Just How Long Airlines Delay Flights


After writing up my recent 14-hour delay on Hawaiian, GS left a comment asking about the average length of delay for each airline as opposed to just the percentage that are delayed. At the time, I responded “it’s a pretty large and annoying project” for me to pull all those flight-level details, but then I decided… ah, what the heck. Let’s do it. I snuggled up with Anuvu on-time data one night, and now it’s time to talk about the results.

I didn’t want to just look at the average length of delay, because that isn’t entirely telling. It can get skewed. Instead, I decided to create bands of delays to show what percent of flights for each airline fell into each group. Just extracting the data was going to take some time, so I settled on looking at just a single month. Naturally I took the heart of the summer and used this past July.

Before we look at the numbers, keep in mind that these are all by marketing airline, so regionals are included. Now, let’s take a look.

Delay-Length Bands by Marketing Airline – July 2026

Data via Anuvu

The lightest red at the bottom shows arrivals within 14 minutes of schedule (or earlier), so that is the number that most people are used to seeing as a measure of on-time performance. And the numbers aren’t all that surprising with Alaska/Hawaiian in first place thanks to their good operations and more benign weather. Then it’s Delta with United close on its heels.

JetBlue has had a miserable summer, and the weather has just been brutal for them. But they’re always at the bottom. Frontier is bad, no surprise there either. But then there’s Southwest which from this metric made American look good.

On a recent episode of The Air Show, I spoke with Southwest COO Andrew Watterson, and he talked about what the airline was doing to fix its issues. But he also noted that many of the delays were short, and that is important, because it is far less disruptive for travelers. This tells that exact story.

As you can see, Southwest by far had the most delays that fall into the 15 to 60 minute category. Remarkably, if you look at A60 — the two bottom categories combined — Southwest jumps to third place behind Alaska and Hawaiian. (Technically, that’s second place, but I broke out the Hawaiian and Alaska operations to show the differences there.) So Southwest’s delays are annoying, but they aren’t going to be as much of a fire-breathing, anger-spewing, “I hate you x airline” kind of delay, at least 91 percent of the time.

A one to two hour delay is enough to make a lot of people made, and this is where we shouldn’t be surprised to see JetBlue at the top with nearly 10 percent of its flights. Like I said, combine an already weak operation with exceptionally poor weather, and you are going to be in trouble.

Frontier and Allegiant follow, and that’s probably because when things do go wrong, they try as best they can to eventually run a flight. I say “try” since Frontier still canceled 2.8 percent of flights and was below average. But Allegiant succeeded, being the only big airline that canceled fewer than one percent of flights. Allegiant does run a better operation than Frontier pretty consistently. Of the big guys, well, American is worst to the surprise of nobody. United did pretty well for itself.

Now let’s look at the really bad stuff, delays of over two hours. This is where gate agents start to need bodyguards. And sure, JetBlue is again the worst performer followed by Frontier and Allegiant. Shocking, I know. But to be fair, American is actually tied with Allegiant.

Breeze stands out to me here. It was doing fine in the upper middle part of the pack in the one to two hour delay band, but over two hours? It has a lot of those. That’s definitely something to keep in mind.

I would see these are the main takeaways from a big picture perspective:

  • JetBlue and Frontier are the biggest problem children
  • Southwest has a delay problem, but you won’t be delayed that long
  • Of the big three, American is the worst
  • Alaska and Hawaiian are doing great, even if my flight was delayed 14 hours

A Regional Look

I did go a little bit deeper on the four airlines that have branded regional flying to see if there were any big differences there. And there were some variations, at least for three of them. Alaska wasn’t all that different whether mainline, SkyWest or Horizon.

For Delta, SkyWest was over 81 percent A14 while Endeavor was under 68 percent. Geography may have something to do with that, but Republic flies all around the northeast too, and it was at 75 percent, so Endeavor looks like a problem. Piling on, Endeavor had nearly eight percent of flights delayed more than two hours while Republic was just over five percent and SkyWest was under four percent.

At United, the script was flipped. Mesa and Commute Air were best at over 80 percent while SkyWest was worst at just over 72 percent. SkyWest and mainline were actually almost the same, actually. Looking at long delays, Mesa was by far the best. It is shocking and somewhat disorienting to be able to say that, but once you’ve retired all those ancient CRJ-900s and focused on relatively new Embraer 175s, I guess that can happen.

And then there’s American which had the biggest swings of all. You know what? Let’s just put this in a visual to look at.

American Airlines Delay Stats by Operating Airline – July 2026

Data via Anuvu

Look at Envoy doing a great job, and then look at both mainline and PSA doing a terrible job. PSA, of course, flies up and down the east coast, so that’s a problem. But guess what… so does Republic.

I know geography goes into a lot of these issues, but Endeavor under Delta and PSA at American look like the regional airlines to avoid if you can.

A New York City Look

I know I keep talking about how geography impacts these results, so let’s take a look at New York City to get a better sense of what I’m talking about. To be clear, everything was bad there, but some were worse than others. For this, I looked at a combined JFK and LaGuardia for American, Delta, and JetBlue. Then for United, I looked at Newark.

NYC Delay Performance by Marketing Airline – July 2026

JFK/LGA for AA/DL/B6 and EWR for UA Data via Anuvu

This tells a story of relative consistency. JetBlue is relatively bad, of course, but that is consistent…. But the big three don’t vary all that much from each other. United is a little better about getting planes in on time, and Delta is a little better at keeping delays shorter. We can pull all of these apart and find their own weak spots, but the point is that systemwide data matters. Just keep in mind that individual airports are going to give different results.

Get Cranky in Your Inbox for Free

How often do you want to receive emails?
Brett Avatar

77 responses to “Just How Long Airlines Delay Flights”

  1. LT Avatar
    LT

    PSA has a third of its flying in DCA which has had significant struggles from a throughput perspective since the accident. Tag on the east coast traffic management issues with terrible weather patterns and you have a recipe for a delayed airline.

  2. 1990 Avatar
    1990

    Brett, thank you for analyzing and writing on this topic.

    As a NYC-based flier, I consider myself an operational reliability hawk because we consistently bear the brunt of airspace congestion, FAA staffing friction, and severe weather across the Northeast corridor. (Oh, dear haters, please spare me the tired ole “no one wants to work anymore” excuses; the real issues are brittle asset utilization and razor-thin crew buffers). The data shows the worst offenders (yikes, jetBlue. This summer has been brutal for them; but, I’ve also experienced regular 3-hour delays on the day of travel with them in the past couple years, too.)

    Looking past sanitized A14 monthly averages to examine actual delay durations exposes the structural rot in domestic scheduling. When you contrast our market with the UK and EU, the lack of accountability is the difference. Under EU261 and UK261, carriers face mandatory cash compensation liabilities for significant delays under their control. In the US, travelers simply absorb the cost of poor operational planning while airlines face no financial deterrent for leaving passengers stranded for hours. (Yes, friends will know, this is the ‘hill I die on’ across the various blogs.)

    Carriers heavily exposed to the Northeast like JetBlue suffer because lean networks lack the frequency to absorb rolling disruptions. Until the DOT implements enforceable passenger rights that mirror European standards, airlines will lack the bottom-line incentive to build true operational resilience.

    For now, personally, when it’s an important trip (a lot of prepaid expenses on the line, etc.), I both purchase comprehensive travel insurance, and sometimes also book multiple back-up itineraries, usually using points, which I can cancel for a full refund, all as protection against these ‘games’ and shortcomings. No one should have to do all that just to make sure they actually make an appointment/cruise/wedding, etc. at their destination. But here we are. (We really deserve better.)

    1. John G Avatar
      John G

      You are correct that these airlines operate at the ragged edge…they have just enough planes and crews to operate the schedules they run. Little if any reserve, of planes, crews, gates, runways…etc.

      These airlines want to fill every seat on every flight. And they want to utilize their aircraft as many minutes they can keep them in the air. All that is great, until problems arise. Machines need maintenance and break down. Humans get sick. Airports have weather issues. And when all that happens, there is no reserve or slop to fall back on, and it becomes a giant mess.

      But here is the problem with your premise. What’s the fix? Cutting the schedules and forcing them to hold more reserve capacity to be able to handle IRROPS better. Right? That’s really the only way. You can’t fix it without having extra crews and planes at the ready when things get messed up. But that isn’t free.

      How much more are you willing to pay in airfares for that reserve? Because that’s who would pay for it. Us, the passengers.

      1. See_Bee Avatar
        See_Bee

        Cutting schedules isn’t the only solution. What about extending block time and ground time? What about more proactivity ahead of weather to set up a better recovery? (e.g., keep aircraft & crews in better positioned stations, whether than be NYC or the outstation)

        I look at the added costs for these levers as something people aren’t willing to pay B6 for now because of reputation, but I think DL would argue they earn a revenue premium to pay for those costs based on their “investment” in the operation

        1. SandyCreek Avatar
          SandyCreek

          WN is making boarding times start earlier, so airlines do have control and do exercise it. However, aviation is a very capital intensive and relatively low margin industry, and efficiency does matter to airlines.

          1. See_Bee Avatar
            See_Bee

            Right, so how are DL and WN’s profitability compared to B6?

            1. 1990 Avatar
              1990

              (This’ll make Tim’s day…) Yes, for the most part, Delta has carved out a competitive advantage because they invest enough back into their operation to command a revenue premium (Oh, and Amex revenue, yeah… that, too.) Travelers are more than willing to pay for reliability; the problem with carriers like JetBlue is that they got caught in a race to the bottom on bare-bones scheduling where a single disruption destroys their reputation and their bottom line. (Again, see my comment about booking back-up flights.)

          2. 1990 Avatar
            1990

            Sandy, well aware of margins and capital constraints in the industry; but, ‘efficiency’ shouldn’t mean running an operation so stripped down that the first afternoon thunderstorm causes a multi-day systemic collapse (*looking at B6*). When airlines face real liability for controllable delays, “efficiency” starts to include building enough structural buffer so that a single rolling delay doesn’t snowball into an operational meltdown.

        2. John G Avatar
          John G

          If you run an operation at the ragged edge, where you have only enough crews and planes to run a schedule on good days, you don’t have room to extend block time or be proactive ahead of weather.

          If you keep only enough planes and crews for a normal day, there is NO reserve for IRROPS. There are no planes, crews, or seats.

          So they would have to hire more crews, and buy more planes…or schedule fewer operations. All of that costs money, but you aren’t adding flights. Where is the added money coming from? Can only come from fares.

          1. 1990 Avatar
            1990

            John, you’re describing the loophole that I’d like to see fixed. In the U.S., there are no statutory delay penalties, meaning airlines can strand passengers and walk away scot-free. Now, if you’re invested in or working for an airline, maybe the status quo is great for you. But, for passengers, it ain’t great.

            Without real liability, airlines externalize all the friction onto the public and treat rolling disruptions as a cheap cost of doing business. Forcing them to internalize those costs is the only way to make operational buffers cheaper than continuous meltdowns.

            That’s why the bailout playbook needs to change. The next time these carriers come to Washington looking for a safety net during a downturn, Congress shouldn’t hand over a blank check—they should staple mandatory passenger protections, real duty of care, and competitor rebooking requirements to the legislation.

        3. Mac Avatar
          Mac

          Extending block time and ground time is exactly cutting schedules. US passengers will only board so early / arrive so late. When you add time to the schedule in the air and on the ground, you’re cutting a flight out of the day aka cutting schedules. And some flights need to operate when they do in order to meet a bank or maintain a frequency that’s competitive and a time that’s in line with business travelers demand

          1. 1990 Avatar
            1990

            Mac, extending block time blindly does cut rotations, but that’s a false choice. True resilience is about strategic buffer windows and reserve aircraft placement so a morning glitch doesn’t cascade.

            Sure, airlines want to keep their schedules packed to look competitive for business travelers. But what good is a “frequent” 7:00 AM flight if it’s delayed six hours because the airline squeezed the plane too tight? Business travelers don’t want a flight time on paper that constantly falls apart in reality; they just want to get there reasonably on time.

      2. 1990 Avatar
        1990

        John G, empirical economic research on passenger protections upends the claim that rules like EU261 trigger severe fare hikes or economic collapse. Peer-reviewed research by economists Hinnerk Gnutzmann and Piotr ?piewanowski (Transport Policy, 2023) tracking flight reliability under consumer regulations shows that when carriers face mandatory delay liabilities, they prioritize prevention over payouts because investing in operational buffers is cheaper than constant compensation. Their data demonstrates that flights covered by strict passenger protections are up to 70% less likely to experience severe (3+ hour) delays compared to the U.S. market, achieving that reliability boost at a negligible per-passenger cost ceiling of less than €2. When carriers operate without financial liability for controllable disruptions, they shift all the friction onto the public. Accountability forces operators to abandon hyper-brittle schedules and build baseline operational resilience.

    2. Bill from DCA Avatar
      Bill from DCA

      1990 – that’s a really interesting post and it is certainly absurd to have to book multiple itineraries but at least you’re able to do that. most of us without massive amounts of points are only able to hope for the best.

      I’m not familiar with the EU px protection laws (other than that they exist) and wonder how much of that would be workable in the US?

      1. 1990 Avatar
        1990

        Yes, Bill. The current workaround—layering backup tickets, relying on elite status, or hoarding points—is a luxury that excludes the vast majority of travelers. When the baseline system is that fragile, flying becomes a gamble for anyone without a fat wallet.

        As for how EU-style protections would work here: the mechanics are entirely straightforward, and U.S. carriers already comply with them every single day when flying westbound from Europe. They fight it here tooth and nail simply because it strips away their free pass.

        Mandatory duty of care and cash compensation for controllable delays didn’t destroy the European market. (Ryanair complies with it daily while printing profits, paying union labor, and slinging dirt-cheap fares.) It just forces airlines to factor reliability into their baseline operating costs instead of externalizing the pain onto stranded passengers.

        1. David M Avatar

          I was wondering if someone would bring up Ryanair. The fact that they continue to operate profitably under EU261 shows that these sorts of passenger protections doesn’t have to mean airfares skyrocket.

          1. Tim Dunn Avatar
            Tim Dunn

            Ryanair also heavily uses non-congested secondary and tertiary airports.

            Delays and cancellations in the US heavily are centered around the largest and busiest airports, esp. in the NE where airspace is so tightly interwoven.

            The appropriate comparison is between US big 4 carriers and the EU3 groups because they have similar rules.

          2. 1990 Avatar
            1990

            You’re welcome, David. Ryanair and other LCCs operating in the UK and EU are a great, reliable example supporting the position that EU/UK261 works there and can work in the U.S. as well. In fact, U.S. carriers already comply with these exact rules every single day on flights returning from the UK and EU.

            And Tim, when you look at the major EU legacy groups operating out of congested tier-one European nodes every single day, they manage to comply with strict statutory protections while running hub-and-spoke networks just as complex as ours. So, it is not differences in airspace geometry. This is all about whether the airline carries the financial risk of its own operational fragility, or successfully dumps it onto the passenger. So far, US corporations have duped the American public (and our elected officials); I think that should finally change, hopefully sooner than later.

    3. Bravenav Avatar
      Bravenav

      Yes, Europe has stricter penalties for airlines, but does that result in better performance?

      1. 1990 Avatar
        1990

        Bravenav, to answer your question directly: yes, the data shows it genuinely does. As I noted above regarding the empirical research, carriers facing mandatory delay liabilities prioritize prevention over payouts. Flights operating under EU261 protections are up to 70% less likely to suffer severe (3+ hour) delays compared to U.S. domestic routes. Accountability fundamentally realigns operational priorities so that carriers build resilience rather than treating passenger delays as a free cost of doing business.

        1. Bill from DCA Avatar
          Bill from DCA

          Even if it didn’t help airline performance, which it certainly should considering airlines don’t like giving money away, at least it provides some attempt to compensate passengers.

          1. 1990 Avatar
            1990

            Bill, I agree. Regardless of the performance and reliability boost, it’s about consumer fairness as well. Right now, when an airline breaks its contract and strands you for hours due to poor planning, you and I (no matter what our background, partisanship, state of residence, etc.) bear near-100% of the financial and personal fallout while they keep your cash and walk away scott-free.

            (Of course, they only risk losing future customers if this isn’t a heavily captured market—or if you aren’t trapped flying out of a fortress hub where they hold the monopoly keys, or if you remember that these legacy carriers are basically multi-billion-dollar credit card marketing companies with a mere ‘side hustle’ of flying airplanes… I digress.)

            Statutory compensation ensures that when the disruption is within the airline’s control, the passenger isn’t left holding an empty bag (or sleeping overnight at Dulles instead of reaching their destination). And, even when it’s weather-related, there should be a baseline duty of care afforded to all passengers (Canada’s APPR falls short here, but EU/UK rules mandate it—things like meals, accommodations, and getting you rebooked).

            1. Bill from DCA Avatar
              Bill from DCA

              Agree completely. Thanks for the interesting insights.

  3. Southside Emil Avatar
    Southside Emil

    Cranky,
    Maybe someday you could run a similar comparison at ORD between UA and AA to see who is worse.

    1. Brett Avatar

      Southside – That’s easy, I have that data anyway. They are almost identical. AA’s A14 is 69.9% while UA’s is 69.2%. Both are 5.9% over 2 hours.

  4. Bill clean Avatar
    Bill clean

    I was hoping to see Air Canada in here as well given their prominent space in the US transatlantic market

    1. Bill from DCA Avatar
      Bill from DCA

      No way, eh? Take off, you hoser!

      1. 1990 Avatar
        1990

        Sorrie? Pro-cess.

        1. Bill from DCA Avatar
          Bill from DCA

          it’s aboot time

  5. See_Bee Avatar
    See_Bee

    The last chart is why I think B6 quarterly earnings calls are comical. They ALWAYS have some line around “oh woe is us, we operate in the most congested airspace…” but when you look on a like-for-like basis (i.e., NYC ops stats) they are always the dead last

    Hmmmm feels like there are some controllable levers you can pull, guys

  6. Bill from DCA Avatar
    Bill from DCA

    I wonder if you could create an overall “operations annoyance” metric using this data as well as cancellations.

    A simplistic approach would be to assign weights to the categories above such that delays of 15-60 minutes are 1x, 60-120 2x, 120+ are 3x-5x, cancellations are 10x, or something along those lines.

    As with many of my best ideas, something like this probably exists already! Regardless this was a very interesting deep dive into the data.

    1. Bill from DCA Avatar
      Bill from DCA

      or perhaps just grafting cancellations onto this graph after the 120+ bar? delays plus cancellations would be a good summary of passenger inconvenience.

      1. Brett Avatar

        Bill – Hmm, you know, I’m going to look into something like this

        1. EasyMoney Avatar
          EasyMoney

          Hi, Ed Bastian here (from Delta). Sorry I went on that other podcast but please don’t do this analysis <3

        2. Bill from DCA Avatar
          Bill from DCA

          A passenger pain index? Passenger pain predictor (aka P3)? Get your creative types on a catchy name, those were my best efforts lol

  7. Teton Bruce Avatar
    Teton Bruce

    I fly a lot – 52 flights so far this year. My total on-time (A15) has been 80.8%. Ironically, American mainline has been the best. Until my last flight, they were at 100% A15. My last flight blew a tire and had a diversion, so we were six hours late. AA overall has been 92.9% on 14 flights. AA Regional has been terrible. PSA was almost an hour late. They had a mechanical on the first turn in CLT and never recovered. My next AA regional was SkyWest. Same thing, same result. 3rd flight was SkyWest…landed 20 early. But the flight from ORD that was late arrived 5 minutes ahead of us and they only had one gate…we waited and arrived 30 late (own goal by AA). 4th flight was SkyWest and was ontime. Regional 25%. My other airline is Alaska. AS hasn’t been great but they fall into that short delay window. A couple of flights left on time and arrived late due to headwinds. AS is 81.3% on 16 flights. Horizon had one major mechanical and they sit at 71% (7 flights). SkyWest hasn’t missed, 100% on 5 flights (83.3% on AS regional). Then I have 6 flights on HA. Only one delay which was a neighbor island flight and it was only 23 minutes late. 10 delays…2 over 2 hours, 1 over 1 hour, 7 under 1 hour. I hope the rest of the year plays out as good or better than my first half with no inflight diversions.

  8. AA-Platinum Avatar
    AA-Platinum

    Two stories from the New York (LGA) trenches:

    LGA-BOS (mid-July):

    Although I prefer American to Delta, I’m open to either…and in this case it was Delta (or whatever regional was operating that flight) that I selected. The previous day LGA (and, I assume, BOS) had both been shut down due to torrential rains.

    Boarding time came and went. The Captain finally came out to the desk and announced that we were waiting for a flight attendant (as this was an early morning flight, its not as if she was coming in on another flight). She finally showed up and boarding commenced. Just before pushback, the Captain came back and, quite apologetically, announced that we would have to burn off fuel for 30-45 minutes before takeoff. Apparently, Delta had fueled all of the plains the day before and, rather than giving us a plane that was scheduled for Boston had given us one that was scheduled for a much longer flight and thus we would have been overweight to land. This was a huge blackeye for operations (not sure if it was Delta or the regional) since they were at fault here. In short, we ended up landing more than 90 minutes late!

    Delta at BOS was a complete melt-down that day (I spent hours in Terminal 1).

    STL-LGA (early August)

    I flew out to STL on Delta but was scheduled to return on American. In both cases, the plane from LGA-STL turns around and returns to LGA (for those who might want to know, Delta’s flight was about two hours earlier on the outbound and an earlier on the return leg). Delta used a CRJ-900 while American was flying an ERJ-170

    Due to a ground stop (storms, again) at LGA, things were delayed. Delta cancelled the first leg…and thus the return leg too. American, however, pushed back a few minutes late and endured over a 90 minute delay on the ground. They got the plane to STL and after doing the fastest turn around I’ve ever seen, had the flight turned around and pushing back in about 25 minutes (the flight was full) thus putting us back in LGA some 45 minutes late.

    In short, when things go right, they are both fine. When things go wrong, American seems to be better operationally

    1. David C Avatar
      David C

      Long report. Informative.

      Try experiencing that American Airlines reliability in DFW and CLT then get back to us.

    2. Bill from DCA Avatar
      Bill from DCA

      NY to Boston? Take the train!!!

      1. AA-Platinum Avatar
        AA-Platinum

        I had a connecting flight afterwards (as I noted I had several hours to wait around Terminal 1)

        1. Bill from DCA Avatar
          Bill from DCA

          That makes sense but I gotta admit that I didn’t even begin to consider why you might have been in T1 for hours.

  9. CraigTPA Avatar
    CraigTPA

    I also wonder how much of JetBlue’s problem could be on the FLL end as well? I don’t have hard data (damn job…) but I hit the FAA website a few times across most days and SE Florida had weather problems in July too. And I saw BOS popping up quite a bit more than usual in July.

    This isn’t to make excuses for B6, it just reinforces that their limited network is unusually susceptable to getting hit very badly if just two or three cities don’t do well on top of their general issues.

  10. David C Avatar
    David C

    CF:
    Thanks for all of the work. Excellent report.

    What is your take on how this particular article lines up with airline profitability?
    I fervently believe that IRROPS handling goes a long way into eating profit or preserving those margins.

    1. Brett Avatar

      David C – I think it’s a contributor, but it’s not going to be a hugely strong direct correlation. Sure, over time, a better operation attracts customers and fares rise, but that’s indirect. You can’t look at one month to see that impact at all. But there’s no question there’s a real cost to running a poor operation as well.

  11. George Nathan Romey Avatar
    George Nathan Romey

    Airlines are constrained by weather and ATC flows. No amount of “compensation” is going to get ATC to release a flight before they are ready to release that flight. It’s a low IQ argument. Once the weather clears ATC can only accept so many flights at once and airport taxiways can hold only so many planes.

    If airlines were forced to pay compensation in the event of weather every airline would end in bankruptcy and the industry would be radically changed.

    A lot goes into how long a delay will be. Airlines do not like delays, delays are costly to the airline. The idea that there is some evil man in a room forcing delays is absurd.

    This being said, airlines having clear communicating between crew, the gate agent and operations and gate agents making timely and proper announcements is a different matter.

    1. David C Avatar
      David C

      But airlines operate in the sky/etc where weather also is present and I get tired of that excuse.
      The weather is as much a condition of their operating environment as oceans are to ships.
      I’m not talking about forcing a plane to land in a windshear situation.
      I’m talking about a schedule and operational philosophy that is not built around everything going perfectly, but one that has the ability to recover and accomodate.

  12. southbay flier Avatar
    southbay flier

    I’m surprised Alaska is that good. I feel like every time I fly them out of San Diego, they are delayed. And the weather is always good there.

    1. John G Avatar
      John G

      Alaska’s main hub is in Seattle, with smaller ones in Anchorage and San Diego etc. In the summer? Zero weather problems. They SHOULD be much better than others. But then again, American should have zero issues in Dallas in the summer. We are now in Day 47 since it last rained. It didn’t rain, at all, in August.

    2. Brett Avatar

      southbay – Actually, San Diego has real issues, especially earlier in the summer. The marine layer rolls in and can be below minimums pretty easily. You’d be surprised how much that can cause trouble.

      1. Patrick Avatar
        Patrick

        Only having one runway may be an issue also

      2. southbay flier Avatar
        southbay flier

        Brett – the odd thing is if I fly in in the morning it’s been fine. It’s leaving in the afternoon when it is sunny that’s been tricky. I don’t think that airport gets enough hate. SFO has horrible delays, but at least the terminal is pretty nice. SAN has an outdated terminal and lots of delays.

        It feels like it’s on the order of EWR, PHL, and the NYC airports in terms of being terrible.

        Parick – The lack of a second runway is definitely an issue as well. What do they do if they have to close that runway for construction?

        1. Brett Avatar

          southbay – SAN has an outdated terminal? When were you last there? The old T1 where Southwest used to operate is gone. I haven’t been yet, but I hear it’s pretty nice.

          1. southbay flier Avatar
            southbay flier

            Yesterday. T2 is an abomination.

            1. Brett Avatar

              Yikes, sounds like you should be flying Southwest ;)

            2. David M Avatar

              T2 is mostly nice. T2 East, where most of the AS flights are, is now the oldest and worst part of the airport now that old T1 is gone. Literally gone, as it was knocked down to build the second part of New T1, which will connect to T2. T2 East is probably the next part of the airport to get some attention once New T1 is done, but I’m not sure how far along the planning is on that. A lounge expansion for Alaska is in process as well (I’ve seen more details on that, but I’m not sure if construction has started); they hold the lease for the Aspire Lounge location but it continues to operate as Aspire; I believe they’re waiting for the expansion to be complete before turning it into an Alaska Lounge.

              And I do mean most AS flights, since I’ve seen AS flights not only in T2 West (the newer part, which includes the FIS where all the Mexico arrivals go), but even in New T1.

            3. southbay flier Avatar
              southbay flier

              Brett – I’d consider Southwest if they flew something other than 737s and that ain’t happening. At least Alaska is E175s. I will freely admit that I try to avoid 737s as much as possible. I really dislike them.

  13. Wany Avatar
    Wany

    It looks like Breeze has improved a lot lately. I have only flown with them in their early days and encountered multiple 6h+ delays. At least they message the customers in advance and tell them to come to the airport a few hours later. I think at that time they had limited aircrafts and crews doing lots of point to point flying. When a flight is delayed there is no back up crew or aircraft to substitute in. I have been avoiding them for some time but I think their current operation is good enough to give them another try.

  14. GS Avatar
    GS

    Great writeup Cranky, always cool when a comment inspires a post.

    I was curious to reframe the table as a derivative since Hawaiian has so few delays overall and here’s the stats:

    Of DELAYED FLIGHTS ONLY, Hawaiian still actually performs the best; only 7% of their delays are over 2hrs. And Alaska follows 2nd, with only 8% of their delays being over 2 hours. Breeze gets whacked here, with 23% of their delays being over 2hrs despite middle-of-the-pack performance overall, while Southwest as you noted performs notably better with just 9% of their delays being over 2hrs.

    The Hawaiian performance genuinely surprises me- you got a true outlier with the 14hr delay! Although to be fair other carriers might have better reaccomm options so an extra long delay wouldn’t actually get flown (And Delta games this anyways, posting 24hr delays to avoid filing a Cxl, but the flight ends up completely empty since everyone was rebooked).

    1. Brett Avatar

      GS – I think Hawaiian is helped by its usually-stellar interisland operation. That wasn’t doing well earlier this year, but they fixed it. And that helps to drown out some of those really long and terrible delays out there. But is that right?…

      I went and pulled just the A330 and A321 aircraft to see how those did. Looking at those, nearly 9% of delayed flights were delayed over 2 hours, so that’s definitely a higher number.

      1. David M Avatar

        I’m pretty sure I said that in the post about the Hawaiian delay as well, but your comments about the interisland operation to align with the conventional wisdom. The interisland operation is such a large percentage of the overall HA operation, that even when the long haul operation isn’t doing well, the overall stats still look good as long as it’s doing ok. It might be interesting to see that first graph with the interisland and overseas HA operations separated.

      2. EasyMoney Avatar
        EasyMoney

        You know, I think weighting delays by stage length or enplaned passenger count would be really interesting. Five interisland hops on-time shouldn’t count the same as five international delays

  15. Common Sense Avatar
    Common Sense

    Nice report. Probably worth comparing a few months to see how patterns change across seasons.

    Also a typo here: NYC Delay Performance by Marketing Airline – July 2016, should be 2026. :)

    1. Brett Avatar

      Common Sense – Or maybe it is 2016 and all the other dates are wrong? Just kidding, fixed. Thanks.

      1. Common Sense Avatar
        Common Sense

        Breeze got blown back to 2016? Would be good news for David Needleman if his airline was still around 10 years later.

  16. John C Avatar
    John C

    Cancellations would also be useful to compare along side the delays.

    My sister and her kids had a flight out of JFK to SFO in July that took rolling delays for six hours before finally cancelling.

    And since airlines have zero extra capacity, they were rebooked for four days later with a truly insane routing (JFK-ORD-BOS-SFO).

  17. Tim Dunn Avatar
    Tim Dunn

    Even though the US has an enormously accessible and massive air transportation system (which carries the equivalent of about every US resident 3 times in year), the US has chosen volume (quantity) over service quality – similar to other parts of US life.

    Flying through the sky is still a pretty novel concept for humans and yet even birds leave the air and hunker down in snowstorms and thunderstorms (the most disruption type of weather event for aviation).
    All of the extra staffing by the FAA or airlines doesn’t change that flight is subject to the whims of nature and no aircraft is going to operate in some types of weather, including the thunderstorms that are so hard to predict.

    The nature of airline systems makes it much harder to make sure everything recovers at identical times for all airlines even at the same airport. The DOT categorizes the reason for delays to include air carrier delays – which includes crew and maintenance – usually at about 7% of total flights which is the second highest reason behind late arriving aircraft which is close to 10%.

    The US could reduce the capacity of the US aviation system (which also includes general aviation which does use airports or airspace that overlaps with commercial airlines) to the average sustainable level of operations in some pre-defined level of bad weather – but there is always weather that is worse than what is modeled.

    and the bigger implication is that fares will rise as there will be fewer flights. The US has simply never been willing to pursue a broad policy of restricted capacity in the transportation system – air or surface – in order to result in fewer delays during peak or weather-impacted periods of time. The US builds highways to create capacity but there are limits of what can be built and delays occur as evidenced by rush hour in just about every city of any size.

    The data CF presents is good to know and evidence that there are differences in how difference airlines operating from the same airport handle delays but the notion that it is possible to eliminate the inconvenience of delays and cancellations esp. during severe weather is likely just not attainable.

    1. John G Avatar
      John G

      Taking away the Delta love affair from Tim, he knows what he’s talking about here. We as a society choose to have a system where we choose to keep the cost down, and trade off efficient operation during bad weather or other IRROPS. We could choose to bring the capacity down to a level where the delays are much more manageable, but in doing so we would price some of us out of being able to us the system at all.

      It’s a balance. I think we actually have struck the right balance here.

      And I say that 48 hours after a very fun experience at LaGuardia with storms canceling everything. 23 hours watching planes circle and cancel is just my favorite way to spend a day.

      That is the price we are going to have to pay for price and convenience. Sometimes, it breaks down a little. Gotta roll with it.

    2. 1990 Avatar
      1990

      Tim and John,

      Nobody expects airlines to control the weather. Under EU rules, severe weather is already a free pass—no one is getting cash compensation for a blizzard. The real issue is that U.S. airlines schedule flights right on the razor’s edge to squeeze out maximum utilization, leaving zero cushion for when things go sideways. When a storm hits, a fragile system turns into a three-day mess where passengers eat 100% of the cost.

      Peer-reviewed empirical studies and data show that when airlines actually face financial penalties for controllable delays, they fix their operations and severe delays drop by up to 70% (See Transport Policy, e.g., Gnutzmann & ?piewanowski). And before anyone says it destroys cheap fares, look at Ryanair: they live under these strict rules, pay union labor, and still pump out low fares. Plus, US carriers already comply with these rules every single night flying back from Europe—they have the capability, they just don’t ‘have to’ use it at home.

      We don’t have to just accept a broken system as the ‘price of doing business.’ You fly, too. We all deserve better.

      1. Tim Dunn Avatar
        Tim Dunn

        Nobody is accepting anything but you don’t seem to want to accept that there are structual limitations to just about everything including transportation systems.

        and you are also simply wrong that US airline customers are left holding the bag for any IROP that they incur. US airlines commit to and are providing significant customer benefit when IROPS occur.

        and you simply cannot understand that no company is going to provide a service that they will lose money on. Paying a customer a penalty that is as large or larger than the fare simply means that airlines will not offer as much service at lower fare levels esp. in periods that are delay prone.

        you also cannot grasp that Ryanair operates the majority of its flights at secondary or tertiary airports. If the big 3 didn’t have to bother with NYC, their operational reliability would be much higher. Charleston, Des Moines and Boise are much easier airports for IROP recovery.

        and, no, Ryanair does not pay labor costs anywhere close to the big 4 in the US and they price their product and offer a level of service that is low enough to pull people away from primary airports. ULCCs in Europe have far less access to the big Euro hubs than they do in the US.

        John G gets it. There is a balance and the US keeps capacity at near maximum levels for ideal weather conditions which allows more people access to the air transportation system INCLUDING at PRIMARY AIRPORTS IN THE US. without that capacity, some people will not be able to fly at all. but when that capacity is removed due to weather, it is taken equally away from all scheduled flights and the entire system contracts.

        tell me. Who do you expect to pay the bill when your commute on the highway is double or triple the “usual” time because of an accident or highway construction? that happens all the time on the ground in the US and affects far more people than are impacted by air delays.

        1. 1990 Avatar
          1990

          Tim, there is no “balance” here whatsoever.

          The scale is tilted entirely in favor of the airlines, not the passengers. You know full-well that carriers are running razor-thin schedules to maximize utilization, externalizing risk onto travelers, then handing out $12 expiring meal vouchers when everything inevitably implodes. That might work great for airline margins, but it fails consumers, and they start to notice after a while… Oh, and do spare us the flawed highway analogy and hub-and-spoke excuses. Public infrastructure and self-imposed capacity jams don’t give a private carrier a blank check to walk away from its contracts.

          EU and UK 261 already treat severe weather as force majeure. The real issue is accountability for controllable meltdowns (Crowdstrike! *gasp*). US carriers comply with strict consumer protections every single day flying out of Europe because they already have the capability—they just refuse to do it at home (because they don’t have to… for now).

          If anyone still thinks statutory rights are an unworkable pipedream, look again at Europe. When the EU passed Regulation 261 back in 2004—and aggressively modernized and strengthened it recently—the airline lobby screamed about mass bankruptcies and soaring fares. Instead, the people of the EU called their bluff, tying consumer accountability to single-market access and telling carriers that treating passengers with dignity is the cost of doing business. The sky didn’t fall. Phew!

          Back in the US, pretending passengers are happy to be stranded for days as the “price of admission” ignores the truth. Our airlines pocket the upside, and passengers absorb nearly all of the risk and cost. Deep down, you know what’s coming. Likely, another economic downturn, predictable cries for federal bailouts, and when that bill comes due, statutory passenger protections have to be the stipulation. Lobbyists can’t hold this off forever. Such rules aren’t perfect, but they’re far better than nothing. And, ultimately, they help airlines to operate more reliably, which is good for business in the long-term (not just a quarterly report). We deserve better.

          1. Tim Dunn Avatar
            Tim Dunn

            you didn’t answer the question about who pays for your (or anyone else’s) delays when commuting by road. It doesn’t happen and you know it never would.

            You see a for-profit company and think that they should pay for inconvenience that would never be compensated in the public sector.

            The simple reality is that NYC – the worst city in the US for air service delays – functions quite well on alot of days.

            Do you remember how good the weather was on 9/11? Those days will come again this year as summer moves on. and then winter will come, as it always does.

            as for your “economic downturn or black swan or whatever is coming and so will be bailouts” I didn’t want bailouts during covid and yet they happened.

            You simply unrealistically want the best of all worlds and want it all for you. Real life proves that simply does not happen because “the people” that you think you advocate for have figured out that what is in your interest is not at all good for them.

            The US airline system far from perfect but it is the most accessible and most reliable for anything even remotely close to its size (which isn’t much because there is no real competition).

            btw, you do realize that even the rails in Europe are struggling to be reliable this summer?

            1. 1990 Avatar
              1990

              Tim, I presumed some of your questions were merely rhetorical. Bah!

              *sigh* Must you so quickly traffic in false equivalences on here? Ok, so, if a private charter bus company or a railroad stranded you for three days due to their own mechanical failure or poor crew scheduling, you would just shrug and call it “public sector reality.” No way. You’d demand clear communication, faster alternatives, reimbursements, etc. (duty of care, compensation, etc.)

              Yet again hiding behind weather to excuse controllable operational meltdowns… it doesn’t wash, sir. You know that US carriers manage to comply with strict passenger rights every single day flying out of Europe because they already have the capability—they just refuse to do it at home where our regulators (or current lack thereof) give them a free pass.

              As for bailouts (I see I got your attention), whether you wanted them during COVID or not, recall that billions were handed over without structural consumer strings attached. That history is why the next time corporate balance sheets implode and management comes cap-in-hand to Washington, statutory accountability has to be the price of admission. Lobbyists can delay it, but they can’t outrun gravity forever.

              I’ve already acknowledged here and before that such rules aren’t a silver bullet, but they do very much force airlines to build actual operational resilience instead of treating passenger inconvenience as an acceptable cost center. We deserve baseline accountability here, too.

            2. Tim Dunn Avatar
              Tim Dunn

              the only false equivalencies are yours.
              A mechanical delay or privately operated bus that is not available because of the operator’s unwillingness to have sufficient equipment to operate their schedule is not at all what we are talking about.

              Every single US airline has many days per year when they run 85% plus on-time and less than 1% cancellation rates.
              But those days happen when there is virtually no bad weather anywhere in their major markets and esp. not in NYC.

              You love to tell the rest of us how we should pay the price for NY/NJ’s inability to come up an airport system that is not susceptible to bad weather; when you have 3 airports w/ intersecting flight paths, intersecting runways and runways that are too close together, disruptions happen.

              Somehow the people of Atlanta, N. Texas and Denver managed to build airports that can operate in bad weather with minimal reductions due to the airfield structure even if those places -just ike every other place else on earth – cannot stop bad weather from coming their way. but they recover from bad weather far faster than NYC.
              and operations return pretty quickly when bad weather ends at those airports.

              and you have acknowledged that major weather disruptions are exempt from some of the very sets of rules you propose – so you haven’t proposed anything other than raise the cost of airfare for alot of people at the expense of “the commons” that will be priced out of air transportation when US airlines are forced to take on costs that don’t really change the root problem which is simply that weather is disruptive no matter what it touches.

              You just might need to call Topeka home. As long as there isn’t bad weather over the field, they do a heck of a job of getting airplanes in and out.

              enough of this with you….

            3. 1990 Avatar
              1990

              Tim, blaming NYC airspace doesn’t absolve the airlines of their controllable failures. Yes, the NY airspace is congested (I live that reality every time I fly out of here, and back home)—which is why carriers shouldn’t jam tight, fragile schedules into JFK, LGA, and EWR if they can’t absorb a simple shift in the winds without melting down for 40 hours.

              And I see you hiding behind “the commons” (since when do people even talk that way?) and fear-mongering over higher fares. Wild that European airlines operate under strict statutory accountability every single day, yet flying remains wildly accessible and affordable across the EU. Huh. The only thing statutory rules price out isn’t the average traveler—it’s airline management’s free pass to externalize the cost of their own operational shortcuts onto passengers.

              If holding carriers accountable for controllable meltdowns means they have to build actual operational resilience instead of pocketing the upside, that’s a trade-off every rational passenger will take. (If this has been enough for you, that’s fine, but I’m just getting started.)

  18. Angetenar Avatar
    Angetenar

    One thing that I see a lot online is that people complain about some AAirlines posting unrealistic expected departure times in the case of delays, leading to rolling delays. Can your data quantify how much this happens with each airline?

  19. Alex Hill Avatar
    Alex Hill

    Nice report.

    It’s striking that the big three airlines with intercontinental operations have around 5-6% of flights delayed 2+ hours, while the narrowbody-heavy (until-recently or still narrowbody-only) big airlines (Southwest and Alaska) have just 2-3%. (Excluding the ultra discount airlines and JetBlue.)

    Is that mostly long haul flights? I’d imagine airlines will be much quicker to cancel one of ten SEA-LAX or MDW-BWI flights than they are to cancel their single daily flight to BCN or SYD. And I too would much rather an airline delay my intercontinental flight three hours than cancel it, whereas on short-haul flights with other options cancelling and putting me on another flight becomes the preferred option much more quickly.

  20. Jason Avatar
    Jason

    WN after it’s historical Xmas IT meltdown Adam and Andrew both successfully merged NOC real Time operations with network planning schedules building. Which resulted in WN having its best OTP and NPR scores EVER in their history since their self proclaimed Triple Crown award.
    Before Covid and their prehistorical Meltdown the GK Era Mike Vander Vin lead “Shave Turn Times” plan sank their OTP flight to the worst in WN entire history.
    Flight were constantly delayed with multiple Snowballs mini meltdown.
    Adam and Andrew revamped the system built in many safe guards eliminated 90% of thru flight increased turn times to reduce delays at Mega airport allowing to isolate major events.
    This elasticity allowed WN network much needed wiggle room and increased efficiency.
    Well Soon as Adam retired a new management shake up happened and Someone Stuck in WN 10 min turn time Folklore nostalgia of the past decided to bring back a Failed plan of “LETS SHAVE TURN TIMES “ and Connection times again plus bring back Thru flights.
    What has it quickly Resulted in and Restored ?.If you answered DELAYS DELAYS DELAYS your a winner.
    I listen to Andrew Watterson Podcast. All the Technology innovation Can’t Give back the time you shaved off the actual Turn time.
    Especially during the middle of a complete Evolution to seat assignments and charging for luggage.
    As much as the golden palace in Dallas doesn’t want to hear it but 45 minute turn times was their Operational Sweet spot.
    Flight often left 5 minutes early and arrived 10 minutes early.
    Now it’s all gone.
    The biggest mind blowing thing to fix morning originating Delays and increase more Crew change flows they finally adjusted the Flight Attendant Gate report times along with Eliminating how many times a crew pair goes through a Domicile eliminating unplanned crew swap by 90% . All that would have further double downed their overall efficiency.
    But as soon as that Switch was made then someone decided let’s Shave Turn Times basically Throwing away all the gains.
    I guess all those millions dollars Ramp display Fids just Remind everyone hey your flights delayed again.

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.