> ## Content Index > Fetch the complete content index at: https://ivn.us/llms.txt > Use this file to discover other available public pages before exploring further. # Think the Election Was Stolen? A Recount Probably Won't Prove It - URL: https://ivn.us/think-the-election-was-stolen-a-recount-probably-wont-prove-it/ - Published: 2026-10-27T16:09:45.000Z - Updated: 2026-10-27T17:13:38.000Z - Description: FairVote's analysis of 7,500 elections over 25 years found recounts changed the outcome just three times. Experts say recounts rarely help trailing candidates and often widen the margin in favor of the winner. - Author: Shawn Griffiths - Tags: 2026 Midterms, Analysis, Election Reform Election Day is almost here. Voters can expect it to be divisive. They can expect it to be contentious. They can expect accusations of cheating. And, they can expect demands for recounts in hotly contested races. [But a report](https://fairvote.org/report/an-analysis-of-statewide-election-recounts-2000-2025/?ref=ivn.us) from the nonpartisan election reform group FairVote offers rare insight that is not often talked about in the national press. Specifically, recounts rarely change the outcome. In fact, they are more likely to add more to the winner’s total. Voters can already see it coming. Democrats are accusing Republicans of [rigging the voting machines in Tennessee](https://www.forbes.com/sites/maryroeloffs/2026/10/23/voters-in-tennessee-say-machines-are-defaulting-to-choose-gop-candidate-marsha-blackburn/?ref=ivn.us). Trump [directed the DOJ to sue](https://www.npr.org/2026/10/25/nx-s1-6016212/blanche-defends-doj-lawsuit-blocking-alaska-senate-results-if-republican-loses?ref=ivn.us) ahead of the US Senate race in Alaska and block the results if incumbent Sen. Dan Sullivan loses. Not to mention the DOJ’s lawsuit to toss 20,000 ballots in California over the allegation that the voters are “likely noncitizens.” The stage has been set for aggressive election challenges and recount demands across the US. Surely, one side lost because the other side had to have cheated. The count must have been wrong. Those counting the ballots are in on it. FairVote’s latest report offers a reality check. "This report shows that US election results are accurate and trustworthy. In nearly 7,500 elections over 25 years, recounts have changed the outcome just three times," Ben Williams, senior policy counsel and researcher at FairVote. > "States should have a clear process in place for recounts to ensure voter trust and confidence – but don't expect recounts to change election outcomes unless the margin is razor-thin." ## Recounts Almost Never Change Who Wins FairVote’s analysis is an updated version of what it has reported in previous election cycles, [which IVN covered in 2024](https://ivn.us/despite-inevitable-calls-election-recounts-research-shows-they-rarely-change-outcome-2024-10/). Researchers examined 7,437 statewide general elections conducted between 2000 and 2025. During that 25-year period, there were just 37 completed statewide recounts. Only three changed the outcome. That is about 8% of races that had a full recount and 0.04% of all statewide elections examined. What’s even more revealing is that the few reversals observed occurred in a race where the original margin of victory was less than 0.06%. ### The Three Elections Where Recounts Actually Changed the Winner **Washington Governor, 2004:** Republican Dino Rossi initially led Democrat Christine Gregoire by just 261 votes. Following recounts, Gregoire emerged victorious by 129 votes out of more than 2.7 million cast. **Vermont State Auditor, 2006:** Republican incumbent Randy Brock initially led Democratic challenger Thomas Salmon by 137 votes. A recount reversed the outcome, giving Salmon a 102-vote victory. Much of the difference involved localities that had incorrectly recorded hand-counted totals. **Minnesota US Senate, 2008:** Democrat Al Franken trailed Republican incumbent Norm Coleman by only 215 votes in the initial tally. Following a recount and a legal dispute over ballots, Franken emerged victorious by 225 votes. Here’s the thing: Oftentimes, recounts can be demanded when the difference between candidates is much larger. Yet these races were not decided by thousands or tens of thousands of votes. They were contests in which a few hundred ballots made the difference. And according to FairVote, no statewide recount has reversed an election outcome since Minnesota's 2008 Senate race. ## Odds Are, Recounts Won’t Help the Runner-Up FairVote’s research found that recounts frequently don’t benefit the losing candidate. In fact, they are likely to do the opposite. A[ ](https://fairvote.org/report/election-recounts-2024/?ref=ivn.us)[previous analysis covering elections through 2023](https://fairvote.org/report/election-recounts-2024/?ref=ivn.us) found that statewide recounts shifted the margin between the top two candidates by an average of just 551 votes, or approximately 0.03% of votes cast for those candidates. And recounts typically widened the vote gap in favor of the winner rather than narrowing it. In other words, they reinforced the initial result that the top performing candidate did, in fact, win. Researchers also found that in 65% of the recounts examined, both sides gained votes. Why? Because recounts often involve ballots that machines initially could not interpret or that election officials must review to determine voter intent. A voter might have made an unusual mark, incompletely filled in an oval, or otherwise marked a ballot in a way that requires human examination. Some votes may be added, others may be removed, and errors in the original tally may be corrected. But these adjustments are generally small, and while they may add to the losing candidate’s vote total, they are just as likely to add to the winner’s as well. ## When a Recount Becomes More About Politics Than Votes Many states have laws that require an automatic recount in the event that an election is decided by a razor-thin margin. However, not every recount is triggered this way. Candidates can often request them, sometimes at their own expense, even when research shows the margin is so large that changing the outcome is highly improbable. FairVote points to several examples of expensive recount efforts that produced little to show for the money spent. In Nevada's 2022 Republican gubernatorial primary, Joey Gilbert spent approximately $190,000 seeking a recount after losing to Joe Lombardo by 11 percentage points. Not 11 votes. 11 percentage points. The outcome didn't change. Lombardo is now the governor of Nevada. In Colorado's 2022 Republican secretary of state primary, Tina Peters paid approximately $256,000 for a recount after losing by 14 points. The result? Peters gained 13 votes. And in Wisconsin following the 2020 presidential election, Donald Trump's campaign paid approximately $3 million for recounts in two counties. Those recounts ultimately resulted in a net gain for Joe Biden. In some cases, the recount was more a political messaging tool than a realistic attempt to change the outcome. A candidate losing by 11-14 points has a near zero chance of winning on recount. But it keeps the allegations of “cheating” and “rigged counts” in the headlines for days, even weeks after an election. FairVote is not saying candidates should be discouraged from challenging genuinely questionable results. However, the evidence supporting a challenge matters. ## Recounts and Election Challenges Are Not the Same Thing It is worth mentioning as Election Day approaches that there is a difference between a recount and an election challenge. A recount examines whether ballots were counted correctly. An election challenge may involve broader legal questions, including whether certain ballots should have been counted in the first place. Voters are already seeing the latter taking place. Disputes over ballot eligibility (like in California), mail ballot deadlines, voter registration procedures, or election administration (like in Tennessee) can become the subject of litigation. This is one reason why FairVote notes in its research that legal challenges can be more consequential than recounts. Because unlike a conventional recount, a successful court challenge could add or exclude hundreds, thousands… even tens of thousands of votes. It is also why FairVote makes it clear that its findings should not be interpreted as proof that election errors, fraud, or administrative problems are impossible. The research measures the outcomes of completed statewide recounts. It does not establish that every election dispute is baseless. The main takeaway is that simply counting the same ballots again is unlikely to reverse an election unless the original margin is exceptionally narrow. Based on its findings, FairVote recommends that states automatically conduct recounts when the margin of victory is within 0.1%. It is a slightly wider margin than what was seen in the three statewide races where recounts changed the winner. But FairVote argues that automatically recounting genuinely close elections helps ensure confidence in the results without forcing candidates to finance the process. At the same time, the organization discourages states from permitting candidate-funded recounts when the victory margin exceeds 1%. For context, in a statewide election with 2 million votes cast between the leading candidates, 1% would be 20,000 votes. If that is what separates the winner from the runner-up, it is highly unlikely a recount will change who won. FairVote also stresses the importance of post-election audits, which examine whether voting equipment and procedures produced reliable results and can help identify problems that a routine recount might not address. ## The Bigger Question: Will Candidates and Voters Trust the Results? Confidence in the electoral process is on shaky grounds with the public. This largely has to do with Republicans and Democrats constantly accusing each other of cheating or suppressing the vote or courting ineligible voters. Recounts this election cycle are inevitable, especially with control of Congress at stake, as well as the possibility that some state offices may flip. Some of these recounts will be justified by extremely narrow margins. Some will be required by law. Others may be driven by candidates unwilling to accept defeat or by campaigns seeking to cast doubt on the process. The challenge before voters will be separating legitimate scrutiny from political theater. There will be both following November 3. One thing is supported by research: Recounts are an essential safeguard in a functioning democracy, but one being triggered or demanded does not, on its own, mean anything nefarious. And, they are not mechanisms for turning a clear defeat into a victory. In this election, the real question isn't whether a recount will change the winner. It's whether the losing side will accept the result when it doesn't.