This document is available in two formats: this web page (for browsing content) and PDF (comparable to original document formatting). To view the PDF you will need Acrobat Reader, which may be downloaded from the Adobe site. For an official signed copy, please contact the Antitrust Documents Group.









Civil Action No. 99-1180-JTM


I. Introduction

American has illegally monopolized and attempted to monopolize airline passenger service on many routes to and from Dallas/Ft. Worth ("DFW"). Section Two of the Sherman Act makes it unlawful for a firm to "monopolize, or attempt to monopolize. . . any part of the trade or commerce among the several States. . . ." 15 U.S.C. § 2. The Sherman Act was enacted to prevent "restraints to free competition in business and commercial transactions which tend[] to restrict production, raise prices or otherwise control the market to the detriment of purchasers or consumers of goods and services." Apex Hosiery Co. v. Leader, 310 U.S. 469, 493 (1940).

The elements of a Section Two monopolization claim are: (1) the possession of monopoly power in a relevant market; and (2) the willful acquisition or maintenance of that power as distinguished from growth or development as a consequence of a superior product, business acumen, or historic accident. The elements of a Section Two attempted monopolization claim are: (1) definition of the relevant market; (2) a dangerous probability of success in monopolizing that market; (3) the specific intent to monopolize; and (4) conduct in furtherance of that attempt. Full Draw Prods. v. Easton Sports, Inc., 182 F.3d 745, 756 (10th Cir. 1999).

Under both claims, the Court must analyze the conduct of the defendant to determine if it is monopolistic, i.e., if it is a "scheme of willful acquisition or maintenance of monopoly power." Eastman Kodak Co. v. Image Technical Servs., 504 U.S. 451, 483 (1992). The Supreme Court has used the terms "exclusionary" or "anticompetitive" or "predatory" to label such willful, unlawful conduct and to distinguish it from competition on the merits. Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585, 602 (1985). The key inquiry in applying the label is whether the conduct constitutes an "abnormal response to market opportunities," Multistate Legal Studies, Inc. v. Harcourt Brace Jovanovich Legal & Professional Publications, Inc., 63 F.3d 1540, 1550 (10th Cir. 1995), cert. denied, 516 U.S. 1044 (1996) (quoting Instructional Sys. Dev. Corp. v. Aetna Casualty & Surety Co., 817 F.2d 639, 649 (10th Cir. 1987)), which entails in part an examination of whether the defendant's actions are ultimately inexplicable except on the basis of the monopoly returns expected as a result of the action's creation or maintenance of a monopoly. See generally Aspen Skiing, 472 U.S. at 605-11; Eastman Kodak, 504 U.S. at 482-85.

In conducting the inquiry, a court must examine the totality of the defendant's conduct in determining that the defendant's overall scheme was monopolistic. Continental Ore Co. v. Union Carbide & Carbon Corp., 370 U.S. 690, 699 (1962); Instructional Sys., 817 F.2d at 649; City of Mishawaka v. Amer. Elec. Power Co., 616 F.2d 976, 986 (7th Cir. 1980), cert. denied, 449 U.S. 1096 (1981).

II. Summary of American's Monopolistic Scheme: The DFW LCC Strategy

American is the second largest airline in the United States, offering service throughout the country. American operates its largest and most profitable hub at DFW, the third largest airport in the United States. American is by far the dominant carrier at DFW, offering over 700 flights daily to more than 100 destinations. In 1998, American's service to and from DFW accounted for nearly $2 billion in annual revenues.

Between 1994 and 1998, American developed and implemented a monopolistic scheme designed specifically to protect those revenues. This scheme, called the "DFW LCC Strategy" by American, focused on certain low cost carriers ("LCCs"),(1) whose entry into some routes to and from DFW American perceived as threatening its monopoly power. American's implementation of its DFW LCC Strategy monopolized or attempted to monopolize DFW routes.

As early as 1994, American was concerned about the threat that LCCs posed to its dominance in routes to and from DFW, some of which had already been entered by LCCs. In particular, American was concerned that an LCC would offer low fare service in multiple DFW city pair markets.(2) Over the ensuing several years, American developed a strategy of capacity, fare and yield management actions designed to suppress LCC competition in DFW routes.(3)

In February 1996, American's senior management reviewed and endorsed what had become known as the DFW LCC Strategy. With the goal of depriving LCC competitors of sufficient passengers to be viable (and so force them to exit), American determined to take whatever actions it deemed necessary, without regard to profitability. In carrying out the DFW LCC Strategy, American disregarded its usual measures of performance, i.e., revenue and profitability, focusing instead on whether its actions were decreasing the LCC's market share and load factor.(4) American investigated the financial resources of LCCs, determined their break-even load factors, and conducted head counts at LCC departure gates to monitor their passenger loads in order to evaluate the success of its actions and to determine whether to intensify its response to drive the LCC from the market.(5) American applied the most aggressive level of the DFW LCC Strategy's capacity, fare and yield management actions against the following LCCs: (1) Vanguard Airlines, from as early as October 1996 through at least October 1997; (2) Sun Jet Airlines, from as early as January 1997 through January 1998; and (3) Western Pacific Airlines, from as early as September 1996 through January 1998. To minimize the likelihood that passengers would fly on the LCC, American saturated the LCC markets with additional capacity, lowered its prices, and adopted an "open" yield management strategy, contrary to its normal policy of holding back seats for last-minute, high fare customers. American also took further steps, such as matching the fares in other city pairs served by these LCCs on a nonstop and connect basis, to keep traffic from the targeted LCCs.(6)

When American applied these most aggressive tactics, its actions were not profitable. Indeed, the costs American incurred by adding capacity were higher than the revenues that American received from carrying additional passengers at lower fares. American was willing to pursue these tactics, however, because it could recoup the earnings sacrificed by raising its prices and reducing service back to monopolistic levels after driving the LCC out of its markets. In addition, American could stifle expansion by these LCCs in DFW routes. Moreover, by making an example of an LCC victim, American could deter other LCC expansion and entry into DFW routes through a reputation for aggressively sacrificing earnings in response to LCC entry. American's reputation would help preserve its monopoly power in DFW city pairs over time.

American's actions on the Wichita-DFW route are illustrative. In February 1994, American announced that it would terminate its jet service between Wichita and DFW and serve the route only with turboprops -- explaining that it was making the change because it had been "losing money" on the route. Dave Higdon, Wichita Air Service Called Healthier, Wichita Eagle, Feb. 12, 1994, at B7; see Dave Higdon, Seeking Jets for Wichita; Officials May Put Up Money to Replace Turboprop Service, Wichita Eagle, Feb. 6, 1995, at 1D. American then told the Wichita Airport Authority that it would maintain three jet flights only if the Authority provided a subsidy guarantee to American.

In the fall of 1996, Vanguard, which had introduced jet service between Wichita and DFW in April of 1995, announced an expansion of its service at DFW. American quickly responded by beginning five daily jet flights (expanding its seating capacity by 35%) between Wichita and DFW -- two more jet flights than it had been willing to put in even with the subsidy. American's costs of adding that capacity exceeded its revenues. In December 1996, Vanguard announced that it was leaving Wichita. Shortly thereafter, American decreased seating capacity by 30% and its average fares increased from approximately $60 to over $90, an increase of more than 50%.

III. American Has Monopoly Power in Relevant Markets

Under both monopolization and attempted monopolization, the United States must prove that (1) a relevant market or markets exist; and (2) that American had market power or monopoly power in the relevant market or markets.(7) See Bright v. Moss Ambulance Serv., 824 F.2d 819, 823 (10th Cir. 1987).

There are two types of DFW-related markets that American has monopolized or attempted to monopolize: city pair markets and nonstop city pair markets.

A. DFW City Pairs and Nonstop City Pairs Are Relevant Markets

Airline passenger service in a city pair and nonstop airline passenger service in a city pair constitute the relevant markets for this case. The particular city pairs at issue are over fifty city pairs emanating from DFW.

A relevant market is the "line or part of commerce" that the defendant has monopolized or attempted to monopolize. 2 Julian O. Von Kalinowski et al., Antitrust Laws and Trade Regulation § 24.01[1] at 24-2 (1999) (hereinafter, "Von Kalinowski"). Definition of a relevant market is a tool to gauge what competition is significant and constrains a firm's behavior. Thus, the relevant market is determined "after a factual inquiry into the 'commercial realities' faced by consumers" that includes an assessment of product function, price, and quality, to determine what products are, in the eyes of consumers, reasonably interchangeable with the products sold by the defendant. Eastman Kodak, 504 U.S. at 482; United States v. E.I. du Pont de Nemours & Co., 351 U.S. 377, 404 (1956); SCFC ILC, Inc. v. Visa USA, Inc., 36 F.3d 958, 966 (10th Cir. 1994), cert. denied sub nom, MountainWest Fin. Corp. v. Visa USA, Inc., 515 U.S. 1152 (1995). Reasonable interchangeability is important because the availability of substitute products from other firms could restrain a firm from raising its prices above the competitive level ("supracompetitive"). SCFC ILC, 36 F.3d at 966. Thus, for products to be in the same market as the defendant's, interchangeability in the eyes of consumers must be sufficiently great that if the defendant charged supracompetitive prices for its product, it would lose not just some of its customers, but a large enough number to make the supracompetitive pricing unprofitable.

American provides airline service in city pairs to and from its DFW hub. Passengers traveling on a particular city-pair route do not view service in other city pairs as a reasonable substitute: A person who wants to travel from Wichita to Dallas is unlikely to substitute a trip from Wichita to Kansas City because the price of the Dallas ticket has increased a small amount. Moreover, except for short journeys, few travelers regard other modes of transportation (e.g., bus, train, or automobile) as a reasonable substitute for airline transportation. Thus, airline passenger service in a city pair constitutes a relevant market.

Airlines may offer city-pair service on a "nonstop" basis or on a "connecting" or "one-stop" basis. Connecting or one-stop service requires a passenger to make one or more stops en route, usually to change planes along the way, and is generally less expensive than nonstop service. For many passengers, connecting or one-stop service is not a good substitute for nonstop service because connecting or one-stop service typically takes significantly longer than nonstop service and the risk of missed connections or lost luggage is greater; time-sensitive passengers, such as business travelers, are unlikely to substitute connecting or one-stop service for nonstop service in response to a small fare increase for nonstop service. Consequently, nonstop airline passenger service in a city pair is also a relevant market.

B. American Has the Ability to Control Prices and Exclude Competition in DFW City Pairs and Nonstop City Pairs

A firm has monopoly power if it can control prices and exclude competition in a relevant market. A finding of monopoly power is based on an evaluation of a variety of factors including of the defendant's share of the relevant market, the ease or difficulty of entry into the market, as well as evidence that the defendant has the ability to control or raise prices in the market. See Reazin, 899 F.2d at 967; Multistate, 63 F.3d at 1554; Von Kalinowski § 25.03[3][a] at 25-27.

The evidence will demonstrate that American has the power to control prices and exclude competition in the relevant markets.

1. Direct Evidence of American's Ability to Control Prices in the Relevant Markets

Empirical evidence will demonstrate that American has been able to charge supracompetitive prices in the relevant markets before LCC entry and after LCC exit: American's fares on DFW city pairs are substantially higher than its fares on otherwise comparable routes where it faces significant competition. These higher fares are often called a "hub premium."

2. Market Shares

American's high market share at DFW is strong evidence of American's monopoly power on DFW city pairs. See Eastman Kodak, 504 U.S. at 464 (market power "ordinarily is inferred from the seller's possession of a predominant share of the market"); Reazin, 899 F.3d at 969-70 (holding that percentages ranging from 47% to 62%, with evidence of market characteristics and evidence of a defendant's power over price and competition, was sufficient to support a jury's finding of monopolization).

American carries: 70% of all passengers who travel nonstop in DFW city pairs; 58% of all passengers who travel in DFW city pairs; 77% of all passengers originating in DFW who travel nonstop in DFW city pairs; and 65% of all passengers originating in DFW who travel in DFW city pairs. On the particular DFW city pairs which are the relevant markets, American's market shares are over 50% on 54 routes, over 70% on 29 routes, and over 90% on four routes; on particular DFW nonstop city pair routes, American's market shares are over 50% on 57 routes; over 70% on 43 routes; and over 90% on eight routes.(8)

3. Entry Barriers

Another indicator of a defendant's ability to control prices and exclude competition is evidence of barriers to entry -- "market characteristics which make it difficult or time-consuming for new firms to enter a market" and restrain a defendant's pricing decisions. Colorado Interstate Gas Co. v. Natural Gas Pipeline Co., 885 F.2d 683, 696 n.21 (10th Cir. 1989), cert. denied, 498 U.S. 972 (1990). Entry barriers include structural conditions and entrenched buyer preferences. See Reazin, 899 F.2d at 968. If a firm -- like American at DFW -- has high shares in markets that have entry barriers, then the prospect of entry is not likely to restrain the firm from charging prices that reflect its dominant position in the markets. Any airline that challenged American at DFW would have to overcome substantial entry barriers.

As a hub carrier, American enjoys significant advantages. American operates "spoke" routes that emanate from its DFW hub to numerous other cities. On spoke routes, American carries both "local" traffic (passengers traveling between the DFW hub and spoke city) and "connecting" traffic (passengers traveling between two spoke cities and transferring at the DFW hub). Any airline without a hub at one end of the market is at a serious disadvantage because it must carry mostly "local" passengers; it cannot achieve the revenue and cost advantages that come from also carrying connecting passengers on a city pair.

There are other advantages American receives from its hub. By providing more departures to more destinations out of DFW, American obtains a disproportionate share of DFW's passengers, that is, its share of passengers is greater than its share of capacity. This happens for several reasons, including the entrenched preference of many travelers to use American at DFW (which is related to brand recognition and American's greater service frequency on DFW city pairs), marketing programs (such as frequent flyer programs) that create incentives for consumers to concentrate their travel on American as the dominant airline at DFW, sales commission practices that create incentives for travel agents to encourage passengers to use American, and contracts with local businesses which commit them to use American for a substantial portion of their air travel in exchange for discounts, impeding the ability of smaller airlines that serve fewer destinations to attract business customers at DFW.

For an airline with a cost structure similar to American's to enter successfully in the face of American's hub advantages, the competitor must build a hub of its own. Building a hub, even one significantly smaller than American's, would be difficult, time-consuming and costly. The need to build a competing hub constitutes an important barrier that prohibits significant entry by other major airlines. Without a hub, an entrant could try entry using a low-cost, low-fare strategy, attracting passengers who otherwise would not travel when faced with American's high fares. However, American's ability and demonstrated willingness to cut fares and increase capacity quickly in response to such entry constitutes a formidable barrier to entry.

IV. Willful Acquisition or Maintenance of Monopoly Power

Under both a claim of monopolization and a claim of attempted monopolization, the Court must analyze the defendant's conduct to determine if that conduct represents a willful acquisition of monopoly power. Courts that have undertaken this analysis have used the terms "exclusionary" or "anticompetitive" or "predatory" as they seek to distinguish defendants' conduct from competition on the merits. Whichever label is applied, the inquiry is the same -- is the defendant's conduct an effort to exclude rivals on some basis other than the defendant's own improved market performance; is the full restrictive impact of the conduct on competition justified as necessary to further legitimate goals of lowering prices, improving quality, or in other ways promoting or expanding consumer choice; and are the conduct's costs to the defendant ultimately inexplicable except on the basis of the monopoly returns expected as a result of the conduct's creation or maintenance of a monopoly. See generally Aspen Skiing, 472 U.S. at 602-11; Eastman Kodak, 540 U.S. at 482-85.

The core of American's monopolistic strategy was that American deployed additional capacity and took pricing and yield management actions in DFW routes in response to LCC competition; the cost of these actions was greater than the revenues that came from carrying the additional passengers. With a similar disregard for the sacrifice of earnings, American cut its prices, reduced restrictions and/or increased its low-fare seat availability for its nonstop service on flights that provided significant traffic support to the connecting service provided by the LCCs, in order to keep traffic away from the LCC. See supra n.6.

The Court must determine whether American made these short-term sacrifices in order to drive the LCC out of the market (or to discipline it), or to deter others from entering the market, thereby enabling American to recoup the losses by raising prices after the LCC exited and to maintain its monopoly on other routes. See Instructional Sys., 817 F.2d at 649 (looking at the evidence to determine whether the defendant engaged in short-term sacrifices "to secure long term monopoly profits"); see also Aspen Skiing, 472 U.S. at 610-11(stating that the "evidence supports an inference that Ski Co. was not motivated by efficiency concerns and that it was willing to sacrifice short-run benefits and consumer goodwill in exchange for a perceived long-run impact on its smaller rival").

In making this determination, the Court must conduct two inquiries: (1) a recoupment inquiry: Did American have a reasonable expectation of recouping the short-term sacrifices it made in its monopolistic scheme? and (2) a cost inquiry: Did American's monopolistic scheme include actions that generated revenues that were below an appropriate measure of American's costs? Brooke Group Ltd v. Brown & Williamson Tobacco Corp., 509 U.S. 209, 222-24 (1993).

A. The Recoupment Inquiry

The purpose of the recoupment inquiry is to determine if the defendant's conduct was a rational business strategy, i.e., whether the defendant had "a reasonable expectation of recovering, in the form of later monopoly profits, more than the losses suffered," after it succeeded in eliminating or inhibiting competition. Id. at 224 (quoting Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 588-89 (1986)). In conducting the recoupment inquiry, the Court should examine evidence of the cost of the conduct and "both the scheme alleged by the plaintiff and the structure and conditions of the relevant market." Id. at 226.

American had a reasonable expectation of recouping the sacrifices entailed in its monopolistic strategy: American often could and did recoup its investment in a market by raising prices and decreasing capacity and the availability of low fares in a market, once it forced the LCC out of that market. American's strategy, however, did not require recoupment through this means to be successful. Rather, American engaged in its strategy to maintain its monopoly power in other DFW markets by stunting the growth of its LCC rivals and acquiring a reputation for aggressive response to entry at DFW that itself served to deter entry. In this way, American insured its ability to charge supracompetitive prices on DFW routes where it faces little or no competition.

B. The Cost Inquiry

The purpose of the cost inquiry -- determining whether the defendant's revenues from its conduct were below an appropriate measure of cost -- is to determine whether a defendant's conduct is inexplicable except by the monopoly returns the defendant expected to reap as a result of the conduct's creation or maintenance of a monopoly.

The Supreme Court refers to "an appropriate measure" of cost without suggesting that there is a single appropriate cost measure for this analysis. Id., 509 U.S. at 222 & n.1. Rather, the approach adopted by the Supreme Court recognizes that various measures of cost may illuminate the ultimate issue of whether a defendant is engaging in exclusionary behavior. The Tenth Circuit also recognizes various measures of cost in determining whether a defendant's conduct was exclusionary. See Multistate, 63 F.3d at 1549 n.5; Instructional Sys., 817 F.2d at 648; Pac. Eng'g & Prod. Co. v. Kerr-McGee Corp., 551 F.2d 790, 797 (10th Cir.), cert. denied, 434 U.S. 879 (1977).

The evidence will show that American's monopolistic scheme included actions that generated revenues that were, viewed under a variety of measures, below cost.

V. Intent

Intent is not an element of a monopolization claim, but evidence of intent is relevant to such a claim because it can assist the Court in determining whether the defendant's conduct is "fairly characterized as 'exclusionary.'" Aspen Skiing, 472 U.S. at 602.

Moreover, specific intent is an element of the attempted monopolization claim. Specific intent is intent to bring about the forbidden objective of a monopoly. See Swift & Co. v. United States, 196 U.S. 375, 396 (1905) (intent "to produce a result which the law seeks to prevent -- for instance, a monopoly"). Specific intent may be proved from direct evidence or inferred from evidence of the defendant's anticompetitive conduct. See Aspen Skiing, 472 U.S. at 609; Spectrum Sports, Inc. v. McQuillan, 506 U.S. 447, 459 (1993).

American's DFW LCC Strategy demonstrates that American had specific intent to monopolize the relevant markets. Moreover, American's prior anticompetitive behavior is also evidence of its intent to monopolize. See United States v. American Airlines, 743 F.2d 1114 (5th Cir. 1984), cert. dismissed, 474 U.S. 1001 (1985); In re Air Passenger Computer Reservations Sys. Antitrust Litig., 694 F.Supp. 1443 (C.D. Cal. 1988)(relying on United States v. American Airlines as evidence of intent), aff'd sub nom, Alaska Airlines, Inc. v. United Airlines, Inc., 948 F.2d 536 (9th Cir. 1991), cert. denied, 503 U.S. 977 (1992).

VI. Conclusion

For the foregoing reasons, American violated Section 2 of the Sherman Act by monopolizing or attempting to monopolize air transportation services in DFW city pair and nonstop city pair markets.

Dated: May 4, 2000

Respectfully submitted,


By: ________________"/s/"________________
Craig W. Conrath
U.S. Department of Justice
601 D Street, N.W. Suite 1200
Washington, D.C. 20530
(202) 307-5779


A copy of the foregoing was telefaxed this 4th day of May, 2000 to:

Edward Soto
Weil, Gotshal & Manges LLP
Suite 2100
701 Brickell Avenue
Miami, FL 33131-2861


Jennifer L. Cihon
United States Department of Justice
Antitrust Division
Patrick Henry Building
601 D Street N.W.
Washington, D.C. 20530 (202) 307-3278


1. An LCC is an airline that has lower operating costs than do major airlines, such as American, and offers fares substantially lower than the fares currently charged on the route, attracting not only consumers who have been paying the higher fares, but also consumers who otherwise would not have be able to afford that trip.

2. As discussed in section III, American provides airline service between a city of origin and a city of destination. These origin-destination combinations are known as "city pairs." For most airline passenger service to or from the Dallas/Ft. Worth area, DFW is the only available airport. The only other airport in the area used for commercial interstate airline service is Dallas Love Field, but the geographic scope and nature of service at Love Field are restricted by federal statute. During the time period relevant to the Complaint, only Texas and contiguous states could be served from Love Field by large aircraft. International Air Transportation Competition Act of 1979, Pub. L. No. 96-192, § 29, 94 Stat. 35, 48-49 (1980) (the "Wright Amendment"). City pairs that could have been served by large jet aircraft from Love Field are referred to in the Complaint as "Wright Amendment city pairs," while other city pairs for which Dallas/Ft. Worth is an endpoint, including DFW-Wichita, are referred to as "DFW city pairs." The Complaint alleges monopolization and attempted monopolization of certain DFW city pairs.

3. In the airline industry, the term "capacity" is used to describe the number of seats that a carrier offers to passengers on a route. Capacity is a function of the size, type, and number of aircraft operated by the carrier. The term "yield management" refers to efforts by carriers to maximize revenue on their flights. Carriers know that business passengers and other passengers often have to make reservations at the last minute or change their reservations. Carriers know that they can charge these passengers much higher fares than they can charge more discretionary passengers, such as vacationers, who are able to book reservations weeks or months ahead of time, can be more flexible, and may even decide not to travel if fares are too high. Carriers therefore have developed "yield management" processes pursuant to which they allocate a number of seats for high paying passengers and offer lower fares -- with restrictions -- only for the remaining seats. Sophisticated computer models determine the number of seats to allocate to high paying passengers on particular flights, and the usual goal of yield management is to "hold back" enough seats so that the carrier will always have a seat to sell at a high fare to those passengers who absolutely have to travel, thereby maximizing revenue on the route.

4. A load factor is a measure of the percent of capacity sold and used on a particular flight.

5. As a rational firm engaged in a monopolistic campaign, American did not intend to sacrifice earnings needlessly and it calibrated its actions. Accordingly, if an LCC posed little threat, American might tolerate its entry to a limited extent. If American believed a response requiring only a moderate sacrifice of earnings would be sufficient to eliminate or constrain a potential threat, American would maintain its existing capacity, but would target the LCC with fare matches and yield management actions designed to take passengers from the LCC.

6. For example, a passenger flying on Vanguard could fly from DFW to Chicago by making a connection in Kansas City. American matched Vanguard's connect fares even though American's DFW-Chicago service, unlike Vanguard's, was nonstop.

7. A monopolization claim requires proof of monopoly power whereas the dangerous probability element of an attempted monopolization claim is "usually demonstrated through the market power" of the defendant. Bright v. Moss Ambulance Serv., 824 F.2d 819, 823 (10th Cir. 1987). "Market and monopoly power only differ in degree -- monopoly power is commonly thought of as 'substantial' market power" and the same evidence relates to each. Reazin v. Blue Cross & Blue Shield, 899 F.2d 951, 967 (10th Cir.), cert. denied, 497 U.S. 1005 (1990).

8. Furthermore, no other carrier accounts for a large percentage of the relevant markets. The next largest carrier serving DFW is Delta Air Lines, Inc., which carries only 16% of all passengers who travel nonstop in DFW city pairs. Delta has gradually decreased the size and scope of its DFW operations over time and is unlikely to expand them. No other carrier accounts for more than 4% of such passengers.