July 22, 2005  ·  Cass Sunstein

First things first: A grateful thanks for the incredibly helpful comments on my ignorant queries about OSS. (More comments on those queries are more than welcome.) The comments prompt the following thought. For OSS, there’s a lot of dispersed knowledge and also creativity, and that’s a big reason for the success. Something similar is true of Wikipedia (though as some people have suggested, the aggregation process is less reliable there). With ordinary product markets, there’s also a lot of dispersed knowledge, both about product performance and about individual tastes. If a watch doesn’t keep time very well, the market will respond, at least in the aggregate. If most people like the looks of a new car, the market will respond too.

Here we have a clue about why prediction markets sometimes fail, as in the cases of the Rehnquist resignation and the Roberts nomination. There just isn’t a lot of dispersed information out there about what the Chief Justice is likely to do or about the President’s particular choice. It follows that we shouldn’t think (as many do) that prediction markets will be able to foresee terrorist attacks. Investors, taken as a whole, probably lack the information that would make for judgments that improve on the conventional wisdom. If this is right, then prediction markets are unlikely to do well in foreseeing a range of international events.

I have a feeling that this is just a rough cut at the question — and some of the comments to an earlier post might do better — but maybe we’re making some progress.

  • http://www.airs.com/ian/ Ian Lance Taylor

    I think one important reason prediction markets can’t work for terrorism is that terrorists are an active enemy. It is very different from a price market. If anybody with the ability to stop terrorism actually pays attention to a prediction market for terrorism, the terrorists can simply monitor the market and change their plans accordingly. So such a market would have to be secret, which means that only a relatively small number of people could play, which means that arguments based on aggregated information are greatly weakened.

    Even the awareness of the existence of a market can make it invalid. Because terrorists have many more potential attacks than they can ever actually execute, they can choose targets by rolling dice, thus invalidating any specific predictions.

  • http://www.airs.com/ian/ Ian Lance Taylor

    I want to add that in general I think market theory does a poor job of integrating second order effects, in which people use their understanding of the market to game the market. For long-term price competition with many buyers and sellers the effect is minimal. But the more power a single player has in the market, the more ability he or she has to distort the market in his or her favor. If there were real money in prediction markets, Justice Rehnquist could have made a killing. And if predication markets ever actually become popular, that is exactly what would start to happen.

  • http://www.chrisfmasse.com/ Chris. F. Masse

    Hello there,

    A couple of comments:

    #1.
    Speaking of both prediction markets and open source software, I would like to point to you that there are now two OSS packages for prediction markets. Try them out. I listed them here:
    http://www.chrisfmasse.com/3/3/software/

    #2.
    Cass Sunstein wrote that “we shouldn’t think (as many do) that prediction markets will be able to foresee terrorist attacks.”

    For those just surfacing out of an Afghanistan cave and landing here at Lessig’s blog, it’s the 2000-2003 DARPA’s FutureMAP/PAM that Cass Sunstein is referring to.

    However, his point is partially valid for FutureMAP (the umbrella program at the DoD’s research arm, directed by a non-economist) but not for PAM (one of the two specific programs that FutureMAP funded, run by a bunch of top-notch economists).
    http://hanson.gmu.edu/PAM/press/SIN-3-17-03.txt

    “The Policy Analysis Market (PAM), however, was a U.S. military research project designed to test the ability of speculative markets to forecast overall geopolitical trends, not terror attack details (Polk, Hanson, Ledyard, & Ishikida, 2003). This two-year-old milliondollar research project began long before Poindexter joined its funder, the Defense Advanced Research Projects Agency (DARPA), and was five months from public trading of under one hundred dollar bets. PAM traders could have, for example, bet on the chance of high levels of civil unrest in Saudi Arabia in the fourth quarter of 2004, conditional on the US moving its troops out of there two quarters earlier. By comparing estimates based on different assumptions, PAM could have advised us on the effect of various US Mideast policies.”
    – Robin Hanson – The Informed Press Favored the Policy Analysis Market – 2005
    http://hanson.gmu.edu/PAMpress.pdf

    More on PAM (Policy Analysis Market), direct from the horse’s mouth:
    http://hanson.gmu.edu/policyanalysismarket.html

    See, PAM was never about “foresee[ing] terrorist attacks” or “foreseeing a range of international events”.

    Best regards,
    Chris. F. Masse
    http://www.chrisfmasse.com/3/3/predictionmarkets/

  • Marcus

    Regarding Ian’s comment on prediction markets and the chance that one individual could make a killing. If PMs were to become popular, wouldn’t the sheer number of traders minimize the ability of a single trader to influence the market? I think that is the case, one trader will not be able to influence the market because of the competition inherent in it. Btw, the Judge voting example is perhaps not a case where PMs can really add value since the actual outcome is ultimately controlled by a select group of individuals (and other traders would have no idea what those individuals are thinking). check out this link for more on that: http://cartegic.typepad.com/mapping_strategy/2005/07/predicting_clos.html