Could Prediction Markets concept be applied to government projects?

Motorola uses a prediction market to gather new ideas from employees and predict their success based on how willing individuals are to buy “stock” in them. They launched a system called ThinkTank in 2003 to collect potential innovations. Employees are given $100K in virtual dollars to vote on ideas at $10 per share.

Perhap government staff could participate in a similar effort to stimulate innovations and involve more people in the decision process. I would suggest that a small group review the finalists, assess scope/cost and make final recommendations.

Has anyone tried this or something like it? Read more about Motorola’s use of prediction markets in CIO article by Kirstin Burnham.

Leave a Comment


Leave a Reply

Al Dominick

FYI: Best Buy did something similar to Mortorola to forecast demand for a new product or idea. The Wall Street Journal ran an article last year on their use of predictive marketing. Also, the University of Maryland’s business school also does a lot in this space and might have some anecdotes about govt’ using such a concept…

Adriel Hampton

This as the idea behind the terrorism futures market, which had very poor PR …

The concept, however, is sound and should be applied more in gov and private enterprise. Election predictive markets are very popular already.

Adam Siegel

Idea markets can be tricky. They often turn in to “beauty pageants” and you’re not quite sure why people are buying those stocks – is it because it really is a good idea? Do they just like the person(s) who submitted it? Prediction markets are intended to have questions that have some sort of objective, verifiable outcome. For government use, predicting performance on key metrics set in legislation, forecasting budgets, and understanding the probability of risk factors occurring would all be great use cases for government. What also may be even more lucrative is the ability to easily cross organizational boundaries to solicit the opinions of people from many disciplines and agencies on a given topic. Take questions related to energy independence. The DoE could easily sponsor markets whose participants were from the various agencies in DoE, but also from Agriculture, Commerce, Treasury, Defense, etc. Try doing that with a pure data-centric approach and it will be years before you get anything out of it.

David Dickerson

THis has been tried – DoD tried to develop a market that included predictions of terrorist attacks – it got shut down. I am a big fan and this concept and believe it is a strong method for senior levels to understand mid-level managements true feeling about things. From my time running federal web sites and developing portals we knew day one which projects were worthwhile and which were a flash in the pan.

Scott Webb

We did something like this at the Monterey Institute of International Studies (MIIS) the last 2 years. The first year, we used “Inkling Markets” which was just OK. We let the market go too long, there was a finite number of “Inklings” to invest, and people were gaming the system to some extent. We also only opened the market to MIIS students, faculty, staff, and alumni. so the market was a little too small. I agree with Adam’s comment about beauty pageants – you have to guard against that. Inklings works well if you limit the time or open it up to thousands upon thousands of people. Shorten the time limit if you’re only surveying a relatively small group, and also try to make the presentation of all the ideas equal (consistent formatting) – so they can be judged equally. Otherwise people can be influenced by campaigning. It all depends on what you’re trying to discover.