Tuesday’s elections will reveal who controls Congress. They will also test faith in polls and prediction markets, both of which agree on a Democratic surge. The RealClearPolling average points to an 8.2-point lead for Democrats in the generic Congressional vote, while Nate Silver’s averages show a similar lead. Traders on Kalshi and Polymarket predict strong Democratic chances for a Senate and House majority.
The consensus among polls and prediction markets poses a risk. Both can rise or fall together because prediction markets largely interpret polling data. This election serves as a stress test for their reliability. In past elections, polls overestimated Democratic support, but have been accurate in some midterms without Trump on the ballot. Robert Erikson notes this only occurs when Trump is not running.
Pollsters aim to identify previous survey misses that shape perceptions of their credibility. Weighting samples by party or past vote has become common. Political scientist Alan Abramowitz warns that last year’s polling errors may not predict future inaccuracies. Past errors were correlated, affecting outcomes in multiple states similarly.
A fabricated poll highlights the issue of accuracy in raw polling data. Prediction markets emerged as an alternative, initially predicting Trump’s success during the 2024 election despite polls showing uncertainty. However, recent events have called their reliability into question. Trading errors in prediction markets, such as ratings for Brazilian elections, have raised doubts.
Some experts see prediction markets not as competitors but as reflections of polls. They add clues like early voting but historically offer no enhanced accuracy compared to polls. Misjudgments like the Brazilian election cast doubts on the markets’ capability.
Kalshi’s and Polymarket’s differences in pricing predictions are curious. Federal regulators plan to address insider trading concerns, with Kalshi suspending several candidates from betting. Polymarket has also referred cases to regulatory bodies.
This election will test both polls and prediction markets. If the House vote aligns within three points of a Democratic lead, polls succeed. If Republicans exceed expectations or Democrats significantly outperform, polls falter. Prediction markets face scrutiny if outcomes don’t match heavily backed favorites. Consistent misses could deter trust in these systems.
Reporter Robert Erikson advises bettors to stick with favorites. The predicted Democratic wave presents a possible outcome. Its arrival would confirm accuracy; failure would raise questions for both systems. Voter and media trust in these predictions hangs in the balance.
