Justin Wolfers in the Wall Street Journal:

[…] Through this process of different people trading based on their own observations about the race, prediction markets prices come to aggregate disparate pieces of information into a single summary measure of the likelihood of various outcomes. Moreover, if this market operates efficiently, it will appropriately summarize all of this information and the price will become the most statistically accurate forecast of the election outcome. […]

If I may, I would like to jot down some thoughts related to my concept of prediction market journalism.

1. The explainer on prediction markets is pretty good.
2. Crappy URL: http://online.wsj.com/article/SB119902559340658043.html?mod=rss_Politics_And_Policy
3. No way to leave a comment.
4. WSJ did list (in one of the sidebar boxes) BetFair along with InTrade &#8212-good point.
5. WSJ didn&#8217-t list NewsFutures and Inkling Markets but listed their own play-money, bots-driven prediction exchange (WSJ Political Market) &#8212-bad point (conflict of interest).
6. No external links embedded in Justin Wolfers&#8217- text &#8212-there are very good resources listed in the sidebar boxes, though (but the links use JavaScript and are not direct).
7. No static or dynamic prediction market charts, even though Justin Wolfers spent a good deal of air time analyzing the recent prediction market events &#8212-a concept he formalized with Eric Zitzewitz.
8. No tips &#8212-&#8221-I can&#8217-t predict what these trends will be […]&#8220-. Sounds like the prediction market approach (declaring that the market is a better forecasting tool than the polls or the experts) kills any anticipation and scenario planning. It shouldn&#8217-t be like that. Prediction market journalism can&#8217-t be only about analyzing the past. More on that in the coming weeks on Midas Oracle &#8212-not in the WSJ.

For all these reasons, I can give more than a straight B to Justin Wolfers&#8217- copy. You can do better than that, prof.