Rehabilitation Reform

Abstract

The employment gap between disabled and nondisabled Americans has been remarkably stable for nearly two decades, and the dominant explanations — impairment severity, employer attitudes, benefit disincentives — cannot account for that stability. This article advances a different claim: the gap is substantially produced and maintained by the administrative architecture of the disability service system itself. Drawing on medical sociology’s account of institutions as distributors of opportunity and on the administrative burden literature in public policy, I argue that the American vocational rehabilitation (VR) system exhibits structural discontinuity — a condition in which statutory entitlement, service capacity, and administrative accountability are lodged in separate institutions with non-aligned incentives, such that the coordination work required to convert eligibility into employment falls on the disabled person. Evidence from federal labor statistics, Rehabilitation Services Administration (RSA) program data, and peer-reviewed VR research indicates that rationing mechanisms, split transition mandates, and geographically uneven service delivery operate not as system failures but as designed features with predictable distributional consequences. The policy implication is that interventions targeting service content will continue to underperform until the coordination burden is shifted from the individual to the state.

Introduction: The Gap That Does Not Move

In 2025, the employment-population ratio for people with a disability was 22.8 percent, compared with 65.2 percent for people without a disability (U.S. Bureau of Labor Statistics [BLS], 2026). Restricting the comparison to working-age adults narrows but does not close the gap: 38.1 percent for disabled people ages 16 to 64. Approximately 75 percent of disabled people were not in the labor force at all, compared with about 32 percent of nondisabled people (BLS, 2026).

A methodological note is warranted before these figures are used analytically. The 2025 annual estimates are eleven-month averages that exclude October, when Current Population Survey data were not collected during the federal government shutdown; BLS cautions that they are therefore not strictly comparable to prior annual averages (BLS, 2026). The direction and magnitude of the disparity, however, are consistent with every annual release since comparable data collection began in 2008.

That consistency is the analytic problem. The period since 2008 has included the ADA Amendments Act, the Workforce Innovation and Opportunity Act (WIOA), the Olmstead implementation push, Medicaid home and community-based services expansion, and a sustained employer diversity apparatus. Interventions have proliferated. The aggregate outcome has not meaningfully moved. When a system absorbs two decades of reform without changing its output, the reasonable inference is that the reforms targeted something other than the mechanism generating the output.

Theoretical Framework

Disability as an institutional product

Medical sociology’s central contribution to disability scholarship is the insistence that impairment and disadvantage are analytically distinct. Link and Phelan’s (1995) fundamental cause framework holds that social conditions persist as determinants of outcomes because resources — money, knowledge, power, beneficial social connections — can be deployed flexibly across whatever proximate mechanisms happen to be operative. Interventions that address a proximate mechanism without redistributing the underlying resource are reproduced around.

Applied to vocational rehabilitation, this predicts precisely the pattern observed: service innovations accumulate, and the gap holds, because the resource that actually determines whether a disabled person converts eligibility into employment is not any particular service. It is the capacity to navigate a multi-institution administrative pathway — to document, to wait, to appeal, to re-present one’s case to successive gatekeepers. That capacity is unequally distributed, and no service innovation redistributes it.

Administrative burden as policy

Herd and Moynihan (2018) provide the mechanism. They demonstrate that the friction citizens encounter in claiming public benefits is not incidental to program design but is itself a form of policymaking, and they disaggregate it into three costs: learning costs (identifying that a program exists and what it requires), compliance costs (documentation, appointments, waiting, re-verification), and psychological costs (stigma, loss of autonomy, the stress of uncertain standing). Critically, they show that burdens are distributed regressively — they fall hardest on those with the fewest resources to absorb them.

The disability context intensifies each cost. Learning costs are elevated for people with cognitive and psychiatric disabilities. Compliance costs are elevated for people with mobility, communication, and fatigue-related conditions, and for anyone without reliable transportation or broadband. Psychological costs are elevated in a service system that requires repeated narration of deficit as the price of admission. Lipsky’s (2010) analysis of street-level bureaucracy completes the picture: counselors facing caseloads that exceed capacity necessarily ration through discretion, and discretionary rationing systematically advantages clients who present as tractable.

The clinical dimension

The psychological consequences are not peripheral to the employment question; they are causally upstream of it. De Dios Pérez et al.’s (2025) realist evaluation of VR for people with multiple sclerosis in the United Kingdom’s National Health Service found that embedding vocational support within health systems requires substantial cultural and organizational change, and that the decisive variable is whether people at risk of job loss are identified routinely rather than incidentally. Where identification depends on the individual recognizing their own eligibility and initiating contact, the people most affected by their condition are the least likely to initiate.

This is where clinical psychology and medical sociology converge rather than merely coexist. Delay and discontinuity are not simply inefficiencies that postpone a service; they are experiences that erode the self-efficacy on which vocational engagement depends. A system that reliably produces demoralization before it produces services has an outcome problem that no amount of downstream job placement can correct.

Systems Analysis: Four Mechanisms

1. Rationing by statute

When a state VR agency anticipates that it lacks sufficient fiscal or personnel resources to serve all eligible individuals, federal law requires it to implement an Order of Selection (OOS), assigning eligible individuals to priority categories by significance of disability and serving the most significantly disabled first (Rehabilitation Act of 1973 § 101(a)(5)(A); 34 C.F.R. § 361.36).

This is the most consequential and least discussed feature of the American VR system. Eligibility and service are formally decoupled. A person can be determined eligible for vocational rehabilitation and receive nothing, indefinitely, without any adverse determination having been made — because no category open to them is being served.

RSA’s Order of Selection listing identifies 16 VR agencies operating under an OOS as approved in their Program Year 2024–2027 WIOA State Plans, of which nine have all priority categories closed (RSA, n.d.-a). State-level activity has continued: Washington’s Division of Vocational Rehabilitation closed all priority of service categories on December 8, 2025 (Washington State Department of Social and Health Services, 2025), and the California Department of Rehabilitation announced an anticipated OOS implementation in spring 2026 (California Department of Rehabilitation, n.d.).

Sociologically, OOS is rationing that preserves the appearance of entitlement. The statutory right survives intact; only its exercise is suspended. This is administrative burden in its purest form — the psychological cost of indefinite waiting imposed without any decision the individual can appeal.

One caution against overreading: Ipsen and Stern (2020) found that rural residents are less likely to apply for VR services than urban residents, and that only a small portion of the cross-state variation in application rates is attributable to Order of Selection rules. OOS is a powerful mechanism, but it is not the whole of the access problem, and analysis that treats it as such will overpredict the effect of simply reopening categories.

2. The transition handoff

WIOA requires state VR agencies to reserve at least 15 percent of their federal VR allotment for pre-employment transition services (Pre-ETS) to students with disabilities, including students who are only potentially eligible (34 C.F.R. § 361.65; see Wisconsin Division of Vocational Rehabilitation, n.d.). The scale is substantial: RSA-911 data show state VR agencies provided Pre-ETS to 203,638 students in program year 2020 and 250,784 in program year 2021 (RSA, 2023).

The set-aside was designed to prevent exactly the discontinuity this article describes. In operation it has introduced a new seam. Pre-ETS sits at the intersection of three institutions — the state VR agency, the local educational agency operating under the Individuals with Disabilities Education Act, and the state intellectual and developmental disability agency — each with distinct eligibility criteria, data systems, timelines, and definitions of successful exit. No entity holds accountability for the trajectory as a whole.

The result is a transition period in which a student may be simultaneously served by three systems and continuously by none. For students with the most complex support needs, the window in which vocational identity and support scaffolding are most modifiable closes while the institutions are still determining whose responsibility it is.

3. A coordination mandate without coordination capacity

Section 511 of the Rehabilitation Act, implemented at 34 C.F.R. part 397, restricts entry into subminimum wage employment. Before a 14(c) certificate holder may pay subminimum wages to a youth age 24 or younger, the youth must document completion of pre-employment transition or transition services, a VR eligibility or ineligibility determination, and career counseling and information and referral. Individuals of any age employed at subminimum wage must receive career counseling and referral every six months during the first year and annually thereafter (Rehabilitation Act § 511(c)(2); 34 C.F.R. § 397.40(c); U.S. Department of Labor & RSA, n.d.).

Section 511 assigns the designated state unit a recurring obligation toward a population it does not serve, does not fund, and frequently cannot locate — people employed by third-party certificate holders. It is a coordination mandate laid over an infrastructure with no coordination capacity, and it depends on the cooperation of entities whose operating model it is designed to erode.

The federal trajectory here has reversed. The Department of Labor published a notice of proposed rulemaking on December 4, 2024, that would have ceased issuance of new 14(c) certificates and established a three-year phase-out for existing holders (89 Fed. Reg. 96466). That proposal was withdrawn on July 7, 2025, with the Department citing concerns regarding its statutory authority to unilaterally terminate certificate issuance (U.S. Department of Labor, 2025). Section 511’s procedural requirements remain in force, now without the substantive endpoint they were designed to anticipate — a durable illustration of how procedural coordination requirements survive the policy consensus that produced them.

4. Geography, race, and the distribution of service type

Fragmentation is not distributed randomly. Analyzing 2015 RSA-911 case service data, Ipsen et al. (2023) found that being Black or living at a long distance from a metropolitan area increased the probability of receiving agency-based services while lowering the probability of receiving purchased services. This is a structural finding, not an attitudinal one: where community rehabilitation provider networks are thin, agencies substitute internal capacity, and the service mix a person receives is determined by the density of the local provider market rather than by assessed need.

The international evidence corroborates the pattern at the program-design level. Mohamad et al.’s (2025) scoping review of vocational rehabilitation for people with acquired physical disabilities identified 16 distinct programs comprising 46 unique service components across vocational, physical, financial, social, and psychological domains — and found that no two programs were alike. A field this heterogeneous cannot generate portable evidence about what works, because there is no stable unit of intervention to evaluate.

Measurement infrastructure compounds the problem. Fleming et al. (2025) note that the RSA-911 captures whether a service was received, by whom, and when, but was not designed to support continuous quality improvement within agencies. RSA has proposed reducing the RSA-911 from 316 data elements to 255, a 19 percent reduction (RSA, n.d.-b). Whatever the administrative merits, a system that cannot observe where people exit cannot be held accountable for the exits.

Policy Implications

The analysis above supports five recommendations, ordered by the strength of the evidence behind them.

Shift coordination costs from the individual to the state. The single highest-value reform is structural rather than programmatic: consolidated intake, warm handoffs with documented receipt, and cross-agency data exchange, so that navigation is performed by institutions rather than by the person least resourced to perform it (Herd & Moynihan, 2018).

Treat Order of Selection waiting-list status as a reportable outcome. OOS currently functions as an accountability void — eligibility without service, generating no performance consequence. Requiring agencies to report time-in-category, attrition from waiting lists, and outcomes for individuals who exited without service would make the rationing visible.

Assign single-entity accountability for the transition trajectory. The 15 percent Pre-ETS reservation commits funding without assigning ownership. One entity should hold documented responsibility for each student’s trajectory from school exit through employment stabilization, regardless of which agency delivers a given service.

Standardize a core service set while permitting local expansion. This is precisely what Mohamad et al. (2025) recommend on the basis of their review, and it addresses both equity and evidentiary coherence.

Measure continuity, not only placement. Time-to-eligibility, time-to-IPE, service interruption, and 24-month employment retention capture whether the system functions as a pathway. Placement counts alone can improve while the pathway degrades.

Conclusion

The National Council on Disability’s 2025 progress report documents continuing barriers across health care, transportation, education, and employment — layered exclusion rather than a single bottleneck (NCD, 2025). That layering is the finding. The American disability service system is not failing because it lacks effective interventions; the evidence base for supported employment, customized employment, benefits counseling, and assistive technology is substantial and growing. It is failing because the interventions are lodged in institutions that do not share accountability for the person moving among them, and because the labor of moving among them has been assigned to the person least able to perform it.

This is not an efficiency argument. Administrative burden is regressive by construction, and structural discontinuity is therefore a mechanism of stratification, not merely of waste. Treating it as an operational nuisance rather than as a distributive policy choice is the analytical error that has allowed two decades of reform to leave the employment gap where it found it.

References

Bureau of Labor Statistics. (2026). People with a disability: Labor force characteristics — 2025. U.S. Department of Labor. https://www.bls.gov/news.release/disabl.htm

California Department of Rehabilitation. (n.d.). Order of selection for vocational rehabilitation services. https://www.dor.ca.gov/Home/OrderOfSelection

De Dios Pérez, B., Booth, V., das Nair, R., Evangelou, N., Hassard, J., Ford, H. L., Newsome, I., & Radford, K. (2025). Vocational rehabilitation for people with multiple sclerosis in the National Health Service of the United Kingdom: A realist evaluation. PLOS ONE, 20(2), Article e0319287. https://doi.org/10.1371/journal.pone.0319287

Employment of Workers With Disabilities Under Section 14(c) of the Fair Labor Standards Act, 89 Fed. Reg. 96466 (proposed Dec. 4, 2024) (to be codified at 29 C.F.R. pt. 525).

Fleming, A. R., Phillips, B. N., & Lockett, M. (2025). Continuous improvement in state vocational rehabilitation: A demonstration and call to action. Journal of Vocational Rehabilitation. https://doi.org/10.1177/10522263241300870

Herd, P., & Moynihan, D. P. (2018). Administrative burden: Policymaking by other means. Russell Sage Foundation. https://www.russellsage.org/publications/book/administrative-burden

Ipsen, C., Jain, K., & Stern, S. (2023). Vocational rehabilitation service receipt, service expenditures, and ruralness. Journal of Vocational Rehabilitation, 59(3), 251–261. https://doi.org/10.3233/JVR-230043

Ipsen, C., & Stern, S. (2020). The effect of ruralness on vocational rehabilitation applications. Journal of Vocational Rehabilitation, 53(1), 89–104. https://doi.org/10.3233/JVR-201088

Link, B. G., & Phelan, J. (1995). Social conditions as fundamental causes of disease. Journal of Health and Social Behavior, 35(Extra Issue), 80–94. https://doi.org/10.2307/2626958

Lipsky, M. (2010). Street-level bureaucracy: Dilemmas of the individual in public services (30th anniversary ed.). Russell Sage Foundation.

Mohamad, N. B. Z., Tan, G. Y. Q., Lim, S. A. Y., Joe, S., & Asano, M. (2025). A scoping review of vocational rehabilitation programs for people with acquired physical disabilities. Journal of Occupational Rehabilitation. https://doi.org/10.1007/s10926-025-10322-2

National Council on Disability. (2025). National disability policy: A progress report, 2025. https://www.ncd.gov/report/national-disability-policy-a-progress-report-2025/

Olmstead v. L.C. ex rel. Zimring, 527 U.S. 581 (1999).

Rehabilitation Act of 1973, 29 U.S.C. § 701 et seq. (as amended by the Workforce Innovation and Opportunity Act of 2014).

Rehabilitation Services Administration. (n.d.-a). Order of selection information. U.S. Department of Education. https://rsa.ed.gov/about/programs/vocational-rehabilitation-state-grants/order-of-selection-information

Rehabilitation Services Administration. (n.d.-b). What’s new in RSA. U.S. Department of Education. https://rsa.ed.gov/whats-new

Rehabilitation Services Administration. (2023). Report for fiscal years 2021–2022. U.S. Department of Education. https://rsa.ed.gov/sites/default/files/publications/ARC%20to%20Congress/RSA-Annual-Report-for-FFYs-2021-2022.pdf

State Vocational Rehabilitation Services Program; State Supported Employment Services Program; Limitations on Use of Subminimum Wage, 81 Fed. Reg. 55630 (Aug. 19, 2016) (codified at 34 C.F.R. pts. 361, 363, 397). https://www.federalregister.gov/documents/2016/08/19/2016-15980/state-vocational-rehabilitation-services-program-state-supported-employment-services-program

U.S. Department of Labor. (2025). Employment of workers with disabilities under section 14(c) of the Fair Labor Standards Act; Withdrawal. Federal Register. https://www.federalregister.gov/documents/2025/07/07/2025-12534/employment-of-workers-with-disabilities-under-section-14c-of-the-fair-labor-standards-act-withdrawal

U.S. Department of Labor, Wage and Hour Division, & Rehabilitation Services Administration. (n.d.). Questions and answers: Section 511 of the Rehabilitation Act. https://www.dol.gov/sites/dolgov/files/WHD/flsa/WHD-RSA_QA.pdf

Washington State Department of Social and Health Services. (2025). Order of selection information. Division of Vocational Rehabilitation. https://www.dshs.wa.gov/dvr/order-selection-information

Wisconsin Division of Vocational Rehabilitation. (n.d.). Policy guide for pre-employment transition services. Wisconsin Department of Workforce Development. https://dwd.wisconsin.gov/dvr/policy-guidance/edu-train/pre-ets-guide.htm