Non-Take-Up of Unemployment Benefit II in Germany: A Longitudinal Perspective Using Administrative Data

Extensive research demonstrates that many households eligible for means-tested benefits do not claim them, a phenomenon known as non-take-up. Empirical studies frequently conceptualise non-take-up as a rational decision, occurring when the perceived net utility of claiming is negative. Theoretically, long-term factors can substantially impact this decision. Despite the potential relevance of longitudinal aspects, evidence on their influence remains limited. This study addresses this gap by incorporating long-term factors in the analysis of non-take-up behaviour relating to Unemployment Benefit II (UB II), Germany's basic means-tested welfare programme. Using data from the German Panel Study Labour Market and Social Security (PASS) from 2008 to 2020, linked with administrative data from Germany's Federal Employment Agency (PASS-ADIAB), this study reconstructs households' benefit receipt and income histories, even during non-survey periods. This allows modelling benefit non-take-up for eligible households using the duration and frequency of past benefit receipt. In addition, the use of administrative data mitigates bias from self-reported benefit receipt. Household eligibility for UB II is simulated using GETTSIM, an open-source microsimulation model, applied to the PASS dataset for the first time. Findings indicate that long-term factors significantly influence the probability of claiming UB II. Specifically, a longer history of benefit receipt increases this probability, whereas higher income potential and positive income shocks reduce it. Including long-term factors substantially affects the estimated impact of traditionally used determinants of non-take-up, indicating a potential misspecification in existing models that neglect them.