We show that the macroeconomic uncertainty series from Jurado, Ludvigson, and Ng (2015) contains information to forecast employment. The results indicate that the uncertainty measure is weak at forecasting the skilled labour but significantly carries forecasting information on the unskilled labour. The forecasting information increases if the sample is restricted to construction and manufacturing industries. Using rolling regressions to conduct a simulated out-of-sample forecasting exercise, we find that the uncertainty measure contains forecasting information for the unskilled labour in those industries for two quarters ahead. By providing detailed information about the forecasting power of uncertainty by skill and industry, this study will be helpful in designing more efficient labour market policies.
All Science Journal Classification (ASJC) codes
- Economics and Econometrics