Monday, January 22, 2007
Posted by Jeff Lipshaw
Cass Sunstein (Chicago, right) hardly needs me to plug or critique one of his articles (another 100 or so articles posted and another 36,000 or so SSRN downloads and I will have him beat), but he has recently posted Deliberating Groups versus Prediction Markets (or Hayek's Challenge to Habermas), which is to be published in Episteme. Here is the abstract, followed by some comments:
For multiple reasons, deliberating groups often converge on falsehood rather than truth. Individual errors may be amplified rather than cured. Group members may fall victim to a bad cascade, either informational or reputational. Deliberators may emphasize shared information at the expense of uniquely held information. Finally, group polarization may lead even rational people to unjustified extremism. By contrast, prediction markets often produce accurate results, because they create strong incentives for revelation of privately held knowledge and succeed in aggregating widely dispersed information. The success of prediction markets offers a set of lessons for increasing the likelihood that groups can obtain the information that their members have.
Much of the description of the problem, like much of behavioral economics, made sense to me. The prescription - to replicate the price mechanism for what would normally be the result of deliberation - is more problematic. Professor Sunstein cites the predictive accuracy of the Iowa Electronic Markets, which allow people to place bets on presidential elections, or the Hollywood Stock Exchange, in which traders use virtual, not real, money to bet on Oscar winners. Orange juice futures predict the weather better than the National Weather Service.
Granted, Professor Sunstein acknowledges the limits of prediction markets: you cannot use a prediction market to determine if a jury reached the wrong conclusion, and "[m]ore generally, it is not easy to see how prediction markets could be used on normative questions." But he claims, nevertheless, that even if prediction markets are not feasible, understanding them can help us understand what goes wrong in, and to improve, deliberation:
It should be possible for deliberating groups to learn from the successes of markets, above all by encouraging their members to disclose their privately held information. When such groups do poorly, it is often because they fail to elicit the information that their members have. Good norms, and good incentives, can go a long way toward reducing this problem.
Thus, as to small groups like corporate boards, the way a prediction market is instructive is by showing group members the consequential value of "dissent and epistemic diversity."
There is a logical leap here that loses me. I have little doubt deliberating groups suffer from a collective action problem for the reasons stated: cognitive errors, withheld information, and cascades. But prediction markets don't "overcome" the collective action problem, they avoid it. Markets create a collective result from individual actions in which there is no gap between actual preference and stated preference - when we place our anonymous bets, we do so in the equivalent privacy of the voting booth. When we step back and look at the result, we are as awed as Hayek that somehow this price mechanism, without deliberation, has created a result that aggregates all of the factors.
When we ask a corporate board member (or a lawyer advising the board) to "disrupt the conventional wisdom" or to overcome "fear of social sanctions," we are not asking her to act atomistically, but autonomously. Prediction markets can tell Costco how many shoppers will buy the twenty pound tub of cashews, but they cannot help me in the decision whether or not to overcome my gluttonous instincts. When all is said and done, the corporate board member will have to invoke some sense of duty to reveal his private information and thereby rock the boat. Learning that from a prediction market is something like pumping up one's courage from the knowledge that electrons will spurn their orbitals and make quantum jumps. If it helps, great, but it's a thin analogy.