How biases can help with decisions
When a judging scandal marred the pairs figure skating competition in the 2002 winter Olympics in Salt Lake City, economist Eric Zitzewitz saw the perfect research subject. Behind the allegations of vote-trading and ties to the Russian mafia was a case study in how organizations make decisions when some members are clearly biased, says Mr. Zitzewitz, a business professor at Stanford University in California.
Surprisingly, his findings suggest that biases may help with decisionmaking. Important information is lost when "outlying" opinions are eliminated, he explains in the spring 2006 issue of The Journal of Economics and Management Strategy. Worse, "truncating" opinions makes it easier for an organization, such as a panel of Olympic figure-skating judges, to fall into vote-trading.
"I think people have this intuition that if someone's saying something very different than what other people are saying, they assume they're wrong or just biased," Zitzewitz says by phone from his Stanford office.
Instead, organizations should extract the useful knowledge from biased opinions while correcting for the bias. "You might imagine designing a process where you give people with strong opinions more weight," he says.
Under the old judging system, national skating federations chose the judges who went to the Olympics. The federations appeared to favor judges who gave their compatriots higher scores, Zitzewitz says. He determined this by analyzing scoring data, employing a technique used to analyze forecasts by stock-market analysts.
In his research, Zitzewitz found that figure-skating judges gave competitors from their native countries higher scores than other competitors - an average of 0.7 positions higher. "The data suggest that countries are divided into two blocs, with the United States, Canada, Germany, and Italy on one side and Russia, the Ukraine, France, and Poland on the other," Zitzewitz writes.
Once assembled at the Olympics, these apparently patriotic judges caused problems, the study says, because of the way scores were counted. If the competition came down to two skaters, their rankings would be determined by which one was higher on a majority of the judges' scorecards. So if a majority was created by judges colluding to reinforce each other's biases - as the French and Russian judges allegedly did in 2002 - then the opinions of more objective judges wouldn't count, even if they more accurately reflected how skaters performed.
This Olympics judging scenario can translate to the boardroom, Zitzewitz says. A committee deciding whether to promote an employee shouldn't discount the extreme opinion of a manager who has worked closely with the employee. Not only does excluding the opinion bury useful information, it can also push that manager to engage in vote-trading with regard to other promotions, as in, "If you vote my way on this one, I'll vote your way next time."
If the decision requires a simple majority, deals are easier to make, Zitzewitz says, so eliminating those outlying opinions eventually means you've lost out on who skated best or would be the better person to promote.
This kind of collusion can come into play in political settings, notes Gary Charness, an economics professor at the University of California, Santa Barbara. "The cautionary note here is that while excluding extremes looks good, it may come at a real cost." he writes in an e-mail.
The International Skating Union voted to change the judging system in June 2002, four months after French judge Marie-Reine Le Gougne said she had been pressured by her country's skating federation to give a higher score to the Russian pair, bumping the Canadian pair down to silver. In the wake of the scandal, the International Olympic Committee awarded the Canadians a duplicate gold medal.
The lesson for organizations, Zitzewitz writes in his study, is that the decisionmaking process is more efficient when its participants are most interested in fairness. The apparent nationalism of figure-skating judges trumped that desire, Zitzewitz says.
A "strong but disinterested" group leader could have prevented judges from exaggerating their opinions, he adds.
"Without that chairperson who's going to say 'All of your opinions have been too extreme lately, we're going to listen to all of them less,' [it can] degenerate into a place where every opinion is extreme," Zitzewitz says.
At the heart of the study is the individual whose actions instigated the 2002 judging scandal: Ms. Le Gougne. Last month, the former Olympic judge told the Associated Press that her suffering over the scandal was worth the positive changes it's brought for the sport. "The judges came to see me and said, 'Marie-Reine, the new scoring system is so great, thank you Marie-Reine because without you, there would not have been a new scoring system,' " she said at a skating event in Paris.
But not everyone is as optimistic about the new system, in which nine scores are randomly selected out of 12 judges. The high and the low are dropped, and then the score is averaged and posted anonymously. No longer can spectators boo an ungenerous judge from a certain country, because no one knows which score belongs to whom - not even the judges themselves.
"That's potentially a less positive thing," Zitzewitz says of the anonymity. The new scoring system is, however, more objective, Zitzewitz says, and judges are now chosen by the International Skating Union rather than the individual nations' skating federations.
The new scoring system's popularity faced its first test over the weekend, with medals for the pairs skaters being awarded Monday night.
Perhaps the most surprising aspect of Zitzewitz's study was how many economists have had a former life in competitive figure skating. "I had students and people from industry conferences approach me and want to help me," Zitzewitz says.