Worker churn in the cross section and over time: New evidence from Germany

Abstract

Worker churn, that is, worker flows in excess of job flows, is procyclical in the German labor market. To understand this procyclicality, we study the plant-level connection of churn and employment growth, using the new Administrative Wage and Labor Market Flow Panel from 1975 to 2014, and find that churn rises in the absolute value of employment growth. Analyzing this V-shaped churn-employment growth nexus by worker skill, age, and tenure, we establish that churn is unlikely to result from plant reorganization but rather from the correction of labor market mismatches. Using a simple dynamic labor demand framework, we argue that the crosssectional evidence on churn can be interpreted as manifestations of idiosyncratically stochastic separation shocks in conjunction with a time-to-hire friction. These shocks become larger and more predictable during booms leading to procyclical churn, which, as we show, (1) increases almost uniformly across the employment growth distribution, and (2) stems almost exclusively from job-to-job transitions. Procyclical churn, thus, reflects a more active reshuffling of workers towards individually better matches in booms.

Monopsony Makes Firms Not Only Small But Also Unproductive: Why East Germany Has Not Converged

“…For some analyses, principally for longer time series, we supplement the SES with the AWFP which is a quarterly plant-level data set based on German social security data and which contains daily earnings, not wages, up to the social security cap. The data covers the universe of private German plants and is available for both West and East Germany from 1993 until 2014 (see Stüber and Seth, 2017;Bachmann, Bayer, Merkl, Seth, Stüber, and Wellschmied, 2021). The AWFP's data source is the Employment History (Beschäftigten-Historik, BeH) of the German Institute for Employment Research (IAB).…”

“…To this end, we study quarterly job and worker reallocation rates as defined and explained in detail in Bachmann, Bayer, Merkl, Seth, Stüber, and Wellschmied (2021). Figure A4 (a) displays the job turnover rates for East and West Germany.…”

Job security, asymmetric information, and wage rigidity

“…In the model with asymmetric information there is no undercutting so that workers do not get replaced in recessions and unemployment rises only via a fall in hiring. The findings of Bachmann et al (2021) for Germany suggest that replacement hiring, as defined by our theory, does not seem to be significant in Germany. If it exists, it would imply that worker churn, due to separations and hires into and out of nonemployment, increases in recessions.…”

“…If it exists, it would imply that worker churn, due to separations and hires into and out of nonemployment, increases in recessions. However, Bachmann et al (2021) show that cyclical variations in worker churn, "which are actually procyclical", are accounted for almost wholly by job-to-job transitions rather than by transitions to and from nonemployment.…”

Structural Estimation of Labor Adjustment Costs With Temporally Disaggregated Data

“…Most importantly, it can account for churning as an active choice of an establishment. Churning can also occur in models with (random) worker‐initiated separations as in Bachmann et al (), but in the context of adjustment costs the exact reason for churning is important if the cost of worker‐initiated separations is different from the cost of establishment‐initiated separations.…”