(AFP) – A US Census Bureau report claiming more than 20,000 noncitizens voted in the 2020 presidential election has raised doubts over the impartiality of the influential but normally non-partisan agency. Unproven allegations of widespread voter fraud — including the casting of ballots by noncitizens, which is illegal in federal elections — have been a cornerstone of President Donald Trump’s insistence that his loss to Joe Biden six years ago was unlawful. And the bureau’s analysis stating with “high confidence” that around 24,000 noncitizens voted in 2020 was pounced on by Trump as clear evidence to back his claims.
However, experts and former Census Bureau officials have raised serious doubts about the accuracy of the document, suggesting the methodology is flawed. Former Census Bureau senior executive Amy O’Hara called the paper an abuse of the agency’s “sterling reputation for high-quality analysis.” She pointed to the lack of attribution and other “red flags on methods and research integrity.” NPR reported that the analysis was not conducted by career civil servants, but rather by a team including individuals affiliated with the America First Policy Institute (AFPI), a think tank started by officials from Trump’s first administration.
O’Hara, who is now director of the Massive Data Institute at Georgetown University, said it is “unthinkable that the carefully negotiated, curated and stewarded data could be exposed for political gain instead of statistical purpose.” The Census Bureau and AFPI did not immediately reply to AFP requests for comment.
– Comparing two datasets –
The report compared federal administrative records that include immigration information with a nationwide voter file dataset provided by the Virginia-based business Data-Clear under a contract it won related to the 2030 census. Company president Carolyn Carlson told AFP: “The voter file records provided and delivered by Data-Clear did not contain an individual voter’s citizenship status.” She said Data-Clear played no role in the analysis, and that any questions about the methodology used should be directed to the Census Bureau.
For Michael Morley, director of the Florida State University Election Law Center, there are several possible sources of error when matching datasets. If the data were collected at different times, for example, some of the information — especially on a fluid process like gaining US citizenship — could be out-of-date, he explained. “It would be a mistake for anybody to try to draw definite conclusions just based on the fact that two databases happen to match,” Morley told AFP.
The Census Bureau analysis, however, does not offer any disclaimer that further cross-checks are needed to determine the actual number of ballots cast by noncitizens. Experts said even if all 24,000 votes were proven to have been cast by noncitizens, that figure accounts for less than 0.02 percent of the roughly 160 million votes in 2020 and would not have changed the outcome of the election. The records also cannot show who ballots were cast for, negating Trump’s implication that all of those votes went to his opponent.
– Marisha GOLDHAMER
© 2024 AFP
