One of the purposes of the surveys overseen by the U.S. Census Bureau is to help the government “decide how to distribute funds and assistance to states and localities,” the organization’s website states. The data gathered helps determine the amount of money and resources to dedicate to various populations on local, state, and national levels.
Logically, then, you would think other federal agencies would be aligned on how they source their population data. But what happens when there is a discrepancy in U.S. agencies’ methods of counting a particular population? Furthermore, how is the community in question impacted when both methods are flawed?
This is a dilemma the disability community faces today, shares Dr. Bonnie Swenor, the founder and director of the Johns Hopkins Disability Health Research Center (JHDRC).
Dr. Swenor explains that the Centers for Disease Control and Prevention (CDC) recently released updated data that shows that the prevalence of disability in the U.S. has increased by about 28%, as was reported in a survey called the Behavioral Risk Factor Surveillance Survey (BRFSS).
The survey uses a set of six questions known as the American Community Survey (ACS-6) which, as Dr. Swenor shares, is a set of six “yes” or “no” questions that ask people whether they have difficulty seeing, hearing, walking, or running errands; difficulty with memory, concentration or decision-making; or difficulty with self-care.
Dr. Swenor says that she and many of her colleagues believe that of the options available, the ACS-6 is the preferable set of questions to use to gather information about people’s disabilities. However, she acknowledges that they’re flawed, particularly because they don’t include a question about the respondent’s ability to communicate.
The alternative set of questions that some people within the CDC are advocating a switch to is called the Washington Group Short Set (WG-SS). Both the CDC and the National Center for Health Statistics (NCHS) use the ACS-6 set on some surveys and the WG-SS for others.
The WG-SS uses a scale asking respondents to report if they have “a lot of difficulty,” “a moderate amount of difficulty,” “a little bit of difficulty” or are “unable” to see, hear, walk, perform self-care, communicate, or perform tasks related to memory and concentration.
People who say they are unable to perform a function or have a lot of difficulty performing a function are classified as having a disability. People who say they only have moderate or a little bit of difficulty with a function are classified as not having a disability.
This method is problematic for multiple reasons, including that respondents aren’t informed that a data analyst determines if they should be classified as disabled, no matter how they self-identify. So, by answering that they have “some” difficulty doing a task, they are potentially putting important decisions about the resources they’ll receive in the hands of strangers.
Per Dr. Swenor, framing questions about functioning as a scale creates a “hierarchy” in the disability community and makes the data easy to weaponize.
The journal Health Affairs (citing the National Survey on Health and Disability [NSHD], which looks at both surveys to gather information about disability),notes that the WG-SS failed to identify 43% of its respondents as disabled while the ACS-6 failed to identify 20% of its respondents.
Additionally, both sets of questions fail to account for people with psychiatric conditions, chronic illnesses, and learning disabilities. So, with two flawed surveys at the forefront of disability data collection, what’s the next step?
“I think we’ve got to go back to the drawing board and we’ve got to do it with the disability community,” said Dr. Swenor, who works with her team at the JHDRC to advocate to include more disabled researchers in the field of disability data collection.
Oregon currently has the most comprehensive approach to collecting disability data. They use the ACS-6 questions and the communication question from the WG-SS in a “yes” or “no” format. Plus, questions have been added about learning disabilities and psychiatric disabilities that other surveys miss.
If a respondent answers yes to one of nine questions, they’re asked what age the disability began. Two additional questions ask if the respondent identifies as disabled and about their access and accommodation needs. But what do we do about the other 49 states and comprehensive national data?
“We need to start with a new, conceptual framework of disability that views disability as a demographic group of people,” Dr. Swenor says. The process will take time, so until then, logic dictates the best course of action until a new framework is created would be to stick with the survey that has the lowest margin of error, which is the ACS-6.
“We’ve got to make use of what we have now, but fully understand and share in the best way possible the limitations. Who are we missing and what does that mean?” asked Dr. Swenor. She added that in the interim, we need to understand the limitations of the ACS-6 and try to adjust accordingly. Dr. Swenor notes that the BRFSS has tried to do that recently, by adding a question about long COVID, which is a step in the right direction.
But why exactly does accurate data about the disability community matter so much? The answer is simple — money.
“If we are undercounting people with disabilities, you can make the leap to saying, ‘We are not properly resourcing the programs and activities and services that are focused on supporting people with disabilities,’” Dr. Swenor says.
There’s a popular saying within the disability community: “Nothing about us, without us.” In the field of data collection, not only is information about us collected without us at the helm, but the standards used to determine who should be counted in the disabled population are flawed, resulting in data that underestimates the size of the community.
The goal of professionals like Dr. Swenor is to get the most accurate estimate of the prevalence of disability in the U.S. possible.
But although there is evidence that the country is currently failing to do that by undercounting, Dr. Swenor says that when she points this out to some people, they tell her that more accurate, comprehensive data would make the percentage of the disability population in the U.S. too high.
“It really is these deep, theoretical conversations around data and power, and who gets to decide who has a disability, and who gets to say that in the data,” said Dr. Swenor. “There’s been comments like, ‘That’s too much disability. That estimate is too high.’ And, I think I’ve literally said on phone calls, ‘Are you afraid we’re going to take over the country? What is the concern exactly here?’”