I’ve already looked at how American, Delta, and Southwest scheduled themselves in May, and now it’s time to round this out by looking at United. A quick peek at schedule data via Diio by Cirium made it clear that United had cut VERY deep.
Let’s put it this way. In May of last year, United operated 34,451 flights. This May it will have a mere 7,142, or about 20 percent of what it flew last year. Considering that United said there were single days last May that had more passengers than the airline expects for the entire month of May this year, this sounds like it could have gone even further. Though to be clear, there is also a massive downgauging here that means United will only fly 15 percent of the seats it flew last year.
During the week I reviewed — as with Delta, it was May 13-19, 2020 versus May 15-21, 2019 — United will not operate a single flight on its A320, 737-900, 757, 767, and 777-300ER fleets. Take a look:
You can’t see it, but there are actually 4 flights a day on the 777-200. Other than that, the 787 is the only widebody flying. And it’s barely flying.
International is Gone
United did not hesitate to absolutely crush its international network in May.
And before you say anything, I know. St Thomas, San Juan, and Saipan are not international flights, but I put anything outside the 50 US states here. And it’s just not much.
Outside the skeleton remains of the Guam network, United will only fly San Francisco to Tokyo/Narita and Sydney over the Pacific. Over the Atlantic, it’s just a handful of flights to Amsterdam, Frankfurt, and London/Heathrow. And outside of Sao Paulo, Latin/Caribbean is all narrowbody flying.
As gutted as this is, it’s really the domestic network that’s more interesting.
Three Hubs/Focus Cities Are Gone
Unlike Delta, United has very few flights that don’t touch its hubs, but what it did have is now gone. And it’s not just that, but two hubs disappear as well. (I’ll talk about one now and the other later.)
Let’s start with the only true focus city, the remains of the Cleveland hub.
With its focus cities, Delta at least kept connectivity to the hubs, but United had no mercy in Cleveland. All that’s left are flights to Chicago/O’Hare heading west and Washington/Dulles heading east. This makes it a rather small spoke.
But as mentioned, Cleveland wasn’t the only one to get the axe. Los Angeles saw something similar, if slightly less dramatic.
LAX now has service only to the other United hubs. That is it. San Francisco is all operated by Embraer 175s. Newark is now on domestic 737-800s, and everything else is on the A319. This too is just a spoke in May.
San Francisco Becomes a Western Regional Hub
Unlike Delta which kept more flying on the West Coast then I would have expected, United has slashed and burned. San Francisco, the airline’s premier California hub and Pacific gateway is now just a regional hub.
Let’s start with what was cut.
If that looks like a lot of routes are disappearing, that’s because they are. In fact, every non-hub route east of the Rockies goes away except Boston.
In other words, San Francisco remains to serve those destinations in the west that mostly aren’t served by other hubs. This is almost entirely about complying with the CARES Act minimum service requirements, outside of the big cities like Seattle or Las Vegas, which have service from multiple hubs.
This is the sole gateway to Hawai’i now, and there is something unique in there. United is flying a once weekly flight from San Francisco to Kahului to Hilo and back to San Francisco. That apparently will satisfy Hilo service requirements even though the market is usually served only from Los Angeles. What’s more, I don’t think they’re even selling the flight to Hilo, only the return.
Denver: A 1,000 Mile Radius
The rest of the country largely appears structured similarly. United is trying to shrink them down to serve their regions. If we look at Denver, United has effectively drawn a circle with a 1,000 mile radius and cut almost everything outside.
You can see that most anything east of the Mississippi is gone with a few other things closer in. The end result is a highly-tailored network.
San Francisco gets the west, Denver gets the mountains, and that means…
Houston Takes the South
At first glance, it doesn’t look like Houston has lost all that much compared to the others.
But once you look at what’s left, you can see exactly what United was doing here.
Houston is basically a Texas-South hub now. Yes, there are large cities in the West that feed into it, and there are some big northeast airports, but this is really about serving the South from Texas and surrounding states. That means, you probably know what Chicago is going to do.
Chicago Takes the North
Just as Houston serves the South, Chicago will serve the Upper Midwest, but it’s more of a North-East strategy. As you can see, most of the cuts here were south and west.
And when we see what remains, it reflects that fact.
As with Houston, big cities in the west retain service, and of course… Florida. But the rest is primarily in the northeast of the country.
And speaking of the Northeast, United has two hubs. Which one would win that battle? Not the one you’d think.
Newark Takes a Back Seat to Dulles
The natural assumption is that Newark would be the primary northeast hub, but that’s only if you don’t watch the news. New York is the epicenter of the COVID-19 outbreak in the US, and so there is no demand. Combine that with the near elimination of Transatlantic service, and what good does Newark do? Not much.
I never thought I’d see the day when the mighty Continental hub in Newark was reduced to only serving other United hubs within the US. There is a 787-8 flying to San Francisco, undoubtedly to feed the airplane between the networks, but everything else is on an A319 or 737. Yes, there are still a few Atlantic flights, but Newark has ceased to exist as a hub domestically.
And that means Washington/Dulles is the winner. That’s not to say it didn’t lose routes. It most certainly did.
But most of the routes that were cut were east-west routes. Those can be served well via Chicago, Denver, or Houston. Then there are routes like Manchester and Providence which, as part of the Boston metro area, are allowed to be cut entirely from the United system. We would have seen more of this if the government hadn’t been so strict with minimum service requirements.
The end result is:
Dulles is now the primarily north-south hub for the airline east of the Mississippi. Most of the flow is still regional; that isn’t a huge amount of service into the South. But this does cover the last remaining piece of the country that United needed.
Wrapping It Up
As you can see, United cut deep. It has ditched Cleveland, Los Angeles, and Newark in May along with most international operations. The rest of the hubs are meant to serve the cities that are closest to them, and that means a whole lot of city pairs will now require two stops instead of one to connect. Despite that, I can’t help but wonder what this would look like if there government rules on serving every city weren’t in place.
It’s incredible to think that United has a mere 20 percent of the flights it had last year… and 15 percent of the seats. But it’s more incredible to think that it should be less. Without the government changing its rules, however, this is probably about as good as United can do.
PQI caught my eye and I had to look it up… I didn’t realize that United sold EAS flights to PQI.
Looks like they are using a 50-seat EMB-145XR operated by CommutAir with the plane staying in Northern Maine overnight, and switched from EWR to IAD a few months ago, just before COVID hit. At 716 miles, PQI-IAD must be one of the longest EAS flights in the lower 48. That big of a plane and that long of a route seem like overkill for EAS, even pre-COVID, but I’m not familiar with the route.
Last graphic says “ORD” instead of “IAD”
Some of those maps look very similar to each hub’s mid-80’s developmental period. Spooky.
It is noteworthy that AA and UA cut RJ flying a lot less than DL in percentage terms. That should be concerning to their labor groups as both try to lean on regional carriers even more than mainline to recover.
UA’s hubs are much more skewed to international and to longhaul domestic local traffic, both of which are greatly reduced.
On top of its balance sheet stress and high capex, there are good reasons why UA was the first and has been the most aggressive to cut capacity.
“Let’s start with the only true focus city, the remains of the Cleveland hub.”
COME BACK UNITED, WE LOVE YOU! MAKE US A HUB AGAIN PLEASE!
(Or at least do something with the closed off D Terminal that you are still paying for; I’ve never got to go in there and I heard there are cool things)
There was no demand that would have justified reopening D prior to this. There’s even less now. The concourse was built on one side for 145s, and turboprops on the other. It’s obsolete.
As for the concourse itself, it’s nothing special. Lots of glass, very open. Much nicer than, and 40 years more modern than the rest of CLE, but that’s really about it.
Maybe it is just home bias, but I like how CLE is styled, sort of an art deco look kind of reminds me of Tim Burton’s Batman or Batman TAS.
Some other airports I’ve been in are just a mess, or they lack soul.
I heard terminal D has some like hanging paper airplane sculptures and I want to see them in person someday.
And that means that it is time for United to come home to Cleveland, and reopen the terminal.
It’s interesting to see just how Newark went from mega hub to single digit flying in such a short period.
I think you misspelled Mississippi, or it was my screen reader.
Well, I got one right. The other is fixed.
After comparing the big 3, it is interesting to see what has happened to LAX. Although LAX has a significant O&D, it’s importance (lack thereof) as a fundamental domestic hub seems to be re-emphasized with the cuts made in this market. Also makes me wonder how profitable this hub has been for any carrier.
I wonder if the better strategy is to focus well on certain markets from LAX instead of trying to serve as many destinations as possible.
It’s a weird time at LAX. The cuts show that United prefers SFO, American prefers PHX, DL prefers SLC and SEA. Now’s a bad time not to be a fortress hub.
Cranky, a fun project would be a map of all the airports that have lost service entirely to LAX in May.
I second that. It would be interesting to see the destinations lost and capacity dropped by airline at LAX in May.
Jimmy – This is a good idea. I’ll put it on the list.
It’s curious that over the Atlantic, both Delta and United kept service between New York and Tel Aviv, a fairly distant spoke with no further connection opportunities. Any idea why? Does it not make sense to serve Tel Aviv through their European partners?
TLV is a fairly significant cargo market; cargo will be the difference in how quickly some markets come back given that half of the world’s cargo capacity normally flies in the belly of passenger aircraft so a lot of capacity has been lost in the global air cargo system.
Also, Israel did a better job of controlling the disease than Europe so a lot of passengers would rather go on a nonstop.
Israel also has a very high (relative to its population) density of medical, pharmaceutical, and life sciences companies, so much of the cargo going to/from TLV these days is likely COVID-related.
From a route scheduling perspective, at what point do you think the UAL planners will revise June downward? With most businesses having already declared no customer travel until mid-June (the large technology company where I work for one), I can’t imagine there will be any sort of material uptick in travel. Between businesses restricting employee travel, 20% unemployment and leisure travel making a VERY slow rebound, one would think they’d want to get ahead of things in June/early July as well.
Also, I’m sure you’ll write about it later, but how stoked are you, Cranky, that Alitalia was nationalized again? It’s the gift that keeps on giving.
Chris – United had been doing weekly updates, so it would roll one week at a time. I’m not sure if that will continue or if we’ll just see a big dump, but June has already been brought down to some extent.
Looks like United’s strategy is more about eliminating risk of empty ASM’s, thus regional HUBs with shorter stage lengths and smaller airplanes, than chasing any revenue. It seems they are pretty convinced that May Revenue will be abysmal, nearly hopeless, so why not batten down the hatches and conserve cash. June may be the same. The third quarter is the big, big question mark.
One of the most interesting things to me on all this that wasn’t captured here (it’s unique, for sure) is how UA is serving EYW in May since they weren’t given an exemption.
An E175 is flying ORD-RSW-EYW-RSW-ORD. There is literally no other shorter route you could serve EYW from commercially. Even MIA-EYW is farther.
Kind of unique and impressive. Talk about a fun route to ride in normal times.
from the look of it, you can’t even buy RSW-EYW as a standalone.
Great Circle Mapper shows RSW-EYW at 136 miles, 10 miles longer than MIA-EYW, but nonetheless that’s definitely an interesting routing.
@Cranky, please do a post featuring some of the more “creative” stops and routings in the schedules now… Would make for some great fun and discussion.
You’re right. I misread the length before. Thanks!
F9 will be flying MSN-MKE again. That should definitely make the list!
That’s an interesting one, especially for Frontier.
There are (or were, traditionally, pre-COVID) a lot of routes connecting small/medium cities, especially college towns, to nearby hub/big city airports that are ~1-2 hours’ drive away. Lansing, Richmond, Charlottesville (VA), and State College (PA) come to mind off the top of my head, but I’m sure there are plenty more.
Finding a short route that connects two (relatively) large cities in the US is a bit harder, especially if you limit it to a route served by a least some mainline aircraft (no RJs or props) in pre-COVID times. MKE-ORD is the classic example that I know of, but there may be others.
I know this is an old thread, but FNT-DTW is even shorter. Growing up, I used to fly AZO-ORD and that was fun because going westbound, you’d arrive in Chicago “before” you left.
Whoa, did they suspend the island hopper? I wonder how the Marshall Islands/Federated States of Micronesia are getting their supplies now..
Air Marshall Islands is still running flights to some places, UA has one or two flights planned to Kwajelein in May.
They are telling the people working on Kwaj that there will be one UA flight to HNL on May 20 and one flight to GUM on the 21st of May.
the one thing that REALLY caught my attention is given the extremely low demand resources, having to pick the most essential long haul routes to fly, UA skips MUC ZRH BRU GVA CDG DUB MAD BCN FCO MXP but continues to fly to….
A Skyteam fortress hub with essentially zero downstream feed for UA. That tells you a lot about how strong a corporate contract must be backing AMS for UA.
AMS is a huge cargo hub and I am sure UA gets some of it.
DL is flying ORD-FRA as a fairly regular cargo flight too so there are plenty of things that don’t otherwise make sense.
I thought AMS was odd too, but it might be cargo related. Flower exports (especially tulips) are a huge part of the Dutch economy and spring is the peak. Perhaps this is the reason behind the flight.
perhaps. in this day and age, you’d think tulips, of all things, would be the least of anyone’s concerns.
Garden stores are doing a booming business in many parts of the country. Home Depot and Lowe’s have managed to get themselves considered essential businesses so people are investing plenty in their yards since they are stuck at home and weather is nice in many parts of the country.
AMS is not just about flowers, though.
From the news reports I have seen, the flower markets in the Netherlands are basically dead.
Remember that social events (weddings, celebrations, etc, along with holidays celebrated together like Easter and Mother’s Day) drive flower purchases. Prices are a fraction of what they were pre-COVID, to the point that growers in Africa are having to pay workers to harvest flowers and dump them in the compost heap, and air cargo rates are way up.
Your comment is correct… which is why I said that AMS is not just about the flower trade. And also why UA’s presence in AMS has continued. Remember that DL has/had a fairly large presence in Germany for a carrier with no alliance partner there including by basically connecting the US and German auto industries – both in DTW and from the SE. AA also has/had a large presence in Paris while DL was the largest carrier in Tokyo even after alliances were formed. Even now, DL is still the largest foreign carrier at HND based on routes allocated.
Not everything about US airline size in foreign markets can be explained by alliances.
Royal Dutch Shell Oil corporate contract maybe? Global HQ at AMS, US HQ at the IAH hub.
“Spooky,” yes, but historic it is. Your maps are wonderful.
Will be nice to see what the load factors will be on these flights. Probably no worry, as I’m sure most of the its revenue will coming from those well-advertised UA Explorer Visa cards, of which I seem to be getting one or two emails a day! I know, times are running out for me to collect my 60,000 miles, but I demure.
Which brings to mind your table of UA flights and how nearly all of them will be 50-seat and 70-76 seaters piloted by regionals SkyWest, ExpressJet, Republic, CommuteAir, Mesa, or AirWisconsin, or whatever regionals are still around in May. I would guess all the mainline flights will be on A319s and 737s, essentially just hub-to-hub, until there is no need even for those flights.
We are making grants, giving loans to an airline that exists in name only. If it avoids bankruptcy, probably come out the other side as an air carrier with none of its own pilots or aircraft, run by a CEO whose degree is in contracting and contract law, never having worked at an airline.
But anyone, a good post!
Thanks for the analysis Brett. Terrific, as usual. How far does this extend beyond May? Or is that not known yet.
As you know, my primary interest is flying UA between the LA area (LAX or BUR) and the Boston area (BOS or MHT). At least for May, it looks like my options are limited. Here’s hoping by late-summer/early-fall, more of the normal options return.
Hajime – This just goes into early June for now.
Well, no nonstops, but should be several options through ORD, IAH, or IAD.
Not sure why you say that all UA flying out of SFO is on E175s. They are also flying CRJ200s. Check out U5623 to TUS on 5/2, for example. There are others.
It makes all your data suspect.
Laura – I would suggest re-reading that paragraph. This is talking about LAX service. All flights to SFO are on Embraer 175s, as stated.
Keep in mind w/ all of these schedule discussions that there is still a big difference between even the big 4 in how much of their schedule they actually fly.
AA and WN are cancelling a lot more flights close to departure while DL is operating most of the flights it schedules, as is AS. WN is cancelling nearly half of its flights within 48 hours.
UA is cancelling close to departure but far less than AA.
The only real way to keep track of capacity is when it is reported as part of the 2nd quarter financial reports – which will be in July
Given the new ‘network strategy’, how do you turn things back on?
It took a decades to flesh out the old United network by announcing routes, selling seats and then flying planes.
Matthew – It shouldn’t be that hard to turn back on. This is really very short term, and I doubt we’ll see it pulled back this much into the summer.
Like the scene in Apollo 13 – turning everything back on in the correct sequence is important.
Pilot Currency: UA is flying each of their narrow-body families as you noted – This likely is to keep their crews current. (Cargo will likely keep their 777 and 787? crew’s current, and 757/767 crew?)
But how does UA (or DL or AA) add back capacity without loosing more money?
(than they are now). Do they add frequency or up gauge on top of the existing model, or do they start to add additional hub-connection points?
Customers will still choose airlines based on departure, duration and pricing. That’s why I think adding back capacity is laden with complexity.
Matthew – I don’t think pilot currency will be as difficult as many expect. They’ll have plenty of slack in the system to get training done.
For adding back service again, that will depend on demand. This recovery should be entirely led by demand. And since I expect the booking curve to compress, they can make better decisions closer to departure.
That begs a good question…
As you note, with booking data you can ’see’ the number of pax that you flew SFO-IAH-STL to see if a SFO-STL makes sense.
But even with a compressed booking curve, are there enough data points in a climate like this to make any making anything meaningful of it? Or do you think we are going to get a lot of published best-guess schedules (and then the noted decisions closer to departure.)
Matthew – I think we see a lot of caution and no rush to add back service.
My guess is they’d try to bulk up the hubs on key routes first. If they see demand, they can just keep adding back. Then over time, if they see enough people flying to St Louis over Houston, they can add a nonstop. But there’s no hurry to do it.
Great analysis Cranky, as always. Typo heads up: IAH green routes map labeled “DEN”
I was surprised to see DEN and IAH keep SNA service while ORD lost it. Orange County has 3 million people and lot of very wealthy premium fliers while being gate limited. It will be interesting to see how the competitive dynamics play out there as this unfolds (in addition to the craziness at LAX – add my vote to the ones curious about an analysis there, Cranky). I fly IAH-SNA often and was concerned DEN and ORD might keep it while IAH was frozen out. UA’s competitive advantage at SNA has always been its EWR nonstop – sadly gone, for now. I suspect the AA-AS alliance will build a substantial advantage there over time with PHX and DFW hubs + intra-CA flying.
One pattern I see developing is airlines serving spokes “usually” to the closest hub or two. SNA-ORD would pass DEN and sort of IAH so this really isn’t a surprise to me.
The label for “IAH remaining routes” has DEN typo
There is one Island Hopper operating on May 20.