What if it is true that: If you make Intel boxes and if you want to sell any MS OS, you must provide the MS OS with every box, or at least pay MS for every box even if some other OS went out the door with the box.
. . . per processor pricing. Handled by the Consent Decree, however inadequately.
Any way, a company does not need to be a bad guy to be a monopoly. The abuse of monopoly powers is the basis for specific action.
I totally agree.
To secure a violation of the anti-monopoly provisions of the Sherman Act, the government must prove that Microsoft has monopoly power in some market and that it has used exclusionary or other illegal means to protect or extend that power.
Monopoly power is defined as a firm's ability to raise prices above the level that would prevail under pure competition and to restrict output below what would prevail in a free market. Traditionally, market share has been used as a rough proxy of market power. Once a firm has 70 percent or more of a relevant market, it is assumed to possess monopoly power. Obviously, if a court uses the traditional test to decide whether Microsoft has monopoly power, Microsoft is going to lose.
To my limited way of thinking, the only way Microsoft has a chance of prevailing on the issue of whether it has monopoly power is to separate the idea of market share from monopoly power by using some sort of theory of Contestable Markets. The theory of contestable markets contends that "neither large size nor fewness of firms necessarily means that markets need perform unsatisfactorily. Impediments to entry and exit, not concentration or scale of operations, may be the primary source of interference with the workings of the invisible hand."
"Contestability theory focuses increased attention upon entry barriers and redefines their character. Economies of scale, for example, have frequently been considered an impediment to entry; contestability analysis shows, however, that they need not permit excessive profits or prices or any of the other manifestations usually associated with market power, even when scale economies make an industry a natural monopoly or an oligopoly. It is the presence of sunk costs rather than economies of scale that is of vital significance for both theory and practice."
Bailey & Baumol, "Deregulation And The Theory of Contestable Markets," 1 Yale J. Reg. 111 (1984). (Baumol also co-authored a book, which I haven't read, W. Baumol, J. Panzar & R. Willig, "COntestable Markets and the Theory of Industry Structure" (1982).)
Under this theory, despite its 80 percent plus market share, if barriers to entry are low, e.g., no legal or regulatory impediments, no sunk costs, then Microsoft's pricing will theoretically be disciplined by the presence of actual competition or the likelyhood of potential competition.
What kind of barriers to entry "count"? According to one source, "until very recently no competition authority . . . has attempted to formulate a coherent and detailed framework for the analysis of barriers to entry, despite the significant degree of effort that has been put into clarifying the related problems of market definition and the measurement of monopoly or market power." Furthermore,
". . . the economic analysis of barriers to entry has been subject to a remarkable amount of controversy since its inception in the 1940s and 1950s. Followers of Bain and the so-called Harvard school of industrial organization in the 1940s and 1950s tended to admit a great many things as barriers to entry and abuses of market power, while the Chicago school, represented in the writings of Posner, Bork, Stigler and others, is sometimes remarked as having had difficulty in finding any entry barriers at all."
This from an interesting survey of some of the recent literature in the U.S. and Europe on barriers to entry, which takes a decidedly more liberal approach than the Chicago approach, Harbord & Hoehn, Barriers to Entry and Exit in European Competition Policy, 14 Int'l. Rev. L. & Econ. 411 (1994).
Thus, some of the same ideological divisions noted by Page in the early history of antitrust law show up here in the literature on competitive vs. contestable markets and barriers to entry.
One thing that is very interesting is the proliferation of econometric models and theories created to justify various outcomes. It seems like every self-respecting law professor has his own model or theory, usually on why Chicago economics is wrong. In addition to "network externalities" (Lopatka & Page, 40 Antitrust Bull. 317), we have "contestable markets theory," "raising rivals' costs" ( ), game theory, and, perhaps one of the most interesting ideas I've seen, econometric modeling based on chaos theory. (Wise, "Antitrust's Newest 'New Learning' Returns the Law to its Roots: Chaos and Adaptation as New Metaphors for Competition Policy," 40 Antitrust Bull. 713 (Winter 1995).) This latter publication goes through a very interesting discussion of the history of marginal-cost economics as an attempt to apply the methodology of physics (and, in particular the mathematic tools used to describe physical phenomena) to the field of political economy. This history sets the author up nicely to argue that, just as chaos theory has undermined traditional static analysis in physics and the other hard sciences, so the same thing is starting to happen in economics, with obvious (at least to the author) implications for Antitrust law.
Econometric models are like MIRVs, easy to deploy and launch, difficult to shoot down and destroy.
The question that comes to mind, obviously, is, how is a judge or jury, who are nothing more than educated laymen and women, going to sort through all the social sciences gobbledy-gook to figure out what the truth is. Applying these new theories will place the judge and jury in an antitrust case not merely in the unwanted position of designing software, but also in the very difficult position of trying to figure out which, if any, of these econometric models and theories accurately describes real behavior by real companies and consumers.
One approach a court could take is to do what the district court in the Consent Decree case in effect did: defer to the DOJ as the agency responsible for enforcing the law, and just adopt wholesale whatever theory du jour DOJ comes up with. After all, DOJ represents the popularly-accountable executive branch. Another is to take the approach recommended by Easterbrook:
"Ignorance is the central problem in antitrust, in litigation, in life. We have trouble telling what business practices do. (When a manufacturer says it has adopted some restricted resale arrangement 'to protect my dealers' margins,' does that mean 'to create monopoly profits' or 'to induce the supply of services customers value'?) We have trouble applying models to the facts of cases. Explanations and data may not be developed until the case is over. Until we learn how to deal with ignorance, we cannot make use of fancy strategic models. And although Professor Hovenkamp tells us that a court 'cannot defer judgment until all the evidence is in,' it can and should wait for enough evidence to be confident. Until data permit a sound judgment that a certain type of practice is harmful, the courts should say that the plaintiffs have not carried their burden. In antitrust, as in other litigation, ties go to the defendant. The proposition that a court must leave the world as it is, unless there is a very good reason to compel a change, is not exactly novel.
Skepticism should rule the day." Easterbrook, Workable Antitrust Policy, 84 Mich. L. Rev. 1696, 1712 (1986).
If DOJ comes up with a convincing response to that argument, it wins.
(I think) |
