Updated June 25, 2015
JOEL I. KLEIN
COMMITTEE ON COMMERCE, SCIENCE, AND TRANSPORTATION
JULY 27, 2000
Mr. Chairman and members of the Committee, I am pleased to appear before you today to discuss recent developments affecting competition in the airline industry and the role that our antitrust laws play in assuring that consumers receive the benefits of competition. Today's hearing takes place against a backdrop of recent reports regarding proposed or possible mergers involving some of our nation's largest airlines. While the Antitrust Division cannot comment on the specifics of any matter that it currently has under review, we fully understand the committee's interest in knowing generally how the Division analyzes airline mergers and investigates other possible antitrust violations in the airline industry.
In my testimony today, I will review the circumstances that have brought us to the present state of competition in the airline industry, as it has evolved around the hub-and spoke system, and identify competitive issues that are presented by that system. I will discuss the cases the Division has brought in recent years against anticompetitive airline practices under the Sherman Act. Then, I will explain how the Division evaluates proposed mergers between air carriers.
Evolution of Competition in the Airline Industry
During the first part of the 20th Century, Congress enacted a number of statutes that subjected major industries to substantial governmental regulation. Building largely upon the statutory regime first enacted in 1887 to regulate railroads, various industries, including other transportation industries such as trucking and airlines, were subjected to restrictions with respect to markets they could enter or exit, prices they could charge, and acquisitions they could make. In most instances, those decisions were subject to prior review and approval by an administrative agency, such as the Interstate Commerce Commission or what became the Civil Aeronautics Board ("CAB").
While the premise of such regulation was that regulatory agencies could restrain anticompetitive behavior by regulated industries and thereby protect the public interest, regulated industries and the public became dissatisfied with regulation. Regulated companies balked at having to obtain regulatory approval every time they wanted to change service or alter price, and consumers complained that agencies often seemed to reflect the views of the industry they regulated, rather than the public interest.
This dissatisfaction culminated in a series of regulatory reform initiatives in the 1970s that reflected a congressional determination that consumer welfare could be enhanced by reducing regulation and allowing consumers -- through their buying decisions in the marketplace -- to identify products and services they desired and the price that they were willing to pay. Thus, Congress enacted a number of deregulatory statutes that curtailed regulation and allowed formerly regulated industries far greater latitude in determining markets to serve and prices to charge.
Following on the heels of a number of deregulatory experiments conducted by the CAB, Congress enacted the Airline Deregulation Act of 1978, which moved the domestic air transportation industry from government regulation to a new era of competition. Carriers were permitted to enter and leave domestic markets without governmental authorization and to set prices and conditions of service. Such behavior would thereafter be subject to the antitrust laws, while the CAB retained jurisdiction over mergers and acquisitions and its authority to prohibit unfair practices.
Industry responses to deregulation were swift. While the prior regulatory regime had resulted in carriers largely providing point-to-point service, with deregulation they began to consolidate their operations at airports, forming what came to be known as hubs. With a hub system, carriers could combine "local" passengers (those originating at or destined to the hub) with "connecting" passengers (those not originating at or destined to the hub but traveling via the hub) on the same flight. In this manner, carriers found they could serve more cities from their hubs (known as "spoke" routes) and offer greater frequency of service with their fleet of aircraft than had been possible with point-to-point service.
Competitive Issues Presented by the Hub System
The hub system has become the dominant business model for most of the major domestic airlines. Such a hub system provides some important benefits for local and connecting passengers. Local passengers benefit because the hub carrier will operate many spoke routes, which means that passengers will be able to obtain nonstop service to many cities. Also, because the hub carrier combines local passengers with a substantial number of connecting passengers on its flights, it is likely to offer more flights from the hub to any spoke city than other carriers (with the possible exception of a spoke city that is another carrier's hub). Connecting passengers benefit not only from the frequency of flights, but also from the ability to choose among routing alternatives offered by various airlines. A passenger seeking to travel from Washington to San Diego, for example, may find that service is offered by several carriers, each via its own respective hubs.
Notwithstanding these benefits, the dominance of spoke routes by hub carriers gives rise to concerns about the exercise of market power by those carriers on those routes. There will usually be at least two carriers providing nonstop service on spoke routes that connect two carriers' hubs, but on other routes there may well be no carrier providing nonstop service other than the hub carrier. Connecting service may be a reasonable alternative for some passengers, especially for those leisure passengers willing to endure the longer travel time that connecting service usually entails, but the absence of competing nonstop service can be especially problematic for business passengers, who often are in a hurry and generally place a higher value on minimizing travel time. Hub carriers can identify such "time-sensitive" passengers and discriminate in the fares they charge them. Studies have shown that carriers generally can, and do, charge higher fares on hub routes, where they face less competition, than on routes that are more competitive.
Once an airline has established a hub at an airport, several structural and strategic factors combine to present high entry barriers to any other airline that might try to enter spoke routes emanating from that hub. By providing more departures to more destinations, the hub carrier can attract a disproportionate share of the hub airport's passengers. This happens for several reasons, including the preference of many travelers to use the carrier with the most flights in a city pair (so that the passenger can change departure times if travel plans change), marketing programs (such as frequent flyer programs) that create loyalty incentives for consumers to concentrate their travel on the dominant airline in their home city, and travel agent commission practices that create incentives for travel agents to encourage their customers to use the hub carrier. A hub carrier often also enters into contracts with local businesses that provide incentives for the businesses to concentrate their travel on the hub carrier. All of these factors serve to discourage entry into a hub carrier's spoke routes, especially by other carriers with similar cost structures.
There is little dispute that hub carriers dominate service at their respective hubs. Today, hub carriers often account for more than 70 percent and sometimes for more than 80 percent of passengers at their respective hubs. There is no reason to think this situation is likely to change in the short run.
Depending on the specific facts involved, there are times when the hub system can present competitive issues under either Section 1 or Section 2 of the Sherman Act.
Section 1 of the Sherman Act prohibits contracts, combinations, and conspiracies that restrain trade. Price-fixing agreements and market allocation agreements are examples of the kinds of collusive conduct that are particularly injurious to consumers. One of the most significant section 1 cases that the Division has recently brought involved the pricing practices of airlines.
In 1992, the Division sued eight airlines and their tariff publishing company for unreasonably restraining trade in violation of section 1. The complaint alleged that the carriers had used computerized fare dissemination services to negotiate fare changes, to trade fare changes in some markets for changes in others, and to exchange assurances concerning implementation of those changes.
Although each of the major domestic carriers offers service in thousands of city pair markets, the Division found that carriers had varying preferences as to the prices that should be charged in any particular city pair. Preferences may differ for any of a number of reasons, including the importance of a route to the carrier's hub operations. A carrier might be very interested in the fare level in city pair A-B if it operated many daily frequencies, and be less interested in the fare level in city pair C-D if it operated only one or two. Yet, city pair C-D might be very important to another carrier, and city pair A-B less so. The Division found that the airlines had used computerized fare dissemination systems to work out trades: "I'll go along with an increase in A-B if you go along with an increase in C-D." A consent decree now prohibits certain practices that the airlines had used to reach these kinds of agreements on fares.
Section 2 of the Sherman Act prohibits monopolization and attempts to monopolize. Unlike section 1, which requires some form of agreement between two or more persons, section 2 focuses on single firm conduct. Generally speaking, even a firm with a dominant share of a market does not violate section 2 unless it engages in some form of exclusionary conduct. The law does not penalize a person for obtaining a monopoly through superior skill, foresight, and industry. However, if a person seeks to maintain a monopoly through exclusionary conduct, or if there is a dangerous probability that a person will obtain a monopoly through exclusionary conduct, the Division may sue under section 2.
In the airline industry, concerns have been expressed that hub carriers engage in exclusionary practices to keep low-cost carriers (LCCs) out of their hubs. The Division takes these concerns very seriously, because LCCs may offer the only realistic prospect of competition to hub carriers in precisely the markets that suffer from a lack of competition. The Division has found that major carriers are not likely to challenge another carrier at its hub by offering point-to-point service (except on a spoke route from their own hubs). The advantages that a hub carrier enjoys at its hub make entry of that sort unlikely. But LCCs, with their lower cost structures, may be able to offer service on a hub carrier's spoke routes notwithstanding the hub carrier's advantages.
A hub carrier may therefore have a strong incentive to engage in predatory practices to drive LCCs out of its hub markets and to send a strong signal to others that might consider entry that the same response awaits them if they try. The airline industry has characteristics that may make such a strategy particularly attractive to a hub carrier. If an LCC begins service on a hub carrier's spoke route and the hub carrier engages in predatory conduct that drives the LCC out, the hub carrier has benefited in many ways. Not only has it driven the LCC out of that particular route, but it has also probably discouraged that LCC from expanding to serve other cities from that hub. And not only has this LCC been driven away, but all other LCCs contemplating entering that hub will see what fate awaits them if they dare to venture in. Thus, predatory practices directed at a single LCC in a single spoke route can protect the hub carrier's ability to charge high fares in other spoke routes it dominates.
The Division has filed suit against American Airlines alleging monopolization and attempted monopolization at its Dallas/Ft. Worth hub in connection with predatory practices directed at LCCs. The attached charts show what happened in some of the markets involved. When the LCC entered the market, fares declined, and the number of traveling passengers went up substantially. After the LCC exited the market, the opposite occurred: fares went up, and the number of traveling passengers fell. The case is still in discovery, and trial is scheduled for next spring. We view this as a very important case, one that can have a significant impact on airline competition and on the Nation's consumers.
Evaluating Mergers and Acquisitions Among Air Carriers
Let me now turn to how the Antitrust Division evaluates proposed mergers and acquisitions among air carriers, starting with some historical background.
During the first years following deregulation, antitrust jurisdiction was divided between the Division and the CAB. The Division could -- and did -- prosecute airlines for price fixing and other violations of the Sherman Act, but the CAB retained sole jurisdiction to review mergers and acquisitions. The CAB was presented with a number of proposed mergers in the late 1970s and into the 1980s. When Congress sunset the CAB in 1985, it temporarily transferred merger review
The DOT's jurisdiction over mergers terminated effective December 31, 1988, after which time the Division assumed responsibility for airline merger review -- although we continue to work closely with the DOT, given its substantial expertise with respect to the airline industry. Since then, there have been very few mergers proposed among the major airlines.
However, in 1998, Northwest, then the fourth-largest U.S. air carrier, sought to acquire a controlling interest in Continental, then the fifth-largest U.S. carrier. The Division has challenged the acquisition, and trial is scheduled for later this year.
In addition to challenges to mergers and acquisitions of stock, the Division has also challenged acquisitions of assets that it concluded would be competitively problematic. The Division has moved to block acquisition of gates or slots when it thought such acquisitions would lessen competition, as demonstrated by its challenges to Eastern's proposal in 1989 to sell gates to USAir at the gate-constrained Philadelphia International Airport and Eastern's proposal in 1991 to sell slots and gates at Reagan Washington National Airport to United.
For most of the 1990s, airline acquisition activity centered on acquisition of international route authority. Here, the Division shares review responsibility with the DOT, which has jurisdiction over transfers of international route authority. Such authority is literally an "admission ticket," since many international bilateral aviation agreements limit the number of U.S. carriers that can provide service to a foreign country, and service cannot be provided absent such authority. Financially ailing domestic carriers with substantial international route authority, such as Pan Am, Eastern, and (at the time) TWA, sold route authority to other U.S. carriers. The Division reviewed these transactions as well and challenged some of them, such as the proposed sale by TWA of its London route authority to American.
Recently, attention has turned back to the domestic scene, with the announcement of a proposed acquisition by United of US Airways and speculation about mergers involving other major U.S. carriers. The Division has announced that it will review the United-US Airways merger carefully, as it will any merger between major U.S. carriers. Meanwhile, there are reports of negotiations between other major airlines.
In reviewing airline mergers, the Antitrust Division applies Section 7 of the Clayton Act, which prohibits the acquisition of stock or assets "where in any line of commerce or in any activity affecting commerce in any section of the country, the effect of such acquisition may be substantially to lessen competition, or to tend to create a monopoly." Section 7 reflects the congressional judgment that merger enforcement should be able to arrest anticompetitive mergers in their incipiency, to forestall the harm that would otherwise ensue but be difficult to undo. Thus, merger enforcement standards are forward looking and, while we often consider historic performance in an industry, the primary focus is to determine the likely competitive effects of a proposed merger in the future.
A major U.S. carrier seeking to merge with or acquire another carrier must provide the Division and the Federal Trade Commission ("FTC") with notice of the proposed transaction pursuant to the Hart-Scott-Rodino Antitrust Improvements Act of 1976 ("HSR"). Although the Division and the FTC share merger enforcement responsibility as a general matter, the Division is the agency that reviews air carrier mergers. The initial HSR filing contains certain basic information, which the Division uses to determine whether more extensive review is appropriate.
The initial waiting period under HSR is usually thirty days. If the Division concludes during that period that the merger is not competitively problematic, the HSR waiting period is allowed to expire or may even be terminated early. The parties are then free to proceed, subject, of course, to any other required regulatory approvals. However, if the Division cannot resolve its competitive concerns within that period, it can issue a request for additional information, known more commonly as a "second request," which defers the ability of the merging parties to consummate their transaction until twenty days after they have provided the Division with the requested information. During this time, the Division will frequently seek or receive information from other persons interested in the merger; these may include suppliers, customers, and/or industry specialists. We work closely with the DOT, which obviously has substantial expertise with respect to the airline industry.
It is not uncommon during this process for the parties to have substantial contact with the Division. The process is confidential and, unlike the procedures in some administrative agencies, competitors do not have access to the merging parties' submissions. Sometimes parties are able to demonstrate that the merger is not competitively problematic, in which case the waiting period expires or is terminated early; again, the parties may then proceed, subject to other required approvals.
If the Division concludes, however, that the merger violates the law, the Division can attempt to stop the merger by filing a complaint in federal court and persuading a judge to enter an order prohibiting the parties from consummating it. It is not uncommon, however, for the parties to make a proposal to address the competitive concerns that the Division has identified, in which case some form of agreed-upon relief may resolve the problem while still allowing the parties to proceed with the overall transaction. In those circumstances, the Division ordinarily files a complaint along with a consent decree that embodies the relief in the form of an order entered by the court. There are times, however, when the competitive problem cannot be cured by any form of relief other than outright prohibition, in which case the Division is likely to seek a preliminary injunction to prevent consummation of the merger pending completion of judicial proceedings and then a permanent injunction prohibiting the merger altogether.
The Division looks for relief that will fully address the competitive problems presented by the merger, which almost always means seeking some form of divestiture. Parties sometimes propose conduct remedies -- usually some form of behavioral restrictions -- but these are generally unsatisfactory for a number of reasons. First, they are often difficult to draft with precision. Second, they require continuing monitoring by the Division. Third, they cannot be enforced without resort to the court on a continuing basis. Finally, they have often proven to be insufficient to remedy the anticompetitive problems presented by a merger.
The particular form of divestiture necessary to solve a competitive problem will vary from merger to merger and involves many inquiries. First, it is essential that the assets to be divested are sufficient to allow a purchaser to be an effective competitor over the long term, i.e. to replicate the competition that would otherwise be eliminated in the markets of concern. Sometimes the necessary assets are easy to identify -- as, for example, when a party agrees to divest a stand-alone business entity such as a subsidiary or a pre-existing operating division. In those instances, the sufficiency of the assets can be evaluated in light of historical performance of the business unit in the marketplace. In other instances, parties propose divestiture of specific assets, sometimes even combining some assets from each of the parties to the merger. It can be difficult to assess whether such assets are sufficient to allow the purchaser to compete on a meaningful basis because there is no track-record to gauge the adequacy of the asset package. A recent study of antitrust divestitures by the FTC "suggests that divestiture of an on-going business is more likely to result in a viable operation than divestiture of a more narrowly defined package of assets," although "divestitures of selected assets can succeed."
Second, the Division will look carefully at the prospective purchaser. In most instances, the proposed purchaser is not selected until after the court has entered an order directing the nature and form of the divestiture, but occasionally parties to a merger will identify a proposed purchaser to the Division during the course of the investigation. In either case, the Division will review the experience, financial resources, and business plan of the purchaser, all in an effort to determine whether the purchaser is likely to solve the competitive problem presented by the original merger.
In performing this review, the Division considers the terms of the proposed contract and any other arrangements between the merged entity and the purchaser to determine whether the purchaser will be an independent competitor. The Division's form consent decree provides, for example, that the merged entity cannot finance the sale to the purchaser. Similarly, the Division is generally skeptical about supply contracts between the merged entity and the purchaser, as well as any other arrangements that tie the purchaser to the merged entity, although there may be circumstances in which such arrangements are warranted. Our concern is that, if the purchaser is dependent upon the merged entity for critical products or services, there are two risks: (1) the merged entity may seek to influence the behavior of the purchaser by manipulating price or supply of such products or services and (2) the purchaser may pull its competitive punches for fear of antagonizing the merged entity.
The Division has, on occasion, refused to approve a purchaser unless and until changes have been made in the terms of the divestiture to assure that the purchaser will be viable and independent. While the Division does not (and should not) seek to ensure the success of a purchaser, it must be confident that the divestiture will remedy the competitive problem that it is intended to fix.
The Division and the FTC have jointly developed Merger Guidelines that describe the substantive considerations for analyzing mergers. "The unifying theme of the Guidelines is that mergers should not be permitted to create or enhance market power or to facilitate its exercise." Merger Guidelines 0.1. As suggested by the language of Section 7 itself, we usually start by seeking to define the relevant product or service markets ("line of commerce") and geographic markets ("section of the country") in which the parties to the merger compete, and then determine whether the merger would be likely to lessen competition in those markets.
The purpose of this inquiry is to ascertain whether, with respect to a product or service offered by the merging parties, there are alternative products and services to which customers could reasonably turn if it were assumed that the merging parties were the only suppliers of the product or service and sought to increase prices. Once relevant markets are defined, we look at various factors in order to determine whether the merger is likely to have an anticompetitive effect.
In performing this analysis, the Division considers both the post-merger market concentration and the increase in concentration resulting from the merger. As a yardstick for concentration, we utilize the Herfindahl-Hirschman Index ("HHI"), which is calculated by summing the squares of the individual market shares of all the participants. The Division will presume that mergers in highly concentrated industries that produce more than a small increase in concentration are likely to create or enhance market power or facilitate its exercise, unless other factors, such as the prospect of entry by other firms, make that unlikely.
We apply this basic approach to analysis of air carrier mergers. In this industry, the definition of product/service market and geographic market converge: relevant airline markets are likely to consist of scheduled airline service between a point of origin and a point of destination, generally referred to as city pairs. This market makes intuitive, as well as economic, sense. A passenger desiring to fly from Washington to San Francisco for a business meeting or a vacation is unlikely to regard a flight from Washington to Minneapolis as a reasonable alternative in the event the fare from Washington to San Francisco is increased. Thus, we should be concerned about a merger that significantly raises concentration levels in city pair markets.
The relevant market may, however, be narrower than all scheduled airline service in a city pair. Carriers can serve a city pair market on a connecting basis or a nonstop basis. If the only available service offered by carriers in a city pair is connecting service, there may be various routes that passengers regard as reasonable alternatives and from which they will choose based on fare, elapsed travel time, and other factors. However, there are many city pairs that are served by some carriers on a nonstop basis and others on a connecting basis, which poses the following question: is a passenger who is able to take a nonstop flight likely to regard connecting service as a reasonable alternative, such that he or she would switch from nonstop service offered by one carrier to connecting service offered by another carrier if the first carrier raised its fare?
Chances are that passengers traveling for leisure -- on vacation perhaps -- are more likely to consider switching; their demand is said to be more elastic. However, passengers making business trips are significantly less likely to regard connecting service as a reasonable alternative -- they are often in a hurry and may place a higher value on getting to their destination in a hurry -- so that a carrier offering the only nonstop service has power to raise fares without losing these passengers to another carrier's connecting service. Thus, there may be circumstances in which a merger will be competitively problematic because of its impact on nonstop service in city pair markets, even if other carriers provide service in those markets on a connecting basis.
Therefore, in considering the antitrust implications of a particular merger, the Division looks at the effect in all city pair markets served by both of the carriers involved in terms of (1) nonstop service and (2) nonstop and connecting service. We have found -- not surprisingly, given the operation by carriers of hubs in the post-deregulation world -- that the mergers most likely to be problematic are those between carriers with hubs at the same airport or at airports in the same metropolitan area. These carriers are likely to serve many of the same city pairs and, especially in spoke markets, they may be the only two carriers, or two of a very small number of carriers, providing service.
That is not to suggest, however, that mergers between carriers that do not have overlapping hubs may not also present problems. Carriers with hubs in nearby cities are often the dominant carriers -- usually on a connecting basis -- for a significant number of city pairs in their region. And even when carriers' hubs are substantial distances apart, it is often the case that they are the only two carriers providing nonstop service between their respective hubs.
The Division has challenged, for example, the acquisition by Northwest of a controlling interest in Continental, even though the carriers do not operate hubs at the same airports. Our complaint alleges that the acquisition would lead to higher ticket prices and diminished service for millions of passengers, especially those traveling on routes dominated by the two airlines. Northwest and Continental are each other's most significant competitors -- and sometimes the only competitors -- for nonstop airline service between cities where they operate their hubs.
Once overlapping city pairs have been identified, the Division looks at the number of other carriers serving each of the markets and at the nature of that service, often by resorting to data that carriers report periodically to the DOT. This allows the Division to calculate market shares and focus further analysis on those city pairs in which pre-merger concentration levels suggest that the post-merger structure would be conducive to the creation or enhancement of market power.
As the Merger Guidelines indicate, however, the analysis does not end there. Pre-merger market shares are a useful tool for predicting future market shares of the incumbents in a market, but they do not take account of the possibility of entry by additional competitors. The prospect of potential competition can constrain the ability of incumbents to raise price or reduce output below a competitive level.
Indeed, the possibility of potential competition was the linchpin for many of the DOT's decisions approving mergers between carriers. Potential competition, it was said, could be relied upon to discipline carriers, even those with dominant market shares: if a dominant carrier sought to raise fares above competitive levels or reduce service below competitive levels, new carriers could easily enter, especially if they already had some operations at the affected airports. Airplanes were the quintessential mobile asset, it was said, and ground facilities could be easily leased or subleased. Knowing that noncompetitive behavior would attract entry, it was claimed that dominant incumbents would price competitively and offer competitive levels of service. Hence, the DOT reasoned that market shares -- and the presumptions of market power that accompany them -- were of relatively little use in airline merger analysis. The airline industry became the poster child for contestable market theory.
The Division does not subscribe to this entry analysis. It simply does not conform to the facts in a post-deregulation world consisting of hub airports. For all of the reasons I mentioned earlier, hub economics are powerful. In these circumstances, carriers with comparable cost structures to the hub carrier generally find it unattractive to take on the hub carrier head-on. Entry by a major carrier on a point-to-point basis into another carrier's hub has become very much the exception. Thus, the hub carrier dominates city pairs it serves directly from its hub, except routes to cities that are hubs for other carriers, in which case the two carriers providing hub service dominate. And without substantial actual competition, hub carriers charge higher fares to local passengers than they do in more competitive markets.
This does not indicate that entry into a carrier's hub is impossible. Carriers with low costs (known as low-cost carriers or "LCCs") may be able to enter profitably, even with point-to-point service. But such entry has tended to be gradual and limited. Under our Merger Guidelines, the Division considers whether entry into the affected markets is so easy -- in the sense that it would be timely, likely, and sufficient in its magnitude, character, and scope -- that it will likely deter or counteract the anticompetitive effects. For a merger between major air carriers with substantial overlaps in markets in which they are the dominant providers of service, it is unrealistic to expect that the prospect of potential competition can fully address the competitive concerns.
Finally, the Division will consider and take into account airline-specific business practices and characteristics that can affect merger analysis, especially those that differ from most other industries. Airline fare data is available instantaneously not only to consumers, but also to the airlines themselves, which can act as a disincentive to fare reductions. Airlines frequently propose general or system-wide price increases, which may be more likely to "stick" as the number of major carriers diminishes. Carriers have developed loyalty programs that tie passengers and travel agents to them at their hubs, making entry into those hubs more difficult. And airlines apply sophisticated computer modeling techniques and ticketing restrictions to identify passengers to whom they can charge higher fares, a form of price discrimination. The Division will consider these and other factors in seeking to determine whether any proposed merger threatens to substantially lessen competition.
Mr. Chairman, competition in the airline industry is critical for the millions of people who depend on air travel in their business life and in their family life. If the Division concludes that hub carriers are engaging in collusive or monopolistic conduct, or that any proposed air carrier merger threatens to deprive consumers of the benefits of competitive air service, I assure you that the Antitrust Division will take appropriate enforcement action.
Mr. Chairman, this concludes my prepared remarks. I will be happy to answer any questions that you or other members of the Committee may have.