For the past five years, the U.S. government has enthusiastically supported airline alliances, and has granted them immunity from the antitrust laws so that they can engage in activities that would be unlawful in this or any other industry. This immunity has been provided because the Department of Transportation perceives that the public benefits of alliances outweigh the substantial damage to competition caused by antitrust immunity. It is time to reexamine that policy.

TWA believes that alliances can provide some public benefits in the form of enhanced interline codeshare service and increased connecting opportunities for consumers. These benefits usually occur in markets that have limited traffic flows and do not justify nonstop service. However, these limited benefits are more than offset by the anticompetitive effect of the combined operation of the alliance carriers on major routes. The elimination of competition by such joint operation does far more damage to the public than the limited benefits received from enhanced interline service.

The government should, therefore, reexamine the issue of antitrust immunity for airline alliances. The first step should be to examine the basic legislation to determine if there is any need for a statute that singles out the airline industry for special treatment under the antitrust laws. The Airline Deregulation Act of 1978 abolished antitrust immunity with respect to interstate air transportation. It is time to extend that principle of the Deregulation statute to international air transportation.

Even if the statute is allowed to remain on the books, the Department of Transportation should remove immunity from existing alliances as their current immunity periods expire, and it should not grant any new immunity applications. It should also limit codesharing on transatlantic segments. Finally, it should either forbid joint frequent-flier programs, or require that they be open to any carrier that desires to join them.

In this paper, we will review the historical development of airline alliances, describe in detail the anticompetitive nature of the alliances, and examine the adverse effects of alliances on consumers and competition. We will also examine the implications of international alliance experience for the proposed domestic megacarrier alliances. Finally, we will discuss the proposed actions which the government should take to eliminate the abuse of airline alliances.

Airline codeshare agreements have developed from very limited circumstances almost 30 years ago into the vast corporate combinations that are operating or proposed today. The first codeshare agreements were established in the late 1960s when Allegheny Airlines, a local service carrier, desired to terminate service to small communities that did not justify operation with jet aircraft. It signed agreements with operators of small aircraft under which they replaced Allegheny, painted their aircraft with Allegheny's colors, and adopted Allegheny's ticketing and customer service procedures. Their flights operated under Allegheny's airline code and connected to Allegheny at its hubs. These commuter codeshare agreements have evolved so that every major carrier has a commuter affiliate providing codeshare service at each of its hubs.

International codeshare agreements originated in the late 1980s as simple linear route extensions, under which a carrier placed its code on the flight of another airline so that it could hold out online service to a destination to which it did not actually operate. The primary impetus for the creation of these agreements was computerized reservation systems. Schedule displays in the systems gave priority to online services, which, therefore, were more likely to appear on a the first display screen of a travel agent or airline reservationist. Thus, in order to achieve prominence in scheduled displays, airlines agreed to share their codes to create the appearance of online service. This was inherently deceptive for travel agents and consumers, but DOT allowed it, provided that the carriers made disclosure of the codeshare operation to the passenger, and sought a Statement of Authorization from DOT.

Early international codeshares were limited to a few routes at a time. One of the first major agreements was between United Airlines and British Airways, which allowed BA to place its code on United flights beyond United's Chicago hub to Denver and Seattle. None of these early agreements involved codesharing in primary transatlantic gateway markets. Given the interline nature of the connecting transportation, carriers could establish joint fares without violating the antitrust laws.

The success of the early agreements led to larger combinations, including proposals for operation between major transatlantic points. In 1992, the United States was highly desirous of achieving its first significant European "Open Skies" agreement with the Netherlands. As a condition of the agreement, the Netherlands required that the U.S. grant antitrust immunity to a codeshare agreement between Northwest and KLM. In its zeal to obtain the "Open Skies" agreement, the U.S. agreed, and issued the final order granting immunity during the last week of the Bush administration. The grant of antitrust immunity corrupted the very concept of Open Skies from the outset. Moreover, it naturally led to similar requests for immunity for other alliances, which are clearly more anticompetitive than the Northwest/KLM consortium, neither of which were major transatlantic carriers. In order to achieve Open Skies with Germany, the Department of Transportation granted immunity to the United/Lufthansa alliance. And Delta similarly received immunity for codeshare agreement with Swissair, Sabena, and Austrian Airlines. In November 1998, the U.S. reached an Open Skies agreement with Italy that is contingent upon the grant of antitrust immunity to Alitalia and whichever U.S. carrier it chooses as a partner.

Anticompetitive alliances reached their apex in the agreement between American Airlines and British Airways, which has been submitted to DOT for approval. The Justice Department has said that this agreement violates the antitrust laws. The European Commission (EC) has found that the agreement violates European competition law, and refuses to approve it unless British Airways gives up enough slots at Heathrow to allow for competitive access. The EC has also proposed several marketing conditions that would limit the power of the alliance to exercise its market power over travel agents and corporations. The UK Office of Fair Trading has found that the agreement violates UK monopoly law, and has proposed conditions similar to those proposed by the European Commission. Yet, the Department of Transportation is willing to consider approval of the alliance, subject to conditions, if the British will agree to Open Skies. Of course, Her Majesty’s Government wants the alliance to be covered by antitrust immunity. However, not even Open Skies would be adequate to ensure competition in the U.S.-U.K. market when Heathrow, the major European airport, is capacity restricted. TWA believes that it would be far more beneficial to competition if the alliance were disapproved in its entirety.

Alliance agreements have several common elements that, except for antitrust immunity, would be per se violations of the antitrust laws:

1. Price-fixing -- The carriers agree on the prices to be established for both public tariffs and off-tariff contract fares. In the proposed American-British Airways agreement, American would establish the prices on behalf of both carriers for itineraries originating in North and South America, and British Airways would set prices for itineraries originating in Europe, Africa and Asia. The parties would establish joint corporate fare discounts and set joint traffic thresholds that corporations would have to meet to be eligible for such discount fares.

2. Revenue pooling and profit-sharing -- One of the most anticompetitive elements of international aviation under the old IATA system was revenue pooling, under which carriers on a route shared revenue without regard to which airline the passenger used. This practice totally sapped all competitive energy from the market. Airline alliances have reestablished revenue pooling. The American/British Airways alliance would go further and allocate profits from the entire transatlantic operation of both carriers to each airline on the basis of a formula.

3. Capacity agreements -- The alliance carriers jointly plan schedules and agree upon the capacity that each carrier will operate.

4. Fix travel agency commissions – Prior to deregulation, the entire industry set joint travel agent commissions, with immunity granted by the CAB. After deregulation, this was one of the first agreements from which antitrust immunity was removed. Now, however, international carriers are free to establish the level of commissions that will be paid to travel agents for the majority of passengers in major international markets. With immunized marketing programs, they can also agree on which travel agents will be favored with special fares or tour packages.

5. Codesharing – While the initial justification for codesharing was to hold out interline service as online service in CRSs, alliance codesharing is different. Alliances also offer codeshare service on nonstop routes where both carriers operate. By placing its code on the alliance partner's flights, each carrier can appear to offer more service in the nonstop market than it actually provides. Moreover, the duplicate display of nonstop flights, by design, crowds services offered by other airlines out of CRS displays. While offering codeshare service in connecting markets may provide some limited public benefit, duplicative codeshare service on nonstop transatlantic sectors can only deter competition.

6. Joint frequent-flier programs -- Since 1981, frequent-flier programs have been a major competitive device to encourage repeat business. They are especially beneficial to large carriers who have greater scope of operations, and can offer passengers the opportunity to collect points on flights to more destinations. By joining frequent-flier programs, alliance carriers magnify this competitive advantage over smaller airlines and increase their protection against new entrants.

While alliance proponents talk in terms of online service, in reality, alliance agreements are pervasive combinations between competitors that result in the establishment of government-sanctioned joint monopolies on major transatlantic routes.

The alleged justifications for codeshare alliances are that they extend route networks of U.S. carriers at limited costs, and provide enhanced interline service that is equivalent to online service. However, such gains are marginal because consumers already have enhanced interline connections under the traditional IATA system. In fact, alliances have had an adverse impact on competition, consumers, and the U.S. flag system.

While codeshare participants gain improved position in CRS displays, they have also provided some improvements in the connecting experience in order to justify claims that the services are actually "online." Alliance carriers may issue through boarding passes, and may work to tighten connecting times. However, these alleged benefits of codeshare service are also available on an interline basis. For decades, carriers have worked with each other to improve connecting services, and have offered joint fares through the IATA system. They have also, through IATA, agreed upon formulas for sharing joint revenue, and have often established separate "prorate" agreements to achieve the same result. Thus, the actual improvement in connecting service offered by codesharing is limited to non-existent.

To achieve these limited gains, the public must pay an enormous amount in the form of lost competition. For example, before the United/Lufthansa alliance, both carriers competed on the Washington – Frankfurt route, with each trying to gain market share from the other by offering better service and lower fares. Now however, the carriers provide the same volume of service, but each places its code on the other carrier’s flights. Because they pool the revenue, it does not matter to them which airline the consumer chooses. They may actually carry more traffic, due to diverting passengers on other airlines to their "online" service beyond Frankfurt, but they no longer need to compete for passengers in the nonstop market. Because of the loss of competition, consumers in that market are less well served.

Moreover, the ability of major airlines on a route to join together enables them to keep other carriers out of their markets. Their ability to combine frequent-flier programs and offer frequent-flier points for services beyond the European gateway makes passengers less likely to choose other carriers on the route. By combining together, they can set higher thresholds for corporations to qualify for corporate discounts. For example, a corporation might be required to provide United and Lufthansa all of its Washington – Frankfurt traffic to obtain any corporate discount on that route as well as others. The same process could apply to travel agent overrides. United and Lufthansa could set extremely high thresholds to qualify for any override commission.

Alliance carriers argue that their ability to raise prices on joint gateway routes will be disciplined by the availability of alternative routings – that passengers will have lower fares available over other gateways. For example, passengers from Washington to Frankfurt may find a low fare over Paris or London, if United/Lufthansa raise prices. This argument is more theoretical than practical. It is certainly not true for the proposed American/British Airways alliance -- alternate gateways on the European continent are usually not price or time competitive for London traffic because of the substantial backhaul involved. Even for routes such as Washington – Frankfurt, alternate gateways impose extensive delay costs. Business passengers who require expedited service will not take an alternate routing even if fares are $100 or $200 lower. Only some leisure traffic may be able to consider the competitive alternatives of gateways that provide inferior service. In any event, why shouldn’t such passengers have the choice of competitive nonstop service between Washington and Frankfurt?

Moreover, codeshares are inherently deceptive. While DOT requires disclosure of the operator of the codeshare flight, many passengers believe they are flying on a U.S. airline, and are surprised when they board a foreign aircraft. Governmental policy should not be premised on practices that amount to consumer deception.

Alliances have undoubtedly been beneficial to their U.S. flag participants. Rather than incur the expense of directly operating international service, the U.S. flag carrier receives a commission from the foreign alliance partner for every passenger it puts on a codeshare flight. A recent Merrill Lynch study found that "alliances have been even more successful than originally anticipated, and are increasingly significant for one simple reason: they are very effective at redirecting traffic to the benefit of the alliance members." However, diversion of traffic from one carrier to another is not necessarily in the public interest, particularly when the traffic is being diverted from U.S. airlines to foreign flag carriers. Such diversion weakens the effectiveness of U.S. carriers and eliminates jobs for employees in the U.S. airline industry.

It is clear that airline alliances have already driven U.S. carriers off some transatlantic routes. The Chairman of American Airlines testified before a U.S. Senate committee that the market power of the Star and Delta/Swissair alliances have driven American from three nonstop routes between the U.S. and Continental Europe -- New York-Zurich, Miami-Frankfurt, and New York-Brussels. Similarly, TWA withdrew from the New York-Zurich and New York-Geneva markets after Delta and Swissair began codesharing in 1994. In early 1998, TWA canceled its New York-Frankfurt service in the face of the growing market power of the Star alliance.

An analysis of frequency growth on the Atlantic demonstrates that alliances have benefited foreign flag carriers at the expense of U.S. airlines. Because of strong traffic and the shift of equipment from larger B-747's to smaller equipment, total frequencies in the U.S.-Europe market increased 42% between 1995 and 1998. However, U.S. carrier frequencies increased only 28%, while foreign flag frequencies increased 53%. The U.S. flag share of frequencies has dropped from 42% to 38%. In the U.S.-Germany market, where United and Lufthansa implemented their Star alliance during this period, there has been no growth in total frequencies. However, U.S. flag frequencies decreased 33%, while foreign carriers increased frequencies by 31%. The U.S. flag share of total frequencies to Germany has declined from 48% to 32%.

Alliances have clearly been great for foreign carriers. U.S. alliance partners have generated increased income by diverting traffic from other airlines and by carrying passengers in conjunction with foreign alliance partners, rather than operating their own flights to secondary European points. However, they have been extremely damaging to other U.S. airlines and have cost jobs for U.S. labor. Thus, alliances are not only anticompetitive, but also, on balance, damaging to U.S. commerce and employment.

The success of international alliances in diverting traffic from other airlines has led to a pairing up of major U.S. carriers. American and US Airways have announced an alliance and begun a joint frequent-flier program. Northwest has purchased voting control of Continental and proposes to operate joint domestic codeshare service with it. Delta and United’s proposed alliance has been stymied for the moment by the opposition of Delta’s pilots union, but the carriers have proceeded with pooling of their frequent flyer programs. Based on the experience with transatlantic alliances, these combinations raise serious questions whether smaller carriers can survive. The impact of alliance joint frequent-flier programs and joint corporate discounts on the domestic market will be enormous, and will be compounded if these carriers are allowed to codeshare in domestic markets.

Northwest and Continental have already closed their transaction, despite the suit by the Department of Justice against it. If DOJ is successful, Northwest may be forced to divest itself of its interest in Continental. The Department of Transportation is examining the integration of Northwest’s and Continental’s frequent-flier programs, but there is no assurance that it will act. If Northwest and Continental are successful, the other major alliances will be forced to respond. The result will be a domestic industry in which 81 % of the traffic is controlled by an oligopoly of three large alliances. At a time when DOT is attempting to protect new entrants and to foster competition in the domestic market, it is most curious that the government would even consider permitting these combinations.

It is imperative in the interest of preserving competition that the government reexamine the entire issue of airline alliances. Their limited public benefits are overwhelmed by their damage to competition, other U.S. flag carriers, and U.S. labor. There are several steps that the government should take:

1. The U.S. should repeal 49 U.S.C., § 41308, which authorizes the Secretary of Transportation to grant exemptions from the antitrust laws. There is no longer any justification for continuation of that statute. The section was originally enacted in the Civil Aeronautics Act of 1938, during the Depression, when the U.S. government was justifiably concerned about protecting fledgling U.S. airlines from the rigors of competition. The U.S. airline industry has long since grown up, and the public policy justification for this sort of special privilege evaporated decades ago. The Airline Deregulation Act of 1978 abolished antitrust immunity with respect to domestic air transportation. Since international air transportation was not deregulated at that time, the power to grant immunity in this area was allowed to continue. Now, however, there has been substantial deregulation in international air transportation, as the United States has reached Open Skies agreements with over 40 countries. It is time to finish the job and eliminate DOT’s power to grant immunity in international air transportation.

2. Whether or not the statute is amended, DOT should remove antitrust immunity from existing alliances. In past approvals, DOT has granted immunity for five-year periods. Renewal of immunity for the Northwest/KLM alliance is presently before the Department. It should be denied. DOT should, of course, refuse to grant immunity to American and British Airways. It should also issue orders to show cause why immunity should not be removed from the Star alliance and the Delta alliances, since these alliances are no less harmful to the public interest. It should also remove the immunity that was improvidently granted to the transborder alliances between United and Air Canada, and American and Canadian International Airlines.

3. DOT should forbid codesharing on transatlantic gateway routes. The alleged benefits of codesharing arise from extension of U.S. carrier systems by placing their codes on flights beyond the European airline’s hub. This benefit can be achieved without the two carriers placing their codes on each other’s flights on transatlantic segments. With modern CRS displays, a Delta flight can be shown as a through service from New York via Zurich to Moscow, without placing its code on the Swissair flight in the New York – Zurich city pair display.

4. The Department should prohibit joint frequent-flier programs, joint corporate discounts, and joint travel agent overrides. All of these programs are designed to take advantage of the size of the alliance carriers, rather than to compete on the merits of airline service. These marketing devices provide no public benefits and should be disallowed. If, nevertheless, alliance carriers are allowed to operate merged programs, they should be required to allow competing carriers to join the merged program on reasonable, nondiscriminatory terms. Failure to do this will leave in place merged frequent-flier programs which will have become anti-competitive and barriers to entry.

Unless the government takes action, it will be faced with a future airline marketplace in which enormous carrier cartels have driven non-alliance airlines out of markets, where foreign airlines dominate the Atlantic to the perpetual detriment of consumers, and where U.S. labor has lost thousands of high-paying jobs. The U.S. should take action now to protect consumers and competition from the vast monopoly power that has been vested in the existing alliances and proposed for the new ones.

(1) This paper focuses on the Atlantic because it is the major area in which immunized alliances are currently operating. Transborder alliances to Canada also have immunity, but there are no immunized alliances on transpacific routes or between the U.S. and South America. The principles set forth in this paper would apply to immunized alliances in those areas, also.

(2) Merrill Lynch, Global Airline Alliances, 25 September 1998, p. 2.

(3) Written Testimony of Robert L. Crandall before the Senate Judiciary Subcommittee on Antitrust, March 19, 1998, p. 8.

(4) OAG, April, 1995 and 1998.