Behind closed doors and without expert input, the Trump administration issued a major policy change to how census data is released. Data experts are concerned the result will be less reliable public data related to redistricting, natural disasters, the workforce, housing, and more.

On June 4, the Trump administration released an order, Disclosure Avoidance for Statistical Products, that forbids “any use of noise infusion” for statistical products. “Coarsening shall be the preferred category of Disclosure Avoidance methods for all statistical products,” the order states. “Suppression shall be permitted as a last resort, only to be used when coarsening is prohibited by law or would substantially defeat the accuracy or usability of a statistical product.” 

In statistical terms, noise infusion is a common and accepted technique for privacy protection when working with data: it creates “fuzz” or random values within a dataset, making the published statistics slightly different from the actual, sensitive data. Coarsening is the process of grouping and rounding data, or reporting it in ranges instead of potentially identifiable specifics. Suppression is what it sounds like: redacting information, replacing it with asterisks, or not releasing the data entirely. 

NPR’s Hansi Lo Wang first reported on the policy change and its implications. People who work with census data and statistical analysis are worried that limiting the ways the Census Bureau and the Bureau of Economic Analysis (BEA) can release data will severely limit what information ends up available to the public.

Data coming out of small communities and industries, especially, could be heavily affected by the change. “Because ‘coarsening’ (grouping, rounding, reporting in ranges) and suppression are the only not-prohibited tools named in the order, it means that to keep information safe, the Census Bureau and BEA need to group small things (like small communities or small business types) into larger ones, or they need suppress the data completely,” Beth Jarosz, a senior fellow at Georgetown University's Massive Data Institute and vice president of the Association of Public Data Users, told me in an email. “Small industries may get rolled into bigger industry categories. Small counties may get rolled into county groups or not reported at all.”

On June 17, five groups — the Population Association of America, Council of Professional Associations on Federal Statistics, Association of Public Data Users, Inter-university Consortium for Political and Social Research, and Association of Population Centers — released a joint statement condemning the order. “This order subverts processes developed over decades to foster transparency and public trust and creates a scenario in which there will either be less privacy for our personal information, or less usable data, or both,” the statement says.

The Director of Science Policy for the American Statistics Association Steve Pierson wrote that the order “handcuffs the Census Bureau and the Bureau of Economic Analysis in terms of the techniques they can use for protecting the privacy of respondents.” 

John Abowd, the former Associate Director for Research and Methodology and Chief Scientist at the Census Bureau, posted a list of data products on Linkedin that this order would affect. These include the OnTheMap for Emergency Management system, a public data tool that provides real-time U.S. population and workforce statistics for areas being affected by natural disasters; Quarterly Workforce Indicators which include data about employment, job creation and destruction, wages, hires, and more; business formation and dynamics statistics; veteran employment statistics; data related to post-secondary educational outcomes, and many more. Many of these use noise infusion, which Trump’s order just banned.

There’s also confusion about how this order will even be enacted in practice. “Regarding the datasets that used noise infusion, it is unclear how this policy will impact public access,” Lynda Kellam, who leads the Research Data and Digital Scholarship team at the University of Pennsylvania Libraries and is a founding organizer of the Data Rescue Project, wrote following the order. “The policy is intended to be retroactive, raising concerns that data might be removed, but how that will play out is uncertain.” 

In the immediate fallout, at least, we’re already losing some public information. As Wang from NPR pointed out on Bluesky last week, multiple webpages related to noise infusion and differential privacy on the Census Bureau's website were removed following the order. Most of those pages have since been restored. At the Data Rescue Project, a team led by Lena Bohman has been proactively collecting and archiving Census Bureau working papers and making them available to the public. 

Jaroz said that along with the risk of unreliable or missing data, the abandonment of long-agreed-upon privacy protection methods can damage public trust in Census data. “When the Census Bureau and Bureau of Economic Analysis gather data, they promise respondents that they will keep responses confidential. When a person responded to the American Community Survey or a business owner provided information about their employees or sales, they expected that the Census Bureau and BEA would protect that information. By taking away tools that those agencies use to protect privacy and confidentiality, people may question whether or not Census and BEA can live up to that promise,” she said. “Similarly, the Census Bureau and BEA are producing information for public benefit. People respond, for example, to the American Community Survey (at least in part) because it will benefit their community. If the new rule results in cutting back how the data can be published and used, it also weakens trust and it is worth responding.” 

As Wang noted, America First Legal, a law group co-founded by Trump's deputy chief of staff for policy Stephen Miller, attempted to force the release of new 2020 Census data in a lawsuit last year, by challenging the Census Bureau's differential privacy system. Judges ruled it was too late to sue, but they refiled the case in February.

As NPR also reported last year, Trump and Republicans in Congress have been pushing to exclude people living in the U.S. without legal status in the 2030 Census. “People who are in our Country illegally WILL NOT BE COUNTED IN THE CENSUS,” Trump wrote on Truth Social in August 2025. This would be a radical change in how the Census has been conducted for more than 200 years. Redistricting and gerrymandering have been a massive fight for the Trump administration for years, and has ramped up ahead of the 2026 midterms, as the Supreme Court recently weakened the Voting Rights Act and allowed for more redistricting that would favor Republican control of the House. 

The data policy change is also happening in light of the Trump administration’s gutting of Census practice test locations in the South. In February, the Associated Press reported that the administration is eliminating four out of the six locations that were slated to test new methods for the 2030 census. “The Census Bureau would be essentially flying blind into communities that need testing most — tribal lands, rural areas with limited connectivity and places with historically low response rates,” Mark Mather, an associate vice president at the Population Reference Bureau, told the AP. “You can’t fix what you don’t test.”