Commerce Secretary Howard Lutnick stirred the pot when he said a federal analysis found more than 24,000 noncitizens voted in the 2020 election. That claim set off a firestorm online — and rightly so. The problem is that the public record does not show a Census Bureau report that reaches that conclusion. What does exist are federal database matches that flagged roughly 24,000 registration records for review. That’s a big difference, and Americans deserve clarity, not spin.
The claim versus the underlying data
Secretary Howard Lutnick posted that a Census Bureau analysis shows “more than 24,000 noncitizens illegally voted in the 2020 election.” But investigators and reporters looking for a Census release that backs that exact wording came up empty. The 24,000 number most closely tracks a DHS/USCIS check using the SAVE database, which flagged about 24,000 possible noncitizen registration records out of roughly 60 million checks. Separately, an administration analysis based on commercial databases produced a much larger, less-documented estimate for a handful of states. In short: flagged records were real, but a public Census Bureau finding that those people “illegally voted” is not.
Why a flagged match is not proof of voter fraud
Database matches are a starting point, not a verdict. SAVE and other automated tools were built for benefit eligibility, not as final proof of criminal voting. Matches produce false positives — input errors at DMVs, name changes, later naturalization, or registrations that never resulted in a ballot cast. State follow-ups in places like Georgia and New Jersey drove the initial tallies down sharply after checking voting histories and correcting data errors. Election experts and fact-checkers warn these technical distinctions matter if we are serious about election integrity.
Hard checks, not headlines — what conservatives should demand
Conservatives want clean, secure elections. I want that too. But vigor and accuracy must go together. It is right to hunt down noncitizen voting where it exists, and it is right to use federal tools to surface potential problems. It is not right to conflate a database flag with a confirmed illegal vote or to misattribute findings to agencies that didn’t publish them. If the administration has proof beyond preliminary matches, show the evidence, state the methods, and let state officials verify. If people broke the law, prosecute. If the flags were errors, say so publicly.
Election integrity burns hot in this country because the stakes are high. We don’t win trust by tossing raw machine matches into the headlines or by reflexive denial. We win it by transparent, careful work that turns flags into facts or clears the innocent. Secretary Lutnick and the agencies involved should stop the theatrical pronouncements and start publishing the data and the follow-ups. Conservatives should press for both accountability and accuracy — otherwise we trade the rule of law for the rule of rumor, and nobody wins.

