So we decided to move beyond asking, “Do prediction markets work?” and instead use them as a tool for better understanding how information flows within a (very cool) corporation.
I am more interested in the accuracy of the enterprise prediction markets than in corporate micro-geography issues.
Related Links: Using Prediction Markets to Track Information Flows: Evidence From Google – (PDF file – PDF file) – by Bo Cowgill (Google economic analyst), Justin Wolfers (University of Pennsylvania) and Eric Zitzewitz (Dartmouth College)
I’m on record as thinking that more research is needed about enterprise prediction markets and the flow of information inside organizations. See, for example, my post here at Midas Oracle last October, titled, Prediction markets and the flow of information inside organizations. That post was more oriented toward information flows up/down the corporate hierarchy. The Cowgill/Wolfers/Zitzewitz paper doesn’t address this particular issue, but the results are interesting nonetheless.
.
I agree with Chris (and Robin Hanson’s view expressed in the related Marginal Revolution post comments) that “prediction market accuracy” is the big money issue. The Google paper only hints at the answer through the discussion of bias. The authors could have said more.
.
But while the accuracy issue remains significant, I see that as no reason that every prediction market paper has to focus on that issue. The Google paper helps broaden and deepen the literature on enterprise prediction markets by digging into new issues.
Chris, you should be more interested in info value, which is the added accuracy the markets provide relative to other mechanisms, times the value of accuracy in improved decisions, minus the cost of maintaining the markets, relative to the cost of other mechanisms. A highly accurate market has little value if other mechanisms can provide similar accuracy at a lower cost, or if few substantial decisions are influenced by accurate forecasts on its topic.
Do Google’s enterprise prediction markets work?
http://www.midasoracle.org/2008/03/26/google-enterprise-prediction-markets-4/