From CommerceNet Wiki

How could you implement experimental economic markets easily?

[edit] Overview

Markets are quintessential decentralized problem-solving systems. The act of trading spreads information about what's valuable, thereby directing people's attention and activity without explicit coordination. If one believes markets are a promising tool for attacking problems that cross organizational boundaries, why are there so few examples of them in use today?

There have been several notable experimental economic markets, such as Hewlett-Packard's future-sales predictions (profiled most recently in Tom Malone's paean to decentralization, The Future of Work), or the myriad politicial markets in elections (apparently, all the way back to the days of Abraham Lincoln!).

We suspect that one reason that markets have not been adopted more widely, or at least explored in a wider range of settings, is the absence of any freely-available standard toolkits. Granted, many kinds of new public prediction markets are hard to introduce given US gambling regulations in each state, but at least internal markets would be more easily constructed if a common foundation were available. A shared infrastructure would also mean that each new application could add to the reusable toolkit rather than rebuilding the ground layer yet again.

A research platform for electronic markets would respond to the call for shared infrastructure in a 1997 NSF workshop report on Netlabs (see our blog post).

[edit] Current Activities

Chris spoke at the Prediction Market Summit on December 2, 2005. The talk covered the current status of Zocalo and presented proposals for improving visibility and liquidity of existing Prediction Markets. Here's the abstract.

Chris Hibbert presented his proposal that CommerceNet should fund development of open source Prediction Market software at the DIMACS Workshop on Markets as Predictive Devices in Rutgers, NJ on Feb. 2-4, 2005, and subsequently wrote a white paper which we have issued as a CommerceNet Tech Report: The Zocalo White Paper.

CommerceNet accepted the proposal, and Chris has been working on development of Zocalo since March, 2005 as Principal Investigator, Prediction Markets.

Chris has been collaborating with [Ryan Oprea], [David Porter], and [Robin Hanson] of [George Mason University] on an experiment that is looking into the effects and effectiveness of manipulators in prediction markets. Here's a [replay of one session] of that experiment (FireFox only).

[edit] Prediction Market Resources

• MarketToolkits, Open Source and otherwise
• BackgroundReading on Prediction Markets
• CommercialPMVentures Commercial ventures in Prediction Markets
• PredictionMarketLinks Many more links on Prediction Markets
• Market Experiments inside companies (Original blog post)
• MarketExperiments inside companies (updated wiki page)