A new national poll found widespread unease about election prediction markets as their political reach expands before the 2026 midterms.
The non-profit organization Partnership for Large Election Jurisdictions (PLEJ) commissioned the survey and SurveyMonkey conducted the poll. It covered 1,037 US adults from July 16 through July 19, and of those questioned, 867 “identified themselves as likely 2026 midterm voters.” While the poll data can provide some useful insights, it’s important to note that PLEJ is a membership organization of election administrators from large jurisdictions nationwide rather than a neutral research party.
Findings among the likely midterm voters:
- 75% said prediction markets create confusion
- Another 60% said the platforms fuel inaccurate election information
- 45% believed some participants may possess inside information about election outcomes
The results also imply confusion over what market prices measure. Nearly four in 10 likely voters thought the odds reflected counted ballots or official government projections.
They do not. Contract prices move with buying and selling by participants. The figures represent collective expectations, not votes, official forecasts or certified results.
That difference matters more now because prediction markets are beginning to work their way into everyday political coverage. Their prices can move quickly as real-time expectations shift around elections, legislation, interest-rate decisions and conflicts.
Reporters watch prediction markets, as do investors, campaign professionals and others looking for a read on changing sentiment. And while this poll supports a need for more education around the distinction between prediction market data and election results, it’s not a definitive indication of how deep the confusion on that particular line runs.
What information do voters actually trust?
PLEJ said contracts linked to the 2026 elections had already drawn nearly $200 million in wagers, a number that is likely understated. For context, Kalshi’s LA Mayor winner market alone has over $87 million in notional volume, and the California Governor winner market is nearing $50 million.
According to the release, “likely midterm voters continue to place substantially greater trust in the officials responsible for administering elections.” That claim is based on the finding that 64% expressed high trust in state election officials, while 61% said the same about local officials. Fewer than one-quarter reported high trust in news media, social media or prediction markets for election information.
PLEJ Executive Director Carolina Lopez said: “When speculation, projections, and prediction markets compete for attention, voters need to hear directly from the state and local election officials responsible for administering elections.
“Clear, timely information from official sources helps the public understand the process and distinguish market expectations from actual election results.”
Methodology note
The poll’s margin of error was plus or minus three percentage points, according to the release. Note: PLEJ has not published the full question wording, order, or subgroup breakdowns, so it’s not possible to independently check for language framing or how sensitive the topline numbers are to how questions were asked.
Additionally, SurveyMonkey’s polling of this kind typically draws from an opt-in online panel rather than a random probability sample — standard industry practice, but one where a stated margin of error carries somewhat less statistical precision than in a true random sample, and where results depend more heavily on how the panel was weighted to resemble the broader public.
Prediction markets continue status as barometer amid public concerns
The survey also found that 38% would lose confidence if official results differed from prediction market odds. That finding points to the potential risk of confusing a tradable forecast with the legal election process.
Insider trading has become a parallel concern. The Senate unanimously changed its rules in April to bar senators, officers and staff from prediction-market trading. Separate bipartisan bills would restrict federal officials from using material nonpublic information or ban certain officials from trading entirely.
States have moved, as well. At least 15 have implemented or considered prediction market legislation this year. California proposed restrictions for officials holding nonpublic information. A handful have barred candidates from betting on races in which they are running, though that is also already against market rules posted by exchanges.
The concern itself is not new. US securities law has policed trading on material nonpublic information for decades. Congress reinforced that principle through the 2012 STOCK Act, which applied insider-trading duties to lawmakers and congressional employees.
Prediction markets present the same old problem in a newer setting. Someone with privileged knowledge can trade before the public learns what happened. The difference is the subject matter, which may now include elections, legislation or government action.