DMV Glitch Floods Voter Rolls

Voters filling out ballots at polling booths — Photo: Rob Crandall / Shutterstock

When a routine DMV transaction quietly turns thousands of ineligible residents into “registered voters,” the story is not just about one state’s glitch—it is about how modern election systems succeed or fail at the basic task of guarding the franchise while remaining usable for millions of ordinary people.

Key Points

New Jersey’s Motor Vehicle Commission software error automatically registered roughly 6,600 self-identified noncitizens to vote between June 2023 and June 2024, despite their indicating they were ineligible.
State reviews show that fewer than 400 of those improperly registered individuals actually cast ballots, with no evidence so far that any election outcome was changed.
The incident has triggered a federal civil-rights investigation, partisan clashes over noncitizen voting, and renewed scrutiny of DMV-based automatic voter registration systems.
Research across states indicates that noncitizen voting remains exceedingly rare relative to total ballots cast, but administrative and software errors do occur and can be politically weaponized.
The New Jersey case exposes weaknesses in vendor oversight, error detection, and public communication—and offers lessons for election administrators nationwide.

What Happened in New Jersey’s Noncitizen Registration Case

In mid-2026, New Jersey’s governor Mikie Sherrill disclosed that a “serious software error” in the state’s Motor Vehicle Commission (MVC) system had improperly registered about 6,600 people who had explicitly indicated they were not U.S. citizens when applying for driver’s licenses or state identification cards. The affected period ran from June 2023 through June 2024, during the prior gubernatorial administration, and coincided with the MVC’s use of a contracted vendor’s software to support automatic voter registration.

Under New Jersey’s system, eligible residents interacting with the MVC can be automatically added to the voter rolls unless they opt out. In this case, the problem was not a lack of an opt-out; it was the software’s failure to respect a clear “no” answer to the citizenship question. Applicants used a keypad to answer whether they were U.S. citizens; for roughly 6,600 individuals who pressed “no,” the system nonetheless transmitted voter-registration data downstream, effectively flagging them as eligible.

Subsequent analysis by the state found that fewer than 400 of those improperly registered noncitizens had actually voted in at least one election. Those ballots were distributed across party lines—Democrats, Republicans, and unaffiliated voters—and spanned multiple jurisdictions. The governor and state election officials emphasized that they had “no evidence at this time that any elections were swayed” by these votes, but acknowledged that every improper ballot is a legal violation and a failure of system safeguards.

Once the error was uncovered, the administration moved to remove all erroneously added registrations from the voter rolls, replace the software vendor, and commission an independent investigation into how the glitch occurred and why it persisted for a year before being corrected. The MVC’s new leadership began transitioning to a different vendor and notified affected individuals that their voter registrations were being cancelled because they were ineligible to vote.

From Software Glitch to Civil-Rights Investigation

The New Jersey incident did not remain a state-level administrative story for long. The U.S. Department of Justice opened a formal investigation into what it described as the “unlawful dilution of American votes,” focusing on how noncitizens came to be registered and how many of them cast ballots in federal elections. DOJ gave New Jersey five days to provide detailed voter data tied to the 6,600 registrations and subsequent voting activity, signaling that it views the matter as a potential violation of federal voting-rights statutes.

Federal scrutiny was reinforced by the Department of Homeland Security, whose Secretary Markwayne Mullin warned publicly that as many as 35,000 noncitizens might still be on New Jersey’s voter rolls, based on prior federal data-matching exercises. That larger number is not yet validated by state-level record review, and election researchers have repeatedly found that broad database matching against immigration files tends to overstate noncitizen registration because of outdated records, name mismatches, and status changes over time. Nonetheless, the claim has amplified calls from national Republicans for aggressive purges of alleged noncitizen voters.

At roughly the same time, federal prosecutors in New Jersey charged four individuals with illegal voting and false statements on citizenship forms, alleging they were not citizens when they registered and voted in federal elections. Separate from the MVC software glitch, these criminal cases underscore that the state has seen both inadvertent administrative errors and instances where prosecutors believe they can prove intentional violations of federal law.

On the political front, the governor’s admission has been seized upon by former President Donald Trump and his allies as evidence that noncitizen voting is widespread and that blue states cannot be trusted to run clean elections. In public remarks, Sherrill has pushed back, arguing that Trump has “zero credibility on the issue of election integrity” and that her administration is demonstrating transparency and accountability by disclosing the error, ordering an independent investigation, and cleaning the rolls.

How Automatic Registration Systems Can Fail

To understand how 6,600 ineligible residents ended up on the voter rolls, it helps to look at the architecture of DMV-based automatic voter registration. In states like New Jersey, when a person applies for or renews a driver’s license, the transaction system captures core identity data—name, address, date of birth—and then either offers voter registration or, under automatic registration, transmits data to election officials by default unless the applicant declines. Citizenship status is often captured with a simple yes/no question, via touchscreen or keypad, and is supposed to control whether any voter-registration record is created.

In the New Jersey case, the vendor’s software appears to have misinterpreted or ignored the “no” citizenship input for a subset of transactions, effectively treating those applicants as eligible despite their self-reported ineligibility. The problem persisted for roughly a year before being patched, indicating weaknesses in quality assurance, monitoring, or both. There is no indication in the reporting that local election offices had visibility into the citizenship answers; they simply received registration records that, on their face, appeared to meet eligibility requirements.

New Jersey is not alone in facing this kind of problem. California’s DMV-based automatic registration program experienced early glitches that added at least 1,500 ineligible individuals to the rolls in 2018, six of whom voted in midterm elections before the errors were identified and corrected. Election administrators and researchers note a recurring pattern: when states implement automatic or streamlined registration tied to DMV data, small but significant bugs in the integration can create erroneous registrations, typically involving noncitizens or people with status changes that the system does not accurately capture.

These errors are almost never the result of noncitizens deliberately trying to register or vote. Studies by organizations such as the Brennan Center and academic researchers have found no documented cases of noncitizens intentionally registering or voting while knowing they were ineligible; instead, the cases uncovered tend to involve misunderstandings, clerical mistakes, or software failures. In New Jersey, the governor expressly stated that the individuals affected answered “no” to the citizenship question and that the system registered them “through no fault of their own.”

How Rare Is Noncitizen Voting, Really?

The New Jersey numbers—6,600 improper registrations and fewer than 400 votes—sound large in isolation. Put next to the total size of the electorate, they look very different. New Jersey has millions of registered voters; those 6,600 represent a tiny fraction of a percent of the statewide rolls, and the 400 ballots constitute roughly one hundredth of one percent of total votes cast in a major statewide election.

That scale is consistent with what comprehensive state reviews have found elsewhere. A 2012 Michigan investigation that initially flagged 4,000 possible noncitizens on the rolls ultimately identified only a handful of cases where noncitizens had actually voted. Louisiana’s review of nearly 3 million voters found 79 individuals believed to be noncitizens who voted, out of tens of millions of ballots cast across years. Pennsylvania reported 544 noncitizen votes over 17 years, out of about 93 million ballots—a literally minuscule fraction.

More recent syntheses, including work by the Brennan Center and the Fair Elections Center, conclude that noncitizen voting is “vanishingly rare” and that federal and state eligibility checks work as intended in the overwhelming majority of cases. One multi-state analysis of voter rolls run through federal immigration databases identified roughly 4,200 suspected noncitizens out of 35 million registered voters—about 0.01 percent—and even that group included many citizens misflagged due to database errors.

This backdrop matters when evaluating claims that the New Jersey incident proves widespread noncitizen voting. It demonstrates that systems can fail and that each improper registration and vote is significant from a legal and moral perspective. But it also shows that, at the population level, such incidents are rare and typically attributable to administrative rather than criminal intent.

Political Weaponization Versus Administrative Accountability

Because noncitizen voting touches immigration, partisanship, and the legitimacy of elections, even small numbers can be politically explosive. In New Jersey, Republican leaders and national conservative media quickly framed the MVC error as evidence of a broader “noncitizen voter scandal,” pointing not only to the 400 improper ballots but also to separate allegations from party task forces that “hundreds” of noncitizens remain on the state’s voter rolls and that at least 30 have illegally voted over many years.

Trump-aligned figures have invoked the case to support federal legislation like the “Save America Act” and to justify threats of withholding federal funds from states that do not conduct aggressive noncitizen purges. Commentators on programs such as Timcast IRL argue that even 400 improper votes could swing close local races and that the New Jersey error is evidence of systemic rot—dead voters on the rolls, felons voting improperly, and lax mail-ballot controls.

Election-law scholars and voting-rights advocates take a different view. They argue that while every improper vote matters, responses must be proportionate and grounded in evidence: targeted cleanup of proven errors, better software testing, clearer procedures at DMVs, and strong but fair eligibility checks when people register. They warn that expansive purges based on imperfect database matching can wrongly strip thousands of citizens of their right to vote, and that broad narratives of “widespread fraud” are often built on unverified numbers or misinterpreted data.

The New Jersey governor’s approach illustrates that tension. Sherrill has condemned the error as “unacceptable,” ordered vendor replacement and an independent investigation, and committed to removing ineligible registrations—all steps aligned with administrative accountability. At the same time, she has refused to share detailed voter data with federal authorities beyond what law requires, citing concerns that it could be misused for immigration enforcement against individuals who did nothing wrong and that it would fuel politicized narratives rather than constructive reforms.

Lessons for Election Systems and Voter Trust

For readers trying to assess what this case means beyond New Jersey, several lessons stand out. First, automatic voter registration through DMVs is not inherently flawed; when properly implemented, it expands the rolls, reduces paperwork, and standardizes data. But the integration between identity systems and election databases is unforgiving: a single mis-coded field or misinterpreted Boolean—citizen versus noncitizen—can silently propagate thousands of errors before anyone notices.

Second, robust vendor oversight and continuous monitoring are critical. New Jersey’s governor has criticized both the software company and the prior administration for allowing the glitch to persist for a year without public disclosure or apparent remediation. That critique is more than political positioning; it highlights the need for routine audits of DMV voter-registration transmissions, test environments that catch logic errors before deployment, and clear escalation paths when anomalies surface.

They Swore Noncitizen Voting Was a Myth … Then New Jersey Admitted 6,600 Registrations, 400 Illegal Ballots, and a Three-Year Cover-Up https://t.co/zOQ9Hdr40F pic.twitter.com/OGcQuXD4Wh

— Flopping Aces (@FloppingAces) July 23, 2026

Third, communication matters. In many of these cases, the noncitizens improperly registered have reason to fear immigration consequences if their names appear in public investigations. New Jersey’s decision to remove them from the rolls while explicitly stating that they did nothing wrong—emphasizing that the fault lay with software—can mitigate that fear. Conversely, broad public rhetoric equating administrative errors with criminal “fraud” risks discouraging eligible citizens, particularly in immigrant communities, from participating in elections at all.

Finally, the New Jersey incident underscores that the health of an election system should be judged on two axes at once: the rate of error and the response to error. A system with zero reported problems but no internal checks may simply be blind; one with rare but well-documented glitches, promptly investigated and transparently corrected, can maintain and even strengthen public trust. In this case, the discovery of 6,600 improper registrations and roughly 400 illegal ballots is serious, but it sits within a broader empirical reality: noncitizen voting in the United States remains exceptionally rare, and where it emerges, it almost always points back to human or software error rather than organized schemes.