Justin Wolfers in March 2007:

For those more interested in the actual research, our original research paper, recently published in the NBER International Seminar of Economics volume, is available here.

I also wrote a more digestible (that is, four page) version for the San Francisco Fed, available here. The NBER have also posted a useful summary of the paper, in the NBER Digest here .

The nice thing about this particular setting is that it really involves a very well-defined comparison between two alternative information aggregators: markets, and averages of expert opinion. And not surprisingly perhaps, the market comes out ahead. Equally, there is some good news for those economists who have used surveys as indicators of economic expectations: they seem to do a pretty good job, too.

Another real-money prediction exchange (betting exchange) should fill the void created by the demise of Economic Derivatives.

I started studying this market over a year ago now, and in a research paper with Refet Gurkaynak, we compared the forecasting performance of the markets, with the median from the usual consensus surveys run by Bloomberg or MMS. While the markets are still young (and hence our conclusions somewhat tentative) our research suggests that the market does a better job in forecasting both actual outcomes, and subsequent bond market responses.

Somewhat surprisingly, about a year later, the press are finally picking up on this story. Look to see a few stories hit the wires this week. Indeed, I did a spot on TV this morning. A copy of the Bloomberg spot is online here. (Warning: Large file.) At best, this is one of the few times you will see me in a tie.

For those more interested in the actual research, our original research paper, recently published in the NBER International Seminar of Economics volume, is available here.

I also wrote a more digestible (that is, four page) version for the San Francisco Fed, available here. The NBER have also posted a useful summary of the paper, in the NBER Digest here .

The nice thing about this particular setting is that it really involves a very well-defined comparison between two alternative information aggregators: markets, and averages of expert opinion. And not surprisingly perhaps, the market comes out ahead. Equally, there is some good news for those economists who have used surveys as indicators of economic expectations: they seem to do a pretty good job, too.

Justin Wolfers