Overview
Manasi Deshpande is an economist at the University of Chicago. Her research focuses on the design and effects of social insurance programs, with particular emphasis on disability insurance, welfare program access, and the long-run consequences of program participation. She is known for combining quasi-experimental methods with large-scale administrative data to study both the effects of receiving benefits and the barriers that prevent eligible individuals from accessing them.
Key Contributions
- Deshpande, Gross, and Su (2021) — "Disability and Distress: The Effect of Disability Programs on Financial Outcomes." AEJ: Applied Economics 13(3): 197–229. First quasi-experimental evidence on DI's effects on financial distress. Documents "Ashenfelter's peak": bankruptcy, foreclosure, and eviction rates peak at disability application date. Vocational grid RD (ages 50, 55) plus novel office classification strategy (RD/Spline/Hybrid DDS offices) yields IV estimates of −31% bankruptcy, −34% foreclosure, −15% net home sale, +14% home purchase within 3 years. Treatment is primarily timing: on average 0.9 additional months of benefits. Optimal annual benefit increases $50–240 with tail risk; MVPF rises from 0.99 to 1.04 with property spillovers. Administrative data linkage (SSA 831 + Gross et al. bankruptcy + CoreLogic + AIRS eviction) is the foundation for Deshpande-Lockwood (2022). See Nonhealth Risk and DI Insurance Value.
- Deshpande and Li (2019) — "Who Is Screened Out? Application Costs and the Targeting of Disability Programs." AEJ: Economic Policy 11(4). Exploits 118 SSA field office closings as a natural experiment; applications −10%, recipients −16%, targeting worsens. Refutes Nichols-Zeckhauser (1982). Net welfare cost ≈ $1.2 billion.
- Deshpande (2016a) — "The Effect of Disability Payments on Household Earnings and Income: Evidence from the SSI Children's Program." Review of Economics and Statistics 98(4): 638–654. Exploits FY2004/05 medical review budget cut as quasi-experiment. Finds parents fully offset child SSI removal with earnings (elasticity ≥ 1, intensive margin only); disability applications by family members fall 50–100% but receipt doesn't (marginal applicants deterred); large asymmetry between exit and entry margins (entry IV: ~0.29, exit IV: ≥1); household clustering of disability applications.
- Deshpande (2016b) — "Does Welfare Inhibit Success? The Long-Term Effects of Removing Low-Income Youth from the Disability Rolls." AER 106(11): 3300–3330. Uses a birthdate RD on the PRWORA 1996 age-18 review cutoff (August 22, 1996). Sample: 81,800 SSI children, 63% male, 73% mental/intellectual conditions, avg parental earnings 9,600.Firststage:−10ppSSIenrollment.Removedyouthrecoveronly⅓oflostSSIincomeinearnings(+825/year vs. −2,170/yearSSIloss);PDVincomeloss−21,000 (−19%); income volatility CV quadruples (0.682 → 1.010). Parents do not offset (contrast with 2016a, where parental offset is complete); household PDV −16,400.Siblingslose−580/year. No effects on education, incarceration, or next-generation SSI receipt. MVPF = 0.90 (risk-neutral) to 1.03 (γ=2), exceeding EITC, food stamps, and housing vouchers. See SSI Children's Program and Causal Effects of DI Receipt.
- Deshpande, Kellogg, Mogstad, and Tseng (2025) — "Explaining the Historical Rise and Recent Decline in Social Security Disability Insurance Enrollment." Develops a bounds-based decomposition of SSDI enrollment across the 1988–2010 rise and the 2010–2019 decline. Key finding: reduced applications — not tighter eligibility — account for ~71% of the post-2010 enrollment fall (after adjusting for applicant composition). Labor demand recovery is the dominant causal mechanism (shift-share IV: −0.368 pp application rate per 1 pp employment-to-population). ALJ reform had essentially zero effect on applications; field office closures account for ~3–6%. See DI Growth Decomposition.
- Deshpande and Lockwood (2022) — "Beyond Health: Nonhealth Risk and the Value of Disability Insurance." Econometrica 90(4). Develops a sufficient-statistics welfare framework (EAWTP) using PSID consumption merged with SSA 831 records. Key findings: DI surplus = 8,700/recipient(642,200 for M-NDI); selective application via the SGA earnings limit — not SSA's award process — is the mechanism; MVPF ≈ 1.42 vs. UI ≈ 0.61. See Nonhealth Risk and DI Insurance Value.
- Deshpande, Gross, and Wang (2017) — "The Effect of Government Cash Assistance on Household Financial Outcomes." Working paper (preliminary). Applies the PRWORA birthdate RD-DD from Deshpande (2016b) to PACER bankruptcy records. Main finding: SSI removal reduces parent bankruptcy filing probability by 12 pp (80% decline from 14.6% base); no effect on SSI youth's own bankruptcies; household combined: −9.2 pp (55% decline). Driven entirely by Chapter 7 (low-income) filings. Proposed credit access channel: SSI income generates creditworthiness; losing SSI means losing credit and therefore the capacity to incur bankruptcy. Paper was preliminary as of September 2017; credit bureau test of mechanism was pending. See SSI Children's Program and Nonhealth Risk and DI Insurance Value.
- Deshpande and Dizon-Ross (2022) — "The (Lack of) Anticipatory Effects of the Social Safety Net on Human Capital Investment." RCT with ~6,000 parents of SSI children ages 14–17. Tests the Dynamic Discouragement hypothesis — that expected future SSI benefits reduce parental human capital investment — by randomizing information about age-18 redetermination. Strong first stage (+20 pp belief shift about SSI removal, F=94); null main result (−0.2 pp, 95% CI rules out +1.5 pp). Rejects expert predictions (+14 pp) and structural model predictions (+11%). Mechanisms: 89% of parents already at maximum investment level; 49% plan to increase own labor supply rather than child investment; wealth effect depresses college plans. See Dynamic Discouragement.
- Deshpande and Mueller-Smith (2022) — "Does Welfare Prevent Crime? The Criminal Justice Outcomes of Youth Removed from SSI." Extends Deshpande (2016b)'s PRWORA birthdate RD design, linking the SSI sample to CJARS criminal justice records (~50% US population). SSI removal increases total criminal charges by 22.7% (IV: +0.464 charges), income-generating charges by 60.8% (+0.380), and annual incarceration likelihood by 61.7% (+2.9 pp). No effect on non-income-generating crimes — consistent with income substitution as the exclusive mechanism. Crime effects grow over the 16-year follow-up via path dependence (criminal record forecloses legal employment), even as the contemporaneous SSI income gap narrows. MVPF = 5.6 (without victim costs) to 16.1 (with victim costs). The paper reframes SSI as a crime-prevention program. See SSI Children's Program and Causal Effects of DI Receipt.
- Deshpande, Fadlon, and Gray (2021) — "How Sticky Is Retirement Behavior in the U.S.?" CWHS administrative data, ~2.76 million workers. Central finding: claiming spikes shift one-for-one with the FRA increase (65 → 66), but retirement is persistently sticky at the old FRA of 65 across 10 post-reform birth-year cohorts. RD using 2-month FRA increments rules out earnings responses above 292–312 (expected ~$1,000+ if retirement had responded like claiming). Employer norms are the most supported mechanism: movers to workplaces where others are 1 pp more likely to retire at 65 are 8.4 pp more likely to do so themselves. Creates a 65–66 gap year with no income for post-reform workers who exit at 65. See Retirement Age Stickiness.
Related
Sources