How Often Do US Voter-List Checks Wrongly Flag Eligible Voters? Recent disputes over voter registration and citizenship checks have raised questions about the accuracy of government databases. I'm looking for court recor…
Wing Li@wingli
The useful question is not whether any noncitizen has ever appeared on a voter roll. It is how often citizenship checks flag or risk removing eligible voters, and whether those errors are caught before an election. Some local reviews show high error rates, some large federal lists remain unverified suspicions, and a smaller number of cases involve people registered after marking that they were not citizens. Those categories should not be collapsed into one claim about illegal votes.
On October 28, 2026, U.S. District Judge David O. Carter denied a request to immediately block California officials from counting the ballots of 20,000 voters whose citizenship the administration questioned. He called the allegations "not credible" and warned that the intervention would likely disenfranchise legitimate voters. The 20,000 figure was presented as a "high confidence" floor, after an earlier claim that there may be more than 190,000 noncitizens on the rolls. Those are suspected records, not confirmed ineligible registrations or proven illegal votes. L.A. County officials said a review of about 200 names found that a substantial portion appeared to have been born in the United States, a claim Homeland Security disputed. Administration officials acknowledged "discrepancies," and a federal prosecutor told the court, "We're never going to stand here and tell you that everything is 100%." Carter said illegal voting can be prosecuted case by case, but setting aside 20,000 ballots at once would "discount the votes of United States citizens."
Documented false positives are clearer when local clerks check the files. A February 2026 ProPublica and Texas Tribune investigation found that the federal SAVE tool made persistent mistakes, especially for people born outside the United States, and that DHS had to correct data sent to at least five states. In Boone County, Missouri, more than half of 74 voters flagged as noncitizens were actually citizens, including a man registered at his naturalization ceremony. In Texas, reporters documented at least 87 wrongly flagged voters across 29 counties. In Denton County, 12 of 84 flagged voters, including naturalized citizen Sofia Minotti, proved their citizenship, an error rate of at least 14%. A survey of 97 Texas counties found that more than 5% of those SAVE identified as noncitizens proved to be citizens. Clerks also found some people who had marked that they were not citizens and were registered in error; that is confirmed ineligible registration, not evidence they voted. After later federal corrections, St. Louis County's list fell from 691 to 133, and Boone County's fell from 74 to 33.
Nevada's dispute shows why a "confirmed" federal match is still only a claim until the state can verify it. In September 2026, chief deputy attorney general Greg D. Ott told DHS that for 185 people the department said it had confirmed as noncitizens, "our records indicate that they were citizens" and that DHS had "provided nothing of substance to suggest otherwise." DHS had earlier cited as many as 15,903 potential matches as a ceiling and later said the 185 were a subset that went through manual review. Secretary of State Francisco Aguilar said the state's records did not support the claims. My view is that list maintenance should require a local check of registration files, naturalization records, or state documents before anyone is removed or a ballot is set aside. Errors can be corrected, as Missouri's later revisions show, but only if a flagged name is not treated as proof.
Note: this post was drafted with AI assistance; I checked each claim against the linked sources and take responsibility for the content.