I think the article in the May 22, 1998 Wall Street Journal that Bork wrote, called "The Most Misunderstood Antitrust Case" is must reading for anyone who wants to understand what this case is about.
He wrote it in response to an article by George Priest of Yale called "A Case Built on Wild Speculation, Dubious Theories," which appeared on May 19.
Without reproducing the article in full here, the most important "choice cut" is the following:
Microsoft's aim is entirely different: to preserve its world-wide monopoly in operating systems by stifling companies whose technology would compete. True, the company makes no additional monopoly profit; in fact, the tactic costs money because Microsoft prices its browser at zero. But that is a rational investment given Microsoft's enormously disproportionate resources and the fact that costs do not rise commensurately with output. These aspects of the software industry, among others, make predation a feasible tactic. The tactic is not feasible in the usual case where the predator and victim have resources in proportion to their market shares, and the predator's increased output causes costs to rise even more rapidly than the output.
Bork identifies in his book a fundamental assumption which makes predation irrational. His analysis rests in part (but only in part) on the assumption, as stated at page 149 in my edition of his book, that
"Losses during a price war will be proportionately higher for the predator because he faces the necessity of expanding output at ever higher costs, while the victim will not only not expand output but has the option of reducing it and so decreasing his costs."
We see from the excerpt in the wall Street Journal article that Bork does not believe this assumption to hold true for the OS/SOftware industry, a conclusion which is consistent with the literature on increasing returns and network externalities (at least what little of it that I have read so far).
However, this is only one of the conditions which, according to Bork, make predation irrational.
At pages 148 to 159 of his book, Bork lays out his overall theory of predation. I doubt it has changed much since 1978, when the book came out. He identifies several different forms of predation: price cutting, disruption of distribution patterns, and what he calls misuse of government processes (litigation and the like).
The most interesting comments for purposes of this case are on page 158, where he says:
[T]here is no doubt that predation can succeed when the distribution pattern [from which rivals are excluded -- GRL] is so much more efficient than the alternative that those forced out of the pattern cannot compete. The technique of predation is the denial of access to an essential economy of scale. Bords of trade, for example, often control such access, and their members may often easily destroy a troublesome rival by expelling him from membership or, perhaps more commonly, may bring a rival into line with the mere threat of expulsion.
And, I suppose that DOJ is arguing that the same thing could be said about channels of distribution in the software industry. Microsoft's exclusion of Netscape from the channels of traditional distribution, such as ISP's, OLS's and advertising by and payments from ICP's, is relegating Netscape to access to distribution over the internet and denying it access to "an essential economy of scale."
However, Bork, in his book, points out that predation is made irrational, and therefore unlikely, by a number of additional factors. He points out that the disparity in capital reserves between predator and prey will not be as great as a simple comparison of resources would suggest because the victim will have still access to capital markets. In addition, he identifies four other reasons, unrelated to rising marginal costs, why predatory pricing is irrational:
(1) in order to succeed, a predator must drive rivals out of business as opposed to merge with them, since the law prohibits monopolization by merger. By driving down the value of a rival's business, Bork says it will make acquisition and resumption of the battle by a better-funded rival more probable.
(2) Ease or difficulty of entry and exist are symmetrical. If entry is easy, then once the strategy succeeds, new entrants will simply replace the existing firm, while if sunk costs are high, making entry difficult, firms will be reluctant to exit the market, making predation that much more difficult.
(3) Future monopoly profits must be discounted by the rate of interest.
(4) Demand becomes elastic over time, reducing the likelihood of future monopoly profits.
I could be wrong, and I won't predict an outcome, but it seems to me this is going to be the real battleground in this case. It seems to me that these factors could apply with equal force to all of the forms of predation Bork identifies. It may require expert testimony to establish whether marginal costs do not rise as quickly, or may even fall, in the software industry, as network effects and increasing returns would suggest, and, if so, whether that fact outweighs these other factors that render predation irrational. |
