Data enthusiasts are taking action.
Following the start of President Trump’s second term, a group of statisticians, demographers, and data scientists banded together to collect and protect datasets. This move was prompted by the observation that significant datasets were disappearing or being altered on U.S. government websites, which was unprecedented.
Their aim? To ensure that national statistics remain accessible in the future. They worry that without reliable data, policymakers might drift further into partisan divides.
“There are knowledgeable and passionate individuals who care deeply about not just the Census Bureau, but all statistical agencies,” said one participant. “Even in these trying times, they inspire hope.”
Since January, threats to the U.S. data infrastructure have expanded beyond just the alteration or loss of gender and health data. Experts highlight that issues have been compounded by job cuts affecting those who traditionally safeguarded access to vital datasets.
“We have trillions of data files; it’s hard to fathom the public money used to gather this information,” noted Jennifer Park, a research director at the National Statistics Committee, during a Public Data Users Association meeting. “Yet, now many of these files are virtually unreachable due to a lack of staff to manage them properly,” she added.
“Gender” has been changed to “Sex”
This February, the official health data portal from the CDC faced major disruptions, which eventually resolved. Concurrently, users trying to access data from the U.S. Census Bureau encountered messages declaring the site was “down for maintenance” for several days before it was back online.
Researchers Janet Freirich and Aaron Kesselheim examined 232 federal public health datasets that underwent changes at the start of the year, discovering that nearly half experienced “significant modifications.” Their findings were published in the Lancet Medical Journal.
A major challenge lies in tracking these changes since many aren’t formally documented.
Beth Jaros, a senior program director who previously felt reassured about having downloaded necessary data from the National Survey on Child Health, later realized that she had overlooked crucial information—specifically, the removal of questions related to discrimination on the basis of gender and sexual identity.
“That’s one corner my team overlooked,” Jaros said during this week’s APDU meeting. “We thought we were compiling a valuable historical record.”
New groups have emerged this year to safeguard federal data, including the American Federation of Scientists, which monitors federal dataset changes. The University of Chicago Library’s Data Mirror actively backs up fragile datasets, while the Data Rescue Project facilitates various data conservation efforts. Additionally, the Federal Data Forum provides updates on the status of federal statistics.
The external data advocates are also encouraging employees within statistical agencies to discreetly connect with them to help preserve restricted public data.
“We can’t take for granted that this data will be here tomorrow,” emphasized Lena Bohman, a founding member of the Data Rescue Project.
Experts’ Advisory Committee Resurfaces Informally
Meanwhile, a group of external experts has informally revitalized the defunct U.S. Census Bureau Advisory Committee, which the Trump administration disbanded in March.
Census Bureau officials plan to skip the upcoming Census Scientific Advisory Committee meeting scheduled for September because the overseeing Department of Commerce shut it down. Nevertheless, the advisory group recently sent recommendations to the bureau, and demographer Allison Pryor expressed enthusiasm regarding the committee’s informal revival, indicating that it’s a valued resource even outside official channels.
“They don’t seem to seek external expertise … it feels like they prefer insights from those they can easily understand,” said Pryor, who leads demography efforts at the New Orleans data center.
