Another interesting article from the NBER Digest: Macroeconomic Derivatives
. Justin Wolfers, of Wharton, and another economist studied the effectiveness of markets to purchase binary/digital options (so-called because you either get the payoff if the event happens or you don't if the event doesn't occur) in predicting macroeconomic events. Their findings were that forecasts based on these markets were at least as good as (and often superior to) surveys of forecasters/economists. A particularly interesting aspect of the research was that it might be possible to construct a probability distribution -- what they call a "density forecasts" -- using the market-based data. Creating actual probability distribution from economic/financial data is very hard to do for a number of reasons, so it is remarkable that one can use these binary options to create such "density forecasts."
A copy of the paper can be found here