Advocates of wide-scale voter fraud in American elections ask how can you even identify voter fraud, let alone prosecute it?
It seems to me you would pursue and prosecute voter fraud the same way you would other kinds of fraud. I find Lorraine Minnite’s 2010 book The Myth of Voter Fraud dispositive. This book is about as rigorous as this topic gets. Minnite did what serious empirical work requires: she went looking for the fraud rather than starting from a conclusion, traced the specific claims, checked the underlying cases, and found that the gap between the rhetoric and the documented reality is enormous. The book is persuasive precisely because it takes the allegation seriously enough to investigate it rather than dismiss it on priors.
The persuasiveness of the messenger matters in David Pinsof Alliance Theory terms, but it also matters in ordinary epistemic terms. When you look at the people who have most aggressively pushed widespread fraud claims, the pattern is consistent. The claims emerge after losses, not before or during. The evidence offered repeatedly fails under scrutiny, including scrutiny by allies. The legal strategy of filing dozens of cases and losing almost all of them on evidentiary grounds is not the behavior of people who believe they have solid evidence. It is the behavior of people doing political work under legal cover.
This connects to something Turner would recognize. The fraud claim is not primarily an empirical claim seeking verification. It is a coalition signal performing a specific function, keeping a base mobilized, delegitimizing future losses in advance, and justifying procedural interventions that benefit one side. The claim does not need to be true to do that work. It needs to be believed by enough people to serve the coalition’s purposes, and the buffered self of the true believer does the rest, filtering contrary evidence as corrupt and treating the absence of proof as proof of how deep the conspiracy runs.
Minnite’s book is the empirical ground. The dubious voter fraud messengers is the Alliance Theory layer on top of it.
Fraud prosecution in general does not require proving the full scale of a scheme before investigating. You follow specific credible allegations, gather evidence, bring cases where the evidence supports them, and the pattern of prosecutions over time gives you a reasonable picture of the scope of the problem. This is how the justice system handles insurance fraud, securities fraud, tax fraud, and every other category where the full universe of violations is never directly observable.
Applied to voter fraud, the question becomes: are credible specific allegations being investigated and prosecuted when evidence supports them? The answer is largely yes. Election officials, state attorneys general, and the Justice Department do pursue documented cases. The Heritage Foundation database, whatever its limitations as a comprehensive count, represents prosecutions and convictions. The system is working roughly as it works for other fraud categories.
What the 2020 fraud claims required was something different and much harder to establish. Not a pattern of individual violations but a coordinated scheme operating across multiple states simultaneously, involving thousands of election workers from both parties, producing no whistleblowers and no documentary evidence, and leaving no statistical trace detectable by the losing side’s own monitors. That is not an ordinary fraud allegation that prosecutors failed to pursue. It is an extraordinary conspiracy claim that failed every evidentiary test applied to it, including by judges appointed by the president making the claim.
The normal fraud framework resolves the dispute more cleanly than the political framing does. Individual violations exist and get prosecuted. The conspiracy claim is a different category of assertion entirely and was treated accordingly by the courts.
The evidence against widespread voter fraud in the US is about as solid as evidence gets in political science. Extensive studies, Republican-led investigations after 2020, court cases that failed for lack of evidence, secretaries of state from both parties confirming the same finding. The Heritage Foundation’s own database of documented fraud cases, assembled by people motivated to find it, covers a few thousand instances over decades of hundreds of millions of votes cast. The numbers do not support a story of systematic fraud.
But the belief in widespread fraud is a perfect case study in the convenient beliefs framework. It is not primarily about evidence. It is a coalition coordination device. It explains away losses without requiring the coalition to update its self-image. It justifies procedural interventions that benefit one side. It generates the emotional energy of grievance and persecution that holds a coalition together between elections. The belief does work that has nothing to do with its truth value.
What makes it interesting from a Stephen Turner angle is the mirror problem on the other side. The belief that voter fraud concerns are entirely bad faith, that there is nothing worth examining, is also doing coalition work. It signals membership, it avoids the awkward concession that any complex system has irregularities worth monitoring, and it lets the believer feel epistemically superior without doing the harder work of distinguishing legitimate procedural concerns from cynical ones.
Both sides have convenient beliefs about fraud. They just serve different coalitions.
If fraud is sophisticated it leaves no easily recoverable trace. A fraudulent vote looks identical to a legitimate one once cast. Chain of custody problems in mail balloting are real in the sense that verification at the receiving end cannot fully reconstruct what happened at the sending end. Non-citizen registration, which does occur in small numbers due to administrative errors at DMV offices, is genuinely hard to detect because the databases that would reveal it are not consistently cross-referenced. The argument that absence of evidence is not evidence of absence has logical standing in any domain where detection is imperfect.
The harder question is whether the detection mechanisms we have are sensitive enough that large-scale fraud would show up anyway. Most election security researchers think yes. Coordinating fraud across thousands of precincts, involving hundreds of poll workers and election officials from both parties, without producing whistleblowers, documentary evidence, or statistical anomalies detectable by the losing side’s own monitors, is an organizational problem of enormous difficulty. The 2020 post-election period was the most scrutinized election in American history, with highly motivated searchers on one side. What they found was vanishingly small.
So the epistemically honest position is something like: the claim that fraud is hard to detect is structurally valid as a methodological point, but the scale of fraud required to change outcomes in multiple states simultaneously would not be hard to detect, and was not detected. The argument works as a reason for reasonable vigilance. It does not work as a reason to believe the specific claims made about 2020.
Turner would note that both the fraud believers and the fraud dismissers have stopped their inquiry at the point most convenient for their coalition.
The case for concern is structural rather than evidentiary. In any system where identity is not verified at the point of transaction, the theoretical vulnerability exists. You cannot know what you did not check. Most fraud detection in voter rolls is after the fact, cross-referencing databases for anomalies, which means some categories of fraud might not surface until long after an election if at all. The honest version of this concern is not that fraud is rampant but that the absence of ID requirements creates an unaudited channel whose integrity rests entirely on the honesty of voters and the accuracy of registration rolls.
The case against the concern is also structural. Impersonating a registered voter in person requires knowing who is registered, showing up before they do, and risking a felony charge for one vote. The cost-benefit calculation is extraordinarily unfavorable for any rational actor. Coordinating enough such fraud to matter in a competitive race would require an organizational apparatus that would be far easier to detect than the fraud itself. Minnite’s research found that documented in-person impersonation fraud is vanishingly rare precisely because the incentive structure makes it an implausible strategy.
The more legitimate fraud concern in a no-ID system is probably not in-person impersonation but registration fraud, votes cast in the names of dead or moved voters, or systematic harvesting operations. Those vulnerabilities exist somewhat independently of ID requirements and are better addressed through roll maintenance and chain of custody procedures than through ID checks at the polling place.
The honest bottom line is that no-ID systems have theoretical vulnerabilities whose exploitation rate appears to be very low based on available evidence, but whose true rate is partly unobservable by design. That is a real epistemic limitation worth acknowledging even if the fraud hypothesis remains unproven.
- https://PayPal.Me/lukeisback
"Luke Ford reports all of the 'juicy' quotes, and has been doing it for years." (Marc B. Shapiro)
"This guy knows all the gossip, the ins and outs, the lashon hara of the Orthodox world. He’s an [expert] in... all the inner workings of the Orthodox world." (Rabbi Aaron Rakeffet-Rothkoff) LATEST POSTS:
- What is a ‘Received Idea’?
- Jordan Bardella: The Manufacture of Normality
- Everyone Became Television: Bourdieu’s Warning and the 2026 Iran War
- Marine Le Pen
- The Coalition-Proximity Rule
- Nigel Farage
- Bernard Haykel: A Life Between the Text and the Gun
- Walker Connor (1926-2017)
- Benedict Anderson and the Nation as Imagination
- Anthony D. Smith: The Student Who Kept the Question and Rejected the Answer
- Ernest Gellner
- Eric Kaufmann: The Man Who Made the Majority Visible
- Dominic Cummings: A Biography
- Steve Lopez: The Last City Columnist
- California Historian Kevin Starr
- Stephen Kotkin: A Life in Power
- William T. Vollmann: An American Life in Excess
- Rod Dreher: A Life in Exile
- The Cross at Sinjar: Tom Holland’s Dominion
- Rick Warren: A Biography
BEST POSTS:
- * The Enlightenment Wasn’t Enlightened (6-23-26)
* Mr. Burge Draws The Line (6-23-26)
* 'Improving on Democracy' (6-17-26)
* People Leak To People Who Are Fun (6-11-26)
* Why Does Australia Produce So Many Great Journalists? (6-11-26)
* Steve Wynn and the Press: Power, Litigation, and the Contest Over Las Vegas (6-3-26)
* Sheldon Adelson and the Journalists (6-3-26)
* The Vigilant Animal: Thinkers Who Reject the Myth of Human Gullibility (6-2-26)
* The Cost of Refusing the Misunderstanding Myth (6-2-26)
* Show Me How It Travels (6-2-26)
* The Norm Explainers (6-2-26)
* Centering Marginalized Voices (6-1-26)
* What would it look like if the Washington Post put its reader first? (6-1-26)
* What would it look like if the Financial Times put its reader first? (6-1-26)
* What It Would Mean for the Los Angeles Times to Put the Reader First? (6-1-26)
* What It Would Mean for The New York Times to Put the Reader First? (6-1-26)
* Why Wembanyama Lives on the Perimeter (5-31-26)
* The Emotional Palettes Of San Diego, Los Angeles, San Francisco & Sacramento (5-27-26)
* The Administrative Capital: Sacramento Legal Culture (5-27-26)
* San Diego - The Quiet Republic (5-27-26)
* The Quiet Bar: San Diego Legal Culture (5-27-26)
* SF v LA Legal Culture (5-27-26)
* Why Talent Travels Poorly Between San Francisco and Los Angeles (5-27-26)
* San Francisco and Los Angeles as Rival Models of Urban Access (5-27-26)
* Social Cliques in New York, 2026 (5-25-26)
* Social Cliques in San Francisco, 2026 (5-25-26)
* The Rival Courts of Washington (5-25-26)
* The City of Private Rooms (5-25-26)
