Krugman seems to think the conventional wisdom that prediction markets are worthwhile is bearing out to be false:
"[P]rediction markets — which you see, again and again, touted as having some mystical power to aggregate information, know no more than the conventional wisdom."
I think Krugman's assertion both false and correct at the same time. There is a substantial body of literature that indicates that prediction markets outperform pollsters in predicting elections. However, in this election, both prediction markets and pollsters seem to be wildly off. For a market to function properly it requires information and in the case of prediction markets this has been recently hampered by restrictions on online gambling. InTrade and TradeSports (both incorporated in Ireland) will not accept any online payments originating from the U.S., now maybe the Iowa Electronic Markets are different, but this would tend to remove those with the best information about U.S. electoral politics (U.S. residents) from the betting process and thus depriving the market of information save for that which is reported in the press.