Rutledge, Wu, Guan, and Trenkamp use Survey of Income and Program Participation (SIPP) panels (1990–2008) linked to Social Security Administration (SSA) administrative records (831 File) to measure the pre-application health status of ~3,250 Social Security Disability Insurance (SSDI)/Supplemental Security Income (SSI) applicants from 1989 to 2013, with health measured 1–3 years before the application date. They find essentially flat trends in most health measures over the period, with one statistically significant decline (work-preventing conditions). The conclusion is that the growth in Disability Insurance (DI) rolls from 1990–2013 cannot be attributed to a deteriorating applicant health pool — economic and demographic factors are the more likely explanation.
"The results suggest that the underlying health of disability applicants has not changed in any perceptible way over the last two decades."
This paper closes one of the key alternative explanations for DI growth: that America simply got sicker. The pre-application health measurement (1–3 years before filing) is a methodological strength — it sidesteps both strategic health reporting and the mechanical fact that awarded applicants must be sick by SSA's criteria. The allowed/denied divergence is interesting: it suggests the examiner screening process has become relatively better at selecting the genuinely sick while the marginal (healthier) applicants piled in during the growth years. The main limitation is sample size (~3,250) and the exclusion of SIPP panels before 1990, which limits precision for the earliest cohorts.