By Vera Messing
The Trump administration’s swift and systematic dismantling of statistical and data infrastructures has largely gone unnoticed. From erasing health and education data to cutting research funding, these actions threaten evidence-based policymaking, mirroring tactics used in autocratic regimes to obscure transparency and suppress inconvenient truths.
The news in February 2025 has been dominated by reports on President Trump and his administration’s attempts to reshape the US and the global order during the initial weeks of his presidency. His executive orders on international security and commerce as well as his actions dismantling US federal institutions and attacking immigrants and LMBTQ+ people have been widely covered by local and international media. However, a less visible yet critically important aspect of his governance has received little attention: the systematic dismantling of US statistical data collection and infrastructure. Statistics are a key foundation of democratic societies. They provide the government and public with necessary information to make evidence-based policies and related decisions. In a recent social science data and methodology conference held in Washington D.C., American colleagues shared their insights and experiences regarding the significant erosion of data infrastructure and suppression of academic freedom. They also shared their concerns regarding the termination of thousands of federal employees in the fields of data infrastructure during the first month of the Trump administration.
The collection of data regarding a country’s population and the performance of its institutions lies at the heart of evidence-based policymaking in modern democracies.
The US served as a prominent example of the value attributed to data collection and infrastructure by the state: a wide array of data – including on health, disease control, education, and equality – has become a model for many other countries.
Thousands of government employees and scholars devote considerable effort collecting and analyzing such data, constructing indicators that show social processes and the functioning of institutions, while enabling the assessment of the societal impact of policies. Much of the data is managed at the federal level, as coordinated and comprehensive systems are crucial for effective governance. Similarly in Europe, despite being composed of individual nation states, the EU operates a unified statistical system, Eurostat, which facilitates the comparative analysis of key policy fields such as economy, employment, health, education, and welfare, to list the most important areas.
The Trump administration’s intervention in the long-standing tradition and expertise for statistical data collection in the US shouldn’t come as a surprise, however the extent and rapidity of the changes are surprising. Within its first month in office, the administration has intervened in and dismantled key data collection systems and infrastructure in areas such as education, health, and disease control, and equality among others.
Academic literature discussing autocratic governance highlights the significance of controlling data collection and management in such regimes.
A common strategy of obscuring the transparency of governance and its outcomes involves halting the collection of certain data, manipulating data or indicators, and imposing censorship over access to data.
This control is vital for maintaining power and legitimacy in such regimes. Therefore, the actions of the Trump administration are not unprecedented; rather, they reflect an effort to exert control over data collection and publication in areas central to its ideology, specifically racial and gender (in)equality, health, and education.
In the weeks following his inauguration, Trump’s administration initiated strategic efforts to limit access to widely used datasets and resources, particularly those documenting economic, social, and health disparities experienced by millions of American people. However, the attacks on data and data collection need to be understood as part of a broader assault on research, knowledge, and evidence-based policymaking. With unprecedented speed, the administration laid off thousands of employees in US science agencies and halted funding for research at both government agencies and academic institutions (including universities). Although these cuts are presented as part of a broader effort to radically reduce government spending and downsize its workforce, its impact extends far beyond fiscal concerns: they threaten to dismantle decades of scholarly knowledge and data infrastructures.
This short article does not aim to address the broader impact on the US academic field – an issue that deserves another analysis. Instead, it focuses on the specific consequences for data collection and infrastructure particularly in two areas that are the most vulnerable to the recent interventions and are of central importance to social sciences: health and education.
Health data
One of the first executive orders issued by the Trump administration was Executive Order 14168, titled “Defending Women From Gender Ideology Extremism and Restoring Biological Truth to the Federal Government”. This order had significant implications for the largest statistical data infrastructures in the United States. Specifically, it mandated the removal of certain “politicized” terms from research data and publications, including the term gender. As a result, the Central Bureau of Statistics, the federal government’s primary statistical agency, has removed gender identity questions from several surveys it conducts in order to comply with the executive order. Furthermore, some of the key data sources maintained by the agency were made inaccessible to the public.
The Trump administration has also removed some of the most commonly used datasets, such as the Center for Disease Control and Prevention’s (CDC) Youth Risk Behavior Surveillance Survey and the AtlasPlus HIV surveillance tool from the web. Additionally, it eliminated resources from the US Census Bureau and other agencies that document economic, social, and health disparities. The CDC also stripped race and ethnicity data from social indicators like the Social Vulnerability Index and the Environmental Justice Index. Over 1,000 pages from the CDC website were removed overnight and the agency withdrew manuscript submissions from scientific journals to revise terms such as ‘gender’ and ‘transgender’. Although some pages were later restored, and publications with revised terminology will certainly be resubmitted,
these harsh and rapid actions represent a significant attack on the freedom of knowledge and publication.
As the Leadership Conference on Civil and Human Rights asserts: “Erasing data does not erase populations […] Data are inherently political because they reflect whose lives, experiences, and needs society values, shaping whose stories are heard, whose realities are acknowledged, and whose communities receive attention and resources. Removing these data sources helps to conceal disparities, making it easier for policymakers to justify inaction”.
Education
The field of education may bear the greatest impact from the Trump administration’s attack on data infrastructure. News regarding plans for the dismantling of the Department of Education are being published at the time of writing. Although the closure of the Department of Education would require Congressional support and may take some time, the administration started to cut budgets for evidence-based education policies and related data collection on day one.
Evidence-based policies require a wide range of high-quality statistical and survey data. As an initial step, the US Department of Education cancelled at least $881 million in contracts within its research division during February 2025, halting research informing students’ achievement in school. Contractors were instructed to halt their work instantaneously before data from research could be analyzed and results published.
Long-standing data collections by the National Center of Education Statistics (NCES) have also been halted, despite the agency’s legal obligation to do this work. The NCES plays a vital role in collecting, standardizing, analyzing, and disseminating data about the nation’s public school systems and student populations. It is an essential component of educational research. The NCES oversees the National Assessment of Educational Progress (NAEP), the US’s most comprehensive longitudinal catalogue of student achievement data in the US. Similar to OECD’s PISA, NAEP involves regular assessments of students in the fourth, eighth, and twelfth grades.
These measurements help inform researchers, policymakers, school administrators, and teachers about students’ performance and the factors contributing to both success and underachievement.
NAEP data is explicitly mandated by the Elementary and Secondary Education Act, which requires schools receiving federal funds to participate. Moreover, NAEP provides critical insights into disparities between student groups, such as students in racial categories, students in low-income households, non-native English speakers and more, making it a key source for studying intersectional inequalities in education.
It is evident that data regarding the performance of students, teachers, and schools can only be collected during the academic year, necessitating careful design and timing. The American Education Research Association and the Council of Professional Association on Federal Statistics issued an open letter on February 10, 2025 voicing their concerns about the measures and warning of their potential long-term impact. “Limiting the important work that NCES does by terminating these contracts will have ramifications for the accuracy of national-level data on the condition and progress of education, from early childhood through postsecondary to adult workforce.”
Professionals, scholars, and various associations are actively resisting the sweeping attacks of the Trump administration on data infrastructures. They are issuing warnings about the potentially devastating consequences these actions could have on evidence-based governance and the harm it will inflict on communities across the United States.
In response, several initiatives have emerged to track the rapid changes to statistical infrastructure and ‘save’ data from potential erasure or barriers to access. The American Statistical Association (AMSTAT) project “Assessing the Health of the Principal Federal Statistical Agencies” monitors the quality, accessibility, and most importantly the changes to data infrastructures. This project evaluates federal agencies’ ability to produce trusted quality statistics as well as their agility, accountability, and trustworthiness. Rather than waiting for their scheduled 2025 report, AMSTAT has decided to continuously publish information on how data collection and infrastructures are evolving. They also launched a SAFE Track website, where government agency employees can anonymously report instances of data cancellation, withdrawal, or impairment. Another noteworthy initiative is the Data Rescue Project, which aims to download and save data that is likely to be canceled or withdrawn from public access.
These forms of resistance to the attacks of the Trump administration on key data infrastructure that have informed policymaking and social processes are and will likely remain very important concerning the access to existing data and tracking politically induced biases. However, they do not affect the halting of data collection in the present and for the future. This might not seem the most important issue in US politics, but it will have important implications for policymaking in the medium run.
Vera Messing is a Research Fellow at the CEU Democracy Institute and a Senior Research Professor at the HUN-REN Center for Social Sciences. Her work focuses on comparative understanding of different forms and intersections of social inequalities and ethnicity and their consequences. She is specifically interested in policy and civil responses to ethnic diversity in the field of education and the labor market; migration and migrant integration; attitudes towards racialized minorities and immigrants; media representation of racialized groups and conflicts; and social science methodology and measurement of social phenomena.
