How the United States Checks That a Voter Is a Citizen

The rule that already exists

A voter in the United States does not simply assert citizenship into the void. When a person registers to vote in a federal election, the federal voter registration form requires a signed attestation of citizenship made under penalty of perjury. A false claim is a felony. It has been a federal crime since 1996 for a noncitizen to vote in a federal election, punishable by fines, up to a year in prison, and removal from the country.

That is the legal floor, and it is the reason the current debate is not about whether noncitizen voting is allowed. It is not. The debate is about verification: whether an attestation backed by database checks is enough, or whether every registrant should first produce a document. Understanding the argument means understanding the machinery that runs underneath it.

The cross-checks in place today

States do not take the attestation on faith alone. Most run new registrations against their own motor vehicle records, where citizenship or lawful status is often already recorded. Many query the federal Systematic Alien Verification for Entitlements database, run by the Department of Homeland Security, which was built to confirm immigration status. States also draw on jury excusal records, in which people decline jury duty by stating they are not citizens, and on naturalization data from U.S. Citizenship and Immigration Services.

Georgia's 2024 citizenship audit, which the state called its most comprehensive ever, is a useful window into how many sources a serious check pulls from. It reviewed records from county courts, the Georgia Department of Driver Services, and U.S. Citizenship and Immigration Services (Georgia Secretary of State, October 2024). The point is that a citizenship check is not a single gate. It is a set of overlapping records that flag a mismatch for review.

What the audits keep finding

When those checks are run at scale, the result is consistent. Georgia's 2024 review found 20 noncitizens on a roll of 8.2 million registered voters, a rate of roughly two in a million (Georgia Secretary of State, October 2024). Of the 20, nine had a record of ever having voted; the state referred them to prosecutors and opened citizenship questions on 156 additional records (CNN, October 23, 2024).

That finding is not an outlier. A 2016 nationwide study by the Brennan Center for Justice examined 23.5 million votes across 42 jurisdictions and identified about 30 incidents of suspected noncitizen voting, a rate near 0.0001 percent (Brennan Center for Justice, 2016). Read all the way down, the audits point the same direction: noncitizen registration exists, it is caught, and it is rare. That is the honest optimistic read, and it is also the accurate one.

The documentary-proof question, in numbers

The SAVE Act would move from database verification to documentary verification, and here the numbers cut the other way and deserve equal daylight. About 48 percent of Americans hold a valid passport, up from just 5 percent in 1990 (U.S. Department of State, 2025). That is a historic high and a genuine American milestone, and it still means roughly half the country would need a birth certificate or another document instead.

There were 183 million valid U.S. passports in circulation in fiscal year 2025, and the State Department issued 27.3 million that year, the most on record (U.S. Department of State, 2025). The Brennan Center estimates that more than 21 million Americans do not have ready access to the documents a proof of citizenship law would require, a group that includes younger voters, married people whose current name does not match their birth certificate, and rural voters far from a records office (Brennan Center for Justice, 2025). Supporters answer that documents can be obtained and that the standard is worth the cost. Both claims are on the table. Neither is settled by the other.

Where this connects to the rankings

Citizenship verification is one stage of a larger system that US Political Rank scores directly. The count of the ballot itself is traced in how the American ballot gets checked, where 98 percent of 2024 votes left an auditable paper record and 48 states require a post-election audit. The quality of the offices that run all of this is ranked in states by election administration.

The legislative fate of the SAVE Act, meanwhile, is a story about the Senate's 60 vote threshold, laid out in the cloture ranking. A reader who wants to weigh the bill fairly needs both halves of the picture: the rarity the audits document, and the documentary burden the critics measure. Hold them together and the argument becomes a real trade to reason about, rather than a slogan to pick a side of.