The Public Health Cost of Eliminating Race and Ethnicity Data
Imagine trying to solve a public health crisis without knowing which communities are being hit hardest. We are moving closer to that reality. In recent months, concerning patterns have emerged that threaten to unravel decades of progress in how the federal government collects and reports race and ethnicity data.
The near-complete elimination of gender identity from future data collection was the first wave of widespread reductions to demographic data since January 2025. We think race and ethnicity is likely to be the second. And while the administration’s approach to gender identity was done in one sweeping executive order, race and ethnicity data are being eroded from several directions.
The erosion began with the removal of information related to diversity, equity, and inclusion (DEI), including topics like health equity and cultural competence, in response to executive orders. Then agencies began walking back their intent to implement the modernized race and ethnicity standards that were a decade in the making. The Office of Management and Budget repeatedly pushed back deadlines for implementation plans. Now, race and ethnicity questions are beginning to be removed entirely from data collections, with stated justifications ranging from legal interpretations and paperwork burden to shifting policy priorities.
The growing threat to the availability and quality of race and ethnicity data has dire public health consequences. It narrows our visibility into who is getting sick, who is dying, and whether the people who need help are getting it. Let’s take a closer look at what these losses could look like.
Our nation’s common denominator
Our nation has collected race and ethnicity data in every census since 1790. Yet a draft Commerce Department rule is under review that would end race and ethnicity data collection in the 2030 Census. The draft rule also excludes undocumented immigrants and H-1B visa holders. The justification for the change is “to protect from any distortions created from the inclusion of personal questions.” Regardless of justification, ending the collection of race data will have far-reaching consequences, especially on America’s health.
Health data depend on the census in a critical way. Imagine only knowing that in 2023, around 85,000 Black Americans and 520,000 White Americans died of heart disease, without knowing how many total Black and White Americans were alive in 2023. If you read those counts alone, heart disease may look like a significantly worse problem for White Americans. We need the denominator from the census,1 the total number of people in each group. Black Americans die from heart disease at a rate 21% higher than White Americans—the highest of any racial group in the country. Without the denominator, we cannot compare one group to another or know if the number of deaths are rising or falling over time.
This issue reaches beyond health statistics. Federal programs, like at HHS, use census counts to distribute funding. For example, some tribal grant programs rely specifically on the census American Indian or Alaska Native category to allocate funding. If you change what the census counts, then you change where the money can help most.
The census is the cornerstone of the federal statistical system, and changes to questions in the census often propagate across federal data collections. With the Census Bureau punting on the timeline for adopting the new race and ethnicity categories, and now possibly pulling race data completely out, health-related data collections are likely to follow suit. Removing race and ethnicity from federal forms and surveys would leave the country unable to measure who is getting sick and dying at higher rates, unable to direct prevention programs to the communities carrying the heaviest burden, and unable to tell whether any of it is working as intended.
While race and ethnicity remain intact in these datasets (for now), the next three examples make clear what is at stake. If demographic data like race and ethnicity are removed, some of the most vulnerable populations would be endangered: new mothers, people experiencing chronic diseases, and those experiencing addiction, to give just a few examples.
Maternal Mortality
The National Vital Statistics System (NVSS) compiles birth and death certificates from every state, giving the federal government its most complete picture of who dies and how, including who dies during or shortly after pregnancy.
The numbers reveal a persistent racial disparity. In 2023, for every White mother lost to pregnancy-related causes, more than three Black mothers lost their lives.This racial disparity has existed for more than 100 years. Health systems use race-stratified data to see the big picture: identify which mothers face the highest risk, put in place interventions like blood pressure monitoring and faster hemorrhage response, and then assess whether those protocols actually narrowed the disparities. An intervention can lower the overall death rate while leaving the disparity untouched, or even widening it, and without race data, no one would know. If NVSS stops collecting or reporting race data, we will no longer be able to see if we are closing that gap.
Diabetes
The Behavioral Risk Factor Surveillance System (BRFSS) conducts an annual nationwide survey tracking chronic diseases and health behaviors in every state, giving health departments a county-by-county picture of where health issues, like diabetes, are concentrated. Respondents also report their race and ethnicity, which gives insight into which populations carry the heaviest health burden.
Diabetes does not fall evenly. American Indian and Alaska Native adults develop diagnosed diabetes at nearly twice the rate of white adults. Black and Hispanic adults also face significantly higher rates. Diabetes drives kidney failure, blindness, and amputation, and costs the country more than $400 billion a year.
The lifestyle change program behind the National Diabetes Prevention Program can cut the risk of developing type 2 diabetes by about a third in people who are at risk, but it does not work equally well for everyone. Without the data showing which communities have higher rates of diabetes, prevention resources flow to whoever is easiest to reach, which is rarely the people who need them most.
Overdoses
The State Unintentional Drug Overdose Reporting System (SUDORS) collects information on fatal overdoses from 49 states and the District of Columbia. It records race and ethnicity for each death along with the circumstances surrounding it: whether a bystander was present, whether anyone administered naloxone, whether the person had been in treatment before. That combination of data can be turned into a prevention plan.
This matters because overdose deaths fall unequally across racial groups than almost any other health issue. American Indian and Alaska Native adults face the nation’s highest overdose death rate at nearly double that of White adults (65 vs. 33 per 100,000). Meanwhile, overdose deaths among Black adults rose to roughly 49 per 100,000 in 2023, nearly 50% above the rate for White adults, even as the national rate fell 4 percent. That decline belonged almost entirely to White adults. Without race data, the drop looks like progress for everyone, and prevention may ease off or continue to not reach the communities where deaths are still climbing.
Implications for Data Policy
With less than four years until the next decennial census, we must keep a close eye on all demographic data, especially race and ethnicity. The census sets the de facto standard for how federal agencies, as well as state and local governments, collect data across thousands of forms as surveys. As goes the census, so goes the rest of the statistical system.
If race and ethnicity are removed from the decennial census, it would set a precedent that extends far beyond 2030, making it easier for other agencies to do the same. This is bigger than a singular elimination of race and ethnicity questions. It could accelerate the second wave of widespread reductions to demographic data, leaving public health officials with even less information to deliver better health outcomes for all Americans.
Removing race and ethnicity data does not make health disparities go away. It merely makes them harder to find, harder to measure, and harder to fix.
Americans Would Trust AI More if Policies Ensuring Fairness Were Implemented. Here are Ten Ways to Start.
By now, you’ve probably heard that most Americans do not trust AI. This distrust is especially concerning given how deeply these systems are already shaping access to healthcare, education, housing, jobs, and public benefits. Too often, these decisions happen without transparency, oversight, or meaningful avenues for recourse. At the same time, confidence in both technology companies and government institutions to manage AI responsibly remains low.
The stakes are clear, and the policy choices we make today will make or break society’s view of AI. We are currently at a critical opportunity to shape how AI is governed before harmful practices and inequities become further entrenched. To meet this moment, the Federation of American Scientists, with the support of the Kapor Foundation, launched a policy sprint, which is an intensive, time-bound effort designed to tackle complex challenges quickly and collaboratively. Policy sprints bring together experts from across disciplines, from academics, technologists, advocates, and practitioners, to develop practical, actionable solutions.
For our SOURCE CODE AI Trust and Fairness Sprint, we’ve developed 10 memos with leading experts that are detailed, implementable policy solutions. We have delved into why fairness is so hard to define and implement, and what is needed to promote public trust in our essay that frames this new policy agenda. These memos are not exhaustive; we know the landscape of challenges and potential solutions is far broader. Instead, we offer them as a starting point: ideas that we hope will not only serve as smart and actionable tools for policymakers, but also inspire the community to build out and advance new, detailed approaches.
- Jae Yeon Kim and Aniket Kesari show how government procurement can be a powerful but underused lever to require transparency, fairness, and accountability in AI systems.
- Nicole Ozer and Brady Hirsch emphasize that in order for AI to work for the people, state leaders must adopt people-centered AI decision making frameworks. hey explain how to ask and answer foundational questions about why and whether to use AI before skipping to the how of implementation.
- JB Branch highlights similar risks in K–12 education, proposing procurement guardrails and impact assessments to protect students’ rights and safety.
- Liza Paudel advances Community Benefit Agreements as a tool for ensuring that communities, especially those disproportionately impacted, secure enforceable protections as AI-driven data center development accelerates.
- Wilneida Negrón examines the rise of “surveillance pay,” calling for stronger regulatory safeguards to protect workers from opaque, data-driven compensation systems.
- Anna Lenhart proposes embedding “Decision Subject Representatives” into AI governance, giving impacted individuals a formal role in shaping high-stakes systems.
- Y. Tony Yang focuses on Medicaid and safety-net healthcare, proposing community-centered oversight, nurse-led audits, and patient protections to ensure AI improves rather than undermines care.
- Ziwei Qi, Tatiana Lin, and Ayokunle Olagoke highlight the risks of “algorithmic invisibility” in rural communities and call for “rural AI proofing” to ensure these populations are not overlooked in AI system design, governance, and implementation.
- Jon M. Peha explores the risks of AI-generated police reports and proposes coordinated federal and local strategies that foster innovation and prevent unsafe deployments.
- Gaurav Laroia and Charlotte Slaiman argue that enforcers must look beyond standard monetary penalties to truly protect communities from algorithmic harms. They show how historic tech settlements can be transformed into an active defense, building societal resilience and preparing the public for the next wave of technological change.
To structure our policy agenda, we have considered how these ideas have employed targeted mechanisms, which we call policy levers, to ensure legitimate use, fair outcomes, and public trust in AI. These levers include government use of AI, public engagement, sectoral considerations, and redress in legal remedies; many of our policy ideas use several of them.
This work was shaped by a multidisciplinary advisory working group of 17 experts from academia, civil rights organizations, think tanks, and beyond. Their insights helped refine the project’s focus, identify the most pressing policy opportunities, and strengthen each memo through expert review. Advisory members provided input in an individual capacity; their participation does not imply endorsement of the findings or recommendations. We are grateful for their support and expertise throughout this process.
- Quinn Anex-Ries, Center for Democracy & Technology
- Alex Ault, Lawyers’ Committee for Civil Rights Under Law
- Annette Bernhardt, UC Berkeley Labor Center
- Miranda Bogen, Center for Democracy & Technology
- Leah Frazier, Lawyers’ Committee for Civil Rights Under Law
- Sanmi Koyejo, Stanford University
- Sarah Myers West, AI NOW
- Tejas N. Narechania, UC Berkeley School of Law
- Cathy O’Neil
- Asad Ramzanali, Vanderbilt Policy Accelerator
- Cody Venzke, ACLU
- Jonathan Walter, Leadership Conference on Civil and Human Rights
The ideas generated through this sprint are practical, actionable, and grounded in existing authority. They demonstrate that policymakers already have many of the tools needed to govern AI effectively; they simply need to be deployed with urgency and intention. There is a narrowing window to shape the integration of AI systems into society. With thoughtful policy action, it is still possible to build systems that are fair, transparent, and accountable, and to earn the public trust that will ultimately determine AI’s future. We hope policymakers are ready to act.