By Timothy Frye, Columbia University

Many have speculated how a Trump victory would affect the U.S., but few have thought about the consequences of a Trump loss. After falling behind Hillary Clinton in the polls, Donald Trump has already developed a narrative for his exit: The election was rigged.

So how likely is a rigged vote?

Full-throated claims

Last week Trump told Fox News: “I’m telling you – Nov. 8, we’d better be careful because that election is going to be rigged. And I hope the Republicans are watching closely or it is going to be taken away from us.”

This is not just an isolated or off-the-cuff statement. Trump confidant Roger Stone recently noted: “I think that we have widespread voter fraud, but the first thing that Trump needs to do is begin talking about it constantly.”

Trump’s campaign manager Paul Manafort noted: “Frankly we think that the situation in the country, just like with the DNC’s primaries, is a situation where if you rely on the Justice Department to ensure the security of elections, we have to be worried.”

That President Obama has dismissed these claims as ridiculous will do little to reassure Trump supporters.

The role of good losers

These charges and countercharges are more than just campaign rhetoric. They raise a central issue for democracy: the willingness of losers to comply with a decision reached via free and fair elections.

Political scientists have long identified this willingness as a critical component of American democracy. The most prominent example in recent memory is Al Gore’s refusal to contest the decision of the Supreme Court awarding Florida’s electoral votes to George W. Bush, effectively handing him victory in 2000. Gore could have easily provoked a constitutional crisis by challenging the results.

Interestingly, the claims of Trump supporters echo my research on electoral subversion in nondemocracies.

In Georgia in 2003, Ukraine in 2004 and Russia in 2011, the “losers” challenged the legitimacy of elections held under less than ideal conditions. The “victors” claimed that their rivals were merely sore losers. The losers then took to the streets and forced a showdown with the government. In Ukraine and Georgia the protests led to greater democracy, at least in the short run. In Russia, they resulted in a more autocratic government. These dynamics are not limited to former Soviet states. Look at the large-scale violence that erupted in Ethiopia in 2005, Kenya in 2007 and Zimbabwe in 2008.

Of course, not all cases of electoral fraud lead to protest and a crisis of legitimacy, but research by Joshua Tucker of New York University and Andrew Little of Cornell University suggests that claims of voter fraud are a powerful tool for rallying protest. In addition, they find that calls to protest are especially likely in close elections where voters believe that fraud may have swayed the outcome.

One could argue that Trump has a point. Elections in the U.S. are “rigged” in the sense that they require candidates to raise enormous sums of money and make candidates dependent on donors.

A tough sell

But it is much harder to argue that American elections can be stolen at the ballot box. For all their great expense, elections in the U.S. are remarkably well-run. Voter identification fraud is extremely rare. One study found that individual reports of vote fraud were less likely than reports of alien abduction. Another found that in 2005 prosecutions for migratory bird violations were more frequent than cases of electoral fraud. As a candidate, I’d far prefer to take my chances against an incumbent in the U.S. than in many other countries where ballot box-stuffing, voter intimidation and the banning of political opponents are the rule.

That said, there is much skepticism toward the conduct of elections in some pockets of the U.S. Researchers from Yale University found that 36 percent of respondents in a national sample in 2010 believed that their ballot was not secret. A 2012 survey from Wisconsin found that just under 40 percent of respondents believed that “a few thousand” fraudulent votes were cast in each election.

Whether the “rigged election” narrative will have bite in November will depend on many factors, including the margin of victory, the reaction of establishment Republicans to charges of vote-rigging, the quality of evidence to support the claim and other contextual factors. For example, it is easy to imagine Wikileaks publishing selectively edited emails to discredit the election results. The extraordinary vitriol and whiffs of violence in the air in this election will only amplify this skepticism.

Maybe Donald Trump will win, Hillary Clinton will concede and the rigged election narrative will be moot. Maybe this is all a bluff and Donald Trump will go quietly in the night, but that doesn’t seem likely.

Whatever the result in November, it is not likely to be determined by vote-rigging.

Timothy Frye, Professor of Political Science, Columbia University

This article was originally published on The Conversation. Read the original article.