This post was originally published here (Urban Institute Research)
In “Generations, disabled,” Washington Post reporter Terrence McCoy tells the story of a three-generation family in Missouri who receives “federal disability checks.”
It’s classic shoe-leather reporting, describing the challenges of family members facing assorted illnesses and disabilities. But the article fails to tell an accurate story about health, disabilities, poverty, and the government programs that support disabled people.
The phrase “on disability” appears in the article 10 times, but it’s never defined. In the United States, several programs support people with disabilities, including the Social Security Disability Insurance Program (DI), Supplemental Security Income (SSI), Medicaid, and workers’ compensation.
Some of those programs, like SSI, are means-tested, meaning that the person (or family) needs to have a disability and very low income. Other programs, such as DI, are work-tested, meaning that an applicant must have a disability and must have worked for a certain amount of time. Without at least discussing these different programs and how they work, the article bundles all disability programs into a single, mysterious bucket.
The story also presents incomplete supporting data without context. A sidebar graph shows the “percent of households with a disabled child and no disabled adults or one or more disabled adults,” as reported by the American Community Survey (ACS). These data are not about families receiving DI or SSI benefits, but about the prevalence of disability. That is an important distinction and one that does not come through in the text.
Without further description, placing the graph next to the article suggests disability and federal disability payments (whatever those are) are the same thing. The casual reader might believe these data are about DI benefit receipt.
It’s difficult to tabulate DI receipt from ACS data because we don’t always know what program people are referring to when they estimate income from “disability benefits.” The Census Bureau asks about Social Security income, which could include retirement benefits, spousal benefits, or disability benefits. Census also asks about health insurance through Medicaid and other sources, as well as retirement income, which could come from a previous job.
Recent research shows growing concerns about the quality of survey data generally. McCoy relies on ACS questions about disability, not about disability income receipt. Thus, these ACS data suggest that in families with disabilities in several generations, those disabilities may have other underlying causes, such as genetics, exposure to environmental hazards, and poor local health care facilities. Poor health may be driven by factors other than a family’s effort to rely on, or abuse, a public program like DI, as McCoy seems to suggest.