In July 2017, Lithuania moved toward flexicurity by significantly increasing the generosity of unemployment insurance (UI) while simultaneously decreasing firing costs. We use administrative data on all Lithuanian workers to quantify labor market effects of i) the reform, which increased the potential benefit duration and the monthly benefit level by 50 percent, and ii) variation in severance pay. We find that the elasticity of the non-employment duration w.r.t. a proportional increase in both policy parameters is 0.4, and the elasticity of the benefit duration is 0.7, suggesting a substantial fiscal cost.

Finally, we find little evidence for a reform effect on re-employment wages.

What is the role of income in driving the choice to claim disability benefits, and how can broader benefit policy affect that? We study three UK reforms which cut non-disability benefits and find that each increased the number of individuals receiving disability benefits, with an elasticity of claims to incomes between -0.4 and -1.1. Our results provide causal evidence that disability benefits provide insurance against non-health related shocks. These effects have implications for the optimal level of non-disability benefit generosity, and the savings policymakers might hope to generate from benefit cuts.

Governments spend heavily on unemployment insurance (UI), yet research on entitlement length focuses almost entirely on the claimant's own job search. This paper asks whether the rest of the family adjusts labor supply when a member gets a longer UI entitlement. The added worker effect shows wives work more when husbands lose jobs, but whether entitlement length shapes this, or extends beyond couples, is unstudied. The net effect is theoretically ambiguous, as this paper's model shows. Using Dutch administrative data, I exploit sharp discontinuities in maximum UI entitlement duration in a regression discontinuity design. Longer entitlement raises actual benefit duration and delays unemployment exits. Spouses and adult children raise employment in response, while parents do not; spouses respond from the start of the spell, children near benefit expiration. Accounting for them raises the MVPF from 0.756 to 0.853. This is the first evidence on UI duration spillovers within couples and across generations.

This paper studies social insurance with imperfect eligibility screening, focusing on Unemployment Insurance during its expansion in 2020-21. We study the extent of imperfect screening by identifying anomalous payments using administrative tax data and UI policies, finding $214 billion in potentially-improper payments—concentrated in the Pandemic Unemployment Assistance (PUA) program—with half detectable ex-ante through improved data sharing. There is substantial geographic variation, and a border design shows this is partly due to state policy decisions. To assess implications for optimal policy, we first conduct simulations that replace PUA with means tested, lump-sum transfers, finding these transfers would have better insured against income losses at lower administrative cost. Second, we develop a model of opt-in versus automatic transfers that shows the targeting advantage of opt-in programs can reverse when ineligible recipients pass the benefit screen. Calibrated to 2020 UI, the model implies that shifting toward automatic transfers would have increased social welfare.