UNITED STATES OF AMERICA et al., Plaintiffs, v. AMERICAN AIRLINES GROUP INC. and JETBLUE AIRWAYS CORPORATION, Defendants.
UNITED STATES DISTRICT COURT DISTRICT OF MASSACHUSETTS
May 19, 2023
SOROKIN, J.
May 19, 2023
SOROKIN, J.
I. INTRODUCTION
This case turns on what “competition” means. To the defendants, competition is enhanced if they join forces to unseat a powerful rival. The Sherman Act, however, has a different focus. Federal antitrust law is not concerned with making individual competitors larger or more powerful. It aims to preserve the free functioning of markets and foster participation by a diverse array of competitors. Those principles are generally undermined, rather than promoted, by agreements among horizontal competitors to dispense with competition and cooperate instead. That is precisely what happened here.
Each of the defendants is a formidable and influential player in the air travel market in this country. American Airlines Group Inc. is the largest airline in the world. It offers more seats and serves more origins and destinations than any other carrier in the United States. It is one of four airlines that control approximately eighty percent of domestic air travel. JetBlue Airways Corporation is the sixth-largest airline in the United States. Younger than its larger domestic competitors and using a lower-cost business model, JetBlue has nurtured its reputation as a maverick airline seeking to disrupt the industry to the benefit of consumers. Until 2020, American and JetBlue were fierce and frequent head-to-head competitors. This was especially so in the northeast, where JetBlue looms large and centers a majority of its operations.
American and JetBlue are two of the four largest carriers operating in New York, and two of the largest three in Boston. Delta Air Lines is the only other carrier with a large presence in Boston. Besides Delta and United Airlines, no other carrier matches or approaches in size the defendants’ respective positions in New York. Challengers seeking to enter or expand in New York would first need to secure gates from which to operate, as well as schedule authorizations from federal regulators who control air traffic in one of the most congested markets in the world. Both the gates and the authorizations are exceedingly difficult to acquire. Challengers seeking to enter or expand in Boston would need to secure gates, which also are in scarce supply. Because of these significant barriers, the defendants’ positions at or near the top of these constrained markets had proven relatively robust and durable over the decade or so preceding the onset of the COVID-19 pandemic.
In the first months of 2020, executives at American Airlines and JetBlue negotiated and signed a first-of-its-kind alliance, in which the two carriers essentially agreed to operate as one airline for most of their flights in and out of New York City and Boston. The partnership is called the Northeast Alliance, or the NEA. This was a sea change in the relationship between two airlines that were direct and aggressive competitors with decidedly different business models and cost structures. There is no doubt that savvy executives representing both defendants earnestly believe the NEA promotes the interests of their respective shareholders and will strengthen American and JetBlue in their rivalry against Delta (and, to a lesser extent, United) in New York and Boston. It is similarly beyond dispute that the NEA
Invoking the Sherman Act, the United States Department of Justice, joined by the District of Columbia, the states of Arizona, California, and Florida, and the Commonwealths of Massachusetts, Pennsylvania, and Virginia, filed suit to enjoin the defendants from proceeding with the NEA. The lawsuit culminated in a month-long bench trial featuring testimony by two dozen witnesses, most of whom were either executives of the defendants or experts paid for their testimony by one side or the other. The trial transcript surpasses 3,600 pages, accompanied by at least fifty binders containing exhibits presented to witnesses. The live testimony was augmented by more than 2,700 pages of excerpts from the depositions of seventeen additional witnesses. More than a thousand exhibits were admitted into evidence. Post-trial written submissions by the parties exceeded six hundred pages. This tidal wave of evidеnce reflects both the state of antitrust litigation1 and the “unprecedented” nature of the NEA. Doc. No. 1 at 2.2
After close attention to the evidence at trial and a careful review of the voluminous submissions by the parties, certain points became clear. Within the highly concentrated airline markets in New York and Boston, where opportunities to enter or expand are vanishingly rare, JetBlue stood largely alone as the only low-cost airline with a significant presence in a domestic
market dominated by larger, higher-cost network carriers. With the NEA, American and JetBlue transformed themselves from competitors to collaborators, joining forces to create a single “optimized network.” They design that network together, jointly determining which airline will fly which routes in and out of the NEA region, how often and on what schedule they will serve each route, and which aircraft (i.e., how many seats) will be used on each route. To further promote the arrangement, American and JetBlue share the revenues each generates within the NEA.
This is no minor shift for the two businesses or the region. Nearly three-quarters of JetBlue‘s overall operations are flights in or out of the NEA. American counts New York among its hubs and is the third-largest carrier operating in Boston. In both locations, the defendants vigorously competed on everything from fares to the features they offered customers. The NEA changes all of that. It makes the two airlines partners, each having a substantial interest in the success of their joint and individual efforts, instead of vigorous, arms-length rivals regularly challenging each other in the marketplace of competition. Though the defendants claim their bigger-is-better collaboration will benefit the flying public, they produced minimal objectively credible proof to support that claim. Whatever the benefits to American and JetBlue of becoming more powerful—in the northeast generally or in their shared rivalry with Delta—such benefits arise from a naked agreement not to compete with one another. Such a pact is just the sort of “unreasonable restraint on
In arriving at the findings of fact and conclusions of law set forth in the following pages, the Court sifted through the evidence and assessed it with a few foundational principles in mind. First, the Sherman Act aims to broadly preserve “free and unfettered competition as the rule of trade.” N. Pac. Ry. Co. v. United States, 356 U.S. 1, 4 (1958). As long as the competitive process is functioning freely, it is not the concern of the Sherman Act (or of a federal court applying it) which competitors win dominant shares in any given market. Next, certain restraints, based on their character or context, pose threats of anticompetitive harm that are sufficiently obvious that they warrant careful scrutiny, if they can be justified at all. See United States v. Topco Assocs., Inc., 405 U.S. 596, 607-09 (1972). Such restraints include agreements between powerful horizontal competitors to control output or allocate markets, which trigger an especially heavy burden on the collaborators to justify what otherwise would be obviously unlawful collusion. Lastly, despite its unusual complexity, this case requires the Court to call upon familiar tools of the judicial trade—observations of witness demeanor, common sense, and a general understanding of human behavior—as it evaluates the credibility and assesses the motivation of people describing their roles in conceiving, debating, and implementing business decisions on behalf of their employers.
Guided by these standards, and for the reasons explained below, the Court finds that the plaintiffs have convincingly established that the NEA violates Section 1 of the Sherman Act.3
II. FINDINGS OF FACT
A. The Industry
The United States passenger airline industry, as we know it today, has been shaped by a series of events that unfolded during the past four decades. After the industry was deregulated in
1978, then-existing carriers began building larger, nationwide networks.4 Meanwhile, new carriers emerged, operating with lower cost structures that allowed them to offer lower prices and compete for market share with the established network carriers. Vigorous competition, an increase in capacity,5 and a reduction in ticket prices followed. The first decade of the 2000s, however, saw the airline industry rocked by external events including the September 11, 2001, terrorist attacks and the 2008 global financial crisis. During the same time period, overall airline capacity fell, and the number of domestic сarriers declined, as struggling airlines—including all of the predecessors to the current three largest domestic carriers—declared bankruptcy or pursued mergers and acquisitions. As a result of
Today, domestic carriers can be roughly divided into four categories based on business model and cost structure. Global network carriers (“GNCs“) possess broad networks that reach a wide range of origins and destinations (“O&Ds“) either directly or through connecting itineraries. A GNC relies on a collection of “hubs“—airports where the carrier operates at a significant scale, with many flights arriving and departing each day—and “spokes“—other destinations the carrier serves on a smaller scale—to create a network that can serve customers going to or from as many places as possible. Three GNCs operate domestically today: American, Delta, and United. Each GNC is the result of one or more mergers.6
Low-cost carriers (“LCCs“) generally rely on point-to-point flying using a single type of aircraft with a single class of service. This simplifies operations by ensuring every pilot, flight attendant, and mechanic can serve every plane in the fleet. Carriers employing this business model enjoy operation costs that are lower than those of a GNC, allowing them to remain profitable while offering lower fares than a GNC. Southwest Airlines was the first LCC and remains the largest domestic LCC. Ultra-low-cost carriers (“ULCCs“), as the name suggests, operate with even lower costs and offer even lower fares. They, too, generally offer point-to-point flying with one class of service and one type of aircraft, often focusing on high-traffic routes with a substantial demand for direct service. ULCCs achieve lower fares by selling an “unbundled” product. That is, a typical ULCC fare includes only transportation from one place to another (often in less comfortable seats); few additional products and services are available, and they carry additional fees. Spirit Airlines was the original domestic ULCC, with Frontier Airlines, Allegiant Air, Avelo Airlines, Breeze Airways, and Sun Country Airlines now joining it in the category.
The last category includes carriers that fall somewhere between the GNC and LCC business models. Alaska Airlines and Hawaiian Airlines are such hybrid carriers, operating what amount to regional hub-and-spoke networks. At trial, different witnesses characterized JetBlue as an LCC, a hybrid carrier, or a carrier presently transitioning between categories. The Court will describe JetBlue‘s status further in the next subsection. See discussion infra Section II(B).
Airlines serving overlapping geographical areas with similar business models tend to compete most directly and most often with one another. Such carriers offer similar fares, fly similar types of aircraft, provide similar levels and varieties of services, and serve similar locations and categories of customers. However, airlines make pricing and scheduling decisions on a market-by-market basis. As to any given market, an airline generally considers the actions of, and views itself as competing with, every other airline serving that market, regardless of business model, as well as other airlines capable of entering that market. If American, Delta, Southwest, and Spirit all provide direct flights from City A to City B, Delta will consider reacting to fare or schedule changes by any of its three direct competitors serving the route. In the airline industry, scheduling
The industry is highly concentrated. Four carriers control more than eighty percent of the market for domestic air travel: the three GNCs (American, Delta, and United) and Southwest. The remainder of the market—less than twenty percent—is generally split among nine smaller carriers. Though most of the smaller carriers formed after the industry was deregulated, the four dominant carriers existed in some form before that time. Each of those four is the product of consolidation in the industry.7
Though the merger of two carriers with complementary networks can result in one bigger carrier with a network that is broader and deeper than what existed before, a merger also requires the investment of substantial time and resources. The process of effectively combining the separate operations and assets of two merging entities is costly and complex. For example, schedules, networks, aircraft, personnel (including unionized groups of workers), technology systems, advertising and branding, loyalty programs, and real estate (from airport lounges to
hangars) all must be combined into a cohesive, seamless, single carrier. American‘s Chief Executive Officer (“CEO“) described the numerous challenges created by mergers, as well as the “inordinate amount of management time and attention” required to integrate two airlines. Trial Tr. vol. 5 at 58-59. Evidence regarding carrier growth from 2009 to 2019 corroborates that testimony and suggests a correlation between growth trends and the merger digestion process. For example, Delta appears to have benefited from completing its merger first and managing the integration expeditiously, paving the way for fairly aggressive and steady growth beginning in 2012. United trailed Delta by two years and effected its merger slightly less smoothly; its growth line turned upward in 2014 but became more aggressively so only around 2016. American‘s merger was later and more protracted, with the integration continuing into 2019. Nevertheless, American‘s growth was on an upward trajectory even before its merger, with a steeper incline beginning in 2017.
The trend of consolidation over the past twenty or so years is subject to differing interpretations. For example, various American executives have praised consolidation as a necessary strategy that created a healthier industry with capacity levels that appropriately balance the needs of consumers and carriers. On the other hand, JetBlue‘s executives have often warned that consolidation leads to higher costs, reduced capacity, and less choice for consumers. The Court‘s role is not to resolve which general view is more apt or to chart a regulatory course for the industry, but rather to resolve the dispute presented in this case. It is enough for present purposes to observe that the number of domestic carriers from which air travelers in the United States may choose has diminished, and market share in this industry has become meaningfully more concentrated, over time.
These trends are evident in the northeast, though with one caveat worth noting. Despite the power it commands nationally,
An airline‘s ability to operate at a particular airport depends on a number of factors, some of which are especially pertinent here. One is access to gates at which passengers can board and disembark flights. The number of gates allocated to a carrier dictates the number of flights it can operate at the airport.9 Like most airports, Logan is gate-constrained; a carrier looking to initiate or expand service there would first need to secure access to gates.10 A handful of airports have
additional limitations on access to air space. In Newark, a carrier must secure accеss to gates and schedule approval from the Federal Aviation Administration (“FAA“), which monitors air traffic demand there. At LaGuardia and JFK—two of the busiest airports in the country—carriers must acquire both gates and “slots.” A slot is authorization from the FAA to land or take off during a particular period of time. Slot control enables the FAA to regulate air traffic in certain congested, high-demand areas. Generally, once slots are awarded, they are treated as the property of the airline obtaining them. A carrier looking to initiate or expand service at LaGuardia or JFK would first need to secure access to gates and slots, both of which are scarce, valuable, and sought-after resources.11 Numerous witnesses explained at trial that operating in New York is a costly proposition, and that opportunities to obtain slots at LaGuardia and JFK are exceedingly rare.12 Though these constraints
The airline industry, like the rest of the world, was turned upside down by the COVID-19 pandemic. In March 2020, demand for air travel in the United States all but vanished. Airlines cut capacity, parking planes in the desert to wait out the pandemic. The FAA temporarily
excused carriers at slot-controlled airports from the usage requirements they normally must satisfy in order to retain the rights to their slots. As travel began to resume—later and much more slowly than expected—airlines altered strategies and schedules to account for changes in the relative demands for leisure and business travel that persist even today. Some executives predicted that business travel might never return to pre-pandemic patterns.
It is against this backdrop of industry consolidation, in this competitive landscape, and amid an industry meltdown during the early months of the COVID-19 pandemic that the agreement at issue here arose.
B. The Defendants
JetBlue is much smaller than American, ranking as the sixth-largest airline in the United States. In 2019, it operated six “focus cities“: New York City, New York (also the location of JetBlue‘s headquarters); Boston, Massachusetts; Fort Lauderdale/Hollywood and Orlando, Florida; Los Angeles, California; and San Juan, Puerto Rico. Approximately three-quarters of JetBlue‘s operations have either Boston or New York as an origin or destination. JetBlue‘s business model has historically centered on pursuing aggressive growth, providing high-quality service, offering affordable fares, and taking market share from other airlines (especially the GNCs). This model has constrained prices charged by other airlines (again, especially the GNCs) and promoted competition. Though often referred to as an LCC—a category into which it once comfortably fit—JetBlue‘s business model and cost structure have evolved over time. The carrier now offers more than one class of service and has a fleet featuring more than one type of aircraft. Some witnesses characterized JetBlue as either a hybrid carrier (akin to Alaska, but focusing its operations along the east coast), or as attempting a migration to the GNC category.
However it is presently categorized, JetBlue plainly occupies a unique position in the domestic airline industry. The carrier prides itself on its “disruptor” status. Its executives have spoken publicly—loudly and often—about the harms they believe consolidation, the GNCs, and coordination via unchecked alliances have wrought on consumers. JetBlue‘s aggressive approach to competing with the GNCs and the responses it has provoked are well documented.13 See, e.g., Doc. No. 325 ¶¶ 32-45 (summarizing evidence of various instances in which JetBlue impacted the prices and service of the GNCs, and American in
It is beyond dispute that, through June 2020, JetBlue vigorously and directly competed with American across all markets both carriers served. See Doc. No. 325 ¶ 236 (describing announcement of new routes by JetBlue on the eve of the NEA‘s signing, including new nonstop overlaps with American). JetBlue‘s Mint sеrvice distinguished it from every domestic LCC and ULCC carrier (all of which offer a single class of service) and enabled it to compete with the GNCs for corporate clients—especially those in the northeast, where JetBlue‘s presence is especially strong—in a way the other non-GNC airlines could not. The competition was not a one-way street, with JetBlue triggering fare responses by American. It worked in the other direction, too. For instance, when American removed capacity in some markets in the northeast due to the grounding of part of its fleet, the competitive pressure arising from American‘s presence eased, and JetBlue raised its fares in response.
American is one of the most powerful airlines in the world. By some measures, it is the largest carrier both domestically and internationally. Headquartered in Texas, American identified the following cities as its hubs in 2019: Charlotte, North Carolina; Chicago, Illinois; Dallas/Fort Worth, Texas; Los Angeles, California; Miami, Florida; New York City, New York;14 Philadelphia, Pennsylvania; Phoenix, Arizona; and Washington, D.C. The carrier achieved its dominant position via a combination of mergers, alliances, and joint ventures, some of which aimed specifically to strengthen American‘s network in the northeast.
American pursued growth by establishing relationships with other domestic and international carriers. It founded the global oneworld alliance, which includes American and thirteen other airlines, and it participates in three smaller joint ventures focused on transatlantic service and transpacific service to both Asia and Australia/New Zealand. An airline based in one country is generally unable to serve routes that begin and end in other countries. For
connecting flight from Madrid to a smaller destination in Spain, or from Madrid to other destinations throughout Europe. Through the oneworld alliance, however, American can rely on one or more partner airlines (for example, Iberia) to complete such itineraries—and members of American‘s frequent-flyer program can accrue or spend miles on all legs of the trips.
Members of these international arrangements generally coordinate schedules, share access to airport lounges, offer reciprocal loyalty benefits, allocate markets, jointly decide on capacity, share profits, and sometimes make joint pricing decisions, all with the aim of providing their customers with access to a global network no member airline alone could replicate. Because of such features, alliances and joint-business agreements like these are reviewed by government regulators and require antitrust immunity in order to operate.
Carriers in the United States have not historically attempted arrangements that intertwine their operations so broadly with other domestic airlines. This is at least partly due to a general understanding across the industry that such coordination would run afoul of federal antitrust law.15 Domestic carriers have cooperated on much smaller scales. For example, some develop interline agreements, which essentially promise that if one carrier must rebook its passengers in the wake of a cancelled flight, it may offer its passengers open seats on its interline partner‘s flights as well as its own. Others have adopted codesharing—whereby one carrier places its own number (or code) on a flight operated by its partner, allowing for customers of both carriers to
locate and purchase seats on the flight through either carrier‘s website—with or without some degree of loyalty-program reciprocity. Delta once had such a relationship with Alaska (before Delta strengthened its own west-coast presence), as does JetBlue with Hawaiian.
Led by Vasu Raja, then its Senior Vice President of Strategy,16 American began contemplating a new domestic strategy in 2019, which involved pursuing deeper partnerships with other domestic carriers. This effort started before the pandemic took hold, and it eventually crystallized into two partnerships—one aimed at addressing American‘s perceived weaknesses on each coast. In February 2020, American annоunced the West Coast International Alliance (“WCIA“) it formed with Alaska.17
The WCIA has the following salient features: 1) it makes Alaska a member of American‘s oneworld alliance; 2) it continues
any routes on which both partners offered competing direct service (“nonstop overlaps“) are excluded from the scope of the WCIA, including its codesharing provision.
Representatives of both Alaska and American described the WCIA as a success, noting it remains in place today and is serving its intended purposes. The WCIA is designed to benefit each partner in a different way. It enables Alaska, an airline without significant international service and with no plans to begin long-haul flying, to provide its customers access to American‘s international flights and those of its oneworld partners. This, in turn, helps Alaska “address a growing threat from Delta in Seattle, Alaska‘s primary hub.” Doc. No. 324 ¶ 10. For American, the WCIA feeds connecting traffic to its international long-haul flights via Alaska‘s domestic service on the west coast (primarily, at its Seattle hub). The relationship is important to American, which viewed itself as operating at a disadvantage on the west coast, where Delta, United, and Southwest each have a more substantial presence.18 Both partners to the WCIA believe it provides their customers with access to a better network and better loyalty benefits on the west coast, and that it does so as seamlessly as possible.
For purposes of antitrust analysis, there are other salient features of the WCIA. American and Alaska, by and large, were not direct competitors. They provided competing nonstop service on few, if any, domestic O&Ds, and Alaska did not offer international long-haul service. In other words, their separate networks were fairly characterized as more complementary than overlapping. The terms of the WCIA largely leave competition between the two airlines intact. They do not coordinate schedules, they do not allocate markets, they share revenue only in a
limited way, and they continue to operate as separate airlines in all respects. Even with these limitations, both partners believe the WCIA accomplishes its purposes—including strengthening their positions with respect to their shared rival, Delta.
Neither the WCIA nor any other domestic airline joint venture has received antitrust immunity. There is no evidence that any domestic airlines have formed relationships involving revenue sharing, pooling of slots and gates with joint decision-making about their use, allocation of markets, coordination of schedules, or broad efforts to operate as one airline in a substantial region of the country. At least, that was the case until American and JetBlue formed the NEA.
By 2020, JetBlue knew that Delta was mounting a challenge to its dominance in Boston. Delta had invested in growth there, ultimately declaring Logan a Delta hub. Meanwhile, JetBlue‘s growth in New York had stalled due to its inability to secure access to more slots at LaGuardia or JFK. Around the same time, American was fretting about its operations in New York. It had a strong historical position there,19 controlled the second-most slots at LaGuardia and the third-most slots at JFK, and counted New York among its hubs. Nevertheless, American did not consider its operations in New York to be sufficiently profitable, and it believed growth by Delta (at JFK and LaGuardia) and United (at Newark) posed a threat to American‘s overall position in the region. By the fall of 2019, American perceived that JFK slot usage was “under heavy scrutiny with the FAA,” and that American‘s underuse of its slots in recent years put those
valuable assets at risk. PX 0148 at 3.20 These atmospherics set the scene for negotiations between American and JetBlue that culminated in the NEA.
In late 2019, the two carriers began discussing a possible lease, through which JetBlue would acquire temporary control over some of American‘s slots at JFK. Though they negotiated an agreement to lease twenty-seven slots on a short-term basis, American subsequently proposed adding more slots for a longer (two-year) term, and internal discussions at JetBlue reflect a belief among its network planners that the leases would continue or renew for longer than a season or two. E.g., PX 0507 at 1; PX 0527 at 1. Talks between the carriers expanded to contemplate a broader domestic partnership focused on the northeast, as envisioned by Raja and modeled after the WCIA. The record establishes that a primary goal—and a significant concern—motivating both American and JetBlue to pursue a partnership was a mutual desire to address the competitive threat they each perceived Delta presented in markets they deemed important.21 See PX 0268 at 2-3 (describing purpose of the initiative that yielded the NEA as improving the competitive positions of American and JetBlue “relative to” Delta and United). The parties had
another set of
Negotiations between American and JetBlue continued despite the COVID-19 pandemic. In April 2020, on the advice of their legal departments, American and JetBlue each designated representatives to a “Clean Team“—a group of individuals with knowledge of scheduling and network planning, but whose daily responsibilities did not involve such work.22 The Clean Team built a theoretical joint network schedule that would allow American and JetBlue to evaluate what the carriers could achieve via a partnership. This process lasted through May 2020. Ultimately, the Clean Team produced a hypothetical schedule for 2023,23 which pooled the resources of both carriers—including aircraft they did not yet possess but, per their respective order books, they expected to receive by 202324—and “optimized” them to create one cohesive NEA schedule. The Clean Team then ran the schedule through a proprietary tool American uses to estimate passenger traffic and revenue.25
Satisfied with the results, and after consulting with experts and lawyers, American and JetBlue proceeded to finalize their partnership. The NEA was established and publicly announced on July 15, 2020. The defendants’ relationship consists of a set of related contracts and is primarily governed by the NEA Agreement, the terms of which take priority in the event of any conflict among the contracts. Together, these agreements create a relationship between American and JetBlue that includes codesharing, schedule coordination, revenue sharing, reciprocal loyalty benefits, and joint corporate customer benefits, all of which extend to most of the carriers’ flights to and from Logan, JFK, LaGuardia, and Newark (“the NEA airports“). In particular, both carriers’ short-haul services to and from the NEA airports are included within the scope of the agreements, as are American‘s long-haul services touching the NEA airports.26
The NEA also includes separate agreements providing for codesharing and reciprocal loyalty benefits by American and JetBlue on NEA routes.27 Though the NEA Agreement contemplates joint bids for corporate customers, no customer had requested such a bid, and the defendants had not sua sponte engaged in such a practice, as of the time of trial.
The NEA‘s revenue-sharing mechanism is established by the Mutual Growth Incentive Agreement (“MGIA“), which was executed concurrently with the NEA Agreement. Modeled after the profit- or revenue-sharing features of American‘s other joint ventures and alliances, the MGIA‘s purpose is to align the partners’ incentives and achieve something the parties call “metal neutrality.” Metal neutrality means American and JetBlue are indifferent to whether a passenger flies a particular NEA route on an American plane or a JetBlue plane. Metal neutrality within the NEA region—that is, on flights to or from the four NEA airports—is a cornerstone of the NEA. The sharing of revenues generated by both partners in the alliance renders each agnostic about which partner‘s aircraft a customer chooses, and also furthers both partners’ shared objective of attracting passengers away from other competitors (here, Delta and United).
The MGIA establishes a complex process for splitting the revenue pool annually. The pool itself, or Net Revenue, is composed of defined categories of passenger-related revenue, less certain selling expenses.28
The defendants urge that the MGIA is designed to incentivize growth by both of them, and to spur continued competition between them. E.g., id. ¶¶ 196-97. Certain features do appear to reward growth. For example, by calculating the Base Revenues using a factor tied to each partner‘s present-year flying, rather than sharing in a fixed proportion based entirely on performance in the base year, the MGIA makes it possible for a carrier to increase its share by adding capacity (in the form of seats or miles). And, of course, splitting Incremental Revenue based on capacity might reward one partner for growing more than the other. These features, however, are not the only incentives at play, and they do not overpower other, stronger influences affecting the parties’ behavior.
For one thing, like partners in nearly any joint venture, both American and JetBlue have a strong and admitted interest in ensuring the NEA succeeds over the long term. Common sense suggests, and the Court finds based on all of the testimony and evidence, that a spirit of partnership between the two carriers will overwhelm any incentive for intra-partnership competition the MGIA might facially appear to create.30 American and JetBlue “now function as a single airline” when they decide how to “optimiz[e] capacity” in the NEA region, and they “don‘t compete with each other directly” on NEA routes. Trial Tr. vol. 1 at 182. It is implausible that revenue-sharing terms designed and intended to align the parties’ economic incentives and foster decision-making in the best interests of the partnership will simultaneously spur the partners to compete vigorously with one another in terms of growth. Unchecked growth by one partner undermines the spirit of partnership and, thus, the entire venture. If one carrier
Apart from the prevailing spirit of partnership, many other factors impact both partners’ ability and desire to embark on a race against one another to add capacity. For example, each carrier is limited by its fleet, its workforce, its access to capital, its business goals, its brand, its fixed cost structure, and its operational needs in regions beyond the northeast—all of which are factors likely to bear more heavily on their joint or separate capacity decisions than the details of the MGIA‘s revenue-sharing formula. The Court rejects the defendants’ suggestion that the MGIA will, in practice, foster or preserve substantial and vigorous competition between American and JetBlue.
The NEA automatically renews for consecutive five-year terms, unless certain defined processes are invoked. If one partner notifies the other within the first five years “that it does not want to proceed to another term absent material changes,” then the parties must discuss possible amendments in good faith. PX 0001-a § 5.1. If they “cannot reach agreement,” then the NEA automatically extends “for an additional two years” and will “expire on the seventh anniversary” of its signing. Id. Additionally, in various circumstances, a party may elect to invoke termination rights defined in the NEA Agreement, subject to specified procedures and fees. Id. §§ 5.2-5.10. After the first five-year term has concluded, any party may decline the automatic renewal process by notifying the other party “at least one year in advance of the end of the then-current term that it does not desire to renew this Agreement.” Id. § 5.1. In other words, the arrangement will be in place for at least seven years and, absent affirmative efforts to terminate or prevent renewal, will continue indefinitely.
The NEA Agreement has been amended three times. The First Amendment, signed September 11, 2020, replaced certain provisions in the original agreement and added language to others. PX 0001-a at 45-47. Of note, it added language limiting the ability of either partner to “voluntarily sell or otherwise transfer any slots” at JFK or LaGuardia to other carriers. Id. at 45. The Second and Third Amendments occurred in connection with a regulatory review process undertaken by the Department of Transportation (“DOT“). That process began shortly after the NEA was announced and continued through the end of 2020. PX 0447 at 1. The DOT terminated its review after securing a series of commitments from American and JetBlue “to address competitive issues arising from the NEA.” Id.
The commitments are memorialized in an agreement signed on January 10, 2021 (“the DOT Agreement“).32 They include: 1) a promise by JetBlue not to exit certain
The Second and Third Amendments to the NEA Agreement, signed shortly after the DOT Agreement, effect the changes the defendants promised in the commitments described above. See PX 0001-a at 48-54. American and JetBlue concurrently amended the MGIA “to remove revenue sharing on routes in which American and JetBlue have a high market share.” PX 0001-b at 31. That amendment defines “Excluded Services“—or, as they were referred to throughout these proceedings, “carve-out routes“—in a manner that described six O&Ds as of January 2021, all of which included Logan as an endpoint.36 See id. § 1.2 (defining term and listing routes between Boston and Phoenix, Rochester, Syracuse, Charlotte, Philadelphia,
Later in 2021, American and JetBlue executed a pair of slot lease agreements. American leased to JetBlue ten slots at JFK and thirty-seven at LaGuardia for approximately one year, ending October 29, 2022. PX 0001-h. JetBlue leased to American eight slots at JFK for essentially the same term. PX 0001-i. In March 2022, American leased to JetBlue thirty-two more slots at LaGuardia and twenty more at JFK for terms beginning on different dates in 2022 and ending in March or October of 2023.37 PX 0001-j.
The NEA, of course, is not a merger. American and JetBlue remain separate entities. Both have operations that fall beyond the NEA‘s reach, and the agreement does not formally embody a complete combination of the partners’ operations even within the NEA region. Nevertheless, as implemented by the parties, its effects resemble those of a merger of the parties’ operations within the northeast in ways the Court will describe next.
D. The Effects
American and JetBlue began implementing the NEA in early 2021. They did so in phases and were approximately eighty percent done by the time of this trial. That the partnership was unveiled, and its implementation undertaken, during a period when the COVID-19 pandemic continued to substantially impact the airline industry makes efforts to assess the NEA‘s effects especially challenging.38 Nevertheless, the parties point to a number of changes arising after the NEA that they attribute to it. The Court‘s evaluation of the evidence adduced at trial leads it to make the following findings about the NEA‘s effects.
First, American and JetBlue no longer compete with one another within the scope of the NEA. Rather, they function like a single airline in the NEA region, as much as possible. The revenue sharing established by the MGIA is designed to make the two carriers indifferent to whether a customer chooses to purchase a flight from American or from JetBlue. The provisions of the NEA aimed at optimizing the carriers’ schedules to produce one cohesive schedule mean that the carriers act as one airline in the northeast when choosing which routes to fly, when to fly them, and which aircraft (and which partner) will do so. Indeed, nearly every witness employed by either defendant testified to some extent about the importаnce of achieving a “seamless” customer experience within the NEA. In pursuit of that goal, the defendants necessarily collaborate to ensure that, throughout their “NEA network,” customers receive service that matches as closely as possible that which would be provided by a single carrier.39
Second, the NEA has caused both American and JetBlue to adjust their overall network priorities, with both carriers now intensely focused on serving and growing in New York, at the expense of at least some pre-NEA plans to devote resources to growth elsewhere.42 For example, though American‘s long-range plans in February 2020 included growth in Philadelphia, by September 2020 it had deprioritized Philadelphia relative to JFK. See Doc. No. 325 ¶¶ 451-52, 455-56. To launch new transatlantic flights from JFK and support growth in the NEA region, American took aircraft out of its hub in Philadelphia. See Trial Tr. vol. 7 at 158-59 (stating American “very much want[s] to put back
These tradeoffs, which the defendants hope will be temporary, are necessary because American and JetBlue are constrained by the size of their fleets. Like any airline, each defendant has a finite number of aircraft presently available, plus additional aircraft it anticipates receiving in the future based on its order book. A carrier, therefore, cannot generally grow everywhere, all at once. Rather, like all businesses, airlines must evaluate their priorities, decide how to allocate their limited resources, and choose where and when to pursue growth. Here, in the short term at least, the record establishes that this internal process has led the defendants to prioritize the NEA region when determining how to allocate resources.43 To be sure, executives of both defendants expressed confidence that the NEA‘s anticipated success would support efforts to add to their fleets by purchasing new planes in the future—a proposition that makes intuitive sense, assuming all else remains constant and the industry is not rocked by another negative demand shock triggered by an unforeseen global event. However, the record simply contains no proof that this optimism has yielded such results to dаte. Of course, the elimination of competition between American and JetBlue is the fuel supporting the hope for future expansion by ending a major risk that would otherwise balance against the possible revenues generated by a new plane.44
As to the first, business records reflect that JetBlue‘s Revenue Management team analyzed various scenarios regarding the retirement of the thirty aircraft—including one in which JetBlue would delay their retirement even without the NEA. See generally PX 0816. Witnesses’ later attempts to suggest that this scenario was included “for completeness,” was only “theoretical,” or was never “really under consideration,” see Doc. No. 324 ¶ 279, are not supported by any similar characterization in any document contemporaneous to the analysis at issue, and the Court does not credit them. No evidence before the Court suggests, let alone establishes, that JetBlue could not have delayed the retirements to pursue its own standalone growth plans. To the extent the defendants also suggest that JetBlue accelerated its existing orders for new aircraft because of the NEA, the record leads the Court to conclude that the new planes at issue were part of JetBlue‘s plan (with or without the NEA) to retire an older, smaller, and less efficient segment of its fleet and replace it with newer, larger, more efficient planes. And, to the extent JetBlue exercised options to order more planes than the number necessary to replace those being retired—a claim supported by evidence the Court finds murky, at best—such planes are not presently part of JetBlue‘s fleet and may not be so for several years, at least. PX 0949 at 36; cf. Trial Tr. vol. 7 at 158-59 (expressing frustration that aircraft American ordered and expected to receive already have not yet arrived).
As to American‘s weaker claim of fleet expansion, the witness who testified most candidly on this topic declined to directly attribute such (potential, future) expansion strategies to the NEA. E.g., Trial Tr. vol. 15 at 107-08, 110; see also id. at 77-78 (conceding “the source of the bigger planes” in New York in “the short term” is to “relocate [them] from other markets across the system“); id. at 94 (saying generally that American “would look to add a number of airplanes . . . by 2026“). The parties have not pointed the Court to any ordinary-course business records of either defendant reflecting a specific pitch to a Chief Financial Officer or lender seeking and obtaining funding for aircraft because of the actual or anticipated effects of the NEA.
Fourth, the NEA‘s schedule optimization and capacity coordination process has led to decreased capacity, lower frequencies, or reduced consumer choices on multiple routes, including some that are heavily traveled. This has happened in at least two ways. In certain markets the defendants both previously served, the NEA has allocated the route to one of them and caused the other to exit. In short, they are dividing the NEA markets between themselves. Examples include thirteen markets touching LaGuardia at one end, including those with Logаn, Philadelphia, Charleston, and Orlando at the other end. Doc. No. 325 ¶¶ 272-73 (citing testimony and exhibits); PX 0322 at 6. In these markets, the number of competitors has literally decreased by one, leaving consumers with one fewer option when traveling between these pairs of cities. On the heavily traveled business route between Logan and LaGuardia, American‘s exit reduced the total number of carriers providing nonstop service from three to two, with Delta providing the only direct service competing with JetBlue‘s. The only reason either defendant stopped service on these routes is
These facts bear on the number of competitors participating, and the choices available, in NEA markets. The defendants’ partnership also has reduced total frequencies or capacity in certain NEA markets that the defendants have allocated to only one of them in their optimized NEA schedule. The same markets identified in the previous paragraph provide examples of such reductions. See Trial Tr. vol. 9 at 48-49 (predicting JetBlue “will eventually have roughly the same capacity” on Logan-LaGuardia “that existed in the market” before the NEA, but not until “some time between now and 2025“); PX 0792 (describing plan for American to stop its four daily Logan-LaGuardia flights at the start of 2022, for JetBlue to continue offering only the same twelve daily flights it already offered until summer 2022, when it would increase to fifteen daily flights); see also PX 0322 at 3 (reflecting that both carriers plan to reduce frequencies in Logan-Reagan, an overlap market, between January 2022 and the time when the NEA reaches its “steady state“). Though the defendants characterize this as a temporary reduction that will be remedied when they have eventually obtained additional aircraft needed to bring service back up to pre-NEA levels, such hope and speculation does not change the fact that the capacity reductions have occurred and (as far as the record is concerned) presently remain.
Fifth, through the NEA‘s reciprocity and codesharing features, frequent flyers and many corporate clients of both defendants have gained broader access to benefits and discounts than they had previously. The travelers for whom these features matter most make up a relatively small proportion of American‘s customers and account for less than half of its revenue. See PX 0009 at 19. Though “power travelers” are the ones American is “most after,” approximately eighty-five percent of its customers are “infrequent” travelers, for
Sixth, because of its partnership with American, JetBlue has increased its operating costs and lost two significant opportunities to expand its own collection of slots and approvals necessary to pursue long-term access and growth domestically and abroad. JetBlue‘s business model—the recipe for its two decades of pre-NEA success and a source of great pride for its executives—depends on its ability to offer high-quality service while keeping costs low. E.g., Trial Tr. vol. 2 at 53-54. The defendants admit, however, that the NEA has increased JetBlue‘s operating costs. Doc. No. 324 ¶ 265; Trial Tr. vol. 1 at 228-29.49 The record also establishes that JetBlue‘s
Similarly, in July 2022, the DOT issued a written determination regarding the reassignment of sixteen “peak-hour runway timings” at Newark. See generally DX 1083. The timings once belonged to United, but had essentially been divested when United merged with Continental in 2010. Id. at 1. The DOT originally awarded them to Southwest, to enhance Southwest‘s ability to provide low-cost competition to the GNCs operating in the New York area. Id. Southwest decided to end its service out of Newark in 2019, and the DOT began a reassignment process. Id. at 2-3. JetBlue, Spirit, and Alaska each applied. Id. at 3. The DOT awarded the timings to Spirit, citing the NEA as a factor that led it not to select JetBlue. Id. at 9-10.52 The Court finds that these effects—the increased costs and the lost opportunities for gaining independent and long-term access to important markets—diminish JetBlue‘s ability to provide disruptive, low-cost competition to the GNCs in the northeast.
Despite departing from the contract‘s requirements so early in the venture, the parties did not elect to amend the MGIA to correct for such scenarios in the future. They also left intact the “Force Majeure” clause in the NEA Agreement itself. That provision—devised by the parties amid the effects of the pandemic—expressly states that “no obligation to make a payment” due under any of the NEA contracts (including the MGIA) “will be excused or limited by virtue of any Force Majeure Event,” a term the parties understandably defined to include a “pandemic.”54 PX 0001-a § 10.9 & App. A. In other words, in agreeing to a lower transfer payment, the defendants disregarded their obligations as defined by both the MGIA and the NEA Agreement itself. This decision not to adhere to their agreement substantially undermines claims by the defendants elsewhere that the terms of their contracts will guarantee certain procompetitive conduct or prevent anticompetitive effects. See, e.g., Doc. No. 324 ¶¶ 188, 197 (noting that the defendants have agreed not to share revenues on certain nonstop overlap routes where their
combined market share poses especially acute competitive concerns, and that the distribution of revenue pursuant
The Court also finds that this same spirit of partnership will diminish competition between the defendants outside the NEA region. See supra note 30 (regarding evidence American competed less vigorously with at least one other joint-venture partner); cf. Trial Tr. vol. 7 at 12-13 (describing the devolution of a more limited collaboration between two other domestic airlines when the spirit of partnership did not prevail). It is neither logical nor possible to view the NEA‘s impact on the defendants’ competitive relationship as confined to the northeast. Both carriers have a strong and vested interest in the NEA‘s success. They intended to form a long-term partnership, and they each have invested substantial time and resources in developing and implementing the partnership.55 All of this materially changes the competitive relationship between American and JetBlue, increasing their mutual interest in the success and survival of the other within the NEA and beyond.
E. Alternatives
The plaintiffs point primarily toward one less-restrictive alternative to the NEA.56 They suggest the defendants could have more closely modeled their partnership after American‘s current relationship with Alaska, the WCIA, and supplemented such a partnership with agreements to lease or transfer slots in New York. See Doc. No. 325 ¶¶ 558-70. Indeed, the defendants devoted time and consideration to whether they should pursue such a partnership instead of the NEA. See, e.g., PX 0268 at 5 (identifying an “East Coast Int‘l Alliance,” structured like the WCIA, as a “Fall Back” scenario in internal presentation pitching what became the NEA). A WCIA-style arrangement would have differed from the NEA in three respects that bear on the competitive impact of the venture. First, it generally would not have included codesharing on domestic O&Ds where the defendants offer competing nonstop service, preserving their status as independent competitors in such markets.57 Second, it would have limited the shared revenue pool to American‘s international
In addition, there was evidence that both defendants have (or have had) more limited collaborations with other domestic carriers, featuring only codesharing and some form of frequent-flyer reciprocity. Though both of these alternatives appear to be less restrictive than the NEA, the plaintiffs have not pursued them in their papers and, following suit, the Court will not (and, given its legal conclusions, need not) further evaluate them.
Though these limitations are features of the WCIA, they have not prevented it from fulfilling that venture‘s central goals—goals that mirror the major purposes of the NEA. In particular, the more limited WCIA still enables Alaska‘s customers and frequent flyers to tap into American‘s international service (on American‘s own network and through its oneworld alliance), while strengthening the feed of passengers to American‘s international flights from the west coast. In doing so, it encourages both WCIA partners to grow and incentivizes them to strive to provide high-quality and seamless customer service. These objectives serve the two partners’ overarching aim: the WCIA enhances their ability to fend off threats from, and compete more fully with, Delta and United.58 Both American and Alaska consider the WCIA a success. There is no evidence before the Court that either WCIA member is dissatisfied with the endeavor, that their incentives are not aligned sufficiently to achieve their mutual and individual purposes, or that consumer benefits would be greater under a different structure than the one they selected.
Though American and JetBlue elected to forego a WCIA-style venture in favor of the broader NEA, projections by American before the NEA was finalized suggest both alternatives offered comparable value and would have spurred comparable growth. See Doc. No. 325 ¶¶ 565-69; PX 0268 at 5; see also Trial Tr. vol. 16 at 125-26 (describing internal American analysis comparing NEA to WCIA-style partnership, showing similar increase in passengers predicted). The Court further finds that a venture modeled after the WCIA would have provoked the defendants’ competitors—in particular, Delta and United—to undertake comparable evaluations, and entertain comparable competitive responses, to those that followed the NEA‘s announcement. Doc. No. 325 ¶ 570.
Neither American nor JetBlue considered—at all, let alone seriously—winding down its operations in the NEA region. That is, with or without the NEA, both carriers intended to continue serving Boston and New York, and both would have done so. For JetBlue, the region contains its two most important focus cities, which account for three-quarters of its overall capacity. See Doc. No. 325 ¶¶ 135, 137-38, 140, 178-79. JetBlue‘s pre-NEA network plans identified goals for growth in the two cities over both the short and long term. See, e.g., id. ¶¶ 211-12, 219, 233, 236 (summarizing
American considered itself “too small to win, but too big to quit” in both New York and Boston. Trial Tr. vol. 7 at 167; accord Trial Tr. vol. 4 at 207. It was the second-largest carrier serving Boston until Delta surpassed it in the year or two before the NEA, and it hoped to win back some of the share it had lost because of its view of Boston as an important market with “a very high proportion of business travelers.” Doc. No. 325 ¶¶ 142-45; PX 0090 at 4. In New York, one of the busiest corporate travel markets in the world, American consistently has been among the three largest carriers for decades, though it admittedly underutilized its resources there (i.e., its substantial slot holdings). Doc. No. 325 ¶¶ 182-83, 189-93, 195. It, too, strategized and discussed internally tactics it could pursue to strengthen its position and grow in the northeast. See, e.g., id. ¶¶ 220-21, 223-24, 230-32 (summarizing evidence of American‘s plan to increase capacity and “fight” back in Boston, and add capacity in the form of more seats and more flights out of JFK and LaGuardia). Those plans included better utilizing American‘s gates at Logan and upgauging its regional jets at JFK and LaGuardia. Id. ¶¶ 221, 231.
There is no evidence that any other carrier was poised to mount a meaningful challenge to the defendants’ positions among the top three and top four carriers in Boston and New York, respectively. In other words, any alternative to the NEA would have included both American and JetBlue continuing to operate and continuing to occupy strong positions among the largest carriers in Boston and New York.60
F. The Experts
Throughout the trial, the Court carefully observed and assessed each witness, considering their credibility generally and as to the specific matters about which they testified. Most witnesses were employed by the defendants. Several hold high-level positions, and many were personally involved in orchestrating the transaction at
The plaintiffs’ primary expert witness was Dr. Nathan H. Miller, a professor and economist who testified about economics and industrial organization.62 In their pretrial filings, the defendants embarked on a quest to strike Dr. Miller‘s analysis and opinions from the record in this case. The Court denied the pretrial motion during the final pretrial conference, Doc. No. 220, and now addresses the renewed objections to Dr. Miller‘s testimony expressed in the defendants’ post-trial papers. Dr. Miller was a thoughtful, credible, and well-credentialed expert witness. He was articulate and methodical in explaining his conclusions and the analysis underlying them, but also candid in acknowledging the limitations of his opinions. More than any other expert testifying in this case, Dr. Miller was consistently measured and precise in his responses, even throughout defense counsel‘s aggressive cross-examination. Dr. Miller expressly accounted for the specific terms of the NEA—that is, he did not ignore the agreements’ terms, as the defendants argue—and explained why he chose the models he used to assess its effects. To the extent the defendants renew their pretrial attack on the admissibility of Dr. Miller‘s testimony, their objections are OVERRULED, and their request to strike his testimony is DENIED.
The defendants’ challenges are more appropriately viewed as directed to the weight the Court should accord Dr. Miller‘s testimony. In that regard, the Court credits Dr. Miller‘s analysis to the extent it suggests the NEA will create upward pricing pressure, a conclusion which is well supported by basic economic principles and incentives, and which confirms the conclusions the Court reaches independently and explains in the discussion of its legal conclusions below. The Court need not, and does not, endorse the specific dollar amount of harm Dr. Miller‘s models predicted (nor did Dr. Miller himself characterize that prediction as a rigid, definite value of literal fare increases). The Court also mentions and credits Dr. Miller‘s testimony and analysis as to other topics where they are material to the Court‘s analysis and specifically noted in this decision.
The apparent bias of the defendants’ retained experts is reason enough to reject the opinions and conclusions they rendered in this case. Certain specific aspects of three defense experts’ testimony further undermine their credibility and merit comment.64 Dr. Darin Lee, an expert on competitive dynamics in the airline industry, expressed opinions that were not soundly reasoned, tailored to this case, or supported by the evidence. Two examples will suffice. First, Dr. Lee “confidently” declared that the NEA is “part of the constant evolution” of an “incredibly competitive” industry where, “clearly,” “the barriers to entry are quite low.” Trial Tr. vol. 12 at 49, 65. This sweeping assertion misleadingly invоkes the language of competition, misunderstands the standards governing federal antitrust analysis, and disregards the specific context in which this case arises. The primary basis Dr. Lee cited to support his assessment that competition is high and barriers are low is the increase in non-GNC options over the last thirty years. But that increase was calculated on a nationwide basis, not with a focus on the unique setting at issue here: the highly constrained northeast. That the ULCC category has grown rapidly says little about the level of competition or barriers to entry and expansion in New York if few such carriers have acquired the tools necessary to operate there.65 Cf. id. at 65 (acknowledging “there are certain places in New York, in particular, . . . where entry is more difficult” but suggesting the alleged size of “the low cost carrier order book” shows “barriers to entry have not . . . constrained [non-GNCs] in any way“).
Similarly, Dr. Lee glibly declared it “completely unplausible” that the NEA would imperil JetBlue‘s disruptive role in the market and cause it to stop “behaving
These are not the only instances in which Dr. Lee conveyed unnuanced and poorly reasoned conclusions. Based on his demeanor and his failure to tailor his opinions to the evidence in this case, the Court rejects both the specific conclusions described above and all other opinions Dr. Lee expressed on topics material to the Court‘s decision.
The defendants’ lead expert, Dr. Mark Israel, suffered from similar problems. In his lengthy testimony, Dr. Israel also demonstrated a misunderstanding and misapplication of antitrust concepts, rendered opinions based on false assumptions, and failed to account for the circumstances presented by the NEA. Again, a few examples make the point. Dr. Israel opined that JetBlue and American—individually and collectively under the NEA—lack market power in Boston and New York. This is so, he says, because market power in the airline industry is linked to hub-carrier status, such that “in no case” can he “imagine a non-hub carrier having the ability to” exercise market power “without ceding share to the hub carrier.” Trial Tr. vol. 13 at 117-18. According to Dr. Israel, market power in Boston lies only with Delta, and in New York it is split between Delta and United, leaving all other carriers asking what they can “do to try to compete with that hub position.” Id.; accord id. at 154. This overly simplistic view is wrong as a matter of fact and law. It ignores American‘s own designation of New York as among its hubs. In other words, American is a hub carrier in New York. If hub status is an indicator of market power for Dr. Israel, then American has market power in New York. Dr. Israel‘s view also would mean JetBlue never could have market power—anywhere—because it is not a GNC with operations it calls “hubs.” This is absurd. JetBlue‘s “focus cities” are its equivalent of hubs. New York and Boston are the two it ranks highest, accounting for nearly three-quarters of its overall operations. They are the centers of its business. Dr. Israel accounts for none of this, adhering to the belief that the NEA lacks market power in the region, essentially because JetBlue is not a GNC (and, presumably, because he either does not realize, or does not believe, that New York is an American hub). The Court emphatically rejects this arbitrary, unfounded view.66
When it comes to Dr. Israel‘s analysis predicting the NEA‘s benefits, his projections are contaminated by his reliance on
For these reasons, and having considered his demeanor and evaluated the basis for all of his testimony, the Court finds Dr. Israel‘s opinions rendered in this case are entitled to no weight.68
The last expert to testify for the defendants was Dr. Dennis Carlton, an economist specializing in industrial organization. His opinions, though expressed more carefully and narrowly than those of his colleagues, were rendered based on a pair of fundamentally flawed assumptions. In assessing whether the NEA had caused fare increases, Dr. Carlton compared fare data from a set of “treatment routes” and a set of “control routes.” Dr. Carlton chose nonstop overlap routes within the NEA (the routes where Dr. Miller predicted the most harm) as the treatment routes, and he used routes within the NEA where JetBlue and American did not offer competing nonstop service as the control routes. Trial Tr. vol. 16 at 19-20. In defining the scope of these two groups, Dr. Carlton made two basic assumptions: 1) that the NEA would have “hardly any [anticompetitive] effect” on the routes in the control group; and 2) that both groups would respond to “industry factors, [including] COVID, in a similar way.” Id. at 20, 64.
The Court finds, based on the overwhelming weight of evidence, that both of Dr. Carlton‘s assumptions are wrong. The
In sum, based on both the content of his testimony and the manner in which he provided it, the Court finds Dr. Miller‘s analysis both credible and helpful. The same considerations lead the Court to reject, entirely, the opinions and conclusions offered by Dr. Lee, Dr. Israel, and Dr. Cаrlton.
G. The Litigation69
Though the DOT ended its informal review of the NEA based on the commitments made by the defendants, the Department of Justice (“DOJ“) undertook its own evaluation of the NEA‘s legality and its impacts on competition. Cf. PX 0452 at 1 (reflecting DOT‘s statement that it would “defer to DOJ” regarding any antitrust concerns arising from the NEA). The defendants had anticipated and prepared for such scrutiny, enlisting the guidance of expert airline economists and antitrust lawyers before the deal was signed and publicized. Doc. No. 325 ¶¶ 93, 95. Ultimately, the DOJ concluded that the defendants’ commitments to the DOT, amendments to the NEA Agreement, and amendment to the MGIA did not eliminate the competitive harms the DOJ anticipated would arise from the defendants’ collaboration. And so, this lawsuit followed.
The United States and the other plaintiffs sued the defendants on September 21, 2021, alleging the NEA violates Section 1 of the Sherman Act. Doc. No. 1. Two months later, the defendants moved to dismiss the complaint. Doc. No. 67. Meanwhile, the parties asked the Court to enter a scheduling and case management order without awaiting resolution of the defendants’ motion, so that they could begin discovery and the Court could calendar their agreed-upon September 2022 trial date. Doc. No. 65. The jointly proposed case-management regime (which the Court accepted) included deadlines for dozens of events, from discovery to motions in limine, but it did not propose a date for dispositive, post-discovery motions. See generally Doc. No. 76. Discovery proceeded, as did some measure of motion practice
The Court subsequently set a trial schedule, Doc. No. 131, and resolved various requests regarding the treatment of confidential materials in pre-trial submissions and at trial, e.g., Doc. Nos. 138, 139. The parties filed, and the Court ruled on, several motions in limine, Doc. Nos. 140, 141, 144, and the Court resolved various disputes pertaining to other aspects of the trial procedure, Doc. No. 221. The defendants sought to compel two Delta executives to appear in person as trial witnesses, Doc. No. 175, the executives responded by seeking to quash the trial subpoenas, Doc. No. 230, and the Court resolved that pair of motions, Doc. No. 248.
Trial commenced on September 27, 2022. Doc. No. 252. At the outset, both parties offered into evidence exhibits to which there was no objection, handing up lists of the items within that category. Trial Tr. vol. 1 at 6-7. Those lists, combined, identified and deposited into the record in a matter of seconds—more than 1,400 exhibits. Many other exhibits were subject to objections, some of which fell into a few categories of objections that the parties identified in pretrial submissions, Doc. No. 157 at 22; Doc. Nos. 157-5, 157-6, and the Court ruled on those objections as the exhibits were offered during the trial. When the plaintiffs were nearing the end of their case-in-chief, they moved for the admission of about fifty more exhibits over defense objections that mirrored objections the Court had overruled during the course of the proceedings as to other similar exhibits. Doc. No. 266. After considering the parties’ positions, the Court allowed that motion and admitted the underlying batch of exhibits. Doc. No. 283. Along with its ruling, the Court admonished the parties that, given the volume of exhibits (and the substantial length of many of them), the Court would not “plac[e] any weight on an exhibit unless it has been explored with a witness at trial and/or meaningfully described or relied upon in any party‘s post-trial submission.” Id. at 2. No party objected to or otherwise expressed concerns about that admonition, and the Court has adhered to it. Ultimately, the parties cited at least four hundred exhibits in their post-trial papers.
Twenty-four witnesses appeared and testified during the trial. The list included eight American executives, six JetBlue executives, and one executive now employed by American after working for many years at JetBlue (who, in fact, oversaw the NEA negotiations on behalf of JetBlue before departing to join American). Executives from Southwest, Spirit, and Alaska also testified. The remaining witnesses were experts—economists with specialized knowledge about antitrust issues, the airline industry, or both. Two experts testified for the plaintiffs, and four for the defendants. The parties were unablе to complete their presentation of these two dozen witnesses in the time originally allotted for this trial, leading the plaintiffs to request additional time. Trial Tr. vol. 12 at 5-7. The Court allowed the request, over the defendants’ objection, adding more than seven hours of trial time to the schedule. Id. at 87-88. Witness testimony concluded on October 27, 2022. See Doc. No. 290.
From the start, the parties intended to supplement the live witness testimony with excerpts from depositions of a number of additional witnesses. Even with the added trial time and despite a Court order that would have compelled one Delta executive to appear in person to testify (a result fervently pursued by the defendants), Doc.
The Court scheduled closing arguments for November 18, 2022. Doc. No. 296. Pursuant to the case management order, the parties’ post-trial submissions were due three weeks after the trial ended—or, as it turned out, the day before the scheduled closings. Doc. No. 76 at 4; Doc. No. 196 at 9. The parties timely provided the Court with sealed copies of their post-trial papers, followed by redacted versions filed publicly on the docket. Doc. Nos. 320, 322, 323, 324, 325, 326. The post-trial submissions were voluminous, exceeding six hundred pages in all. See Trial Tr. vol. 18 at 4 (reflecting the Court‘s “confess[ion] that [it] wasn‘t capable of reading all 500 pages“—a sizeable underestimate—“between last night and this morning“). On November 29, 2022, the parties provided to the Court hyperlinked versions of their post-trial submissions—accompanied by more than 1,500 files consuming approximately twelve gigabytes of disc space—facilitating more efficient review and cross-referencing of the exhibits, testimony, and authorities cited throughout the papers. Cf. Doc. No. 196 at 9 (requiring hyperlinked versions a week after the original submissions).
Shortly thereafter—that is, after the testimony had concluded and the post-trial submissions were received, and while the Court had this matter under advisement—the first in a series of lawsuits was filed by an individual seeking to represent a proposed class of consumers alleging antitrust injuries arising from the NEA. Compl., Berger v. JetBlue Airways Corp., No. 22-cv-7374 (E.D.N.Y. Dec. 5, 2022); see also Compl., Kupferberg v. Am. Airlines Grp. Inc., No. 23-cv-1220 (E.D.N.Y. Feb. 15, 2023); Compl., Buehler v. JetBlue Airways Corp., No. 23-cv-10281 (D. Mass. Feb. 2, 2023); Compl., Guerin v. JetBlue Airways Corp., No. 22-cv-7423 (E.D.N.Y. Dec. 7, 2022). Those lawsuits remain pending.71
One final event bears mention. In July 2022, JetBlue announced it had reached an agreement to acquire Spirit. See Compl. ¶ 29, United States v. JetBlue Airways Corp., No. 23-cv-10511 (D. Mass. Mar. 7, 2023). That proposed acquisition, referenced at times during the trial in this case, was also subject to regulatory review, a process which recently yielded a separate antitrust suit by the United States seeking to prevent the merger based on alleged violations of the Clayton Act. See generally id. ¶¶ 1-86. Another session of this Court has scheduled trial in that matter to begin in October 2023.
III. CONCLUSIONS OF LAW
A. The Sherman Act72
Section 1 of the Sherman Act prohibits the formation of any “contract . . . in restraint of trade or commerce among the several States.”
In decisions spanning a century, the Supreme Court has characterized the “true test of legality” under the Sherman Act as “whether the restraint imposed is such as merely regulates and perhaps thereby promotes competition or whether it is such as may suppress or even destroy competition.” Bd. of Trade of Chi. v. United States, 246 U.S. 231, 238 (1918); accord Ohio v. Am. Express Co., 138 S. Ct. 2274, 2284 (2018) [hereinafter Amex]; see Nat‘l Collegiate Athletic Ass‘n v. Bd. of Regents of Univ. of Okla., 468 U.S. 85, 104 (1984) [hereinafter Bd. of Regents] (“Under the Sherman Act the criterion to be used in judging the validity of a restraint on trade is its impact on competition.“); see also Nat‘l Soc‘y of Prof‘l Eng‘rs v. United States, 435 U.S. 679, 692 (1978) (describing “the purpose of the analysis” as formation
The Sherman Act distinguishes between concerted action by separate entities and independent action by a single firm. Copperweld Corp. v. Indep. Tube Corp., 467 U.S. 752, 767-68 (1984). A single firm faces liability only if it engages in conduct that “threatens actual monopolization.” Id. at 767. Otherwise, “an efficient firm” may act vigorously to “capture unsatisfied customers from an inefficient rival, whose own ability to compete may suffer as a result,” because such “competitive zeal” by a single actor “is precisely the sort of competition that promotes the consumer interests that the Sherman Act aims to foster.” Id. at 767-68. Where concerted activity is involved, however, it is “judged more sternly,” and “it is not necessary to prove [a threat of] monopolization.” Id. at 768. The reason for this distinction between unilateral and concerted behavior is “readily appreciated“:
Concerted activity inherently is fraught with anticompetitive risk. It deprives the marketplace of the independent centers of decisionmaking that competition assumes and demands. In any conspiracy, two or more entities that previously pursued their own interests separately are combining to act as one for their common benefit. This not only reduces the diverse directions in which economic power is aimed but suddenly increases the economic power moving in one particular direction. Of course, such mergings of resources may well lead to efficiencies that benefit consumers, but their anticompetitive potential is sufficient to warrant scrutiny even in the absence of incipient monopoly.
Horizontal restraints—that is, agreements among “actual or potential rivals” that “eliminate[] some avenue of rivalry among them“—“have traditionally received antitrust‘s highest level of scrutiny.” 11 Phillip E. Areeda & Herbert Hovenkamp, Antitrust Law: An Analysis of Antitrust Principles & Their Application ¶ 1900, 1901b (4th ed. 2018) [hereinafter, Antitrust Law].
Areeda on Antitrust].73 Such restraints “as a class provoke harder looks than any other arrangement“—harder, even, than mergers—because they generally pose a greater “threat of a market output reduction” than do other classes of restraints. Id. ¶ 1902a. One reason for this is that “the costs of forming a horizontal agreement are typically less than the costs of a merger involving the same firms.” Id. ¶ 1902c. As a result, competing firms might be tempted to pursue a horizontal
Generally speaking, a plaintiff alleging a violation of Section 1 must prove the existence of (1) an agreement between or among separate entities that (2) unreasonably restrains trade and (3) impacts interstate commerce. Here, it is beyond dispute that the NEA constitutes an agreement between two separate entities (American and JetBlue), and that it impacts interstate commerce (air travel from one state to another). The focus of this litigation and the Court‘s analysis, then, is on whether the NEA amounts to an “unreasonable” restraint on trade.
B. Mode of Analysis
“In evaluating . . . claims under the Sherman Act, one of the first considerations a court faces is determining the appropriate framework for its review . . . .” Am. Steel Erectors v. Local Union No. 7, Int‘l Ass‘n of Bridge, Structural, Ornamental & Reinforcing Iron Workers, 815 F.3d 43, 60 (1st Cir. 2016). The parties express a surface-level agreement that the rule of reason should provide the framework for the Court‘s analysis of the NEA. E.g., Doc. No. 322 at 19; Doc. No. 326 at 20. They part ways almost immediately, though, when it comes to defining the contours of that framework. Compare Doc. No. 320 ¶¶ 25, 132 (explaining the plaintiffs’ view that the first step can be satisfied by proof of likely anticompetitive effects or harm to the competitive process itself, and if the analysis proceeds beyond identifying harms and benefits, the final step requires a weighing of such effects whether or not less restrictive alternatives exist), with Doc. No. 323 ¶¶ 10, 118 (describing the defendants’ belief that the first step requires proof that the restraint already has caused identifiable anticompetitive harm, and that weighing harms and benefits, if ever appropriate, would occur only if a less restrictive alternative is proven). The Court finds that the plaintiffs’ recitation of these legal standards is the correct one, and it will proceed to outline the mode of analysis required by those standards here.74
When tasked with determining whether a restraint is “unreasonable,” courts apply a level of scrutiny that varies depending on the nature of the restraint. See Cal. Dental Ass‘n v. Fed. Trade Comm‘n, 526 U.S. 756, 780 (1999) (invoking Areeda on Antitrust and endorsing the view that the amount and quality of proof required depends on the circumstances); accord Alston, 141 S. Ct. at 2160. Some agreements, including those fixing prices or allocating markets, “are thought so inherently anticompetitive that each is illegal per se without inquiry into the harm it has actually caused.” Copperweld, 467 U.S. at 768; accord Palmer v. BRG of Ga., Inc., 498 U.S. 46, 49-50 (1990) (per curiam); see Stop & Shop Supermarket Co. v. Blue Cross & Blue Shield of R.I., 373 F.3d 57, 61 (1st Cir. 2004). Others, including many mergers and joint ventures, “are judged under
Restraints arising in the context of joint ventures ordinarily are subject to the rule of reason, which involves some form of burden shifting but is not a rigid framework. See Alston, 141 S. Ct. at 2160 (admonishing that the rule of reason‘s steps “do not represent a rote checklist, nor may they be employed as an inflexible substitute for careful analysis“). The level of scrutiny and the magnitude of each party‘s burden depends on “the circumstances, details, and logic of a restraint,” yielding “an enquiry meet for the case.” Cal. Dental, 526 U.S. at 781. That said, the typical stages of the inquiry can be summarized as follows.
First, the plaintiff must make an initial showing that the challenged agreement “has a substantial anticompetitive effect.” Amex, 138 S. Ct. at 2284. This showing can be made with direct proof of actual harm to the competitive process—including, though plainly not limited to, evidence that price has increased—or by indirect proof that such harm is likely to arise from the restraint. See Doc. No. 320 ¶ 26 (collecting cases); cf. Addamax Corp. v. Open Software Foundation, Inc., 152 F.3d 48, 53 (1st Cir. 1998) (allowing that “a sufficiently high risk of an anticompetitive effect, coupled with marginal benefits . . . might justify condemnation under the rule of reason“). As explained further below, though this step can require an evaluation of market power, it need not always involve such an assessment.
If the рlaintiff succeeds, the burden shifts to the defendant “to show a procompetitive rationale for the restraint.” Amex, 138 S. Ct. at 2284. The measure of proof required of the defendant depends on the strength of the plaintiff‘s initial showing; that is, the more significant the anticompetitive effects, the heavier the defendant‘s burden to justify the restraints with evidence of procompetitive benefits. See Bd. of Regents, 468 U.S. at 113. This step of the analysis “does not open the field of antitrust inquiry to any argument in favor of a challenged restraint that may fall within the realm of reason“; rather, the defendant must remain “focuse[d] directly on the challenged restraint‘s impact on competitive conditions” as it seeks to justify the restraint. Nat‘l Soc‘y of Prof‘l Eng‘rs, 435 U.S. at 688.
Should the defendant satisfy its obligation, the ultimate burden returns to the plaintiff. A plaintiff can prevail at this point with proof that “the procompetitive efficiencies could be reasonably achieved through less anticompetitive means.” Amex, 138 S. Ct. at 2284. Absent such proof, the plaintiff may alternatively seek to establish that, on balance, the restraint‘s
As noted above, not every case within the reach of the rule of reason “is a candidate for plenary market examination.” Cal. Dental, 526 U.S. at 779; accord United States v. Apple, Inc., 791 F.3d 290, 329 (2d Cir. 2015) (Livingston, J.). If an agreement “does not quite fit the bill of per se liability,” but bears some hallmarks of such a restraint (e.g., apparent market division among the parties), its “anticompetitive effect on customers and markets” might be so apparent at first glance that a court may appropriately shift the burden promptly to the defendant “to show empirical evidence of procompetitive effects.” Am. Steel Erectors, 815 F.3d at 61 (quoting Cal. Dental, 526 U.S. at 775 n.12). That is, where “an observer with even a rudimentary understanding of economics could conclude that the arrangements in question would have an anticompetitive effect,” a more abbreviated analysis suffices.77 Cal. Dental, 526 U.S. at 770; cf. Nat‘l Soc‘y of Prof‘l Eng‘rs, 435 U.S. at 692 (noting some restraints “are so plainly anticompetitive that no elaborate study of the industry is needed to establish their illegality“). For example, a “naked restriction on price or output,” even in the context of an otherwise legitimate joint venture, “requires some competitive justification even in the absence of a detailed market analysis.” Bd. of Regents, 468 U.S. at 109-10.
A scenario hypothesized by Professor Areeda and invoked by the Supreme Court provides a helpful illustration of such a restraint, neither “categorically unlawful in all or most of its manifestations” nor “universally lawful“:
[J]oint buying or selling arrangements are not unlawful per se, but a court would not hesitate in enjoining a domestic selling arrangement by which, say, Ford and General Motors distributed their automobiles nationally through a single selling agent. Even without a trial, the judge will know that these two large firms are major factors in the automobile market, that such joint selling would eliminate important price competition between them, that they are quite substantial enough to distribute their products independently, and that one can hardly imagine a pro-competitive justification actually probable in fact or strong enough in principle to make this particular joint selling arrangement “reasonable” . . . .
Phillip Areeda, The “Rule of Reason” in Antitrust Analysis: General Issues, 37-38 (Fed. Jud. Ctr. 1981) [hereinafter Areeda Monograph]; accord Bd. of Regents, 468 U.S. at 109 n.39; 7 Areeda on Antitrust ¶ 1508. In such circumstances, it is possible to perform the rule of reason and balance the relevant considerations “in the twinkling of an eye.” Areeda Monograph at 38.
As the following sections explain, the parties’ presentation of this case places it within the realm of the rule of reason. Given the dearth of instances in which other courts have confronted an arrangement like this involving domestic airlines, that placement is almost certainly appropriate. That said, after a deep and searching review of the voluminous record in this case, the findings of fact set forth above compel the Court to conclude that the NEA is situated “at [one] end[] of the competitive spectrum,” Alston, 141 S. Ct. at 2155, and that no deep and searching analysis is required in order to discern its unlawfulness. See Doc. No. 320 ¶¶ 28, 30-35 (identifying cases in which some form of abbreviated review was deemed appropriate where horizontal restraint at issue involved obviously anticompetitive features).
C. Substantial Anticompetitive Harm
“Every antitrust suit should begin by identifying the ways in which a challenged restraint might possibly impair competition.” 7 Areeda on Antitrust ¶ 1503a (4th ed. 2017). Here, notwithstanding the defendants’ vigorous claims otherwise, the harms are considerable and obvious.
1. Direct Evidence of Actual Harm
To start, the plaintiffs have proven that the NEA already has caused actual and substantial harm to competition. The defendants have expended many pages in an effort to resist the evidence showing the NEA has already produced anticompetitive effects.78 E.g., Doc. No. 322 at 20-48. Nevertheless,
First, the NEA has eliminated the once vigorous competition between two of the four largest domestic carriers in the northeast, replacing it with broad cooperation in pursuit of the shared interests of their partnership. American and JetBlue no longer operate in the northeast as two distinct carriers engaged in direct competition with one another. Rather, they work together as one combined carrier with a single “optimized” network. As JetBlue‘s CEO conceded on the first day of trial, American and JetBlue no longer “compete with each other directly” within the NEA region; they collaborate and make capacity (i.e., output) decisions “as a single airline” would. Trial Tr. vol. 1 at 182.
The two carriers pool their resources and decide together what routes to fly within the NEA and on what schedule. As to each O&D, they decide together which partner should operate the route—a decision which determines what aircraft will be used and, thus, what capacity (and what amenities) will be available. In other words, they coordinate in a way that impacts their output within the NEA. If both carriers will operate the same route, they do so on an “optimized” schedule they jointly design to minimize (and, ideally, eliminate) instances in which the two offer flights departing at or around the same time. This optimized NEA schedule emulates that of a single carrier. Though American and JetBlue do not discuss the fares they will set, they otherwise strive to provide a seamless product by operating, as much as they can, as a single airline would. Put simply, the NEA has replaced direct and aggressive competition between American and JetBlue on nearly every front (including routes, schedules, and capacity) with cooperation. This, in and of itself, is a fundamental assault on competition and an actual harm the Sherman Act is designed to prevent. See Impax Labs., Inc. v. Fed. Trade Comm‘n, 994 F.3d 484, 493 (5th Cir. 2021) (“Eliminating potential competition is, by definition, anticompetitive.“); Doc. No. 320 ¶ 36 (citing additional cases).
The combination of two powerful competitors within the northeast effectively reduces the number of market participants—and the number of distinct choices for consumers—by one. Though various executives claimed American and JetBlue continue to compete in some fashion within the northeast, there is no credible support for those claims. There is no evidence the defendants have engaged in meaningful competition with respect to fares, schedules, service, advertising, or anything else within the NEA region (or even on the carve-out routes touching an NEA airport) since implementation of the partnership began. The absence of such evidence makes perfect sense, as the NEA is designed to render American and JetBlue agnostic as to which of them a customer selects, so that they will work together in pursuit of their shared, rather than individual, interests. The executives’ testimony, at best, describes a philosophical hope or preference—and certainly a sincere one—that customers will choose their carrier. Such hopes, however, mean little in the face of overwhelming evidence that the NEA is meant to align the defendants’ incentives and render them “metal neutral,” and that it has done so. A hallmark of a free market is the incentive to fight for revenue and customers against one‘s direct competitors. As between American and JetBlue, that incentive is eliminated by the NEA.
Conditions such as these make a horizontal collaboration like the NEA especially tempting; the competitive landscape is challenging, and the possibility of increased profits without the full cost of a merger is appealing, as is the prospect of operating with one powerful rival removed from the mix. See 7 Areeda on Antitrust ¶ 1902c. The NEA not only pools for shared use valuable resources that the defendants once used to directly compete with one another, it also limits the ability of either partner to transfer slots to competing carriers, fortifying the barriers to outside competition. In this way, the NEA “deprives the marketplace of the independent centers of decisionmaking that competition assumes and demands,” and “suddenly increases the economic power moving in one particular direction” in Boston and New York. Copperweld, 467 U.S. at 768-69. In constrained and concentrated markets like those within the NEA region, the reduction in competition and choice resulting from the defendants’ cooperation is a real and substantial harm.80
Second, and relatedly, by aligning its interests with a powerful GNC, JetBlue has sacrificed a degree of its independence and weakened its status as an important “maverick” competitor in the industry.81
These effects are neither theoretical nor predictive. Already, JetBlue has twice lost out on valuable assets because regulators have found (and this Court agrees) the NEA entangles JetBlue with American in a way that diminishes its status as an independent low-cost player in the market. Regulators in the U.K. found JetBlue no longer eligible for highly valuable, long-term remedy slots that were available at Heathrow, and the FAA similarly awarded Spirit (rather than JetBlue) a set of runway timings at Newark earmarked for a carrier that would provide low-cost competition against United. These unfavorable decisions were expressly tied to the NEA, and they were rendered despite JetBlue having devoted time and advocacy to the pursuit of both assets. Thus, the Court concludes that the NEA has caused JetBlue to lose opportunities to independently secure greater access to busy, constrained markets where its “maverick” competition would otherwise operate as an important competitive check on the conduct of larger carriers.
Though JetBlue has secured some access to Heathrow and a second London-area airport through other means, such access does not offset the harm flowing from the NEA. For one thing, the access JetBlue has obtained is less desirable in terms of number of slots, cost, location, and duration compared to the CMA slots it had sought. JetBlue‘s own advocacy powerfully articulates why access to London at Heathrow is preferable to the other London-area airports: it provides the strongest platform from which to challenge GNCs and their international alliance partners and thereby both gain market share and impact transatlantic fares. E.g., Doc. No. 325 ¶¶ 244-45 (citing such evidence). The CMA slots were more numerous, would have been awarded at no cost to JetBlue, and carried a defined and longer duration than the slots JetBlue is leasing now.83
Furthermore, JetBlue‘s motivation to compete aggressively with American‘s transatlantic fares is inevitably skewed by the MGIA. Though JetBlue‘s flights to
Third, though the plaintiffs have pursued a rule-of-reason challenge, at least one of the NEA‘s core features closely resembles a restraint that is per se illegal: the defendants’ assignment of various routes to either American or JetBlue as part of their “optimization” of the combined NEA network. In New York‘s slot-constrained airports alone, American has exited more than a dozen routes that both defendants served before the NEA. This includes the high-traffic Boston-to-LaGuardia route—a market in which Delta is now the only carrier besides JetBlue providing nonstop service. Moreover, the evidence suggests that the defendants’ ultimate objective, when the NEA is fully implemented, is to continue identifying which carrier should fly which routes—with one carrier per market wherever possible. Cf. Trial Tr. vol. 9 at 10 (“I think one of the things with the NEA is we‘re always looking to figure out which airline would be best to serve a certain route.“).
This is a straightforward example of market allocation.85 The Supreme Court has held, on more than one occasion, that “an agreement between competitors at the same level of the85 market structure to allocate territories in order to minimize competition” is among “the classic examples of a per se violation” of the Sherman Act. Topco, 405 U.S. at 608; accord Palmer, 498 U.S. at 49-50. Here, American and JetBlue are horizontal competitors who have agreed to make network decisions together within the northeast. In doing so, they decide which routes to serve, and which partner should operate the planes serving them. Already, that process has led to the allocation of markets to one or the other defendant, either by one partner exiting a market both previously served or by one partner tabling plans to enter a market already served by the other.86 This
In sum, the NEA has materially altered the competitive landscape in a highly concentrated industry, and in a region with significant barriers to entry. It has accomplished this in at least three ways. It has reduced the number of competitors (and, thus, choices) by one in a setting where such a reduction is especially harmful. It has reduced JetBlue‘s independence and undermined its status as an important “maverick” competitor. And, it has allowed the defendants to engage in horizontal market division, a practice historically and consistently invalidated as a matter of antitrust law. Each of these three actual effects already hаve resulted from the NEA and, considered together, they amount to a powerful showing of serious anticompetitive harm.87
2. Indirect Evidence of Actual and Likely Harm
Beyond the actual competitive harms already discussed, the plaintiffs also adduced evidence to alternatively satisfy the first step of a fuller rule-of-reason analysis by showing indirectly that the NEA threatens substantial and imminent harm to competition. Though the above finding of actual harm is sufficient to shift to the defendants the burden of showing procompetitive benefits, the Court briefly addresses the plaintiffs’ alternative showing for the sake of completeness.
To establish an antitrust violation based on indirect evidence, the plaintiffs must identify the relevant market, offer proof that the defendants have power in that market, and supply “some evidence that the challenged restraint harms competition.” Amex, 138 S. Ct. at 2284. Here the parties agree, and the Court finds, that the relevant product market is “scheduled air passenger service,” and the relevant geographic markets are O&Ds “in which Defendants compete or would likely compete absent the NEA.” Doc. No. 320 ¶¶ 44; Doc. No. 323 ¶¶ 33-34; see Doc. No. 320 ¶¶ 40-44 (outlining legal standards governing market definition). Here, the geographic markets at issue are those O&Ds that include Logan as an endpoint, and those that include New York as an endpoint. The parties vehemently disagree regarding the definition and scope of the “New York” endpoint. The plaintiffs point to evidence, including analysis by Dr. Miller, that the geographic market should be limited to those markets beginning or ending at JFK or LaGuardia only. See Doc. No. 325 ¶¶ 111-34 (detailing the evidentiary basis for the plaintiffs’ geographic market
Ultimately, this dispute amounts to much ado about nothing. For one thing, the tests used by economists in the antitrust context are neither intended nor required to identify a single relevant market. See Doc. No. 320 ¶ 60 (explaining that there may be multiple relevant markets or sub-markets, any of which might appropriately be examined to assess the competitive effects of a restraint). In other words, both sides might be right, and appropriate geographic markets might exist with New York defined both ways. The Court need not choose between the two for purposes of this decision. Besides the fact that this issue has no bearing on the finding that the plaintiffs have produced direct evidence of actual harm, the record suggests that neither the defendants’ market power nor the indirect proof of harm here depends on whether Newark is considered within a relevant geographic market.
However the geographic market is defined with regard to New York, the Court finds that the defendants plainly have the power to influence prices. This conclusion is overwhelmingly supported by evidence ranging from various exhibits expressing the relative market shares of the defendants and other carriers operating in Boston and New York, to business records showing that strategic moves made by each defendant individually before the NEA (e.g., changes in fares) unsurprisingly elicited competitive responses from the other major participants in those markets (i.e., proof that the defendants have, in fact, exerted control over prices in the past). See, e.g., PX 0885 at 4-5 (reflecting estimates of market share based on 2019 seat capacity in all three New York airports, with the defendants’ combined shares exceeding thirty percent at JFK and LaGuardia, and totaling twenty-five percent in the region with Newark included); see also United States v. Visa U.S.A., Inc., 344 F.3d 229, 239-40 (2d Cir. 2003) (sustaining finding of market power based on “evidence of specific conduct undertaken” demonstrating “the power to affect price” and, alternatively, based on “large shares” such as twenty-six percent in “a highly concentrated market“). The Court rejects as patently incredible the defendants’ assertions to the contrary.88 E.g., Doc. No. 323 ¶¶ 57, 61.
One more point is worth noting. The antitrust implications of the NEA are, and always were, apparent and understood by the defendants. Though this partnership is in some ways unprecedented, alliances among airlines that otherwise would compete are not. As their papers repeatedly concede, international partnerships like those upon which this one is based routinely provoke scrutiny by regulators and require the conferral of antitrust immunity before they may proceed. That is because, absent such immunity, collaboration among competitors in this manner—sharing profits or revenues and coordinating schedules and output—violates antitrust law. Such collusion is unlawful whether or not the participants believe it is beneficial to themselves or the public. If regulators agree that the benefits outweigh the harms, then those regulators immunize the otherwise unlawful cooperation. They do not decide it was lawful in the first place.
Numerous executives representing different airlines said as much at trial, evincing industry-wide awareness about the legal limits on their ability to even discuss certain topics with their competitors. The defendants also were concededly aware of this, as it is the reason they consulted with antitrust lawyers and consultants and formed a “Clean Team” when devising the relationship. The fact that antitrust immunity sometimes is bestowed upon such alliances by regulators (often with concessions, such as the divestment of slots or other resources to entities the regulators believe will provide a competitive check on the alliance) has little bearing on the questions before the Court. The Court is not a regulator. It is not empowered to excuse a pact that contravenes
D. No Legitimate Justification or Cognizable Benefits
Given the strength of the plaintiffs’ showing of anticompetitive harms, American and JetBlue are obligated to produce substantial, credible, and empirical evidence establishing the procompetitive benefits they claim arise from the NEA.91 Doc. No. 320 ¶ 104. They have not done so. The Court will address each benefit the defendants have identified, explaining why none are sufficient—independently or collectively—to overcome the plaintiffs’ proof of the NEA‘s illegality.92 See Apple, Inc., 791 F.3d at 330 (finding that agreement‘s “anticompetitive effects are easily ascertained,” and promptly “shifting the inquiry directly to a consideration of the defendant‘s procompetitive justifications“).
The defendants place great emphasis on “the underlying purpose of the NEA,” which they characterize as a “procompetitive” effort to “to efficiently pool some of JetBlue and American‘s resources and maximize customer value.” Doc. No. 323 ¶ 105. They even peppered the introductory clauses in the NEA Agreement with self-serving declarations of their good intentions. E.g., PX 0001-a, “Recitals” ¶ 1 (claiming “aim [is] to deliver significant customer benefits” and “to enhance the experience of passengers” in the region). This, however, is simply not a fair characterization of the NEA‘s “underlying purpose,” based on the weight of the evidence presented at trial. It is abundantly clear to the Court that the defendants’ primary motivation in establishing the NEA was to strengthen their own сompetitive positions against Delta (and, to a lesser extent, United) in Boston and New York. Their own witnesses, business records, and submissions to the Court have repeatedly described this purpose. E.g., supra note 21 (citing multiple instances of testimony describing purpose as competing more effectively against Delta); DX 0037 at 2 (reflecting American‘s description of a potential collaboration with JetBlue as intended to “[a]ddress AA/B6 ‘incomplete’ customer proposition relative to DL/UA in NYC” (emphasis added));93 Doc. No. 322 at 51 (describing NEA as a “solution” allowing American to “become a stronger competitor against Delta and United,” and as “a way” for JetBlue “to answer the competitive threat from Delta in Boston“).
The problem for the defendants is that this purpose—strengthening their own position against one or two rivals—is not a valid justification, and cannot render an unreasonable restraint
on trade reasonable, under the
Thus, the central thrust of the NEA—to strengthen the defendants with respect to Delta and United—is anticompetitive for purposes of the
American and JetBlue also endeavor to fit the NEA within the realm of productive or otherwise legitimate joint ventures with restraints that are ancillary to their larger, permissible purposes. They invite the Court to approach the NEA the way other courts have approached various sports associations, or collaborations by two competitors with different skills or resources to produce a new product. See Doc. No. 322 at 51-52 (describing NEA as “a classic pooling of complementary assets that . . . results in an attractive, saleable new network“). This is an effort to force a square peg into a round hole.
The NEA is not a venture with an overarching legitimate purpose under the
Similarly, the defendants have not established their pooled assets are “complementary,” cf. Doc. No. 325 ¶ 444 (citing evidence that, in 2019, JetBlue served only four airports American did not already serve), such that they enable the defendants to create an innovative product. They are not, for example, pooling resources to engage in research they could not independently fund, with the aim of developing a new, more efficient way to train pilots or service aircraft. They are not combining capital to fund the renovation and expansion of a terminal at an airport in the northeast—a project neither might undertake independently, and which will result in an increase in both output (more daily flights from the expanded terminal) and customer service (better facilities
But, of course, the NEA does none of these things. Its anticompetitive features are at its core, ancillary to no overarching legitimate objective. The defendants cannot evade scrutiny of their deliberate decision to eliminate competition between them by calling it a “joint venture” and pointing out that other joint ventures have often produced efficiencies. The Court is concerned only with the NEA, the purpose and effects of which are to suppress, rather than enhance, competition.98
Beyond the venture‘s primary objective and essential character, the defendants identify certain benefits or efficiencies they claim the NEA has produced. None are sufficient to satisfy the defendants’ burden in the face of the strength of the plaintiffs’ showing of anticompetitive harms. Some benefits the defendants claim amount to variations on the theme that they can be stronger in the northeast if they can combine, rather than compete with, their resources.99 As explained already, that theory is not “procompetitive.” Examples of this type of claimed benefit include: the creation of a larger and “seamless” NEA network with “better connectivity” by adding together and “optimizing” the defendants’ separate networks; the creation of “better schedules” by allocating various routes to only one or the other defendant and spreading out their respective flights on shared routes such that they no longer compete head-to-head; and the defendants’ perception that they can compete better for corporate clients if they rely on their combined network (and, potentially, seek joint contracts) instead of having to compete separately. These features arise only if the defendants mimic one carrier, elect not to compete with one another, and cooperate in ways that horizontal competitors normally wоuld not. This elimination of competition negatively impacts the number and diversity of choices available to consumers in the northeast. As such, “benefits” arising in
Other claimed benefits lack evidentiary support entirely or find support only if an artificially narrow lens is applied. For example, as the Court‘s findings of facts explain, see supra note 44, the defendants have claimed, but not proven, that their fleets have actually grown as a result of the NEA. Likewise, they have claimed, but not proven, that corporate customers in the northeast previously viewed Delta and United as their only choices, and that such customers are now available to the defendants only because of the NEA.100 Though they insist the NEA has provoked competitive responses from Delta and United (the only “competitors” about whom the defendants are concerned), their actual evidence of such responses is milquetoast, at best.101 Their claims that the NEA has led to growth and increases in capacity rely on evidence that examines only the northeast.102 They also disregard evidence that capacity growth within the NEA comes at the expense of resources and output by the defendants elsewhere,103 as well as evidence the defendants each would have pursued at least some of this growth with or without the partnership.104 See, e.g., Doc. No. 323 ¶ 110 (acknowledging that “funding” NEA growth required a “small service reduction elsewhere and a modest number of redeployed aircraft” by American, pending the acquisition of new incremental aircraft at some point in the future); Doc. No. 325 ¶¶ 464-70, 497 (describing JetBlue‘s fleet constraints, its reallocation of resources to the
To be sure, the defendants have identified new service they have launched from the northeast since the NEA was implemented. See DX 1087 (listing in two slides markets the defendants served from NEA airports in late 2022 that they had not served in 2019 or 2020). But even this is not supported strongly and unambiguously by reliable evidence. For example, the chart the defendants most often cite as proof of these routes does not itself characterize them as having been launched “because of” the NEA; rather, it lists new services that began “since” the NEA. Id. It says nothing about whether either defendant served the routes before 2019, whether the routes were included in long-term plans the defendants independently developed and would have pursued without the NEA, or whether they arose from the substantial shift in flying patterns occurring during and after the pandemic.106 It also does not identify the sources of the planes allocated to these new markets. As the Court already has explained, the defendants have not shown they presently possess new aircraft with which they can fund such expansion, so the only possible conclusion is that the new flying is funded by planes pulled from other locations.107 Testimony of witnesses confirms that at least a few of these new routes—American‘s long-haul service to Colombia—were not profitable and already have been cancelled.108 No objective or helpful corroboration is provided by citations to the defendants’ own internal slide decks pitching the benefits or success of the NEA without providing reliable sources or support for the claims contained therein. E.g., DX 0055 at 3 (pitching NEA in a November 2021 “Customer Announcement” sales
Because the anticompetitive effects of the NEA are clear, it is the defendants’ burden—and a heavy one—to justify their collusion. The evidence they have provided is not sufficient to support a finding that they have added new flying they could not otherwise have pursued,110 without meaningful reductions in flying elsewhere, and in a manner that produces benefits substantial enough to undermine the plaintiffs’ step-one showing.111
All that remains are claims of more flexible loyalty benefits. The ability to earn and spend frequent flyer miles on either carrier might increase the defendants’ appeal for some customers—likely the “power” travelers American deems most valuable. However, it is the role of the marketplace, not the Court, to judge whether such a feature is appealing or worthwhile. And even if this feature is cognizably “procompetitive,” it is de minimis compared to the anticompetitive harms the Court has found and, thus, insufficient to save the NEA from condemnation.112
In sum, focusing on the short term, American and JetBlue have not adduced credible evidence demonstrating that the NEA has yielded procompetitive benefits capable of justifying its substantial anticompetitive harms. The features they identify are “benefits” only if the Court disregards the free-market principles underpinning the Sherman Act and artificially limits its review to those places and categories of customers to which the defendants have reallocated their resources (ignoring the fact that they did so at the expense of other places and other categories of customers). The Court cannot, and will not, do so.
As the above sections establish, the plaintiffs have made a decisive showing of anticompetitive harm, against which the defendants have offered insubstantial evidence of cognizable procompetitive benefits. Because of this, the NEA fails after the first two steps of a rule-of-reason analysis, however calibrated, and the inquiry need not proceed further. As such, the Court will not address in detail reasonable alternatives or balancing of effects, commenting only briefly on each of those questions.
As noted in the preceding discussion of the benefits alleged by the defendants, the objectives American and JetBlue sought to realize via the NEA could have been achieved by one or more less restrictive alternative arrangements. For example, American and JetBlue could have employed a more limited WCIA-style arrangement to leverage any complementary aspects of their networks, better compete with Delta, and use JetBlue‘s domestic traffic to feed American‘s international service out of the northeast. In such an arrangement, the two carriers would not coordinate with one another on scheduling, network, or capacity decisions, and they would not share revenue on any markets where they provide competing nonstop service. American and Alaska entered just such a partnership, and they both consider it successful despite its limitations. Though the defendants have offered testimony that this alternative was not the one they deemed most appealing, they have not established it would not have been viable. See Doc. No. 320 ¶¶ 125-30 (correctly outlining the standards governing the step-three analysis).
In fact, the record suggests many of the objectives cited by the defendants could be achieved via some degree of codesharing and loyalty reciprocity, both of which are features of the WCIA. Domestic carriers—American and JetBlue included—commonly use such arrangements to provide their customers with access to additional destinations (e.g., to smaller airports in Hawaii), and to allow their frequent flyers to earn and use rewards on their partner‘s flights. Supplementing those features with a slot lease in New York, such as the one the defendants were negotiating before the pandemic, would enable growth by JetBlue in New York, more efficient use of the leased slots, and the reallocation of American‘s own resources to new markets (including those potentially fed by JetBlue‘s increased New York operation).
Such an alternative, which the plaintiffs have identified and supported with evidence, would be less restrictive than the NEA and would likely satisfy any burden the plaintiffs bear at step three of a full-fledged rule-of-reason analysis. Of course, that burden would require proof that the posited alternative would produce only those benefits that are cognizable under the
Finally, if the defendants’ meager showing at step two were somehow sufficient to warrant further analysis, and if the alternative identified by the plaintiffs were not enough to warrant a finding in their favor at step three, the Court would proceed to balance the established harms to competition against the established and cognizable procompetitive benefits. See id. ¶¶ 132-35 (describing legal basis for ultimately engaging in weighing). Here, in
IV. CONCLUSION
The question before the Court is whether the NEA suppresses or promotes competition. The record supports only one answer. The NEA, operating as it was designed and intended by American and JetBlue, substantially diminishes competition in the domestic market for air travel. It does so by combining the Boston and New York operations of two airlines that are among the most significant competitors in that region. These two powerful carriers act as one entity in the northeast, allocating markets between them and replacing full-throated competition with broad cooperation. The plaintiffs have convincingly established that this arrangement immediately and substantially upsets the competitive balance in a highly concentrated industry, not only on a single overlap route or a handful of O&Ds, but throughout the northeast and beyond. The defendants have offered minimal evidence of any cognizable procompetitive effects arising from the NEA. Accordingly, having carefully parsed the record and evaluated the evidence in light of the governing legal standard, the Court concludes that the NEA plainly violates Section 1 of the
V. ORDER
In light of the foregoing Findings of Fact and Conclusions of Law, it is hereby
ORDERED:
- That the defendants are PERMANENTLY ENJOINED from continuing, and restrained from further implementing, the Northeast Alliance, effective thirty days after the date of this Order.
- That within twenty-one days of this Order, the parties shall submit a proposed order reflecting their joint or separate positions regarding the text of the injunction.114
- That any exhibits which were not presented to a witness who testified in person at trial, cited in any party‘s post-trial papers, or cited by the Court in this decision are hereby STRUCK from the record. The parties shall confer and, within twenty-one days of this Order, file a list on the docket containing a complete list of the exhibits that remain in the record in light of this ruling.
- That to the extent any prior Order in this case has sealed any of the information contained herein, such prior Order is AMENDED only to the extent that such information is
UNSEALED,115 based on the Court‘s determination that any confidentiality or other consideration favoring sealing is substantially outweighed by the public‘s right to access and understand the Court‘s decision. - That the Motion to Correct Trial Transcript (Doc. No. 334) is ALLOWED insofar as it is UNOPPOSED—i.e., as to the changes identified in Appendix A to the Motion (Doc. No. 334-1). The motion is DENIED as to the one change the defendants oppose, though the Court notes the disputed testimony was immaterial to its decision.
SO ORDERED.
/s/ Leo T. Sorokin
United States District Judge