Prediction Markets Don't Actually Harness Crowd Wisdom for Reliable Forecasting
Source: Rafferty Thompson. "Gambling on the Wisdom of Crowds Is a Bad Bet." February 20, 2026. jacobin.com
The Gist
The author argues that prediction markets don't actually tap into crowd wisdom to make good predictions. Instead, they're just gambling platforms where the odds are shaped by how the market is set up, not by people's real ability to predict the future.
Conclusion
Political prediction markets (PPMs) should not be liberalized based on the 'wisdom of crowds' theory because they don't actually generate reliable forecasts through crowd wisdom, but rather through speculation shaped by market design and infrastructure
Premises
- The wisdom-of-crowds theory requires expert intervention to collect, tabulate, and validate crowd judgments against independent standards
- Prediction markets depend on market infrastructure and design rules that can be poorly constructed, undermining information reliability
- Many modern prediction markets involve non-repeatable events with no alternative benchmarks to validate their probability estimates
- Market operators like Polymarket make arbitrary decisions about outcome definitions that affect payouts and undermine reliability
- The evidence for PPM superiority over traditional forecasting methods like polls is mixed at best
- Current prediction markets follow a 'financialize everything' approach that extends betting to topics where crowd wisdom cannot be meaningfully validated
Assumptions
- Reliable forecasting requires independent validation standards
- Market design and infrastructure significantly influence outcomes
- The original wisdom-of-crowds theory has been misapplied to modern prediction markets
- Speculation and gambling are fundamentally different from knowledge generation