In Reply to: About Kalshi and other prediction markets. posted by mh on May 11, 2026 at 14:09:27
Frankly the smartest economist I listen to on podcasts would absolutely disagree with you. The nuance is not as cut and dry as you write.
Let me give you a Claude summary of 2025 elections predictions - betting markets vs polls.
For 2025, markets and polls broadly agreed on the winners, but polls again got the margins wrong (underestimating Democrats this time). Kalshi called the outcomes correctly on the big races, but academic scrutiny suggests its accuracy advantage over good poll aggregators is real but modest — and inconsistent across platforms. The honest answer is that neither tool is reliably superior; they work best when used together.
In 2024 prediction markets were more correct on Harris vs Trump. It worked best on that election because that is where all of the money (i.e. data) was.