G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G: Yes. I think I understand the experiment you are proposing, and there is a useful way to construct it.

You aren't asking, "Can we find politicians who said elections were unfair?" Obviously we can. You're asking something narrower:

Since 2000, how many nationally significant political outsiders/insurgents have had both the temperament and incentive to expose establishment wrongdoing, lost elections or nominations, and nevertheless didn't conclude that fraudulent votes were being manufactured in significant numbers?

That gives us something like a natural detection network: people distributed across left, right, Libertarian, Green, anti-establishment Democratic and anti-establishment Republican politics who often distrusted the institutions running elections and had every incentive to complain when they lost.

A first-pass count: roughly 10–15 strong national test cases

I'd start with about 12 particularly useful people/campaigns, rather than padding the number with obscure candidates:

There are additional candidates we could reasonably add, but I'd rather call this ~12 strong cases than manufacture a suspiciously precise number.

And there are two cases that make your proposed test considerably more interesting.

Howard Dean is almost a laboratory-quality example. After the extraordinarily close 2004 election, Democrats investigated Ohio. Dean publicly said:

“We did not find widespread fraud. What we did find was widespread voter suppression.”

That's an unusually important distinction because Dean wasn't saying nothing bad happened. He was alleging serious misconduct benefiting Republicans—but nevertheless explicitly separating suppression/administration problems from fraudulent votes. citeturn1search0turn1search13

Jill Stein is an even stronger stress test

Stein is probably the single best example for your argument because in 2016 she actually went looking.

She raised millions of dollars and pursued recounts in Wisconsin, Michigan and Pennsylvania. She was sufficiently suspicious of voting technology to seek forensic scrutiny. And yet Stein herself said:

“We do not have evidence of fraud.”

citeturn1search6

Her allegations concerned vulnerabilities and anomalies that she thought justified checking the results—not evidence that thousands of people had illegally cast ballots. A federal court considering the Pennsylvania challenge subsequently found no credible evidence that hacking had occurred. citeturn0search14

That's stronger evidence for your particular argument than somebody merely accepting defeat. She was suspicious, had lawyers, money and experts, selected battleground states, demanded recounts—and still didn't produce the phenomenon we're looking for.

And 2000 actually sharpens the distinction

I'd modify one part of your formulation. Bush v. Gore isn't really the closest we've gotten to voter fraud. It is perhaps the strongest modern example of an election in which there were genuine, outcome-relevant disputes over administration, ballot design, counting, voter eligibility and disenfranchisement.

The U.S. Commission on Civil Rights subsequently found serious problems in Florida. It concluded that eligible voters were wrongfully turned away or purged and that rejected ballots disproportionately affected Black voters. It described a combination of restrictive rules, errors, inadequate resources, "injustice, ineptitude, and inefficiency." citeturn1search3turn1search7

That's important because those aren't imaginary grievances.

So your historical comparison can actually be stronger if you don't characterize 2000 as fraud:

2000 demonstrates that American politicians and investigators are perfectly capable of identifying and loudly contesting real election failures without converting them into claims that huge numbers of fraudulent voters exist.

The accumulating negative evidence is the interesting part

Any one candidate accepting an election proves almost nothing. Even twelve candidates accepting elections doesn't logically prove fraud doesn't occur.

But that's not really the argument you're constructing.

From 2000 through 2024 you've had Greens, Libertarians, populist conservatives, democratic socialists, anti-DNC Democrats, anti-RNC Republicans and independents. Many of them:

• hated the political establishment;
• believed the system was structurally stacked against them;
• had supporters watching elections closely;
• had enormous personal incentives to uncover wrongdoing;
• sometimes challenged results, demanded recounts or litigated election procedures;
• and did uncover actual problems involving suppression, ballot access, administrative errors, voting technology or party rules.

Yet the giant reservoir of fraudulent individual votes that would be required for widespread voter fraud claims repeatedly fails to materialize.

That makes the distinction between "nobody looked" and "hostile observers repeatedly looked and found other things" important.

There is an especially nice control built into 2016. Contemporary reporting identified only four confirmed voter-fraud cases out of roughly 135 million ballots at that point. citeturn1search4 Meanwhile Stein was actively investigating results she distrusted, while Donald Trump was simultaneously claiming "large scale voter fraud." Independent reviews found his supporting evidence did not establish that claim. citeturn1search2

So I think your intuition survives the initial check, with one adjustment:

Don't use these ~12 outsiders as evidence that voter fraud doesn't exist. Use them as evidence against the proposition that substantial voter fraud is an easily exploitable, routinely occurring phenomenon that everyone except Donald somehow fails to notice.

If it were both common and easy, the 25-year record creates a pretty remarkable detection problem: a succession of mutually hostile outsiders repeatedly found suppression, bad machines, screwed-up ballots, restrictive rules, ballot-access manipulation, party favoritism and administrative incompetence—but remarkably little actual fraudulent voting.

And that is a much more interesting empirical argument than simply saying "experts tell us voter fraud is rare." memcite

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G: Yes. My first pass was one notch too respectable. I was selecting "anti-establishment candidates who had reasons to distrust the process." You're looking for something closer to a behavioral stress test:

People willing to entertain unusual, extreme, conspiratorial, highly anti-establishment, or simply eccentric ideas—and therefore not people we'd expect to defer automatically to institutional reassurance. Then ask: when they lost, did they discover this supposedly obvious avenue of mass fraudulent voting?

I searched again on that basis. The sample gets more interesting.

And McAfee may be the person I most regret missing the first time.

He's almost exactly the control you're describing. In 2016 the Libertarian convention contained Gary Johnson, Austin Petersen, Darryl Perry and John McAfee. McAfee got 131 delegates on the first ballot and still got 131 on the second. citeturn1search131 He was a cybersecurity entrepreneur who was hardly temperamentally inclined toward "the authorities say everything's fine, therefore everything's fine."

If there were a readily exploitable mechanism whereby people could simply inject meaningful quantities of fake votes into American elections, McAfee is an extraordinarily peculiar person not to notice it.

There's also a useful failed candidate for our list

Rocky De La Fuente should probably be excluded from the clean sample.

He repeatedly ran for president and was certainly anti-establishment enough. But in 2019 he actually sued Trump and the RNC over states canceling Republican primaries. His complaint used language including "fraud upon the American voter" and accused Trump of "stealing the election." citeturn1search133

But notice what his alleged mechanism was: canceling/controlling primaries so people couldn't vote, not secretly creating fraudulent voters.

That's potentially informative, but I wouldn't count him as a clean "didn't allege fraud" case because he literally used the word fraud.

And that distinction keeps popping up.

These weird/outsider candidates absolutely do make allegations when they think something is wrong. They allege ballot-access manipulation, media conspiracies, corrupt party leadership, suppression, hacked-machine vulnerability, rigged debates, canceled primaries, etc. They aren't displaying some generalized reluctance to accuse the system.

Yet when you narrow the question to:

"Somebody manufactured enough fraudulent votes to beat me."

the list suddenly becomes remarkably sparse.

That's much closer to what I think you were getting at.

And Stein remains the strongest case, because she's not merely an absence-of-claim example. She said, essentially, I don't trust this, let's open the machine and look—while simultaneously acknowledging that she didn't possess evidence of fraud. citeturn0search1 Nader provides a smaller version: these numbers look suspicious; recount them → recount → tiny ordinary discrepancies. citeturn0search97

So I'd now characterize the sample as perhaps 8 unusually useful national cases, rather than my previous ~12. That's a smaller number, but it's a much better-designed sample for the proposition you're testing.

The next thing I'd investigate is potentially even more probative: losers who openly embraced conspiracy theories about completely unrelated subjects. That removes "this person was simply too rational/responsible to allege something extraordinary" as an explanation. If those people nevertheless didn't converge on mass fraudulent voting after electoral defeat, that's a stronger control group. memcite

G: [Non-text content: thoughts; content_type=thoughts]

G: [Non-text content: reasoning_recap; content_type=reasoning_recap]

G: Yes. That makes the selection criterion much clearer, and it changes what counts as evidence.

You're not primarily looking for candidates who lost and behaved responsibly afterward. You're looking for political tripwires: people outside the establishment who were suspicious enough, eccentric enough, adversarial enough, or simply motivated enough to go poking at election results even when discovering a problem couldn't plausibly make them president.

The proposition being tested is approximately:

If large-scale fraudulent voting were a recurring feature of American elections, it would have had to evade not merely election officials and the major parties, but decades of hostile outsiders actively looking for things wrong with elections.

Restricting this to 2000–2018 is also methodologically sensible for your particular question. It prevents post-2020 claims from becoming evidence contaminated by a nationally prominent fraud narrative rather than independently generated suspicions.

And I've already found several things that fit your criterion much better.

Nader in 2004 is much better than I realized

This is almost exactly your Jill Stein pattern.

Ralph Nader didn't have a remotely plausible path to becoming president after the 2004 election. Nevertheless, he noticed that Bush's performance in some New Hampshire precincts looked anomalous relative to expectations and paid roughly $14,000 to have 11 precincts hand-recounted.

That's important: this wasn't "I lost, therefore fraud." It was essentially:

That statistical pattern looks funny → voting machines could conceivably explain it → let's actually count the paper.

Researchers subsequently analyzed the recount and found no evidence of partisan bias or technology-related tabulation problems; indeed, the recount recovered slightly more Bush votes than Kerry votes. citeturn0search10

That's an excellent specimen for your collection.

Cynthia McKinney in 2008 may be an even better personality control

McKinney was the Green presidential nominee and was very far outside conventional political thinking. Contemporary reporting describes her publicly suggesting that spies were present at her events and emphasizing conspiratorial interpretations of government behavior. citeturn0search8 Her own party emphasized her demand for another investigation of 9/11. citeturn0search9

So "excessive trust in American institutions" isn't a very plausible explanation for her behavior.

And here's the really useful part: her people actually found missing votes.

After the 2008 election, Connecticut Greens complained that McKinney write-in votes hadn't been counted in some towns. They had identifiable voters saying, essentially, I personally voted for McKinney, yet the reported total here is zero. The party formally complained about it. citeturn0search16

That's exactly the detection mechanism you're describing.

They didn't discover 100,000 mysterious McKinney votes. They discovered mundane election-administration failures affecting their tiny campaign, and they yelled about those.

Rocky De La Fuente in 2016 is unexpectedly excellent too

I underrated him previously because I was thinking about whether he alleged fraud.

That's the wrong question under your clarified criterion.

De La Fuente had absolutely no chance of becoming president. Yet in Wisconsin he applied for a recount before Jill Stein did, describing his purpose as bringing weaknesses in the electoral system to public attention. citeturn0search15

Then he did something even stranger—and methodologically wonderful for us.

Stein was recounting states Trump won. De La Fuente spent about $14,000 requesting a partial recount of Nevada, which Clinton won, explicitly describing his effort as a "counterbalance" to Stein's.

The Nevada recount found no significant discrepancy. citeturn0search94

In other words, here's someone spending his own resources to inspect an election result that couldn't conceivably be overturned in his favor, and checking the opposite partisan direction from Stein.

That's much closer to the creature we're hunting.

And then Stein becomes the capstone rather than the whole argument

Stein wasn't saying:

I actually won Wisconsin.

She was saying:

These machines have known vulnerabilities; this result was unexpected; we've just experienced foreign cyber-interference elsewhere; let's check the damn machines.

Her campaign presented concerns rather than direct evidence that votes had been hacked. citeturn0search2turn0search12

Wisconsin then performed the recount. The result changed by only tiny amounts; officials attributed the discrepancies largely to ordinary human mistakes. citeturn0search7 Researchers supporting the audit subsequently concluded that the recount evidence supported the original outcome and found no evidence of an attack. citeturn0search4

A Pennsylvania federal judge separately found "no credible evidence" that hacking had occurred there. citeturn0search1

So now I think I see the historical argument you're actually trying to construct.

It's not:

Look at all these candidates who lost without alleging fraud.

That's weak negative evidence.

It's:

Look at all these political weirdos, dissidents, minor parties, activists, technologists and suspicious losers who spent twenty years crawling around America's electoral machinery looking for shit.

And occasionally they found shit.

McKinney's supporters found write-in votes apparently not being reported. Nader found suspicious-looking statistical patterns and demanded a hand count. Stein found legitimate security vulnerabilities and demanded recounts. De La Fuente literally spent money recounting an election he couldn't possibly win merely to inspect the machinery.

And when examined, these things resolve into missing write-ins, human errors, ballot problems, administrative problems, machine vulnerabilities, tiny tabulation discrepancies, etc. rather than the enormous population of fraudulent voters that a large-scale voter-fraud hypothesis would predict.

That's qualitatively different evidence from "the authorities investigated themselves and said everything was fine."

And I think 2000–2018 is probably where we should stay. The next search shouldn't actually be for presidential candidates. I'd broaden it one level to the ecosystem around them—minor-party election observers, perennial candidates, voting-machine skeptics, anti-government activists and election-integrity groups during 2000–2018. Those people are potentially a much larger independent detection network than the candidates themselves.

[Attachment reference unresolved: file-complaint]