Essays in labor economics with applications to Germany

This dissertation consists of three essays in LaborEconomics. The first chapter estimates the costs for establishments of hires andseparations for a panel of German establishments. The panel records the start and thetermination of the employment of all employees in the surveyed establishments, allowingestimation of adjustment costs under different assumptions of how frequentlyestablishments revise their labor demand. Under the assumption that establishments revisetheir labor demand every month, estimates suggest hiring costs per employee ofapproximately 5,000 Euros, and costs of separations of 1,000 Euros. Hiring costs varyconsiderably between skilled (8,000 to 28,000 Euros per hire) and unskilled (4,000 to8,000 Euros) labor. Spatial aggregation (large establishments) is associated with lowercost estimates, and only monthly adjustment frequencies yield estimates consistent withtheoretical predictions. The second chapter analyzes the role of regional ethnic capital - defined as the averageyears of schooling of ethnic groups - in the educational attainment of young secondgeneration immigrants in Germany using information on naturalization and country of birthin a nationally representative survey. I find evidence for externalities of ethniccapital for ethnic groups. A higher average education of ethnics makes attendance ofhigher-quality secondary schools more likely. Moreover, the effect is mainly mediatedthrough the ethnic concentration in the region. However, if higher than regionalaggregates are used for the measurement of ethnic capital, no externalities are detected.The third chapter analyzes the impact ofregional own-ethnic concentration on the language proficiency of immigrants in Germany.It solves the endogeneity of immigrants' location choices by exploiting the fact thatguest-workers in Germany after WWII were initially placed by firms and labor agencies. Wefind a robust negative effect of ethnic concentration on immigrants' language ability.Simulation results of a simultaneous location and learning choice model confirm thepresence of the effect and show how immigrants with high learning cost select into ethnicenclaves. Under the counterfactual scenario of a regionally equal distribution ofimmigrants the share of German-speakers increases only modestly.