on
Airline Mergers and Concentration
Before the
Committee on the Judiciary
U.S. Senate
February 7, 2001
Mr. Chairman and Members of the Committee: Thank you for giving me the opportunity to testify before you today at what I believe is a critical point in the development of the deregulated airline industry. I testify at the invitation of the Committee as a private citizen and not on behalf of any airline, industry group or other organized interest. My reason for testifying is simple: I have dedicated most of my career first to bringing about a competitive deregulated airline industry and then to demonstrating through my own personal efforts that it is possible for a well-managed airline to survive and prosper in a competitive environment. I see a threat to the continued success of airline deregulation, and I hope to play some part in countering that threat.
I am at present a member of the faculty of the Harvard Law School, teaching courses in regulation and international joint ventures. I have attached a detailed biography to this testimony for your information, but let me say briefly that I have had the unusual opportunity to study, to regulate and to work in the airline industry. This experience has included work as a dean and scholar who has advocated and continues to advocate deregulation at USC, Caltech, Yale and Harvard. It also included a position as the senior staff member at the Civil Aeronautics Board under Alfred Kahn and then Marvin Cohen during the most pivotal deregulation period. And I also have had the opportunity to participate in the industry as a CEO or senior executive of a transitional network airline (Continental), a new entrant airline (New York Air) and finally at the fourth largest airline in the United States (Northwest).
I am very concerned about the consequences for industry competition and ultimately for consumers of the proposed division of US Airways between United Air Lines and American Airlines.
Before I discuss that transaction I should make clear that the “companion” merger between American and TWA on its own presents no serious competition problems. That TWA is a failing company seems beyond doubt. The TWA deal may present difficulties for American in terms of labor, fleet and systems integration. Those problems may present service problems for the traveling public but if they materialize, the public can deal with them by avoiding American. They will still have that choice because the American-TWA transaction will not change the structure of the industry and does not present a threat to the competition that is necessary for deregulation to succeed as a public policy. This matter should be left to the marketplace and the bankruptcy courts.
American has justified its merger with TWA on its own merits at the same time that it has presented it as part of a strategic package that includes American’s agreement with United to divide US Airways. It seems clear to me that the most important purpose of the TWA deal is to help give a “failing-company” cast to the whole four-airline transaction, and to provide political cover (preserving 20,000 jobs and a large-airline hub presence at St. Louis) to politicians and government officials as they consider a total transaction much more difficult to justify on competition grounds. The second major benefit to American is not the chance to operate a St. Louis hub, but rather to use TWA’s slots and facilities at congested East Coast airports to bolster American’s New York and East Coast strategic position and to use TWA aircraft to achieve market share parity with United as part of the Big Two strategy discussed below.
The significance of the TWA transaction is that a closer look at it raises suspicions about American’s strategic motives. On its own, the TWA transaction is difficult to justify commercially. TWA has been carefully examined as an acquisition candidate by every major airline (more than once, in many cases), and I believe that those studies all came to the same conclusion: while St. Louis is well-located and can support a hub of some size, it would be very difficult for a “normal” network airline to make any significant profit there.
First and most important, operating a hub on top of Southwest Airlines means that normal hub economics are impaired by the inability to charge normal hub fares to short-to-medium haul business travelers, and as Southwest’s system continues to evolve out of its previous short-haul, point-to-point mode, that effect becomes more and more severe. Just ask America West, which has had considerable difficulty maintaining at Phoenix a revenue base adequate to support a significantly profitable hub operation, even at its very low costs. When you add into this equation American’s labor costs and the transition costs of labor, systems and fleet integration, it’s difficult to believe that American’s better credit and better fuel purchase position and the overhead savings from eliminating TWA’s management infrastructure make this transaction taken by itself additive to American’s earnings or worth the risk. I know these numbers didn’t work for anyone else, and would be surprised to learn that they suddenly make sense on their own for American.
Second, this is clearly a case where American is acting in concert with United to achieve jointly-shared strategic goals. If United was only interested in solving the Washington, DC part of the antitrust problem presented by its own US Airways deal, any number of other airlines would have been willing to help them out. But rather than Continental or Airtran, who have publicly indicated a willingness to work with Robert Johnson to produce a DC Air that would be a full-blooded competitor to United (or rather than the couple of other airlines who are rumored to have expressed serious interest), United has chosen to work with the airline that is its supposed arch-rival and that should be its most difficult competitor from the standpoint of network coverage (“scope”). In fact, when the transaction is taken as a whole United has cooperated in fashioning a deal that represents a giant step forward for American in achieving its stated goal of network ubiquity even as it impairs United’s attempt to build a uniquely ubiquitous position. Why would United do this? To understand, I think we need to look at a bit of history.
American and United are what remain of the prederegulation “Big Four”. Eastern has gone to its reward and TWA, shrunk to a shadow of its former self, is about to follow. Both were victims not only of their own managements’ strategic mistakes, but also of their inability to persuade their own labor forces to adapt proactively to the changed circumstances of deregulation. United and American, facing the same concerns about their ability to survive deregulation given their high costs, adopted a different management strategy: they persuaded their labor forces in the postderegulation period to reach accommodations that lowered marginal labor costs (“B”-scales, ESOP, periodic scope relief, etc.) and allowed fleet and system flexibility in return for assurances of growth, producing more job security and richer lifetime career paths for employees. They coupled this with adoption of a “ubiquity” strategy, in which the size and reach of their networks would allow them to meet almost every air transportation need of every airline customer. This ubiquity would be used to differentiate themselves from new entrants for business travelers and to gain a revenue advantage over other network competitors. United announced shortly after deregulation that it had become the first airline to serve all 50 states. American moved to Dallas so that it could serve a very large, centrally located, facility-unconstrained O&D market as a national hub. The idea for both American and United was that they would ultimately overwhelm smaller network competitors as customers and travel agents chose to sign contracts with and use the frequent flyer benefits of the airline that could satisfy the largest portion of their needs.
On their way to unchallenged ubiquity, two things happened. Other network competitors saw what was happening and refused to roll over quietly. First Texas Air, then Delta, Northwest, Allegheny/USAir (remember the Piedmont merger and the name change?) and Continental on its own attempted expansions designed to enhance their own ubiquity and thus survivability. A sort of ubiquity arms race ensued, which caused severe self-damage to more than one participant and nearly destroyed the entire industry when the economic expansion of the 1980s segued into the recession of the early 1990s. In the process, Delta became large enough to approach American and United in size, but more important, the recession-induced stunting of the growth process evolved the industry into an “almost-national” mode, with each successful network airline building and defending regional core positions that supported a large but incomplete national hub system. The traveling public benefited hugely from this process (shareholders benefited less!). The almost-national systems were very large and provided many of the benefits of complete network scope. People in spoke cities often had a choice of as many as half a dozen competing hub carriers that could meet a particular trip need, hub-located travelers could get nonstop service to 80 or more destinations comprising most of their travel needs and most travelers could meet virtually all their needs by concentrating their business on two systems, for which they were rewarded with frequent flyer benefits they valued greatly.
But from United’s and American’s perspective, this was not such a splendid state of affairs. They had built their labor strategies around paying labor for growth and the ability to use their network strength to capture revenue premiums (monopolistic rents). Growth was slowing as it had become clear that capacity expansion would be defensively matched and there was not enough new business to support profitable expansion for American and United relative to the rest of the industry. The national market became more concentrated among the top five network airlines and Southwest, but almost all of the incremental share went to Southwest, Delta, Northwest and Continental. The development of alliances by smaller airlines as a way to achieve many of the benefits of network size without the risks of overcapacity further eroded their revenue premiums. The net result of twenty years of deregulation was NOT that American and United had become uniquely ubiquitous airlines, but rather that they had come to share the network industry with several competitors that not only wouldn’t go away, but which constrained the possibility of further share expansion. For American and United, the strategic question became: how can we (either American or United or both) gain a network size advantage that can’t be duplicated and eroded and which will yield monopoly rents to support our very high costs?
Both airlines came to the conclusion that the key was the East Coast: United already dominated network service on the West Coast, but the West Coast has relatively few cities and while those cities wouldn’t support more than one network (as American repeatedly found out through expensive tests – the Air Cal and Reno acquisitions and the San Jose north-south hub), its relatively uncongested, separated airports were ideal for expansion by Southwest. Further competitive shifts toward American/United were unlikely there. Delta’s Atlanta hub operation along with expansion by Southwest and Airtran made the Southeast unpromising. The midline of the country provided as many opportunities to Continental and Northwest as to American and United, especially given the constraints at Chicago-O’Hare.
By contrast, the East Coast has a variety of interesting features which might allow it to underpin a sustainable network size and scope advantage which could be leveraged into a dominant position: a large part of the nation’s population and travel origin is located there. Airports are congested and facilities tight, making substantial matching expansion by network competitors difficult and substantial discount competition at the primary business airports nearly impossible. Four major population concentrations are the focus of much of the business traffic: Boston, New York City, Philadelphia and Washington. Northwest has no presence there except through the Continental alliance. Continental’s and Delta’s strength is largely limited to Newark (Continental) and north-south and transatlantic service (Delta). Transcontinental business is already dominated by American and United. Continental has only been able to build a significant transcontinental business from its Newark hub using narrowbody aircraft and Delta has been unable to make a significant dent in these markets. United has built a hub at Dulles and American has made a significant effort to build its presence at Boston, but neither of these efforts have produced a sufficient increment in East Coast presence to allow unduplicable network expansion that could cast a halo over the entire United States system.
American started to build an alliance with US Airways, the only airline with strategically-located sufficient mass that could make a difference to its network strength. The alliance involved codesharing, a frequent flyer deal and computer systems integration which lowered American’s costs. Northwest and Continental built an alliance which made Northwest a much stronger competitor to United in the Midwest and over the Pacific and strengthened Continental’s position in New York. These developments concerned United greatly. United was offered the opportunity to do something decisive in response by US Airways management’s conclusion that its structural and cost problems couldn’t be overcome without major flexibility by its unions, and its consequent decision to save its shareholders by bailing out after an attempt to reach union accommodation failed. The result was the United/US Airways deal.
What United expected to get out of the deal was an effective monopoly in Washington and Philadelphia, a greatly enhanced position in Boston and New York, and a major frequent flyer presence in the very important Shuttle markets. It hoped simultaneously to strengthen its revenue position vis-à-vis American, achieving through system market power what it had never been able to achieve through service and operations and to finally separate itself from the increasing competition offered by Delta, Continental and Northwest. That United paid too much is a tribute to Stephen Wolf’s bargaining skills. That it did the deal without getting the union consents that would have helped manage transition costs is a confirmation of the priority that United’s management gave the deal and how much impact on competition they expected it to have. There are many who think that this transaction might have in the end cost so much that it wouldn’t have made a profit for United. That the costs of integrating the two airlines might have been such that its shareholders might not ultimately have benefited does not mean that there were no monopoly profits to be made, but only that the monopoly profits would be distributed among US Airways shareholders, United’s labor force and Robert Johnson.
The only problem with all this is that the United/US Airways deal, despite its beautifully prepared political campaign, appeared to be in danger of failing. The DC Air “cure” to the Washington problem was not passing the laugh test. No one seriously believed that a United-supported DC Air with a large commuter component was likely to provide significant stand-alone competition to United in Washington. Offers of “help” by Continental and Airtran put United between the devil and the deep blue sea with respect to its transaction goals. Giving Continental a strong Washington position was the opposite from what United was trying to achieve in redistributing network system strength away from its pesky pursuers. And allowing a discount airline like Airtran to operate from the business revenue heart of its East Coast hub strength (bad enough to have Southwest at BWI!) would be very damaging to United’s Washington economics and would make the transaction even more expensive by a substantial margin (in much the way that Southwest’s presence at St. Louis makes the TWA transaction expensive for American).
American, with the prospect of losing its US Airways relationship and of seeing its United rival get a structural lock on a superior network position, offered United a brilliantly-conceived truce that was much more valuable to United than a failed deal and a continued war with Delta, Continental and Northwest. In effect, it offered to jointly share ubiquity, establishing a Big Two protected from imitation by East Coast facilities constraints and antitrust barriers to further merger. With the TWA deal and the deal as American and United have structured it, American and United would be almost exactly the same size at about 25% of the national market. Each of the Big Two could sustain a revenue premium relative to Delta, Continental and Northwest and generate network monopoly premiums to help stave off the economic impact of Southwest. Neither would have the incentive to erode those rents through price competition with the other (because little relative share gain would be possible), so pricing discipline would be maintained without collusion. While there would be a possibility that Delta or Continental might try to defend itself by combining with Northwest, none was a failing company and the Justice Department could be expected to be hostile, given its record in the Northwest/Continental control case. Paradoxically enough, the United/American joint monopoly position could be defended with the antitrust laws!
Even if their rivals could merge, no one would have the combination of Boston, Philadelphia and Washington strength available to the Big Two and could achieve the same system leverage. American could make itself stronger in New York through the TWA deal, achieve near-parity in Washington and Boston, and concede Philadelphia. It could make excellent network use of the Washington and other Northeast slots and gates it gets in this deal because of its success in using regional jets to maintain presence on mainline routes. Its ability to sustain a network advantage over “the others” would be assured. United would strengthen its position in Washington, Boston and New York, gain control of key facilities and slots, and build an East Coast North/South system. For both American and United, rivalry with each other along nonprice dimensions while each had market power relative to the rest was an attractive alternative to the status quo.
The Big Two position that these transactions would create is likely to last a very long time. The large pool of customers available in the Northeast and the ability to use the scarcity of slots and gates at its congested airports to lock them up will make it impossible to duplicate the Big Two position that American and United will share. No comparable opportunity will be available to other big network airlines and therefore no other network airline will be able to match United’s and American’s ability to offer corporate contracts, travel agency and internet incentives and frequent flyer benefits. Over time, Delta, Northwest and Continental will find it increasingly difficult to capture East Coast business passengers, providing less flow at their hubs and supporting less service than American and United will be able to sustain. The gap between American and United and the “others” will grow.
Among the strongest pieces of evidence that this narrative captures what the participants predict and intend in this deal is the treatment of the US Airways Shuttle, which is a crown jewel in any network scope strategy. The Shuttle is used primarily by a group of business travelers who are also the ones most likely to buy high-priced tickets to elsewhere from Boston, New York and Washington. In Delta’s hands, the other shuttle is one of the assets most valuable in its efforts to move toward network parity with American and United. As a potential source of monopoly dominance, the US Airways shuttle is wasted in US Airways’ hands because US Airways doesn’t have the complementary system strength to take advantage of it. In fact, the Shuttle doesn’t even serve Philadelphia, which is US Airways focus for much of its valuable business flying! American had a temporary advantage over United through its alliance with US Airways. United grabbed it back. United’s giving up exclusive control of the network value of this Shuttle only makes sense in the context of a shared-dominance strategy in which both airlines see its principal value as enhancing their ability to suppress competition on the rest of their networks. This view of the transaction is confirmed by the fact that United gets to keep all of the Shuttle if American concludes an acquisition that makes it bigger than United!
This discussion doesn’t deal with all of the potential objections to this transaction, some of which are common to the United/US Airways transaction as well. For example, public vulnerability to labor disruption is increased as more of the system falls into fewer hands. The public consequences of a job action on an airline so big that the rest of the system simply cannot absorb its business are very serious, as are the consequences of the associated imbalance in bargaining power. I have tried instead to focus on the subtle and complex competitive dynamics that underlie this transaction in an attempt to explain why this is not just another merger and just another rescue of some threatened airline jobs. (On that subject, I should say that the notion that US Airways is, like TWA, a failing company is entirely wrong. Faced with no alternative, management and labor could work together at US Airways to achieve costs and revenues that would enable it to survive, although some surgery might be necessary. But that’s another story for another time.)
What can be said in favor of this transaction? Only that if consumers prefer to concentrate their business on one very large system, we should accommodate them. And there is no doubt that some consumers would prefer to do so, especially if all other things were equal. But all other things will not remain equal. This convenience will come at the price of choice and long-term competition. There are often conveniences to monopoly, as anyone who used to have only one number to call when they wanted to discuss their phone service will attest. But there are benefits from competition which have generally been judged superior as a matter of public policy. If one compares the utility to consumers of having competitive choices among airlines, almost any two of which can satisfy almost all their needs, with the “convenience” of one-stop shopping in a duopoly, I believe that most consumers would prefer competition. That comparison is reflected not only in our antitrust laws, but in the regulatory policies of the past twenty-five years.
It has been urged by at least one observer that we need not be concerned about loss of competitive pressure in the network business because Southwest in particular and other low-cost airlines in general represent a large enough share of the business to discipline United and American. I suppose that the first rebuttal is American and United clearly don’t agree with him. It’s difficult to justify the cost commitments and vulnerabilities which this transaction entails for American and United without assuming that they believe that they will earn substantial monopoly benefits from the transaction. There are good reasons for thinking they may be right, even if in the end the transition and labor costs of the deal are so large that it ultimately doesn’t benefit their shareholders:
First, although Southwest and its ilk offer a valuable service to their passengers, it is not a service equally valuable to all passengers. These airlines do not have significant presence (indeed, Southwest has no presence) at the very congested and constrained airports that are the principal focus of this transaction. Business travelers value and will pay for airport convenience, which is why, for example, business fares are much higher from Boston to Reagan National than they are from Providence to Baltimore-Washington International.
Second, these discount airlines do not maintain networks that are easy to use for complicated itineraries or which afford easy access to airports close to smaller cities. They rely on the willingness of a traveler to drive to reach an airport where fares are low. For many travelers, this is an excellent tradeoff, but for a substantial number of business travelers, it is not. Third, Southwest may be second in the nation in the number of passengers it carries, as some are fond of noting, but it is much smaller in terms of its overall volume of business, which is ultimately how economic impact is measured. Southwest is seventh in the number of Revenue Passenger Miles (the standard measure of output) and even if it grows as rapidly as analysts assure us it will, it will still be responsible for a substantially smaller share of industry total revenue or industry total output than its large network rivals, not to mention the Big Two.
Finally, Southwest itself is not a charitable organization, fully conceding Herb Kelleher’s legendary benevolence and charm. Its pricing is constrained by network carriers, just as network carriers constrain it. If the pricing umbrella is set higher by the Big Two, Southwest itself can charge more. Southwest claims that its main competition is the car, but that is only true in the short-haul, point-to-point markets that are no longer the mainstay of its system or the source of its growth. In fact, the car has become much more a complement for travel on Southwest than a substitute. Its customers drive significant distances to get to its uncongested airports. If the Big Two price higher, Southwest can charge more and still make it worthwhile for its customers to drive to its flights. Each rise in Southwest’s price level would cost the public a very great deal. Southwest and its brethren are a very valuable part of the U.S. airline system, but its existence is certainly not a substitute for strong competition among network airlines.
In conclusion, this is not just another merger and not just another bailout of a failing airline. The American/United/US Airways transaction is an attempt to undermine the competition created by deregulation. It will do this by building a wall of scarce East Coast infrastructure around a fortress occupied by a Big Two, who will use the protection of that fortress to attack their pursuers. With all its imperfections, deregulated airline competition has served the United States well. The Big Four of the CAB, protected from each other by regulation, is now a group of six highly rivalrous network airlines in which at least three of the smaller players are gaining on the larger two, supplemented and disciplined by a large and growing discount airline system. Congress and the Administration should not allow those who have the most to lose from this evolution to put a halt to it.
What should be done? In my view, the current Justice Department merger guidelines are inadequate to deal with network airline competition. By focusing on city-pairs and national markets, they are simultaneously too narrow and too broad. Creating duopolies in affected nonstop city-pair markets ought not to be the goal of merger policy. Counting remaining networks as one by one gets picked off without considering the degree of effective rivalry that the survivors can offer is equally shortsighted.
By considering mergers only one at a time, DOJ’s policies allow eggs to be scrambled without gaining an understanding of the recipe of which they are a part. The Department should focus here on the fact that effective national network competition requires adequate traffic density to support competitive frequency and roughly comparable network scope, either directly or through alliances, to be able to meet competitively the transportation needs of a significant portion of the national market. Justice may wish to argue that the East Coast assets of US Airways and TWA when combined with existing American and United assets create a barrier to entry in national network competition because of the impossibility without them of building a network of the size and scope necessary to compete with the American and United that will emerge from this transaction.
American and United will achieve Big Two dominance by using together network scope much larger than their competitors and an access lock on the Northeast markets that serve as a source and destination for much of the nation’s traffic. If these transactions go ahead, it is to be expected that as the remaining firms fall behind, they will try to combine their Northeast strengths and then try to offset their Northeast weakness relative to American and United by building a national network even larger in scope. The ultimate result would be an industry with three network competitors, two jointly dominant and one large and struggling, instead of six. DOJ should develop a merger test, at least for network businesses, that allows or requires them to look at and take into account both the implications for network competition and the probable further transactional consequences of mergers before them. If it doesn’t believe it can do that within the current statute, the Congress should help them by making clear that DOJ has the right and obligation to do so.
I have already indicated why I don’t think non-network discount airlines like Southwest represent a “cure” for the competitive illness this would produce. And I don’t think that American and United should be given a “failing company” license to dismember US Airways. US Airways is a troubled, but not failing, company. Its management and labor have elected to sell out at a good price to a duopoly that will pay off shareholders and maintain above-market labor arrangements rather than to accept the hard choices that would be necessary to make their airline healthy again. It’s the public that’s expected to pay the bill.
In short, we have reached a critical point. We can either preserve competition among four or five or six network competitors, none of which have the potential of achieving a level of dominance which makes the others unable to compete. Or we can turn the deregulated airline industry into the preserve of two powerful airlines who have pulled up the ladder of access to the East Coast and who expect to watch their rivals fade away as they struggle to overcome an impossible competitive disadvantage. The authorities and the Congress should not stand by while this happens.