Key Points
- Delaware County has required approximately 2,500 election workers to affirm that they hold no direct or indirect interest in election bets, wagers or prediction markets.
- An election-administration survey found that nearly four in 10 likely midterm voters mistakenly believed market odds represented counted votes or official government projections.
- A new analysis found that $133 million had already been traded across 7,466 markets covering 2026 congressional races as of August 10.
- Kalshi and Polymarket say they have strengthened surveillance and restrictions intended to keep election insiders out of political markets.
Election Offices Are Rewriting Their Rules Before November
Election officials are preparing for prediction markets to become a new source of confusion and integrity concerns during the 2026 midterms.
Kalshi, Polymarket and Polymarket US now carry markets on races across the country, including individual House and Senate contests, party control of Congress, turnout, victory margins and candidate endorsements.
An August report from the Anti-Corruption Data Collective identified 7,466 markets related to the 2026 congressional elections across the three platforms. The researchers calculated that $133 million had been traded as of August 10, already exceeding the $92 million recorded across the entire 2024 congressional cycle.
That scale is forcing election administrators to consider questions that were barely relevant during previous midterms: whether employees can hold positions on races they help administer, what happens when market odds move before results are released and whether voters understand what those percentages actually represent.
Delaware County Adds Prediction Markets to Worker Oaths
In Delaware County, Pennsylvania, elections director Jim Allen raised prediction markets during poll-worker training after previously having little reason to discuss them.
One worker asked whether a small wager on turnout could be allowed to make the day more interesting. Allen rejected the suggestion.
The Delaware County Board of Elections subsequently amended its worker oath to include an affirmation that signatories have no direct or indirect interest in bets, wagers or prediction markets. Approximately 2,500 people have signed it, covering permanent election employees, poll workers and temporary staff processing ballots.
Pennsylvania law already prohibits election officers from betting directly or indirectly on an election result. The county’s change explicitly identifies prediction markets, removing any argument that event contracts sit outside the existing restriction.
“The last thing we need is the referee in elections being accused of having a financial stake,” Allen said in an April interview with Votebeat and Spotlight PA.
Other election offices have begun taking similar precautions. The Arizona Secretary of State’s Office adopted an internal policy in July prohibiting employees from using or disclosing nonpublic election information for financial gain through event wagering or prediction markets. Violations can result in termination and referral to law enforcement, according to the official policy.
Many Voters Do Not Understand What Market Odds Represent
Election officials are also concerned that market prices may be mistaken for actual voting data.
A national survey commissioned by the Partnership for Large Election Jurisdictions found that nearly four in 10 likely midterm voters believed prediction market odds represented either votes already counted or official projections from state and local officials.
The survey covered 1,037 adults between July 16 and July 19, including 867 likely midterm voters, and reported a three-percentage-point margin of error. It found that 75% believed prediction markets create confusion, while 38% said a result that differed from market odds would reduce their confidence in the official outcome.
“Prediction markets are becoming part of the election information environment, but they are not election results,” PLEJ Executive Director Carolina Lopez said in a statement announcing the findings.
That confusion becomes particularly dangerous when markets move sharply during ballot counting. A price change may reflect a new trade, news report or shift in sentiment. It does not indicate that election officials have counted additional ballots unless public results supporting that movement have actually been released.
Thin Markets Can Move on Very Little Money
The most heavily traded national markets may be difficult for any single participant to move for long. Hundreds of smaller congressional markets have considerably less protection from an aggressive trader.
The Anti-Corruption Data Collective found that 87% of the Polymarket election markets it examined had less than $10,000 in trading volume. In 69% of those markets with active order books, a $100 trade would have moved the displayed price by at least five cents.
Its analysis also found that 80% of Polymarket’s congressional markets had no more than 100 participating wallets. The largest 1% of wallets accounted for 68% of trading volume, while 10 wallets alone generated 17%.
The displayed probability is therefore most vulnerable in obscure primaries and lightly followed House races. A trader does not need to alter the election itself to produce an eye-catching price movement that is repeated by political accounts, campaigns or media outlets.
Kalshi has argued that financial incentives help markets correct distorted pricing because other traders can take the opposing side.
That mechanism works best when a market has enough participants and capital waiting to challenge a bad price. A market with a shallow order book may display a distorted probability long enough for it to circulate well beyond the platform.
Why This Matters for Bettors

For political bettors, the immediate concern is market quality. Thin liquidity produces weaker prices, greater slippage and odds that may reflect one large position rather than a broad assessment of the race. A contract displaying a 70% probability can look authoritative while representing the judgment of remarkably few traders.
Eligibility rules are also becoming harder to ignore. Election workers, campaign personnel, government employees, candidates, contractors and even relatives of political insiders may be prohibited by platform rules, employment policies or state law from trading particular contracts. Bettors with access to nonpublic information face possible account suspensions, financial penalties and regulatory investigation.
The Commodity Futures Trading Commission has already declared that its authority over designated contract markets covers insider trading, fraud, manipulation, wash trading and other disruptive practices. In February, the regulator publicized two Kalshi cases involving a political candidate trading on his own candidacy and a media employee trading with likely advance knowledge of unpublished content.
Kalshi has stated that donors, campaign staffers and certain family members are among those prohibited from trading in its election markets. The company also said it uses surveillance software to block some prohibited traders before they can place a position. Polymarket said it had expanded the team monitoring election activity ahead of the midterms.
Those controls reduce straightforward abuse, but election administration creates a much larger circle of potential insiders than a conventional corporate market. County workers, voting-system contractors, campaign staff, media decision desks and officials processing mail ballots may all receive meaningful information at different points before the public.
The broader industry risk is reputational. Political markets have traditionally been defended as forecasting tools that aggregate dispersed information. If voters begin viewing them as vehicles for insiders or wealthy traders to manufacture election narratives, operators could face narrower contract approvals, tougher participant restrictions and pressure for outright state or federal bans.
Longtime bettors have seen bad lines, thin markets and public overreactions before. Elections add a less forgiving audience. A temporarily distorted football line does not encourage millions of people to question whether the final score was legitimate.
What Happens Next
More state and local election offices are likely to adopt explicit employee policies before November, particularly in jurisdictions where workers receive partial results before public reporting begins.
Wisconsin has taken a more aggressive approach. Its Elections Commission warned in July that state law disqualifies a person from voting in an election if that person has a direct or indirect wager on its outcome. Intentionally voting while disqualified can be prosecuted as a felony.
Platforms will face pressure to show that their prohibited-trader lists and surveillance systems work across thousands of individual races, not only the national markets attracting the most money. That may mean more preemptive account blocks, position reviews and referrals to the CFTC as Election Day approaches.
Election administrators will meanwhile need to communicate results quickly enough to prevent unofficial market movements from filling an information vacuum. The largest test may come after polls close, when partial returns, delayed mail-ballot counts and thinly traded contracts all begin moving at once.

Cole cut his teeth as a sportswriter in Texas, covering everything from Longhorns games to small-town Friday night lights. A lifelong bettor stuck with offshore books for over a decade thanks to Texas' slow path to legalization, he eventually found his way into the world of social sportsbooks - where he uncovered a fast-growing, community of bettors.
Today, he writes for the millions of Americans in states without legal books, helping them explore safe ways to bet without running afoul of the law.
As editor-in-chief, he aims to keep BettingScanner honest, human, and grounded in what bettors actually care about: fairness, fun, and finding your lane - even when the state won’t give you one.







