Autor and Duggan argue that Social Security Disability Insurance's (SSDI) binary eligibility standard — total inability to engage in substantial gainful activity — is systematically mismatched to the heterogeneous residual work capacity of disabled workers, and that this structure, combined with a 12-month average adjudication delay and no early-intervention mechanism, transforms short-term work limitations into permanent exit from the labor force. They document that the SSDI caseload doubled (2.3% to 4.6% of adults 25–64) between 1989 and 2009 despite flat disability prevalence, driven by rising replacement rates and program expansion rather than worsening health. They propose universal mandatory private disability insurance (PDI) as a 24-month front end to SSDI, providing early vocational rehabilitation and wage replacement while the worker either recovers or transitions to permanent disability.
"The DI program is structured as though applicants either can or cannot work when, in fact, most have some residual work capacity that diminishes with program tenure."
"Unlike workers' compensation, DI provides no incentive whatsoever for employers or workers to engage in the early intervention and job accommodation activities that could prevent a short-term disability from becoming a long-term or permanent one."
The paper's central insight — that SSDI's binary structure is mismatched to the heterogeneity of disability — is compelling and well-supported by the flat prevalence/rising-rolls decoupling. The Dutch reform comparison is the strongest empirical anchor: a near-natural experiment showing that employer-side liability and early intervention can cut inflows by 60–80% over four years. The cost-benefit math (~$150–$250/year breaks even if 1 in 11 diverted) is plausible given the long-term disability (LTD) market data documented in Autor, Duggan, and Gruber (2014). The paper is primarily a policy advocacy document (co-published by Brookings Hamilton Project and Center for American Progress), not an identification exercise, so the causal evidence is indirect. The Dutch reform is an imperfect analogy: the Netherlands' broader sickness-insurance system, universal health coverage, and labor market institutions differ substantially from the U.S. The paper predates the SSDI natural experiments (Maestas, Mullen, and Strand 2013; French and Song 2014) that would later confirm large employment effects, providing stronger causal foundations for the welfare case against SSDI expansion.