Throughline
Throughline

Prediction markets are making a 150-year comeback

Prediction market sites allow users to put money on everything from the war in Iran to the winner of the Super Bowl. But where did these markets come from? And what can that history tell us about where they might be going? Today on the show, how betting on popes and presidents long ago planted a see

Topics Discussed

Episode Summary

Executive Summary: The episode traces prediction markets from 16th-century election betting to today’s billion-dollar platforms Polymarket and Kalshi. It explains how markets often forecast outcomes better than polls, but also how they repeatedly collide with ethics, regulation, and fears of insider trading—especially when bets involve war, assassination, and terrorism.

Main Topics: Modern prediction markets explode in scale (Priority: 5/5): The show opens with the rapid rise of Polymarket and Kalshi, fueled by high-profile bets on geopolitics and elections, and framed as a new form of public forecasting and speculation. The Iowa Electronic Markets prove the concept (Priority: 5/5): A trio of University of Iowa economics professors created one of the first modern academic prediction markets, showing markets could outperform traditional polling in elections. Historical roots in election wagering (Priority: 4/5): The episode shows prediction markets are centuries old, with betting on popes, city-state elections, and U.S. elections long predating modern polling. Robin Hanson and DARPA’s failed terrorism market (Priority: 5/5): The program’s second act follows Robin Hanson’s attempt to build more ambitious markets for geopolitical forecasting, culminating in DARPA’s Policy Analysis Market and its collapse after public outrage over ‘betting on death.’ Commercialization and regulatory conflict (Priority: 5/5): Intrade and Betfair turned prediction markets into businesses, but U.S. regulators repeatedly pushed back, culminating in shutdowns, fines, and bans that shaped the industry’s trajectory. Ethics, insider information, and the future (Priority: 4/5): The episode ends by debating whether prediction markets are useful intelligence tools or morally troubling platforms that may reward private information about war, death, and public harm.

Key Arguments: Prediction markets can be remarkably accurate because they aggregate dispersed information and incentives better than polls in some cases. The idea is not new; people have wagered on political outcomes for centuries, and modern prediction markets are a reorganization of an old practice. The collapse of markets like DARPA’s was driven less by technical failure than by public disgust with betting on violence, death, and terrorism. Commercial growth intensified the ethical stakes: bigger markets mean more money, more scrutiny, and greater risk of insider trading or abuse. Regulation has largely determined whether prediction markets survive, with the CFTC acting as the central gatekeeper in the U.S. Today’s market boom suggests the public may be more accepting than before, but the same backlash cycle could recur if the markets keep expanding into sensitive domains.

Data Points: Profit from a political bet: $400,000 to $500,000 - A trader reportedly made six-figure gains betting the U.S. would strike Iran by Feb. 28, 2026. Odds of U.S. strike on Iran: 7% to 26% - The bet that paid out heavily was made when the strike seemed unlikely. Total trading volume on Iran-related events: More than $2 billion - Since the beginning of the year, traders wagered heavily on Iran-war outcomes. Combined trading volume on Polymarket and Kalshi: More than $60 billion - The two biggest prediction markets have seen massive trading activity since the beginning of 2026. Iowa Electronic Markets sample size: A couple hundred people - The original academic market was launched with students and faculty at the University of Iowa. Accuracy vs. polls: 74% of the time - Between 1988 and 2004, the Iowa Electronic Markets beat traditional polls in presidential elections. Polling miss on Michigan caucus: 53% vs. 29% - Jesse Jackson won the Michigan caucus with 53% of the vote while Dukakis got 29%. Market prediction error: Within two-tenths of 1% - The Iowa market predicted the popular vote on the eve of Election Day with very high accuracy. Maximum allowed investment in Iowa Electronic Markets: $500 - The CFTC’s no-action letter allowed only small-stakes trading. Intrade user decline after legislation: Big drop in users - U.S. online gambling crackdowns contributed to Intrade’s downfall. DARPA individual investment cap: $100 - The Policy Analysis Market limited each participant’s investment. Time from media firestorm to shutdown: 24 hours - DARPA’s Policy Analysis Market was killed within a day of the public controversy.

Pivotal Quotes: "The whole world, unfortunately, has become somewhat of a casino." — Donald Trump: Used in the episode to capture the president’s attitude toward prediction markets and speculative culture. "The idea of a federal betting parlor on atrocities and terrorism is ridiculous and it’s grotesque." — Sen. Ron Wyden: Critique of DARPA’s Policy Analysis Market during the 2003 congressional backlash. "We were having markets betting on death." — Robin Hanson: His characterization of the outrage that ended the Policy Analysis Market.

Implications: Prediction markets may become a major forecasting and financial tool, but their future depends on whether regulators and the public accept betting on sensitive real-world harm. The more powerful they become, the more likely ethical backlash and oversight will intensify.

🔓 Sign Up for Unlimited Episode Search

About Throughline

View all episodes from Throughline