Corleone vs. Shemp

It’s ironic that on the same day Intel announced it was settling the Federal Trade Commission’s lawsuit against it, Advanced Micro Devices–Intel’s chief competitor–announced it was going to report “significant” losses in its first quarter because of a major design flaw in its key microprocessor.

What the AMD announcement underscored was the flimsiness of the central pillar of the FTC’s case against Intel, which is that the chipmaker has monopoly power in the microprocessor market. Intel’s dominance is not monopolistic dominance in any sense that matters. Its dominance is the product of its own manufacturing and design excellence and of the manufacturing and design incompetence of its competitors. In settling the suit before it went to trial, Intel may have made the right move from a business standpoint. But from an intellectual and legal standpoint, it conceded more than it needed to.

The Intel case centered on the company’s aggressive tactics against smaller rivals with whom it was engaged in disputes over patents and copyrights. These tactics, most notably the cutting off of contracts and of access to information about Intel’s next generation of chips, would generally be considered legal if a normal company pursued them. But if a monopolist uses these kinds of tactics they become–justifiably–illegal. If Intel really had a monopoly, then cutting off smaller companies was tantamount to a death sentence, one that these companies would presumably do anything–including yielding in patent disputes–to avoid.

The problem, though, is that there is no sense in which Intel can meaningfully be considered a monopoly. Although the company’s Pentium chips are the industry standard for PCs running Windows, you can buy PCs with chips made by AMD or by Cyrix that run as fast as Pentiums and that let you do everything with a PC that you could want to do. In the fourth quarter of 1998, in fact, Intel’s share of the microprocessor market fell to 76 percent, down from 85 percent a year earlier, and some recent studies suggest that AMD chips now actually have a larger market share of the retail desktop PC market in the United States than Intel chips. And Intel has been able to retain its huge overall market share only by cutting prices more aggressively than in the past and introducing a whole new line of chips–Celeron–to compete in the low-end market. These are not, needless to say, strategies normally favored by monopolists.

One might argue that 76 percent of the market is a lot for any one company to control. But Intel’s hold on that market is precarious–far more precarious, for instance, than Microsoft’s hold on the operating-systems market–and it has been able to sustain that hold in no small part because its competitors have been so bumbling. AMD has been supposedly on the verge of real profitability for years now. With each new generation of chips, its management has promised that finally the company has achieved technological and production-line equivalence with Intel. But with each new generation there’s a whole new set of problems. If Intel over the past ten years has been as crisp and ruthless as Michael Corleone, AMD has been as clumsy as Shemp. And to say that Intel should somehow be punished because it can execute when it has to and its competitors can’t is to turn antitrust law into what it was never meant to be: an instrument for equal results.

You might say that Intel’s very decision to settle the suit was a testament to its lack of monopoly power. Based on Microsoft’s experience in the courtroom, Intel was apparently concerned that its image would suffer from revelations about its rough business tactics, and that this might hinder its rollout of the new Pentium III chip. But surely if Intel was really a monopoly, this wouldn’t matter, because consumers would have nowhere else to turn. It probably does make sense for Intel to have settled. But the FTC shouldn’t take the decision as confirmation of its dubious logic.