FinanceBuzz presents some interesting data looking at airline schedules versus taxi time and time in the air to come up with how much, on average, each airline “pads their schedule” – schedules flights for longer than it actually takes to travel from push buck to arrival. Their headline result is that “Southwest Airlines is the worst offender for artificially inflating its flight times” but this is highly misleading.
First, here’s the data:
| Airline | Time in air | Actual time | Scheduled time | Padding time | % padding |
| Southwest Airlines | 62 | 79 | 90 | 11 | 13.9% |
| Alaska Airlines | 90 | 115 | 128 | 13 | 11.3% |
| United Airlines | 102 | 128 | 141 | 13 | 10.2% |
| American Airlines | 99 | 124 | 136 | 12 | 9.7% |
| Delta Airlines | 84 | 106 | 116 | 10 | 9.4% |
| Jetblue Airlines | 71 | 93 | 101 | 8 | 8.6% |
| Spirit Airlines | 111 | 136 | 147 | 11 | 8.1% |
| Frontier Airlines | 89 | 117 | 125 | 8 | 6.8% |
| Allegiant Air | 83.5 | 109.5 | 116.5 | 7 | 6.4% |
| Hawaiian Airlines | 26 | 43 | 45 | 2 | 4.7% |
Southwest’s greater percentage of time they say is padded doesn’t mean they have the most padding per flight. The data above clearly suggests they do not. Instead, it’s an artifact of Southwest having shorter stage lengths than the other data they’re comparing to. Southwest pads each flight less than Alaska, United, and American, according to FinanceBuzz’s data. It’s just that their flights are, on average, shorter.
Southwest also has shorter turn times. They probably need to ‘pad’ flight times a bit more. Remember they board 30 minutes prior to departure versus 35 or 40 for other major carriers.
And what matters most to customers is an airline doing what it says it’s going to do – boarding when they say, and arriving when they say, not the absolute number of minutes it takes. Schedule padding is actually good when it allows an airline to deliver on its promised schedule. Without it, a single delay for an aircraft early in the day can’t be made up for throughout the day as easily and delays stack. The only other way to do it is more scheduled ground for an aircraft, but that also means suboptimal schedules.
What’s interesting is how schedule padding has changed over time, and the claim here is that padding has increased from 8 minutes on average in 2012 to 11 minutes in 2022. A three minute variance in 10 years isn’t alarming on its own, though it sounds like a much bigger deal when it’s trumpeted as 27%!.
Schedule padding increases, though, point to less of an airline issue and more to facilities and air traffic control – the throughput of the system. There’s greater variance in how long it will take an aircraft to go from one place to another because of bottlenecks in the system, and airlines schedule to account for this. That variance is not a good thing. It is costly to airlines. They wouldn’t do it if they didn’t believe they needed to.
The way to address this is to improve throughput.
- We need greater airport capacity (gates, taxiways, runways). The major reasons we don’t have this are (1) airlines don’t want to pay, they want taxpayers to pay, and (2) the U.S. has become bad at doing public infrastructure projects, with too many veto points along the way guided by the National Environmental Policy Act.
- We need better air traffic control. The FAA’s ATO doesn’t do a good job with IT projects, only just retiring paper flight strips. We need better management and better funding mechanisms, so that long-term capital expenditures aren’t subject to the vagaries of annual congressional budget cycles.
Faster travel is better for airlines, it means better fleet utilization and lower costs. It’s better for passengers too. But railing on the airlines for adjusting to the reality of bottlenecks in a sclerotic system isn’t going to get us there.
Some of it just comes down to hub congestion. If you’re a short haul Delta flyer in the Southeast, you’re probably very well acquainted with regular 10-15 minute ‘hold at origin’ orders from ATC because things are starting to back up at ATL. And Delta has long baked that into scheduling since it’s not like ATL is in a position to add more runway space at this point.
I’m perfectly fine with this. Allows a little leeway and always better IMHO to get there early instead of late. I could care less about the padding since I book based on the schedule and it really doesn’t matter to me if an extra 15-20 minutes are added.
Frankly not sure why anyone would view this as a negative.
I’m surprised that it’s so few minutes on average. But when considering it’s an average of good and bad times, it makes sense. Not least as in good times when leaving “on time”, you can often arrive 20+ minutes “early”. And in even in slightly bad, you’ll be very lucky to arrive “on time”.
I have always wondered about this. Flying between Dallas and Seattle uniformly takes 20-30 ,minutes less than the schedules of both American and Alaska call for. Consequently it is not at all unusual to depart as much as 15 minutes late and still arrive 10 minutes or more early! Thanks for the informative post.
I recently flew a segment on SWA from DIA to ABQ. I swear the longest part of the flight was the taxi time from DIA’s gates to the actual runway to take off! I think SWA could have done a drink service in the time it took! The flight itself, in the air, seemed “right up, and right down.” LOL
Although schedule padding is now a fact of life I have a few questions…
1. Although nobody mentions this, schedule padding helps the airlines on time performance since reports to the DOT are based on actual vs scheduled arrivals. So when you see that airline x is on time for a specific percentage of their flights one has to remember that the on time percentage is also padded a bit.
2. Most airline crew contracts have (or maybe had) a provision called ‘schedule or actual’. This means that the operating crew is paid the greater of the scheduled flight time vs the actual flight time. So if the schedule is padded by 11 minutes that means the crew, on average, is paid for 11 minutes more then what it actually took to get there. Multiplied by thousands of flights what does that do to the ticket price?
3. Finally what does schedule padding due to the gate planners at the receiving airport? If you show up 11 minutes early how much time do you spend in the penalty box, as your gate is occupied, and then when you do get to the gate you are still arriving ‘on-time’
Picard
So, Gary, in short you are saying that the problem is really not a bad one. And to fix it, we mostly just need to print more money. As one of your readers aptly stated, he knows what the schedule is when he books the flight. All his plans are based upon that. With so many problems in the world are dire, this is not one of them.
What I love is when a pilot says “and we got you here 9 minutes early” as he pats himself on the back. I’m thinking great, you still owe me 32 days 21 hours and 13 minutes before we are square.
I make my plans based on the arrival time, and apppreciate the just in case time. Better to have a few extra minutes to make a connection or just be early. And as we say in manufacturing tell me what you grade me by and I the decide what to measure. It’s a natural reaction to add buffer I order to maximize the on time arrival metric
On the other hand, arriving 5 minutes early and waiting 20
Minutes for a gate “gets my goat”
“Multiplied by thousands of flights what does that do to the ticket price?”
Basically nothing. A few extra minutes per flight in padding (let’s say 10) X the number of flight attendants (let’s say 5) X incremental labor cost per minute (let’s say $5 although I doubt FA’s are making $120/hour incl benefits) = $250. Divide that by the number of people on a 737 (roughly 180 or so) and you get… an additional $1.39 or so on the ticket price to pay for all the padding.
Nobody is going to notice a price increase of $1.39.
Agree that the FinanceBuzz analysis is misleading and shallow. The variability is key, as Gary mentioned. If the authors can tell the airlines exactly what days the jetstream is going to blow harder, or when thunderstorms or blizzards are going to roll through country, six months ahead of time – then the airlines could run a pretty tight schedule. Transcontinental actual flight times can vary by an hour or more depending on the winds – how do you keep your airline ontime without being conservative and planning towards the higher end of the distribution?
Jean Luc – on the gating problem, airline ground managers aren’t stupid. If the airplane showed up 11 minutes early consistently every day, then planning would be easy, and you’d have a lot fewer waits for a gate. The problems happen on the occasions when the incoming flight has favorable winds or ATC flow and the flight time is on the lower end of the distribution. A more consistent ATC would help, but the variabilities of wind and weather will always make perfect planning elusive.