. To do this, we consider the short- and long-term unemployment rates jointly in a Phillips curve, while also considering institutional variables and without using any measure of the unemployment gap or the NAIRU. Using panel data for 25 OECD countries for the period 1970–2016, we test whether there is a statistically significant difference between the coefficients of short- and long-term unemployment rates in this version of the Phillips curve. Our findings indicate that there is no significant difference between long and short-term unemployment rates in wage-setting, and provide a critique of the mainstream explanation of the hysteresis mechanism.