Summary
Davies, Rupp, and Wittenburg (2009) apply a life-cycle human capital framework to Supplemental Security Income (SSI) children, synthesizing program rules, National Survey of SSI Children and Families (NSCF) 2000–2001 descriptive data, and new longitudinal Social Security Administration (SSA) administrative records for 1980 and 1997 award cohorts. The paper documents the institutional history (Sullivan v. Zebley, 1996 welfare reform), characterizes the dual challenges of disability severity and family poverty, and presents the first longitudinal comparison showing ~40% of childhood SSI awardees still receiving benefits at age 45 and a flat adult earnings trajectory far below the general population. The paper frames the basis for front-end Youth Transition Demonstration (YTD) interventions as a response to prior back-end demonstration failures.
Key Claims
- Life-cycle dual challenge: SSI children face simultaneous constraints from (1) severe health problems and disabilities and (2) family poverty. Human capital development is determined by three inputs: nature/severity of disability, family environment, and public programs. Effective interventions must address all three; employment-oriented programs that ignore upstream constraints are likely to produce limited results.
- Sullivan v. Zebley (1990) and the 1996 welfare reform: Zebley decision found SSA's "Listings-Only" approach violated the "comparable severity" standard; created the Individualized Functional Assessment (IFA); ordered redetermination of all childhood denials since 1980 → caseload surge. The 1996 welfare reform eliminated the IFA, required age-18 redeterminations under adult criteria → 100,000+ children removed; 45% of the first cohort of age-18 redeterminations resulted in benefit cessation. The caseload initially dropped but subsequently resumed growth, surpassing its pre-reform level.
- NSCF 2000–2001 family profile (N≈792,000): 64% male; 47% White, 46% Black; primary impairment: 33% mental retardation (MR), 29% mental/behavioral disorders (other than MR), 25% physical. 56% single-mother families; 44% below poverty level; 47.7% of family income from SSI. 48.9% of households contain another member with a disability. 90% Medicaid coverage. 70% ever enrolled in special education; 68% ever had an Individualized Education Program (IEP). Average monthly SSI benefit: $672. 67% of transition-age (14–17) recipients entered SSI before age 10.
- Long-term SSI participation — 40% still on at age 45: For the 1980 award cohort (age 17 at award), SSI participation declines monotonically but slowly: still ~40% at age 45. Rupp and Scott (1998) estimate mean SSI duration (including multiple spells) for 1974–1982 childhood awardees at ~27 years before reaching age 65. The 1997 cohort shows faster decline in participation, suggesting the 1996 welfare reform produced a structural shift toward shorter durations.
- Employment: flat earnings trajectory, large gap vs. general population: Childhood SSI awardees: ~40% have positive earnings in adulthood vs. ~80% general population. Average earnings conditional on employment: 10,000–13,000 for ages 24–38 vs. 33,000–48,000 for the general noninstitutional population in the same age group. Critically, the age-earnings profile is flat for SSI awardees — no wage growth with age — unlike the steep rise in the general population trajectory. Employment peaks around early-mid 20s for the 1980 cohort and slowly declines thereafter.
- 1997 vs. 1980 cohort improvement in employment: 1997 cohort shows employment (positive earnings) peaking at ~50% around age 20, substantially higher than the 1980 cohort's ~20–28% for comparable age groups. This is consistent with the faster SSI exit trend.
- Exits driven by excess income, not death: Rupp and Scott (1998) found >1/3 of 1974–1982 childhood awardees had no exit in the first 10 years; among those who exited, ~half exited due to excess income (earnings above the means test). Death is a rare reason for exit: fewer than 7% completed their first SSI spell via death. SSI data for 1999–2007 confirm that suspensions for excess income/resources accounted for 50–60% of all childhood SSI suspensions.
- Prior SSA demonstrations: "somewhat disappointing": Three back-end employment demonstrations — Transitional Employment Training Demonstration (TETD; 1985–87, job placement for adults with mental retardation), Project NetWork (1992–94, case management for disability insurance (DI)/SSI; some youth outreach ages 18–24), and State Partnership Initiative (1999–2004, benefit counseling) — produced limited employment gains. Motivated the strategic shift to earlier, front-end YTD interventions.
- Youth Transition Demonstration rationale: YTD (ages 14–25) combines intensive employment supports with SSA work-incentive waivers to retain benefits while working. Even a small reduction in caseload is cost-effective given large lifetime expenditures. Random assignment design. Key tension: voluntary participation enables self-selection ("cream skimming") vs. mandatory participation may harm those not suited for employment services ("do no harm").
Concepts Introduced or Extended
Entities Mentioned
Quotes
"A substantial fraction appears to be headed for a lifetime of reliance on SSI benefits and weak labor market outcomes."
"The data show no improvement in the work patterns of childhood SSI awardees past their mid-20s."
"Because most young adults who received SSI benefits as children are not investing in their human capital through work, rehabilitation, or education for a variety of reasons, they are at high risk for a lifetime of reliance on disability benefits and Medicaid."
My Take
This is primarily a descriptive-institutional paper that serves as the contextual anchor for a special Journal of Vocational Rehabilitation (JVR) issue on SSI children in transition. Its most durable empirical contributions are the longitudinal cohort charts (Figures 1–2) documenting SSI participation and earnings into middle age — these remain among the few direct estimates of lifetime SSI reliance for childhood awardees. The findings are sobering: 40% still on SSI at 45, a flat earnings ceiling of ~$12,000/year, and a 27-year expected SSI duration. The paper predates the Deshpande regression-discontinuity (RD) studies, so it cannot identify causal effects. Its strength is in the breadth of the life-cycle framework and the longitudinal administrative data. The Sullivan v. Zebley and 1996 reform history is well-documented and provides institutional context missing from most of the causal literature. One limitation: the figures are largely unreadable in the PDF extraction (right-to-left axis labeling), so the cohort participation charts must be read from the text descriptions rather than directly from figures.