Job market paper
Abstract: I develop a sufficient-statistics framework that determines the optimal design of disability insurance by explicitly incorporating employer behavior. Standard analyses balance insurance benefits against workers’ labor supply incentives, but I account for a novel type of firm-side moral hazard: because employers do not bear the costs of workers’ disability insurance, they may underprovide accommodation. This framework allows me not only to identify the optimal level of experience rating—a policy linking firms’ DI costs to premiums—, but also to show how the standard policies of DI eligibility and benefit generosity are altered when firm behavior is endogenized.
Publication
Abstract: This paper studies the impact of stronger employer responsibilities for facilitating work resumption of sick or disabled workers on employers' workplace accommodation efforts during sick leave. We exploit a reform in the Netherlands that altered experience rating – i.e., shifting the costs of sick leave and disability insurance to the firm – both for permanent and non-permanent employees. Using unique Dutch survey data on workplace accommodation of long-term sick-listed workers, we show that experience rating has no significant impact on accommodation efforts. Moreover, we provide evidence that the reform led to more firms opting for self-arranging both the sick leave benefits and the reintegration process of sick non-permanent workers, instead of using the public insurance scheme.
Working paper
"Application and Employment Effects of Reduced Screening for Disability Insurance Benefits of Older Workers" (with Max Groneck, Pierre Koning, Raun van Ooijen, and Marcel Spijkerman). 2025, working paper.
Abstract: We examine application, award, and employment responses to a 2022 reform in the Netherlands that allowed long-term sick-listed workers aged 60 and above to bypass the medical assessment and receive full Disability Insurance (DI) benefits. Using administrative data and a difference-indifferences design that compares ages 60-64 to 55-59 across pre- and post-reform cohorts, we find that DI applications rose modestly (+1.2 pp relative to the pre-reform baseline of 39.6%) while awards increased markedly (+6.5 pp relative to the pre-reform baseline of 28.7%), driven by higher acceptance and a shift to full benefits. Effects are larger for women, service-sector workers, and the lowest quintile of pre-disability earnings. One year after potential receipt, employment fell by 2.1 pp, and earnings responses imply a crowd-out of 31%. Overall, moral hazard responses are more pronounced in labor supply than in D applications, with application responses emerging only later in the post-reform period, consistent with learning by applicants, employers and caseworkers.