us-ftc-advocacy-filings-6bae168b3c9d71be08352e755042d2c6•FTC Staff Comment Before the Department of Transportation Concerning Proposal to Give Limited Antitrust Immunity to Discuss Scheduling Overlaps at Congested Airports
FTC Staff Comment Before the Department of Transportation Concerning Proposal to Give Limited Antitrust Immunity to Discuss Scheduling Overlaps at Congested Airports
us-ftc-advocacy-filings-6bae168b3c9d71be08352e755042d2c6FTC Staff Comment Before the Department of Transportation Concerning Proposal to Give Limited Antitrust Immunity to Discuss Scheduling Overlaps at Congested AirportsAdvocacy Filing
..J C ~~ f; r:' ,~' ~ ill ........ "" V ;'k':':- ~tV; Ii ~ f;:, f7 ~~ >:t.~. Ynt:"~uV~~Un~ BEFORE THE A PPR 0 V'~ ~DEPARTMENT OF TRANSPORTATIOP i;;,D;'WASHINGTON, D.C. ~" DISCUSSION AUTHORITY FOR AGREEMENT TO SHIFT SCHEDULES ) ) ) ----------------) Docket No. 44634 Comments of the Bureaus of Economics, Competition, ard Consumer Protection of the Federal Trade Commission Introduction In these comments, the Federal Trade Commission staff addresses a Department of Transportation (DOT) proposal that air carriers be given limited antitrust immunity to discuss scheduling. overlaps at congested airports. We urge that this proposal not be implemented. We believe that scheduling overlaps • could be eliminated more efficiently if the DOT imposed limits on takeoff and landing operations as needed, and then made the resulting operating rights, or ·slots,· marketable through cash sales. We suggest that this approach would be more consistent with the Federal Aviation Act's preference for market-oriented solutions, and would better serve the interests of the flying public. 1 These comments represent the views of the Federal Trade Commission staff, and do not necessarily represent the views of the Commission itself or of any individual Commissioner. The Commission has, however, voted to authorize the Bureaus to submit these comments.
The Federal Trade Commission (WFTC Wor wCommission W) is an independent regulatory commission, created in 1914 by the Federal Trade Commission Act, and empowered to prevent unfair methods of competition and unfair or deceptive acts or practices in or affecting commerce. One of the Commission's principal responsibilities is enforcement of the antitrust laws. Government regulations that create barriers to entry or that weaken competition are of particular concern to the Commission. The Commission's Bureaus of Economics, Competition, and Consumer Protection have a continuing interest in policy issues involving airport access, including the allocation of slots to airlines at restricted airports. FTC staff research has led to the publication of a report on slot allocation,2 and the FTC staff has also participated in earlier ·administrative proceedings raising many of the same issues involved here. We filed written comments before the Civil Aeronautics Board in 1984 in response to Eastern Airlines' request for antitrust immunity,3 and 2 3 The Report presents evidence that competition in air transportation markets would be encouraged and that consumers would receive significant net benefits if restrictions on the transfer of slots are removed. See D. Koran and J.D. Ogur, Airport Access Problems: Lessons Learned from Slot Regulation by the FAA, Bureau of Economics Staff Report to the Federal Trade Commission, May 1983 (FTC Staff Report). Such restrictions were ended by DOT in April of 1986 (Docket No. 24105, 50 Fed. Reg. 52180). See Application for Discussion Authority and Prior Board Approval of Carrier Agreements to Integrate Schedules, Docket No. 42410, Answer of the Bureaus of the Federal Trade Commission (August 27, 1984). -2-
1- , . participated that same year in response to an FAA notice" of proposed rulemaking concerning airport delays.4 The current proceeding is similar to the 1984 Eastern Airlines request. Both proposals share the common goal of reducing the frequency of delayed flights. 5 The tentative DOT plan, like the Eastern Airlines request, would grant air carriers an antitrust exemption to discuss scheduling overlaps as a temporary means of reducing airport flight volume during congested periods. Adoption of the DOT proposal is likely to harm both competition and consumers. To avoid these undesirable effects, we offer the following recommendations: 1) DOT should not grant antitrust immunity for carrier schedule discussions: 2) instead, DOT should adopt the less anticompetitive and more efficient alternative approach of imposing slot limitations whenever 4 5 Federal Aviation Administration, Notice of Proposed Rulemaking Regarding Elimination of Airport Delays, 49 Fed. Reg. 33082 (1984). According to statistics complied by the FAA, over 417,000 flights had a departure or arrival qelayed by more than 15 minutes in 1986. About 75 percent of the total delays for 1986 (313,568) occurred at ten of the nation's largest airports. Moreover, according to the Official Airline Guide schedule tapes, carriers have scheduled operations at a much higher level that FAA Engineered Performance Standards would indicate are possible at Atlanta, Boston and Newark airports. S7e DOT Order 87-1-54, pp. 1-3. -3-
necessary to eliminate congestion, and then making these slots transferable by expanding the existing cash market. 6 Both restrictions on slots and their subsequent marketing have been successfully implemented in the past. Slot restrictions on total operations per hour were originally imposed at four airports in the late 1960's. In response to the air traffic controllers' strike in 1981, the FAA decreased the number of slots at the four original locations and imposed slot restrictions at 18 additional airports. In 1982, the FAA modified the program by permitting implementation of a barter system for the trading of slots at all slot-constrained airports. 7 Shortly thereafter, it permitted cash slot sales on an experimental basis for six weeks. During that period over 190 slots were bought"and sold. As ATC system capacity slowly increased following the PATCO strike, restrictions were removed at most airports. In 1984, the FAA imposed slot-like restrictions on operations during specified time periods at certain major airports. Two years later, DOT permitted the buying and selling of slots at the four original slot-restricted airports. We believe that more extensive use of 6 7 Our recommendations follow from our responsibilities to promote competition and the interests of consumers. We recognize that there may be other concerns of public policy which should be taken into account in a proceeding such as this one. However, Order 87-1-54 made no finding that safety is affected by the current situation. By Rslot-constrained R we mean that there are legal restrictions on the number of takeoff and landing operations that are allowed to be scheduled. -4-
this market approach will be a better solution to airport congestion problems than DOT's proposed solution. 8 The Proposal The DOT proposal arises out of the apparent scarcity of air traffic control (-ATC-) capacity at several u.s. airports at certain times of the day. In other words, at the prevailing (zero) price for the use of ATC services, airline demand sometimes appears to exceed supply. The Order observes that delays have recently increased sharply and asserts that in certain peak periods airline scheduling has exceeded airport capacity. (Order, pp. 1-3). This excess demand results in the imposition of congestion costs on carriers and travelers. In its Order, DOT proposes to grant antitrust immunity to carriers to allow them to engage in multilateral discussions for the purpose of integrating their schedules in order to reduce the number of flights during congested periods and thus reduce the incidence of delayed flights. The Order covers thirteen airports, ten of which are not now slot-constrained. The grant of antitrust immunity would apply to the following airports: Atlanta, Boston, Chicago O'Hare, Denver, Minneapolis-St. Paul, 8 An alternative solution would impose a congestion charge on carriers during peak periods to equate marginal private cost and marginal social cost. See Borins, -Pricing and Investment in a Transportation Network: The Case of the Toronto Airport,- Canadian Journal of Economics, November, 1978. -5-
New York LaGuardia and Kennedy, Newark, St. Louis, San Francisco, Dallas-Forth Worth, Philadelphia, and Washington Dulles. 9 The DOT's proposed plan would place both substantive and procedural limitations on any multilateral agreement reached by carriers. Carriers would not be authorized to determine the maximum number of flights that could be operated within any specific time period. Rather, their discussions would be confined to spreading flights over a longer period of time or to distributing them more evenly within current peak periods. Carriers could not discuss fares and other pricing matters; nor could they discuss services in individual city-pair markets. All scheduled carriers serving an airport and any prospective entrants would be permitted to participate. Notice ·of ~eetings would be given to all·carriers serving an affected airport and to the airport proprietor, the Department of Justice, the FAA and the DOT. Meetings would be opened to all interested parties, including government agencies, and reports listing the attendees and summarizing the discussions for all meetings would be filed with the DOT. In order to better facilitate government oversight, the meetings would be held in the Washington, D.C., metropolitan area. The Problem and Its Solution Airport congestion is an example o~ a ·common property" resource problem, which has the essential element that no one 9 DOT Order 87-1-54, p.4. -6-
entity controls access to a resource (in this case, the ATC system) that is available only in fixed supply. As a result, competing users (air carriers) overuse the resource. This causes an inefficiently high level of congestion. IO The end result is that each airline imposes costs on other airlines. ll For example, each carrier with an aircraft in a queue waiting to take off imposes congestion costs on carriers with aircraft further back in the queue and on the passengers in those other aircraft. An efficient solution to a common-property resource problem includes setting a limit on the use of the resource and allocating the use rights to parties who value those rights most highly. The DOT proposal embodies one method of making this allocation. It suffers, however, from two disadvantages. First, 10 11 When we speak in this comment of reducing congestion, we do not necessarily imply that DOT should attempt to reduce congestion and delays in all cases to zero. Such a course might not be in the best long-term interest of consumers. While eliminating the last increment of congestion might be a benefit to them, the costs to them of attaining it, including the costs of expanding ATC capacity or of eliminating desired flights in order not to exceed current capacity, may well exceed those benefits. We will accordingly speak here of the optimal level of congestion, which weighs all these factors and maximizes net consumer welfare. Put differently, the optimal level of congestion exists when carriers are paying all the costs they create, inclUding the congestion costs they impose on other carriers. In other words, the marginal social cost of each use is greater than its marginal private cost. Private users make decisions regarding use based on marginal private cost and thus overuse the resource from society'S point of view. -7-
it may create market power, and second, it may lead to an inefficient allocation of operating rights. 12 Anticompetitive Effects The first drawback to the proposed plan is the risk that an agreement might produce anticompetitive results. The carriers could use the planned discussions as a forum for reaching agreements to reduce capacity, which would enable all of them to reap higher-than-competitive profits, at least as long as DOT acted to enforce the agreement. 13 12 13 As a further threshold problem, the carrier discussions provided for in the Order may not produce agreement in any event. Evidence to support this concern is provided by the experiences of the scheduling committees at the high density airports and the committees set up to deal with flight delays in 1984. After deregulation opened entry into air transportation markets, committees at the high density airports experienced increased difficulty in reaching agreement. In some instances the result was total stalemate. In 1984, under the threat of an imposed shifting of flights by the' FAA, the carriers were finally able to reach agreements, but only after extensive discussions. Because the committees created by the DOT Order would operate under the same decision rule (that of voluntary agreement or unanimity) and because the possible enforcement role of the FAA is less clear than it was in 1984, the proposed committees may experience similar difficulties. In the event of stalemate, the committees would obviously provide no solution to the problem of congestion. The DOT Order does not clearly address the question of enforcement. Private enforcement of agreements would probably be ineffective, however. Each participating carrier would have an incentive to cheat on the agreement by restoring flights that it had dropped, since this would increase that airline's profits. Other airlines would have an incentive to enter with new flights. While such cheating might be easily detected by the airlines that were faithful to the agreement, they would lack DOT's power to punish the violators. -8-
The proposed DOT Order addresses this potential danger, by providing that carriers may not discuss or agree on the maximum number of flights in any time period. It is not clear, however, that this prohibition can be effectively enforced. 14 Any process of reducing peak-period congestion necessarily implies shifting some flights to other, adjacent, and less desirable time periods. In turn, the supply of flights in the ~ore desirable peak periods is reduced. To the extent that off-peak travel is not a perfect substitute for peak-period travel, this situation may permit airlines holding peak-period slots to raise prices. If coordination is perfect, then the number of flights will be 14 We recognize that the procedural limitation which DOT would impose on carrier discussions are intended to strengthen its prohibition against agreements setting the maximum number of flights. Should Order 87-1-54 be issued, we would encourage vigorous oversight of carrier discussions by DOT and other government agencies. However, for reasons set forth in the text, we nevertheless believe that anticompetitive results would be difficult to avoid, due to the inherent nature of the situation and despite the best efforts of the DOT. The order acknowledges the problem where it states that DOT is -reluctant to authorize carrier schedule discussions since they could affeqt airline competition.- Order 87-1-54, p.4. -9-
reduced to the monopoly level, and each carrier will share in the monopoly level of profits. ls In sum, scheduling discussions could enable carriers to exe"rcise (or at least move in the direction of) monopoly power, something that they are otherwise unlikely to accomplish. The result would be fewer flights than necessary to reduce congestion to the optimum level, and an increase in fares. 16 The loss to consumers from higher fares and fewer flights could outweigh the current costs of congestion-related delays. Any market power exercised by the committees will be augmented if the newly limited operating rights cannot be freely bought and sold. In the FTC Staff Report (pp. 17-25), and in · 15 16 This result follows most easily if each carrier maintains its relative market share unchanged. In that case, each carrier cancels the same number "of flights," a result that may be easier to achieve than one involving unequal sacrifices. With the same share of the larger total group profits, individual carrier profits will also be higher. If shares are unequal, coordination may be less than perfect, but still good enough to achieve supracompetitive profits. For example, if carriers give up the same percentage of their original (differing) numbers of flights, each carrier can earn higher profits. In any case, the requirement that agreements be voluntary gives each carrier the power to veto any agreement that does not increase its profits. The incentive of some carriers to restrict flights below the level that would reduce congestion efficiently at one airport is constrained by the ability of other carriers to add flights at competing airports. For example, carriers at O'Hare would not increase their profits as much by restricting flights, if other carriers could readily add flights serving the same cities fro~ Midway, or through alternative hubs. The ability to add flights at O'Hare during off-peak hours further constrains the incentive to engage in monopolistic restriction of flights during peak hours. However, because off-peak flights and flights using other airports are probably less than perfect substitutes for peak-period flights, carriers will probably have the ability to raise fares above the efficient level to some extent. -10-
previous comments before the FAA (Slot Transfer Methods, Docket No. 24105, Comments of the FTC Bureaus, pp. 6-12), we have shown that barriers to entry are greatest where each newly created slot is allocated to a specific carrier and cannot be transferred to another carrier at all. Carrier entry barriers will be lower, but still present, if the FAA permits the slots to be transferred through barter but not for cash. Barter systems are an imperfect solution because, in contrast with cash sales, new entrants will be forced to incur supracompetitive costs for the slots they need. Costs will be supracompetitive because the transaction costs of barter arrangements are higher than for cash purchases, and because, in some instances, a more valuable slot must be exchanged to obtain a slot necessary for entry into a particular city-pair market (see FTC Bureaus Comments, pp. 8-10). The net result is that carriers already in that market may be able to raise fares above the competitive level without inducing entry. Inefficient Allocation The second drawback to the proposed plan is the risk that the carrier agreements will lead to an inefficient allocation of operating rights. Unless those newly restricted rights can be bought and sold, the flights that remain after reducing the total number may not be those of greatest val~e to consumers. Assume for the sake of illustration that two carriers have ten flights each during a time period when ATC capacity can handle only sixteen flights efficiently. Even if congestion is -11-
reduced efficiently at a level of sixteen flights, thrs outcome is optimal only if the four cancelled flights were the lowest valued of the original twenty flights. If carrier A has three flights that are lower in value than carrier B's least valuable flights, for example, an optimal reduction would cancel those three carrier A flights and only one of carrier B's flights. However, this result will probably not be reached by voluntary agreement between the carriers. Carrier A is likely to insist on preserving its 50 percent market share by sacrificing only two flights. If A could receive payment in return for surrendering an additional flight, the efficient result would more easily be achieved. 17 Because carrier B can provide a more valuable flight than carrier A, B will be willing to pay A more for its third slot than A can- earn using -that slot for its relatively low-valued flight. Hence, A will sell its slot to B and cancel its third flight, while B will cancel only one flight. Both carriers would gain from the transaction, congestion would be reduced, and consumers would receive the flights they value the most. However, unless the existing slot market is extended to the ten airports covered by the proposed order that are not currently slot constrained, such desirable results will not be possible. • 17 In the absence of the possibility of such payments, this situation is equivalent to two rival firms, one less efficient than the other, attempting to maximize joint profits without side payments. See F.M. Scherer, Industrial Market Structure and Economic Performance, Rand McNally, 1980, pp. 159-160. -12-
A Procompetitive, Efficient Alternative Proposal The procompetitive, efficient solution to airport congestion is for DOT to impose additional limits on the number of slots during peak periods at the affected airports as needed,18 and to permit carriers to buy and sell the restricted slots. The anticompetitive effects and inefficiencies of the proposed DOT plan would be-avoided because carriers would not meet in committee, and DOT would not be involved in enforcing carrier agreements. The benefits of permitting slot sales ·are illustrated by comparing the results of the FAA's administrative allocation after the Professional Air Traffic Controllers (PATCO) strike with the allocation that a free market in slots would have achieved. In the FTC Staff Report (pp. 10-14), we performed this comparison for certified carrier slots at St. Louis." Some cost to consumers was unavoidable after the strike, because the FAA reduced the number of certified carrier flights to St. Louis by 27 percent. This reduction was needed due to the decreased capacity of the ATC system. FTC staff estimated the unavoidable loss due to the reduction in the number of flights at $12 million per year. However, according to our estimates, an additional consumer loss of approximately $4 million per year resulted because the FAA's administrative methods of allocation gave slots 18 At the three airports subject to the DOT Order where slots are already limited, further restrictions may entail tightening the time period during which a slot is usable. For example, slots that are now usable for takeoffs between 9 and 10 a.m. at O'Hare might be restricted to (say) a 15- minute period during that hour. -13-
to some relatively low-valued flights, and certificated carriers were unable to switch those slots to more highly valued flights. Because St. Louis was one of 22 slot-constrained airports at the time of the strike, the total consumer loss due to the persistence of low-valued flights was some multiple of $4 million. 19 If slot sales had been permitted, consumers would have been spared this loss. We have not estimated the benefits of the slot market that has been in operation since April 1986. However, we expect that such benefits have been and will continue to be substantial. An example of the competitive benefits of the existing slot market is the entry of the Pan Am shuttle between New York and Washington, which was facilited by the purchase of slots in the cash market. Long-Term vs. Short-Term Solutions Because the proposed DOT plan calls for a prompt solution to current concerns regarding airport congestion, it is appropriate to address whether slot restrictions and a slot market can be implemented quickly. The evidence indicates that both slot restrictions and a slot market have been implemented in the past without great delay. For example, the FAA rapidly expanded slot restrictions from four to twenty-two airports in response to the PATCO strike in September 1981 (see SFAR'No. 44-1, 46 Fed. Reg. 19 These estimates are for the period immediately after the PATCO strike when FAA slot restrictions were the tightest. Estimates for a period of lesser scarcity would be smaller, other things equal. -14-
44424). This indicates that the-extension of slot restrictions is a feasible short run response to congestion problems. There is also evidence that the existing slot market can be extended rapidly to handle transactions in these newly created slots. In May 1982, the FAA temporarily removed restrictions on slot sales (SFAR NO, 44-3, 47 Fed. Reg. 19989). In the following six weeks, carriers bought and sold over 190 slots without apparent difficulty, and despite the simultaneous existence of a slot barter alternative (FTC Staff Report, pp. 9-10). In April 1986, the FAA permanently ended restrictions on slot sales at four airports to permit a cash market (Docket No. 24105, 50 Fed. Reg. 52180). Between that time and mid-December 1986, nearly 700 slots were exchanged in uneven transfers. 20 Extensive experience thus suggests that a slot market solution to congestion problems can be implemented quickly. The Legal Test for Antitrust Immunity Finally, a program of transferrable slots, such as we have suggested, is the approach best in keeping with the philosophy of the Federal Aviation Act. Any decision on the proposed plan must be judged in light of section 412(a) (2) (A) (i) of the Act. That section states in pertinent part as follows: 20 An example of an ·uneven· transfer is a trade of 3 slots by carrier A for 1 slot of carrier B. Such transfers, which often involve cash on the side, are recorded by the FAA. Data on cash transactions are not broken out by the agency. -15-
[T]he Board may not approve ••• any ••• agreement • • • which substantially reduces or eliminates competition, unless it finds that the • • • agreement • • • is necessary to meet a serious transportation need or to secure important public benefits ••• and it does not find that such need can be met or such benefits can be secured by reasonably available alternative means having materially less anticompetitive effects. DOT should thus ensure that the most procompetitive and market- oriented means are employed to alleviate airport congestion. Indeed, in reviewing section 412 of the Act in the context of a joint operating agreement, the Eighth Circuit stated that antitrust immunity wis the exception and not the rule.- Republic Airlines v. CAB, 756 F.2d 1304 (8th Cir. 1985). Moreover, in 1982 the Civil Aeronautics Board itself recognized that there was a wcl ear presumption w that competition can be relied on to serve the public interest. 2l Such a presumption reflects the national policy favoring competition that is embodied in many other statutes, including 21 ~ generally CAB Docket No. 36595, Investigation into the Competitive Marketing of Air Transportation -- Agreements Phase, Slip Ope at 25 (served December 29, 1982). -16-
·. .: the antitrust laws. Restrictions on output and higher prices are the core target of the antitrust laws. 22 Modern antitrust jurisprudence also recognizes that some restraints and joint activities may enhance competition. The Supreme Court, for example, has stated that a joint selling arrangement may be procompetitive if it increases sellers' aggregate output,23 and noted that a restraint in a limited aspect of a market may actually enhance marketwide competition. 24 These exceptions are not applicable here, however. Carrier negotiations respecting non-transferrable time . slots could unnecessarily restrict output, exacerbate barriers to entry, and lead to higher prices for consumers. Such risks need not be accepted when, as here, a less restrictive and more practical alternative is available. 22 23 24 Senator Sherman, for one, sought to act against: [c]ombinations and individuals ••• forming a league and covenant • • • with power to suspend the production of some and enlarge the production of others, and absolutely control the supply of the article which they produce • • • • 21 Cong. Rec. 2457 (1890). Modern antitrust jurisprudence adheres to this same durable standard: A restraint that has the effect of reducing the importance of consumer preference in setting price and output is not consistent with this fundamental goal of antitrust law. NCAA v. Board of Regents, 104 u.S. 2948 (1984) (citation omi ttted) • Id. -17-
The most practical alternative is the use of transferable slots. Such an approach is most beneficial to consumers, and is just the sort of market-oriented solution that section 412 was intended to encourage. Conclusion In 1986, DOT/FAA permitted cash sales of slots at the four high-density airports. In its current order, DOT seeks to reduce congestion at thirteen airports, including ten that are not now slot-constrained. We recommend that the FAA impose additional slot restrictions as needed, but extend the benefits of market competition to all airports that require slot restrictions. As shown in the FTC Staff Report, permitting a market for slots is a procompetitive approach that would yield significant net benefits to consumers. By contrast, the proposed DOT Order is likely to yield an inefficient and possibly anticompetitive outcome. For these reasons, we recommend that the DOT not grant antitrust immunity for carrier schedule discussions. -18-
Wondering how this text applies to your situation?
Free consultation, no commitment
Describe your situation. Our team will contact you by email about your consultation.
Continue your research in ChatGPT or Claude
Connect Omnilex to search the legal corpus from your AI assistant.