Unemployment Insurance Benefit Changes via Regression Kink in Benefit Schedule by State
Introduction
Unemployment insurance in the United States is administered on a state-by-state basis, with funding coming from both state and federal sources, along with contributions from payroll taxes paid by workers and firms. Both the amount of benefits and eligibility requirements vary by state. Generally speaking, benefits are available to workers who have lost their jobs for reasons outside of their control.
Landais (2015) studies kinks in the UI benefit schedules for five states in the late 1970s and early 1980s.
Pattison (2025) draws upon the estimates from Landais (2015) to compute the MVPF of a marginal change to UI benefit levels. The calculation builds on the work of Schmieder and Von Wachter (2016), Hendren and Sprung-Keyser (2020), and Le Barbanchon et al. (2024) by estimating the administrative costs associated with an additional dollar of UI benefit expenditure.
MVPF = 0.8
Net Cost
The net cost is estimated as $1 of additional UI benefit, administrative costs, as well as the fiscal externalities from reduced search effort for a new job. Reduced search effort means people will be covered by UI for longer, leading to more total UI payouts. In addition, longer periods of non-employment lead to a fiscal externality through reduced income taxes.
The paper estimates $0.09 in administrative costs per additional $1 of UI benefits. The paper uses a calculation from Schmieder and Von Wachter (2016) to get a fiscal externality of $0.40. This leads to a total net government cost of $1 + $0.09 + $0.40 = $1.49.
Willingness to Pay
The willingness to pay for $1 of additional UI benefits is estimated as
to capture that individuals are willing to pay more than $1 to the extent to which they are risk averse. This is measured by the coefficient of relative risk aversion, gamma, and the extent to which they experience a consumption drop upon unemployment. The paper assumes a coefficient of relative risk aversion of 2 and an 8.5% estimate of consumption drop at job loss from Hendren (2017). This leads to a willingness to pay of $1 + 2*$0.085 = $1.17.
MVPF Summary
With a willingness to pay of $1.17 and a net government cost of $1.49, this leads to an MVPF of 0.79.
References
Hendren, Nathaniel and Ben Sprung-Keyser (2020). “A Unified Welfare Analysis of Government Policies.” The Quarterly Journal of Economics, 135(3): 1209–1318. DOI: https://doi.org/10.1093/qje/qjaa006
Landais, Camille (2015). “Assessing the Welfare Effects of Unemployment Benefits Using the Regression Kink Design.” American Economic Journal: Economic Policy, 7(4), 243-278. DOI: https://doi.org/10.1257/pol.20130248
Le Barbanchon, Thomas, Johannes Schmieder, and Andrea Weber (2024).“Chapter 6 – Job search, unemployment insurance, and active labor market policies.” Handbook of Labor Economics, 5: 435-580. DOI: https://doi.org/10.1016/bs.heslab.2024.11.006
Pattison, Nathaniel (2025). “Administrative Costs and Optimal Unemployment Insurance.” Working Paper. https://pattison-nate.github.io/files/adminUI_pattison_oct2025.pdf.
Schmieder, Johannes F. and Till Von Wachter (2016). “The Effects of Unemployment Insurance Benefits: New Evidence and Interpretation.” Annual Review of Economics, 547-581. DOI: https://www.annualreviews.org/doi/abs/10.1146/annurev-economics-080614-115758
Policy Features
- Category
- Social Insurance
- Sub-Category
- Unemp. Ins.
- Beneficiary Type(s)
- Adults
- Average Age
- 34
- Average Income
- 48737
- Country of Implementation
- United States
- Year of Implementation
- 1980
- Empirical Method
- Regression Discontinuity
- Research Type
- Secondary
- Peer Reviewed
- No
- MVPF Publication Link
- pattison-nate.github.io/files/adminUI_pattison_oct2025.pdf