I don't know what the literature agrees on, my own view is that switching costs are so large they constitute a de facto barrier. Imagine the overall cost of switching all spreadsheets to a new product world wide.
In his reply to your post, Regimond points out that Microsoft has created software to facilitate switching from the Novell/Lotus/Word Perfect networks to its own products. As this shows, competing companies can reduce the switching costs and, essentially, monetize them in the cost of their products.
This supports the view that path dependency and network effects are really manifestations of efficiency. Once it becomes sufficiently cost-effective, i.e., efficient, for consumers to switch to a new network or path, they will. In Bork World, these types of "barriers" are not a legitimate concern of antitrust enforcement.
I don't understand the difference between "natural monopoly" and monopoly well enough to comment.
Industry structures can be classified in two types. In one type, a typical firm's marginal cost of production rises above the competitive price level before all of the demand at that price is satisfied. Because of this characteristic, the market into which this industry sells its output will naturally support two or more firms and can be said to be "competitive."
In the second type of market, a typical firm's marginal cost of production will not rise above the competitive price level before all of the demand in that market is satisfied. In other words, even a single firm is capable of meeting all the demand for that market at a price higher than what it costs to produce the next unit of output. Such an industry structure gives rise to a market structure known as a "natural monopoly."
Garden variety monopolies arise in the first type of market. Garden variety monopolists are able to erect or use "artificial" barriers to entry to exclude more efficient competitors from what would otherwise be a competitive market. Since they face a sloping demand curve, they can restrict output and charge a higher than competitive price for what they sell. Such a firm will set output at the point where marginal cost equals marginal revenue and collect the difference between the competitive price and the higher monopoly profit-maximizing price as monopoly profit. (There is a nifty little chart on page 100 of Bork's book that lays this all out.) Bork and the rest of us agree that antitrust's mission is to break down barriers to entry and restore competition in this type of market.
Natural monopolists behave just like regular monopolists. They also face a sloping demand curve, and can restrict output and charge a higher than competitive price for what they sell. Such a firm will also set output at the point where marginal cost equals marginal revenue and collect the difference between the competitive price and the higher monopoly profit-maximizing price as monopoly profit.
The difference is that, due to the efficiencies of a natural monopolist's industry structure, the natural monopolist does not face an upward-sloping marginal cost curve and can produce a lot of output before marginal cost exceeds the price at which all of the demand in the market will be satisfied. It is thus more efficient at meeting all of the market's needs than any potential or actual competitor could possibly be (since the competitor will be producing at a higher point on the downward sloping part of the marginal cost curve). Hence, by dint of sheer efficiency, a natural monopolist will drive all competitors from the field and will produce 100 percent of the market's output. Any competitor that tries to enter the market will necessarily be less efficient than the firm that is already there, and will cause the monopolist itself to become less efficient to the extent the monopolist must restrict output to accommodate the new entrant.
Antitrust law arguably is not well suited to regulating a natural monopoly. Any effort to break up the natural monopolist in order to create a "competitive" market will exact a toll in efficiency, and the market will tend to revert back to control by a single monopolist. You can use antitrust law to curb predation by a natural monopolist, but, to the extent the natural monopolist is using predation to defend its monopoly, the behavior is efficiency-enhancing because it will (a) bring prices down to "competitive" levels (where price equals marginal cost), thus giving consumers back a share of the monopolist's profits and/or (b) deter potential entrants from trying to enter a market that can only support one firm.
Now, you might say, "Oh, but, if the monopolist engages in predatory behavior, it excludes rivals from having a fair chance to replace the monopolist through competition on the merits of its products."
My response would be that, in Bork World, in the case of a natural monopolist, that is not a legitimate concern of antitrust. In Bork World, antitrust is solely concerned with enhancing consumer welfare, aka efficiency -- not technological innovation, protecting rivals or any other value unrelated to efficiency. By prohibiting predatory conduct by a natural monopolist, the government is, in effect, favoring one competitor (or class of competitors) over another, which, in Bork World, is not a legitimate purpose of antitrust. In Bork World, antitrust should also be indifferent to the pace or direction of technological innovation, except to the extent that technological innovation enhances consumer welfare. This, of course, is objectively unmeasurable, except to the extent that consumers choose the entrant's products over those of the natural monopolist, and if the natural monopolist is willing to reduce prices to make the choice of his product more efficient than choice of the rival's, antitrust law should not care.
The appropriate regulatory response in the case of the natural monopolist is not to file an antitrust lawsuit, but to set up a government agency to set the natural monopolist's prices at competitive levels.
The government simply has to show that Microsoft is a monopoly.
I would argue that a case based on the theory that Microsoft is a "garden variety" monopolist in an OS market that would be able to support more than one firm is a very different case than the one the government has pled. By pleading that part of the barrier to entry that protects Microsoft from competition is network effects, the government is pleading facts which tend to show that Microsoft has a natural monopoly.
If the government wanted to, it could amend its complaint to take out the part about network effects. But then, how would the government explain how Microsoft arrived at its present monopoly position? If, for example, the government argued that Microsoft did so through predatory conduct (the usual way), that would require proof of a pattern of predation throughout the company's history. I'm not saying that historical pattern isn't there, just that that's a much broader and different kind of case than the one I think the government is trying to bring now.
Evidence of network effects would be irrelevant, indeed detrimental, to such a case, since such evidence would tend to show the presence of a natural monopoly.
And, if Microsoft is a natural monopolist, I would argue that the implications for what conduct should be illegal and what remedy is appropriate are very different than if Microsoft is a "garden variety" monopolist, for the reasons explained above.
The government has probably given these issues a far greater amount of consideration than I have, and I am not saying that the law makes the distinctions I am making here. It probably doesn't, and you may be right that the source of Microsoft's monopoly power is irrelevant.
What I am giving is a devil's advocate response to your argument. |
