Technology, Unemployment, and Inflation

Abstract

We explore the response of employment (unemployment) skill differentials to skill-biased shifts in demand touched off by the new and spreading technologies. We find that skill differentials in unemployment follow at least in part the same pattern as skill differentials in wages: They widen initially but decline after a roughly 5-year lag, allowing time for training and learning to handle the new technologies. In the micro (PSID) cross-section the differentials show up as sectoral differences defined by technology. In the aggregate time series relative unemployment is defined by educational unemployment differentials. We find that the pace and turnaround in the unemployment gap' is twice as fast as in the wage gap'. Apparently, the hiring and training response is quicker than the wage response. We also observe in time series that the pace of technology has unclear effects on aggregate unemployment in the short run, but appears to reduce it in the longer run. In addition to technology, maturing of the workforce, and growth of international trade reduce unemployment in the longer run. The same variables also significantly reduce inflation in both the short and long run. Given the actual changes in these factors in the early 90's we are able to predict a little over a half of the decline in unemployment and about 70% of the reduction in inflation in the latter half of the last decade.