Essays on Spillover Effects and Unintended Consequences of Pension Policies

This dissertation consists of three independent studies that examine different types of spillover effects and unintended consequences that may arise from pension policies. The first study (Chapter 2) focuses on spillover effects on peer co-workers induced by the 1999 German pension reform. The reform abolished the old-age pension for women, causing a substantial increase in employment rates among women born after 1951. The chapter examines whether workplace social interactions generated spillovers among peer co-workers employed in the same establishment and occupational group. Using administrative linked employer-employee data, the study applies an instrumental variable approach based on the share of reform-affected co-workers to identify causal peer effects. The results show sizable peer effects: women are more likely to extend their employment when a higher share of their peers delayed their retirement due to the reform. Heterogeneity analyses suggest that information exchange, the transmission of social norms, and employer characteristics shape the magnitude of these peer effects. The second study (Chapter 3) analyzes spillover effects within married couples. Specifically, the chapter investigates whether the extension of older women’s working lives induced by the 1999 pension reform affected their husbands’ own retirement behavior. To address this question, the study uses administrative employment records and couple identifiers based on geocoded addresses and surnames within a regression discontinuity design centered around the reform cutoff. The estimates show that husbands responded to their wives' prolonged employment by extending their own employment as well. To investigate the underlying mechanisms, the study examines effect heterogeneity and conducts additional analyses with the German Socio-Economic Panel (SOEP). The findings provide suggestive evidence that joint preferences for shared leisure and adherence to traditional gender norms play an important role in explaining husbands’ responses. The third study (Chapter 4) shifts the focus from spillover effects on older workers’ retirement decisions to unintended consequences for the career outcomes of younger employees. Using linked employer-employee data and an event-study difference-in-differences design, the chapter compares firm adjustments across establishments according to whether they employed women affected by the 1999 reform. The results show that the reform increased the retention of older female workers in affected establishments. At the same time, this increase was accompanied by reduced hiring and fewer wage increases among workers from younger age groups. Further analyses suggest that firm adjustments vary in both magnitude and type depending on the extent of firm-specific human capital accumulated by the affected older female workers.