Amid deletions and lawsuits, groups race to save federal data

After President Donald J. Trump’s executive orders on gender identity; diversity, equity, and inclusion; and foreign aid, some scientific data, information, and tools are being deleted from US government websites. Groups made up of librarians and scientists are banding together to try to save copies of government data for scientific use before they disappear. In addition, some groups are filing lawsuits challenging the administration’s actions.

The deletions are a problem because scientists use some of these data to do their research, says James R. Jacobs, US government information librarian at Stanford University and one of the cofounders of Free Government Information, a group that works to preserve free access to government information. For example, the US Census Bureau has a lot of statistics and data that scientists and historians rely on to inform their research. “Going forward, they might not be able to get data on trans people or any of those kinds of issues,” he says.

During the first Trump administration, some data on the environment and climate change were removed from federal websites. But this time around, things are happening much faster and more broadly, says Lynda Kellam, one of the organizers of the Data Rescue Project, a group formed to act as a coordinator to save US federal data in danger of being removed from public view.

“It was a week after the administration [change], and suddenly data was starting to be either taken down completely or taken down, modified, and put back up,” Kellam says. “We were really surprised by the pacing,” she says.

For example, in late January, US Centers for Disease Control (CDC) web pages and datasets that mentioned gender identity were removed or edited. After a federal judge’s order on Feb. 11, information was restored, although as of Feb. 26, yellow banners warning that changes are in process or that the administration rejects the reality that there are more than two genders stretch across the top of some CDC websites.

In addition, a farming group and environmental advocacy groups filed a lawsuit Feb. 24 in the US District Court of the Southern District for New York to challenge the removal of climate data from the US Department of Agriculture websites.

In another example, the White House Council on Environmental Quality’s Climate and Economic Justice Screening Tool—interactive maps that let the public track variables such as air quality, pollution, traffic, and education of an area—went dark sometime after Jan. 21, screenshots from the Internet Archive show. As of Feb. 26, the tool’s page gives an error message. The Environmental Data Governance Initiative, another group working to preserve data, created an unofficial copy of the tool that it made available to the public.

Some changes in web pages are normal when a new administration takes over, Jacobs says. “At 12:01 a.m. on Jan. 20, the National Archives and Records Administration (NARA) encapsulates the whole Whitehouse.gov site, puts it on an archival space, and creates a new Whitehouse.gov for the new administration,” he says. But what’s happening now goes beyond that Jacobs says, into removing information and data on topics such as gender identity or climate. “That’s problematic,” he says.

Beyond the government-based NARA’s efforts, the End of Term Web Archive group works to make copies of federal web pages that a new administration might alter or take down. The Internet Archive is another collective effort that preserves some federal websites through its long-running Wayback Machine, although this doesn’t specifically focus on US government data.

But saving data is a different problem than saving a copy of a website, Kellam says. “You need a lot more information about a dataset than you would need for a website,” she says. For example, often you need documentation that goes with the data, such as scripts, PDFs, read-me files or other information that describes the data, and the metadata.

“Going back to [the] Wayback Machine is not going to get you a dataset that's going to be usable for the long term, even if the Wayback Machine could capture it,” Kellam says. “Web archiving is great, but it's not this. It's just not the same as the kind of things that we would need for data archiving.”

Another issue is the usability of the data, she says. There are many people helping grab data, but without a central location to put the information in, how are people who need to use the data going to find it? “We have people who are going and rescuing data and putting them in random places,” including personal websites, Kellam says.

The Data Rescue Project now has a Data Rescue Tracker to keep tabs on what group is working on what data, what has already been collected, and where the data are now. But Kellam thinks that no matter what the group’s efforts are, some federal data will be lost, such as some US Agency for International Development (USAID) data on foreign aid, which was taken down before any data-rescuing groups could mobilize.

In addition, Jacobs wonders what’s in the future for federal data and whether federal agencies will continue to collect data on gender identity, climate, and some of the other topics that have been removed or altered. “There'll be a gap, a black hole in the data,” Jacobs says.

Gaps from the removal of US government data are one of the biggest potential impacts to scientists and researchers, he says, as a lot of academic work looks at how data change over time. This data purge by the federal government may cause ripples into the research of the future, resulting in less-robust studies or even a lack of studies on a subject with missing data.

The White House did not respond by press time to requests for comment about federal data removal.