Paul Rhode's "skeptical perspective" on corporate prediction markets:
In many ways, tapping of the wisdom of crowds within the firm is intended to overcome the information barriers created by the bureaucracy. It is obvious that the upper management might want better access to selected information available down the organizational ladder, to stop having to listen to the self-serving lies of middle management. But it seems to me, the individuals in an organization derive their power from the information under their exclusive control and will not easily give up this monopoly position. What models we economists have about hierarchies largely concern controlling information flows, both up and down the organization. This includes both having the higher-ups monopolize the firms’ secrets and strategies and preventing them from being overwhelmed by the day-to-day minutia.
With internal prediction markets, key questions include who will set the agenda, who decides what questions will be answered and how? It seems authority matters in whether this is done in a top-down or bottom-up manner. If the question is what is the best forecast for demand growth, will this deadline be met, or how will the product rank in quality tests, it is clear that upper management, the “deciders,” would be happy to learn from the collective wisdom of employees in contact with customers or doing the design work. If the questions posed address how long before the company president is fired, whether this product is found defective and has to be recalled, or when the mass layoffs will begin, then upper management will be unhappy.
Prediction markets provide more information, but they do so in a public way. What prevents competitors from spying, from gaining access to company secrets? Besides making private information common knowledge, prediction markets undermined the mystique, the information monopoly of those in charge.
I agree completely and have said similar things many times. So why am I called a "hyperbolic" optimist? Today I speak at a corporate prediction markets summit in NYC.
See also the SNAFU principle.
Rhodes' comments on controlling information reminded me of Janis' book on Groupthink. One of the problems leading to groupthink is that information flows within decision making institutions are not linear (passing easily up and down the chain of command). The use of prediction markets, I think, allows for the use of information that some stakeholder may have and not wish to disclose. Without disclosing the information, the stakeholder can use that info to make their prediction. When a large enough group of people make use of their private information as predictions, then it may be possible to reach the best decision, without the stakeholder giving up control over their privately held information.