As the past month aptly demonstrates, political pollingoften swings wildly. Researchers, such as the late political scientist Seymour Martin Lipset, have advised against putting much stock in their reliability, warning everyone to be very skeptical of polling on presidential elections in particular. Unfortunately, they have offered no alternative. Yet one exists, if laws do not prohibit it: Betting markets have the potential to outperform polling, and achieve far more—yet a government agency wants to shut them down.

　　Lipset observed major discrepancies in the polls concerning domestic issues, presidential candidates, and foreign policy too.Examples abound: On December 15, 1975, Harris poll showed that Hubert Humphrey led Gerald Ford 52 to 41 percent and Ronald Reagan by 50 to 43 percent. Three days later, Gallup poll showed Ford leading Humphrey 51 to 39 percent, and Reagan leading 50 to 42 percent. Before the elections, between November 1974 and March 1975, four different polls across the nation asked people whether they considered themselves “liberals” or “conservatives.” Harris and Gallup found many more conservatives than liberals. The two other polls, both associated with universities, reported 50/50 division—though the academia-based polls worded the questions somewhat differently from the other two. Gallup and Harris simply asked people how they identified themselves, whereas the academia-linked polls asked the following: “We hear a lot of talk these days about liberals and conservatives. I am going to show you a seven-point scale etc. etc.” to place yourself.

　　In November 1975, pollsters asked if people favored giving New York City federal funds to rescue it. The answers varied, and were significantly different when pollsters got results from face-to-face interviews as opposed to over the phone. In January 1975, Gallup and Harris asked people to choose which method they preferred for reducing gasoline consumption: rationing or raising prices. The wording of the questions was different—and so were the replies: In one poll, 37 percent wanted rationing, in the other 60 percent. At best, the media mentions the different polling methods in tiny letters, in a footnote.

　　Such differences, Lipset found, were typical of every topic pollsters raised, both in surveys he examined and in those he did himself with colleagues. Even when using similar methods of selecting responders, done at the same time but differing in the wording, the poll results differed drastically. Clearly, subtle differences in the way a question is presented can impact responses. It matters if relevant information is supplied. It matters if respondents are asked what they think is right, or how they feel about certain decisions or events. Obviously, the framing of a question can have a dramatic impact on the answers respondents give.

　　Particularly in an election year, it is distressing to realize how fickle and misleading polls can be. But is there an alternative?

　　There is: letting people put their money where their thoughts are and bet on a range of questions, ideas, and events. In particular, they could bet on the outcomes of elections. This could become a source of invaluable information.

　　To see why such markets, when they are deep (meaning they have significant volume), offer better, more objective information than polls, journalists, or political scientists, consider first how betting markets actually work. They have been working well for a long time, in sports in particular, being more accurate in predicting odds than sports analysts or fans. In fact, a closer look reveals that such betting markets are already implicit in existing stock and bond markets too.

　　First to clarify: betting markets do not really predict the outcome of the game or event. They do not tell us what will happen. The numbers generated in these markets reveal how much money people are willing to bet on particular outcomes, whether they like that outcome or not. Statisticians must then extract the implied probabilities from such prices and knowledge of the parameters that allow for such exchanges to exist.

　　Setting up such markets is not much different from creating one for an IPO (Initial Public Offering). In the latter case, investment bankers first do thorough due diligence to price it. For an event contract, bookmakers and traders with broad experience would use a variety of data to frame the market and then set the initial pricing for the event. Once the contract on the event trades, the traders and the bookmakers adjust the price as the money flows in—or flows out.

　　Politicians dislike deeper financial and betting markets because they make it harder to cover mistaken decisions with obscure, pretentious jargon.

　　Bookmakers do not care what the traders think about the event, whether they buy or sell. In this respect, the organization of betting markets is no different from an insurance company. The insurance company assesses the likelihood of a particular event occurring, and the amount they will pay if it does, and sets the insurance fee accordingly.

　　To illustrate the point, compare traders in betting markets to someone buying fire insurance. The person makes a bet on an event at the existing price. The other side of the bet, the bookmaker in the first case and the insurance company in the second, are pricing events (the fee in the insurance company case) to stay in business. The insurance company treats the fires as uncorrelated events, and must find a critical mass of people wanting the insurance. (There are no insurance models for dealing with the consequences of, say, active volcanoes, such as Vesuvius. In this case “all bets are off” as an expression says, and no private insurance can be offered.) Once there are many traders, and provided political agitators do not finance or otherwise manipulate—betting markets on political events would not differ from the wide variety of insurance markets that exist now, and have existed for centuries.

　　After all, stock and bond markets’ reactions to a wide variety of political events have been public for centuries. In other words, markets for betting on events do exist: they are just less visible, and it is harder to extract signals from them. It is difficult to isolate the implied price of a specific event, since many things are happening at the same time, changing bond and stock prices. Financial life is not a laboratory experiment. Bets on events would offer a specialized financial instrument, complementing and deepening existing financial markets by pricing separately some events. The existence of these markets in the trillions also eliminates the possibility of cornering betting on political events markets, since gross deviations would be instantly visible, as nobody is wealthy enough to have an impact on these deep markets.

　　Potentially then, electoral betting markets could supply information far more accurate than what we are getting now from polls. Why then is there so much opposition to allowing such markets to emerge?

　　For centuries, millennia, rulers and governments have opposed all “betting on ideas” markets, relating to stock, futures, insurance, sports—you name it—on false premises. It is not only a case of responding to the pressures of groups who suffered or expected to suffer from new competition. The reality is that these markets had the impact of destroying the foundations of much orthodoxy and of decreasing concentration of powers, financial powers in particular. They led to “producing prices” which could substitute for centralized bureaucratic and political meddling.As in so many areas of life, politicians and bureaucracies do not like that.

　　Just this year, the Commodity Futures Trading Commission (CFTC) has been rejecting applications of traders, political analysts, and investors to bet on a wide range of events, elections among them, saying that such contracts on events are undistinguishable from “gambling.” Now, as throughout history, words bias thoughts—with facts out of sight and minds. The CFTC is hardly an exception.

　　CFTC’s Democratic Commissioners, Kristin Johnson and Christy Goldsmith Romero, among others, voted for prohibiting such markets. Republican Commissioners, Summer Mersinger and Caroline Pham, voted against the proposal, though Mersinger said that she is not a “fan” of event contracts; she merely thought that the CFTC did not have the authority to prohibit them. Meanwhile, the agency is moving to shut down the PredictIt platform, and rejecting an election-betting bid by derivatives exchange startup Kalshi. Both PredictIt and Kalshi launched lawsuits to challenge the agency’s decisions, and Polymarkets is an offshore entity that can trade in cryptocurrencies only. However, the legal status of American citizens making bets on this offshore site is unsettled.

　　The CFTC is making the same mistakes that courts, regulators, and academics made for centuries when arguing against stock and futures markets, drawing on ancient prejudices and zero evidence. The evidence always and everywhere contradicted the arguments brought up by those opposing their existence. Ideas have very long lives. However, it is true that shallow events-contracts markets, unable to expand because of heavy regulatory clouds, may not produce reliable prices. They could be manipulated.

　　It is also true, since nothing and nobody is perfect, that gambling and betting markets had very marginal negative side-effects—greatly exaggerated by those opposing them—with the media predictably trumpeting them on front pages too. Since the side effects were exceptions they sold news, whereas normal, routine events do not. True too that trial and error has always been required to establish the proper legislative working of these new markets, to hold all parties accountable, whether insurance, futures, sport, or political betting. Yet once governments and the courts dismissed legal and regulatory obstacles, such markets settled into responsible routines, with rare exceptions. Expect much opposition to this change. Bureaucracies have always been against deepening financial markets, realizing full well that these markets diminish their roles. Politicians dislike deeper financial and betting markets because they make it harder to cover mistaken decisions with obscure, pretentious jargon.

　　But law, morals, liberty, and deepening financial markets are and have always been closely connected, a connection often obscured by the heavy veils of language. With so much justified concern about fake news and unreliable polling, now would be a good time to make those connections more transparent.