Definition
Program spillovers occur when a policy change in one public program shifts caseloads and costs to other programs rather than reducing total government expenditure or total recipient hardship. In the disability insurance (DI) context, the welfare-shifting hypothesis holds that tightening DI eligibility criteria causes denied or deterred applicants to flow onto substitute programs — Temporary Assistance for Needy Families (TANF), Supplemental Nutrition Assistance Program (SNAP), unemployment insurance (UI), general assistance, Supplemental Security Income (SSI) — leaving aggregate public spending and applicant outcomes largely unchanged.
Key Ideas
- Rejected DI applicants do not primarily return to work; instead they shift onto other social support programs.
- The fiscal savings from a denial are partially or fully offset by increased costs in downstream programs not budgeted under the Social Security Administration (SSA).
- Total social cost is not eliminated — it is redistributed across agencies, often to programs with lower per-recipient benefits (SNAP, general assistance) and no health insurance component, worsening recipient welfare.
- The effect is an upper bound: a portion of the social-support increase reflects health deterioration common to the applicant population rather than the denial decision itself.
How It Works
Chen (2014) documents the mechanism using Survey of Income and Program Participation (SIPP) 1990–2008 data matched to SSA administrative records. Rejected DI applicants' probability of being on social support programs rises sharply at the filing date and remains elevated across a 10-year window. Crucially, by t+7 to t+10, the gap between rejected and approved applicants collapses to at most 7 percentage points (pp) — statistically indistinguishable from zero — meaning that denial does not meaningfully reduce long-run program dependency; it merely determines which program the individual depends on.
Labor supply does not recover: employment is 23 pp below pre-filing levels at t+10 and earnings remain roughly $8,800 below trend, ruling out work-restoration as an offsetting mechanism.
Why It Matters
- Fiscal arithmetic of DI stringency: advocates of tightening the DI standard often cite projected savings to the Social Security trust fund. Spillover evidence implies that those savings are partially illusory — they appear in one budget while being absorbed in others (TANF, SNAP, Medicaid, state general assistance).
- Program design: if denial shifts rather than reduces costs, the policy question becomes where on the program spectrum applicants land — and whether the substitute programs provide adequate health and income support.
- Cohort sensitivity: Chen (2014) finds larger spillover effects for 2000s filers vs. 1990s filers, suggesting the applicant pool composition (more severely impaired, lower education) interacts with spillover magnitude.
Open Questions
- Do spillovers fully offset the DI fiscal savings, or only partially? Chen's estimates are upper bounds; a cleaner causal design (instrumental variable (IV) rather than difference-in-differences (DiD)) is needed.
- Are spillover effects symmetric? Loosening DI access may reduce other program caseloads — the mirror question to Chen's denial analysis.
- How do spillovers vary by state, given that state-level programs (Medicaid generosity, TANF rules) differ substantially?
Related
Sources