Many disabled people are not included in official U.S. data. This is because there is “No Box to Check” to indicate their particular disability on surveys from the U.S. Census Bureau and other federal agencies. The questions used to identify people with disabilities are missing millions.
Two question sets are most often used to assess disability in U.S. federal surveys: the American Community Survey Six (ACS-6) or the Washington Group Short Set (WG-SS). Combined, they are used in at least 17 U.S. federal surveys, so these disability question sets have an enormous effect on people’s lives. For instance, they appear in the Census Bureau’s surveys, which provide information that “generates data that help inform how trillions of dollars in federal funds are distributed each year.” They are also used in surveys fielded by the Centers for Disease Control and Prevention that help monitor health and health care in the U.S. population.
In both sets, questions focus on respondents’ limitations on specific tasks or activities, such as difficulty seeing even while wearing glasses.
Because of this approach, both sets miss large groups of disabled people. They miss between 23% and 59% of people with mental health or psychiatric disabilities, between 13% and 33% of people with intellectual and developmental disabilities, and between 32% and 53% of people with chronic health conditions, such as long Covid.
We recently led efforts to push back against a proposal for the U.S. Census Bureau to change how disability is measured in the ACS. Had it been enacted, the Census Bureau would have changed from using the ACS-6 questions to the WG-SS questions to measure disability in the American Community Survey. This proposed change to the WG-SS questions would have underestimated the prevalence of disability in America by over 40%. Advocacy efforts from the disability community successfully halted this change. But this issue is far from being resolved.
We have been frequently asked which set of disability questions, the ACS-6 or WG-SS, is best. Which set should be used? These questions echo a long debate among researchers, advocates, and the federal statistical community about which of these measures are superior. The honest answer is that both are bad options. Both exclude many disabled people. The federal government must invest in providing something better.
The first step in moving disability measurement forward is recognizing that defining disability solely by someone’s functioning is inadequate. While measures of functioning are important to understand the limitations certain disabled people experience and could identify disabled people who may benefit from specific programs or benefits, these surveys still fail to capture many disabled people. Using this data to determine national estimates of disability is akin to using data on languages used in the home to determine the national race and ethnicity estimates.
Second, disability must be viewed as a demographic variable. As such, it should be collected everywhere we collect other demographics such as age, gender, and race-ethnicity.
There is a federal precedent making this case. The Affordable Care Act (ACA), passed in 2010, set data standards for collecting demographic data in federal surveys, including race, ethnicity, sex, primary language, and disability status. The ACA further required the collection of demographic data, including disability demographic data, in “any federally conducted or supported health care or public health program, activity or survey.”
The ACA is not the first time disability was conceptualized as a demographic by the federal government. As early as 2002, staff from the Census Bureau and Centers for Disease Control and Prevention (CDC) stated, “Disability is not an outcome, it’s an input. … Disability is a demographic variable — just like age, sex, racial ethnicity, socioeconomic status — and it needs to be seen that way, as opposed to a negative health outcome.” Collecting disability data as we collect other demographic data is essential to better understanding the U.S. population, improving the allocation of resources across a wide range of programs and areas, and removing barriers to economic opportunities from education and employment to housing and small business development.
Third, we must clarify how disability is defined. The term “disability” is used to describe different things to different people in different settings, making the development of a standard measure challenging. Being disabled is an integral part of some people’s identity; for others, a medical diagnosis, a natural part of aging, or even a fate they think is worse than death. Simply measuring functional limitations does not get at the diversity of these experiences.
The fact that there are so many different ways people define their disability will make the task of improving measurement challenging. But variation in how people define their disability is not reason to accept the continuation of inequitable disability questions. It is the reason we need to invest in and support collaborations with disabled people to understand these differences and create inclusive disability survey questions.
As researchers with disabilities ourselves, we understand the hard work ahead and what is at stake if action isn’t taken. Disability data is essential for supporting evidence-based policies, allocating funding and resources, and better understanding the needs of all our communities.
However, our research has been stymied by disability data limitations. Because we cannot determine accurate national prevalence estimates of people with disabilities in the United States, it is challenging to understand and address the barriers this population faces across all sectors of life, including education, employment, health and health care, housing, transportation, and food access. Without improved measures of disability and collection of this data as a core demographic, we can’t answer simple questions: How many teachers have disabilities? What percentage of public transit riders are disabled? What is the fall rate of patients with disabilities in nursing homes or hospitals?
We can no longer consider disability data a niche issue. Disability is part of all of our lives. People with disabilities are among our families, friends, and co-workers. We, ourselves, may be disabled now or in the future. Our national data must reflect the broad experiences and communities across America and must include people with disabilities.
Bonnielin Swenor is the founding director of the Johns Hopkins Disability Health Research Center and the endowed professor of disability health and justice. Her work uses data-driven approaches to advance equity for people with disabilities. Scott Landes is an associate professor of sociology and O’Hanley faculty scholar at Syracuse University. His research focuses on health trends among disabled adults and disability measurement. Jean Hall is director of the Institute for Health and Disability Policy Studies at the University of Kansas, where she conducts research focusing on health, health policy, and access to care for people with disabilities, as well as on disability measurement.
