A Closer Look at American’s Transatlantic Operation at JFK

American, JetBlue

The Northeast Alliance (NEA) between American and JetBlue was supposed to do a lot to solve American’s troubles in the New York City area as a distant competitor behind Delta and United. Part of this plan was bolstering the domestic network, but one of the more visible pieces was the rapid expansion of American’s transatlantic flying.

Back in 2019, American flew from New York/JFK to Barcelona, London/Heathrow, Madrid, Milan/Malpensa, and Paris/CDG with Rome in the summer only. London was the mainstay of the operation, primarily because of the joint venture with British Airways. The same could be said for Barcelona and Madrid since Iberia is part of the joint venture as well. Milan and Paris are, well, they’re just key markets that American apparently needed to serve. And Rome? Well, Rome in the summer is just an easy one to fill.

With the NEA, American expanded. It added Delhi, Doha, and Tel Aviv as year-round flights. It also threw in Athens as a summer-only operation. Though there was a lot more happening in the domestic world, these flights were the headliners that were meant to show the world that the NEA was good for competition.

In reality, JFK has been a terrible struggle for American for ages. It was hopeful that the partnership with JetBlue would not only make these new routes viable but also help improve performance on existing routes.

Summer is the easiest time to make money in these markets, so I looked at the winter. Specifically, I looked at ARC/BSP data from Cirium for January through March 2023 to see how the NEA had come together in these markets.

During that period, it looks like just over 6 percent of people flying on American over the Atlantic had at least one connecting segment on JetBlue. That may sound like a small number, but it’s good for around 250 people a day.

There is pretty wide variance between the cities here. Take a look.

Traffic Breakdown on American JFK Transatlantic Flights – January to March 2023

Data via Cirium

I left out Athens and Rome, obviously, since we are looking at winter when they mostly don’t operate. It is unsurprising to see JetBlue’s contribution be smallest in London/Heathrow where it’s all about British Airways. Still, 4 percent of those travelers touch JetBlue which is more than I would have guessed, especially since JetBlue operates its own London flights.

At the other end, we have Barcelona which had a whopping 10 percent of travelers touching JetBlue. Doha, of course, stands out in that there is almost no local traffic. That’s because the bulk of this traffic (nearly 60 percent) is connecting on to Qatar Airways beyond Doha. That means it’s a New York-centric market, as much as Paris is. As for Paris, it may be largely local because it’s much easier to connect there via other American hubs.

Regardless of all the individual details, you can see that JetBlue is not insignificant here. Let’s put that into even more perspective. The Doha flight is the clear load factor winner with over 88 percent of seats filled. Tel Aviv followed at almost 84 percent. Nothing else was above 80 percent. What’s that tell us? It means there are seats to fill. And without that JetBlue feed, it gets harder to fill those seats.

I think it’s helpful to put some of this in context here versus American and United.

% Local Originating Traffic on Competitive NYC Transatlantic Markets – January to March 2023

Data via Cirium

American, by nature, has to take more local traffic in New York. It doesn’t have the connecting hub that Delta has at JFK/LaGuardia or United has in Newark. That’s why in the majority of these markets, you see American with a higher percentage of traffic that starts or ends in New York.

London is a clear outlier here, and that’s really because so much traffic is connecting beyond London on to partner British Airways. The BA partnership is what really enables American to be in that market, though with Q1 loads under 60 percent, it’s hard to even call that a success. It’s more like a strategic necessity.

You’ll notice Delta’s unusually strong position in Milan as well. I wonder if that may be due to corporate contracts, but I don’t know. And Tel Aviv is another outlier. But overall, you can see the pattern. American is heavily reliant on the local New York traveler, but the flights really benefit from that JetBlue feed.

With that feed presumably going away, it makes it even harder for American to fill those seats. That means more discounting and worse results. Is it worth it for American to keep flying them? That’s going to be a strategic decision, not one based on actual economic performance.

Of course, London doesn’t go anywhere, but all the rest are up for grabs. If American decides it needs to be a player in New York, then it will need to keep some of these, like Paris. But others can go away and not have that same negative impact on an overall New York presence.

There’s a lot more to be said on this topic over the coming months. American really has to figure out if it wants to stick with a New York strategy or move on. Maybe it will defer its decision until after the NEA appeal is heard, but I would not be surprised at all to see some of these long-haul flights get pulled down as the already meager connecting feed dwindles.

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 “A Closer Look at American’s Transatlantic Operation at JFK

  1. AA’s lack of follow through at JFK has been evident in their Terminal 9.
    It is a very large, fairly new Terminal which they built in the early 2000’s,
    which for years was incredibly underutilized. A white elephant if you will.
    And despite being better utilized today, the majority of the increase came
    from foreign flag carriers moving in and not AA operations. They are very
    Lucky to now have BA in the Terminal to help pay the bills.

    1. Totally correct except you’re referring to T8. T9 was destroyed when the new T8 was built. At this rate, T8 will be referred to as the British Airways terminal instead of the American terminal.

  2. “Of course, London doesn’t go anywhere, but all the rest are up for grabs”

    Why wouldn’t Barcelona and Madrid fall into the same bucket as London given the JV with IB?

    1. Eric – They just aren’t huge or hugely important markets. Iberia is already in those with 2x daily to Madrid and 1x daily to Barcelona (on LEVEL). Maybe there’s a rationale to stay in them, but they don’t have the same core level of importance as London.

  3. I know the benefits are much less than the NEA but why not code share with B6 to keep asses in their xatl seats to make those flights more viable? It’s better than nothing.

    1. Exactly. And that is what AA and B6 had in the mid-2000s. That should return but with added benefits and a more seamless experience that came with the NEA left intact.

      1. I guess a code share between AA & B6 similar to what AA has with AS is the only option if the NEA is unwound. Notice how AA needs a partner to fly north/ south routes on both coasts while UA & DL don’t.

    2. Bill – There are plenty of options out there, potentially. The judge indicated an Alaska-style alliance that even shared revenue on long-haul would be feasible. But JetBlue isn’t likely to do anything before it gets the Spirit merger sorted out. And even then, it remains to be seen if it’s an airline that wants to keep playing ball or not with American. Maybe American will wait around to find out. Maybe it won’t.

      1. Cranky,

        I guess even if it doesn’t make financial sense for B6 at first, working with AA still may make business sense in the long run despite the NEA.

  4. NYC is one of a few US cities in witch both Domestic & foreign flag carriers fly the same route. L A, SF, Chicago, Boston & Miami are the others that come to mind.

  5. Great analysis as usual. The other factor to consider, I think, is the role of JFK in AA’s system. Even DEL/TLV may not be as high as CDG for local passengers, JFK would still be preferred option in AA’s system to place the long haul. In addition to New York having higher O&D than PHL/DFW for DEL/TLV, the current Russian airspace restriction essentially means JFK is among the only few viable options for India.

  6. BCN, CDG, MXP were all there long before the NEA. CDG (originally it was ORY, then moved to CDG in 1999) has been flown continuously since it launched in 1987, with the exception of 2020 due to the pandemic.

    That said, there is no question AA struggles at JFK on its own, and it has for 25+ years. The NEA will possibly (and quite likely) be reincarnated as a code share, similar to, but with more benefits to B6 and AA customers, what the two airlines had in the mid-2000s at JFK.

    DOH is, like LHR, a connector to points beyond DOH. AA doesn’t need to be in the market, and could easily have QR do the flying, but for now, I don’t see the route going anywhere. FCO is currently running 2 x daily this month and next. It is a gold mine.

    Having recently flown JFK-BCN-JFK on AA, the flight is full and often oversold. Am a regular on JFK-CDG too. It always goes out full and has for years. There are contracts on this flight, as well as MXP. DL in no way owns MXP. It simply added a second seasonal nonstop in 2022 and kept it this year. Whatever corporate traffic DL gets on JFK-MXP-JFK is attributable to AZ being in SkyTeam and that’s likely about to end with AZ going over to LH.

    1. Regarding DOH, it’s not so easy asking QR to pick up the flight. QR has an even bigger issue aircraft availability than AA. In fact, this is the reason why QR is having Finnair fly a lot of international routes from Europe to DOH. Finnair has slack in their fleet because a lot of their routes to east Asia have been impacted by the Russian airspace issue.

  7. TLV, DEL, and probably connections beyond DOH are markets dominated by travel agencies. AA’s recent retreat from this type of traffic is going to only make these routes worse and unsustainable.

    1. I could be wrong on this, but I’ve noticed far more residence in my area & other areas around NYC from India & I wonder if the JFK – DEL AA flight has enough traffic realizing it’s not the only flight on the route.

      As for TLV, There’s always demand from NYC as more Jews live around there than any other north American city & four airlines fly the route.

      1. Air India services JFK-DEL,  and with multiple A380’s a day, Emirates skips DEL with service DUB to multiple cities in India. Sent from Samsung tablet

    1. Emac – I think it’s pretty well known! United flagged London specifically on its last earnings call. It’s just a weak market but with slots, no airlines can pull down flying without losing their place in the future. So, the garbage fire rolls on…

      1. Sorry for the ignorance, Cranky, but do you mean routes like EWR-LHR aren’t earning at present? I’m flying home regularly and every (every) LHR flight on UA has been oversold for months and months. I realize bums on seats still doesn’t mean it is profitable, but how on earth can the airlines be losing money when the fares have been eye-waveringly high for at least 18 months?

        1. yes, the 763 in UA’s high J configuration has higher seat costs than the 772ER in AA and UA’s “standard” configuration but the 777-200ER and 767-300ER actually have pretty similar operating costs according to DOT data when the 763 is in a 215-225 seat configuration which is the way DL operates it and AA used to operate the 763. The 772 has higher fuel burn per seat but spreads pilot and other fixed costs over more passengers.

          In this discussion, we need to keep in mind there are some historic and geographic/market realities that impact AA’s transatlantic strengths and weaknesses. AA and UA got into LHR fairly early – ahead of DL and other legacies at the the time – and both still have more slots on their own metal than DL. UA also developed a relationship with LH – which morphed into the LH Group – giving them strength in both the UK and continental Europe. Iberia became part of the AA joint venture but Spain is a much smaller market than the UK, France and Germany and those 3 markets are the “backbones” of the 3 alliances and JVs. DL’s European presence was built from the Pan Am continental Europe system, then the Swissair Group, and then Air France which later merged w/ KLM. DL’s partnership with Virgin Atlantic helps round out its historically weaker position in the UK.

          DL and UA have more strength in more of the biggest markets in Europe and also use their partnerships and JVs to strengthen where they are weaker. oneworld is heavily UK focused which gives AA less ability to offset its weaker presence in continental Europe. Southern Europe has much bigger seasonal demand swings which translates into seasonal average fare differences. Market strength translates into the ability to sustain off-peak periods, something DL has developed over decades while Germany is well-suited for covering Eastern Europe and beyond.

          Any carrier’s ability to command revenue premiums and serve markets in lower demand periods of the world is driven by their strengths. While feed at JFK is part of the equation, the other part is market strength in Europe.

          AA simply does not have the market strength in continental Europe that DL and UA have and there aren’t any likely developments that will change that reality which means that AA’s ability to fix its NYC situation is hindered as much by its market position in Europe as fixing its feed problem in NYC.

        2. Bobber – Those London flights are not full. Maybe they’ll do better for the summer, but there have been a lot of empty seats for the first quarter of the year where I currently have data. United was at 64%.

          1. Cheers, Brett – stats are stats, I guess. Because EWR is such a shitshow, I routinely check availability on all EWR-LHR flights on the days I travel in case I miss connections from DTW – they never show any available seats in Y.

  8. Always helpful to see data like this.
    What you don’t include but you presumably can access even if you can’t post it on a public website is that AA underperforms DL and UA in most NYC local transatlantic market average fares which is driven in part because they have to take more local traffic on 777-200ERs which are roughly the same size as DL and UA aircraft.
    and AA is at a profit disadvantage because they don’t use the 787 which is their most efficient aircraft at JFK while DL and UA each have new generation aircraft which they can and do use.

    Just a reminder that AA has very few aircraft on order relative to DL UA and WN so it will be much harder for them to commit to a large rebuilding of their NYC operation – and doing so will involve much more financial risk than from growing CLT and DFW where another aircraft can easily generate positive financial returns.

    It is also worth noting that the value of LGA and JFK slots has gone up over the last few weeks as prolonged and repeated rounds of bad weather have led to multiple days of extensive ATC delays for all of the NE. A high percentage of regional jet flights from all 3 airports have been cancelled, leaving mainline aircraft to operate a very reduced schedule. Add in that AA and DL have the majority of their NYC operations spread out over 2 airports while UA is highly dependent on just EWR and the value of LGA and JFK being slot controlled is higher.

    1. For UA, EWR-LHR is almost exclusively on 763s (with maybe one 772 per day) – although they probably own these planes outright, neither are anywhere as efficient as the 787s.

  9. Where do you see the very limited JFK-Deep South America operation (EZE & GRU year-round, GIG seasonal) going? At least EZE & GRU are money-makers for AA.

    1. DCA – South America is a key point of strength for American overall, so it has a better shot of keeping those flying. The Sao Paulo flying is the most important business market in South America, and presumably that won’t change. Buenos Aires, I’m not as sure about and Rio in winter is like Rome in summer, not a bad place to put an airplane. (Ok, it’s not anywhere near as good as Rome, but the idea is the same.)

      1. It’s Sao Paulo with a u, not Paolo with an o. Paolo is an Italian spelling, and the city of Sao Paulo and country of Brazil in which Sao Paulo is located is Portuguese speaking.

        AA and its predecessors have flown NYC to Buenos Aires for decades. It’s a long term successful route.

      2. Believe it or not, UA used to fly JFK-GRU nonstop back in the early 2000’s using a DC-10. I know because I flew that route. Hard to imagine now, but this was before the merger with CO gave UA a fortress hub across the Hudson.

        1. stogieguy – According to Cirium that last flew in Dec 2001, so it was a 9/11 casualty. But the data shows it being on a 767-300ER, not a DC-10.

          1. United use to fly nonstop from JFK to Sao Paulo (continuing to Rio de Janeiro) And Buenos Aires, for a good part of the ’90s right up to 2001/ 9/11. I lived in Buenos Aires for a year from 1999-2001 and took the JFK-Buenos Aires flight several times, right up until July 2001. It was technically a through flight – Buenos Aires – New York JFK – London Heathrow. Same flight number all the way through. There were always people on the Buenos Aires – JFK flight connecting to UA’s flight to Tokyo and, in 2001, to Hong Kong, when UA briefly flew that flight.

            Those flights were always on 767-300s. If they were ever on a DC-10 it was before I started flying them in 1999.

            I think at a certain point United even flew from JFK to Caracas, but my memory is a little hazy on that point.

            1. I recall seeing UA JFK-Caracas on a Hemispheres route map from the 90s; don’t recall if it was future service that never started or if it actually flew.

        2. Sorry.
          UA flew 767-300’s JFK to BUE, EZE and SAO in the 1990’s and early 2000’s after buying PA’s South American Operation out of Chap 7 in the early 1990’s. JFK-BUE had a tag at the end EZE-MVD. There wera also Latin American flights from PA on UA that operated from MIA.
          .
          The DC10’s were long gone to the scrap yard and FED EX by the 2000’s.
          United,as the launch customer of the 777, started accepting the new 777’s in 1994 and 1995 which was about the same time the DC-10’s were being phased out.
          Quite of few of UA’s DC-10’s were sold to Fed Ex and converted to freighters with 2 pilot cockpits. I think some referred to them as DC11’s after the conversion

          You may be thinking of UA flying JFK-CCS with a DC10 for a very short time in the early 1990’s.
          CCS flights ceased as Venezuela fell deeper into political and economic turmoil in the 1990’s with the rise of Chavez, followed by the Nationalization of many private assets in Venezuela, leading to money losing flights.

          In the mid 1990’s JFK-EZE was one of United’s most profitable routes. But as political and economic decisions were made in Argentina, especially the delinking of their currency from the US dollar, the flights quickly made a U turn and began to become unprofitable as the Argentinian economy spiraled out of control. At the same time AA was plowing as much as possible into MIA to grow Latin America as Varig, VIASA, Aerolineas Argentinas either went out of business or pulled back dramatically, especially on long haul flying.

          After 9/11, profitability of UA’s JFK – EZE, BUE and SAO grew steadily weaker. As a last ditch effort UA moved the flight to their IAD HUB hoping to strengthen them with IAD feed traffic. But the South American economies, especially EZE, never cooperated, and the flights were discontinued out of IAD.

          Then AA, although faced with the same challenges as UA, had reached critical mass in Latin America and the field had been cleared of many competitors .

          1. Looking at T-100, the last United DC-10 flight was in Feb 2001 flying Chicago to both Honolulu and Las Vegas that month.

            Here’s some fun from JFK though, T-100 goes back to 1990.
            *Caracas was a 757 from 1992 until Sep 1998 when it went 767-200. Then in May 1999 is went A320 until discontinued in Dec 1999.
            *Buenos Aires was a 767-300ER from 1993 until Oct 1997 when it went to a 777-200. It stayed there until May 1999 when it flipped back to a 767. It was canceled in Jun 2022.
            *Rio was a — wait for it — 747SP (!) from 1992 until Apr 1993. It was then a 767-300ER until Sep 1993 when it was canceled.
            *Sao Paulo was a 767-300ER from 1993 all the way until it canceled in Dec 2001.

            The only other Latin flights from JFK were Port of Spain and San Juan.
            Port of Spain flew Nov 1992 through Jan 1994, mostly on a 757 but there were 10 trips in Dec 1993 on the DC-10. San Juan was a 757 from Dec 1997 until Feb 2003 when it flipped to an A320 until Aug 2003 when it canceled.

    2. AA has outlasted everyone that has tried (and largely failed) on NYC Area to Buenos Aires. AR, in spite of its Sky Team membership and overall service improvements still operates an atrocious schedule and does not generate huge point of sale on the NY end. AA has strong sales teams in Latin America overall and balances JFK-EZE sales with strong cargo and passenger sales on both ends. It also seems to navigate Argentina’s strong capital controls and limits to currency expropriation better than most. UA could not make EWR-EZE work and succumbed to Argentina’s unrelenting boom and bust economic cycles, but it performs well on IAH-EZE.

      Delta’s partnership with LATAM will do little to help DL on JFK-EZE and AA has already responded by upping the frequency for several weeks. My sense is DL will be run off JFK-EZE. It is a route it doesn’t need to fly.

      GRU is strong performer for AA and will continue to be. GIG is a seasonal, leisure route, but it works, given the limited options from NYC.

  10. Also take a look at the surprisingly high number of sales on JFK-TLV via JetBlue.com. Another interesting angle to the NEA.

    1. Wonder if as the NEA is unwound, JetBlue would consider a joint venture with El Al as El Al is unaligned & would allow the feed to continue at JFK outside of American.

  11. I’m not sure why you lumped LGA in as a connecting hub for Delta. There’s no Delta transatlantic service out of LGA and how many travelers are going to fly to LGA, exit the airport, and schlep their luggage over to JFK? And does that really count as a connecting flight?

    Also, if Doha is performing well, why wouldn’t American keep that route alive given the low dependency on connecting traffic?

    1. Delta has a split hub operation in NYC. Not the most efficient way to do things, but they make it work despite that. Now if there was a landline style bus between JFK & LGA where there’s no need to exit & reenter through security, that may improve things. But of course you cant escape the highway traffic on the GCP or I-678.

      1. Cranky – what percentage of people actually do this? Is it really not trivial? I would be very surprised if it’s anything worth even considering as a factor in this type of analysis. (Excellent write-up, by the way – as per usual!)

        I am a lifelong New York City-area flyer, and I must say I believe I have observed this happening maybe twice? three times? in my entire life (from hearing people complain at check-in – both times they did not realize they had to switch airports on their itinerary). In fact, the only times I have observed it happening “regularly” is for crews, not passengers (again, from hearing crews speaking about it on the JFK AirTrain or the Q70 “LaGuardia Link” bus).

        My question is simply this: do we have data anywhere that shows how many people/what percentage of flyers actually travel between JFK and LGA? Because I feel that calling it a “split hub” may really be overstating the relationship between JFK and LGA by an order of magnitude.

        Again – great post. Why I read the Cranky Flier!

    2. There’s plenty of domestic to domestic connections at LGA on Delta. I’ve flown DCA-LGA several times and many people get off and are rushing for connections to varied cities such as DFW (surprisingly) and points in New England and elsewhere.
      My parents are from Louisville, KY, and would travel to Boston frequently. Before DL and AA starting Louisville-Boston nonstop, they’d frequently fly from Louisville to LGA and connect to Boston. There are tons of domestic connections at LGA on Delta.

  12. 1) Why specifically do people think London is a “dumpster fire”? I am guessing it is just too much high end, corporate capacity in a time with less business travel structurally.

    2) Spain (BCN/MAD) is the most important markers for transatlantic leisure traffic in the country – I wouldn’t call them “unimportant”. Any major US airline needs to fly to these markets from their international gateways to capture that high end leisure travel

    3) Where does credit card revenue fit here? AA pulling down in JFK, cutting leisure flights like JFK to BCN, will reduce the appeal of AA cards. That is the big deal

    1. Anthony – Credit card revenue is the great black box. We don’t know how much revenue comes from the New York area, nor do we know how much would be a risk if American had a smaller presence. It could be significant or it could be minor.

  13. Why do so many people feel American has to have a major presence in New York? Southwest has a minimal presence in the area, yet it’s been the most consistently profitable airline over the last 30 + years.

    The world’s only perfect airline, which has a major presence at both LaGuardia and JFK, had to file for Chapter 11 bankruptcy protection. Meanwhile, li’l ole Southwest, with its measly market share at LaGuardia, just kept reporting profit after profit after profit.

    Maybe Southwest is onto something?

    1. I would say something akin to international prestige. As in you can’t be the world’s largest airline and not have some sort of meaningful footprint in your home country’s largest city.

  14. I think a lot of this is due to the inability of AA to use the correct gauge at JFK as compared to DL and UA. The later have the ability to use anything from a 757 up to a 773 for UA and a 359 for DL while AA is stuck with the 777 because they can’t make the economics of a multi plane pilot base work at JFK due to its size. Maybe if they move the Dreamliner back to JFK from PHL, especially the 788, they could make things work without the NEA, but I am sure there are some routes which warrant the 777 due to cargo and J contract guarantees.

    As Delta shows us, the 763 is the perfect airplane for most of their JFK operations.

    1. Delta’s use of the 763 for most JFK long haul intercontinental flying illustrates the limits of its profitability (and the challenges). Huge market, lots of fragment. The DL 763 isn’t a competitive product overall.

      1. 1. The A330 in its various versions, is more commonly used on DL’s transatlantic system, including from JFK, than the 767-300ER
        2. UA’s 757s don’t have any better premium product than DL’s 767-300ERs. DL, like AA, does not use the 757 on longhaul (over 8 hours) transatlantic service.
        3. DL and UA have similarly sized 767 transatlantic fleets since about a dozen of DL’s 767-300ERs are used for transcon service.
        4. DL has suites with doors in Delta One on its entire A330-900 and its original delivery A350s while neither AA or UA have any suites w/ doors.
        5. DL execs just said they will reduce the 767-300ER fleet in the next year, likely using their A330-900 and A321NEO deliveries including to reduce the number of 767-300ERs on transcon routes.
        Airlines determine if a product is competitive if it delivers revenue comparable to other airlines in the same or similar markets. DL uses its 767-300ERs as UA uses its 757s – in secondary markets esp. where there is not strong or high paying premium cabin demand. The 767 can carry cargo while the 757 essentially cannot.

        Delta has consistently operated the most profitable transatlantic network and that is true based on the most recent data.

        There is no US airline that “needs” to operate any route (comment above). They do so based on their expectation of a profit and they operate at a loss if that does not happen.

        AA, DL and UA and their most recent merger partners, have all reorganized under chapter 11. It is irrelevant to any current issues.

        AA should be focused on profitability above all, including what it gets from credit card revenue but they likely cannot overcome operational losses and also have yet to prove they can deliver revenues that DL and UA can from NYC.

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