February 22, 2024
Today, the Bureau of Labor Statistics (BLS) released its annual estimates of the labor force characteristics of the disabled population in the United States. To complement this, we are releasing the following preview of our forthcoming chartbook on disability and economic justice, an update of the 2022 chartbook authored by Hayley Brown, Julie Yixia Cai, and Shawn Fremstad. Analyzing different data than the BLS, we hope this preview will add to the important ongoing conversation on disability and work. Like the previous edition, the new chartbook will contain a variety of tables and figures that help to tell the economic story of disability justice (or lack thereof) in the United States. The chartbook will cover disability prevalence, disparities in employment and earnings, disparities in income poverty and material hardship, and more.
Disability is a complex and multidimensional concept. Definitions vary substantially by context; the definition that the Social Security Administration employs to determine who qualifies for benefits is different than the one established by the Americans with Disabilities Act, and both of these differ from definitions used by academics and in various surveys. Even across US government surveys, the definition of disability is far from settled, and the questions used to assess disability vary (see Appendix Table). As a result, there can be big differences in the estimated size of the disabled population between surveys and even within surveys, depending on the definition used.
As Figure 1 shows, depending on the survey, the question set, and the way the questions are interpreted, annual government surveys produce estimates of disability prevalence among US adults that range from less than 10 percent to over 40 percent. Many national surveys in the US use the Standard Short Set from the Department of Health and Human Services (HHS), which consists of six yes or no questions about different types of difficulty. Other surveys use a different set of questions referred to as the Washington Group Short Set (WG-SS), which includes six questions about type and degree of difficulty. This set of questions can yield different prevalence estimates depending on what degree of difficulty one sets as the threshold for disability.
Figure 1
The first two columns in Figure 1 reflect different potential definitions of disability within the National Health Interview Survey (NHIS). The NHIS uses the WG-SS; because this question set allows for more than one possible estimate of disability prevalence, two of these potential estimates are shown in the figure. The first column (NHIS 1) includes only those who reported "a lot of difficulty" or "can't do at all", while the second column (NHIS 2) additionally includes those who said they had “some difficulty”. For the remaining surveys — the Behavioral Risk Factor Surveillance System (BRSFF), the American Community Survey (ACS), and the Current Population Survey Annual Social and Economic Supplement (CPS ASEC) — determination of disability is based on answers to the Standard Short Set. The Standard Short Set is also used in the CPS Basic Monthly survey, the aggregated results from which are featured in today’s release from the BLS.
For disability overall, the National Health Interview Survey yields both the highest and the lowest prevalence depending on the definition used. However, it’s also worth noting the variability in prevalence among the three surveys that use the same set of questions. This suggests that other aspects of survey design may significantly influence the extent to which surveys include disabled people and accurately report disability status among those they include. Research indicates that both the Standard Short Set and the WG-SS fail to capture large swathes of the disabled population compared to other, more comprehensive sets of questions, with the more restrictive interpretation of the WG-SS missing the most.
The Bureau of Labor Statistics release uses information from the monthly Current Population Survey (CPS) aggregated over the year. The latest release notes a significant employment gap between people with and without disabilities, one that has persisted since the CPS began asking about disability in late 2008. However, people with disabilities are not a monolith, and there is considerable variation in employment and labor force status by type of disability (Figure 2). Those with more than one type of disability are also less likely to be employed or to participate in the labor force (Figure 3).
Figure 2
Figure 3
Even among those who are employed, there is a substantial pay gap between those with and without disabilities (Figure 4). This pay gap remains even for those who work full-time for the entire year.
Figure 4
One potential remedy for the pay disparity between disabled and nondisabled workers is union representation, which is especially beneficial for disabled workers. Even after controlling for other factors, union representation provides enough of a wage boost that disabled workers with union representation actually enjoy a wage premium relative to nonunion workers without union representation (Figure 5). The disability wage gap within unionized workplaces is also much smaller than it is in nonunion work environments.
Figure 5
Unfortunately, union membership currently sits at a record low (though the rate for workers with disabilities did tick up slightly in 2023). Rebuilding the labor movement should be considered a key part of the fight for economic justice for those with disabilities, and it must be done in a way that includes and lifts up disabled people.
Appendix Table
