The End Of Management?

(3 of 4)

Ultimately, "you may someday see someone in a desk job or a manufacturing job doing day trading, knowing that's part of the job," says Thomas Malone, a management professor at M.I.T. who has written about markets. "I'm very optimistic about the long-term prospects."

But no market is perfect. Economists are still unsure of the human factor: how to get people to play and do their best. In the stock market or even the Iowa prediction market, people put up their own money and trade to make more. That incentive ensures that people trade on their best information. But a company that asks employees to risk their own money raises ethical questions, so most corporate markets use play money to trade and small bonuses or prizes for good traders. "Though this may look like God's gift to business, there are problems with it," says Plott, who ran the first HP experiments. Tokyo-based Dentsu, one of the world's largest advertising firms, is still grappling with incentives for an ad forecasting market it will launch later this year with the help of News Futures, a U.S. consultancy.

And even if companies can figure out how to make their internal markets totally efficient, there are plenty of reasons that corporate America isn't about to jump wholesale onto the markets bandwagon. For one thing, markets, based on individuals and individual interests, could threaten the kind of team spirit that many corporations have struggled to cultivate. Established hierarchies could be threatened too. After all, a market implies that the current data crunching and decision-making process may not be as good as a gamelike system that often includes lower-level employees. In a sense, an internal market's success suggests that if upper managers would just give up control, things would run better. Lilly, which is considering using a market to forecast actual drug success, is still grappling with the potential ramifications. "We already have a rigorous process," says Lilly's Bingham. "So what do you do if you use a market and get different data?" Throw it out? Or say that the market was smarter, impugning the tried-and-true system?

There could be risks to individual workers in an internal trading system as well. If you lose money in the market, does that mean you're not knowledgeable about something you should be? "You have to get people used to the idea of being accountable in a very different way," says Mary Murphy-Hoye, senior principal engineer at Intel, which has been experimenting with internal markets. "I can now tell if planners are any good, because they're making money or they're not making money."

That is one reason Intel has been bandying markets about for more than two years but has yet to implement them in a real-world scenario. It's not for lack of good results. In a laboratory experiment run with M.I.T.'s Malone, Intel used a market to make a coordination decision: which factories should produce computer chips and when. In the experiment, a centralized, strategic plan was replaced with a market in which salesmen and a plant manager traded futures contracts representing chips. The result was nearly 100% efficiency in allocating manufacturing capacity. That experiment echoed another, real-life market triumph. In 1998 oil giant BP set out to reduce company-wide greenhouse emissions 10%. Instead of issuing plant-by-plant dictums, the company let plant managers trade permits to produce emissions. Managers who could quickly get their plants into compliance and reduce emissions even further could sell their permits to other plants. BP hit its reduction target — nine years early.