(Up to: Economics )

From my original posting on Tribe:
Is there any work out there on some kind of hierarchical/meta-prediction markets?
i.e. if you have a relatively large number of markets, each predicting some facet, could you then combine the results of the markets, along with the inputs for those markets (i.e. the emergent demographics for the personalities contributing to the market) to gain more accurate predictions?

This, I guess, will depend on just how much variation there is between markets and their demographics, although this variation could occur as an outcome of various factors such as global location, technology used and community-of-interest-style groupings. For example, a prediction market restricted to face-to-face trading may differ to one on the internet, but based on a particular interest group, which may differ to one that has developed tools to assist their trading...

OK, I admit I'm a little lazy, but I'm also not sure what to search for in Google for this... metamarket.com is some thinktank thing, and the closest I can see so far is some Real Audio thing, which I'll try to listen to later:

www.gsb.stanford.edu/CEBC/Se...r_Archives/speaker4.htm

Perhaps, we also have a version of this in the global economy already, to some extent. But I'd be interested to see if it's been studied/implemented in a more academic manner.