Alfred Kahn on predation at air hubs: an AAI column

The following essay by Dr. Alfred Kahn, AAI Advisory Board member, appeared in the December 7, 1998 edition of FTC: Watch as the first of a series of AAI columns. These columns represent the views of the author and not necessarily the position of the AAI.

The promulgation by the Department of Transportation of a proposed new set of criteria for the identification of "exclusionary practices" in the airlines industry, accompanied by descriptions of incidents of apparently flagrant predation, has inspired intensified attention to the adequacy of the antitrust laws’ proscriptions. There are in fact strong reasons to believe that, at least so far as the airlines business is concerned, the Supreme Court’s view that "predation is rarely attempted and even more rarely successful" is simply incorrect1; and also that the dominating Areeda-Turner test provides an inadequate basis for identifying it.

I suggest three criteria, all of which would have to be satisfied if the response of an incumbent airline to competitive entry and/or price cutting is to be condemned: was the response (1) geographically and temporally pinpointed, (2) aggressive rather than merely defensive and (3) likely to drive (or to have driven) the competitor out and weaken competition.

The first and third are obvious: Robinson Patman prohibits price discrimination "where the effect … may be substantially to lessen competition or tend to create a monopoly." A recent report in The Wall Street Journal (November 23, 1998, p. B4) of Delta Airlines "last year…matching AirTran fares between Atlanta and Mobile but then raising fares to $404 round trip from $59 the day after ValuJet [sic] announced it was dropping service there" certainly seems to meet both of those tests.

As for my proposed second criterion, Section 2b of Robinson Patman exempts price reductions made "in good faith to meet" (but not beat) the equally low price of the competitor. The clear intention is to permit purely defensive responses. I suggest the critical test should be whether the incumbent carrier cuts its fares and/or increases its offerings of discount seats sufficiently merely to retain its previous traffic volume or increases its capacity and/or discount fare offerings substantially beyond that point. (In an apparently similar incident, cited by DOT, an incumbent increased its offerings of $75 discount tickets—on a route where its previous sales had been almost exclusively in the $300-$350 range—from fewer than 1,500 in the previous quarter to almost 50,000 during the quarter in which it faced the new competition, then cut them back to fewer than 1,000 after departure of the interloper.)

The logic of Areeda-Turner, which compares the reduced prices with the supplier’s variable costs, is that if the incumbent were taking losses out-of-pocket, it could be only in the expectation of driving the rival out and thereby permitting recoupment. When, however, the response involves a temporary expansion of capacity—which is possible almost uniquely in the airline industry, simply by moving planes into and then out of the contested market—the proper economic measure of costs is the net revenue that that capacity would have earned in its previous or alternative uses. Determining those opportunity costs is likely, however, to be an administrative or judicial nightmare. The substantial, temporary expansion of capacity and increased offer of deeply discounted seats should suffice to distinguish defensive from predatory meetings of competition.

DOT’s proposed rules go beyond Areeda-Turner in identifying as objectionable also pricing and capacity additions by incumbents that produce "lower local revenue than would a reasonable alternative response"—a clear and conceptually correct recognition of opportunity costs. The logic is the same: why would incumbents sacrifice profits that they could achieve by some alternative, less aggressive response, except in the expectation that the response actually selected would suppress competition and so permit recoupment?

Spokesmen for the major carriers profess to be shocked by the DOT’s suggestion that they be prohibited in these circumstances from expanding their flight frequencies, which would deny them the right to satisfy the expansion of demand for their services set off by the newly introduced price competition. In effect, they protest, the DOT would force them to ration their services, to turn away willing customers. The simple and sufficient answer is that the "yield management" (i.e., price discrimination) practices of the major carriers had already entailed just such rationingrestricting the offerings of discounted seats, both explicitly (limiting their number on each flight) and by the introduction of conditions (most prominently a Saturday night stayover) designed to confine them to demand-elastic customers.

Areeda-Turner assumes, implicitly, that sellers stand ready to satisfy whatever demand appears at prices that meet that test. That is simply not the case in the airline industry. In these circumstances, a sudden, dramatic and geographically highly selective change in a carrier’s rationing policy in response to competitive intrusion could surely be predatory in intent or effect—which is in my opinion of at least 45 years the proper antitrust standard.2

As I have already observed, the logic of the profit-maximization test implicit in the proposed DOT rules is the same as the logic of Areeda-Turner, extended to embrace opportunity costs. But by the same token, it could still exonerate responsive increases in capacity that should signal an aggressive, exclusionary intent and likely effect, because abandonment of the incumbent’s previous policy of rationing discount seats could be the most profitable course of action available to it in the altered circumstances, without regard to the possibility of successful predation: Before entry of the low-fare competitor, the marginal opportunity cost of offering however many discount seats travelers demanded would be sacrificed sales of full-fare tickets; after entry it might fall to something close to zero. Before entry, therefore, profit-maximizing would have dictated strictly rationing the availability of discount seats; afterward it could well call for the incumbent offering however many seats it could sell at more than marginal supply costs. And yet both the intent and the effect could be anti-competitive.

While I would therefore view a failure of an incumbent carrier to pass either the DOT test or the Areeda-Turner as prima facie predatory—at least in the presence of a dangerous possibility of an anti-competitive effect—I question the sufficiency of both these rules: predation may occur also in situations in which those conditions are satisfied.

In an industry that regularly practices direct quantitative rationing of low-price services, the test of whether a competitive response is "in good faith" must logically take into account not merely the price but whether it is accompanied by a change in that rationing policy. Rate reductions by incumbents designed to retain traffic that they would otherwise lose to competitors should be regarded as acceptably defensive; but expansions of capacity on the contested routes—whether or not justified under a profit maximization test—should be regarded as predatory or exclusionary in intent and/or raising at least a prima facie presumption of a dangerous probability of anti-competitive effect.

These generalizations do not make allowance for the likelihood of errors--in

particular, 'false positives' (identifying as predatory actions that are really not, and thereby discouraging beneficial competition). In the present condition of the U.S. airline industry—its progressive concentration at the national level, the emergence of monopoly power at hubs and the recent sharp increases in unrestricted fares—the greater danger is that the errors consequent on failing to make increased efforts to preserve opportunities for competitive entry outweigh the dangers of depriving consumers of the benefit of highly-pinpointed, geographically discriminatory and ephemeral competitive responses.


1 Ironically, the opinions of the Court on this subject clearly reflect what has come widely to be identified as the University of Chicago position at the very time when the consensus of academic economic opinion has moved to a "post-University of Chicago" position-a move from which I, as a "premature post-Chicago" economist of several decades' standing, take particular satisfaction. See, e.g., "The Economics of Antitrust," The Economist, May 2, 1998, pp. 62-64.
2 See my "Standards for Antitrust Policy," Harvard Law Review, Volume 67, November 1953, pp. 28-54 (also reprinted in Homewood-Irwin, Readings in Industrial Organization and Public Policy, American Economic Association, 1958, pp. 352-375.)