Disappearing Data at the U.S. Federal Government

Document Type

Article

Publication Date

3-26-2025

ISSN

1533-4406

Publisher

Massachusetts Medical Society

Language

en-US

Abstract

Health data hosted by the U.S. federal government have been disappearing. According to our analysis, between January 21 and February 11, 2025, the Centers for Disease Control and Prevention (CDC) removed 203 datasets (13% of its online datasets, a reduction from 1519 to 1316). The situation is dynamic; various datasets have been removed, restored, manipulated, or not posted as planned.

A dataset with county-level information on mortality associated with heart disease, for example, which is used for research on demographic and geographic determinants of heart disease, was unavailable starting on January 31 and was restored on February 10, with the term “gender” replaced with “sex.”1 The Social Vulnerability Index, which includes information on race and gender and is used to predict susceptibility to natural disasters, was unavailable beginning on January 31 and was restored only after a February 11 court order. The CDC reportedly stopped giving researchers access to the Pregnancy Risk Assessment Monitoring System, a maternal and infant health database that is widely used to study maternal mortality, and ceased collecting new data for the program.2 The Trump administration hasn’t provided definitive explanations for these changes, but many of them may reflect agency responses to its directive to remove information “that promote[s] or otherwise inculcate[s] gender ideology.”

The consequences of these changes are expected to be far-reaching. Without access to accurate and timely data, scientists’ work will become more difficult, and we will understand less about the world. Clinicians and public health practitioners will need to make the same challenging decisions as they otherwise would have, but with less information. Public health experts will have more trouble identifying trends in disease outbreaks or overdose deaths and assessing or advocating for various interventions. The problem isn’t only federal: state and local public health agencies often rely on federal data or work with federal agencies and will be similarly affected.

Data suppression can directly support policy goals. The 1996 Dickey Amendment and the 2003 Tiahrt Amendments, for example, blocked data sharing and federal funding for gun-violence studies, with what public health officials believed was an aim of scaring agencies away from collecting data that could provide evidence of harms associated with gun ownership.3 As we discuss below, removing government databases can also transfer power from public to private entities, strengthen monopolies, hobble innovation, and promote autocracy.

This document is currently not available here.

Link to Publisher Site

Share

COinS