Should United De-Hub Los Angeles? It Already Has.

United

There has been a consistent drumbeat regarding United for years in the analyst community: it has too many hubs and needs to close at least one. This mantra is flaring up again with talk that United should look at closing its hubs at Washington/Dulles and Los Angeles. Regarding Los Angeles, that’s sort of a foolish statement. It’s already happened.

Whether United wants to call Los Angeles a hub or not is irrelevant. The reality is that over the last 10 years, United has dramatically shrunk its operation there and shed a lot of flights which were more attractive to connecting passengers. A friendly party with a mi.diio.net subscription helped me out with the schedule data so you could see just how much has changed. I compared schedules (including mainline and regional operations) from July 2016 to those of July 2006, 10 years earlier. In 2006, I included Continental with United for comparison purposes. I did the same with Northwest and Delta, America West/US Airways and American, and AirTran and Southwest to put this chart together.

Percent Change LAX Flights Seats

If that doesn’t just say it all right there…. Overall, United scheduled 45.7 percent fewer flights this past July than it did in July 2006. It also had 26.2 percent fewer seats. Why the difference? Well many of the cuts have been on the regional side where fewer seats are impacted. But no matter how you look at it, this is a substantial decrease. At the same time, it has eliminated 22 nonstop destinations from LAX while adding 11. Here’s that list.

Dropped Routes Added Routes
Aguascalientes Ontario Austin
Bakersfield Oxnard Bozeman
Carlsbad Portland Hilo
Durango (Mexico) Philadelphia Jackson Hole
Guatemala City San Salvador Melbourne (Australia)
Imperial St George Minneapolis
Inyokern San Jose Oklahoma City
New York/JFK Santa Maria Shanghai/Pudong
Morelia Orange County Puerto Vallarta
Modesto Yuma San Jose del Cabo
Oakland Zihuatanejo Vancouver

To see the trends, I think it’s helpful to put this into a map. Thanks to Great Circle Mapper, here are the routes United has dropped.

United Abandoned LAX Routes Since 2006

And here are the routes United has added.

United Added LAX Routes Since 2006

What you notice is that the routes that have gone away fall into two categories. They’re either short regional routes that were full of connections, or they were Latin America routes. Meanwhile the routes that were added are either leisure routes or international, for the most part. But let’s get more specific here by region and see how capacity has shifted.

Intra-California
When I was growing up, United was a strong west coast airline. Regardless of where the headquarters was, people in the west considered United their own. That continues to be the case in San Francisco but it has been largely dismantled in both the Pacific Northwest and in Southern California.

From LAX to airports within California (excluding to the SFO hub), flights are off an incredible 80.8 percent with seats down 72.7 percent. Of the 22 eliminated markets from LAX, exactly half are cities in California. Some of this is because there just isn’t a good airplane for United to use to serve these markets. For years these were the bastion of the Embraer 120 turboprop. But once those were retired, it was 50-seat-jet or bust.

Excluding San Jose and Oakland, two cities that are no longer served because United just couldn’t compete with Southwest, these markets were nearly entirely connecting markets. These were full (or not full) of people flowing through LAX to fly elsewhere. In the case of the absurd flights from Ontario and Orange County, there couldn’t have been a single local traveler.

Even the markets that have kept LAX service have seen dramatic cuts. Fresno is down from 6 or 7 a day to 2 or 3. Palm Springs is down to only 1 flight a day. In markets like that, it still has to be about the connection, but it’s just such a minor presence now.

Intra-West
This isn’t just a California problem. It’s also an intra-West problem. In those markets, United flights are down 64 percent with seats down 60 percent. St George (Utah) and Yuma (Arizona) behave like California markets. They were Embraer 120 cities that disappeared when the fleet left. The only other market to go away is Portland, which I still find incredible.

But you don’t have to kill a route entirely in order to cut capacity. Tucson had 5 flights a day. It now has 1. Albuquerque had 3 flights a day. It now has 1. It’s the same story in all these markets where frequencies have collapsed. Vegas is down from 7 to 3, Seattle is down from 4 to 2. The only new market here is Vancouver, which is probably related to the strengthened partnership with Air Canada. To me, these look like markets that used to have a schedule designed for local and connecting travelers. Now, the semblance of a schedule that remains doesn’t really serve either group well.

Latin American Business/Ethnic Markets
It’s important to break out the business/ethnic markets in Latin America from the leisure markets because they cater to different travelers. On the business/ethnic side, United has effectively walked away. Flights are down 62 percent with seats down 60.9 percent. And 5 destinations are no longer served. It’s now just León and Mexico City. This is a market that United has simply decided not to bother playing in anymore.

Hub Markets
I pulled other United hub markets out of the geographic categories, because they should behave differently. After all, if LA becomes less of a hub, then it should need more connectivity into the other hubs, right? Apparently not.

Hub flights are down 5.4 percent and seats are down 4.2 percent. (I included Cleveland in here since it used to be one.) But really, this is a tale of shifting hubs. Denver service has been decimated (flights cut in half, seats down a third) while SFO has shrunk as well. The only hub that has gone up is O’Hare… except for Newark. Newark, however, is a special case because when United stopped flying to JFK, it shifted that over. So it better have grown.

Leisure Markets and Hawai’i
When it comes to leisure travel, United has a split personality. Hawai’i flying is down, a lot. That actually surprises me. United has 30.3 percent fewer seats on 27.2 percent fewer flights to the islands.

But in other leisure markets, United has grown flights by 127.7 percent with seats growing similarly. What’s going on there? Well there are new markets like Puerto Vallarta, Cabo, Bozeman, and Jackson Hole. Other markets like Aspen and Cancun have been strong growers as well.

Without looking at the data, my guess is that Hawai’i is down because connecting options are down while competition is up. Meanwhile these other markets survive more on local traffic and have less competition.

Transcon Flying
If we forget about JFK, which went away in favor of Newark, then United serves a mere 2 transcontinental markets. Philly is gone, so that leaves Baltimore (seasonally) and Boston. (Orlando is considered a leisure market, though I know there is Disney business that you could argue should have it elsewhere.) Those existing markets are down.

Cats and Dogs
I wasn’t really sure how to categorize these flights to the middle of the country, but I see this as a list of cats and dogs. Existing markets have shrunk (Dallas, New Orleans, San Antonio) but three new markets have been added (Austin, Minneapolis, Oklahoma City). For these specific new markets, my assumption is that they have a very specific need for the local market.

So, has anything grown? Yes! It’s the international world. United has added Melbourne and Shanghai. Meanwhile it has kept Tokyo, London, and Sydney, albeit with small airplanes. Melbourne isn’t really “new” service in that United used to just serve Melbourne via Sydney on a tag flight. It’s just now decoupled thanks to the 787. As for Shanghai, I can’t believe it’s still flying. United either wants to hold on to the authority or it has some big corporate clients that need it.


Ultimately, United has decided to shrink itself in LA dramatically. I imagine there is room to shrink more. Can one flight in Tucson and another in Albuquerque really do well enough to bother keeping them around?

United can’t de-hub LAX, because it’s already done it. The LAX operation looks like one that’s largely designed for the local market, even though it’s hard to see how United can serve the area as well as Delta or American now. Maybe the question analysts should be asking is… should United shrink Los Angeles further?

Get Cranky in Your Inbox!

The airline industry moves fast. Sign up and get every Cranky post in your inbox for free.

81 comments on “Should United De-Hub Los Angeles? It Already Has.

  1. The question is why would United abandon one of the largest air travel markets in the world? Especially when it had the natural advantage of being one of the first to lay down stakes. Obviously, the numbers have been crunched and they believe that they can’t be profitable at LAX. Why ihas United decided to build up SFO instead? The Bay Area is also a very competitive market.

    1. MarylandDavid – The Bay Area is far less competitive where United is the only legacy carrier with a dominant position by a long stretch. In LAX, it’s everyone fighting for the same thing. LAX is also a worse hub geographically and it has a ton of Asian carrier capacity that’s sold on the cheap.

  2. Has the drop in UA seats/flights matched the gains from all other airlines? By the look of your graphic one would think DL has built up a fortress hub in LAX when I know that’s not the case. Coastal cities have always perplexed me as “hubs” for domestic travel. If I’m east of the Rockies why would I go to LAX to connect to a destination that I flew over already? This is where DEN makes sense to me as a hub. Totally understand the massive O/D market in LA and the international flights but the traditional hub and spoke seems much better suited for major market mid-continent locations. That said, looking at the dropped flights it looks like they cut under performing routes to small cities and leisure destinations. I’m assuming a lot of the drop was in seat miles to destinations they still serve.

    1. A – Just looking at the big four (so not including the massive growth from Virgin America, Alaska, and all the foreign carriers), flights are up 1.6% with seats up 17.9% and available seat miles up 28.1%. So the other guys have replaced United and more.

      Specifically United flew 8,468 flights in July 2006 with 855,572 seats. In July 2016 it flew 4,596 flights with 631,413 seats. Delta flew 2,015 flights with 380,181 seats in July 2006 with 5,132 flights and 749,797 seats in July 2016. American flew 5,771 flights in July 2006 with 675,762 seats. In July 2016 it flew 6,893 flights with 934,560 seats.

      1. Thanks for clarifying. That begs the question: why hasn’t another legacy carrier attempted to build up SFO to compete with United?

        I realize that this is inviting thread drift….

        1. davidp627 – Well Virgin America has. But SFO was an overcrowded, expensive airport for years. And United dominated. It’s a lot easier to start a hub in a place where nobody dominates. At SFO, it would be an uphill battle.

  3. How big is Delta compared to UA today at LAX? The change in service graph is interesting, but could be misleading if the amount of service is an order of magnitude different between the carriers.

  4. Interesting analysis. Do you not count IAD-LAX as a transcon market? They are doing 7/day on that route.

      1. A little hub history…

        When hubs were growing after deregulation, CO was at LAX, UA was in CLE & both were in DEN. Fast forward a decade & a half & CO leaves DEN & UA trades CLE for LAX with CO. Once UA leaves CLE, the hub reopens at IAD. When CO chose to grow along the east coast, they placed their hub at EWR buying a small airline in the process who’s name I cant recall at the moment.

          1. Ah yes – thanks I thought that was it. In fact, CO almost left EWR. This was about 15-years ago & look at it now.

      2. My thought is that IAD counts in the UA hub category rather than the transcon category. Thus EWR doesn’t appear either.

    1. CP – OhioExile is correct. I view hubs differently, so those are all in their own category. They should behave differently than any other market, so it’s worth carving out.

  5. another great topic, Cranky.

    Regarding the question upstream, Delta this month has 12% more flights and 18% more seats than United at LAX. The terminal swap at LAX should provide space for Delta to continue to grow.

    United is doing at LAX what American is doing at JFK: trying to say they can be a niche market carrier in a city where other carriers have much larger hubs. In most other hubs, the evidence shows you can’t compete in someone else’s hubs outside of to your own strength markets (usually other hubs) if you aren’t overall equal to or larger than your competitors. We have yet to see how the competitive environment will play out at JFK and LAX but I expect that AA’s hub at JFK will shrink just as UA”s will at LAX because both will increasingly be left to discount local traffic in markets where other carriers can fill a flight with local and connecting passengers. That is exactly what happened at JFK for UA where UA had no feed on the JFK end but AA, B6 and DL all did. VX faces the same challenge on its transcon routes which are point to point flights competing with other carriers’ hub flights.

    It is also likely that the Haneda route case which United supported will result in the need to cut international service at both LAX and IAD. At LAX, UA is trying to operate its own Narita flight against 3 Haneda flights, one of which is from its partner NH. NH also got the only route to the NE from HND which means that UA’s EWR and IAD to NRT flights will be competing for a shrinking base of NRT local traffic while the more lucrative passengers will go to Haneda. In time, perhaps by 2020 and the Tokyo Olympics, all premium Tokyo flights will move to Haneda. NH’s (and UA under the JV) larger share of the US-Haneda market will result in the DOT attempting to balance competition by giving new HND routes to AA and DL when the Japanese open up more slots at HND. Reducing UA’s presence to Tokyo from LAX and IAD will weaken their position in the overall market

    On the west coast, UA has taken the strategy of hubbing just at SFO which is a great market and they have a very powerful hub there. LAX and SFO did compete against each other for flow traffic which is part of why they had to reduce LAX. However, given that AA and DL are both committed to building LAX and other carriers are simply adding back what UA is cutting, the effect will be that UA will be weaker on the west coast. Add in that if DL’s dual hub strategy at LAX and SEA works, and the Port of Seattle has to pour a lot of money into that airport in order for SEA to grow, UA’s SFO hub could remain the most powerful on the west coast but be sandwiched between DL at LAX and SEA and AA at LAX. And of course, neither of them are likely to displace WN in flying all of the point to point routes on the west coast or in connecting all of the secondary west coast airports to the US any better than WN does. and as WN adds short-haul international routes on the west coast and eventually gets around to Hawaii, UA will gain a competitor even in the markets where it has been able to hold onto a share of the LAX market.

    1. I agree the similarities between UA at LAX and AA at JFK are striking. Who knows if it will work, but in reality those are probably the only two cities (maybe Chicago) where you can have a reasonably comprehensive O&D only focus and make money, nowhere else has a chance.

      Remember that NH & UA are in a JV, they are allowed to coordinate routes and compare data legally, so any changes to TYO will be informed between them and serve the JV as a hole, for the all important financial side, it doesn’t matter which metal is flying.

      1. While from a financial standpoint, JVs share revenue, there are very different impacts on employees. If NRT flights have to be cxld as local, more valuable Tokyo local traffic is pulled to HND, there will be a shift in capacity operated by US carriers and their Japanese partners because the same-sized pot is split between 2 Japanese but 4 US airlines.

  6. Does it really matter if UA has cut flights or other airlines have added flights, or does it matter more if those flights are making money?

    If UA only had 20 flights out of LAX and all 20 make money and say DL has 50 flights but only 19 made money, who is really better off. I’d rather have one flight between LAX and TUS as an example that made money then have 6 flights to compete, but only one still made money.

    1. David,
      AA DL and UA all have very similar average fares at LAX with UA being a little bit higher – likely because they have dropped a lot of west coast flights. The big 3 have very similar cost structures now so the chances are high that LAX is profitable for all 3 and for WN. Even though WN’s average fares are considerably less, they focus on intra-west flights far more than the big 3. The relative lack of gates at LAX overall is going to make LAX an attractive market; the fact that AA and DL are both committed to growing validates the money that exists in S. California. It is also worth noting that a higher percentage of Los Angeles area revenue flows through LAX than through any single airport in NYC or Chicago. .

  7. Is anyone working on an Embraer 120 replacement, or are those markets essentially hosed forever unless they can support 50-seaters (and what happens with those as the fleets get reduced).

  8. Hey Cranky – you forgot to mention Pittsburgh in the dropped routes list, as well as no mention to the New Orleans route ending this past week.

    As to my theory why United still operates LAX–PVG, I believe that’s due to the fact that J/V partner Air China does not have the authority to operate that route themselves (since China Eastern already does, and under current rule no two Chinese carriers can operate the same route to the US), but allowing United to operate it fills that hole.

    1. Kurt – Well Pittsburgh started after 2006 for United, so that’s a wash in this particular analysis. But indeed it’s a market that came and went. And New Orleans, oh yeah. I forgot that was going away soon. That’ll just be a further cut in the “cats and dogs” bucket.

      As for LAX-Shanghai, Air China isn’t a JV partner. (That won’t be allowed unless an open skies agreement is signed.) But it is a codeshare partner, so it does get some benefit. It seems more like squatting on rights to me – nobody wants to give up China rights because they are so precious right now.

        1. …which is all the crazier once you realize that UA offers the worst product on the route. As a Star Gold member I go out of my way to keep collecting miles with them but in this case I take the ANA detour via NRT.

  9. If I lived in a market where I could choose any airline to fly and they are all equally convenient, UA would be my last choice. That’s the reality they are facing in LAX. Their hard and soft product have been behind what AA and DL offers. Why would you fly UA unless you had to?

    1. I live in the LA area and am not tied to a particular carrier, it’s been awesome getting so many new/additional choices due to the AA/DL build-up.

  10. You seem to miss the point that they already have another, very significant hub very close to LAX (SFO), It makes no sense to have two hubs in close proximity to each other and have short flights from regional destinations to both hubs to then connect. It defeats the purpose of a hub which is to funnel everyone through the same connection point. SFO obviously makes better sense than LAX as the West Coast hub, for one because there’s less competition than SFO which are hubs for both DL and AA.

    1. AW – Well United has already retreated. Whether the merger has any impact on what remains of United at LAX is unclear. I don’t expect it would change all that much.

  11. No expert here by any means, but why expand international outbound without feeder traffic into LAX?

    Side note UA terminal at LAX ongoing renovation incl UA lounges, or is that not on UA’s tab?

    1. I’m no expert either, but I’d say the second largest metro area in the country ought to generate enough traffic by itself to support quite a few international flights.

    2. MB – Well, it’s more of an accidental expansion. United used to fly 747s from LA to Sydney and then people could go on to Melbourne from there.
      (United couldn’t carry local passengers between Sydney and Melbourne, so it was an expensive flight.) Now it can use 787s for both routes and go nonstop. I assume there’s some important client that needs to go to Melbourne so that’s why they bothered. Otherwise this flight probably would have gone from SFO.

      The other flight is Shanghai, which was really just a part of the “we need every China slot we can get” race. Other than those two flights, the rest has shrunk in terms of seats.

      1. Although SFO is counterintuitively closer to Australia than LAX, SoCal is a much, much larger market for Aussie traffic (and the demand is both ways: SoCal is the #1 market to Australia while SoCal is the #1 market for Aussies coming to the US). Therefore it makes sense for UA to continue connecting both SYD and MEL to LAX.

  12. I travel to Tokyo quite often–mostly for connections. I dislike HND if i am going downtown–NRT is more convenient. I don’t know why people keep thinking HND is more convenient; however, HND does have good connections on ANA and a great hotel within the airport. The flight from LAX to NRT is always nearly full and nearly always the cheapest (I am coming from Montana). I can go via DEN to NRT but only do this a quarter of the time because it’s usually much more expensive even though much more convenient. I have noticed that I fly more out of SFO than LAX when I go to Asia and for my money it’s friendlier and has better lounges so given equal fares I always pick SFO.

    UA is doing much better this year than AA and DL’s fares are so outrageous I won’t consider them in first.

  13. I’d like to have a dollar for every pundit who’s said X airline is going to, or should, de-hub Y city. I’d make Bill Gates look like a pauper.

  14. Brett hinted at corporate traffic for LAX-PVG. AA historically has had most of the entertainment contracts, but Disney was one where internet armchair CEOs say it’s an UA account. They are building a Disneyland in Shanghai (and unlike Tokyo’s Disney not a franchise) which most likely required flights on their own metal. 787 is a good fit given the competition and a need to serve a corporate account.

    Brett – great analysis and a very interesting read.

    St George Utah I would say is a bit different than the others. It is Skywest’s head quarters and with all the EMB-120 flying it was likely a flight more to skywest corp shuttle with any passenger demand being gravy. Skywest recently announced at risk flying to Phoenix on AA for likely the same purpose.

  15. Being so near the southwestern corner of the continental US, LAX isn’t well-situated to be a domestic hub, but it’s still surprising to me that UAL is cutting service to those transcon and international destinations. LA just seems too big & important for a big-three carrier not to be there with abundant nonstop offerings. But I was equally surprised when UAL dropped JFK service, and my thinking was the same: how can a major carrier not have some presence at all three NY airports?? Obviously, UAL management didn’t think the same way.

  16. Why is everyone so obsessed with UNITED in Los Angeles? Investors’ unreasonable obsession with PRASM clearly demonstrates that the ‘bottom line’ is ALL that matters in the U.S. airline industry. The days of flying certain routes or aircraft for ‘prestige’ are over in the states so I find it hard to believe that LA’s sub-hub doesn’t reap financial return. I’ve read a lot of commentary that suggests UNITED could reallocate LA capacity to SF, but what gleaming opportunities are they missing out on in SF? Adding a fourth 2nd tier Chinese city in one year? Doubtful. Latin America is a graveyard covered via. a laggard Houston hub which could easily cope with capacity reallocations, if need be. The trans-atlantic market is saturated and well covered via. Chicago & Newark. The one major opportunity I see for UNITED is a 787 SFO-DEL flight (possibly BLR to make UNITED an innovator again). The Newark-India flights sufficiently cover the large VFR market, whilst an SFO flight could better capture the only high yielding U.S.-India traffic. And let’s be real here, while the Chinese carriers have dumped loads of cheap fares into the market, their premium products are garbage. When pitt against China Eastern & the other U.S. carriers, the long established LAX-PVG appears formidable. The slot holding theory is totally valid as well…an interesting conversation all-around!!

  17. Cranky So, has UA decided to abandon the US and focus on international? How do maintain an image and prominence when you walk away from all the major markets?
    Marv Zwerin

  18. Cranky, do you have data showing numbers of passengers instead of percent change? I imagine that Delta was much smaller than United at LAX in 2006, so the large percentage increase may not mean that much. Even with the shrinking, I think United still has an operation of comparable size to the others on your chart.

    1. Jim – I put that in another comment. To recap:

      Specifically United flew 8,468 flights in July 2006 with 855,572 seats. In July 2016 it flew 4,596 flights with 631,413 seats. Delta flew 2,015 flights with 380,181 seats in July 2006 with 5,132 flights and 749,797 seats in July 2016. American flew 5,771 flights in July 2006 with 675,762 seats. In July 2016 it flew 6,893 flights with 934,560 seats.

      So United was 4 times the size of Delta in 2006 and now it’s 10% smaller.
      United was also nearly 50% larger than American in 2006 but now it’s about 50% smaller. It’s a huge change.

      1. Brett, I know that DL and AA have grown much faster, but I don’t think it’s fair to say that UA has been “dehubbed” just because it’s shrunk a bit and other airlines have grown more. LAX is still a major market for UA, both for O&D and connections. If DL eliminated 75% of its flights at ATL, it would still have a hub there.

        1. Jim – LAX is still a major market for United in terms of local traffic, but it has cut connecting opportunities significantly. Many of the connections that still exist go on to other hubs. (I think I was looking at Tucson or Palm Springs which had most connections going on to SFO, Denver and maybe Dulles.) LAX is being scheduled for local traffic and those markets that primarily served connections have been cut. LAX is most definitely an important focus city for United, but it really isn’t a hub in the traditional sense. And connecting opportunities only continue to shrink.
          I certainly wouldn’t call it a hub.

  19. Where do I read more about the goofy Orange County to LAX flight that United used to run? That would be a fascinating trip report…

    1. AW – Well, I actually flew it once. Back when I worked at United, I was flying out to LA. I had the choice of either flying in coach on the nonstop or I could get First into Orange County. I had no way of getting from Orange County to where I needed to be (right near LAX), but I could just take the Brasilia up. I had the time, so I went with that option.

      I actually have some notes from the flight. It had 24 out of 30 seats filled, strangely enough. Here’s what I had to say:

      This is just a strange flight. I really want to know who takes this. It was definitely an odd group. One woman got to the gate and asked if she could just drive to LAX instead (why would you go to the gate?). Another was freaking out because she had an 80 minute connection in LAX and thought she was going to miss it. When the gate agent finally asked, “What do you want me to do?” the woman asked if she could just book her on a later flight so she wouldn’t miss it! So weird.

      Anyway, we walked out on the tarmac, shifted pax for weight and balance and then we were off. We took off out of SNA, headed over the water, turned around and came over LGB. Then, while still seeing SNA out the window, we started final descent, looped around into the pattern and landed at LAX.
      We parked at the remote terminal, and it was ridiculous – we had to wait about 15 minutes for a shuttle to T7, and the lady who was worried about missing her connection started to make a little more sense! It ended up fine.

      1. Brett, when I first was hired at AA, we used to fly DFW/DCA or DCA/DFW with a stop at IAD. DCA had not opened up the restrictions in 1988 and as a new Flight Engineer, I would send the ACARS Change Over for arrival info at either airport while taxiing out at the preceding airport. Flight time could be as low as 10 minutes with a north departure out of DCA up the river, left turn for right base on to 1R at IAD.

        Ben Rich

      2. Pretty sure that AA Eagle used to run this route too in the 80s. I remember an article (maybe NYTimes) about the author flying the run just to push him over the top with miles for a trip to Europe.

  20. Lots of conjecture based on non correlated data points, and not much real analysis.

    The overall tone of this thread is overly negative, without any real focus on the positive aspects,

    1) All the prop flying from LAX was dropped with the EAS agreements and I believe “at risk” flying by Skywest. This flying was a large segment of the ASM drop.

    2) You ignore the fact that UA is spending $500M in terminal improvements to T7/8, which includes new lounges (United Club and Polaris).

    3) You didn’t mention that United is adding another LHR departure from LAX, or the vast increase in TCON flat bed seating to New York (EWR)

    4) MEL is hardly, “just like SYD with a tag.” Having a nonstop service, without having to share capacity with SYD is a large increase in ASM to MEL.

    5) You fail to really discuss the increase in transit and capacity increases at SFO to the North. SFO is the premier hub in the West Coast, DL and AA are flooding capacity into LAX to increase market share. UA is smart enough to not play that game (there are no winners) and focus primarily on higher yielding O&D.

    6) The Central American and Mexican market is yield poor, higher yield traffic in those markets have fallen off the cliff, and these markers are unlikely to contribute to much downline feed to United at LAX. Any demand that’s available can be handled by IAH, plus I haven’t seem AA or DL pick up any of these markets.

    United is focusing on building a profitable operation at LAX. Combined with JV with LH/NH (where GPUs can be used) in addition to NZ and LX, United has a large POS presence in profitable long haul and high yield business markets. If you live in PIT, TUL, OKC and need to get to NRT, SYD, SHA, you can transit ORD or SFO. LAX is redundant for that market segment.

    There’s no story here. Markets change, demand changes, competitors change, technology changes. That’s all that’s happened here.

    1. The “story” is to say that there is no need to “de-hub” LAX (as other analysis articles around the internet suggest) since UA has done it already. Regardless of what reasoning is pointed out, what Brett is trying to highlight here is that UA has reduced capacity at LAX and is effectively serving the local market.

    2. There is no negative tone to Bret’s story. He’s just reporting the facts. United is doing the smart thing in focusing on where it can make money rather than having capacity for ‘bragging rights’ in a location where others are going crazy. There have been several articles speculating about the status of the IAD and LAX hubs for United. Bret is simply stating that LAX has already been de-hubbed.

  21. DL has had a habit of building and tearing down its LAX operation. I can think of at least three times it’s grown and shrunk. If I remember correctly, the last trough of their operation was right around 2006/07. They may have shrunk even more into the recession in 08/09. So the growth of DL between 2006 and 2016 looks really big, but it starts from a place where they purposefully got a lot smaller at LAX. A fun look might be the ebb and flow of DL (and AA) at LAX over the years. DL, along with AA have a had big changes in flying there. AA in 2001 after the takeover of TW was much bigger than they were in 2006.

  22. Steve,
    there actually is real fact to support most of what has been discussed and not conjecture.

    First, average domestic fares between LAX and SFO are within a few dollars of each other – but UA is higher at SFO, likely a product of UA’s larger network to the eastern US.

    Second, AA, DL and UA all have domestic average fares within a few dollars of each other from LAX. The notion that they are dumping capacity to gain market share is not supported by facts.

    Third, UA’s strength at SFO is its transpacific network from which they gain a very healthy premium relative to other markets. However, AA and DL are adding capacity on the west coast – both from LAX and DL from Seattle that is competing for flow traffic that UA historically has carried. They aren’t going after UA’s core SFO market but they are going after everything else in the western US and their share is increasing.

    Fourth, five years ago, United was the largest carrier of LOCAL traffic from LAX to Asia both by share and by revenue; today that title belongs to Delta, even with the devaluation of the yen. DL carries LAX local traffic via Seattle in addition to its nonstop from LAX just as UA does via SFO.

    Fifth, UA’s share of the LAX local market -international and domestic has shrunk relative to AA and DL. given that the LAX local domestic market alone is 25% larger than SFO, no company can afford to say or justify that LAX is not important. UA simply supported a lot of routes from LAX with connecting traffic and those routes can’t be sustained now that UA is not trying to push as much connecting traffic over LAX.

    Sixth, UA does not have a point of sale premium from LAX relative to other AA or DL. All 3 carriers garner similar points of sale; the average fares differ by carrier and market but the POS is within a few percent.

    Sixth, all carriers are spending money on facilities at LAX. They aren’t going anywhere but terminal upgrades are a cost of doing business there. It is a gate-constrained facility which means that the ROI on those investments will be high and AA and DL will gain a long-term market position that likely won’t be challenged because any further growth that comes will be at high terminal costs – where room can be found.

    Len,
    AA and DL both were involved in mergers in the western US back in the mid to late 80s. Both pulled back LAX in the 90s and esp. over the past 10 years focused their efforts elsewhere – AA predominantly on Latin America and DFW while DL focused on the east coast, esp, NYC. Remember that UA once had a Latin America hub in MIA alongside AA while UA bought CO as the largest carrier in NYC; now DL carries more local passengers from NYC. It would appear that AA and DL both succeeded at the strategic initiatives that took their focus off of LAX but have refocused on the LAX market at the very time when UA seems more interested in its other hubs.

  23. You really have to be familiar with the CO mindset to understand what happened at LAX.

    CO uses a narrowminded approach to assessing things. In the case of flights, each flight itself is measured as a profit center, and if the assets can be deployed elsewhere for more return, then that’s what they do. Little consideration is put into the value of the whole, the network, or what it would do to other customers.

    An example of this in another industry would be a grocery store not selling milk. Hey, it takes a lot of valuable fridge space, and we sell other, higher margin items that make us more money, so milk has got to go. Problem with that is you lose the grocery business of all those customers who need milk as well.

    A more egregious example would be a gas station that stopped selling gas. Profits are razor thin on gas, but convenience store items are high margin, so might as well stop selling gas, and expand the convenience store, right? Wrong, if you’re not selling gas, nobody will come to the store.

    Same thing happened with UA at LAX. At some point, they lost a critical mass of flights to become the strongest player in the market. Combined with the service issues post-merger, and the growth by DL and AA, the alternative became roughly comparable in network, and arguably better in service. This drove away corporate contracts to the point where LAX suffered the largest losses in traffic in the United network, which combined with the fact that the merger had eroded UA’s revenue premium, and they now had a much stronger need to reduce flights. The revenue premium issues were even more amplified at LAX because it is such a competitive market. So now with reduced earnings, the profitability of LAX was harmed further than other hubs, necessitating the need to cut more with the CO approach. A vicious cycle.

    The way to come out of such a rut is to spend money and beef up the network and service to earn back business. But that takes a while and UA didn’t have the patience or the desire to burn cash with their financial performance (look how long it’s taken DL flying unnecessarily big RJs on short haul markets just to advertise there is always F on your flight out of LAX?). Meanwhile they relinquished gate assets to AA at one of the most competitive and gate constrained airports in the country.

    Normally this would be really shocking behavior, but if you understand the small-minded CO mentality, it really makes sense. After all, why does a gas station need to sell gas if they make more money selling Doritos and lottery tickets inside?

    1. 7-Eleven would argue operating convenience stores without a gas station is a perfectly valid (and profitable) business model.

  24. I think this is great analysis and agree that UA have moved LAX into more of a focus city than a hub but (and similar to the IAD story) i really feel people are creating more of story here than exists. I agree that United has reduced hub flows through LAX and favors SFO, just as it has shown more love for EWR than IAD. It makes sense when one is smart and rational about it: SFO is a superior place for UA to hub (cf. LAX) with the one exception of fog and the harm that does to your operational reliability. It’s geographically better situated for north-south flows on the west coast, the small market overflying (the Fresno, Bakersfield, Reno, Eugene, etc markets) is often shorter from SFO, plus it’s better situated for trans-pac flows, even Hawaii flows are better for many US destinations. When you add that UA has roughly 50% marketshare at SFO and the incredible wealth of the metro area and the high-margin business traffic, you get the picture of why UA would favor SFO over LAX. It makes almost no sense to run two mostly overlapping hubs purely to boost market share in this age of consolidation. So this all leaves UA offering 630,000 seats in July from LAX mostly aimed at O&D traffic. Given how much connecting traffic that DL in particular (but also AA) push through LA, its likely that UA is #2 in O&D. The analogy to NYC is an apt one, but I feel people are using the wrong airport: LAX is to UA what LGA is to DL and AA. Its a large operation that is aimed for local traffic with some limited connecting flows where they make sense. That’s not a bad place to be for UA and I’ll bet you every dollar in my wallet that DL and AA would trade west-coast hub operations with UA if they could.

    1. About 30% of the passengers flying United at LAX are connecting which is in line with AA and DL. The percent of connecting passengers can be calculated from local passengers from the DOT O&D survey and the number of passengers boarded by carrier.
      LAX is a predominantly O&D airport for AA DL and UA but each connects thousands of passengers per day through LAX

  25. There are some real fundamental misunderstandings here. American and Delta grown a lot at LAX. But to increase their Asia presence what were their choices? Also, when you look at the routes UA has added at SFO, UA is quickly becoming the Asia powerhouse airline. They are adding unique routes AA and DL will never be able to duplicate. Xian? Chengdu? Singapore? Routes the competition won’t be adding in our lifetimes. LAX will never be a fortress hub for any airline. If UA is going double daily on LAX-London, that doesn’t exactly follow the narrative of closing the hub does it?

    United has walked away from ethnic markets from LAX to Latin America for one simple reason. They merged with Continental, a much more powerful airline in the region. UA is more powerful and committed to Latin America than ever before actually. If these niche flights were so profitable from LAX, they would still be flying them. Routes have to make money. The article suggests they did this for no reason at all.

    When an airline completely dehubs an airport, they are not adding ANY new destinations. That is what dehubbing means. We get that UA dropped more cities than they have added at LAX. The article makes a big deal of this. But gee whiz, a real hub draw down? Nowhere close. When UA drops CLE as a hub, how many new destinations were they adding? Zero, correct? This article completely loses sight of this. The author is unaware of what a hub closure actually entails. For the author’s information, a hub closure means no new cities at all.

    To claim a nonstop to Melbourne Australia isn’t a new market shows real ignorance. A tag on after is not the same as a nonstop on a 16 hour flight. Would you prefer a nonstop or a tag on after being on a plane for 16 hours? What rational human would say this is the same?

    Okay so UA dropped Imperial, and Bakersfield, and Inyokern. This does not constitute a full hub draw down.

    The author doesn’t know that Dulles is a transcontinental market? How sad. Why isn’t there a list of markets UA does serve from LAX. The weaknesses in this article are serious. Yes, UA dropped JFK routes to LAX and SFO. But now they have hourly flights to EWR from both places.

    1. Ken,
      United became the Asian powerhouse very soon after they bought Pan Am’s Pacific network in 1985 or so. Pan Am had long had a Pacific route system that was not focused just on Japan as Northwest did so United was in a very great position to be able to make San Francisco into a transpacific gateway. While it is hard to believe that either AA or DL will ever have a transpacific hub as large as UA at SFO, they both have done a very good job of growing other hubs. DTW is the largest US carrier transpacific gateway east of the Mississippi while AA’s transpacific hub at DFW is just about the same size as DL in Detroit. In time it is possible that DL’s combined SEA and LAX transpac hubs could be as large as UA at SFO and LAX combined. DL has asked for a lot of space at both LAX and SEA and is willing to pay the price to get it.

      Still, UA’s size and experience in Asia will allow them to add secondary cities in China – but I would hardly believe that AA or DL will never be in a position to add nonstops to either secondary cities in China or major cities like Singapore.

      regardless of what UA does in LAX, they likely can continue to sustain their PVG flight because of the size of the local market and the size of their codeshare operation at PVG. While AA and DL certainly are growing at LAX, all 3 have and will continue to have enough of a presence in the largest markets to PVG that are connected to LAX.

      The bigger issue is that UA is not about to drop any China route until the US and China either signs an Open Skies or adds enough flights that UA is not at risk of losing any flights. And let’s not forget that UA added its own LAX-PVG flight immediately after AA announced it was starting that market. Now that DL is also in that market, I believe it is the only market to China that AA DL and UA all serve.

  26. No, LAX is not a UA focus city. CLE is absolutely a focus city for UA. They have hub flights with a number of other flights to places like DCA, LGA, BOS, CUN, and SJU. They also have seasonal flights from CLE to Fort Meyers and Fort Lauderdale. That is truly a focus city. UA has a broad portfolio of LAX flights, including China, Japan, London, and Australia. UA does connect people through LAX to these destinations, so it’ s pretty straightforward in being a hub. It’s still a hub even though it may not be the #1 in the market.

  27. Well it looks like someone posting this to airliners.net (http://www.airliners.net/forum/viewtopic.php?f=3&t=1340703) has brought out a few new commenters here on the site. I take pride in having good, civil discussions here, so hopefully our newest participants can keep it that way. Now, for some general thoughts, some of which may be targeted more at the Airliners.net commenters, but I don’t participate over there anymore.

    1) There is some strange belief that I am somehow attacking United for shrinking Los Angeles. That’s ridiculous. In fact, I say there may be more room to shrink at LAX. There’s no question SFO is the more powerful hub and United should be trying to funnel traffic through there. LAX for United should be focused on local traffic, as it is. I have no idea why people think I’m attacking United for the draw-down.

    2) There’s a lot of angst about me suggesting LAX is no longer a hub, but this is a game of semantics. If you consider something a hub because it has a fair number of flights, then fine, call LAX a hub. That’s not how I’d define a hub. A hub is something that’s built to handle connections. Of course a mix of local and connecting traffic makes for a great hub, but you have to design for connections. Think about the race United and American have had in Chicago to make connections more efficient from a scheduling perspective. Hubs are meant to connect, and that’s not what United has done in LA.

    Here’s a great example. United’s flight from Tokyo gets to LAX at 1115a. Technically it allows 1h10m connections as legal in LA, so what has United scheduled after 1225p for up to 3 hours past arrival? You have the usual flights to United hubs which operate all day long anyway. Other than that, today there’s a 1245p to Vegas and another to San Luis Obispo. There’s a 1255p to Monterey. There’s also a 120p to Oklahoma City and a 125p to both Santa Barbara and Leon. Those are the only connections scheduled within 3 hours of the flight arriving. 6 non-hub flights.

    Now let’s look at SFO. The Narita flight was scheduled to get in at 1040a today. Minimum connect is 1h20m, so how many non-hub flights were scheduled between 12p and 140p? 23: Tucson, Burbank, Santa Maria, Arcata/Eureka, Bakersfield, Kansas City, Salt Lake, Orange County, Reno, San Diego, Santa Barbara, San Luis Obispo, Mexico City, Orlando, Boston, Portland, Redmond, Seattle, another Orange County, Northwest Arkansas, Medford, Dallas, Phoenix.

    SFO is being scheduled like a hub. LAX is not. LAX is a big focus city for United, and the airline is investing in it to keep it that way.

    3) Dismantling a hub does not mean connections disappear. There are still people connecting over St Louis, Pittsburgh, Memphis, Cincinnati, etc. The options are more limited, but any time flights can connect up, people will take advantage. The same thing will happen in LA. There are a couple pure connecting flights still around, like the Santa Barbara and San Diego flights, but it wouldn’t surprise me to see those disappear over time as United refines it strategy. Frequency has already been reduced on those significantly.


    Some people are way too emotionally invested in this. The point is very clear. United has dramatically reduced flights and capacity at LAX, and it has reoriented the operation to serve the local market. To me, that is a dismantling of a hub. How is this different than Pittsburgh or St Louis? The huge LA local market can support a whole lot more flights via local traffic than Pittsburgh or St Louis can. But the idea is the same. Serve the markets that the local market wants and supports without routing them through your other hubs. That is the definition of a focus city.

    1. Cranky,
      Your example in this post did a whole lot to explain your point. I doubt if many people, me included, bothered to look at the connectivity of major international flights as you did.

      the overarching principles are still the same..
      1. UA chose to reduce many flights using smaller aircraft esp. small RJs and turboprops because those flights added very little marginal revenue but a lot of increased costs. The competition for pilots is making it harder to support flights that operate for the primary purpose of feeding hubs instead of being able to generate profits even in the local market. United’s hub at LAX shrank as a result of UA choosing to focus western US connecting flights over SFO. UA still connects a similar percentage of traffic over LAX as AA and DL but they do it from a smaller base of total flights and, as you note, choose not to operate some flights in/out of a banked hub operation.
      2. SFO is a less competitive and larger hub to many global regions which provides opportunities for UA to increase efficiency and revenue at little incremental cost. IN and of itself, reducing one hub in favor of a nearby hub makes sense.
      3. The problem UA has faced is that in many former hubs, another carrier very quickly moves in to replace that capacity. DL has added capacity at LAX as UA has cut -but DL’s additions have been in markets other than what UA cut. at JFK, AA cut a number of markets esp. to the Caribbean but DL jumped in and added them. At PIT, STL, SJC RDU and BNA, WN added a flights as the former hub carrier reduced its presence.
      4. It might or might not prove to be true that UA can make LAX work as a smaller carrier compared to AA and DL just as the jury is still out regarding AA at JFK. Hub dynamics in most markets including DFW ORD and DEN have repeatedly shown that one or two large carriers and one or more smaller niche carriers cannot coexist. Someone blinks.
      5. Your last paragraph is spot on. The airline industry results in a lot of emotion about business topics and a lot of people come to conclusions devoid of the facts even when presented with them.

      Please continue to tackle the difficult issues on here including business-related issues that generate emotion but also have strong foundations in fact.

      1. Tim – So what’s really interesting to me is that someone wrote a post on a different site (http://upgrd.com/matthew/united-de-hub-los-angeles.html) disagreeing with me, and that made me dig in further on connectivity. It gets even more interesting when you look at Melbourne. Here’s what I wrote over there:

        Melbourne is the one international destination from LAX that United doesn’t serve from any other airport. Since it’s timed almost exactly at the same time as the Sydney flight, that should probably be the most important connecting bank at LAX for United. But the bank is actually quite weak. Let’s look, still using my 3 hour rule that I applied to the Narita flights.

        First, yes, there are flights from the hubs that connect as you’d expect would happen. There are flights from the hubs all day long and that’s going to allow connections to any other flight at LAX just because of the frequency involved. Looking beyond the hubs,[there are 12 flights that arrive at LAX that connect well on to Australia]. On the return from Australia, however, only 7 of those 12 cities have legal connections within 3 hours of arrival. Further, there are another 6 that connect only in that direction (Austin, Bozeman, Colorado Springs, Minneapolis, Monterey, and Prescott if you want to count that one). There are a couple others that would connect, but since the Australia flights arrive at the Bradley Terminal due to the early hour (not exactly a compact and easy connecting experience), a long connecting time is required.

        Why is this scheduled so oddly? It’s because from my vantage point, I see flights that are scheduled for the local market, in most cases (excluding San Diego, for example). These are flights that operate at the times that locals want them. Sometimes they connect internationally, sometimes they don’t. I have no doubt that the Melbourne flight relies heavily on connections, but I’d bet most of it comes from other United hubs and maybe Vegas. I wish I had that data in front of me.

        1. I think you have more than provided enough examples to make your point. The number of flights does not define a hub; connectivity does. I’m honestly not sure to/from what flights/markets UA does connect at LAX and I could figure it out but I think your point about the largest international markets is more than telling.

          If you want to further validate your point and have the time, perhaps you could compare the number of connecting flights AA and DL have to/from the same international markets (SYD being a proxy to BNE since AA and DL don’t serve it but the schedules are similar as you note).

          It’s up to you but I am certain both have much more connectivity to/from their largest int’l flights

          The 99 cent (10 billion dollar) question is whether the LAX market is large enough that UA can sustain a point to point operation against AA and DL which do operate their int’l flights as part of banks. Banking allows them to carry more passengers on the same flight – or at the least have more potential passengers to choose from so that they can choose the highest revenue passengers to carry if there is more demand than capacity on even some of the flights.

          Delta decided not to try to operate a focus city at ORD alongside AA and UA’s hubs or to have a junior-sized hub at DFW alongside AA. F9 blinked in DEN against UA and WN. Likewise, DL has got to be betting that its larger hub will make a difference at JFK as it competes against AA and B6 while AA and DL both hope that their own larger size at LAX will help them offer a larger schedule that will bring more passengers at higher value. That is how hub theory has long worked. It is only now as we have large hubs on the coasts that we have some carriers trying to prove that a few large coastal cities can serve as exceptions where differing sized hubs or niche point to point operations in other carriers’ hubs haven’t worked in interior US cities before.

          again, good work and a good conversation.

          1. Tim – I’m happy to pull that for flights to Sydney. I used the same rule looking for legal connections of 3 hours or less that aren’t coming from another hub. Delta has 18. (BNA, DFW, PHX, MEX, SFO, RDU, MCO, DEN, SMF, LAS, MSY, SAN, TPA, GDL, PDX, BOS, YVR, MIA). American has 22 (MRY, FAT, SLC, SFO, SEA, ATL, MSY, SMF, PDX, SJC, RNO, LAS, YYZ, IAH, MCO, IAD, DEN, RDU, AUS, ABQ, BOS, SAN).

    2. This topic a good discussion. I’ll just add that the topic is really so complex, yet nuanced, that it can be difficult grasp.
      From my perspective (former UA employee 1990-1995, contract work in 2004) there’s a litany of past corporate, fleet planning & network decisions, along with competitive responses and post-9/11 concerns, that would truly lend color to the discussion.
      Having said that, I’d suggest that post BK-1 UA management realized that it was competing against itself by hubbing at both LAX & SFO. Despite a large international schedule presence at both LAX & SFO, the vast majority of the capacity remained (and remains) domestic flying.  The geographical constraints of being on the coast, as well as having a dearth of major traffic generating spokes in the western-third of the U.S. was also a handicap. At a majority of the western spokes, UA/UAX was already the market share leader, so there was little opportunity for traffic stimulation at those cities served by both LAX & SFO.

  28. Looked up UA mainline stats comparing 8/2006 with 8/2016. UA operated lights are down from 3,486 to 2,984. Seats are down from 564,117 to 527,950, a decline of 6.4%. ASM’s are up from 1,008,903,166 to 1,080,303,305. Average passenger stage length has grown from 1,788 miles to 2,046 miles. Source is Diio based off of DOT stats. Having worked in network strategy for a rival, I think the best way to look at the hub vs focus city argument for UA at LAX is what share of passengers departing a city are connecting vs local origin. A hub has a higher percentage of connecting. I used the first quarter of 2006 vs 2016 in comparing T-100 data (all passengers on a flight) versus DB1B DOT data which is local passengers originating as this is the most recent available. In 2006, United passengers through LAX were 43% local versus 57% connecting, where in 2016 the percentage has only shifted to 46% local versus 54% connecting. This largely reflects the loss of local UAX prop service and some flights going to codeshare partners offset by some new international flying in my estimation. Overall it looks like UA at LAX is still very much a hub based on the share of connecting traffic remaining high, LAX is just a 17% smaller hub in 2016 versus 2006. Comparing connecting opportunities is a weak measure of a hub as many markets you mention as being dropped were pretty small. United’s connecting percentage of 54% in Q1 of 2016 is still higher than Delta at 51% and AA at 47%. United at LAX seems to have plenty of hub connectivity left and has not restricted to a locally targeted schedule.

  29. Cranky, I have appreciated this discussion and thank you for your input on my blog. A user named Rhys just posted some interesting info that you should check out — comment #6.

    ETA: Oh, looks like he has posted the same to your blog.

    What do you think?

  30. Have posted to both of your blogs.

    I filtered out connections in other hubs originating at LAX (such as LAX-IAH-MIA). Was a little quick with the unoriginal stats as Inwas in a rush at lunch. United’s LAX connecting percentage (DEN-LAX-SYD) is 30% in Q1 of 2016 compared to Delta at 33% and AA at only 24%. United in 2006 Q1 connected 41% of paxs at LAX. United at LAX seems to have plenty of hub connectivity left and has not restricted to a locally targeted schedule. Part of this may be due to AA and DL having better shedule for locals. UA may just have kept short and medium haul flights that best feed longer haul.

  31. Southwest, just to be thorough, is 73% local O&D pax at LAX during the first quarter of 2016, 13% connecting or 1-stopping through LAX, and 14% originating in LAX connecting elsewhere.

    1. Thanks for pulling the data Rhys. Just one thing to confirm. Are you including CO data from 2006 in this? (I assume so.)

      Not sure if you’re up for more, but… can you pull what percent of traffic is connecting through LAX from a hub? So in other words, pull DEN/IAH/SFO/EWR/IAD/ORD – LAX – XXX vs connections originating (or terminating) elsewhere. Of course someone going to MEL has to go over LAX, but people connecting from the hubs to Hawai’i/Asia/West Coast would undoubtedly have better options either nonstop or through another hub that has better options (like SFO). So I’d imagine most of that connecting traffic coming from the hubs is spill and could likely be better served by distributing capacity from LAX to other hubs. It’s also probably only going through LA because of price.

      And that’s what alarms me about high connecting percentages at LAX. It’s not really being scheduled like a hub anymore, from what I can tell. Since it’s being scheduled more like a focus city, that means people are more often than not picking these connecting routings for price since I’d bet they’re rarely going to be the most convenient options. That tells me that United can’t get enough local traffic and has to settle for cheap connecting options. Certainly something to be concerned about.

  32. I can dig further. In my rapid lunch time exercise, I may have left out CO in 2006, but will double check. That would show a bigger decline in hub size if I had left CO out, maybe a 25% decline instead of 16% in UA seats, but would lessen the decline in LAX connecting percentage over the last 10 years. I would be able to filter our other hub originating UA traffic between 2006 and 2016. It would be interesting to compare fares through LAX connecting versus DL and AA performance between 2006 and 2016. I’m not sure what connecting strength DL and AA would have had back then, particularly DL, as AA has at least had the Eagle system there for sometime, but will look and see what fare strength was in 2006 versus 2016 connecting through LAX for those airlines and get back to you.

    1. Just to give a little more color on how United schedules LAX, I’ve put together a chart showing arrivals and departures at LAX on a summer Monday (tomorrow) excluding flights to other United hubs (DEN/EWR/IAD/IAH/ORD/SFO).

      http://crankyflier.com/wp-content/uploads/2016_08_16-UALAXops.png

      What I see is more in the mornings and evenings with the rest scattered during the day. Now, I can imagine someone looking at this and seeing a semblance of a bank with arrivals in the 7am hour and departures in the 8am hour. But that’s not really the case. Here’s how it breaks down:

      Arrival from: Salt Lake, San Antonio, Dallas/Ft Worth, San Diego, Austin, San Luis Obispo, Colorado Springs, Monterey

      Departures to: Honolulu, Salt Lake, Albuquerque, Bozeman, Reno, Austin, Boston, Vegas, Orlando, Minneapolis, Colorado Springs, Maui, San Diego,

      I guess first it’s good to point out that even though this is the heaviest time for operations, the total numbers are still pretty small.

      For the flights in, most come from the east, with the only exceptions being San Luis Obispo and Monterey. So except for those two, the connecting options would be Honolulu, Reno, Vegas, Maui, and San Diego. The rest would be major backtracking connections that are highly unlikely.

      The Sydney and Melbourne flights arrive earlier, and those certainly would feed the rest of the connections going east in the 8am hour (most of them, anyway). But I’ve already talked about that before. It’s just not being scheduled like a hub.

      One other thing to note. About half of United’s operations at LAX go to other hubs. So on this day, it’s about 73-76 each to hubs and to non-hubs.

  33. Spent a little more time working on UA LAX traffic stats and ironed out the data. There had been an error that excluded CO from UA’s 2006 stats, I added HP to AA’s stats for 2006, they had yet to merge in the DOT stats, and WN would get credit for Air Tran’s 3 flights.

    UA’s share of LAX traffic connecting through LAX dropped from 36% in 1st Qtr 2006 to 30% in 1 qtr 2016, higher than AA’s 24% in 2016 and just three points behind DL’s 33%. Based on connectivity, UA’s LAX operation resembles DL and AA which are both considered hubs. For comparison, only 13% of Southwest traffic connects or 1-stops through LAX.

    On the question of whether UA at LAX has tanked fares to maintain low yield connections, that is more difficult to ascertain, but at first glance it doesn’t look like they are. Two caveats for pulling connecting fare data based on marketing code. Stage length can change over time, but for the average UA passenger, despite dropping many short-haul UAX routes, UA/UAX’s average stage-length actually dropped from 1,484 in 2006 miles to only 1,466 miles in 2016. The second item that clouds a comparison, is that it is subjective on who should get credit for a fare on a codeshare flight, or a mixed airline connecting flight. A UA codeshare ticket flying on Star Alliance partner Lufthansa might only have a short segment, LAX-SAN, on an expensive international itinerary. This issue should affect both 2006 and 2016 for many carriers so just take that with a grain of salt.

    Given all of that, United’s average LAX connecting fare was up 49.2% from 2006 to 2016. To put it in perspective, AA was up 37.1% over the same time period, Southwest with their lower connectivity was up 30.8%, and oddly, Delta was only up 5.4%. Northwest only had one NRT flight back in 2006, so that couldn’t explain too many of their 1,057 daily average connections that DL/NW had in LAX back in 2006. Maybe a greater share of DL LAX connects were on higher fare international connections on Skyteam partners in 2006. There is AF to PPT, Korean, China Airlines, China Eastern, etc. that could have connected traffic onward to DL hubs. Anyways, at first look it doesn’t look like UA has had to tank fares to maintain connecting traffic.

    Overall, based on levels of connectivity, UA at LAX still walks and talks like a hub, just one that is 29% smaller (based on passenger traffic) in 2016 compared to 2006. (All stats drawn from Diio which is sourced from DOT T-100 and DB1B traffic data that is in the public domain.)

    1. Rhys I think you’ve done fantastic work, especially since you’ve been assessing the network of a competitor (and, apparently, on your own time!) This is going to sound like I’m attempting to diminish your work (I’m not) – but it appears to me that we’re we getting a bit ‘weedy’.

      One of responses CF posted earlier was that (to paraphrase) defining a ‘hub’ is somewhat subject to semantics. From my reckoning there are, at least, three definitions of what constitutes a hub from an airline network perspective:
      (1) A hub that funnels traffic on an on-line basis (UA to UA flights/passengers)
      (2) An alliance hub, where the aligned partners create a semblance of a hub network by connecting passengers at key gateway airports.
      (3) The gateway where a domestic carrier can leverage a large number of off-line (interline) connections as feed for the domestic carrier network from that gateway.

      A limitation of D1B1 is that it will indeed identify all connections – whether true online, codehared, thus appearing as an online connection, or other airline interlined as connections.

      To digress a bit as I think about this, how is this discussion of the UA/LAX hub (or former UA/LAX hub) being framed? Is this a discussion of online (own metal to own metal) connections? Of online + alliance connections? Of online + alliance + unaffiliated carrier (interline) connections?

      The percentage of connecting traffic moved over LAX will differ based upon how the discussion is framed.

    2. Rhys – Thanks for pulling all of this. It’s fascinating to me because if you look at the schedule, it doesn’t look like a hub. So who are these people who are paying more for inferior connections? Some of it could very well be connecting to partner carriers, I suppose. Could it also be double connections through a different hub and LAX? The numbers just don’t add up to me.

      I may need to go pull down this data from DOT and play around with it. The numbers make it sound like a hub, but it’s not being scheduled that way so there’s a strange disconnect here.

  34. I disagree that a nonhub is a city oriented around local passengers only, and that LAX is not a hub city by that definition. This is what CLE has become for UA. On a side note last summer I did actually connect on CLE on UA. I was traveling from CUN-CLE-DEN. But CLE is not a hub. It should be called a focus city. I am living proof UA still connects people through CLE. Does that make CLE a hub? No. There are probably very few customers connecting on UA in CLE. But for some odd reason UA does not even call CLE a focus city. I don’t believe LAX is the same, as many of you are asserting.

    Do the readers here actually believe UA is just squatting on LAX-PVG to confront the competition? This is the assertion that is presented here, and call a spade a spade, it’s laughable. Long prestige routes are a thing of the past friends. It is also suggested here that maybe UA has a client that wants a nonstop to Melbourne Australia from LAX specifically, and UA has that one flight for this unnamed client. Honestly, these are pretty flimsy reasons to be flying such long flights. Maybe the flights are profitable and generate a lot of local tickets sales in addition to connections. So,why don’t you come out and just say that rather than making some weak and unsubstantiated assertions. Since the UA LAX operation is so similar to AA at JFK, why don’t you present a breakdown?

Leave a Reply to chris Cancel reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Cranky Flier