Why International Alliances Are Good for You

Cranky is on vacation, but I’ve lined up some excellent guest bloggers for you while I’m gone. Today I have Aviation Policy Blog author Evan Sparks.

One of the big stories in aviation thus far in 2009 is how international airline alliances have come under heavy fire on Capitol Hill. The “big three”–Star, Skyteam, and Oneworld–have added members and expanded their global reach in recent years. “I have become increasingly concerned with the decline of competition in international markets, particularly between the United States and Europe,” said Jim Oberstar (D-Minn.), the powerful chairman of the House Transportation Committee. “Combined, the . . . alliances account for almost 80 percent of the total world airline capacity, 78 percent of world revenue passenger kilometers and 73 percent of passengers carried. These three alliances control over 87 percent of the traffic between the United States and Europe.”

Back in February, Oberstar introduced legislation to limit the power of these alliances. The international alliances offer more than just codesharing and interline connections. They cooperate in marketing, scheduling, and route planning, and (with an exemption from antitrust regulation granted by the federal government) ticket pricing. The key partners in Skyteam (Delta, Northwest, KLM, and Air France) and Star Alliance enjoy immunity on transatlantic routes. Just last week, the Department of Transportation tentatively approved Continental to join United, Air Canada, and Lufthansa in fare-setting immunity. American, British Airways, and Iberia, the core carriers of Oneworld, are seeking an antitrust exemption themselves.

Oberstar’s legislation–which, according to Aviation Week, was recently inserted into the House version of the FAA reauthorization bill–would constrain these international alliances by limiting the grants of antitrust exemptions to three years, after which the airlines would have to reapply for another three years. This would insert politics into airline regulation decisions. The actual effect would be to discourage airlines from seeking to form alliances, period. Unpredictability of policymaking can sometimes harm businesses even more than bad policy. “Having the sort of ex post facto law that would undo this is a very disconcerting principle to have in the business community,” US Airways executive C. A. Howlett said recently.

By now, you may be thinking that airlines in alliances collude on prices to gouge the little-guy traveler, and that’s certainly the sort of rhetoric you hear coming out of Washington. But that’s not what studies of international codesharing indicate. In a 1998 paper [PDF here], Jan K. Brueckner and W. Tom Whalen found that “alliance partners charge interline fares that are 18-28 percent below those charged by nonallied carriers”–although they found a slight but statistically insignifant uncompetitive effect on flights between “gateway” cities. Why? “First, joint pricing internalizes the negative externalities from the uncoordinated choice of subfares, leading to lower fares. This fare reduction then stimulates traffic, which in turn lowers marginal costs via economies of traffic density, leading to further downward pressure on fares. Our results, which confirm this prediction, are extremely robust to changes in the measurement of competition and the handling of market-specific eff ects.”

Basically, it works like this: Delta sets fares on the Atlanta-Louisville route on its own, as does its alliance partner KLM for the Amsterdam-Bratislava run. Then, the immunized partners collaborate (or collude) on pricing the Atlanta-Amsterdam run. If they didn’t have immunity, they wouldn’t be able to collaborate and individual fares would stand. But because KLM needs passengers on the transatlantic flight from Louisville and because Delta wants to send its passengers on to Bratislava, they need to set an overall fare that makes sense for both carriers in the market. If Delta set its fare on ATL-SDF too high, it would depress feed traffic for the transatlantic flight and thus hinder the joint efforts of both airlines. “Cooperative behavior, however, internalizes the externality, so that a lower fare is chosen.”

Brueckner’s 2003 article in the Review of Economics and Statistics, “International Airfares in the Age of Alliances: The Effects of Codesharing and Antitrust Immunity,” [PDF of early version here] updates his earlier model and finds that “the presence of antitrust immunity reduces the fare by 13-21 percent,” and that the “combined effect [of codesharing and immunity] ranges between 17 and 30 percent.” And what would happen if, as Oberstar wants, alliances were to risk losing immunity every three years? “The estimates show that, when the DOT grants antitrust immunity, substantial benefits arise for an alliance’s interline passengers. . . . [T]he aggregate loss [modeled for Star Alliance passengers] from withdrawal of immunity ranges between $17 and $22 million per quarter, suggesting that the annual gain to the STAR alliance’s interline passengers from the presence of immunity is on the order of $80 million per year.”

Bottom line: airline alliances don’t drive up fares. They do the opposite. Given that alliances actually smooth the travel experience, allowing greater ease in baggage transfer, terminal transfer, through ticketing, and other amenities, it is striking that they actually reduce fares in most cases. Why would policymakers want to change that? Well, that’s a question for another post.


Evan Sparks writes the eponymous Aviation Policy Blog at www.evansparks.com. He lives in Washington, D.C., where by day he is an editor at the American Enterprise Institute. Evan’s writing on aviation policy and other subjects has appeared in the Wall Street Journal, Forbes.com, and The American.


10 Responses to Why International Alliances Are Good for You

  1. The Traveling Optimist says:

    The economic double-talk always throws me, so thanks, Evan, for illustrating the language with the DL/KL example.

    My question though, stems from this example. If you say DL sets ATL-SDF on its own but sets it too high then it depresses the ATL-AMS portion where DL and KL collude? I thought the example would be that DL sets ATL-SDF wherever it likes but both airlines collaborate on ATL-AMS so one doesn’t outprice the other.

    It seems that, without ATI on this route, independent pricing (which Oberstar wants back) would leave the “external negativity” in place of DL and KL pricing the overwater independently of each other. If DL is higher one day, KL gets the business and vice versa, but the customer is still served because, with KL and DL in SkyTeam, the mileage and ground services remain coordinated.

    Did I misunderstand the economic double-talk or am I missing something else here?

  2. David says:

    I need more time to think through your argument over ticket pricing being reduced…. but you seem to make little mention that mutual codesharing allows an increase in apparent schedule frequency which is of benefit to the little guy flying on a non-full fare ticket but who still wants to fly at a particular time of day on what may be a relatively thin route.

    Has this aspect been debated at all on Congress, or am I making a fundamental error in microeconomics ?

  3. Carla says:

    This definitely got me thinking. I was completely unaware that the government was starting to get involved in this.
    My favorite part of the alliances is being able to earn FF miles for using partner airlines.

  4. enplaned says:

    Sounds a little too simple. There’s more than one type of alliance — The original KL-NW alliance was pretty clearly beneficial for all concerned because it was “end-to-end” — there was not much overlap between the KL and NW route structures, so there was not much elimination of competition when KL and NW joined hands — in fact the opposite.

    Since this was the only meaningful alliance prior to 1998, it’s no surprise that a 1998 paper showed that alliances were beneficial — but it was really a statement saying specifically the KL+NW alliance was beneficial.

    On the other hand, more recent alliances have more overlap, including LH+UA and especially the still-born AA+BA. In these alliances there’s a meaningful amount of potential elimination of competition. These alliances have been more prevalent recently. It would be interesting to redo the paper today — I would bet it would show a reduction in benefit (or even negative benefit) from alliances today.

    At the end of the day you have to ask why alliances are so popular. Most of the significant airlines have sought the shelter of an alliance. Now you’re saying that the effect of this is to reduce the fares received by airlines in alliances?

    Seems unlikely, companies generally seek to do things that increase revenues, whereas reducing fares is generally consistent with reducing revenues (at least for the stodgy, traditional airlines that seek to be part of alliances).

    Some skepticism seems warranted here.

  5. MathFox says:

    I agree that codesharing and other ways of opening the reduced fares of several airlines in one trip are a direct economic benefit to the passenger. I add through-checking of luggage and a single check in for the trip as another benefit to the passenger.
    The question that remains is whether there is enough competition between the three alliances (and the independents) to keep prices low. At the moment I don’t think airfares can get much lower with many airlines losing money. However, with only three alliances having 87% of the transatlantic market my gut feeling is that we are close to the point where (informal) price collusion could be effectuated.

  6. Alex says:

    I think its a great idea, despite the studies you cited (and by the way its not really a surprise that US airways executive doesn’t want any government oversight) i think the only real way to test if allowing more price collusion is to allow the airlines to do it and then see the results. But to suggest that there shouldn’t be any oversight i think is irresponsible. I think this does need looking at in 3 years, to sit down and see if the average customer (both business and leisure) is better or worse off and adjust the regulatory framework as required.

    Didn’t we learn anything from the banks? Sure its not a bad idea now given the state of the economy but many of the airlines (for example BA/AA) are rushing this through using the economy as the excuse. In reality there proposal hasn’t changed much since the last time the regulators told them it was bad for consumers but it looks like they will get anti trust due to the downturn.

    The last thing the consumer wants is airlines that come out of the downturn in more monopolistic situations than before. I don’t see the problem with telling them that you can collude on prices for now but we want to review it when the economy has recovered to see if it really is good for consumers.

  7. The Traveling Optimist says:

    Let’s look at Australia.
    With three global alliances there are really only two that compete to that corner of the world, Star (UA/NZ) and oneworld (QF/AA/BA). SkyTeam has no local traffic support from New Zealand or Australia now that V-Blue is feeding traffic to V-Oz.
    The current economy, the newest Virgin product and NW re-entering the market via the DL takeover have all colluded to suppress fares to the lowest levels ever.
    Across the Atlantic, oneworld is heavily UK oriented but with reasonable pressure from Star via United and the fact that the UK is all but a coach bucket market thanks to the high volumes, again, two alliances here seem sufficient for now.
    As far as beyond traffice flowing over the competing alliance hubs, Carla’s point about FFY miles being the base driver of consumer choice will continue to make pricing competitive. Take South Africa.

    The example would be ORD-JNB. AA/BA via London, a choice with Star of UA/SA via IAD or UA/LH via FRA or a SkyTeam option of DL direct, DL/AF via Paris or DL/KL via AMS.

    Tons of other examples out there but the bottom line is, if FF miles help drive your decision, you’re covered. If you don’t care about mileage but want low fares there remains a lot of metal to choose from for your end-to-end travel. At least 12 carriers operate between the US and South Africa either directly or via Europe or the Middle East, with or without alliances to capture your business.

    In today’s economy there’s not much worry about prices. In a strong global market, right about three years from now as Oberstar suggests, that may be a different story. Adjusted for inflation I’d say prices will still be affordable and competitive.

    There may be only five US carriers flying to Europe but there are currently 11 flag carriers operating between the US and South Africa directly (2) or via Europe and the Middle East. Plenty of choices out there, alliances or not.

  8. The Traveling Optimist says:

    Bad day at the editing office. My apologies.

    The 12 carriers between the US and South Africa that I found are:

    Direct: SA and DL
    Europe: AF, AZ, BA, IB, KL, LH, OA and SR
    Middle East: Emirates and Etihad
    Pacific (If you wanted to take the long way): CX, MH, QF and SQ

    If I missed anyone please share and thanks!

  9. Pingback: Rebuttal: Ending Airline Competition on the North Atlantic is Bad For You >> The Cranky Flier - Airlines|Airports|Air Travel Blog

  10. JBM says:

    Optimist – for east coast departures you missed ET, MS, SV and LY. Europe you missed VS and TP. The bigger point is that with reasonable circuity between the US and South Africa, there are literally dozens of intermediate connect points that discipline economy fares.

    Business fares are another story. Schedule plays a bigger role, as does product, which is why BA/VS/LH can drive significant premiums over second-tier options. It’s also why SAA had to switch the JFK flight back to a nonstop – high-yield fares were leaking to European carriers even though the SAA option was direct.

Join the Conversation