Connect with us

Current Affairs

Dead Voters in US Elections: What the Research Actually Finds

Academic studies on dead voters in elections reveal the actual scale differs sharply from viral claims. Peer-reviewed research documents the real prevalence and impact.

Published

on

Dead Voters in US Elections: What Studies Actually Show
Dead Voters in US Elections: What Studies Actually Show
Image: Numbersinstitute (via Wikimedia Commons) (CC BY-SA 4.0)

Claims that large numbers of deceased individuals cast ballots in US elections have circulated for decades, intensifying sharply around every major election cycle. These narratives spread quickly on social media, are amplified by prominent political figures, and often carry an implicit suggestion that dead voters in elections represent a systemic, outcome-altering form of fraud. Yet when researchers have gone back to the primary data — actual voter rolls, death records, and cast-ballot files — the picture that emerges is strikingly different from the one painted in viral posts and campaign speeches. Understanding that gap between perception and evidence is essential for any informed conversation about the integrity of American elections.

Why the Dead-Voter Narrative Gains Traction

The claim that dead people vote is intuitively alarming because it sounds simultaneously plausible and sinister. Voter rolls in the United States are maintained by thousands of separate county and state jurisdictions, each operating under different laws, budgets, and update schedules. When a registered voter dies, the process of removing their name from the rolls depends on timely death-record reporting from county clerks, state vital-statistics agencies, and sometimes the Social Security Administration. Delays in that chain are common and entirely mundane — but they create a surface-level statistic that critics of election administration frequently weaponize.

An NPR report from 2012 highlighted a study finding that approximately 1.8 million deceased individuals remained on active voter rolls across the country. The same report noted that roughly one in eight US voter registrations is invalid or inaccurate in some way — covering not just deceased registrants but also people who have moved, duplicate entries, and clerical errors. That figure sounds alarming in isolation, but the report was careful to note that the presence of a name on a voter roll is categorically different from a ballot actually being cast in that name. Few documented cases of fraud were identified.

This distinction — between registration status and actual voting — is the single most important conceptual point in the entire debate. A dead person remaining on a voter roll is an administrative lag. A ballot cast in a dead person’s name is fraud. Conflating the two, whether deliberately or carelessly, produces a misleading picture of election security.

What Peer-Reviewed Research on Dead Voters in Elections Actually Found

The most rigorous quantitative work on this question comes from academic researchers who obtained and cross-referenced large datasets. A study associated with Stanford’s Institute for Economic Policy Research (SIEPR), conducted by researcher Andrew Hall and colleagues, examined approximately 4.5 million voter records in Washington state covering the period from 2011 to 2018. The methodology involved matching voter participation files against death records to identify cases where a ballot appeared to have been cast after the registered voter’s recorded date of death.

→ Read also: Campaign Ads Use Out-of-Context Votes to Attack Opponents

The result: only 14 possible cases were identified across the entire dataset. That translates to roughly 0.0003 percent of the voters examined. Hall’s summary conclusion, as reported by SIEPR, was that concerns over ballots cast in the name of dead people represent overblown worries about election-day fraud. The study did not claim that zero fraud of this type ever occurs anywhere in the country; rather, it found that the phenomenon is so rare as to be statistically negligible, and certainly not of a scale capable of affecting election outcomes in any jurisdiction studied.

It is worth noting what even those 14 cases represent. In studies of this type, researchers typically acknowledge that a small number of apparent matches may result from data-entry errors, timing mismatches between death-record filing and ballot processing, or ballots submitted by a voter before death that were counted after the recorded death date — a scenario that is legal in many states for mail-in ballots. The actual number of fraudulent ballots, as opposed to administrative anomalies, is likely lower still than the already-tiny figure the study identified.

How Specific Election-Fraud Claims Have Held Up to Scrutiny

Academic research is one lens; investigative fact-checking is another, and both point in the same direction. FactCheck.org published a detailed examination of claims circulating after the 2020 presidential election under the headline Thin Allegations of Dead People Voting. The piece evaluated specific assertions made by supporters of President Donald Trump that named deceased individuals had cast ballots in key swing states. In case after case, the allegations dissolved under examination: some named individuals turned out to be alive; others involved ballots cast by family members with similar names; still others reflected clerical errors in public databases that had been misread by people without training in voter-file analysis.

Snopes similarly rated the claim that dead people voting has swayed US elections as False, after examining the available evidence across multiple election cycles. The pattern across both fact-checking organizations is consistent: specific, named allegations of dead-voter fraud almost never survive contact with primary source documents, and the aggregate numbers involved — even accepting every allegation at face value — fall far short of margins that could plausibly alter outcomes in contested races.

This does not mean election administration is perfect. Voter rolls across the country contain outdated information, and the patchwork of state and local systems makes uniform data hygiene difficult. The Stanford SIEPR summary of Hall’s research is explicit that the administrative problem of outdated registrations is real, even as it concludes that actual fraud exploiting those outdated records is extremely rare.

→ Read also: Political Ads in 2026 Midterms Contain False or Misleading Claims About Opponents

The Gap Between Registration Errors and Actual Fraud

Understanding the mechanics of how a fraudulent dead-voter ballot would actually have to work helps explain why documented cases are so scarce. In virtually every US jurisdiction, casting a ballot requires presenting identification, signing a poll book or envelope that is matched against a signature on file, or both. A would-be fraudster attempting to vote in a deceased person’s name would need to know that the person was registered, know their polling place, present plausible identification, and produce a matching signature — all without being recognized by poll workers who, in many precincts, know their neighbors. Mail-ballot fraud faces a similar set of practical obstacles, including signature verification and, in a growing number of states, additional authentication steps.

None of this means the system is impenetrable. It means the cost-to-benefit ratio of attempting this form of fraud is extremely unfavorable for any would-be fraudster, which aligns with what the data show: it essentially does not happen at scale. Organized, outcome-altering dead-voter fraud would require coordinating hundreds or thousands of such individual acts across multiple jurisdictions simultaneously, with each act carrying criminal exposure — a logistical and legal challenge that no credible evidence suggests has ever been mounted.

What Responsible Election Administration Should Look Like

Acknowledging that dead voters in elections represent an extremely rare form of fraud is not an argument for complacency about voter-roll maintenance. Accurate, up-to-date voter rolls benefit election administrators, reduce the potential for confusion at polling places, and remove any theoretical vulnerability — however small — that outdated records might create. Many states participate in interstate data-sharing compacts that cross-reference death records more efficiently, and automatic voter registration updates tied to vital-statistics reporting are increasingly common.

The legitimate policy conversation is about how to modernize voter-roll maintenance in a way that is accurate, timely, and does not inadvertently purge living voters — a risk that aggressive purge programs have historically created. That is a real administrative challenge worth taking seriously. What the evidence does not support is the claim that deceased individuals voting represents a widespread, organized threat capable of determining election outcomes. The studies that have examined the question most carefully, using the largest available datasets, consistently find the phenomenon to be vanishingly rare.

Elections are complex systems administered by tens of thousands of local officials, subject to extensive legal oversight, post-election audits, and, increasingly, rigorous academic scrutiny. The evidence accumulated across multiple research efforts and multiple election cycles tells a coherent story: voter rolls contain errors, as any large database does, but the leap from an outdated registration to a fraudulent ballot is one that very few people appear to be making. That conclusion deserves as much circulation as the alarming claims it contradicts.

This article was produced with AI assistance and reviewed editorially.