Decision Market Predictions: Separating Efficiency and Bias

Information markets provide a revolutionary mechanism for revealing aggregate knowledge. They continue to gain popularity in academic circles and continue to prove themselves through successful prediction for practical applications. Policy makers have shown to be wary of depending upon their results exclusively for decision-making, but there is evidence that information markets could be used for relevant advising. One particular type of information market uses contingent contracts that speculate about unrealized states of the world. These estimates hold an inherent bias. This study develops a model for a more efficient estimator and how to acquire it through conventional contracts. * I would like to thank Eric Zitzewitz for all of the suggestions, guidance and resources he provided throughout the project. I would also like to thank Sinead Kelliher www.tradesports.com for generous help with the data. Glenn Bean: Undergraduate, 2005, Stanford University; [email protected]