Markets for standard, bulk commodities such as coal, lumber, and processed food operate according to diminishing returns: If a company were to keep expanding, eventually it would run into less demand for its brand or into higher costs of meeting demand. Such markets are shared and stable. But markets for high tech products-for pharmaceuticals, computer hardware and software, aircraft, missiles, telecommunications-operate according to increasing returns. If a product gets ahead in the market it gains further advantage.
There are several reasons for this. Technically sophisticated products, such as the B-2 bomber, or Microsoft Word 6.0, or Sun Microsystems' Java, have high upfront R&D costs, may be difficult to learn to operate properly, and may hitch people into a network of users. Larger market presence amortizes such upfront costs, familiarizes workers with use of the product, and makes it more attractive to join the product's network. Market success breeds further success.
As a result, if a company or a product gets sufficiently ahead, it may go on to lock in the market. And so in high tech, we often see near dominance of a single product-think of Lotus 1-2-3, the Boeing 747, Prozac, Windows 95, America Online, etc.-at least until the next wave of technology comes along or the product becomes a commodity.
In high tech, monopoly fortunes can be made, and are made. Management becomes a series of quests to find and lock in the Next Big Thing.
If monopolized markets abound in high tech, shouldn't the government step in?
Actually, I favor little direct government regulation of monopolies in high tech. But I also favor a much heavier emphasis on fairness in these markets.
Here's why.
If a standard commodities-market, for electric power, say, or airline transportation, falls under monopoly, potential over-pricing is the main issue. In high tech, monopoly pricing is less a worry. It may happen to some degree. But what counts in high tech is not pure price, but performance per dollar of price. And performance, whether it is measured in multiple instructions per second, or engine thrust, or pharmacological effectiveness, can change greatly as a function of future upgrades, of future technological developments-it can change greatly with future innovation.
Potential future innovation then is the key issue in high tech. I believe a government policy that might outlaw all monopoly of technical markets would set the wrong incentives and would severely damage innovation in the US. High tech firms put in day-and-night efforts to innovate precisely because lock-in, with its huge cash windfalls, is a potential prize for their endeavors. As with the great land rushes of the 1880's in the US, the one who puts in the most effective effort takes the prize.
But there is a danger. If a company wins land rush after land rush and can parlay its wins into great advantage in the next competition, it may be starting with a Landcruiser while the others have horses and wagons. In the 1994 Wilson-Sonsini white paper objecting to Microsoft's planned acquisition of Intuit (which I helped write) we complained that Microsoft's dominance of Windows gave it the potential of levering Windows' 60 million users into a potential First Microsoft Bank by bundling Intuit's Quicken with the next update of Windows. This would have been a ten mile start in the race for digital banking that no one else could have matched. Companies should not be able to lever winning in one market into the next.
In bringing Microsoft under investigation, the US. Department of Justice may find itself writing the rules for playing the game of high tech in the next century. I favor policies that do not penalize winners. I favor policies where, to whatever degree practical, each company starts behind the same starting line in each new race. And I favor policies that severely punish companies that hobble their competitors' horses in the dark.
The wonderful, wild spirit of innovation in America needs to have free rein. Let us adopt government regulations that do not penalize innovation. But let us keep these horse races fair.
W. Brian Arthur is an External Faculty Member at the Santa Fe Institute, Santa Fe, New Mexico. From 1983 to 1996 he was Morrison Professor of Economics at Stanford University.
He author of Increasing Returns and Path Dependence in the Economy,
U. Mich. Press, 1994, and of "Increasing Returns and the New World
of Business," Harvard Business Review, July-Aug. 1996.