Using prediction markets to solve complex, controversial problems

At his talk at the International Symposium on Forecasting, J. Scott Armstrong will discuss the use of prediction markets for solving complex, controversial problems like climate change. 

Abstract. Some scientists have proposed manmade CO2 as a cause of dangerous global warming. We examine the predictive validity of using CO2 as a policy variable. In doing this, we look at other possible relationships with global mean temperatures, including spurious predictors. We have previously shown that forecasts from a no-change benchmark  model — which predicts global mean temperatures will be the same as last year’s temperature for the next 100-years  — are quite accurate. Do prediction markets agree? We address this question and discuss whether prediction market  forecasts of global mean temperatures are useful. Finally, we describe our use of structured analogies to forecast the  outcome of the social phenomenon of alarm over dangerous manmade global warming.