The U.S. Census Bureau on Tuesday halted plans to change how it asks people about disabilities after facing a growing backlash.

Advocates for disabled people had argued that proposed changes to disability questions on the bureau's American Community Survey would artificially reduce their numbers by more than 40%, limiting the ability of some to get vital resources for housing, schools or program benefits. They also argued that they weren't properly consulted on such a major overhaul.

“Good news. Good news. Good news,” said Scott Landes, an associate professor of sociology at Syracuse University, who is visually impaired. “They got the message that we need to engage.”

The bureau plans to meet with advocates in the disability community and determine what changes to the questions are needed to better capture the range of disabilities while keeping the current questions about disability on the 2025 American Community Survey, said Census Bureau Director Robert Santos.

“We will continue our work with stakeholders and the public to better understand data needs on disability and assess which, if any, revisions are needed across the federal statistical system to better address those needs,” Santos said in a blog post.

The American Community Survey is the most comprehensive survey of American life, covering commuting times, internet access, family life, income, education levels, disabilities and military service, among other topics.

The existing questions ask respondents to answer “yes” or “no” if they have difficulty or “serious difficulty” seeing, even with glasses, or are blind; hearing, or are deaf; concentrating, remembering or making decisions because of a physical, mental or emotional condition; walking or climbing stairs; dressing or bathing; or performing everyday tasks because of a physical, mental or emotional condition. If the answer is ″yes,” they are counted as having a disability.

Under the proposed changes, which follow international standards, respondents would be allowed to answer most of the same questions with four choices: “no difficulty,” “some difficulty,” “a lot of difficulty” and “cannot do at all.” A person would be counted as disabled if they answered “cannot do at all” or “a lot of difficulty” for any task or function. The changes also would have added a query on whether respondents have trouble communicating.

Supporters of the proposed changes said the revisions would have provided more nuanced data and given officials better details about disabilities that can inform how services or resources are provided.

In a test run, the percentage of respondents who were defined as having a disability went from 13.9% using the current questions to 8.1% under the proposed changes. When the definition was expanded to also include “some difficulty,” it grew to 31.7%.

The proposed changes to the disability questions were among several tweaks to the American Community Survey that the Census Bureau was planning to submit this year for approval to the Office of Management and Budget. As part of that process, the Census Bureau solicited public feedback and got more than 12,000 responses, with the majority expressing concerns about changes to the disability questions.

Advocates said Tuesday that they will be focused on working with the bureau on coming up with questions that capture the range of disabilities, including people with mental health problems, developmental disabilities or chronic health conditions, such as those faced by many people living with long COVID.

“While this is a win for our community, we must stay committed to the long-term goal of developing better disability questions that are more equitable and inclusive of our community,” Bonnielin Swenor, director of the Johns Hopkins Disability Health Research Center, said in an email.