Research

Unemployment insurance (UI) eligibility requires a claimant to have lost their job through no fault of their own. Approximately 10\% of claims are deemed ineligible solely based on the job separation reason. Using the systematic variation in separation-based eligibility approval rates across UI claim processing offices and examiners in California from 2002 to 2019, we show that receiving any UI benefits causes approximately 2 additional weeks of nonemployment. These effects are the smallest for lower-income claimants. By replicating existing research designs for other UI policy margins within the California data, we conclude the efficiency costs of UI benefit expansions through separation-based eligibility criteria are lower compared to those of expansions through monetary eligibility, weekly benefit amount, or potential benefit duration.

We use a panel of survey responses linked to administrative data in Germany to measure the depreciation of skills while workers are unemployed. Both the reemployment hazard rate and reemployment earnings steadily fall with unemployment duration, and indicators of depression and loneliness rise substantially. Despite this, we find no decline in a wide range of cognitive and noncognitive skills while workers remain unemployed. We find the same pattern in a panel of American workers. The results imply that skill depreciation in general human capital is unlikely to be a major explanation for duration dependence.

The empirical literature on unemployment insurance (UI) robustly finds significant effects of UI benefit generosity on unemployment duration. We systematically collect estimates and contextual factors from 52 studies, documenting publication bias favoring increases in unemployment duration based on all research designs. Publication-bias corrections nevertheless imply that UI benefit increases lengthen unemployment duration. We additionally provide suggestive evidence on sources of heterogeneity and highlight key practices for the ongoing literature to follow in order to facilitate building evidence across contexts.

We use linked administrative education and employment data from Colorado since 2010 to study the impact of reverse transfer, which allows students who did not complete a 4-year college degree to apply their credits towards a 2-year degree. Our selection on observables identification strategy leverages idiosyncratic variation in the transferability of courses across schools.