After trailing Harris in the betting markets for 26 consecutive days, Trump first pulled ahead Sunday before getting a bigger boost Monday. As of Wednesday, the markets give Trump a 51.3% chance to win the White House, compared to 47.3% for Harris.

Just a couple of weeks ago, Trump was trailing by 7.3%, but he cut his deficit to 2.5% by the start of October. On Sunday, Trump pulled 0.1% ahead. He shot up to a 3.5% lead Monday, corrected down to a 3.2% lead Tuesday, and bounced back up to a 4% lead Wednesday.

After being heavily favored against Biden in the betting markets, at one point leading by 48.2%, Trump initially led Harris by a wide margin until she started gaining steam at the end of July. Since Aug. 8, Harris had led Trump with bettors for the vast majority of the race while maintaining a two-to-three-point lead in the polls.

Why are the sharps switching sides?

This is not the first time Trump has erased a large betting odds deficit to Harris, as he briefly overcame an 8.8% hole in August. He eventually surged to a 5.2% lead before the Sept. 10 presidential debate, which immediately tanked his odds, resulting in him trailing for 26 days consecutively.

While Trump’s national polling numbers haven’t changed in any meaningful way, as he has been steadily two to three points behind Harris, the swing state landscape has sharply put its money behind the former president in recent days.

There are seven swing states within two polling points for either candidate, and if things shake out exactly the way the polls suggest, Harris will secure 281 electoral votes and the presidency. These states are Arizona (Trump +1.3), Michigan (Harris +1.8), Nevada (Harris +0.9), Pennsylvania (Harris +0.4), Wisconsin (Harris +1.6), Georgia (Trump +1), and North Carolina (Trump +0.9).

With many of these races in a dead heat, voter registration numbers have Republicans optimistic about their chances in at least four of those states as Democrats have lost significant ground to the GOP in Pennsylvania, Arizona, North Carolina, and Nevada on the voter rolls. 

Swing state voter registration has spiked in 2024, with more names being added to the registry than last cycle’s margin of victory in these battlegrounds. In many cases, Republicans are out-gaining Democrats. This may signify greater Republican turnout than the polls suggest, and if it does, a commonwealth such as Pennsylvania, with its 19 Electoral College votes, could swing the election.

While the swing state voter registration figures set the stage for Trump’s surge, the catalyst may have been the release of Harris’s 60 Minutes interview that aired on Oct. 7. After evading media interviews for the better part of her campaign, Harris launched a media blitz this week, which included an interview with CBS’s Bill Whitaker. She has faced criticism for not giving direct answers to critical questions and offering “word salad” instead. 

Do betting odds reflect voter sentiment or realistic probability?

Those surveyed in polls are often asked who they believe will win the election, regardless of their own preference. In 2020, most voters thought Trump would beat President Joe Biden from that summer all the way until the beginning of October, yet Biden was crushing Trump by around 9 points nationally the whole time. Meanwhile, bettors were with Biden the whole way.

A poll from July 12 to July 14, 2020, had Biden up 9 points, while Trump led by a point on the question of who voters actually thought would win the election. A poll from right before Election Day, Oct. 31 to Nov. 2, had Biden up by 10 nationally but only up by 1 against Trump on the question of who will win. On July 14, 2020, betters gave Biden a 21% advantage, and by Nov. 2, 2020, that rose to 28.2%.

Betting odds reflect the sentiment of a specific group of people: the bettors. Ideally, without a political motive, election bettors are financially incentivized to be correct. The people surveyed in polls are giving their best guess with nothing to lose for being wrong, and 20% to 30% often say they are unsure, while the bettors are incentivized to be objective and well-researched. As a result, election betting odds have been historically reliable at predicting races.

Since former President John F. Kennedy did it in 1960, only two betting underdogs have won the presidential election: former President Jimmy Carter in 1976 and Trump in 2016. Kennedy was only a slight underdog, at +110 odds, which is a 47.62% probability of victory, and Carter was placed at +100.

Trump won despite having the second-longest odds for a presidential election winner in U.S. history at +375, or a 21.05% chance of winning. Former President Harry Truman’s 1948 +1500 odds place it as the most unlikely electoral victory since at least 1872, beating the odds with a 6.25% win probability.

While the odds haven’t been perfect at predicting elections, a 2022 study found that only 60% of polls conducted in the week before an election contain the correct vote share outcome. In 2020, Biden was supposed to beat Trump by nearly 10 points, according to the polls, yet he won by 4.4%. Many polls also had Sen. Mitt Romney (R-UT) winning in 2012, and most had former Secretary of State Hillary Clinton winning in 2016.

In 2020, the bettors rarely faltered in their confidence. At the beginning of September, Trump pulled within 0.5% of Biden but remained at least 20 points behind for the majority of the election cycle from July on. This time around, bettors haven’t given either candidate even a 10% advantage since July, when Harris took over for Biden at the top of the Democratic presidential ticket.

The Washington Examiner conducted an analysis of polling and betting data starting in July of both cycles and found that fluctuations in betting confidence over the last two election cycles have not significantly correlated with the results of the national polls. However, the tighter margins in who voters think will win the election correlate with tighter national polls and tighter election odds.

In the newest polls, most voters think Harris will win. One poll from Big Village CARAVAN conducted from Oct. 2 to Oct. 4 showed voters thought Harris would win by a 3% margin. Another from I&I/TIPP had it at 6%, and another from Yahoo/YouGov at 4%. The betting odds had Harris up by 0.6% at that time.

The most recent poll from Oct. 6 to Oct. 7 from the Economist/YouGov showed voters split on who they think will win, and betting odds gave Trump a 3.5% advantage on Oct. 7. Meanwhile, all of these polls had Harris up nationally by 2 or 3 points.

While polling data and voter sentiment do share some positive correlation with overall betting confidence, additional factors appear to be at play as there remain volatile fluctuations in betting odds even when the polls do not change. Both 2024 presidential debates significantly shifted the odds, as Biden’s June 27 performance dropped him 16.6% relative to Trump in one day, while Trump’s Sept. 10 debate cost him 6.6% relative to Harris.

Trump lost 5.8 relative points on the day of his July 31 interview with the National Association of Black Journalists in Chicago, while Harris lost 3.4 relative points on the day her Oct. 7 60 Minutes interview aired.

The regulatory battle: ‘Threatens our democracy’

The idea of betting on elections is controversial. Election betting in the United States was frozen until last Wednesday when a federal appeals court allowed the revival of the first fully regulated election-betting market in the country. As the market expands, there will likely be more battles in the courts and legislatures on election betting.

Critics of election betting point out the possibility of corruption. For example, Sen. Jeff Merkley (D-OR) has warned that malicious actors could someday decide, “Hey, I will spend millions of dollars smearing some candidate to make sure the candidate I bet on wins.”

Cantrell Dumas, director of derivatives policy at Better Markets, said the practice “threatens our democracy” because “we could witness scenarios similar to what happened with GameStop” where “speculative traders turned the stock market into a frenzy, political betting could incentivize manipulation, disinformation campaigns, and wild speculation on election outcomes.” 

Commodity Futures Trading Commission Chairman Rostin Behnam, who is currently tasked with regulating election betting, has pursued regulation to ban betting on election outcomes on commercial exchanges as long as the CFTC oversees such transactions. Behnam said the agency does not have the capability to serve as “an election cop” that can monitor markets for fraud and manipulation.

The CFTC has noted two big examples of attempted market manipulation, including a fake poll showing Kid Rock was leading Sen. Debbie Stabenow (D-MI) in a Senate race, which affected the price of reelection contracts despite the fact that the singer didn’t even run for office. It also cited a 2012 case where one person bet millions of dollars on Romney in an effort to make the election appear closer than it was.

“These examples are not mere speculation,” the CTFC wrote. “Manipulation has happened, and is likely to recur.”

The Washington Examiner reached out to the Trump and Harris campaigns for comment.