What the study found
The study found that published estimates of how unemployment benefits affect unemployment duration are likely overstated because statistically significant findings are eight times more likely to be published. After correcting for this publication bias, the average elasticity is about one-third smaller, and the implied optimal replacement rate in the United States is 28 percent.
Why the authors say this matters
The authors say meta-analysis, which combines results across studies, can be used as a data-driven way to generalize sufficient statistics methods for policy design. The study suggests that correcting for publication bias changes the policy conclusion compared with existing consumption drop-based approaches.
What the researchers tested
The researchers systematically reviewed studies on how unemployment benefits affect unemployment duration. They used meta-analysis to combine estimates across policy contexts and examined publication bias as well as the relationship between 'micro' elasticity, meaning effects seen in smaller individual-level studies, and 'macro' elasticity, meaning effects at the aggregate level.
What worked and what didn't
Statistically significant findings were eight times more likely to be published. Correcting for publication bias reduced the average elasticity by about one-third. The authors also report that they were unable to reject the hypothesis that micro elasticity is equal to macro elasticity.
What to keep in mind
This is a review article based on existing studies rather than a new experiment. The abstract does not describe specific study-by-study limitations beyond publication bias, and the findings are presented in terms of the evidence summarized there.
Key points
- Published studies with statistically significant findings were eight times more likely to appear in the literature.
- Correcting for publication bias reduced the average elasticity estimate by about one-third.
- The corrected estimates implied an optimal unemployment insurance replacement rate of 28 percent in the United States.
- The authors were unable to reject the idea that micro elasticity equals macro elasticity.
- The paper uses meta-analysis to combine evidence across policy contexts.
Disclosure
- Research title:
- Unemployment insurance reduces unemployment less than published estimates suggest
- Authors:
- Jonathan P. Cohen, Peter Ganong
- Institutions:
- University of Chicago
- Publication date:
- 2026-02-25
- OpenAlex record:
- View
Get the weekly research newsletter
Stay current with scholarly research without reading academic papers — one filtered digest, every Friday.