Friday, January 8, 2010
Posted by Herb Hovenkamp
The idea that FTC Act’s §5’s “unfair methods of competition” has a kind of penumbral reach to practices that are not covered by the Sherman Act makes historical as well as some logical sense. The framers of the FTC Act very likely held that view. While both Sherman Act provisions are quite open ended in their coverage, they also have limitations. Section 1 of the Sherman Act requires an “agreement” while §5’s “unfair methods of competition” does not. While application of §5 to anticompetitive interaction that does not satisfy the agreement requirement has been attempted several times with little success, there are ways that it could be given new life in the future. Indeed, explicit agreements are most likely to be found in markets where they are least likely to work. A second limitation is that §2 of the Sherman Act reaches unilateral conduct only if it “monopolizes” or creates a dangerous probability of doing so. Section 2 has no application to “leveraging” situations that European competition law’s “abuse of a dominant position” formulation (Article 82) permits.
There are also compelling institutional reasons why the FTC can move into areas where the Sherman Act has not. It has advantages in procuring discovery that can avoid the pleading problems with respect to covert activity that the Supreme Courts Twombly and Iqbal decisions have produced. The FTC commands more expertise and does not use juries. Most importantly, because private parties cannot enforce §5, the FTC can move more aggressively without worrying about the excessive litigation and overdeterrence that often results from private plaintiff lawsuits. The social cost of an error can be much lower when the only intervention is a cease and desist order against a challenged practice that cannot be shown to be beneficial.
Historically, the FTC itself has not realized these possibilities, mainly because of its tendency to condemn practices where there was no serious injury to competition under any reasonable definition. This was true of its exclusive contracting claims in Brown Shoe and Motion Picture Advertising in the 1960s. Unfortunately the FTC may be about to make the same mistake again in the Intel case, pursuing pricing practices that are very likely essential to cost minimization in the computer processor industries. Two problematic elements in its suggested relief are a requirement that pricing include an unspecified percentage of fixed costs, and prohibition of most market share discounts. Processor chips are characterized by relatively high development costs, which are both fixed and sunk, fairly short product lifecycles, typically on the order of two or three years, and relatively low variable costs. As a result, two things are true: first, a firm facing these constraints can minimize per unit costs by selling as many units as possible during the product’s life; and second, any incremental sale at a price sufficient to cover variable costs will make a positive contribution to fixed costs. As a result, the FTC may be ordering Intel to price less aggressively, which will keep Intel’s own costs higher. In addition, loyalty discounts are important in this setting because prices are bid up front and the price that can be bid is critically dependent on the number of units that will be sold. Quantity discounts will not work because they discriminate against smaller OEMs and will thus serve to concentrate the computer manufacturing market. In sum, before the FTC proceeds it must examine the economics of this market very carefully. The most likely result of its proposed relief will be the creation of a duopoly in which Intel and AMD both earn high returns. The FTC will have succeeded in protecting AMD, but at consumers’ expense.