Former Census officials say the breadth of data used, as well as the lack of clear explanation of the methodology employed to analyze it, raises significant red flags.
“We don’t know whether or how the sources are updating their citizenship info,” Amy O’Hara, a former official who used to run the administrative data linkage unit at Census, tells WIRED. “It is really hard to join and deduplicate these files, especially across a range of years.”
In order to match people between surveys and records without revealing their identities, the Census Bureau uses what are known as Protected Identification Keys (PIKs), which replace identifying information with a unique code. Sources who spoke to WIRED say that they are unsure which data points were used to match PIK data across the many datasets, though something like an address could be used for cross-referencing.
Though PIKs have been used for years to allow academics and government employees to conduct studies without compromising the privacy of the people whose data it is, they are not foolproof. A January 2026 study published by the Census earlier this year found that “while PIK assignment is generally accurate, some person records are not successfully assigned a PIK, which can lead to sample selection bias in analyses of linked data.”
“The brief says ‘high confidence,’ which indicates some probability,” one bureau source who spoke on the condition of anonymity over fears of retribution by the administration, tells WIRED. “However, no confidence intervals in the tables of the brief are shown. The supposed count of non-voters is small relative to the total state population.” The source alleges that there are probably PIK errors and not actual noncitizens who voted.
Abowd says that false match rates—in this case, when a voting record is assigned the wrong PIK, and then when it is “matched” with citizenship data–are likely to be higher when using commercial data like those used in the report. This means that the analysis is more likely to return a “false positive” showing that a noncitizen had voted. “You can’t estimate the false positive rate without more technical details” that were not included in the report, Abowd tells WIRED. “Because the false match rate is much greater than the estimated noncitizen voting rate, the alleged noncitizen voters are more likely to be data errors.”
All of this, Boyd worries, could undermine more than just trust in voting and elections—it could make academics and researchers doubt the quality and reliability of the products and data that the Census Bureau produces. “They’re putting it out with the imprimatur of the Census Bureau so it looks legitimate,” she says. “But there’s no authors, no methods, no discussions of the limitations of the data, or the Bureau’s approach.”
Republicans have already attacked the accuracy of the 2020 census, alleging that an algorithmic process used by the Bureau to protect data privacy had disproportionately favored Democrats in terms of the drawing of congressional districts. In August, WIRED reported that an internal memo circulated by the Department of Commerce, which oversees the Census Bureau, proposed excluding immigrants from the count, a move that could radically reshape how congressional districts are apportioned.
The report is also just one part of a broad attack on the integrity of American elections by the Trump administration, which includes weaponizing the Department of Justice, dismantling voter protections, attempting to wrest control of election administration from states, and pushing an endless barrage of baseless conspiracy theories about voting in America.
