Share 
Prediction markets did not begin in a sleek startup office or a Wall Street boardroom. They started in 1988 at a bar in Iowa City, where three University of Iowa economists were swapping frustrations over lunch. Polls kept getting elections wrong, and that year’s shocker made it hard to ignore.
Civil rights leader Jesse Jackson had just defeated Michael Dukakis in the Michigan presidential primary, even though most polls indicated Dukakis would win comfortably. Robert Forsythe, George Neumann, and Forrest Nelson began discussing a simple question: What if people could put money behind their predictions instead of just answering a survey?
Somewhere between the first and third beer, the idea clicked. If markets are efficient at setting prices, they might also be effective at setting probabilities.
A closet server and the 1988 Bush-Dukakis bet
Later that year, the trio launched what they called the Iowa Political Stock Market. It was not glamorous. The server sat in a tiny faculty office that barely fit two chairs; many students had to trek to computer labs to place trades because home computers were still uncommon.
The first real test came during the 1988 presidential election between Dukakis and George H.W. Bush. Fewer than 200 traders bought and sold contracts tied to each candidate’s expected share of the popular vote. On paper, it looked like a small academic exercise.
The results, however, were significant. The market predicted Bush would receive 53.2% of the popular vote and Dukakis 45.2%. The actual results were almost identical. For a tiny experiment run out of a cramped office, the accuracy was stunning.
Nobody got rich in that first round—the biggest profit was just $13.54 on a $250 investment, though the bragging rights were worth far more.
The CFTC’s strict academic sandbox
As word spread, the project evolved into the Iowa Electronic Markets. It operated with the Commodity Futures Trading Commission‘s blessing, but only under strict conditions. The markets had to remain academic and experimental, with no outside advertising and tight limits on individual investments.
That arrangement allowed the Iowa team to continue testing their theory while private companies struggled to find a legal foothold. Platforms such as Intrade and TradeSports later tried to commercialize similar ideas but ran into regulatory roadblocks. For years, universities were among the few places where prediction markets could exist openly in the United States.
According to the original report by NBC News, researchers were intrigued by the findings. When people risk their own money, they tend to seek out better information. That simple incentive structure appeared to sharpen forecasts in ways traditional polls sometimes could not.
The leap to billion-dollar digital platforms
Decades later, the concept has exploded into a full-scale industry. Companies like Kalshi and Polymarket now offer contracts on everything from elections to Super Bowl side bets. Trading volume has soared, reaching levels the Iowa founders could hardly have imagined.
A federal court ruling in 2024 cleared a significant hurdle for the sector, opening the door to faster growth. A more welcoming regulatory environment in 2025 further accelerated expansion, even as debates over oversight and consumer protection continue.
