The Southwest Effect on Disappearing Change Fees


Domestic change fees disappeared so quickly and with such little resistance — except you, JetBlue — that I haven’t been able to stop thinking about this. The industry built change fees up to absurd heights over the last three decades, yet they fell in less than a week, at least with Alaska, American, Delta, Hawaiian, and United. There are undoubtedly several factors that went into the decision, and today I’m going to explore one of them. Let’s talk about Southwest Airlines — an airline that has never had change fees — and its impact.

When I was working for America West a couple decades ago, we were trying to figure out the right away to compete with the ever-growing Southwest threat. At one point, we decided to get rid of our change fees in key short-haul business markets like Phoenix – LA. I think I still have the pin with $0 on it somewhere in the garage.

This plan did not last for long. America West didn’t see the needle move enough to bother keeping the change, and so everything went back to normal. There were several issues with this test — including the very limited scope — but time and time again, airlines have resisted actually dropping change fees to match Southwest… until now.

What exactly has changed? I’d argue there are two things that have altered the landscape. One thing that we all know and hate — COVID-19 — changed the industry overnight. The other — Southwest’s size and influence — has changed slowly but surely year by year. And that’s where I’ll start.

Southwest Has Dramatically Increased Overlap with the Big Three

Once again, I turned to my old friend Cirium to help me decipher the data. We all know that Southwest has grown over the years, but there is an important nuance in this growth. Let’s start with a look at seat share within the lower 48 states over time.

Data via Cirium

For this look, I included all predecessor airlines. So American includes America West and US Airways, United includes Continental, Delta includes Northwest, and Southwest includes AirTran.

Back in 2004, Southwest had about 15 percent of the seats flying in the domestic US market. Soon after, Delta went bankrupt and cut capacity significantly. That meant Southwest’s share continued to grow. You can see more shifting around the Great Recession, but you can also see both American and United declining after their mergers. By the end of 2019, Southwest had increased its seat share to above 20 percent. , right in the same range as American and Delta.

Now, this might not look like a huge gain, but the key is understanding where the gains were made. Southwest moved into major metro areas quickly. It took over AirTran and got its place in Atlanta. It moved into LaGuardia, Newark, and Washington/National when the feds forced slot divestitures from the others. So, it’s really the overlap with each legacy airline that matters.

To look at this, I had to make some decisions. To get true overlap, you have to decide which airports can act as substitutes for the others. If I didn’t combine airports in, say, Chicago, Dallas, and Houston, then overlap would look artificially low. In the end, I settled on these as co-terminals:

  • Chicago – Midway, O’Hare
  • Dallas – DFW, Love Field
  • Houston – Hobby, Intercontinental
  • Los Angeles – LAX, Burbank, Long Beach, Ontario, Orange County
  • Miami – MIA, Fort Lauderdale, West Palm Beach
  • New York – JFK, LaGuardia, Newark
  • San Francisco – SFO, Oakland, San Jose
  • Washington, DC – Baltimore, Dulles, National

You can feel free to quibble with these all you like, but I decided these were true secondary airports to serve large markets. I didn’t include places like Manchester, Providence, or Islip, because those felt like poorer substitutes that never really caught on. Just look at Southwest’s network now and you’ll see those have shrunk dramatically because of their inadequacies.

I then put together head-to-head match-ups. United was the first to roll out the change fee cut, but American was working on it before United announced. That means these two both came to simliar conclusions, so that’s where I put my energy. Let’s start with American.

Data via Cirium

This may be a little confusing, so let me explain. The little gray columns are the percent of domestic routes American flies nonstop that Southwest also flies nonstop. Since American has more of a hub-and-spoke system, you would expect this to have less overlap on the route basis. Still, it nearly doubled between 2004 and 2019.

The red columns are the percent of domestic flights that American operates in markets where Southwest also flies nonstop. What this tells us is that Southwest overlaps with American in bigger markets where more flights exist. That’s not a surprise. Notice how much the gap widens between 2004 and 2019. Nearly half of American’s flights operate in markets where Southwest has nonstops.

Lastly, we have government data showing what percent of passengers that flew American within the lower 48 could have flown Southwest. This doesn’t take into account times or seasonality, but it does include connecting traffic. That’s why it’s so important from a base competitive standpoint.

In 2004, about half of American’s passengers could have opted to fly Southwest. By 2019, however, that number had soared to 80 percent. A mere 20 percent of American’s domestic passengers didn’t have a Southwest option, and those are going to primarily be in small to small-medium markets like, say, Harrisburg or Fresno, or something like that. There aren’t many mid-size or larger markets where Southwest doesn’t fly.

When I turned to United, the overlap was even greater.

Data via Cirium

Southwest has become a major network competitor that can compete for the vast majority of the legacy airline customers. From a business perspective, that had to push pressure on the legacy airlines to be more competitive with fees.

The Right Opportunity Presented Itself

The problem, of course, is finding a way to scrounge up the $2 billion in domestic change/cancel fees that the industry collected in 2019 from some other source. I don’t care how big your couch is… you aren’t going to find that much loose change. Then COVID-19 hit, and the pandemic created the perfect cover.

Overnight, airlines had no choice but to roll out change fee waivers to let people shift their plans. That meant change fee revenue dropped to zero as did, well, ok, pretty much all revenue dropped to zero, or close to it.

With no change fees in the market, airlines could do little tests to see if those who were traveling behaved any differently with change fees out of the market. There were people who had essential travel needs that kept flying, and of course, we saw people start flocking toward leisure destinations in June. So United and American could look to see if they were able to shift share. Unfortunately, that data isn’t out yet publicly, so we’ll have to wait. But I assume the airlines could start to make some justification for removing change fees based on what they were seeing.

Ultimately, however, the lack of change fee revenue made it an easier sell. Wall Street couldn’t destroy them for getting rid of a fee that wasn’t bringing in any money. Simply the expectation that changeability would hold an important role in travel decisions for the near to medium-term future meant that it was a justifiable move. And now, the deed has been done.

United said that it would have no fee “forever,” calling it a “permanent” move. I have trouble believing that. When attitudes start to change, and people begin to forget the horrors of the pandemic, behavior will shift too. Some airline will try to bring back change fees, and then we’ll see if it sticks this time around. If it does, I imagine it will be at a much lower level that won’t cause such high levels of anger. But we’re far from that day actually arriving.

Get Cranky in Your Inbox!

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

56 comments on “The Southwest Effect on Disappearing Change Fees

  1. Quite sure I’ve commented in the past that I personally have never picked an airline based on the change fee policy. I don’t book a flight with the intention of changing it. Yes, I’ve changed my flights (a lot) usually when on business trips and I want to get home early or something changes. For leisure travel….I can’t think of a time I’ve changed my flights. Assume most leisure travelers are in the same camp. Yes, it’s good marketing to say change fees are gone but for the once a year to MCO family I’m betting a free checked bag is a bigger draw.

    Secondly, didn’t everyone waive change fees when you reached a certain status? I know Delta has done this for me and even before I was anointed the level of precious metal X, Y or Z it wasn’t uncommon for an agent to waive the fee anyway. If the flight I was originally booked on looked anywhere near close to full it’s a no brainer to free up that seat and save an overbooked situation. At least most of the agents I dealt with saw this and wouldn’t ruthlessly try to wring me (my employer) out of any extra $$$. I can’t speak for other airlines…been ages since I’ve changed a flight on AA or UA, but I never found it an issue. The vastly superior perks of flying Y on DL vs. WN pre-covid was more than enough to make up for a flight change I never intended to make in the first place. And yes, I flew about a dozen flights with WN in 2019 so I know both airlines well enough. Just my 2 cents.

    1. Free same day changes have generally been a perk for mid-tier elites (United and Delta Gold etc.) – those are the ones with a fee around $75. This sounds like what you’re talking about, rather than the $125-$200 change fees Cranky is addressing in this article (for changing dates or destinations at any time, or changing flight times prior to checkin or day of departure)

      For the latter, as far as I know Alaska is the only carrier that had waived change fees on paid tickets as a published benefit (though it might also be the case for invitation-only status or certain corporate clients on the big 3)

      1. also of note is that Southwest has notoriously been far worse than the big three for same day changes since Wanna Get Away fares are not eligible – if you’re on one, you have to buy up to an Anytime or Business Select fare, which is usually far more than the $75 that United, Delta, or American would have charged you.

        1. Funny you should mention WGA fares and their ineligibility. Four times I’ve had same day “situations” where I tried to stand by on an earlier flight, all on WGA. Twice WN agents told me I had to pay the difference and I decided to stick to the original itinerary. Once at LAX and once at DEN. The other two times they just put me on the earlier flights no questions asked. Once at SJU and once at PWM. I’m not sure if those two smaller stations are staffed by WN employees or contracted out. LAX/DEN: By The Book. SJU/PWM: Sure! No Problem.

      2. Thanks for clarifying. Yes – same day changes is what I was mostly referring to. Guess I should ask is that waived too? That’s what has way more value to most people.

        I’ve done the $200 changes many times as well but it has always been waived for weather or some other mitigating circumstance. Again, my point remains, I don’t book a flight with the intention of changing it.

        1. American and United are going to allow free same day standby. I believe confirmed same day changes are going to remain free only for elites, though iirc United is going to make it free for Silvers (previously it was Gold or higher)

    2. Depends on the nature of your business. At my previous company, we traveled quite a bit (domestically) and project issues sometimes dictated that flights be changed or cancelled. For this reason, Southwest was the company’s preferred airline. Admittedly, it was unusual for us to get the cheapest “Wanna Get Away” fare but – even if you did – the competing airline’s fare would have been the same and a change would still result in paying the difference in fare PLUS the change fee. Company hated that. Plus they made a deal with WN for us to use SWABIZ which I’m sure resulted in kickbacks.

      Nowadays, I mostly fly UA for work (well, pre-Wuhan). And I’ve liked them. But I do still like WN very much, they’re the easiest airline to deal with when you have an issue, they don’t charge for checking baggage (unless it’s excessive), and they tend to be pretty punctual. Also, they have decent pitch and you’ll never be stuck on an RJ.

  2. Your graphs above highlight precisely why UA was the first to move and WN’s growth at DEN made it all the more necessary for UA to act.
    There are business travelers that do select their airline based on total costs which might include change fees – although those are often changed if there is any kind of negotiated agreeement.
    Change fees were disliked even among leisure travelers which is part of why Delta was talking about eliminating them even before covid. Southwest’s one size fits “most” pricing model is much more palatable even if it leaves some revenue on the table.
    Finally, the real shift – not just for the big 3 but also for WN – is competing with ultra low cost carriers. WN does not do as well against ULCCs as they might want you to believe – and that will play out in LAS just as it has in FLL.
    Of course, the larger issue is the overall financial health of AA and UA and those two airlines’ ability to compete against WN regardless of change fees.

    1. WN hasn’t been competitive with the big 3 for quite a while too. You can massage the #’s by saying they don’t have the fees but as I already commented, those could be waived. For years I’ve been checking DL flights vs. WN and on routes where they directly competed there often was no cost advantage to Southwest and many times it was a premium to fly them. The only real savings on WN was when I had to inconvenience myself with a layover at MDW or STL, etc. I’m just one flier going to where I need to go but I think the luster of Southwest being this inexpensive airline to fly wore off somewhere in the 00’s.

      1. Which market are you in?

        Here, Southwest operated, and operates, a schedule that provides more nonstops and sufficient frequency to a large number of markets, including overlaps with legacies.

        How much I’ve used them varies by year, depending on which destinations I hit…while I have their credit card I’m not a WN loyalist…but there are a bunch of dots on the map where they’re comparably, or more, convenient than the alternatives. Keeping in mind that AA/UA are unimpressive from a domestic product perspective.

        I say this as someone who’ll pick Delta, all else equal. But in my market at least all else tends not to be equal, so Delta gets reserved for flying into their hubs, where thanks to Southwest (and sometimes others) their pricing is comparable.

        1. MSP….so my non-stops are limited but I still price WN when I’m out and about. Did a project in Houston where HOU was just an easier airport to use. DL through ATL or WN through STL to get home. Prices were about par at that time. Still cheaper to take a non-stop into IAH and just drive through Houston traffic. Routes like MSP-DEN where there is a ton of non-stop competition WN is outdone by Frontier so if you’re truly shopping lowest ticket price in this market Southwest isn’t the winner. They do have very loyal fliers and I’ll grant them that.

      2. “Those fees can be waived”…..well yes they can be in theory, but in reality are not often waived. That includes baggage fees and ticket change fees. Even is Southwest fares are equal or slightly higher than competitors, the baggage fees alone will tilt total cost in favor of Southwest.

  3. The red columns are the percent of domestic flights that American operates in markets where Southwest also flies nonstop. What this tells us is that Southwest overlaps with American in bigger markets where more flights exist. That’s not a surprise. Notice how much the gap widens between 2004 and 2019. Nearly half of United’s flights operate in markets where Southwest has nonstops.

    Should that last sentence be “American’s” and not “United’s”?

  4. From today’s AA marketing email:

    For a limited time, book refundable fares to select destinations in Europe, Asia and South America and travel anytime. If your travel plans change and you wish to cancel, your fare is refundable minus *a $100 administrative fee*.

    That didn’t take long, did it?

    1. I’ve always wondered about the demographics of who changes their flights. The most common example in my industry is traveling technicians — people who fix large machinery that either buy one-way tickets, or buy roundtrips in hopes of actually satisfying the customer in a known timeframe. That’s easier said then done and I’ve seen cases where a guy will change his flight multiple times a week. I have to believe that leisure travelers are less likely to change flights — “I know I am going to Florida this week and nothing is going to stop me.”

      Point is how much does this change consumer demand? Are people buying more tickets now that they know they can change the flight? Conversely, were change fees a driver for corporate travel bucks?

      1. At this point leisure traffic is a *lot* more change fee sensitive than before. Between pandemic uncertainty and remote work/schooling, folks need, and have, more schedule flexibility, where before they could/would/had to build vacations into a rather fixed schedule.

        To provide a bit of anecdata, my family speculatively booked two trips a few months ago: one where we would fly into MCO, and one where we would drive three hours. Some of the Rapid Rewards points my family is now sitting on are from that trip…we opted for the trip that didn’t involve flying to Florida in the middle of a case spike.

        The fact that leisure travelers need flexibility but flat-out won’t buy a ticket without it, versus businesses where there was some willingness to shell out for a higher fare class or flat-out pay the change fee in return for added flexibility, certainly plays in here.

        1. I had similar thoughts: Eliminating change fees is designed to attract customers to buy tickets while uncertain about whether they’ll be able to go. Sure they’re waiving them short term early, but removing it permanently entices people to book further out as well, getting more cash in the bank to help cover their ongoing costs.

    2. AA’s new policy doesn’t apply to flights to Europe, Asia, and South America, so this is a totally different, limited-time policy. So no, they’re not walking back the changes that CF was talking about with this promo you’re mentioning.

      Here’s the policy change where it specifically says “No more change fees for all domestic and short-haul international flying on Premium Cabin and Main Cabin fares.”:

      1. I suspect we’ll be waving that in AA’s face in a couple of months, just as Democrats are waving Republican’s assertions from four years ago to wait until after the election to name a new SC judge. And with a similar outcome.

  5. It would be interesting to see the same graphics and follow-on discussion in a Southwest/Delta comparison. In a sense, it would be a more “apples-to-apples” study in that they compete head-to-head in both SLC and ATL, not having to factor for surrounding airports.

    Last year, Delta took in over $830 million in change fees, nearly all of it pure profit. If we assume that both WN and DL borrowed $10 billion for COVID and Delta is paying 2% more in interest (as revealed by Mr. Dunn elsewhere), then that’s another $200 million a year Delta has “lost” to Southwest. So, between the lost change fee profit and the additional interest expense, this is robbing over a billion dollars a year from Delta’s bottom line.

    Where will they make that back? Increase baggage fees? Increase ticket prices? Renegotiate their deal with American Express for a bigger slice of the credit card pie? Pay and benefit reductions for their employees? Fleet commonality savings by going to a nearly all-Airbus fleet (except for the 737)? More third-party maintenance work for Delta Tech Ops? We just don’t know.

    One thing we do know is that without another PSP, Delta begins furloughing 1,721 pilots in 10 days. Mr. Bastain, Delta’s CEO, is in favor of the additional PSP. Mr. Dunn, our resident Delta expert, is against saving those Delta Pilots as he is anti-Cares Act 2. Is this due to Mr. Dunn knowing what’s better for Delta than Mr. Bastain? Or could it be that Mr. Dunn’s near-obsession with American’s demise trump his compassion for his own airline’s employees?

    1. no, we really don’t know what Delta will do because Delta and ALPA are still negotiating. The latter doesn’t want the company negotiating in public and I have repeatedly said both sides should figure out how to address total costs which is what furlough for any airline – including tens of thousands of non-pilot American and United employees are about. Where is your outrage that American and United aren’t apparently doing anything to eliminate their non-pilot furloughs?

      and Delta and Southwest are ALREADY gaining more revenue from American to more than offset the additional interest expense both will pay – which is a fraction of what American and United are paying.

      1. Give me a break. Delta furloughed their entire non-pilot company’s salary via mandatory reduced hours by 25% starting in April And continuing to the end of the year, at least. That was never a choice it was a mandatory reduction in pay. Delta has been equivalently furloughing since April while feeding their poor employees their traditional anti-union bs.
        That alone is the equivalent of furloughing 25% of their workforce even during The middle of the Cares Act support that Delta pocketed instead and said “screw you” to their people. They’re sending flight attendants to be caterers and forcing one month on and one month off for some to avoid the dreaded word “furlough” even though they’ve been hurting their people far more than United and American since April.
        Where’s your outrage? How dare you suggest anything Delta has done reducing salaries is positive or short of furlough. It’s just traditional delta anti-union sleight of hand.

        1. THANK YOU, Jack! You are so right. Delta has imposed what amounts to mandatory pay cuts since nearly Day 1 of this crisis. Just today, Ed Bastain said Delta payroll is down 40% YOY. Wasn’t the Payroll Support Program supposed to make EVERYONE whole until September 30th?? So, where did the money go??? No one has held Delta accountable on that.

          Another perspective here is that if Delta has reduced its payroll by 40% and is still losing $27 million a day, then that speaks to Delta’s vulnerability. If payroll is your largest expense and you reduce it almost in half and you’re still losing $27 million a day, well that’s downright ugly. In June, Delta was messaging the Plan was to break even by the end of the year (with or without a vaccine). Furloughing the 1,721 pilots who are at the bottom of the pay scale won’t cover $27 million a day.

      2. Just back from “Town Hall” meeting with Ed. He realized that another round of PSP prior to his man-made furlough date of October 1st is becoming untenable, so he pushed it back to November. He keeps saying he wants a deal with his pilots, but treating them like other Delta work groups (ie. Stay home. No pay.) instead of the way other carriers are treating theirs will only further fan the flames of their growing animosity.

        One other interesting tidbit is that Uncle Sam wanted the SkyMiles program as collateral for the Loans part of the Cares Act. However, they only wanted to loan $5 billion against it. So Delta leveraged it for $9 billion in the private sector, but alledgedly at a higher interest rate than Uncle Sam wanted. The good news there is not only the extra $4 billion, but also Delta can resume its stock buyback program much sooner than the carriers that take the Loan.

        1. So in other words, your first statement was wrong?
          “One thing we do know is that without another PSP, Delta begins furloughing 1,721 pilots in 10 days”

          When you and everyone else tell us why you aren’t outraged that American and United aren’t saving their non-contract employees, then you might have reason for throwing stones.

          And, Jack, the CARES Act never was designed to replace 100% of employee salaries and benefits and it didn’t.

          and, secondly, are you aware that both Delta and JetBlue reduced employee hours before the CARES Act was even approved?

          I’m not sure, once again, what any of this has to be with change fees other than your compulsive need to throw dirt at Delta while ignoring everything else that matters specifically related to this discussion.

          And you can’t seem to understand or accept that the pilots elected a union and have asked the company to not negotiate in public but yet you have appointed yourself as their spokesman.

          I’m not going to give you the time of day if you want to keep arguing. The topic is change fees.

          1. The question remains, Mr. Dunn: How will Delta replace the $830 million PROFIT it made from change fees last year? Today, Mr. Bastain admitted to losing $27 million a day while lowering his single biggest expense – payroll – by 40%. That number is so staggering to me that I genuinely believe the only reason that anyone would loan Delta even a dime is that United and American are in worse shape than Delta, and even that assumes that business travel rebounds quickly post-vaccine, which isn’t necessarily a given, either.

            1. change fees.
              If you want to discuss financial analysis, start your own blog or contribute to one where that is the subject. But you will probably have to use your real name. It’s a credibility thing.

              It doesn’t matter what you think about why anyone would loan Delta or any other company money.

              Delta pulled off the largest debt financing in the history of commercial aviation by actual financiers.

              Now back to change fees.

              CF’s analysis – not mine or anyone else’s – shows that AA and UA have heavy exposure to WN; if you don’t think that datapoint matters, address your argument to him.
              He has in fact noted the key factors that are in play with the change fee discussion. You would do well to focus on those issues and not a million unrelated side issues.

            2. Look at Tim Dunn recommending people take their analysis to their own blog. Please do, Tim.

            3. This may be an oversimplification but DL can’t reproduce the 830M in change fees simply because they don’t have the pax to produce it right now. Pretty easy for everyone to say no-change-fees when hardly anyone is flying and the potential profit from those fees is nil.

              In a non-Covid world I don’t think we’d be having this conversation right now.

      1. Clearly it would be the fault of those nuns, not Delta… They didn’t look both ways because they were rushing after they arrived to the city late on an AA flight! :-)

  6. I understand change fees for cancellations made 7-10 prior to the flight when flights are flying more than 90% full as it frees up seats that others may want to book.  But if flights are only 70% full, no chance that those reservations that would cancel will prevent other reservations.   But SWA’s policy has made me a firm passenger.  I make many reservations that I am unsure of knowing I can cancel.  Of course, 85% of the time I do fly so SWA gets my business.

  7. As soon as the pandemic is over, change fees will return. The flying public will be reminded once again of the hypocrisy of the companies who bragged about eliminating change fees during a period where they were essentially forced to do so.

    1. My guess is that the more likely case is changeable tickets (anything above Basic Economy) will get proportionally more expensive as the economy improves, shuffling customers back toward Basic Economy, while maintaining “no change fees”. United may have to make Basic Economy more attractive to avoid losing market share when playing this game, but it’s a way to recapture the revenue that change fees provided earlier without flat-out reintroducing them.

  8. Mr. Snyder…..For the sake of comparison, could you please do a similar study between Southwest and Delta? I know you have many other matters on your plate, but for the sake of having an even more complete industry comparison, could you possibly share the data with your readers?

    I know you are concerned about civility on your blog. Each time I post here, I try to use facts. From those facts, perhaps we can divine the industry’s direction, as a whole as well as each individual carrier’s. Your analysis of “The Southwest Effect” as it relates to change fees begs the question: Where will the participating carriers recoup that sizable profit loss? This is particularly relevant in these times when the entire industry (Southwest included) is desperate for revenue.

    Thank you for creating space for ALL your readers, from industry insiders to frequent flyers to casual observers. I learn things from ALL of them.

    1. I was curious as well. I’m assuming by not having a chart, that Delta has less exposure to Southwest than United or AA. But, that’s just an assumption.

      1. Southbay Flier…..Initially, I thought the same thing. But then I remembered Southwest’s purchase of AirTran, which resulted in considerably more head-to-head competition with Delta than in 2004. So that made me curious to see the Southwest/Delta numbers as well.

    2. I would absolutely hate for you to not have an answer so I did a preliminary look at AA vs. WN compared to DL’s US metro area domestic hubs and focus cities. CF can run the data using his own criteria but WN overlaps AA’s network at more than twice the rate at DL’s network.
      More significantly, WN grew in AA’s hubs over the past 5 year while it was 7% smaller over the past 5 years in DL’s hubs.

      IN case you missed it, a big reason why DL has controlled a higher percentage of market share and revenue – is because it controls its metro areas better than any other airline except for WN. ATL, DTW, MSP, SLC, CVG, BOS, RDU and SEA all have ONE commercial airport in contrast to Chicago, Dallas, S. Florida and Washington DC where AA has hubs.

      I’m not sure what readers you think are being excluded.

      Market share WAS mentioned in the article and is a fair discussion. Labor costs and furloughs are not.

      as for your question about how ANY airline will cover fee revenue, that has relevance if you’d like to discuss ALL of the industry, including B6 which CF noted has resisted saying their change fees are gone. But since neither you, I, or CF have any insight as to how any airline will find revenues 2 weeks from now, let alone in 2 years when passenger demand might return enough for the absence of change fees to matter, any comments would be speculative opinion at best.

      And, Delta’s CEO has said that 80% of its corporate travel accounts do have customers traveling now although at small numbers. Travel demand is slowly returning which is why, as CF notes, reworking policies now makes sense rather than once demand returns. Again, Delta publicly discussed removing change fees long before covid. UA just happened to act first – likely because they are seeing more booking erosion to WN now – due to the fact that WN has added back and is growing far more in UA’s hub metros that UA itself has returned capacity.

      I do financial analysis on airlines elsewhere under my own name.

      1. Tim – SEA has a 2nd metro airport in Paine Field (where Delta has zero presence). Also tough to argue that DL has much of a “hold” on SEA when they’re shrinking rapidly and ceding share back to AS, all under the shadow of AA’s international expansion there. Controlling other hubs might be good for short-term revenue, but it also gives you a reputation for being expensive, annoys your customer base, and spark interest from new competitors entering the market. I get that you Stan DL, but please call a spade a spade and stop fudging the numbers on here all day long.

        1. In other words, data doesn’t matter if it conflicts with your pre-selected bias.

          PAE account for far less than 1% of all seats that SEA offered. PAE is a rounding error. It can’t and won’t ever compete with SEA.

          Delta carried a far larger percentage of traffic from Love Field – an airport that American and United chose to abandon – than ALL carriers did from PAE.

          As for your statement about DL’s size in Seattle, DL in 2019 had 48% of the flights that AS had. In 2004, DL had less than 5%; even including NW, DL had fewer seats than WN and just 2/3 of what UA had in 2004.

          Thanks for making my point.

          DL has gained seats in the ENTIRE Seattle metro area at the expense of Southwest and United.

          Want to carry on? I have real data and I am not afraid to use it.

          and I don’t manipulate my data to come up with some pre-conceived notion that is not supported by data.

          DL’s growth since deregulation – just like WN’s – has come at the expense of American and United.

          If market share had anything to do with the decision to get rid of change fees, then DL, not AA or UA, was in by far the best position to dictate the terms of how removing change fares would be handled.

          But UA rushed a decision solely because its network was already the most exposed to WN and, because UA has sat on its duff in adding back even domestic capacity, WN has been free to add capacity in the covid environment.

          And as a result of AA”s multiple mergers in its post 9/11 lineage has lost more share than any other airline to WN – and that trend continues as AA slow rolls its return to PHL and DCA among other cities while flat out choosing not to rebuild LAX.

          Facts and data, not my opinion or anyone else’s, support the understanding of why AA and UA needed to respond to WN while DL was in a position of growing network strength not just relative to WN but also AA and UA.

    3. MissTheMasters – I didn’t bother running Delta, because Delta didn’t lead on this. It wasn’t, as others have suggested, anything to do with a lower level of overlap. That being said, I did figure DL would have a lower level of overlap. I tried to run this again, but I got some weird numbers. And frankly, it’s too time consuming to do again right now.

      What I do show is that in 2004, route overlap was 6%, flight overlap was 7%, and pax overlap was 56%. You can see why I’m skeptical. But in 2019, I have route overlap of 24%, flight overlap of 36%, and pax overlap of 76%. It seems like something may be off with the pax numbers, but from the rest, it is clear that there is less overlap. That being said, it has grown mightily.

      1. Thank you. As I suspected, Southwest’s direct competition with Delta has grown tremendously, due primarily to the AirTran buy. Of course, those numbers would be even higher had Delta not abandoned its DFW hub years ago. As you deciphered, WN will pursue market share in the West, where it is strong. But, if I’m Gary Kelly, I’m also watching to see if any opportunities open in the East, particularly in slot-controlled airports.

        1. I thought in Cranky’s analysis that AirTran was included in Southwest’s numbers even before the merger. Plus, IIRC, Southwest shrunk Atlanta a bit after the merger. My initial impression was that Delta would have less overlap because their hubs in Detroit, Minneapolis, and SLC have a much smaller Southwest presence than their competitors hubs in Chicago, Detroit, Dallas, Houston, and Phoenix. But, that’s all just a guess.

          1. southbay – No, AirTran wasn’t included in the 2004 numbers, because the point was to compare overlap with airlines that have no change fee. Since AirTran had a change fee, I left that out.

  9. (Insert emoji of head banging against wall here). Mr. Dunn, The Cranky Flier’s comparison goes back to 2004, not 5 years ago. You are cherry-picking data after the Southwest purchase of AirTran, when Southwest decreased Atlanta departures. Let’s compare apples to apples here.

    1. 2019 is still more recent than 2004. and if the point is to show a larger WN, then going back to 2004 only shows how much further WN has shrunk in Atlanta.

      If you noticed, he included the merger partners for all airlines which is the correct way to do a long-term comparison of market share.
      FL/WN is far smaller today in Atlanta than they were at any time since 2004.

      Add in that WN has shrunk in BOS, DTW, SEA and SLC while DL has grown in those cities and yet WN has grown in Washington, Dallas, Denver and other cities and actual data does show that WN has far more overlap in AA and UA’s hub metros than for DL.

      And I’m still not sure what your fixation is on the point; DL said it was looking at eliminating change fees even before covid so they not only have looked at the competitive environment but have thought through the financial implications.

      Delta has the highest market share in its hubs of any carrier except for WN. That translates into better pricing power including the ability to get higher average fares.
      and more significantly even in the covid era, WN’s seat share is growing in AA and UA markets but not in DL’s.

      No, I am not cherrypicking data. The point is the same for 2004 as it was for 2015.

      in case you haven’t figured it out, AA and UA have lost market share to other airlines including DL and WN for all of the 40 years of deregulation. that trend is not changing and will accelerate as covid recovery continues.

      Again, CF can post his own analysis but a real mindset of seeking understanding means accepting facts and data and not manufacturing your own theory whether data supports it or not.

      1. Tim – I did NOT include AirTran because AirTran had a change fee. My point was to show the overlap with an airline that had no change fees.

        1. Even if you did not include FL in 2004 but did include WN later, WN operated around 100 flights/day at the end of 2019.
          They have cancelled far more flights in other Delta hubs – BOS, SEA, DTW, SLC and more – than even the 100 flights/day they gained at Atlanta.

          And even if you chose to only include true WN internal growth, then you can’t include most of WN’s flights from Atlanta because most are in markets which FL served. You can’t choose to exclude FL’s flights but then count WN operating in the exact same markets.

          There is an enormous difference in the network overlap in metro areas between Delta and American and United. Delta and Southwest share focus cities/hubs in just one Delta hub (Atlanta) and even there Delta was more than 10X larger than Southwest – plus LA. In contrast, American shares hubs/focus cities in metros with Southwest in Chicago, Dallas, Phoenix, Washington DC and South Florida while for United it is Chicago, Houston, Denver, Washington DC, the Bay Area and LA for both AA and UA.

          The direct network overlap between Delta and Southwest is nowhere near in the same league as between Southwest and both American and United.

          and I’m not sure the whole hub argument makes sense anyway. WN competes with all carriers at the O&D level anyway. Most passengers choose airlines based on nonstop flights but Southwest most certainly does compete on the basis of its network even in markets where it does not operate on a nonstop basis just as the legacy carriers do.

          You can compare domestic O&D growth since 2004 and it clearly shows that DL lost a smaller amount of system market share than AA or UA has to WN over the past 15 years. AA’s share loss has been the greatest precisely because its lineage includes two mergers compared to one each for DL and UA.

          But if you wish to do that exercise, you need to exclude FL’s market share from WN’s 2019 numbers; FL had 2.5% of the US domestic market in 2004. You can’t decide not to include FL’s market share but then give WN credit for the same O&Ds which it inherited as part of the merger.
          Excluding FL’s domestic O&D share, AA, DL and WN all had nearly identical O&D market share but WN’s share generated far less revenue.

          THAT is a true apples to apples comparison.

          And, honestly, what is the point in including and excluding markets to create a share story specific to change fees?

      2. WN doesn’t have the highest market share of its hubs than other carriers. I have the highest market share of the airport it operates out of. But DAL, HOU, BWI, MDW and OAK are all low fare airports compared to their legacy rival airports. Not that it really matters that much, since WN still gets great margins due to keeping its unit cost low.

        AS at SEA is the most dominant non-network carrier hub in terms of pricing power, connectivity and market size. That no doubt has to do with the fact that it is also a legacy carrier that had decades to build that hub.

        Any look at DL needs to break its hubs/focus cities to 3 groups:
        1) Core hubs – ATL/MSP/DTW/SLC – great fortress hubs that generate better margins than any other hubs
        2) Large coastal hub/focus cities – NYC/LA/CVG – I’m assuming they will become largest at LA here. They are the largest or close to the largest carrier in fragmented market or small market. Likely to be below average margin hubs for them but can still be quite profitable in good time.
        3) Small coastal hubs/Focus cities – SEA/BOS/RDU – Vastly under-performing market here compared to rest of their network

        Sure they can keep growing market share in 3), but there is a limit to how large they can get and they will always be vastly below system wide margin.

        1. FC,
          As always, I enjoy chatting with you on here.
          I’m not sure that your statement about WN came through as you intended.
          Feel free to show us the data you have but WN most certainly has a higher share of its “hub” airports if they are viewed solely on the basis of airports. It is precisely because WN has such a high share of MDW, HOU, DAL, BWI etc that their overall system market share is so much higher than any other carrier.
          If your point is to compare metro area to metro area, then WN does not have the highest market share precisely because nearly all of their largest “hubs” are in multi-airport metro areas where another carrier has a hub at the other airport.
          WN’s strategy from day one was to dominate the airports where it operates and to keep other carriers out – which is why it is still fighting to kick Delta out of Love Field where WN already has 95% of the traffic.
          If the focus is on metro area dominance, then Delta has the highest share of its metro areas – and that is precisely because its core 4 hubs plus BOS and SEA are in single airport cities (again noting that PAE is a vanity project that does not and never will move the dial in the PNW).
          As for your statement about AS and their SEA hub, looks all the world to me like you wrote a definition to define AS as the only player. AA, AS, DL, HA and UA are all legacy carriers and hub and spoke operators, even though WN carries a higher percentage of traffic through some of its “non-hubs” including MDW and BWI than legacy carriers do from some of their hubs including their NYC and LA hubs. Compared to other legacy carrier hubs, AS at SEA is at best mid-tier in terms of pricing power, connectivity and market size.
          Other than to continue to rehash internet lore which is not substantiated by data, I’m not sure what your point is regarding Delta’s hubs and focus cities. Your statements about margins can’t be proven any more than hub margins for any carrier can be proven – and they can’t because airlines don’t report hub specific data; costs differ considerably from hub to hub. Delta’s network from BOS, RDU and CVG for example is far more point to point than their system which means trying to compare system costs doesn’t logically work. Then you get to specific aircraft costs; half of Delta’s flights from BOS are on E175s which are a lower CASM aircraft than JBLU’s E190s, which also make up about half of their flights, based on DOT filings, not my opinion. The 737-900ER was the most commonly used aircraft at JFK by DL in 2019 and it has a lower CASM than B6’ A320s.
          On the revenue side, the whole reason why Delta has been able to rapidly grow in both Seattle and Boston against lower system cost competitors is because Delta has not chased every local passenger at the lowest fare but instead won over the highest revenue passengers. Delta carried 75% of AS’ SEA local market revenue and 100% of B6’ BOS local market revenue in the summer of 2019 even though Delta was much smaller than AS in SEA or B6 in BOS; a few people love to point out that B6 gets a $2 higher fare than DL to a dozen Florida markets but ignore the fact that DL gets far, far higher average fares to its core 4 hubs which makes up a much higher percentage of DL’s revenue. AS and B6 both saw their margins drop as they tried to flood their hubs with capacity and pushed down their own yields; DL focused on the international market, strength to cities where it is dominant and still managed to get about equal fares in the dozens of leisure markets where Delta CHOSE to offer far less capacity than AS or B6 but still have enough to be relevant in the market.
          Delta’s system margins which were neck in neck with Southwest, both of which were better than any other airlines, proves that Delta’s strategies did work better than AA, UA, AS B6 and NK etc pre-covid; trying to argue that DL must be making boatloads of money in its core 4 hubs such that it can lose money or underperform in BOS, CVG, LAX, RDU, SEA etc doesn’t pass the logical sniff test since DL grew faster outside of its core 4 hubs even as its margins increased.
          And again, I’m not sure what all of this discussion will accomplish esp. relative to finding any proof or causation regarding change fees.
          What is clear is that no airline’s covid strategies have yet been proven to be sustainable which means that airlines have to pick what markets they want to hold onto and what markets they are willing to sacrifice in the name of survival. WN gained so much share in DEN post 9/11 because UA didn’t have the financial strength to try to fight back – and that same scenario is setting up once again not just in DEN but also in EWR with B6 and others. DL has not given up any seat share in any of its hubs and focus cities based on current schedules; they offer the same or more capacity in BOS, SEA, RDU etc as their competitors compared to pre-covid. DL has gained seat share in LAX as well as BNA, one of their newest focus cities, as well as in NYC. AA clearly does not have the financial strength to compete in the biggest markets any longer and has retreated to a subset of its “core” hubs.

          Change fees won’t matter in whether WN or other carriers gain share at AA and UA’s expense; AA and UA’s lack of seats and flights in competitive markets will make the difference.

          1. Right, I think it’s more important to measure market share as part of catchment area rather than the airport itself. That’s one thing DL has that most airlines don’t have. AA has dominant market share at DFW and CLT, but it has such a high cost that the pricing power in these markets haven’t enabled them to generate higher yield. So, DTW/MSP end up probably having the highest margins of any hubs in the country. While WN has 70+% market share in quite a few airports, they are generally going to be in metro area that have at least 1 high fared airport. Just comparing DFW/DAL, the fare level of AA and WN is quite a contrast.

            i think you are little too stubborn when it comes to DL and don’t see where they are under-performing. A while back, I did a comparison of DL’s yield from LGA/JFK to a set of airports around the country and compared them to BOS. The stage length were all longer out of BOS, but the yields out of NYC airports were all higher outside of MCO. In many cases, the yields were more than 40% higher out of NYC. Now, part of that is the O&D vs connection. But even adjusted for that, the yields out of NYC was quite a bit higher than BOS. Now, we know NYC was at best a system average station for Delta. So it’s quite reasonable to conclude Boston was doing pretty poorly on the routes I tracked. It seemed like DL did well in many of the routes that faced no JetBlue competition, but it’s just a matter of time before JetBlue adds stuff like IND/CVG/CMH.

            And out of SEA, I did an analysis of 15 to 20 routes and compared connection adjusted yields between AS/DL. Aside from DTW/MSP/ATL, I can’t see any other routes where DL could’ve been generating higher margins. There is no other DL hubs where they are dominated that much by another carrier. Now, I see SEA as a station that DL views importantly for network and strategic reasons, so I don’t see them giving it up.

            As for your argument about revenue, keep in mind that a large part of DL’s revenue gain is due to their international flying and their JVs. Just think about the amount of ASM on 1 daily widebody flight to ICN on KE and how much revenue that would bring in. Of course, the cost for that is also quite high. Similarly, think about the amount of ASM on 4 daily widebody flights to CDG.

            At some point, you just have to admit that you have hubs that are obviously below system average in margin. That’s inevitable. In the case of SEA/BOS/RDU, these are quite a bit below system average. DL has been very public that it will be shrinking quite a bit. They and other legacies are likely to be 20 to 30% smaller by the end of next year when LCCs are likely to be back to 2019 size. You should expect them to cut back in places that are under performing and not strategically important. It doesn’t appear to me that DL has figured out how they want to re-orient their network based on the actions they’ve taken so far. Say what you will about AA, at least they have a clear cut strategy of throwing in the towel in NYC/LAX and growing DFW/CLT.

            1. No, I don’t have to admit that there are hubs that are below average because I haven’t seen the data to support it.
              When you tell me that you have looked at 10-15 markets from a hub that serves scores of cities, you haven’t seen the whole picture.
              When I pull up DOT data for all of 2019 for SEA and DL has higher average fares in the local market for LAX, SFO, HNL, ANC, SLC JFK and all of those are major markets for AS, then you clearly are not choosing to see the data that tells the whole story.
              As for BOS, just like for AS in SEA, B6 threw tons of capacity into markets and depressed the average fares. The number of markets where DL has higher average fares in BOS than B6 is not a small number and the difference in average fares is not insignificant either.

              I get it. You and others want desperately to believe that DL has expanded into a bunch of new markets, doesn’t make money, so as to be making a fortune in their core 4 hubs in order to fund all of their aggressive expansion.

              Problem for you is that full and complete data does not support your theory either at the market level or on the system level.
              As hard as it is for you and others to admit, DL’s net income margin for 2019 was significantly higher than AS or B6. DL was growing its presence in these hubs and focus cities that you think are low margin and yet their margins went up faster.

              I am sure you will believe what you want to believe but there really is not evidence that DL is walking away from its pre-covid strategies or that any other carrier is going to succeed at wresting market share away from Delta.

              As hard as it is for you or anyone else to admit, the two most vulnerable carriers in the industry right now are AA and UA; everyone’s strategies including growing into AA and UA markets.

              If UA thought that getting rid of change fees would make a difference in what is shaping up to be another market share capitulation in its top markets, then I am pretty sure they will be in a for a rude awakening. As for AA, change fees don’t matter if they continue to withdraw from major markets leaving customers to choose airlines that continue to serve those markets.

            2. FC
              just a few more notes to respond to the last part of your post.
              the reason the big 3 MIGHT be down more than LCCs in the future is because of the longhaul international markets that are going to be slower to return.
              LCCs don’t serve those markets.
              The true apples to apples comparison is domestic markets and on that basis, based on current schedules, LCCs as a group are not adding back more capacity than legacy carriers.
              CF regularly posts capacity comparisons and B6 is actually returning less capacity than some legacy carriers.
              B6 just released capacity projections and they do not envision more capacity returning than AA or DL; UA is by far in the lowest amount of capacity being returned.
              Your expectation that DL will drop what you think are strategically unimportant or underperforming routes is flawed because you are fixated on DL’s statement that will be smaller.
              In fact, EVERY airline is and will be smaller; DL’s admission of the reality that everyone else knows – said or not – does not make Delta or anyone else in any better position.

              The mere reason why no airline has managed to significantly return more capacity than any other is because every airline is chasing the same domestic passenger right now; there is far less demand than can fill all of that capacity so it comes down to what carriers can hold out and compete. As much as you want to believe otherwise, B6 is operating far fewer frequencies in their previous top markets and are spreading out their flights over more markets, hoping they can fill planes in markets they have never flown.

              You are entitled to your opinions but when you write things that clearly are based on biased interpretation rather than data, you can’t possibly come to an accurate conclusion.

              I’m sure we will beat this discussion until there is a clear answer in place – but I still believe that DL and WN will be the two megacarriers that are capable of holding onto their key markets and repelling competitors that try to invade their markets while AS will largely stick to its core markets. B6 will continue to flail around trying to find markets that will work for them but very likely will not end up any better than anyone else.

              I am sure we will have enough data 6 months, 12 months and 18 months from now to see who was right.

              I still enjoy the conversation.

  10. I’ve never understood the love affair with Southwest and it’s been even more difficult to choose them in recent years. I live in Chicago and am very lucky to live roughly equidistant from both MDW and ORD. Chicago is a great market as 3 of the 4 majors have a huge presence here so there’s lots of competition. Southwest is rarely the low cost option. Maybe there used to be a “Southwest effect” but that’s long gone. Instead I find that prices are 10-15% more expensive. Moreover, the boarding process is horrid. There’s no power on flights. Most of the passengers act as if they’ve never been on an airplane before. And Rapid Rewards is a joke. Do I like the flexibility? Sure. But there’s no reason to save up the points as there’s nothing aspirational in their route network. I’ll gladly give my $ to United or AA and enjoy a more expansive network, greater connection options if there’s a delay, and points that can get me to Asia, Europe, or South America. But I’ll give Southwest credit – their marketing machine is great and so many of us still think they’re the cheap, friendly airline.

  11. Wow, didn’t realize how far behind the big 3 (Delta, SW and AA) United was in share of seats.

    My other thought – If only hotels would get rid of resort fees.

Leave a 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