Midterm Prediction Markets Hit $431 Million in Volume Yet Most Congressional Contracts Shift on Modest Sums

With polls opening Tuesday, $431 million has been wagered on congressional prediction markets. Yet the Anti-Corruption Data Collective reports that 96% of those markets with live order books can still be moved five cents for $3,500 or less. The analysis covers 28,845 markets across Kalshi, Polymarket and Polymarket US, where volumes have nearly tripled since the last congressional cycle and the number of contracts has grown 61-fold.

Michelle Kendler-Kretsch, research fellow at the Anti-Corruption Data Collective, told Gambling Insider the low cost to influence prices persists even as Election Day nears. "We were actually surprised when we did our initial research in September how cheap it was to move many of these markets." Most markets lack sufficient volume for quick counter-trading.

Persistent Thin Liquidity Across Platforms

66% of live congressional markets can be moved five cents for under $100. Only 244 require more than $10,000. Even the hardest market in each of the 470 races can typically be shifted for less than $25,000. 90% of all contracts show trading volume below $10,000, the exact threshold Kalshi applies for low-volume warnings introduced in October.

Trading concentration is stark on Polymarket, where the top 1% of wallets now account for 75% of congressional volume, up from 68% in August. A typical market attracts just 48 bettors. The report also flags 591 suspicious round-trip trades across 22 Polymarket markets that moved $618,395 while generating only $34 in net profit.

Counterpoints on Influence and the Limits of Current Tools

Kalshi's new warnings highlight thin markets but do not disclose individual account activity or order flow. Kendler-Kretsch contrasted this with Polymarket's public wallet data, which lets researchers track patterns across trades. She questioned whether platforms should list so many low-volume contracts if warnings are required for most.

Paul Krishnamurty, chief political oddsmaker at BetOnline.ag, countered that five-cent moves are routine, rarely newsworthy, and lack evidence of shifting voters. "I have never heard any evidence that betting odds or narrative shifts voters." Matt Cepulis of 3betmedia observed that cheap price jumps can still generate screenshots that circulate far beyond those reviewing order books.

A Brennan Center survey cited in the report found respondents 10 to 20 percentage points more likely to expect victory for a market-favored candidate. Post-election review of pricing versus actual results may reveal more about repeated anomalies, yet order-book data alone cannot prove intent. Greater access to platform records would allow independent testing of reliability claims before the next cycle.