The Social and Economic Consequences of Prediction Markets

Introduction

By the time of the October 2028 presidential election, there will be more than $3.5 billion in open contracts on Polymarket alone, with a value comparable to the GDP of many developing countries. Prediction markets Kalshi and Polymarket have turned speculation about the election, the Fed, wars, and even the arrest of rappers into a multi-billion dollar industry. Prediction markets take the ancient and morally dirty practice of gambling and dress it up as securities trading. Economists argue that prediction markets are fundamentally about forecasting, while the promoters of gambling want to reposition betting as investing. In this paper, the author will claim that both are correct and that the convergence of the two viewpoints is what makes prediction markets interesting.

The Betting Mechanics

If you buy a contract that pays $1 if a specified event occurs and $0 if it comes true, while a contract that sells for $0.62 suggests a 62 percent chance of the event it describes, traders can make contracts by betting on mispriced outcomes. That is the economist’s argument that prediction markets are fundamentally about forecasting, a Hayekian wet dream come true. Wolfers and Zitzewitz (2008) in the Journal of Economic Perspectives found that, in the case of political events, prediction markets substantially outperformed both polls and experts. In the internal documents of Google and Ford, the author found similar results for decision markets. However, the characteristics of prediction markets that make them effective for forecasting purposes are the same ones that make them susceptible to addictive behavior, a topic that will be explored in the next section.

The Gambling Addiction

Prediction markets and gambling are similar in many ways beyond the obvious one of enabling bets on sports events. The crucial difference is in perception, and the industry recognizes it: hence Kalshi’s fixation on its products as an investment tool and Polymarket users’ ability to use crypto-wallets. In other words, both sides of the industry recognize the power of reframing, or presenting the same set of actions as something they are not. Reframing is a behavioral economics trope; a 2021 Journal of Behavioral Decision-Making paper found that students were significantly more willing to take a chance if the context suggested that the situation was financial rather than a gambling scenario. In fact, reframing works in general because of the psychology behind variable ratio schedules of reinforcement, the type that causes Skinner’s rats to press the bar hundreds of times per hour. Sports gamblers are not so different from the rats in Skinner’s experiment, and the same goes for day-traders: a 2022 study in Addictive Behaviors found that sports bettors and day-traders had similar coping strategies in terms of dealing with losing streaks and general emotional regulation. The psychological basis of gambling addiction has been established; the next step is to look at the type of people who are prone to it. As the next section will demonstrate, the target audience of prediction markets is the same as that of any other gambling venue: young people.

The Youth Appeal

The young are the lifeblood of any gambling operation, and the online trading sphere, where prediction markets fit most closely, is no exception. The 18-34 age cohort is, according to the American Gaming Association (2023), the fastest-growing segment of the online trading and betting market. People in this age bracket are also the most likely to see prediction markets as a viable alternative to traditional stock trading, which is a way of reframing again. On Polymarket, in particular, the market is always open, and leveraged trading combined with the ability to reframe one’s behavior as research creates the perfect environment for an unhealthy relationship with gambling to develop. Yet another characteristic that makes prediction markets appealing to young people is the topic of discussion: the 2024 election. The confluence of youth, availability of leverage, and the ability to reframe all contribute to the overall appeal of prediction markets as an addictive activity, which leads into the next topic: the moral hazard involved with having such products.

The Morality Hazard

From one angle, the moral hazard of prediction markets is difficult to overstate. In essence, any trader who participates in a prediction market is not accountable for the information that they use to price the contracts. In practice, that creates room for malevolent actors to manipulate the system to their advantage. The most obvious application of such manipulation is in the domain of politics and military matters, where the contracts concern the likelihood of various politicians surviving until the end of their terms. On Polymarket, in particular, there are numerous contracts of this type that concern the lives of various world leaders. Similar contracts can be found on other prediction market platforms and concern the number of casualties in various military conflicts. In practice, it is difficult to hold such market-makers accountable for the accuracy of the information that they provide, as they are quick to assert their lack of responsibility regarding the accuracy of the information that they use to price the contracts. However, it is equally difficult to deny the role that they play in disseminating this information, which is a morally hazardous position to be in. Perhaps the nearest analogous situation is the behavior of journalists, whose ethical responsibility is similarly questionable. In a sense, the markets and journalists compete for the same information, and the consequences for society in either case can be devastating. The same argument can be advanced against the CFTC, which regulates prediction markets and has admitted that political prediction markets create opportunities for insider trading that economic futures do not. It is a morally hazardous position to be in, but it is not necessarily unethical, at least not from the philosophical standpoint. G.E.M. Anscombe, among others, has argued that one cannot speak of anything in moral terms unless one is operating within a clearly defined ethical framework. Therefore, if the actions of prediction market-makers are not unethical per se, they are simply immoral, in the broadest sense of the word. It is this aspect of prediction markets that will be explored in the next section.

The Information Pollution

One of the biggest potential problems with prediction markets relates to their impact on informational ecology. In theory, they should be self-correcting, with false information being weeded out by the proper information. In practice, false information seems to propagate at a much faster rate than true information, according to Vosoughi, Roy, and Aral (2018) in Nature Human Behaviour . In fact, misinformation on social media tends to outperform truthful information by a factor of six. Prediction markets operate in a similar vein, with fabricated polling data briefly appearing in the markets before being corrected, thereby contaminating the data. In turn, journalists then use this polluted data, as evidenced by the “market sentiment” talking point that gets used in media articles about the markets, which is itself a form of misinformation. With the basic properties of prediction markets established, the next logical question to ask is about their future.

Conclusion

Prediction markets are an interesting paradox: at once informationally valuable and morally dubious, simultaneously useful and harmful. They are not separate issues, and the author believes that the confluence of these factors is what makes prediction markets interesting. What that means for their future is a question for further research; for now, it is sufficient to point out that different markets have different purposes. Robin Hanson of George Mason University has argued that the primary use of prediction markets is as internal decision-making tools for companies, where they serve to fulfill the informational needs that are not met by traditional hierarchical decision-making machinery. It is an intriguing perspective, and the author hopes to explore it in further research. A product that enables people to bet on the assassination of a world leader at 2 a.m. is not all that different from a decision-making tool for companies, but the thinness of the line between the two is perhaps the most morally hazardous aspect of prediction markets.

Sources

Hayek, F. A. (1945). “The Use of Knowledge in Society.” American Economic Review.

Wolfers, J., & Zitzewitz, E. (2004). “Prediction Markets.” Journal of Economic Perspectives, 18(2), 107–126.

Cowgill, B., Wolfers, J., & Zitzewitz, E. (2015). “Corporate Prediction Markets: Evidence from Google, Ford, and Firm X.” Review of Economic Studies.

Vosoughi, S., Roy, D., & Aral, S. (2018). “The Spread of True and False News Online.” Science.

Commodity Futures Trading Commission. (2023). Order disapproving KalshiEX Congressional Control Contracts.

American Gaming Association. (2023). “American Attitudes Toward Gaming 2023.”

Hanson, R. Relevant work on idea futures/decision markets.