Jane Tavares, PhD | Marc Cohen, PhD | Sara Rosenbaum, JD | Alison Barkoff, JD
June 2025
In a new post, Kevin Corinth of the American Enterprise Institute claimed that between 40 and 56 percent of nondisabled adults without minor children and who are not in school don’t work, and furthermore, that they spend over 180 hours per month “relaxing.” The implication: these individuals could readily “reallocate their time” to meet proposed 80-hour monthly Medicaid work requirements. His analysis contains a fundamental problem, however—a serious underestimation of disability among the population he claims to analyze.
To support his claim, the author links data from two national surveys – the American Time Use Survey (ATUS) and the Current Population Survey Annual Social and Economic Supplement (ASEC). By merging these files, he asserts, it is possible to know how nondisabled working-age Medicaid beneficiaries without minor children and who are not in school spend their time.
But the article contains a basic flaw that makes its findings and conclusions virtually useless – how it defines "disability." Obviously, the meaning of disability is critical to his analysis, which purports to examine the behavior of working-age adults without disabilities. Unlike more comprehensive measures of disability among an adult population, the ASEC data set he uses narrowly defines disability by whether an individual receives Supplemental Security Income (SSI) or has a health condition that prevents their ability to work.
For two reasons, this results in a gross undercount of disability among a low-income adult population who experience a sufficient level of activity limitation to impair their ability to work.
First, by law, SSI benefits are reserved only for the most deeply impoverished adults with severe disabilities. Nationally, only about 14% of low-income individuals with disabling conditions receive SSI—meaning millions of Medicaid beneficiaries with serious functional limitations are misclassified as “non-disabled.”
Second is the definition of disability itself. In order to constrain its reach, the SSI program uses a definition of disability so severe that work must be virtually impossible for an individual to qualify (although some people who do receive SSI may be able to work in limited amounts in substantially supportive environments). By contrast, many adults (especially older adults) with serious health conditions still may not qualify for SSI because they are judged able to do at least a limited amount of work. For example, a 54-year-old woman with a serious heart condition may be able to work a few hours each week and therefore not be considered disabled for purposes of SSI; yet she may be seriously limited in how many hours she can work and may need a good deal of rest. But ASEC data would consider her an able-bodied adult who is “relaxing”. Using her “leisure time” to justify a work requirement grossly misrepresents her reality.
Any meaningful policy analysis must be grounded in comprehensive, nuanced data that reflects the complex health realities of the Medicaid population. This AEI article fails that basic standard, stripping any value from the information it purports to produce —and in doing so, distorts the truth in ways that could do real harm.