Political prediction markets are a public good

Review Editor Rating: Star Tribune Opinion publishes a mix of national and local news comments online and in print every day. To contribute, click here.


Recently, the Commodity Futures Trading Commission, or CFTC, withdrew its no-action letter from Victoria University of Wellington, which since 2014 has been licensed to operate the PredictIt political prediction market.

The original no-action letter allowed Victoria to exploit limited scope markets for political events for research purposes. Since its inception in 2014, PredictIt data has been made freely available to researchers in a wide range of disciplines, including economics, statistics and political science. Research from these markets has provided insight into a wide range of topics, including the impact of Brexit on trade between countries, the extent of Russian trolling on Twitter, and what makes Brexit financially successful. a trader.

From the perspective of Victoria’s initial CFTC tenure, PredictIt has more than lived up to its end of the bargain. But beyond their use in public policy and social science research, political prediction markets serve an invaluable public good.

On the contrary, the CFTC should lift its restrictions on PredictIt and other markets, rather than closing them. At a time when the media is exposing rampant “misinformation” and “misinformation,” especially around elections, prediction markets play a crucial role in the public interest. Unlike experts and polls, with their quirky methodologies and unspoken partisan biases, prediction markets are far more transparent and accurate.

At the time of writing, the political event “Who will be Speaker of the House of Representatives at the next Congress?” had contracts on Kevin McCarthy at $0.64 each, Nancy Pelosi at $0.15, Hakeem Jeffries at $0.13 and a host of others at $0.05 or less. Once the next House Speaker has been chosen, each of these contracts will be worth either $1 (if the person in question is chosen as the speaker) or nothing. At a high level, this indicates that traders believe a Republican house after midterms is more likely than a Democratic house. But more than that, the precise values ​​assigned to each candidate simply indicate that the most recent trade on McCarthy involved a trader selling a contract to another trader for more than $0.60. No mysterious algorithms or political posturing.

Although it seems comparable to what we can learn from polls, political prediction markets are more accurate and always up to date. Research on the 2020 elections, by Harry Crane and Darrion Vinson, comparing the reliability of PredictIt’s market prices to that of polling analysis conducted by FiveThirtyEight, finds that PredictIt’s market prices are a more accurate predictor of election results than competing polls and data analysis widely quoted in the media. The results for 2020 confirmed the same conclusions from an analysis of the 2018 mid-term reviews.

In addition to predictions, prediction markets make civic engagement educational and fun. They provide a platform in which people of different ideologies can interact, an increasingly rare place in an era of growing polarization. Since users focus on the objective outcome of who will win an election, they focus more on facts and data than on emotion and vitriol. The extensive commentary section associated with each PredictIt market is testament to this.

There’s no doubt that hardcore PredictIt traders know more about the ins and outs of elections than many political journalists. Market participants are urged to dig into polling methodologies, read between the lines of political ads, research outdated election laws, and be aware of their personal biases. Even new traders benefit: there is evidence that trading on PredictIt markets can increase students’ interest in the political process.

Critics point out that if prediction markets were to attract more attention, then the possibility of market manipulation would open up opportunities to manipulate public perception and negatively affect election results. But these criticisms rest on theoretical concerns that have not materialized in practice, and further overlook the fact that opportunities for manipulation are already endemic for the media and polling stations.

When prices move for no apparent reason, traders step in to pull them back. While some have argued that profit would lead to market manipulation, in practice it is negated by the same motive.

Prediction markets are an important alternative to our degraded public discourse. Anyone who has ever criticized “liberal media,” “right-wing pollsters,” or the “toxicity” of our political climate should hope the commission reverses its decision as soon as possible.

We need more platforms like PredictIt to make markets work and promote discussion of public interest events, not less.

Harry Crane is professor of statistics at Rutgers University and co-founder of Researchers.One. Koleman Strumpf is the Presidential Burchfield Professor of Political Economy at Wake Forest University and associate editor of the Journal of Prediction Markets. This article first appeared in the Chicago Tribune.

Leave a Comment