As I have mentioned before, I have been watching the bird flu developments with concern. Seeing how things have spread (e.g., here and here [animation dated on later link]) – well the graphics help people to visualize what has been going on.
But what draws me to write this post is how markets and polls sometimes seem to diverge substantially in terms of predictions. For example, CNN reports that 60% in the US worry about the bird flu but that less than one-third think it will show up in the US this year. In December of 2005, InTrade (one of the exchange markets carrying futures contracts on whether bird flu will hit) announced, "Trading on Bird Flu — 65% probability of U.S. case by March 2006!".
Although I’m no expert in reading the financial stats associated with the InTrade contracts, what I glean from information is that the predicted probability of bird flu hitting the U.S. by March 2006 (as per the InTrade market) has fallen substantially since December. Of course, I’d venture to say that much of this has to do with the fact that the expiration period of the futures contract is approaching.
There is a general belief that an incented market (e.g., the Intrade market) where parties are financially motivated to make good predictions generates better predictive results than pure polls where people have no vested interest to be right or wrong in their predictions. In any case, betting for or against bird flu seems weird. But I suppose there is value in using these types of markets for planning purposes and for greater understanding.