It sounds jaw-droppingly callous, not to mention absurd: An Internet
gambling parlor, sponsored by the U.S. government, on politics in the Middle
East. Anyone, from Osama bin Laden to your grandmother, can bet over the Web
on such questions as whether Yasser Arafat will be assassinated or Turkey's
government will be overthrown.
If the bettors are right, they'll win money; if they're wrong, they'll lose
their wagers. The site itself will keep numerical tallies of the current
"odds" for various events.
Why not just ask the guys at the corner bar whether or not we should invade
Jordan, or play SimCity to make foreign policy decisions?
But experts say the DARPA-backed Policy Analysis Market (www.policyanalysismarket.org) is based on a legitimate theory, the Efficient
Market Hypothesis, that has a proven track record in predicting outcomes.
Basically, the idea is that the collective consciousness is smarter than any
single person. By forcing people to put their money where their mouth is, the
wagers help weed out know-nothings and give more weight to the opinions of
those in the know.
"Markets are a great way of aggregating information that a lot of different
people have," said Eric Zitzewitz, an assistant professor of economics at the
Stanford Graduate School of Business. "One of the big issues with intelligence
that was gathered before 9/11 was that information wasn't aggregated within
the intelligence community. This is directly aimed at addressing that."
Although the idea sounds offensive to some, "to the extent this has even
a small probability of using valuable information to help prevent tragedies,
that's got to be the overriding ethical concern," he said.
In a paper published in February about the effect of the looming war with
Iraq on the stock market, Zitzewitz and two Stanford colleagues used
predictions from Web site Tradesports.com, which allows people to bet on
current events.
"We were able to get minute-by-minute estimates of the probability of war,"
he said. "When Colin Powell gave a speech at the UN that was hawkish, the
probability would blip up and the stock market would blip down."
Similar markets have predicted the box-office potential of movies or the
future prices of petroleum. The Iowa Electronic Markets, run by the business
school at the University of Iowa, lets investors buy and sell "shares" in
candidates. Some studies have showed it does better than pollsters at
predicting election results. Similarly, the orange juice futures market is
better than the National Weather Service at predicting Florida weather, some
experts said.
"It's a powerful idea and it works a lot of the time," said Mark Rubinstein,
a finance professor at the Haas School of Business at UC Berkeley. Still, he
cautioned, the government's application of the theory for intelligence
gathering has a number of potential drawbacks, ranging from whether terrorists
or others could easily manipulate the system, to whether government employees
might engage in "insider trading," leaking proprietary information to make a
quick buck.
Robin Hanson, an economics professor at George Mason University in Fairfax,
Va., has spent over a decade researching decision markets. Two years ago, he
joined up with Net Exchange, a San Diego firm, and the Economist Intelligence
Unit, an arm of the London magazine The Economist, to come up with the Policy
Analysis Market, in response to a government request.
"The basic idea of intelligence is to find out about bad things that might
happen. You pay people to find out about those things," he said in a phone
interview. "These markets are doing the same thing, offering to pay people for
information about bad things."
As compared to government studies, the online trading can provide real-time
information, he said. Hanson said the betting process helps attract more-
knowledgeable players.
He hopes to draw players who are well-informed, although he said "it's hard
to attract the people who know the most, because they are very busy people.
It's easier to attract the underlings who do a lot of the work and don't get a
lot of credit."
Hanson spelled out his ideal players: "People who work for the government
who are in intelligence organizations, people in think tanks, university
professors, people who are hobbyists in Mideast foreign policy, financial
traders or analysts who keep track of certain countries, the travel industry."
He welcomed the controversy that erupted Monday when two Democratic
senators denounced the project as "grotesque" and "unbelievably stupid."
"This is a low-budget operation," he said, noting that it's so far cost
less than $1 million. "This sort of national media attention is far and away
more than we can afford to pay for advertising."
E-mail Carolyn Said at csaid@sfchronicle.com.
