Prediction Markets

Topic Owners: Matthew, Elisabeth

back to syllabus

Precis

The most high-profile examples of prediction marketsthe Iowa Electronic Markets and Intradestarted by focusing primarily on predicting election outcomes and related political and financial events. Now they have expanded to cultural (Oscars) and technological (X Prize) events as well. The status of the commercial prediction markets is uncertain; for example, Tradesports announced recently that it is closing. And questions remain about the legal status of prediction markets, whether the CTFC will regulate them, and whether they will be taxed.

Rather than focusing on the traditional markets, however, we want to focus on future applications of prediction markets, particularly their possible use by government or by government-industry collaboration. We'd like to explore in particular applications that are likely to be controversial.

The focus will be three cases that we think raise interesting legal and ethical questions.

• Crime rate predictions, a la this proposal

• Google's flu-tracking application (where, as Professor Zittrain noted, the predictors aren't even aware that their knowledge is being harvested)

• the failed DARPA terrorism futures market

Possible Guests

We propose a panel (if possible) to discuss these ideas:

Justin Wolfers, Economist: Crime Rate Predictions

Bo Cowgill/Hal Varian, Google: The Epidemiology of the Flu

Guest TBD, DARPA: The Past and Future Possibility of an Intelligence Futures Market

Concrete Questions

• Which of these applications are most likely and desirable?

• Should the government be involved in administering prediction markets at all? Should it regulate them?

• What ethical concerns do we have about prediction markets of the future, and how might we address them?

Possible Experiment

Mostly just for fun, and to give the class a sense of how prediction markets work, we'd like to see whether we could get Google to let us participate in one of their internal prediction markets (e.g., when's the next iPhone coming out). Perhaps they'd never agree, but it could be interesting to see how our results match up to their internal results. This isn't necessarily tied to the focus of the class, which will be prediction markets of the future, but it could be cool.

Possible Readings

• academic literature on prediction markets, either generally or focusing on particular applications

• some of the discussion and congressional hearings related to the DARPA terrorism futures market

• relevant chapters from Professor Sunstein's Infotopia

• Michael Abramowicz's book Predictocracy

OLD STUFF

• To what extent should the government be engaged in the regulation of prediction markets; should it and how might it change current structures to be more accommodating?

• To what extent should government be involved in administering or using prediction markets (e.g., a la Hanson's suggestions)?

• For ethical or other reasons, should we be skeptical about using prediction markets for purposes such as predicting terrorist attacks and the like? What about for predicting regular crime (see this proposal)?

• More generally, if we think prediction markets are a useful tool, and yet it seems clear that they generate a considerable amount of unease, can we think about why and how policymakers might respond? Can design of the markets (reducing inaccuracy, or reducing concerns about rewarding misbehavior that might crop up if we have terrorism or crime futures) solve these problems or are some more fundamental?

Some more helpful material:

• Our very own Prof. Sunstein gives his comments on prediction markets and group deliberation.

Could prediction markets transform how we govern ourselves? Robin Hanson proposes Futarchy. The idea in brief:

"Democracies often fail to aggregate information, while speculative markets excel at this task. We consider a new form of governance, wherein voters would say what we want, but speculators would say how to get it. Elected representatives would oversee the after-the-fact measurement of national welfare, while market speculators would say which policies they expect to raise national welfare. Those who recommend policies that regressions suggest will raise GDP should be willing to endorse similar market advice."

Some tentative guest ideas

• Michael Abramowicz
• Justin Wolfers
• Bo Cowgill, Hal Varian: Google prediction markets
• Robin Hanson

Other ideas

One obvious thought is to see whether the class can play around with using prediction markets, though more thought needed on what we'd want to predict. Incentives for accurate predictions like t-shirts?

Will Harvard give us some small amount of money to invest for the semester? We could have an auction to determine whose investment ideas we use. The incentives would work so that you would only bid more to control the investment if you actually thought your investment idea would generate more net return to you (minus what you spent on the auction), despite it being divided up among the class.

Looks great. Harvard will not give us money to gamble (!), but some form of experiment would be fun, and maybe we can design one that draws in audiences outside the class. Might be good to think about prediction markets of the future -- perhaps ones where the ants making the predictions aren't even aware that they're doing so. (Consider Google's recent cooperation with the US Centers for Disease Control to associate a spike in searches for flu medicine with a prediction that a particular region will experience a health emergency.) JZ 16:28, 15 December 2008 (UTC)