Sunday, 16 July 2006

USA proposes to reduce IATA's exemption from antitrust law

I've spent the last week with a severe summer cold and maybe a touch of the flu -- just the thing to put me in the best or worst frame of mind, depending on your perspective, to analyze the details of several long, arcane legal and regulatory documents of significance to travellers that have appeared in the Federal Register and my e-mail in-box. I'll try to catch up shortly.

On 5 July 2006 the USA Department of Transportation issued a proposal to eliminate the long-standing exemption from USA antitrust law of the International Air Transportation Association (IATA), insofar as it pertains to joint airline "traffic conferences" to fix industry-standard airfares between the USA and Europe or the USA and Australia. (If the proposal is put into effect, IATA would remain exempt -- at least for now -- from USA rules against acting as a price-fixing cartel with respect to airfares on routes everywhere else in the world, including for example between the USA and Asia, Africa, and Latin America.)

I've been among the most vocal critics of airline oligopolies and IATA's antitrust exemption.

But as I've said previously:

Exemptions from antitrust law to permit airlines to fix prices and routes together... have gotten at least some critical scrutiny. But they are far from the worst of the government policies in favor of airline oligopolies and against the interests of air travellers, especially in the USA.

On its face, the DOT proposal sounds like a good thing: Won't prices for airline tickets be lower if airlines aren't allowed to collude to fix industry fares? That's the conclusion the DOT comes to, or so they say. Unfortunately, it's not that simple.

R. Paul Marcelin-Sampson, who was the first to bring the DOT "Order to Show Cause" to my attention, got it exactly right:

The Order to Show Cause says that interline arrangements are likely to remain in place without IATA involvement. This may be true, but the Order definitely emphasizes airline alliances. Speecifically, the Order implies that today's covert, airline-determined codeshare itineraries are preferable to yesterday's overt, customer-determined interline itineraries. I am worried about itineraries that cannot be accomplished conveniently, cheaply, or at all -- within a single airline alliance.

The DOT makes several key mistakes in its analysis:

  1. The DOT proposes to eliminate the USA antitrust exemption for price-fixing discussions between all IATA member airlines, while retaining the antitrust exemptions -- entirely unquestioned and unexamined in the DOT proposal -- that the DOT has already issued to members of the largest global airline marketing alliances. The result would be to enhance the oligopolistic power of the major alliances, and the airlines in the USA that participate in them, to the detriment of airlines that aren't part of these alliances, and to the detriment of travellers wanting more choices and lower fares. The consequences would likely be worst in those markets where the members of a single alliance between themselves have sufficient market share to exercise oligopoly power, such as "Oneworld" marketing partners American Airlines and British Airways come close to doing between New York and London.

  2. The DOT compares apples and oranges in evaluating the impact on consumers and travellers of the "normal" interline fares fixed by IATA "traffic conferences". (Despite errors later in the document, both the introductory section of the DOT proposal and IATA's rules for traffic conferences , which were inserted in the DOT docket, provide a useful primer in the concept of "interlining" -- issuance of a single ticket for travel involving multiple airlines -- and the IATA price-fixing system of traffic conferences.)

    The DOT recognizes that what it calls "the IATA Interline Product [normal industry or "YY" mileage fares allowing unlimited interlining] is likely to have significant value only for those travellers who have relatively complex itineraries that involve multiple airlines and a number of stopovers" (p. 40 of the DOT proposal). But instead of comparing those industry fares set by IATA with airline-specific fares for the multistop journeys for which they are used, the DOT compares them with point-to-point carrier-specific fares (without stopovers, or with limited stopovers) between the endpoints of the journey. Not surprisingly, DOT concludes that airline-specific fares from A to Z without stopovers are typically less than the IATA fare from A to B to C to ... to X to Y to Z.

    Duh. Stopovers cost money. So what? The relevant comparison, for anyone actually using such a fare, would be between the IATA through fare (including all the stopovers) and the sum of the carrier-specific fares for each of the sectors needed to construct the entire journey. In almost all cases with a sufficient number of stopovers, the through IATA normal fare is less than the sum of the lowest carrier-specific sector fares. Indeed, the only situation in which an adequately-informed traveller would purchase a ticket at an IATA fare is when it is less costly than any available carrier-specific fare or combination of fares.

    The DOT claims that, "the IATA fare product in many markets is much more expensive than the unrestricted fares offered by individual airlines for equivalent levels of service" (p. 40). But the relevant "market" and basis of comparison is the stopover market, particularly the multiple stopover market, in which even cursory comparison quickly shows that the DOT's assertion is utterly unsupportable.

  3. Rather than being overused, at higher prices, to travellers' detriment, IATA fares are grossly underused, to travellers' detriment, because neither airlines nor the pricing software used by computerized reservation systems typically suggests most routings or pricing that qualify for IATA fares, even when they would be more direct, more efficient, and/or less expensive than carrier-specific routes and fares. The DOT claims (p. 24 of the DOT proposal) that "the computer systems operated by the global distribution systems, which are used by travel agents and by the on-line travel agencies accessed directly by consumers, automatically construct and price interline itineraries. In doing so, they combine the best available fares and create airline itineraries that typically do not rely on the IATA Interline Product." But in fact, as I've previously shown in analyzing examples from The Amazing Race , GDS's/CRS's typically don't display or price optimal interline routings . Indeed, a major reason for the fraudulent airline labeling practice of codesharing is to game the CRS/GDS bias in favor of online routings and fares.

  4. The DOT misrepresents (probably because it misunderstands) the rules of carrier-specific fares in its claims that "the IATA Interline Product sits side-by-side with competing services offered by individual airlines that are comparable to the IATA product" (p. 33 of the DOT proposal); that "Neither IATA tariff conferences nor a similar industry group has ever established interlineable fares for routes within the United States or in the U.S.-Canada/Mexico transborder markets, yet travellers can readily obtain interline service in those markets" (p. 39); and that "airlines offer comparable products in the domestic U.S. market and U.S. transborder markets even though no IATA tariff conferences exist for these markets" (p. 40).

    In reality, carrier-specific point-to-point fares -- virtually all of which limit stopovers, interlining, transfers, and/or routings -- are neither comparable to nor competitive with IATA mileage fares that allow unlimited stopovers and interlining. The DOT commits one of the typical errors of inexperienced travel agents in failing to realize that even if the stopover rule of a fare allow "unlimited" stopovers, that means only that stopovers aren't limited to a specific number. All of the other rules of the fare -- such as those restricting the allowable routings, transfers, interlining, and in many cases flight number restrictions -- must also be complied with if the fare is to be applicable, significantly restricting which stopovers, and how many of them, are actually possible within the permitted routing(s) and other rules.

    Carrier-specific fares allowing unlimited interlining and stopovers and "mileage" routings are extremely rare in any market, and entirely unheard of within and between the USA, Canada, and Mexico. Yes, as the DOT says, there are carrier-specific fares within the NAFTA zone that allow some interlining with specified airlines on specified routings, but they are in no way comparable to fares with unlimited stopovers and interline transfers and mileage routings. I defy the DOT, or anyone else, to show me a published carrier-specific fare between, say, New York and Los Angeles that allows unlimited stopovers and interline transfers on an otherwise-unrestricted mileage routing. It just doesn't exist.

  5. Clearly betraying its bias toward the major global airline marketing alliances, the DOT asserts without any basis that an airline marketing "alliance ... should enable ... travellers to obtain transportation that would be more integrated than interline services offered by unaffiliated airlines" (p. 7 of the DOT proposal) and that "alliance interline services ... enable travellers to obtain service that can be comparable to on-line service and that offers passengers significantly more flexibility than did traditional interline service between unaffiliated airlines" (p. 41). It's not clear what the DOT could possibly mean by this, since a ticket that can be used (within the mileage limits, and subject to the other rules) on any of the flights of several hundred airlines is manifestly more flexible than one that can only be used on the member airlines (ten or twenty, at most) of a single marketing alliance. And because of the degree to which codesharing (which is most common within alliances) misleads travellers, interline connections between members of the same alliance are often much more problematic than properly-disclosed and honestly labelled interline connections between unrelated airlines.

    On my most recent trip to the East Coast, for example, I was on an American Airlines flight from San Francisco to Boston with several people who had originated in Seattle, and who had almost missed their connection in San Francisco because the codeshare labelling of the Alaska Airways flight from Seattle with an American Airlines flight number had hidden the fact that Alaska and American operate from different terminals -- the two most distant ones from each other -- at SFO. Had their flight been properly labelled, as it would have been between unrelated airlines, the interline connection would have been apparent, and they would have been better prepared for the necessary race to make their connection.

  6. Despite conceding (p. 6 of the DOT proposal) that many airlines are drastically reducing the numbers of other airlines with which they have interline ticketing agreements (i.e. whose tickets they are able to accept) -- a trend with, as I've previously detailed, severe negative implications for travellers -- the DOT says, without specifying its rationale, that "We doubt that our proposed termination of antitrust immunity for the IATA by-laws for U.S.-Australia/European Union markets could significantly reduce the availability of interline service" (p. 39) and that "growth of alliance services in these markets has reduced consumer needs for traditional interline service" (p. 41). The DOT implicitly, but incorrectly, assumes that passengers ticketed for online travel have no interest in interline agreements, or only in those between members of the same alliance.

    In fact, the main use of interline ticketing agreements -- particularly between airlines that aren't members of the same marketing alliance or a code-sharing agreement -- is to accommodate each other's passengers in the event of flight cancellations, delays, missed connections, overbooking, and the like. Airline alliances utilize their antitrust exemptions (which would be retained under the current DOT proposal to eliminate the IATA exemption) to coordinate schedules, typically beginning with the elimination of duplicate flights. If one member of an immunized alliance has to cancel a flight, or a late inbound connection causes a passenger to miss it, it is especially unlikely -- as a direct result of the alliance antitrust exemption -- that there will be an available alternate flight or set of connections not too much later on another alliance member airline. So minimizing the delays and inconvenience to passengers who have to be rebooked or rerouted is especially likely to depend on the existence of the widest possible array (as has traditionally existed, but as is now being cut back rapidly) of non-alliance interline agreements.

The complete docket will eventually include all the comments received.

If you want to voice your opinion:

  1. Go to the comment submission Web page.
  2. Enter "25307" in the "Docket ID" box.
  3. Choose "OST" [Office of the Secretary of Transportation] from the "Operating Administration" pull-down menu.
  4. Click either the "Enter a Comment" or "Attach a File" box next to "Submission Method".
  5. Fill in as much of your personal information as you like. (Keep in mind that whatever you fill in will be posted on the Web with your comment.)
  6. Click "Continue". Depending on which option you chose, you'll either get a box to type in your comments (4000 characters maximum), or a page to upload a file of comments of any length that you've prepared on your own computer (in text, PDF, RTF,TIFF, Wordperfect, or MS-Word format).

To be considered, your comments must be received by 5 p.m. Washington, DC, time on Friday, 18 August 2006.

[Update, 18 August 2006: At IATA's request, the DOT has extended the deadline for comments until 5 p.m. Washington, DC time on Friday, 20 October 2006.]

[Addendum, 20 October 2006: I've filed comments with the Department of Transportation restating what I said in this article.]

Link | Posted by Edward on Sunday, 16 July 2006, 17:39 ( 5:39 PM) | TrackBack (0)
Comments
Post a comment









Save personal info as cookie?