Airways Magazine

Delta to Trim Narita hub, to Axe New York, Bangkok, Osaka, Los Angeles and (Possibly) Minneapolis

 Breaking News

Delta to Trim Narita hub, to Axe New York, Bangkok, Osaka, Los Angeles and (Possibly) Minneapolis

Delta to Trim Narita hub, to Axe New York, Bangkok, Osaka, Los Angeles and (Possibly) Minneapolis
August 10
22:48 2016

MIAMI — Delta Air Lines (DL) is once again removing transit links between the U.S. and Asia at its Tokyo-Narita (NRT) hub, which it inherited after merging with Northwest Airlines in 2008, by removing nonstop service from New York (JFK), Los Angeles (LAX) and Minneapolis/St. Paul (MSP) (conditionally) in the U.S. and Bangkok (BKK) and Osaka (ITM) in Asia.

Though technically Delta’s LAX and MSP links will be shifted to Tokyo-Haneda (HND) airport, which the carrier won in the latest round of slot distribution awards from the U.S. Department of Transportation (DOT), this will effectively remove connection opportunities from either city to the remaining Asia-Pacific markets that Delta offers out of NRT: Shanghai, Manila, Taipei, Singapore, Koror, Guam and Saipan.

Delta has stated that the future of its MSP–NRT link is uncertain and being “evaluated” in the context of the slot it received to offer nonstop service to HND, but the likelihood of Delta offering two daily flights from MSP to Tokyo serving both airports is slim.

The Atlanta-based carrier has previously tried operating side-by-side service from its west coast hubs—Seattle (SEA) and Los Angeles—to both Tokyo HND and NRT, with little success.

This does not bode well for Minneapolis given that both LAX–Tokyo and SEA–Tokyo are much larger local markets than Minneapolis–Tokyo, even though Delta is uncontested in the MSP–Tokyo market, while LAX – Tokyo and SEA–Tokyo are much more saturated among competing carriers.

In its application to the DOT advocating for service from Los Angeles, Minneapolis/St. Paul and Atlanta to HND (it only received two out of the three), Delta stressed that the lack of a Japanese joint venture partner at either NRT or HND airports place the carrier at a disadvantage relative to American Airlines (AA) and United Airlines (UA), who were competing for the same share of limited landing slots at HND.

It was implied, if not overtly, that receiving the slot pairs for LAX–HND and MSP–HND would come at the expense of the NRT flights from each airport.

Delta defended that, in its application, both LAX and MSP were poised to capture U.S. point of sale from the West Coast and Desert Mountain (LAX) and from the Midwest, Central Plains and Northeast (from MSP)



Minneapolis/St. Paul, Los Angeles, New York JFK and Bangkok are all long-standing routes that Delta inherited from Northwest

Delta’s ending of New York JFK–Tokyo and Bangkok–Tokyo service dates back to the 1950s, when Northwest received permission to fly to the Japanese Capital (along with Pan Am) as part of the bilateral aviation treaty ratified by the U.S. and Japan.

Delta had previously announced that it was cutting its LAX–NRT route earlier this year when it received approval to begin operating daytime slots into HND, beginning October 30, and therefore concluding its LAX–NRT flight on October 29. New York JFK–NRT will end on October 3, 2016 while Bangkok will end on October 30 as well.

Delta was the only remaining U.S. carrier serving Bangkok after United cancelled its NRT–BKK service in 2014. The commercial aviation market in Thailand has been crippled by economic unrest as well as the rise of Gulf carriers offering deeply-discounted fares between Europe, North America and Africa and Southeast Asia.

Osaka is also an interesting route that Delta has served off-and-on from the U.S. mainland and Japan since the merger with Northwest. It was formerly linked to Seattle, but dropped in 2013 as Delta wanted to focus on alpha-tier Asian markets in its transpacific push from Seattle, which included an expansion to Seoul and Hong Kong the following year.

Delta resumed service to Osaka in March 2016 utilizing a 757 from Narita, which served as a continuation of its JFK–NRT route. Osaka is a notoriously expensive airport to operate into with steep landing fees, which likely rendered the route as an unworthy expenditure. However, Delta will remain.

Finally, Delta’s withdrawal from New York–Tokyo will leave the market entirely to Star Alliance and oneworld. Presently, four carriers (Delta, Japan Airlines, ANA and United) offer 42 weekly services from New York area airports (JFK and Newark) to Tokyo Narita.


Beginning October 30, ANA will move one of its two daily services from JFK to HND from NRT, thereby maintaining the same number of seats and frequencies, but simply changing operations. That same day, Delta will effectively leave the market.


This will reduce ASMs in the New York–Tokyo market by approximately 21%, represented by the market share that Delta presently has in the New York–Tokyo corridor. Effectively, Star Alliance’s market share will grow from roughly 50% to 63% during that period.

At Haneda, Delta will face extremely limited codeshare and connecting opportunities, particularly for its MSP route

Delta does not have an expansive list of codeshare partnerships between it and its fellow SkytTeam carriers on any of its nonstop routes from the U.S. to Asia, let alone between the U.S. and Tokyo.  Among its Asian SkyTeam partners, it carries codes for the following carriers on the routes below:


The codeshare partnerships that Delta offers on routes out of Tokyo airports are even more limited. Currently, Delta only places its code on three routes from Narita to onward Asian markets (China Airlines to Kaohsiung, and Korean Air to Seoul and to Busan), and does not offer any codeshare services on Skyteam partners operating out of Haneda airport.

There are going to be significant limitations for Delta’s LAX and MSP services given that there are only eight flights on SkyTeam carriers from Haneda airport each day, offered on China Eastern (Shanghai Pudong and Hongqiao), China Airlines to Taipei–Songshan, China Southern to Guangzhou, Korean Air to Seoul Incheon and Gimpo and Garuda to Jakarta).

Of these, only two out of the eight can realistically connect to the arrival and departure times of Delta’s services to LAX and MSP (China Eastern to Shanghai and Korean Air to Seoul – Gimpo).

Los Angeles – Tokyo Haneda
DL007 LAX1050 – 1545+1HND 777 D
DL006 HND1725 – 1020LAX 777 D

Minneapolis – Tokyo Haneda
DL017 MSP1145 – 1405+1HND 777 D
DL016 HND1635 – 1335 MSP 777 D

From LAX, this will not be as much of a problem given the abundant SkyTeam presence from Asian carriers. Excluding Delta’s daily flights to Shanghai and Tokyo, SkyTeam carriers together offer 91 weekly flights from LAX to Taipei (China Airlines), Guangzhou (China Southern), Seoul (Korean Air) and China Eastern (Nanjing and Shanghai).

Similarly, the void left by Delta between New York and Tokyo will be backfilled somewhat by the presence of SkyTeam carriers that fly nonstop between JFK and Asia, including China Airlines (five weekly to Taipei), China Southern (14 weekly to Guangzhou), Korean Air (14 weekly to Seoul) and China Eastern (14 weekly to Shanghai).

For Minneapolis, however, the situation will be far more challenging.

Delta’s Asia-Pacific operations are stagnating relative to United and American

Delta has not seen any organic growth in its transpacific network other than the proposed service from Los Angeles to Beijing, which is going to be partially offset by a competing bid from American to serve the same route. It has not added any additional service from Seattle since 2014, and its Narita operations continue to wind down.

United, on the other hand, has expanded aggressively out of San Francisco, and to a lesser extent, Los Angeles, while American has added links from Dallas/Ft. Worth and Los Angeles to Asia in recent years.

Delta’s challenge will be to balance its middle-ground position between two competitors who are playing catch-up in the Asian markets (American) or who are augmenting an already entrenched position (United) while both arguably have far better-standing relationships with their alliance partners.

For Delta, the opportunity costs of abstaining from a more traditional alliance-partnership structure with some of its SkyTeam peers may no longer be worth the effort for the sake of being number three in Asia.


About Author



A Global Review of Commercial Flight since 1994: the leading Commercial Aviation publication in North America and 35 nations worldwide. Based in Miami, Florida.

Related Articles


  1. Tim Dunn
    Tim Dunn August 11, 12:29

    There are several errors to what has been said here and are repeatedly in discussions about Delta and Japan.

    – Northwest knew that NRT was not viable as a hub even before it merged. They could not sufficiently cover the western US to Asia other than to Japan because there were nonstop flights on many carriers that made and still make connections in Japanese low yielding. Delta has cut its dependence on Japan almost in half in the 8 years since the merger.
    – Haneda access is about the local market no matter what AA, UA, JL and NH say. There simply are not enough seats to Haneda to allow any meaningful connections to be sold without displacing local revenue. MSP-HND will do well because it is one of only 3 US-HND flights that are not on the west coast. DL will have the only US carrier non-west coast/non-Hawaii Haneda flight.
    – HND will pull the highest value passengers away from NRT and will kill NRT as a hub whether AA/JL and UA/NH admit it or not.
    – The Japanese government has slowly opened HND and in the process has limited how much access AA and UA can get. Based on the original HND route case and the results of this one, Delta will get at least as much new access as AA and UA get with Hawaiian very much taking a chunk of access. The end result between the decline of Narita as a hub and the slow access to Haneda is that Japan will become less and less of a connecting hub in Asia. New flights from the US to other points in Asia only ensure that Japan will become less and less meaningless in TPAC aviation.
    – Northwest did not operate New York to Tokyo service at the time of the merger. Delta restarted it and it will operate for 8 years. American started and ended 2 JFK to Tokyo routes – one each to Narita and Haneda. Delta simply did not see it had a chance of profitably filling planes from JFK to Tokyo competitive with NH. It is very likely that some of JL and ANA’s flights from New York to Tokyo will end and UA’s EWR-NRT flight will also become a lot less profitable.
    – LAX is even more dicey for NRT services. All 3 alliances will operate HND flights and there is far more service to the rest of Asia. It is very likely that within a short time, the amount of capacity from LAX to Japan will fall dramatically.

    Delta has its code on more Korean flights from Seoul than AA or UA do from either NRT or HND.

    DL has its code on more flights beyond PVG than any other US airline does beyond any gateway in Asia. Delta’s partnerships with carriers in Asia are more than sufficient to provide access to the markets.

    Finally, this is only one phase of Delta’s restructuring of its Pacific operation. The “cut” phase will be followed by growth phases. Conclusions about the size of Asian operations cannot be made until the complete impact of what Delta adds and how other carriers Japan operations are impacted are known. Delta’s announcement about what it is cutting is just the first phase of several phases by several carriers.

  2. Rohan Anand
    Rohan Anand August 11, 14:06

    @Tim: thanks for the comments and feedback. to answer some of your points:

    1. I don’t believe I had ever implied that DL was trying to preserve the NRT hub after it was inherited from the NW merger. I was indeed incorrect about the JFK-NRT route timeline in that it was dropped by NW before re-started by DL.

    2. Your point about MSP-HND performing “well” because it is not located on the West Coast is arbitrary because there are no proven data points (yet) for a daytime flight that has been successful from the U.S. to HND. This is resting on the assumption that *all* of the US-TYO traffic will fly on DL via MSP. Given that there are multiple remaining gateways that fly nonstop from the U.S. to NRT that are not located on the West Coast (ATL, IAD, JFK, BOS, ORD, DTW, DFW, YYZ and IAH), some with multiple daily services, the number of options to fly into Tokyo Narita will still outnumber Haneda by a lot. Furthermore, your theory rests on the assumption that the vast majority of US-point-of-sale travelers are aware that they can fly into both airports.

    3. You point of HND pulling HVF customers away from NRT on ANA/UA and AA/JL is negated by your earlier statement that HND doesn’t have the bandwidth to provide meaningful connections with the restrictions in place.

    4. There will still be a need to connect flow passengers over Japan to dozens of markets in Asia Pacific. The notion that TYO serving as a transit hub will become completely obsolescent over time is illogical given that it is literally located at the cross roads of the world’s most populous regions (Asia-Pacific) and high-volume traveling markets (USA).

  3. Rohan Anand
    Rohan Anand August 11, 14:22

    @Tim: CONTINUED:

    Per your comment on the codeshare agreements, your data is incorrect. Per Diio Mi, for the current week, Korean carries the Delta code on approximately 415 weekly flights to the U.S. and Asia, and that includes *all* Korean Air flights from ICN and PUS. JAL, on the other hand, carries the AA code on 510 weekly flights between Japan and the U.S. and Japan and Asia, of which 385 are from NRT and 13 are from NRT. ANA also carries the UA code on 226 weekly flights from HND to the U.S. and Asia, and 427 from NRT to the U.S. and Asia. Put simply, AA and DL are roughly on-par between NRT/HND and ICN and UA is miles ahead of both.

    Additionally, your data on DL placing its code on more flights beyond PVG is also inaccurate. Data for the current week shows that PVG offers 329 flights to Asia carrying the DL code. Out of HKG, AA places its code on 301 weekly OneWorld flights out of HKG (virtually all on CX and a few JL). Out of PEK, UA offers 701 codeshare flights on Star, more than double the amount of Delta/SkyTeam from PVG.

    The point I was trying to make, moreover, in the codeshare analysis was specific to the MSP flight, which will face extremely limited opportunities out of HND for beyond traffic.

  4. Tim Dunn
    Tim Dunn August 13, 11:32

    Thanks for your response and for addressing the issue in an article that could be discussed. Hopefully more people will jump in.
    There is indeed data that shows how well US-Haneda flights do relative to Narita; all are from the west coast and they show that Haneda and Narita generate average fares close to parity – and that is with Haneda flights operating at night.
    The reason why no Haneda flights have worked from other than the west coast or Hawaii is because of the slot restrictions from the Central and Eastern time zones. Both AA and DL operated flights from the US to PEK that had poor slot times that weren’t quite as bad as Haneda but received better slots and resulted in very obvious revenue performance improvements.
    The value of Haneda vs. Narita is a question that will only be answered after we have at least a quarter or more of daytime flights – but public data for the west coast to Haneda flights will be available in about 8 months – close to when MSP-Haneda starts.
    Delta simply believes, and based on the experience at Gatwick, flights to Narita simply will not perform as well when passengers have a choice of going to Haneda. Even if some people don’t agree that Haneda is a more convenient airport, a valid alternative will impact NRT. There also is no example of a global airline operating two competing longhaul hubs at two airports in the same city as NH and JL say they will do at HND and NRT. The Japanese government’s policy is for long-haul premium carrier flights to move to Haneda so at best the discussion is about the transition time between when a few flights are opened at HND until when HND is the only option for most global carrier flights from the US to Tokyo.
    DL’s MSP and NH’s ORD flights will be the only two that connect to markets throughout the eastern and central US. Those two flights do represent a fraction of the capacity that currently operated east of the Rockies to Narita. DL is not trying to support HND based on feed at the HND end but rather its own connections at MSP. NH/UA will have the advantage of feeding the ORD-HND flight at both ends but the JFK-HND flight will have no US codeshare feed. The west coast to Haneda flights already carry a disproportionate amount of local Tokyo traffic compared to their Narita counterparts which says they very likely will pull passengers from existing NRT flights. DL’s decision regarding JFK-NRT likely is because they won’t try to compete against NH which will operate JFK-HND flights while DL will not operate MSP-NRT alongside MSP-HND.
    I believe there will be an impact to other carrier’s NRT operations; DL is simply making the first move but other NRT flights will be impacted including at LAX, ORD, and in the NE where there are multiple NRT flights that will now have to compete with daytime HND flights. AA’s displeasure with the DOT’s decision in not granting it DFW-HND service is because they know they will be at a strategic disadvantage with not having a central US HND gateway compared to DL and UA. AA/JL also have to compete against another carrier with all of their HND flights while DL and UA each have one or more HND flights where they do not have to compete directly against another airline.
    Finally, I accept the differences in data that you highlighted about codeshares. I’m not sure we are using the exact same baseline but that is ok because the overall picture is the same. DL is not trying to support HND service based on codeshares. DL does have multiple codeshare partners in Asia at other major gateways outside of Japan. UA has a multi-carrier/multi-national joint venture while AA has a joint venture/joint business arrangement only with JL which is functioning as a multi-national JV. UA was the largest carrier to Asia as a result of the acquisition of Pan Am’s Pacific network 30 years ago and they have been building from it for years. DL merged with NW and its Pacific system and has spent 8 years refocusing away from Japan with a focus on growing China which likely will pass Japan as the largest US transpacific market. AA has been engaged in aggressive internal growth alongside its JL JBA but size and transpacific yields are still in the order of UA, DL, and then AA. Access to Haneda and the gateways involved will likely not change that.
    Japanese government policy combined with growing airline networks to other points in Asia will negatively impact the role that Japan has historically had in transpacific aviation as a connecting point. Japanese airlines will be weakened compared to airlines from other countries that have faster growing local markets, lower labor rates, and aren’t trying to support two competing long-haul international hubs, even if on a temporary basis.

  5. Rohan Anand
    Rohan Anand August 14, 19:44

    Hey Tim –

    Thanks again for the comment/follow-up. I enjoy healthy debates as always!

    I agree with you that the data shows revenue parity for flights from the US to NRT and HND. In fact, this was something I raised in a previous article: for Q4 2015, per airline performance data, both routes averaged load factors above 84%, and yields were nearly identical with $0.01 variation. LAX – HND obviously managed higher local load factors given that Haneda is a terminating point for Delta, but overall, Narita generated higher average fares, revenue per available seat mile (RASM).

    Now, this was, of course, WHILE Delta had poor slot times from HND, so those metrics will most definitely evolve. We will see how LAX/SFO-HND perform in the 8 month period you’ve specified. However, I do not buy-in to the Gatwick comparison based on the fact that LGW was, and never has been, the primary/preferred airport into London. It has ALWAYS been Heathrow, and when Bermuda II dissolved, virtually every US carrier shifted to LHR. Whereas, from Tokyo, Narita has always been the gateway airport to global carriers, and still very much is despite some adjustments.

    MSP-HND will have to rely solely on local traffic and feed from the US point-of-sale. We will have to see how that does. I don’t think those situations matter as much for ORD and JFK, with or without codeshare feed, because both of those are MUCH larger local markets to TYO than MSP.

    AA is displeased about being displaced by DL for the DFW-HND route because its a hub-to-hub route that makes far more business sense than MSP-HND. DFW is a OneWorld hub, with far greater connection opportunities than DL at MSP. As-is, DFW-NRT is offered thrice daily between AA and JL. American also supports multiple long-haul flights to Latin America, and there are hundreds of Japanese companies with major operations in North Texas.

  6. Tom C
    Tom C August 16, 14:39

    Delta’s connection flight between Osaka and Narita actually originates and terminates at Kansai International Airport (KIX). Itami Airport (ITM) near downtown Osaka no longer has international flights.

  7. Frank
    Frank August 16, 20:41

    There are other costs to this for Delta that no one has mentioned, namely, losing flyers who do a huge amount of US domestic and Japan travel (Tokyo is not the only major business city in Japan). When you lose Osaka flyers you don’t just lose their Japan business. Northwest used to have great service to Osaka and other connections within Japan (Fukuoka) that Delta has slowly eroded. Tokyo is essentially useless for many people working in Japan, and even for tourists who want to go to Kyoto as their main destination (close to Osaka). Delta has lost a large number of high mileage flyers who commute regularly to Japan because Osaka is a major business hub and part of the Kansai region, which includes Kyoto, Kobe, and Wakayama, all of which are becoming hi-tech havens. Most of these folks just switched airlines and partner teams to United or American. I stopped flying Delta to Japan and all my domestic routes when they cut SEA to Osaka and switched to United/ANA and then back to Delta when they added NRT-KIX. I will now probably switch to American/JAL. A lot of the former Osaka flyers didn’t come back because they were worried Delta would cut the route again.

    The Korean Air codeshares Tim mentions are often made cost prohibitive by Delta so that it is more efficient to just book on Korean Air through Seoul or China Airlines through Taipei.

    Delta’s DTW-NGO route is great and can be convenient to Kansai, but depending on who is paying for my ticket it can be cost prohibitive because Delta’s fares to Japan are rarely competitive in economy or business class.

    Finally, anyone who has done a lot of flying into or through PEK or PVG will tell you that customer service is nowhere near what it is in Japan, Korea, or Taiwan.

Only registered users can comment.

Current Issue


Subscribe to our weekly Newsletter


Would you feel safe flying on a Boeing 737 MAX once the grounding is lifted?

View Results

Loading ... Loading ...


Airways YouTube Channel


Would you feel safe flying on a Boeing 737 MAX once the grounding is lifted?

View Results

Loading ... Loading ...