Reading the Tea Leaves

Could an online terror-futures market predict where Al Qaeda will strike next? Early last week the Defense Advanced Research Projects Agency at the Pentagon was moving forward with trials of an Internet exchange for futures contracts on economic, civil and military events in eight Middle Eastern countries. Traders could bet, for example, on the prospects for Jordan's economy, the next suicide bombing in Israel or the assassination of Yasir Arafat. Then word got out. Sen. Byron Dorgan called the idea "unbelievably stupid" and "offensive." The New York Times called it "ridiculous" to think markets can predict the future. It didn't help that one key man behind the trial was retired Adm. John Poindexter, who last hit headlines in association with the Total Information Awareness program, an Orwellian-sounding plan to link all computer databases in a hunt for terror suspects. The day after Dorgan and Senate ally Ron Wyden publicly exposed the terror-futures market, the Pentagon killed it, and by the end of the week Poindexter had resigned from DARPA.

It's a wonder that Poindexter, who beat a charge of lying to Congress in the Iran-contra scandal, ever got a critical job in the war on terror. Still, he was on to something. Farfetched as terror futures sound, DARPA argued that markets have proven surprisingly "effective and timely aggregators of dispersed and even hidden information." That's the only kind of information available on terrorist plans, and the latest evidence on 9/11 shows that it just might have been prevented if U.S. intelligence had been quicker to act on scattered and hidden clues. "There is some amazing evidence to support markets as a predictor of future events," says Vernon Smith, the 2002 Nobel Prize winner in economics, adding that it is "ridiculous" to dismiss ideas like the terror-futures market out of hand.

Academic studies, for instance, show that orange-juice futures are often a more accurate indicator of weather than meteorologists. Game sites like the Hollywood Stock Exchange have a good track record in picking Oscar winners. Since 1988, the Iowa Election Markets (IEM), conducted by the University of Iowa, has consistently bested pollsters when it comes to predicting presidential-election results. Its record piqued Pentagon interest and was one inspiration for studying terror futures.

Why do markets work as crystal balls? It's not entirely clear, but part of the reason may be their impartiality. They tend to allow the best information to rise to the top and to override the bully factor, which is huge in the intelligence community. In any sector some voices will be forceful, others intimidated, "and all of that will influence the opinions that are shared," says Prof. Thomas Rietz, one of the directors of the IEM. "It's a leap from elections to geopolitical risk, but you could see why someone would make that leap."

Born during the cold war, DARPA has a mandate to dream up creative ways to defend the United States, and has worked on everything from unmanned vehicles to the prototype for the Internet. Nonetheless, Pentagon higher-ups quickly killed the $8 million terror-futures program as "too imaginative." One objection: that terrorists could in-filtrate an anonymous market and profit by betting on their own future acts of murder. But why not limit the market to experts, or at least vet the players? "We're surprised that we were never given a chance to respond to our critics," says George Mason University pro-fessor Robin Hanson, who was working with Net Exchange, the group of academics and entrepreneurs developing the market for the Pentagon. "You can ask any economist in the country about this idea. They may not like it, but they won't say it's crazy."

The project may have been politically tone-deaf. But it's the government that may lose out. DARPA was funding other market trials, including one designed to forecast changes in the new color-coded terror alert system, which could help states manage their security budgets. Now all these projects have been canceled, and we'll never know if they might have worked. At least Net Exchange will persevere: the DARPA controversy has inspired a flood of calls from business people who want to know more about how markets can tell the future.