Worker flows in Germany: Inspecting the time aggregation bias

• The paper studies the effects of time aggregation in the measurement of worker flows.
• Daily information on worker flows is provided by German administrative data.
• Monthly time aggregation leads to an underestimation by 10%.
• The bias in the job-finding rate shows procyclical behavior.
• The bias in the separation rate appears to be rather acyclic.

Employing daily information from German administrative data, this paper analyzes the effects of time aggregation in the measurement of worker flows. Time aggregation that is based on a comparison of monthly labor market states leads to an underestimation of total worker flows by approximately 10%, which is larger than the prediction of a standard correction approach. Multiple labor market transitions within a month induce a procyclical bias in the job-finding rate but not in the separation rate. Rather, during upswings, workers are more likely to change jobs with short intervening transition periods. Relative to previous findings, this reconsideration of German worker flows reveals that the job-finding rate plays a larger role in explaining unemployment fluctuations, although the additional dynamics of the time aggregation bias are shown to be less significant.