Summary
Using National Health Interview Survey (NHIS) data from 1984 to 2000, Lakdawalla, Bhattacharya, and Goldman document a divergence in U.S. disability trends: while the elderly were becoming less disabled, reported disability rose significantly among Americans ages 18–59, especially ages 30–49. The trend holds after adjusting for sex, race, Hispanic origin, education, and employment composition, and it runs across all demographic subgroups. Three candidate explanations are evaluated — obesity, medical technology advances, and expanding disability insurance (DI) incentives — with obesity the best partial explanation, particularly for ages 50–59.
Key Claims
- Main finding: Routine-needs disability rates rose 15–66% for ages 18–59 from 1984 to 1996. Ages 30–39 experienced the sharpest increase (>50%); ages 60–69 actually declined (−28% in routine-needs disability). This young/old divergence holds throughout the sample period.
- Two disability measures: "Personal-care disability" (unable to attend to personal care needs) and "routine-needs disability" (broader: also includes those limited in other routine tasks). Both show rising trends for ages 18–59; personal-care growth was mostly confined to ages 18–39, while routine-needs growth extended to 40–59.
- Not compositional: Adjusting for sex, race, Hispanic origin, educational attainment, and employment status increases the estimated disability growth — demographics actually dampened observed rates. Disability rose universally across whites and nonwhites, all education groups, and people both inside and outside the labor force.
- Obesity as partial explanation: Rising obesity explains approximately 50% of disability growth for ages 18–29, ~25% for ages 30–39, and ~10% for ages 40–49. For ages 50–59, all of the disability increase was among the obese; the non-obese were flat. Key weakness: elderly obesity also rose, yet elderly disability fell.
- Medical technology as optimistic interpretation: Lifesaving treatments may keep chronically frail people alive who would otherwise have died (survival-disability tradeoff). Hard to test, but consistent with the data.
- DI incentive explanation weakly supported: The Autor-Duggan rising replacement rate for less-skilled workers is a candidate, but its prediction — disability growth concentrated among the less-educated and non-workers — is not confirmed. Disability rose equally across all education groups and within the labor force.
- Forward projection: Rising young disability projected to produce a nursing home population 10–25% larger than baseline and Medicare expenditures 10–15% higher.
Concepts Introduced or Extended
- Compression of Morbidity — documents the mirror image of elderly compression: young disability expansion on the same NHIS data source; the young/old divergence is real and survives compositional adjustment
- Disability Prevalence Measurement — introduces the NHIS personal-care/routine-needs disability categories used throughout this paper; distinct from Current Population Survey (CPS) work-activity-limitation or American Community Survey (ACS) six-question sequences
Entities Mentioned
Quotes
"Even as the elderly have become less disabled, reported disability has risen for younger Americans, especially those ages 30–49."
"Whatever its sources, rising disability among the young could have adverse consequences for public programs such as disability insurance, Medicare, and Medicaid."
My Take
A compact but important descriptive paper that establishes the young-disability expansion as a cross-demographic empirical fact, not an artifact of changing population composition. Its value in this wiki is primarily as the factual backdrop for the DI growth literature: while the compression-of-morbidity story reassures about the elderly, the young are on a different trajectory. The obesity mechanism is credible but partial; the paper is honest that no single explanation fits perfectly. The medical-technology explanation (lifesaving treatments sustaining frail survivors) is underappreciated in the subsequent DI literature but aligns with the "disability insurance chronology" showing that DI caseloads shifted toward more severe impairments over the 1980s–1990s. Data end in 2000, limiting its relevance to the post-2010 DI decline, but for documenting the growth phase it is a clean reference.