Our research at Boston Indicators relies heavily on high-quality, publicly accessible data from the federal government—data that enable our analysis of the economic, social, and public health issues that face our region. It’s no surprise then that we are alarmed at the increasing politicization of federal data and the lack of transparency surrounding recent alterations to key datasets from government websites. And we are not alone.

In the early weeks of the current presidential administration, federal agencies took down thousands of datasets and public web pages, including critical information from the Census Bureau, the CDC, and the EPA. These removals affected public health surveillance data, mapping tools, and various publications, as well as other highly valued resources. While most of the affected data have since been restored—due in part to public outcry and legal challenges—data on gender identity are unlikely to be collected going forward. This episode has raised serious concerns about the reliability of federal data collection and dissemination under this administration.

The sudden disappearance of these datasets without any communication, even if temporary, has shaken our confidence in the federal government’s commitment to maintaining nonpartisan, high-quality data. Looking ahead, the bigger issue is not just what was taken down, but whether we can continue to trust federal agencies to collect and report data without political interference.

The Elimination of Gender Identity Data and Targeting of DEI in Data

One of the most damaging aspects of the data removals has been the Trump administration’s explicit effort to eliminate federal data collection on gender identity. This move was part of a broader executive order directing federal agencies to dismantle diversity, equity, and inclusion (DEI) initiatives. In practice, this has led to:

• The removal of all gender identity-related questions from at least four major federal surveys, including the Household Trends and Outlook Pulse Survey, the National Crime Victimization Survey, and the National Health Interview Survey.

• The scrubbing of federal websites containing research and data on LGBTQ+ health and demographics, including the Behavioral Risk Factor Surveillance System (BRFSS) and the Youth Behavioral Risk Survey System (YRBSS), which were later restored but with missing supporting materials such as codebooks and questionnaires.

• The permanent suspension of efforts to expand federal data collection on transgender populations, despite previous bipartisan efforts to improve this critical data infrastructure.

The justification for these actions is clear in official government statements. A memo currently posted on several federal data websites that were reinstated following a judge’s February 11 ruling states that while the page was restored “per a court order,” the Trump administration “rejects gender ideology” and claims the information “does not reflect biological reality.” This language makes it evident that the removals were ideologically driven rather than the result of a standard, evidence-based review process. It is one thing to debate whether the government should be asking these questions in the first place, but it is another to take the additional step of censoring these data entirely, erasing their existence from public records. By doing so, the administration is not just changing policy—it is preventing current and future researchers, policymakers, and the public from accessing crucial information that reflects the lived realities of transgender individuals and other marginalized groups.

“Move Fast and Break Things” Should Not Be Applied to Federal Statistical Agencies

Beyond the targeted removal of gender identity data, the administration’s broader approach to federal data management has been chaotic and damaging to public trust. The lack of communication surrounding the removals meant that researchers, policymakers, and the public were left in the dark about which datasets had disappeared and why. This approach also led to the sudden disappearance of data from sources as varied as the US Census Bureau, Centers for Disease Control and Prevention, Department of Health and Human Services, Office of Justice Programs, Office of Scientific and Technical Information, and the Department of Housing and Urban Development.

Crucially, this pattern of disruption has not been limited to website takedowns. Reports indicate that internal federal agency operations have also been affected, with employees struggling to access data-sharing platforms, such as US Census staff responsible for planning the 2030 Decennial Census who lost access to critical internal platforms for multiple days. Such disarray raises legitimate concerns about the future reliability of federal data collection efforts.

The Integrity of Public Data Is Built on Trust

Public trust is the foundation of federal data collection. Reliable and transparent government statistics are essential for everything from urban planning and economic forecasting to public health interventions and climate research. The complexity of federal data collection—which requires meticulous planning, methodological consistency, and years of coordination—means that even temporary disruptions can have long-term consequences.

The shadiness surrounding these removals has already eroded confidence in federal data systems. Professional organizations such as the Council of Professional Associations on Federal Statistics and the Association of Public Data Users have called for assurances that federal data will not be altered or removed for political reasons. Meanwhile, research organizations have scrambled to archive federal data on independent servers, fearing that critical information may disappear again without warning.

Beyond immediate concerns about access, there are deeper fears about the administration’s future actions. Reasonable people can debate which specific questions belong on federal surveys, and have—for instance, in past discussions about adding a citizenship question to the Decennial Census or more recent ones about how best to ask about race and ethnicity. However, such decisions should be made through careful deliberation, not through abrupt and politically motivated edicts. The recent removals set a dangerous precedent: If one administration can erase entire categories of public data overnight, what prevents future administrations from doing the same to datasets they find politically inconvenient?

While the immediate response from local organizations—working swiftly to archive invaluable resources—is encouraging, the long-term implications remain uncertain. No single entity can replace the various federal institutions that make all these rich and important data possible. It is therefore of utmost importance that these institutions remain fully functional, politically independent, and follow established laws regarding the collection and dissemination of their invaluable data.