The Danish Matched Employer-Employee Data

The Danish data constitutes a unique source of information: it covers and links together the universe of persons, establishments and firms for more than 30 years. In addition workers’ histories are constructed at the weekly level. The data has the potential to become a benchmark in social research, but access has been limited by the lack of systematic documentation. The chapter provides an introduction, with a particular emphasis on those portions more commonly used in labor market research. We establish a wide variety of descriptive statistics that can serve as a reference for future studies. The second part is dedicated to a more detailed analysis of wages and flows, for which the data is uniquely well suited. Returns to seniority are low on average and even lower for workers coming from unemployment; wage changes between jobs are large, but often negative and in these cases the wage remains persistently lower. Labor market turnover is high, the unconditional job and unemployment hazard display strong negative duration dependence. Also, higher wages are associated with a lower probability and duration of unemployment. Finally, we use the data to revisit the recent debate on the “ins and outs” of unemployment: the reduced form decomposition assigns 2/3 of unemployment volatility to the ue rate; the ue rate is procyclical, while the eu rate countercyclical and leads unemployment; the ee rate is procyclical, so the separation rate is essentially acyclical. These results are remarkably consistent with findings from U.S. data.