New software, more employment?

Abstract

We evaluate a major software change in the placement software of the German Federal Employment Agency in 2006. This change was remarkable as it aimed at improving key factors of the placement process by increasing the accuracy and number of job placements. As a result of this effort, we would expect more transitions into employment leading to shorter unemployment durations and thus increased lifetime earnings by improved matching rates between unemployed and jobs. We take advantage of the staggered implementation of the new software over time to estimate its effect on employment rates. By employing a dynamic difference-in-differences approach that takes into account the time under treatment, we show that the software implementation led to an increase in both, overall employment and employment stability.