Bill Gates and the Irish potato famine
forbes.com
(have to go to archives, Feb 23)
By Peter Huber
Wheat farming: classic equilibrium. Potato famine: kinky demand curve. Information Age: supply-side surprises, in which a lower price elicits a greater supply. THE IRISH POTATO FAMINE, Bill Gates and your local phone company have something in common. Or so it seems to people who distrust free markets and favor government intervention.
It goes back to classical economics, which is built around a supply curve that slopes up and a demand curve that slopes down. Price changes push supply one way, demand the other; markets converge on the price where the two curves intersect.
The lines point the way they do for a reason. From the farmer's perspective, the supply side, the last bushel of potatoes is harder to grow than the first, because he plants good land before bad. On the demand side, people won't pay as much for the last bushel--they don't just want fries and more fries, they want a burger, too. That's what you see in the first graph above.
A monopolist, or a colluding cartel of firms that would otherwise compete, can, in theory, set price wherever it likes. So the government intervenes to bust up "unnatural" monopolies or to regulate the prices of "natural" ones.
But sometimes supply and demand curves take strange turns all on their own. They curl back on themselves, get snarled or race off toward the edge of the page without crossing. During the Irish famine, a perplexing thing happened. As the price of potatoes climbed, demand for them rose rather than fell (second graph). Why? People were so impoverished by the rise in the price of their most staple food that they could no longer buy meat. They had to spend every last farthing on starch.
Famines are history. Today's issue is the monopolization of desktops by Microsoft and Intel. These companies are said to operate businesses in which the supply curve takes on a strange slant, so that greater supplies are associated with lower prices.
Why should this be so? Because of the huge fixed costs in the business. If it costs you $1 billion to create a new operating system, and nothing to make copies of that software, then the usual rule about more supply requiring higher prices simply doesn't apply. To a lesser degree this is true of microprocessors, too. The first chip off the line is the costliest, and all subsequent copies just get cheaper and cheaper to make. The more chips you sell, the more widely you can spread your development cost, and the cheaper you can sell them. And the cheaper you sell them, the more chips you sell.
What about the demand curve? That, too, may take on a strange shape in the information economy. As with potatoes in the famine, though for quite different reasons, demand for Wintel products is said to slant up. Once it gets past some critical point, demand just keeps rising, because everyone wants to use the same platform, interface and language as everyone else. It was written in the curves that somebody had to be a monopolist. Intel and Microsoft just got lucky.
Okay, but then the Wintel duo pursued their luck with intelligence and determination. Is such a mix of luck and profit illegal? Economically intolerable? In markets like these, say a new generation of litigators and regulators, supply and demand balance only when the market has been saturated and monopolized, end to end. And who knows where the end really ends? Antitrust lawyer Gary Reback foresees Microsoft and Intel winding up in "total, undisputed control of the technology upon which the country's citizens and economy will depend."
People like Reback do not say, of course, that they are opposed to the free market; they profess to be saving it. They just need to shield competition from the pernicious threat of "increasing returns" and "network externalities."
Behind these high-sounding phrases lurks the old bureaucratic urge to have government take charge. An "increasing return" describes how Microsoft's profit margin on a software package goes up as the number of copies sold goes up. The "externality" is the economist's way of saying that people may buy the program not because it's better than competing ones but because so many other people already have it. Both notions are used to prove that Microsoft's success is not entirely deserved. Snarl your supply and demand curves just right, and you can predict any economic pathology, and justify any new economic prescription, that you like.
What Gary Reback is saying about the desktop software market, policymakers were saying a century ago about another high-tech industry-- telephone service. They were saying it on the strength of exactly the same two theories: "natural monopoly" skewing the supply side, "network externalities" skewing the demand. A monopoly phone company could provide service cheaper than two competitors with duplicate poles down the streets. And by connecting all to all, a monopoly would provide more valuable service, too. So competition was outlawed, and the anointed monopolist was regulated in microscopic detail for the next 80 years. It was a big mistake. We could have had all the savings of competition, in both local and long distance markets, decades earlier if competition hadn't been against the law.
Social engineers have been predicting chaos from abnormal supply and demand curves far longer than most of them realize. Thomas Malthus invented the business two centuries ago, with the famous prediction that food supply and demand would forever be in disequilibrium. French revolutionary theorists had announced that rational men would soon live prosperously without private property. In response, Malthus dashed off an anonymous pamphlet arguing that population increases geometrically, food supplies only arithmetically. Demand outpacing supply would always push mankind toward war, famine and misery.
Yet somehow, Malthusian prophecies never pan out.
As Peter Drucker recounts in Innovation and Entrepreneurship, in 1908 a Bell Co. statistician projected forward and concluded that every American woman between the ages of 17 and 60 would have to work as a Bell Co. switchboard operator by the year 1930. Thus, it was theoretically clear that a shortage of women would eventually limit talk on the telephone. In 1910, however, Bell engineers put into service the first automatic switchboard.
In the 1950s Herbert Grosch, a onetime IBM employee and later head of the National Bureau of Standards, concluded that a computer's power would inevitably increase with the square of its size. That meant relentless economies of scale. IBM projected that computing power for the entire planet could be supplied by 55 mainframes. Intel proved otherwise.
Markets often transform the demand side of the Malthusian equation, too. In prospering market-based economies, parents prefer raising fewer, better-educated children. Thus, it turns out that producing food abundantly seems to be the best way to curb population growth--a strange twist on things that Malthus never anticipated. Telecom regulators believed that people wanted no more service than the well-regulated monopolist had to offer. But competition has revealed a cornucopia of richly varied demand for wireless, Internet, on-line, and other fancy services that competition cooked up. In computing, the arrival of microprocessors exposed a huge reservoir of unimaginably diverse demand never dreamed of by Grosch.
Now the entire planet is supposedly going to be supplied forever by a Wintel monopoly of chip and operating system. Perhaps. But if history is any guide, consumers will end up buying silicon and software in far more varied arrays than drably uniform theorists can imagine. |
