Scott Sumner wants one:
We need a subsidized prediction market in futures contracts based on global average temperatures and atmospheric CO2-equivalent levelswith contracts having 10, 20, 30, 40 and 50 year maturities. (Aaron Jackson and I did a paper on this.)
The paper Sumner appears to be talking about, "Using Prediction Markets to Guide Global Warming Policy," can be downloaded from Aaron Jackson's website. The authors address an obvious problem with the idea:
One obvious concern is that powerful special interests that have an economic exposure to climate policy might try to corner the market, by buying or selling massive quantities of global temperature futures, as a way of affecting the equilibrium price of these contracts. Evidence from the field, and from experiments by Hanson (2006b) and Hanson, Oprea, and Porter (2006), suggest manipulation in prediction markets is difficult at best. Even so, policymakers may still find it desirable to limit trading by companies with an obvious financial stake in global warming policy, and also limit the exposure of any individual trader.