Earlier this month, The Washington Post ran a front-page story about Social Security disability benefits in rural counties, followed this past Sunday by an editorial calling for a wholesale restructuring of Social Security Disability Insurance.Often called SSDI, this is the plank of Social Security that replaces some of your lost wages if you become disabled before reaching retirement age. Several SSDI experts, including our colleague Rebecca Vallas, as well as Kathleen Romig of the Center on Budget and Policy Priorities and Dean Baker of the Center on Economic Policy Research, published responses explaining what the Post missed in their reporting. But it turns out the article’s problems go even deeper than they thought. Not only does the Post’s reporting paint a misleading picture about SSDI, but the data analysis they published is just plain wrong.
The Post’s central assertion—flanked by an interactive map—was that as many as one-third of working-age adults in rural communities are living on monthly disability checks. But the data analysis supporting this argument doesn’t hold up.
Get TalkPoverty In Your Inbox
In a sidebar to the article, the Post says they used publicly available county-level data from the Social Security Administration (SSA) to count “every working-age person who receives benefits through the Supplemental Security Income (SSI) program, the Social Security Disability Insurance (SSDI) program or both.” But the Social Security Administration doesn’t publish the data needed for that calculation. In an email response to our request for these data, the SSA confirmed that these data are “not readily available.”
The Center for American Progress also reached out to the Post to ask about their data. The Post confirmed in an email exchange that they did indeed rely on publicly available data, and identified the specific reports, tables, and figures they used.
We tried to replicate their analysis, and here’s why their numbers are flat-out wrong. (Warning: We are about to dive head-first into the weeds.)
The analysis overcounts working-age people receiving disability benefits by nearly 500,000. The SSA doesn’t publish county-level data on SSDI beneficiaries in the age range the Post defines as “working age” (18 to 64). SSA’s OASDI Beneficiaries by State and County report does provide county-level data on SSDI beneficiaries (Table 4), including disabled worker beneficiaries. However, of the 8,909,430 disabled worker SSDI beneficiaries whom the table breaks down by county, 472,080—or about 5 percent—are age 65 or older. Including these older disabled workers would inflate the share of working-age people with disabilities.
It overcounts “disabled adult children” by about 750,000. About 1 million SSDI beneficiaries are disabled adult children (DACs)—people whose disability onset occurred before age 22 and who are insured for SSDI benefits based on a parent’s work record. Since the Post claims to count working-age people receiving SSDI, SSI, or both, they need to include working-age DACs. But—contrary to the Post’s data sidebar—there are no data available on working-age DACs at the county level.
The same SSA table from above does provide county-level data on one group of “children” receiving SSDI—totaling 1,755,276 in 2015. The problem is, these children aren’t disabled adults—they’re actually the offspring of disabled workers. Most are under age 18, and most are not disabled. Not only does erroneously using these data mean including minors without disabilities, it also inflates the number of DACs by about three-quarters of a million, since the total number of DACs aged 18-64 is 977,776. What’s more, offspring of disabled workers and DACs are likely differently distributed across counties, creating problems in county-level comparisons.
It can’t accurately adjust for double-counting the 1.3 million working-age people who receive both SSDI and SSI (a.k.a. “concurrent beneficiaries”). About 1.3 million working-age Americans receive a small amount in benefits from both SSDI and SSI—generally people with very low incomes and limited resources. To avoid double-counting these folks, the Post would need county-level figures on concurrent beneficiaries. But here they run into another problem: SSA doesn’t publish county-level data on working-age concurrent beneficiaries. The Social Security Administration does provide the number of people receiving both SSI and Social Security benefits of any type (Table 3), but that figure also includes people receiving any other kind of Social Security benefit (like survivor or retirement benefits). What’s more, they also include concurrent beneficiaries who are children and adults 65 and older. Both of these issues make it impossible to calculate for working-age beneficiaries receiving both SSDI and SSI at the county level. So these county-level figures can’t give the Post what they need to accurately mitigate their double-counting problem.
It’s missing data for a whopping 106 counties. Mostly because of small population size, SSA doesn’t publish county-level data on SSI beneficiaries for 106 counties. This would be problematic for any county-level analysis. But it’s especially notable given that the Post’s article focuses on rural counties—as some 97 of the counties with missing data are rural. It’s unclear how the Post treats these counties in their analysis.
This might seem like a lot of trouble to go through to explain two inaccurate newspaper articles. But the thing is, misleading media reports have consequences—particularly in political climates like the one we’re living in right now. Just this week, White House budget director Mick Mulvaney once again opened the door to cutting Social Security Disability Insurance, despite President Trump’s pledge not to cut Social Security. Misleading media reports based on inaccurate data analysis risk giving Mulvaney and others cover to slash critical programs like SSDI.
Media covering this important program should get their facts straight before going to press.