Short-Term Forecasting of Unemployment Rate (Evidence from EU Countries)

Nowadays unemployment rate is an important indicator that is used to assess the labour market situation. European authorities pay special attention to labour market policy as it may foster business activity by using extra labour resources. Nonetheless, the best strategy to monitor the dynamics of unemployment is yet to be discovered. This paper attempts to describe the potential to use advanced short-time forecasting methods as a tool to track changes using more affordable data than the results of labour force surveys. The research is also focused on testing the hypothesis that Internet search data may improve the quality of unemployment forecast in most European countries. We expect that internet search data indicates the real demand for jobs, so that, this statistic can significantly improve the quality of the forecast. The main contribution of the study is the coverage of potential improvement of short-term forecast accuracy in 28 European countries using an extensive set of model variations and metrics to evaluate models are used.