A comparative analysis of self-identification and functional measures of disability

Background: Disability is typically measured in surveys using functional limitation questions rather than asking respondents to self-identify as having a disability. Little is known about the characteristics of those who self-identify with a disability and how they compare with those identified via functional limitation questions. Objective: To compare the prevalence and characteristics of people with disabilities measured by both functional and self-identification measures, and to assess the overlap between these populations. Methods: Using nationally representative survey data from 2023, we conduct bivariate comparisons of demographic, health, functional characteristics and financial needs between populations captured by self-identification question, the Washington Group Short Set on Functioning (WG-SS) and the American Community Survey Six (ACS-6) Questions. We estimate a multivariable regression to explore predictors of self-identification. Results: Adding a self-identification question doubles the population with disabilities relative to measuring disability with the WG-SS alone, and increases the population by 30 % relative to the ACS-6 alone. People who self identify are less likely to be female or Hispanic/Latinx, are in worse physical health and more likely to be LGBTQIA + compared to those identified only by functional measures. The group identified by both the self-identification and either functioning question set are in the worst health and are more likely to participate in disability programs. Conclusion: Including a self-identification question in addition to functioning questions expands the prevalence of disability to varying degrees depending on which functional questions are used, but use of functioning and self-identification questions together enhances identification of the subgroup with highest needs.