GCW Consulting LLC

Chairman Hyde, Ranking Member Conyers and other Members of this distinguished Committee, on behalf of Northwest Airlines' 50,000 employees worldwide, thank you for inviting me to testify today.

Northwest appreciates the opportunity to discuss the state of competition in the airline industry. I would like to address three main issues today: (1) the overall financial and competitive situation in the industry today; (2) the competitive implications of the recently announced alliances; and (3) the DOT guidelines on predatory behavior.

There Is Robust Competition Today in the Airline Industry.

"[E]very review of objective evidence -- by the Department [of Transportation], the Brookings Institution, the Transportation Research Board of the National Research Council, the GAO and a host of independent studies -- has concluded that overall the network-dominated domestic system provides superior, competitive service."^⁽¹⁾ I quote from the Secretary of Transportation.

Consumers have benefited tremendously from free and open competition among airlines. Today many more people are flying and they are flying more often. Consumers board a plane bound for some destination approaching a half-billion times each year, more than twice the number in 1978 (when deregulation began). Appendix A at 2. New and expanded service continuously is being offered, and passengers have available to

them a wide range of low price options. Indeed, about 95% of all tickets sold in 1997 were discounted tickets. Appendix A at 9.

People previously not in a position to afford air transportation are now enjoying the benefits of air travel to the far reaches of this country. All of this has been achieved by freeing the airlines from economic regulation.

Contrary to what airline critics might have the public believe, average yields (that is, revenue per passenger mile) have continued to fall between 1990 and 1998 from over 15 1/2 cents per passenger mile to under 13 cents per passenger mile. Appendix A at 8. Nor have fare trends differed as between hub and non-hub markets. Appendix A at 11-14. This is true both for unrestricted fares and peak and off-peak deal fares, and over the range of flight distances. Appendix A at 15-16.

Moreover, claims that unrestricted fares have dramatically increased in recent years are untrue and misleading. Unrestricted fares in fact have grown only moderately over the past seven years, averaging 3.1% growth per year between 1990 and 1997. Appendix A at 9. Only 5.4% of the public purchases these tickets, typically businessmen and women who purchase their tickets at the last minute and require maximum flexibility in their travel plans. Id. On the other hand, it is usually overlooked that the higher prices charged to business people ensure that the seats they need at the last minute are available for purchase. If the seats they need were priced at discount levels, they would have been purchased earlier by leisure travelers willing to pay a lower fare, but unwilling to pay the higher "business" fare. The higher prices charged for last minute seats also are necessary to compensate the airline for the risk it takes by holding the seat for sale at the last minute. An airline that does this may fail to sell the seat at any price.

THE AVAILABLE EVIDENCE DOES NOT SUPPORT THE ALLEGATION THAT NETWORK AIRLINES EXTRACT A "HUB PREMIUM

Notwithstanding this remarkably pro-consumer record of achievement by the deregulated airline industry, airline critics stubbornly adhere to a dark view that would have the U.S. public believe that fares are skyrocketing out of control. In particular, it is often alleged that the major carriers have constructed fortress hubs which enable them to charge premiums of upwards of 40 percent in hub markets.^⁽²⁾ The Department of Transportation believes that such fare premiums exist, and further, that such premiums are evidence of market power of major carriers at their hubs, market power that supposedly permits the predatory conduct which it seeks to eliminate through its proposed policy statement on predatory pricing.

The Department does not identify the source of its claim that hub fares are significantly higher than fares in non-hub markets, nor does it explain why any such fare difference must be a reflection of market power. The studies of hub fares of which Northwest is aware, however, do not demonstrate that hub-market fares reflect market power.

Studies on this subject often simply compare fares for flights from certain hub airports to fares for flights from other airports. They thus ignore several factors other than alleged market power that demonstrably affect fare levels and may explain why fares by major carriers for flights to or from hub airports appear to be higher than fares from other airports.

Morrison and Winston, in their Brookings Institution published study, found several factors that need to be reflected in any comparison of hub and non-hub market fares. These factors include (1) length of flight (costs on longer flights are less on a per-mile basis than costs on shorter flights); (2) the number of stops or plane changes on the routes being compared (one-stop flights and routings requiring a change of planes are less desirable, and therefore must be priced less than non-stop flights); (3) passenger mix (flights with a higher percentage of passengers flying on unrestricted coach fares will yield more revenue per mile than flights with a higher percentage of passengers flying on discount fares); (4) identity of the carrier at issue (a carrier may charge higher fares at all airports it serves, not just hub airports, thus negating an inference that higher fares are a result of market power at a hub); and (5) the number of passengers flying free as a result of frequent flyer programs (most frequent flyer miles are earned and redeemed on flights originating or terminating in hub markets and hub fare levels may be set to cover those costs).

Their analysis demonstrated that any hub premium was approximately 5 percent in 1993 (the most recent year for which data were available), far less than the 40 percent figure thrown out without support in the Proposed Policy Statement.^⁽³⁾ The "problem" the Department envisions, therefore, virtually disappears when it is viewed in a more analytical manner.

Indeed, Northwest knows from its own experience that these factors exist in the real world, not just in academic papers.

Average Flight Length. The average trip distance for domestic passengers originating in Northwest's major hubs (Detroit and Minneapolis) is 901 and 944 statute miles, respectively. The average for domestic passengers originating and terminating in all non-hub cities is 1,005 miles, approximately 10 percent more than the Detroit figure and 6 percent more than the Minneapolis figure.^⁽⁴⁾ Because many costs (such as take-off and landing expenses) do not vary with the length of flight, shorter flights are more costly on a per-mile basis than longer flights.

Number of Stops or Connections. Approximately 90 percent of all domestic passengers originating in hub airports arrive at their destination on either a non-stop or direct (single coupon) flight. For Northwest's major hubs, the figures are 91.6 percent (Detroit) and 89.6 percent (Minneapolis). For non-hub passengers, however, the number is 82.8 percent.^⁽⁵⁾ In other words, almost twice as many passengers originating in non-hub airports (approximately 17 percent) must change planes for a domestic destination than hub passengers (about 10 percent). Any pricing difference between these different services reflects the reality that non-stop service is the most convenient form of service and that changing planes is less convenient. It is neither surprising nor unreasonable to find that the marketplace reflects this reality.^⁽⁶⁾

Passenger Mix. Many airlines, Northwest included, have located hubs in major business centers. It is not surprising, therefore, that a larger percentage of tickets purchased for travel originating at hub airports are at unrestricted fares. At Minneapolis, for example, 30 percent of Northwest's originating passengers are flying on unrestricted fares, while at Detroit, the number is 27 percent. In comparison, only 17 percent of the traffic originating at Los Angeles International is flying on unrestricted fares. The comparable number for Orlando is 11 percent. The higher unrestricted fares skew average fares upward at hub markets, and this factor must be taken into account in any comparison of hub and non-hub market fares.

Identity of Carrier. Any comparison of hub and non-hub market fares also must take into consideration the identity of the carrier at issue. Deregulation means that carriers are permitted to set their own fares, at levels reflecting their own costs and marketing strategies. Thus, one must compare hub and non-hub fares for the same carrier to determine whether hub fares are higher than non-hub fares. (Although, because of the presence of the other factors discussed in this section, such a comparison does not constitute a complete analysis of this issue.) Northwest's fares for hub and non-hub market routes are virtually identical, whether one looks at restricted fares or unrestricted fares.^⁽⁷⁾

Frequent Flyer Programs. From Northwest's experience, a very significant number of passengers on hub flights are flying free. Approximately 8.4 percent of Northwest's passengers originating in Minneapolis and 5 percent of its passengers originating in Detroit are flying free on frequent flyer programs.^⁽⁸⁾ Because these numbers are likely higher than the number of passengers originating in non-hub cities that fly free, this factor must be taken into consideration in comparing hub and non-hub fares.^⁽⁹⁾

Thus, there is no hub market premium. Moreover, even if a premium existed, it would not mean that the hub carrier has market power that permits it to engage in predatory activities. The higher fares may be simply reflecting the additional value to passengers of the hub service. An airline operating a hub generally offers a passenger more frequent flights to major cities and serves more domestic and foreign locations than its competitors. "[C]ontrolling for population, hub airports offer non-stop flights to nearly twice the number of cities as airports that are not hubs and they have more than 35 percent more daily departures per city served."^⁽¹⁰⁾ Almost all international flights operated by U.S. carriers originate or terminate in their domestic hubs. Moreover, major carriers often fly newer aircraft than other carriers, particularly in their hub markets. All of these factors are valuable to passengers and legitimately justify higher fares (whether higher fares are charged or not). None of them demonstrates either the existence or the exercise of market power.

ALL AVAILABLE MARKET STRUCTURE AND FINANCIAL DATA DEMONSTRATE THAT THE AIRLINE INDUSTRY CONTINUES TO BE VERY COMPETITIVE

All indicia of industry structure and performance confirm the existing vibrancy of competition in the airline industry and the ill-wisdom of any effort to reregulate it. The

number of competitors serving markets of almost all sizes has increased over the years of deregulation. Appendix A at 3-5. Since 1989, the share of domestic passenger miles flown by new entrants has continued to increase and in 1996 and 1997 reached nearly 18%, its all-time high.^⁽¹¹⁾

The experience of Northwest Airlines is no different. An ever increasing amount of revenue on an ever increasing number of routes are subject to competition from low fare carriers.^⁽¹²⁾ In 1997, $1.074 billion of Northwest's total revenues of $5.88 billion were exposed to such competition, an increase to over 18% in 1997 from under 3% in 1992. In addition, inter-hub competition has intensified which, in turn, has intensified the need to create the passenger flows that generate the hub network efficiencies required to remain competitive. The industry's financial performance stands as a stark reminder that competition is alive and well. Through 1997, the industry continued to suffer a cumulative net loss approaching $3 billion over a period that began in 1990. Appendix A at 66. Negative net profit margins have prevailed in five of the last eight years. Appendix A at 67. A record net profit in 1997, one of only three profitable years in the last eight, equaled less than 5%. Id. Even at that, it is worth noting that the airline industry ranked among the lowest of all industries in profitability. Appendix A at 69-71. These dreadful financial results are not the type one would expect from an industry populated by rapacious monopolists. If we are monopolists, we give new meaning to the term.

II. Alliances That Do Not Restrain Competition Deliver Enormous Consumer Benefits and Should Be Welcomed by the Traveling Public.

Airline alliances significantly benefit the traveling public. Our pioneering end-to-end alliance with KLM has become the model which all other alliances seek to emulate. In addition, our proposed end-to-end alliance with Continental Airlines also will deliver enormous efficiencies and consumer benefits without reducing competition.

An alliance can produce two or more services more cheaply than if those same services were produced separately. It allows the carriers to integrate operations with those of its partner in a way that will enable both to reduce costs, increase revenues, and use combined assets more productively.

Alliances afford millions of travelers new on-line service which they strongly prefer. The NW/KLM alliance has benefited consumers with new on-line service in more than 36,000 city-pair markets.

WHILE END-TO-END ALLIANCES PRODUCE SIGNIFICANT PUBLIC BENEFITS, HORIZONTAL ALLIANCES CAN HARM CONSUMERS

There is a tremendous difference, however, between procompetitive "end-to-end" alliances, and "horizontal" alliances, which can be decidedly anticompetitive.

In "end-to-end" airline alliances, the participating carriers each serve routes, regions, and cities that do not extensively overlap. By forming an alliance, the partner airlines essentially link together these disparate service areas, gaining access to routes and city-pairs to which the partners of the alliance would not have access individually.

In the international arena, our alliance with KLM is perhaps the purest example of a true, procompetitive, end-to-end alliance. Both the Departments of Justice and Transportation acknowledged this fact when they first approved the NW/KLM alliance some five years ago, with a grant of antitrust immunity. At the time the alliance was formed, there was virtually no overlap between the Northwest and KLM systems. In granting Northwest and KLM antitrust immunity, the DOT found that the two carriers were minor players in the transatlantic market, with respective market shares of 4.1% and 3.9% and would become, when combined, only the fifth largest U.S.-European carrier, behind Delta, BA, American and United.

The alliance has enabled Northwest and KLM to initiate new flights that previously were not possible. In July 1992, before immunization, KLM operated 58 weekly U.S.-Amsterdam flights and Northwest operated 7 weekly flights to Amsterdam (from Boston, a city that KLM did not serve), for a total of 65. Based on July 1998 schedules, KLM will operate 72 weekly U.S.-Amsterdam flights and Northwest will operate 56, for a total of 128.

This virtual doubling of flight frequencies reflects a doubling of traffic between the U.S. and Amsterdam carried on the two carriers. In 1992, Northwest and KLM combined carried 1.4 million passengers between the United States and the Netherlands. In 1996, the NW/KLM alliance carried a total of 2.5 million U.S.-Netherlands passengers, and we estimate that the alliance carried 2.9 million in 1997, more than double the number carried before the alliance was formed.

Most of this service would not have been operated without the alliance, including service in the hub-to-hub markets. The largest of these, Detroit, generates only 58 passengers per day each way, not enough to support any service much less the three daily flights currently operated.

The NW/KLM alliance succeeded by attracting passengers behind and beyond the hub -- so called "bridge" traffic. Approximately 85% of the traffic moving on the hub-to-hub routes is behind/beyond traffic.

In contrast with the procompetitive "end-to-end" alliance of Northwest and KLM is the "horizontal" alliance in which carriers overlap in many of the same markets. These alliances increase market concentration without producing offsetting consumer benefits. The Department of Justice has warned of the dangers of such horizontal alliances. American airlines' proposed alliance with TACA is an example of such an anticompetitive horizontal alliance.

EC REGULATION SHOULD APPLY, IF AT ALL, ONLY TO HORIZONTAL ALLIANCES, NOT END-TO-END ALLIANCES

American's alliance with British Airways also has certain horizontal aspects. The European Commission is trying to "fix" these horizontal problems by applying competitive rules - on airline capacity in hub-to-hub markets, frequent flyer programs, and corporate discounts, among others.

Whatever may be the merits of seeking to "fix" the AA/BA alliance, these competitive rules should not be applied to us. Our hub-to-hub markets exist only to serve the bridge traffic moving between the respective networks of Northwest and KLM. These bridge passengers have multiple choices. If our capacity is restricted on the hub-to-hub markets, the pipeline for the bridge passengers will be constricted and these passengers will lose convenience and competitive choice. The need to provide ample capacity for our 'bridge" passengers ensures that we would never restrict output on the hub-to-hub routes.

In a recent letter to Mr. Van Miert of the EC, Assistant Secretary Hunnicutt confirmed the critical role hub-to-hub routes place in fostering network competition:

U.S. approval of these alliances is based on the conclusion that the appropriate reference frame for evaluating their competitive impact is their overall effect on competition in the transatlantic market, not merely on traffic between any given city-pair. In our analysis of specific alliances we have found that while the alliance partners may reduce competition between themselves for time-sensitive point-to-point local traffic on the relatively few routes where they offer overlapping services, the alliances increase network competition. Further we have recently reviewed data in several alliance markets, and found that many business travelers use connecting services even where non-stop service is available.

Each of the large transatlantic alliances now carries passengers on a connecting basis in thousands of city-pair markets, and connecting services of two or more alliances already compete in hundreds of connecting markers. As the alliances expand their reach through hub expansion, improved coordination, and better marketing, alliance competition will increase, benefiting even more consumers. It is important that this pro-consumer aspect of alliances not be blunted by restrictions aimed at protecting a relatively small number of passengers in hub-to-hub markers.^⁽¹³⁾

END-TO-END DOMESTIC ALLIANCES SUCH AS NW/CO ALSO PRODUCE SIGNIFICANT CONSUMER BENEFITS WITH NO RISK TO COMPETITION

These same principles apply to domestic transportation and domestic alliances. They certainly apply with force and clarity to the alliance announced last January between Northwest and Continental. Under our proposed alliance with Continental, we plan to engage in a variety of joint marketing and operational activities to provide increased utility to our customers and to enhance the competitiveness of both firms. Thus, we plan to provide reciprocal frequent flyer participation; reciprocal executive lounge access; coordination of connecting flight schedules to improve and enhance

connectivity between the two networks; airport facility coordination for the same purpose; and code sharing, both domestically and internationally. The carriers will coordinate to provide for common physical product and service standards, and seamless service including one-stop check-in for seat assignment and boarding passes, baggage transfer and operations systems compatibility.

A NORTHWEST-CONTINENTAL ALLIANCE DOES NOT REDUCE COMPETITION

The NW/CO alliance will not reduce competition as there is virtually no overlap between the two carriers' networks. On the other hand, the alliance dramatically enhances consumer convenience and economic efficiency. We will be able to provide a better, more convenient product, and to more effectively compete with the three largest airlines - Delta, United and American.

The NW/CO alliance will substantially enhance the competitiveness of the domestic airline industry. Today, Northwest's and Continental's domestic networks are approximately half the size of each of the big three: Delta, American and United. An alliance between Continental and Northwest will result in a combined market share of 16.3%, which would still leave NW/CO fractionally smaller than each of Delta, United and American.

Combined, Northwest and Continental provide service domestically in approximately 18,500 city pair markets. The two carriers overlap in only 7 nonstop routes -- the hub-to-hub routes. These 7 routes constitute .03% of the combined NW/CO system -- three hundredths of a percent. There are 168 additional one stop and two stop routes where there is more than trivial overlap between Northwest and Continental, but these constitute less than one percent of the total markets served by the two. In most of these 168 markets, there is another airline in the market that either is at least the size of Northwest or Continental, or that has at least 10% of the market. These data prove that there is virtually no competitive overlap between the networks of Northwest and Continental, and, as a result, there can occur no elimination of competition between the two systems after the alliance is formed.

Moreover, the hub-to-hub routes do not present any competitive problem. These routes get far more capacity than the local market would justify to keep the pipeline open for connecting traffic. Between Detroit and Minneapolis/St. Paul, for example, Northwest offers 15 daily round trips with a total of 2,405 daily seats. This represents a 50% increase in nonstop flights from 1992 through 1996. But the local market size between Detroit and Minneapolis/St. Paul is only 296 passengers per day each way ("PDEWs"), enough to fill only 12% of the seats provided. The rest are filled with connecting passengers. (Appendix C) Between Detroit and Memphis, we provide 8 daily round trips, a 45% increase in capacity between 1992 and 1996. The local market provides enough passengers, however, to fill only 9% of the seats. Between Memphis and Minneapolis/St. Paul, the story is the same. The local market provides enough passengers to fill only 8% of the seats. (Id.) There is no reason to be concerned, therefore, that hub-to-hub flying constrains output, leading to higher prices. The opposite is true. Hub-to-hub flying leads to a substantial increase in output, well beyond what would be justified by local demand alone, and providing local passengers with far more service then they could ever expect to enjoy were it not for the hub-to-hub nature of the route. The same increase in output will occur on the hub-to-hub routes connecting Northwest's with Continental's networks.

Based on the foregoing, it is clear that the NW/CO alliance will not reduce competition in any relevant market. In fact, the alliance will produce rich and abundant efficiencies and consumer benefits.

A NORTHWEST-CONTINENTAL ALLIANCE PRODUCES LARGE AND SIGNIFICANT CONSUMER BENEFITS

In North America, by connecting the networks of the two carriers, Northwest and Continental will create some 2,000 new routes on which will travel some 3.4 million passengers. An example might bring this home more clearly. Let's consider a typical two-stop market, say Duluth, Minnesota to Lubbock, Texas. Today, there is no on-line service in this market. A passenger must make an interline connection between Northwest and either American or Delta. There is no interline connection available between Continental and Northwest, because the flights are mistimed and do not connect. After the alliance is created, however, Northwest and Continental will be able to retime their networks to serve the Duluth-Lubbock market, by flowing passengers over the Houston-Minneapolis hub-to-hub route, thereby creating an entirely new on-line opportunity for passengers between Duluth and Lubbock, and on thousands of routes like it. As a result of this new demand for the services provided by the combined networks of Northwest and Continental, we anticipate that service between Houston and Minneapolis/St. Paul in the above example, and on all hub-to-hub routes will increase. In addition to these new service opportunities, consumers will experience all of the benefits which flow from on-line service, such as coordinated check-in and baggage handling, common ticketing and reservations, and common frequent flyer programs.

In addition, by linking their networks, Northwest and Continental will provide substantially increased frequency on existing NW/CO routes: 17,500 new on-line flight opportunities on 10,500 routes conveniencing some 78.9 million passengers. On average, these 10,500 routes will enjoy a 31% increase in flight frequency. Finally, Northwest and Continental combined will provide additional routings for millions of passengers. We will in total offer some 9,300 additional routings on some 5,800 routes, conveniencing some 78.5 million total passengers. Over 250,000 passengers on these routings will enjoy shorter block times.

NORTHWEST-CONTINENTAL PRODUCES SIGNIFICANT BENEFITS IN INTERNATIONAL MARKETS

The foregoing competition analysis of NW/CO covers the domestic market only. When one examines the international markets affected by the alliance, one becomes even more enthusiastic. Northwest does not serve Central or South America at all. Continental has a significant presence in that part of the world. The alliance, therefore, will provide on-line access to Latin America for thousands of Northwest's customers, and more competition for American, the dominant force in Latin America. Similarly, Continental has virtually no service between the U.S. mainland and Asia; Northwest is a major player in that market. The alliance will provide new on-line access to Asia for Continental's customers, and enhanced domestic strength for Northwest in its challenge to remain competitive with the newly announced alliances between American and JAL, and United and ANA.

Thus, it is clear that an alliance between Northwest and Continental will produce large and impressive consumer benefits and utility, with virtually no elimination of competition in any relevant market.

RECENTLY ANNOUNCED ALLIANCES BETWEEN DELTA/UNITED AND AMERICAN/US Airways SHOULD BE CAREFULLY SCRUTINIZED BY THE JUSTICE DEPARTMENT

Last month, American and US Airways announced their own domestic alliance, as did United and Delta. These airlines have claimed that they need their own alliances to compete with Northwest/Continental, but these claims ring hollow. Today, Delta is the largest domestic carrier with 17.4%. of the market. United and American are tied for second place, each with 16.5%.^⁽¹⁴⁾ By comparison, Northwest and Continental will have only 16.3% of the domestic industry, smaller than each of the big three today. Whatever may be the motives of Delta, United and American for announcing their alliances, surely it cannot be to remain competitive with Northwest and Continental. They do not need to become twice our size to be able to compete with us.

These proposals would create entities much larger than anything our country has ever seen. United and Delta combined would have 34% of the domestic RPMs, and American and US Airways would have 25%. Each of these transactions must be judged on its own merit, just as the Northwest/Continental transaction must be judged on its unique merit. We are confident that allowing Northwest and Continental to combine forces in an alliance that gets us to the size of the big three as presently constituted is pro-competitive. Today, Delta, United and American dominate flow market traffic because of the size of their networks. An alliance between Northwest and Continental will enable us to compete effectively in these flow markets. (Appendix E.)

THE JUSTICE DEPARTMENT HAS THE NORTHWEST/CONTINENTAL TRANSACTION UNDER COMPREHENSIVE REVIEW

Mr. Chairman, I know that you and the Committee are very concerned that these airline alliances all be scrutinized closely by the appropriate antitrust authorities to insure their conformity with the antitrust laws; to ensure that they not harm consumers by increasing concentration in relevant markets. We agree. Last February, we submitted the Northwest/Continental transaction to the Justice Department under the Hart-Scott-Rodino Premerger Notification Act. We have complied fully with all Justice Department rules and requests, and specifically, we have made a full and complete response to the Department's "Second Request" for documents and responses to extensive interrogatories. Justice now has in its possession thousands of documents and analyses describing in great detail all of the competitive and efficiency implications of our proposed alliance. And although Northwest and Continental have structured their transaction to maintain absolute independence on all competitively sensitive business decisions, such as pricing, capacity allocation and market selection, we have nevertheless invited the Justice Department to apply the toughest standard possible to the transaction, and to assume that for purposes of competitive analysis that we will have fully merged when we consummate the transaction.

Based on the rich evidence of efficiencies, and the virtual absence of competitive overlap, we are confident we can pass this high hurdle. We are less sure that the other alliances can pass such a hurdle, but we are sure of this: Delta/United and

American/US Airways, no less than Northwest/Continental, should be subjected to the most searching scrutiny by the Justice Department to ensure that they do not sacrifice the interests of consumers in a competitive airline industry. And if the Big Three choose to structure their transactions in a way that avoids the reach of the Hart-Scott-Rodino Premerger Notification Act, then the Justice Department should use its ample enforcement powers and issue CIDs to demand the same kinds of materials that we have supplied to the Department under Hart-Scott-Rodino. It would be perverse, indeed, for the Justice Department to devote extensive enforcement resources to investigate the competitive implications of Northwest/Continental and then turn around and give a free pass to transactions that are twice our size.

III. The DOT's Proposed Rules Against Predatory Pricing Are Dangerous and Should Not Be Implemented.

On April 6th of this year, DOT issued a proposed statement of enforcement policy,^⁽¹⁵⁾ which will reserve important segments of the airline market to a new class of federally favored airlines. DOT proposes to accomplish this through the regulation of the central pricing and capacity decisions of all major airlines on literally hundreds of the most important routes emanating from their hub markets. Under the new regulatory scheme, DOT would threaten enforcement action under Section 411 of the Federal Aviation Act, 49 U.S.C. § 41712, against any major airline that competes too hard against certain other airlines. To qualify for this federal protection, an airline must:

Pursue a strategy of "low fares"

Be less than ten years old.

DOT'S PREDATORY PRICING GUIDELINES ARE A SOLUTION IN SEARCH OF A PROBLEM

DOT has proposed this drastic departure from competition policy to remedy a "problem" it has failed to demonstrate exists. DOT claims that consumers are being hurt by a deliberate industry-wide practice on the part of the established carriers to implement "drastic" price cuts and to "flood" the market with low fare capacity in order to drive new entrant carriers out of the market and to deter others from entering. The large carrier supposedly recoups its losses by raising fares to very high levels after the new entrant is gone.

The objective evidence does not demonstrate that there is an industry-wide problem. The practice of reducing price and adding capacity in response to low fare competition that is assumed to be permanent is often the profit maximizing response for a network airline.

On occasion, Northwest has responded to low fare entry with price cuts and capacity increases. On other occasions, we have responded very differently. Between 1993 and 1998, there were 32 occasions on which low fare carriers entered a Northwest route. Sixteen ultimately exited, and sixteen remain today. We matched the new entrant's fare on all 32 occasions. Of the sixteen that exited, we added capacity in ten. Of the sixteen that remain, we added capacity in ten. In every instance, our commercial response was dictated exclusively by what would maximize revenues and profits for Northwest Airlines on a network basis. On no occasion did we assume as part of our profit maximizing analysis the exiting of the new entrant. And as you can see from the record, the outcome of our competitive responses to new entrants has been totally unpredictable. Half the time, they exit; half the time, they keep right on competing. And there is no difference in the outcome arising from whether we added capacity or not.

The DOT seems to believe the industry treats Southwest differently. Professor Morrison, however, detects no such difference:

Airlines may believe that they can successfully repel the entry of most post-deregulation new entrants. However, it does not seem reasonable that they would believe this applies to all new entrants, especially the premier "new entrant," Southwest Airlines. The anecdotal evidence presented below suggests that the strategy of matching a new entrant's price is followed in both cases. Sometimes the result is the new entrant leaves the route, sometimes the carriers continue to compete for several years, and in other cases the incumbent firm abandons the route to the new entrant.^⁽¹⁶⁾

Nor is it true that Northwest treats Southwest differently. When Southwest Airlines entered the Detroit-Nashville market in 1997, for example, we matched their price and added one flight because it was the profit maximizing thing to do for our network. In excess of 90% of our seats are offered at the Southwest levels. 68% are sold at Southwest's lowest levels. This is further evidence that such a strategy is not, by itself, evidence of predatory intent. No airline manager in his or her right mind expects to drive Southwest out of a market.

DOT'S PROPOSED RULE IS A DRASTIC, DANGEROUS AND MISGUIDED DEPARTURE FROM EXISTING US COMPETITION LAW AND POLICY

DOT's proposed action should not be viewed as a mere tinkering or fine tuning of the competitive marketplace. Rather, it is a fundamental departure from the core rules of competition. Its impact will be dramatic. The major airlines will be hobbled in their ability to compete on the merits for those passengers upon whom they are highly dependent to maintain the volume, efficiency, and sustainable revenues necessary to maintain fully effective hub and spoke systems. Reservation of passengers to the new protected class of airlines inevitably will lead to a reduction by the major airlines in the quantity of service they can provide, a diminished ability viably to offer service on the least dense routes, and to higher prices throughout their hub systems.

DOT will take into account neither the comparative efficiency of the major airline and its smaller rival, the comparative quality of their service, nor the cost-based justification of the major airline's competitive initiatives in performing its analysis. Instead, DOT will disregard its regulatory mandate, along with well established principles of antitrust law and policy, and simply assume a predatory purpose and effect of actions by a major airline that are plainly motivated to benefit consumers and thereby preserve the major airline's competitive position in the market. To protect smaller airlines from competition and reserve to them a passenger base that will aid in their survival, DOT proposes to prohibit major airlines from:

increasing capacity and reducing fares, if to do so would result in the major airline providing low fares to passengers that otherwise would have been willing to pay higher fares; or

reducing fares and carrying a number of local passengers that exceeds either the seat capacity or the number of low fare passengers carried by the smaller airline, if to do so would result in the major airline providing low fares to passengers that otherwise would have been willing to pay higher fares.

Adverse DOT action may be avoided, but to do so a major airline must be able to demonstrate that there was no "reasonable alternative response" that could have resulted in lowering fares to fewer passengers that would have been willing pay higher fares.

The extent of the intrusiveness of this pervasive regulatory scheme, not to mention the costs and uncertainty it would impose on the industry, hardly can be overstated. The array of pricing and capacity decisions normally left to firms in an open and unregulated market will be severely circumscribed. To be sure, DOT is aware of the implications of its proposed new scheme. DOT has advised that a major carrier can minimize DOT oversight and intervention by:

". . . matching the new entrants' low fares on a restricted basis (and without significantly increasing capacity) and relying on its own service advantage to retain high-fare traffic."^⁽¹⁷⁾

In short, a major carrier must leave to the new entrant any passengers that it cannot accommodate on its existing flights, and it may accommodate additional passengers on its existing flights only to the extent that passengers who are willing to pay higher fares are not offered lower fares. Thus, for example, were a fare decrease to result in a small carrier filling 120 seats, the major carrier ordinarily would be limited to selling no more than 120 seats at prices lower than its previous fares. But even 120 may be too many low fare network carrier seats for the DOT regulators if they conclude, based on their own post hoc analysis (after the predictable complaint from the new entrant) that the major airline could have charged some of the 120 passengers higher fares by relying "on its own service advantage."

What's worse, the major carrier has no way of knowing how many seats the low fare new entrant is offering at any particular fare, nor does the major carrier have any way of knowing how many passengers it actually carries at any particular fare. The major carrier must guess at both of these figures at its peril, for if it guesses wrong, it will be found, after the fact, to have violated the DOT rules.

To the extent that a major airline concludes that on some routes it may make sense to develop shuttle-type low price service through a low cost "airline within an airline" operation, that operation will be hobbled by the same preference rules as apply to its parent's hub operation because it would not qualify as an "independent" airline. Accordingly, the major airlines' incentive to invest in such cost reduction strategies will be diminished.

In light of the legion of uncertainties in the rule (What is a "low fare"? What is a "very low fare"? What is a "low fare strategy"? What is a "reasonable alternative strategy"? How many seats can be offered? How many passengers can be carried? What is "substantially below the major's previous fare"? ), the only safe strategy for the major is to do nothing and cede the market to the new entrant. If this is what the majors are obligated to do, the DOT will have unwittingly put in place a policy that dismantles the hub and spoke network system, for without access to local traffic, service on many spokes will have to be reduced or eliminated altogether.

This proposed new regulatory scheme of accommodation in place of competition is borne of basic misconceptions and faulty policy choices, a few of which I hope I might briefly address today. Implicit in the proposed new regulatory scheme is a fundamental misunderstanding of the economics of the hub systems that are common to the major carriers. Smaller new entrants are fast to attribute their failures to predation rather than to the superiority of their competitor's service offerings or the integrative efficiencies that allow (if not mandate) an aggressive competitive response. It thus is particularly problematic that the DOT proposes to scrutinize the pricing and capacity decisions of major airlines without taking into account the revenue effects of their decisions on their overall hub systems, or whether the major airlines are offering service below some appropriate measure of cost. These conceptual oversights are all the more troubling when applied to an industry that operates on razor thin margins in much of its network and overall hardly can be characterized as producing monopolistic returns.

Over the coming weeks and months, we at Northwest Airlines hope to expand and elaborate on these points. Our sincere hope is that after a thorough airing of the issues, we will retain the paradigm of airline deregulation: open competition policed by reasonable and well-established antitrust principles. Critics of airline competition point to the failure of the Justice Department to file a single case of predatory pricing as evidence that the laws against predatory pricing have failed. Justice has had hub airline practices under continual scrutiny for the past several years. The fact that Justice has yet to file a predatory pricing case ought not to be judged as a failure of antitrust enforcement, but rather as evidence that no antitrust violation has yet been found. Assistant Attorney General Klein has announced that the Department will complete its current investigations later this fall. If Justice finds no violation, that should be the end of the matter -- for DOT no less than Justice.

THE HUB SYSTEMS OF THE MAJOR CARRIERS ARE EXTRAORDINARILY EFFICIENT

The airline industry discovered in the early years of deregulation that tremendous efficiencies flow from the establishment of a network. Through a hub system, airlines are able to offer passengers from many origins one-stop and two-stop flights to many destinations. The result is a more efficient use of facilities and, as the network grows, the increasing provision of service to otherwise under-served, thinly traveled destinations.

Northwest Airlines has invested heavily in its hub system, dramatically growing its hubs. At Minneapolis/St. Paul, since 1990, we have increased domestic seats offered by 27%, and international seats offered by 260%. Northwest now serves 134 nonstop domestic and international destinations with 487 daily departures.

At Detroit Metro, Northwest has, since 1990, increased domestic seats offered by 42%, and international seats offered by 84%. We now serve 117 nonstop domestic and international destinations with 512 daily departures.

It is quite important to understand that hub and spoke systems create significant demand and cost inter-relationships among city-pair routes. This network system creates economies of scope by combining passengers traveling nonstop to a particular destination with those that originated at, or are traveling to, a different destination. Economies of scope, in turn, allow hub carriers both to increase the frequency of service and to achieve economies of scale associated with the use of larger aircraft.

In light of these economic realities, it is not surprising that, in addition to increases in the frequency of service on its existing hub routes, Northwest continues to add new destinations to its system. Since 1992, we initiated 38 new destinations from

Minneapolis/St. Paul and 26 new destinations from Detroit Metro. Each new spoke served from the hub results, through connection at the hub, in service to hundreds of markets.

At least two very important criticisms of DOT's proposed new regulatory scheme arise from a proper understanding of the hub system. First, the failure of many new carriers in competing with hub carriers should be of no surprise. The major carriers have become increasingly efficient and the airline industry is vibrantly competitive (a point to which I will later return). It is less than surprising that many passengers decide that, fares being equal, the hub airline with its better schedules, frequent flier programs, and in-flight services offer a better value than the new entrant carriers. It is equally understandable that a failing new carrier will complain and seek from the regulators shelter from the forces of competition. But, it is perverse in the extreme for the government to grant those pleas to preempt competitive market outcomes by promulgating special rules to protect infant and frequently poorly managed airlines from the vigorous and lawful competition promoted by the antitrust laws.

Second, and equally important, evaluation of the competitive response of a hub carrier to its smaller rivals must take into consideration the dynamic interrelationship of demand and costs in a hub system. The loss of passengers by a hub carrier to a competitor affects in complex and interactive ways not only the economics of a single flight on a single city-pair, but the entire interrelated hub system. The benefit of winning back from a new entrant a lost customer can generate revenue benefits beyond those attributable to a particular flight on a particular route.

Conversely, restricting a major hub airline from competing effectively (in order to protect a favored class of smaller airlines) produces the untenable outcome of preventing the hub airline from taking competitive action necessary to achieve system-wide efficiencies. At the same time, the restriction preempts market forces in testing whether the protected carrier is more efficient and offers service more desirable to consumers. Even worse, the protected carriers will game the regulatory system to deprive consumers to the maximum extent possible of the competitive benefits they otherwise would have derived from free and open competition with the major airlines. For example, it will calibrate its fares to a level no lower than is necessary to create demand equal to its capacity plus the equal number of passengers that the major airline is permitted by DOT to carry at the lower fares.

Given the importance to major airlines and the consuming public that the forces of competition be allowed to choose winners and losers,^⁽¹⁸⁾ it is particularly troubling that, under DOT's proposed new regulatory scheme, a major airline's pricing and capacity decisions will be scrutinized and sanctioned without regard for whether they truly are predatory under well-established revenue/cost standards.

DOT PROPOSES TO ESCHEW ESTABLISHED PRICE/COST TESTS FOR PREDATION

Mr. Chairman, the wisdom of the DOT's proposed policy need not be evaluated in a vacuum. We are informed by the Airline Deregulation Act of 1978, the legislative history of the Federal Aviation Act, and by some 80 years of judicial precedents involving real-world allegations of predatory pricing, the result of which is a well developed analytical framework designed to protect the interests of the consuming public.^⁽¹⁹⁾

In its Policy Statement, the DOT states that under Section 411 of the Federal Aviation Act, it may "stop carriers from engaging in conduct that can be characterized as anticompetitive under antitrust principles even if it does not amount to a violation of the antitrust laws." Policy Statement at 7. For the DOT to move beyond well-defined and considered rules against predatory pricing as established by the courts under the antitrust laws would be both bad public policy, and unlawful.

IMPORTANT COMPETITIVE AND CONSUMER INTERESTS WILL BE HARMED BY THE DOT'S PROPOSED RULE

Predatory pricing like other types of anticompetitive conduct has long been the province of the federal antitrust laws -- in this case, the Sherman Act, aptly described by the United States Supreme Court as "a comprehensive charter of economic liberty aimed at preserving free and unfettered competition as the rule of trade."^⁽²⁰⁾ Those laws rest on

the fundamental premise that "unrestrained interaction of competitive forces will yield the best allocation of our economic resources, the lowest prices, the highest quality and the greatest material progress," while at the same time providing nothing less than "an environment conducive to the preservation of our democratic political and social institutions."^⁽²¹⁾

Pricing is predatory and unlawful under the antitrust laws when used willfully to acquire and maintain monopoly power.^⁽²²⁾ There is nothing wrong with competing aggressively, even if in the hope of becoming a dominant firm. As Judge Learned Hand explained long ago: "The successful competitor, having been urged to compete, must not be turned upon when he wins."^⁽²³⁾ It is the acquisition or maintenance of monopoly power by unreasonable means that concerns the antitrust laws. And one of those unreasonable means that has long been addressed by the antitrust laws is the act of a monopolist to sustain short-term losses in a pricing war of attrition against a competitor, only to reap long-term monopoly profits once its would-be rival is driven from the market.

But competitors often complain that their rival's low prices are "predatory," when in fact they reflect nothing more than vigorous competition by an efficient firm offering a product or service more desired by consumers. For that reason, the courts have crafted a specific set of rules when a claim of monopolization rests on alleged predatory pricing. In its recent treatment of the issue, the United States Supreme Court established a

requirement of proof of two critical factors, both of which are glaringly absent from the DOT's proposed policy:

First is that the challenged pricing be below an appropriate measure of cost.^⁽²⁴⁾ Although courts may debate what measure of costs is most appropriate (e.g., whether they be average variable cost or marginal cost), no court will condemn pricing unless it is below a defined measure of a firm's cost.^⁽²⁵⁾

Second, the antitrust laws require a dangerous probability /that the monopolist will "recoup" its investment in below-cost prices.^⁽²⁶⁾ This requirement is founded on the common sense principle that no firm will knowingly set a price below cost if there is no rational prospect of recovering the lost revenues once a rival has been successfully driven from the market.

DOT's proposed enforcement policy contains neither of these elements. Nowhere in the policy is there a requirement that a carrier's prices be below cost to be condemned. To the contrary, the DOT is quite clear in its intention to condemn an airline that lowers its fares in response to competition even with profitable, above-cost fares. Its proposed policy is to initiate enforcement proceedings against a carrier that lowers its fares "substantially below the major carrier's previous fares" or simply to the level of the "new entrant's low fares," so long as the pricing results "in lower local revenue than would a reasonable alternative response." In short, a major carrier that responds to competition from a smaller rival with lower prices acts at its peril if DOT decides in its discretion that the carrier made an "inordinate sacrifice in local revenues," regardless whether the new, lower fares are profitable.

And because a carrier need not price below cost to run afoul of DOT's regulatory scheme, there is no need to prove, as the antitrust laws require, that the carrier can recoup its losses. Indeed, there is not even a requirement that the carrier have or threaten to acquire monopoly power in any defined market, as the antitrust laws require.

It is not surprising that, against this backdrop, the DOT contends that the activity it seeks to condemn "is analogous to (and may amount to) predation within the meaning of the federal antitrust laws." Policy Statement at 7. Nor is it surprising that the DOT seeks to cast the price and capacity decisions of the major airlines with which it is concerned as "unfair exclusionary behavior." No amount of wordsmithing, however, can alter the fact that unilateral price and output decisions by major airlines either are predatory or they are not, and that well-established standards under the federal antitrust laws have been developed to make that assessment. DOT's proposed policy represents a radical, unprecedented departure from those federal antitrust laws that inevitably will undermine consumer welfare.

DOT announced that it intends to work closely with the Department of Justice in implementing its new regulatory scheme. However, the Justice Department has long made clear its view that "fares themselves cannot be used to establish predatory pricing unless they are shown to be 'below cost.'"^⁽²⁷⁾ "[W]here the price is not below cost, no further inquiry is required." ^⁽²⁸⁾

The carefully tailored standards of the antitrust laws were created by design; they are central to the very purpose of the antitrust laws and our marketplace economy. The goal of the antitrust laws, after all, is to foster vigorous competition among rivals, especially price competition. As the Supreme Court put it: "Cutting prices in order to increase business often is the very essence of competition."^⁽²⁹⁾ Condemning a firm for competing too aggressively, with prices that are too low, should be done only with extreme hesitation, lest the consuming public lose the very competitive benefits that we are here today to promote. Mistaken condemnations under the antitrust laws, as the Supreme Court has explained, "are especially costly, because they chill the very conduct the antitrust laws are designed to protect."^⁽³⁰⁾

Until promulgation of these guidelines, the DOT faithfully had followed the law, and the airlines were free to compete aggressively for the benefit of consumers. Thus, in 1982, the CAB explicitly held in Capitol Air, Inc., Order 82-7-107 at 5 that it would "not consider marginal cost pricing to be anticompetitive or otherwise illegal under Section 411 of the Act." It recognized that "[i]t is a generally accepted economic principle that

firms in competitive markets price at marginal costs. Such pricing is desirable because it maximizes consumer welfare and results in the optimum allocation of resource." The Board cited Alfred Kahn, The Economics of Regulation, Vol. 1 p. 67, to support this basic economic principle.

DOT's proposed enforcement policy poses a particularly severe threat to competition in the airline industry. By replacing the antitrust law's cost-based standard for predation with an undefined standard of "reasonableness," airlines no longer will be able to judge in advance the legality of their pricing decisions. Under the proposed policy, price cuts will be made at a major airline's peril, and will be outright prohibited if too many consumers benefit from them. Senior citizen discounts will be at risk; internet "cybersaver" fares will be at risk; bereavement fares will be at risk; supermarket discount coupon fares will be at risk; free or deeply discounted seats for infants under two years of age will be at risk. Indeed, all price discounting will be at risk.

At risk, also, will be fleet planning. Today, a major airline will review its plans ten to twenty years out, and order large number of new aircraft to fund anticipated demand on specific city pair routes. Under the DOT's rules, the airline will be precluded from installing purchased equipment on those of its planned routes where the increased capacity at the prevailing fare level would violate the DOT's guidelines.

The result only can be to chill competition on all sides; to protect competitors at the expense of competition; to put the interests of new rivals above the financial interests of the flying public; and to deprive consumers of the very benefits that new competition is supposed to bring.

THE DOT'S PROPOSED GUIDELINES ARE UNLAWFUL

The DOT is wrong in its belief that the Airline Deregulation Act authorizes it to substitute for the definition of predation contained in the antitrust laws its own sense of what types of competitive behaviors are "reasonable". The Airline Deregulation Act of 1978 expressly specifies that the DOT is not to find unlawful a low fare unless it would be the type of activity that a district court would find unlawful under the antitrust laws. In the definitions section of the Airline Deregulation Act, Congress specified that "'predatory' means a practice that violates the antitrust laws. . ." 49 U.S.C. § 40102(a)(34). This interpretation is consistently applied throughout the legislative history and the CAB and DOT decisions in the area.

The Congress instructed the DOT not to utilize complaints of predatory pricing by poorly managed firms as a pretext to reintroduce price and capacity controls:

It is the express intention of the committee that the [DOT] not utilize its power to use the rubric of 'predatory' to find lower fares unlawful unless such low fares are truly predatory. . . . Thus, the committee would not expect the [DOT] to strike down a low-fare level which represents genuine competition simply because it would tend to decrease the revenues of less efficient carriers in the market or perhaps force from a given market carriers who were not able to provide the price and service mix which the passengers in that market desired. It is for this reason that the bill amends section 101 of the Aviation Act to define 'predatory' as behavior which would constitute a violation of the antitrust laws. In other words, the [DOT] is not to find unlawful a low fare unless it would be the type of activity which . . . a district court would find unlawful under the antitrust laws.

S. Rep. No. 631, 9th Cong., 2d Sess. 107-108 (1978) (emphasis added). See also Air Florida v. Eastern Air Lines, Inc., Order 81-1-101 (Dkt. No. 37313) (Jan. 21, 1981), pp. 3-4, n.4 (acknowledging that "Congress did not intend us to hold fare reductions unfair which do not violate the antitrust laws.").

Thanks to deregulation, competition is alive and well in the airline industry. Under the Airline Deregulation Act of 1978, prices are down, output is up and travel is up. Market forces have done for American consumers precisely what the congressional framers of deregulation intended.

Consolidations that expand network coverage without eliminating competition in overlapping markets -- like Northwest/Continental -- should be favored because they provide consumers with greater value and more competition. Similarly, aggressive price competition among airlines delivers massive consumer benefits and must be promoted on all occasions. Vigilant antitrust enforcement has a central and critical role to play in ensuring that only consolidations that do not harm competition, and that promote consumer choice are permitted. Similarly, vigilant antitrust enforcement is necessary to ensure the preservation of vigorous price competition. But it is equally important that we not stray from sound antitrust principles, established over the years by the Congress and the courts, in assessing airline alliances and mergers, and airline pricing. It is critical that we adhere to the rigors of the antitrust laws, because substituting vague bureaucratic notions of "reasonableness" for vigorous application of hard antitrust analysis inevitably will harm commoners and the economy.

2. ⁰ 63 Fed. Reg. at 17,920.

3. ⁰ Morrison & Winston, The Evolution Of The Airline Industry at 46-49.

4. ⁰ Appendix A at 42.

5. ⁰ Appendix A at 43.

6. ⁰ Indeed, the Department has previously recognized that "the superior service offered passengers traveling to and from network hubs is a counter-balance to the higher fares." The Low Cost Airline Service Revolution at 26.

7. ⁰ Appendix A at 15-16, 40.

8. ⁰ Appendix A at 23-24, 32.

9. ⁰ Morrison & Winston, The Evolution Of The Airline Industry at 47.

10. ⁰ "New Entrants, Dominated Hubs, and Predatory Behavior," Statement of Professor Steven A. Morrison, Hearing before the Subcommittee on Aviation, Committee on Commerce, Science and Transportation, United States Senate, April 23, 1998, at 4.

11. ⁰ Statement of Professor Morrison, at 1.

12. ⁰ Appendix B shows for each of the years 1992 through 1997 local and connecting revenues of Northwest Airlines that are exposed to competition from airlines pursuing a low fare strategy.

13. ⁰ Letter from Charles A. Hunnicutt to Karel Van Miert, dated April 17, 1998.

14. ⁰ See Appendix D.

15. ⁰ Docket No. OST-98-3731.

16. ⁰ Statement of Professor Steven A Morrison before the Subcommittee on Antitrust, Business Rights, and Competition, Committee on the Judiciary, U.S. Senate, April 1, 1998 at 4.

One exceptional source of benefits [of airline deregulation] is provided by competition exerted by Southwest Airlines and other new entrants. . . . [A]ny policy to combat cases of alleged predatory behavior needs to develop a basis for distinguishing a normal competitive response from a predatory one -- a difficult task in a network industry like the airlines where route-level costs are difficult to define and measure.

Testimony of Steven A. Morrison, Hearing before the Subcommittee on Antitrust, Business Rights, and Competition, Committee on the Judiciary, United States Senate, at 5, April 1, 1998.

19. ⁰ As early as 1911, the Supreme Court identified predatory pricing as an offense within the proscription of monopolization in Section Two of the Sherman Act. United States v. American Tobacco Co., 221 U.S. 106, 182 (1911); Standard Oil Co. v. United States, 212 U.S. 1, 43 (1911).

20. ⁰ Northern P. R. Co. v. United States, 366 U.S. 1, 2 L.Ed.2d 546, 78 S.Ct. 514 (1958).

23. ⁰ United States v. Aluminum Co. of America, 148 F.2d 416, 430 (2d Cir. 1945)

24. ⁰ Brooke Group, Ltd, v. Brown & Williamson Tobacco Corp., 113 S. Ct. 2587 (1993) (the plaintiff must demonstrate that the "prices complained of are below an appropriate measure of its rival's costs").

25. ⁰ It would be ironic indeed if the antitrust laws could be used to protect a less efficient firm against aggressive price-cutting by a firm pricing where marginal cost equals price, the level that results "in the optimum allocation of resources." Alfred Kahn, The Economics of Regulation, Vol. 1, p. 67.

27. ⁰ Business Review Request, British Airways PLC, issued by the Antitrust Division of the Department of Justice (Dec. 20, 1984), p.2. (A copy of this Business Review Letter is attached as Appendix F.)

28. ⁰ Id. at 2, n.1. See also Remarks by Roger W. Fones, Chief of the Antitrust Division's Transportation, Energy, and Agriculture Section, Before the American Bar Association Forum on Air and Space Law (June 12, 1997), explaining that the basic principles of an antitrust analysis of alleged predatory pricing in the airline industry require that:

"The incumbent's prices must be 'below an appropriate measure' of its own costs.

Corollary A: An appropriate measure of costs should not establish a price umbrella for inefficient firms.

Corollary B: An appropriate measure of costs should minimize the risks of condemning legitimate competitive behavior.

Corollary C: An appropriate measure of costs should be reasonably measurable with a high degree of confidence and predictability."

29. ⁰ Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 594 (1986).