In two recent posts, guest-blogger John McGinnis makes a strong case that the US government should legalize prediction markets, that these markets are valuable sources of information, and that claims that prediction markets can easily be “manipulated” are overblown. I agree completely on all counts. Nevertheless, I am somewhat pessimistic that prediction markets will do a lot to overcome the problem of political ignorance.
As I discussed in my previous post commenting on John’s excellent new book Accelerating Democracy, voters are rationally ignorant about politics, and are often unaware of even very basic information about government and public policy. Thus, even if prediction markets make high-quality data readily available for free, most voters may pay little or no attention to it, just as they currently ignore most of the information already available from other sources. In this context, it’s worth noting that there’s little if any evidence that more than a tiny fraction of voters have been paying attention to prediction markets over the last few years that the data has been available. Although I haven’t seen any survey data on the subject, I would not be surprised if a majority of Americans have not even heard of prediction markets, much less been paying attention to the data they generate. Obviously, as John points out, voters who actually place substantial bets on prediction markets do have a strong incentive to pay attention to them. But that is likely to be only a small proportion of the population, just as currently only a small proportion are highly active sports bettors.
Even those voters who do pay some attention to prediction markets may not evaluate their results with anything approaching objectivity. Just as voters tend to overvalue other types of information that confirm their preexisting views, while undervaluing any that cuts the other way, they could do the same with prediction market results. The very nature of prediction markets makes many people’s reactions to them particularly susceptible to cognitive bias. Because they seem like frivolous “gambling” on life and death political events, people may be tempted to ignore inconvenient results as unserious. In addition, political activists can easily use demagoguery against them. To borrow one of John’s examples, if prediction markets indicated that a capital gains tax increase would harm the economy, liberal activists and politicians could dismiss that on the grounds that the result must have been manipulated by Wall Street fatcats eager to protect their ill-gotten gains. If the markets predicted that a military action proposed by a Republican president is likely to fail, conservatives could blame the result on manipulation by liberal financiers like George Soros and other “unpatriotic” and anti-American left-wingers.
Finally, political ignorance makes it more difficult for prediction markets to function at all. In part because most of the public is either unaware of prediction markets or does not understand their value, there was little outcry over the US government’s recent crackdown on Intrade, one of the largest prediction market sites. The Commodity Futures Trading Commission forced Intrade to stop taking bets from Americans.
If more people appreciated the benefits of prediction markets, the CFTC might have been met with a massive political backlash and compelled to back down. So far, political ignorance has merely allowed the CFTC to act against prediction markets without facing much public opposition. In the future, however, opponents of prediction markets might be able to use populist rhetoric to actively mobilize public opinion against them. If, for example, government started relying on prediction markets to inform policy decisions opposed by liberals, one can easily imagine an Occupy Wall Street-like campaign against it. If prediction market-driven policies seemed to tilt leftward, conservatives could resort to populist attacks of their own. And such campaigns might well fall on receptive ears.
This is not to say that prediction markets are useless. Far from it. People can still use the data to make private sector decisions and for “voting with their feet” between states and local governments. In these situations, rational ignorance is far less of a problem. Some of the prediction market data people learn for those purposes might also eventually affect their ballot box decisions, though it’s important to remember that the types of data useful for these two purposes are often very different from each other. Prediction market data can also be used to good effect by government officials and policy experts. But, at least in the short to medium run, the potential policy benefits of prediction markets are likely to be significantly curtailed by political ignorance.
UPDATE: I have made some stylistic and organizational changes to this post in order to increase readability.