In this paper we evaluate the forecasting performance of employmentexpectations for employment growth in 15 European states. Our data coverthe period from the first quarter 1998 to the fourth quarter 2012. With in-sampleanalyses and pseudo out-of-sample exercises, we find that for most of the Europeanstates considered, the survey-based indicator model outperforms commonbenchmark models. It is therefore a powerful tool for generating more accurateemployment forecasts. We observe the best results for one quarter ahead predictionsthat are primarily the aim of the survey question. However, employmentexpectations also work well for longer forecast horizons in some countries.
Employment forecasting European business survey employment expectations Granger causality