I recently received a note from a reader pointing out something curious. He noticed that some flights on American were now showing some very long flight times, much longer than in the past. I decided to take a look myself, and sure enough, this wasn’t just a one-off. In the last year, American has significantly increased its block times despite there being far fewer flights in the air.
To look at this, I dove into Cirium schedule data to find the average block time by route. “Block time” is the amount of time the airline schedules for a flight to go from gate to gate. It’s the time you see on the schedule when you book a flight. I decided to pick a sample week in March from this year and compare it to a sample week last year. I wanted to use March since the schedules from last year were pre-pandemic, and I had to make sure to use a very similar week so that seasonal variation wouldn’t throw me off.
I pulled out the routes that didn’t operate last year but did this year, and vice versa. That left me with 1,680 routes to compare. I should note that these are directional routes. Block times can vary a lot depending upon the direction. For example, Miami to JFK on average had a block time of 3 hours and 2 minutes. But JFK to Miami was blocked at 10 minutes longer. So I looked at each direction separately.
I added up all the block time for all the flights on each route during the week and divided by the sum total of flights. Then I just looked at the percent change year over year. Here’s how those routes all plot on a chart.
American Average Block Time Change by Directional Route
March 15-21, 2021 vs March 16-22, 2020
As you can see, of the 1,680 routes, only 275 saw a decrease, 21 saw no change, and 1,384 saw an increase. This really surprised me. After all, with traffic down, shouldn’t block times decrease considering there are fewer air traffic control delays and less gate congestion?
Before I dug in further, I wanted to do a sanity check to make sure that this wasn’t an industry-wide phenomenon. I looked at other airlines, and found the data to clearly show that it was not.
Average Block Time Change by Directional Route by Airline
March 15-21, 2021 vs March 16-22, 2020
While there was some variation, sure enough American has increased block on a higher percent of routes that anyone else by far.
I went back to the American data to see if I could break this down further and find some trends. I started with the routes that were seeing decreasing block time. Of the 20 biggest decreases, 6 of those involved flights to Key West. I’m guessing something happened in Key West that enabled them to reduce their block time. Or they just realized that they had put more than they needed after analyzing performance.
Another five of them involved departures from Washington/National. Some were on routes like DC to LaGuardia or Philly, where I assume the congestion used to just crush those flights. But even these gains were fairly modest. Still, this wasn’t the real story. It was more interesting to consider why other routes saw block times go up so much.
A clear picture started to emerge quickly. There were 23 routes that saw increases of more than 20 percent. Of those, a whopping 17 involved short-haul departures from Phoenix. Here’s the Great Circle Mapper map:
Once you get under 20 percent, Phoenix still dominates. So clearly there’s something going on with the airline’s Phoenix operation, but it’s a puzzle. And these block times appear very heavily padded. For example, American has a flight at 1:25pm from Phoenix to Long Beach that’s blocked at 1 hour and 50 minutes. The flight time itself rarely exceeds an hour. Meanwhile, Southwest has a flight at 2:55pm that’s blocked at only 1 hour and 15 minutes. Though Southwest usually does have a shorter taxi time in Phoenix, it’s not 35 minutes more. It goes both ways, the 7:04am return is blocked at 1 hour and 48 minutes on American while the 9:10am on Southwest is 1 hour and 15 minutes.
Phoenix may be the most aggressive, but it goes well beyond the Valley of the Sun. United has a flight at 2:09pm from O’Hare to Cedar Rapids that is blocked at 1 hour and 8 minutes. American has a flight at 2:15pm that’s 1 hour and 21 minutes. You might argue that with terminals in different places, block times can vary by airline and that is true. But last March, American’s average block time in this market was 1 hour and 8 minutes. United’s was 1 hour and 12 minutes.
Even though block times were up everywhere, I asked American specifically about Phoenix and what was going on there. I was given this statement.
We have increased PHX block times recently, but as you look over the past year American has the lowest block performance when compared to our major competitors.
That may or may not be the case, but American has purposefully run its business that way, keeping block time lower so that it could squeeze more flying out. But now, it has very clearly altered its strategy. Maybe it’s because the airline has more crew and airplanes sitting round, so they don’t need to do it. Or maybe they’re trying to mask operational issues. Regardless of the reason, this seems counterintuitive.
(On a separate note, your tips/intel can help produce really interesting posts. You can send your tips to email@example.com or if you want to send the intel anonymously, use this form.)
Could they be reducing flight speeds to a lower fuel burn? I always assumed they did all along but maybe they were focused on achieving block times on par with the days the convair was flying. I’m not sure the increase in labor costs is worth it but maybe they found some sweet spot.
If you’re going to have to pay the pilots and FAs anyway, due to the strings attached with taking Uncle Sam’s money, and you have more than enough pilots/FAs on the payroll than you have the need for, the additional labor costs may not really be very significant or relevant.
Jet fuel isn’t expensive now relative to longer-term history, but it’s double the price now compared to what it was a year ago, and 3-4x the ~ 40-50 cents/gallon that it bottomed out at in late April/early May of last year. (Source: https://www.eia.gov/dnav/pet/hist/eer_epjk_pf4_rgc_dpgD.htm )
My guess is that operational performance is the driver here and that any fuel savings are just gravy on the top, but nonetheless, when many flights are/were unprofitable, saving a little fuel by flying a plane 10 knots slower in cruise may make a difference. When you have excess labor that you’re paying anyway and can’t shed,
As a reference point, during economic contractions shipping lines often direct their ships to cruise at slower speeds instead of laying them up, as doing so reduces their costs per voyage while also reducing the total supply (in terms of container-km moved per month) of shipping.
I suspect the longer scheduled block times are often ignored by crews and flight planners and the number of early arrivals has also gone up dramatically avoiding any increase in fuel burn cost.
You’re probably right, as that would be consistent with increasing block times to juice on-time performance and pushing to get the planes to arrive “on time”, no matter how early.
To your point in another post, it will be interesting to see a before & after analysis (of both block times and on-time arrivals) of what happens when the Cares Act support goes away, especially if air travel demand continues to increase.
On another note, once airlines are forced to increase the utilization of their planes again, it will be interesting to see (and to hear, anecdotally) to what extent the “increased cleaning and sanitizing” procedures between flights increase turn times compared to pre-COVID. My money would be on the effect being only slightly more than a rounding error, or barely noticeable at most, but we’ll see.
Flyingcat – Flight times don’t seem to be any different at a glance, but I haven’t done a full search to see. Still, it doesn’t really seem like that’s a cost saving measure.
flyingcat: “…with the days the Convair was flying.” 990? 240?
Before I reached the portion of your post about Phoenix, I assumed that AA had less competition on its routes and increased block times in order to boost their on-time performance. If your competitors don’t have a competing flight, your “longer” flight won’t get pushed down too far in the flight search rankings. Even your “slow” non-stop will still win vs. a 2 leg itinerary.
This is an interesting post and something I have noticed too specific to AA and notably, on JFK-MIA-JFK, which has a longer block time than it has in the past. This is surprising on the JFK side of the route, given the drop in traffic there since the pandemic. On most of the flights I have taken though, with either exact or not more than 10 minute late departure times, arrivals in both directions have been substantially early, by as much as 30 minutes. I have also been on AA flights between the two cities where the departure (on the JFK end) was between 10 and in one instance, 15 minutes early.
I remember as a child & teen traveling between LGA & FLL & the block time was a consistent 2:30. Same for if the flight was headed to MIA or PBI. now it’s a consistent 3:00. Granted that was in the 70’s & 80’s, but what explains the dramatic block increase between then & now? Also the block time was constent between airlines such as Eastern & Delta .
I think this may be due to the reduction of banks in Phoenix which leads to gate shortages during the remaining banks. They are also doing some tarmac work right now that is adding, what looks like, improved drainage systems. This is causing about 2-3 gates to be out of service at a time. Pure speculation on my part.
Although since crews are usually paid by scheduled block, or actual flying time, whichever is longest, I guess it doesn’t matter from a cost perspective since with the Cares payroll support program. Crews are already flying below their minimums, with some not flying at all, but are getting paid their minimums anyway because Uncle Sam’s actions.
I wonder if the block time shrinks once the Cares Act support is gone.
I was kinda thinking the same thing, that it had to do with crew pay.
Less of a factor, but aircraft utilization was (still is?) also way down so there was no need for a fast turnaround to get the airplane to the next destination.
Is there data to analyze how many flights end up in the penalty box because they are running early have landed and no gates are available?
That would be an interesting follow-up.
With such dramatic schedule reductions, I suspect a gate is always available when they arrive early. Penalty box holds are probably minimal to non existent.
SubwayNut – That data does exist out there somewhere in a sense. You can look at taxi times to see if they’re longer or not. But I don’t have that info at my fingertips.
I think a big part of it is that PHX has been reduced significantly in terms of the number of banks – from what I recall, it’s really two banks heading east and two banks heading west each day. Any misconnect really hampers the ability to connect those passengers through the hub. To ensure flights make their connections, from my view there’s two options – either lengthen the scheduled time that the plane is on the ground in PHX (of which, the downside to doing this is a passenger might see it as a two hour connection in PHX and think that’s not as pleasant of an option over an hour in DEN/SLC) or artificially pump the block times – to a passenger, they still think they’re getting an hour connection in PHX, and if you’re into the gate 15-20 minutes early, then you can announce that as an early arrival and passengers will be pleased (I remember a couple of years ago Delta LOVED to promote their early arrivals over the PA once the plane touched down).
It would be interesting to see, especially with PHX, how much of this is Mainline vs. Express, especially in markets which are served by both.
Sadly, the mostly unnecessary block/gate time buffers (padding) Brett identifies in this article costs individual large airline like American, and their shareholders, over $2 Billion annually.
Add in the other costs of inefficiency created by the airline’s 1950s “day of”, curb to curb production process and the waste rises above $3 Billion annually for large individual airlines.
And few seem to care that this is an easily solvable problem (3 years and $25 million dollars).
Quoting Rami Goldratt’s recent LinkedIn post, “The more that we have invested, the harder it is to abandon it”. The airline industry’s current 1950s, “day of”, curb to cub production process is no longer valid.
Do you have any sources for your claims at all?
FAA (Task J), Embry-Riddle, GE Aviation (Dubai) and others have independently validated, in actual airline operations (ATL, MSP, DTW, CLT, DXB, etc.), that airlines can simultaneously reduce scheduled block/gate time, emisions, delays and costs, starting in 2021. All of the operational tools (process AND technology) to rapidly make these a reality already exist, right now, today.
No grandiose, hugely expensive, highly complex plan (think Billions). No new technology. No more studies. No decades more waiting and waiting for others to act. Just, as other industries have done for decades, system focused, “day of”, internal operational blocking and tackling at a whole new level (right part, right place, right time), starting with real time, tactical management of the movement of the aircraft from a system perspective.
Please email me and I will forward the financial analysis.
So basically airlines know how to save billions of dollars per year but simply choose not to out of spite? That’s a bit difficult to believe.
I don’t care about the detailed financial analysis (I have no stake whatsoever in the airline business), but if the numbers are real you could post or explain them. The links you posted are all fluff/opinion pieces to advertise yourself rather than actual facts or data.
It is clearly not out of spite. And, yes, individual large airlines could save Billions each and every year. Look at GM, Ford and Chrysler in the 1980s, when what Toyota was doing was staring them in the face. Unfortunately, it took 30 years and bankruptcy before they took the time to understand the benefits of Operational Excellence.
Also consider Southwest airlines statement that “It would cost us approximately 8 to 10 airplanes of flying per day if we were to add just a couple of minutes of block time to each flight in our schedule.” (Greg Wells, Chicago Tribune, Mar 3rd, 2011). Given that the airlines pre-covid schedule padding was about 20 minutes per flight, half of which is easily recoverable, shaving10 minutes from each flight equates to 80 to 100 extra aircraft in Southwest’s schedule just for the pleasure of doing what they have always done.
Like yourself, airlines are obviously skeptical, but the facts I presented have been independently validated in FAA’s Task J report.
The roadblock to change is that the airline’s current decisions are based on their 1950s assumptions, i.e., silo over system, only ATC can fix this, capacity/weather/labor/fuel is the problem, the movement of the aircraft is not the airline’s responsibility, nothing can be done, etc., which have been invalid for decades.
Excess schedule padding, delays, costs, operational inefficiency, unnecessary greenhouse emissions, etc., can be rapidly and inexpensively reduced.
As Mark Twain has been reported to say, “It ain’t what you don’t know that gets you into trouble. It’s what you know for sure that just ain’t so. ”
Little late to the convo. Thanks Brett for the post.
These block times, especially in PHX, are very bizarre. AA is scheduling it’s flights anywhere from 20 to 40 minutes longer than it’s competitors, even on short to medium length flights.
Is there taxiway, runway, apron or terminal construction projects planned? Is there a concourse that’s closed?
On-time or early arrivals are good, but arrivals more than 20 minutes early can be problematic from an operational standpoint.
Seems like a very inefficient way to operate.
@cranky : ahhhh i see you’re a fellow enjoyer of color brewer’s palettes too.
dVery late to the post and I am not sure how many people will read this.
I don’t have the knowledge or capability, but how does the on time departure and arrival times compare to the increase block times? I.E. MIA-JFK, are the actual flight times consistently shorter than the block times? Is there any rule that says the Airlines must present their schedules accurately?
I used to fly PAL MNL-JFK before Covid-19 and we would always leave late, but seemed we would always arrive early anywhere by 30 minutes to on time. The turnaround in JFK did seem to take long and when I inquired, I got the excuse of customs needing to inspect the plane before cleaning could be done.