AAI Files additional comments with DOT about predation at air hubs


September 10, 1998


Many of the comments in this docket argue that D.O.T.’s proposal must be rejected as incompatible with the antitrust laws because it is not cost-based, because it doesn’t require a showing of recoupment, and because it is too vague. In response, we say that within the peculiar circumstances of the air transport industry, a predatory (i.e., illegal) method of competition may be recognized and dealt with quite consistently with the antitrust laws, without specific reference to the predator’s costs and without proof of recoupment in the particular local market. Problems of vagueness can be reduced in a variety of ways.


  1. Cost-based tests are not appropriate for this industry. It is consistent with the spirit of the antitrust laws to provide alternative tests for predation. The Supreme Court has said that there are two elements which must be proven in acase of predatory pricing: that the predator has set a price below an appropriate measure of its costs, and that there is a reasonable danger of it being able to recoup the investment in below-cost losses.1 To understand what would be an appropriate measure of cost, one must look to the purpose of even discussing cost: namely, to help define a recognizable line between aggressive, competitive price cutting, which is a normal and desirable part of competition; and anticompetitive price cutting, which is harmful not only to particular competitors, but to the competitive process. As the Court noted, there is no consensus on what constitutes an appropriate measure.2

    Today’s world of scheduled air carriers, working out of a hub-and-spoke network, is fundamentally different from the world in which Areeda and Turner and others debated about cost-based tests of predation. In this world, the very concept of “a” cost for a particular spoke is a chimera. And when the airlines argue that we already have antitrust laws that deal sufficiently with price predation, on the basis of a cost-based test, they are wrongly implying that the antitrust laws offer a more certain method for coping with air hub predation.

    According to the declaration of Laura D’Andrea Tyson, the major airlines today are multi-product companies.3 Even for a given airplane, each seat may represent a different product, reflecting its own costs and its own demand. This means that each seat on the plane can have its own separate price. (Though we speak of “buckets” grouping seats into many categories, in theory each seat could constitute a separate bucket.) The price would represent a package of variable costs (it costs more if a ticket is refundable and is purchased at the last moment) and demand (a business traveler needing to attend a hastily scheduled meeting will pay more than one who can schedule in advance, and will pay even more if there are no competitive carriers for the route in question). Thus, we have an industry that is arguably characterized by each seat having its own cost/price justification.

    Next, we have an industry in which the marginal costs or average variable costs (the Areeda-Turner substitute for marginal costs) are extremely low.4 If the plane is going to fly and it has an empty seat, the marginal cost for that seat is next to nothing. But this makes no sense as a test for predatory pricing because it is so much less than the average cost for seats on the plane or within the company’s overall operations. In other words, if the airline had to charge the marginal cost of one seat for all of its seats, it would quickly be bankrupt. A newcomer whose costs are not amortized over an entire large system and lacking external sources of deep-pocket capital, could rarely if ever match a national airline’s marginal costs, no matter how efficiently it operates. Thus, the marginal cost test would not reflect a test of relative efficiencies; it would be a sentence of death to new entrants.

    Finally, we have an industry which is characterized by networks. We are told that if one spoke in the network is damaged by a new entrant’s pulling away passengers from the dominant carrier, the entire network is endangered.5 Thus, the costs of operating the network are joint, and cannot be isolated to one particular leg.

    Cost-based tests for predation have proven to be difficult even in the most stable and simply-structured of manufacturing industries. But in an industry characterized by multiple products, extremely low marginal prices, and joint network costs, the cost of a particular seat on a particular leg is impossible to calculate.

    Under these circumstances, two approaches seem possible. One is to construct a carefully defined version of “cost” which makes sense for the air transport industry’s peculiar characteristics. But this would ultimately be arbitrary. The other is to find substitutes for a cost measurement that capture in a common sense way the distinction between fair and unfair competition. We believe that DOT is attempting to do this, and that the attempt is within the spirit of what the antitrust laws are intended to accomplish.

    The statute under which DOT makes its proposal6 is based on the FTC Act, and DOT can therefore look to the flexibility of the FTC’s “unfair methods of competition” language.78 Although the language is not open-ended, it gives the FTC authority to go beyond the literal limits of the Sherman, to interpret and apply the “spirit” of that law, to deal with incipient violations, and to fill in gaps.8 Generally, courts give great weight to the agency’s expertise in analyzing an industry. We think a strong argument can be made that DOT has authority to apply its expertise in the air transport industry in such a way as to determine that a recurrent pattern of behavior which has the result of restricting new entry on a basis other than comparative efficiency, represents an unfair method of competition.




  2. In the particular context of this industry, it could be consistent with the spirit of the antitrust laws for an expert agency to find a reasonable likelihood of recoupment from a pattern of alleged predation.


Without being able to focus on specific cost data, the requirement to find recoupment cannot be one of simply accounting for the future recovery of below-cost losses. Rather, the focus must be on foregone revenues.9 In our original comments, we spoke about the strategic ways in which a predator can recover its investment. Here we simply emphasize that the context of a major airline operating in multiple markets with a complex network provides many opportunities for recoupment of foregone revenues. It should not be necessary to recover in the same market where the predation occurred. Merely scaring off future entrants elsewhere in the system through building a reputation of playing hardball may represent a terrific return on investment.

We do not see why DOT, expert in the air transport industry, could not make a blanket determination, based on facts that it finds, that a particular pattern of behavior is reasonably likely to facilitate recoupment of foregone revenues.



III. Problems of vagueness can be dealt with in a variety of ways.

While we sympathize with the airlines’ complaints that the DOT proposal is in some ways too vague, such that a competitor can only determine after the fact what responses to new entry are illegal, this needs to be taken in context.

First, the antitrust laws themselves typically proceed by “rule of reason” analysis such that their application is difficult to determine ex ante. This is all the more the case when the issue is price predation and the test is cost-based. Even in relatively simple cases, it is an expensive, complicated, and ultimately somewhat arbitrary task to determine not only what is an “appropriate measure” of cost, but what the cost actually is under the particular method of measurement that is adopted. Reliance on antitrust prosecution after the fact would have less predictability than the DOT proposal.

Second, there are a variety of ways in which DOT can reduce the element of uncertainty.

    1. It can construct safe harbors that eliminate elements of uncertainty. In our original comments, we proposed a safe harbor based on semi-permanent price reductions. This was criticized as “a return to regulation”, but it is no more regulatory than a cost-based approach. In our safe harbor, if a dominant carrier wants to cut its prices, it can do so—without regard to cost and without regard to the DOT tests. But it must commit to keeping the reduced prices in effect for a substantial period. In other words, the reduction cannot be temporally tied to the length of time it takes to force a new entrant out of the market. There is little uncertainty, because (like the DOT proposal) everything is based on price, which is easily knowable, rather than cost.
    2. Definitions can be tightened in response to issues raised in the comments.
    3. Examples can be provided by DOT, not only in the rules themselves, but in public speeches, staff discussion papers, etc. The types of patterns which are deemed most problematic can be further spelled out.
    4. Enforcement policies can be phased-in as the industry adjusts to a new level of scrutiny.
    5. The whole issue can be revisited on a regular basis.




We applaud the DOT for its initiative in trying to preserve competition at air hubs, and we encourage it to keep the process moving forward by taking steps to reduce the elements of uncertainty.

Respectfully submitted,

Albert A. Foer

1 Matsushita Electrical Industries Co. v Zenith Radio Corp., 475 US 574 (1986); Brooke Group v Brown & Williamson Tobacco Corp., 509 US 209 (1993).
2 509 US at , 113 S.Ct. 2578 at 2587.
3 Statement at 27 et seq., attachment A to Comments of Northwest Airlines, Inc.
4 See Comments of Frontier Airlines at 53: “An additional passenger on a scheduled flight costs the airline “peanuts”, literally and figuratively.”
5 See, e.g., Statement of Tyson, note 3 supra, at 21.
6 49 USC 4172.
7 See United Airlines v CAB, 766 F2d 1107 (7th Circuit 1985).
8 See Neil Averitt, The Meaning of Unfair Methods of Competition in Section 5 of the FTC Act, 21Boston College Law Review 223 (1980). The Supreme Court in FTC v Sperry & Hutchinson Co., 405 US 233 (1972) held that Section 5 empowers the FTC “to define and prosecute an unfair competitive practice, even though the practice does not infringe either the letter or the spirit of the antitrust laws.”
9 “[P]ost-Chicago economics refuses to rule out the possibility that successful predatory tigers may lurk in the marketplace jungle. The primary new theoretical development is the identification of a new type of predator, which cuts price in a handful of markets and creates a reputation as an aggressive competitor. This type of predator recoups the costs of predation not merely in the markets in which it engaged in a price war, but also in other markets to which its reputation has spread, by intimidating rivals in those markets to act less aggressively toward it.” Jonathan Baker, Predatory Pricing After Brooke Group: An Economic Perspective, 62 Antitrust Law Journal 585, 592 (1994).