Editor's Note: This essay is part of the Democracy Project's Elections 2026 series.

Debates over non-citizen voting present a microcosm of the political polarization and dysfunction that have developed over the past few decades concerning election administration issues that otherwise appear to be dry and technocratic.  On the one hand, the Trump Administration is often criticized for exaggerating the scope of the problem.  This past summer, Department of Homeland Security Secretary Markwayne Mullin announced that “more than a quarter of a million noncitizens were registered to vote in four states.”  President Donald J. Trump suggested “the real number is actually much higher” and went on to declare that the “election system [is] so broken and vulnerable that no one can possibly defend it.”  Consistent with such concerns, Republicans have advocated the adoption of new safeguards against non-citizen voting.  Arizona is currently defending its documentary proof-of-citizenship requirement for voter registration before the U.S. Supreme Court. 

On the other hand, Democrats and left-leaning groups have dismissed concerns about potential non-citizen registration and voting on the grounds it is “vanishingly rare” and non-citizens “rarely, if ever, vote.”  Consequently, they have generally rejected efforts to identify potential non-citizens on voter registration rolls as discriminatory “purges” and opposed measures to identify or prevent improper registrations. 

Available empirical evidence demonstrates that thousands, and almost certainly tens of thousands, of non-citizens have appeared on voter registration rolls over the course of the past decade, and many of them never cast a ballot.  These figures reflect only a tiny fraction of one percent of the more than 270 million people currently registered to vote. 

The debate over non-citizen registration and voting shares the same structure as many other political disputes over election administration issues.  Republicans focus on ensuring that non-citizens cannot cast invalid ballots that dilute the weight of other votes and could potentially impact electoral outcomes.  Democrats, in contrast, prioritize minimizing potential barriers to voting to ensure eligible voters are not deterred or prevented from casting ballots.  They generally oppose prophylactic requirements or restrictions absent evidence that the problems they target frequently occur.  Thus, the parties fundamentally disagree over whether an electoral rule’s validity should depend primarily on the relative number of eligible voters who refrain from casting ballots due to the rule, as compared to the number of improper or fraudulent votes the rule prevents.  In the case of non-citizen voting, such utilitarian cost-benefit analysis tilts sharply against the adoption of new safeguards due to the relatively limited number of non-citizens proven to have registered or voted over recent years.   

A well-functioning election system must balance three fundamental factors.  First, every eligible voter must have a reasonable opportunity to cast a ballot safely and have it counted.  Second, the system must minimize the possibility of mistakes, uncertainty, irregularity, illegality, or fraud.  Third, the system must cultivate and preserve public confidence in both the integrity of the electoral process and the accuracy of electoral results. 

Even when an electoral system is functioning well in practice, the absence of safeguards against various harms may reasonably undermine public confidence.  A system may appear vulnerable, insecure, unreliable, or otherwise problematic even if potential harms have not materialized.  Although the frequency with which a particular problem actually occurs is relevant to assessing the adequacy of a system’s safeguards, it is not the only—or even most important—factor.  The desirability of an additional safeguard or requirement may also be impacted by how readily most people can comply with it, the extent to which similar requirements exist in comparable contexts, the accuracy or efficacy of the requirement, the logical connection and fit between the requirement and the underlying goals to be achieved, and more generalized perceptions of “common sense.” 

The Census Bureau estimates that approximately 24.4 million non-citizens currently reside in the United States.  According to the Pew Research Center, this figure is estimated to include nearly 14 million undocumented aliens, almost 10 million of whom participate in the labor force.  Federal law requires states to provide opportunities for voter registration to people applying for driver’s licenses or social programs.  In many jurisdictions, the primary safeguard against non-citizen registration is an applicant’s self-certification of their citizenship status on a checkbox on their registration form.  Under these circumstances, it seems reasonable for voters to be concerned about the possibility of non-citizens being registered, whether through inadvertence, lack of attention when completing the form, indifference, language barriers, processing or programming mistakes, problems stemming from automatic voter registration, or misconduct in third-party voter registration programs, even without regard to the possibility of intentionally fraudulent registrations by non-citizens themselves.

The National Voter Registration Act (“NVRA”) of 1993’s failure to explicitly address non-citizen voting offers additional grounds for potential concern.  First, the statute allows election officials to remove a person from the voter registration rolls only if that person requests removal, is convicted of a crime or declared mentally incompetent, dies, or moves.  Read literally, the NVRA bars states from removing ineligible non-citizens.  Second, the NVRA requires election officials to “conduct a general program” to remove the names of people who become ineligible because they die or move.  Again, the statute does not expressly require election officials to make any affirmative effort to identify non-citizens who have been erroneously registered to vote. 

Third, the NVRA generally allows voter registration forms to require only information that is “necessary” to enable election officials to determine a person’s eligibility to register.  Courts have consistently held that requiring documentary proof of citizenship is not “necessary” because election officials can accept applicants’ self-certifications of their citizenship on their voter registration forms. 

Fourth, lower courts have held that election officials may not conduct programs to “systematically remove” the names of ineligible non-citizens from voter registration rolls for up to 180 days during presidential or congressional election years; the Supreme Court is considering the issue this term.  One reasonable compromise during this “blackout period” would be for states to designate potential non-citizens who appear on the voter registration rolls as “inactive” voters rather than removing them.  With this approach, eligible voters would not be at risk of missing registration deadlines or be required to re-register if they are erroneously flagged as non-citizens.  Rather, they would simply provide confirmation of their citizenship status the next time they cast a ballot or seek to have a provisional ballot counted. 

Federal courts have blocked recent efforts to facilitate the identification of potential non-citizens on voter registration rolls.  The Government has been enjoined from consolidating citizenship information from several federal databases to provide to states, even though states are not under any express legal obligation to do anything with that data.  The Government has likewise been barred from requiring states to provide their complete voter rolls to allow the Government to attempt to identify potential violations of federal law.  Yet state efforts to identify and remove potential non-citizens are often enjoined at least partly because the citizenship information in their databases is not sufficiently reliable.  

An average voter may reasonably conclude that the system as a whole seems indifferent at best to the prospect of non-citizen registration and voting.  Ensuring that elections are open only to members of the American political community should be an uncontroversial, non-partisan objective.  At a minimum, the federal government should be permitted to share the most accurate and up-to-date citizenship information in its possession with states in a convenient manner.  State election officials, in turn, should be required to make a reasonable effort to identify and request confirmation of citizenship from potential non-citizens on their voter registration rolls. 

Even under our current rules, however, candidates should not undermine public confidence in the accuracy of election results through unsupported claims that an election has been “stolen” through non-citizen voting. Generalized concerns about potential non-citizen voting are insufficient grounds for seeking to overturn election results.  Our political process often makes reform impossible unless a problem is presented as a “crisis.”  But participants in a democratic system must help preserve public faith in the process even while recognizing that important opportunities exist for further improvement.  In short, it is reasonable to believe that the system lacks adequate safeguards against non-citizen registration and voting while also acknowledging that, based on current evidence, non-citizen voting is unlikely to occur frequently or impact the outcome of any federal or state election.