The Los Angeles Times asks, “Are airlines padding flight times to improve on-time performance?” (HT: Paul H.)
About a decade ago, Joe Nolan, a semi-retired electrical engineer from Palm Desert, could expect to hop on a flight at Palm Springs International Airport and arrive in San Francisco 55 minutes later.
Now the flight is usually scheduled for about 90 minutes. Nolan suspects that airlines are allotting more time for each flight to make it easier to meet their arrival schedule.
“It tells me that the on-time statistics are worthless,” he said.
This gets it exactly backwards.
The LA Times cites a study suggesting “On average, the allotted time for flights between Los Angeles and San Francisco increased 8% from 1996 to 2015, the study found.”
Airlines adjust their schedules to reflect how long it will take to fly from one airport to another. And they absolutely hate it. Just a couple of extra minutes per flight means several more aircraft are needed to operate the same schedule for a large domestic carrier.
- If it were about marketing, airlines would show shorter flight times and sacrifice on-time performance. Many booking systems will display the quickest itineraries first. Customers are more likely to choose the shortest trip, while few dive into the on-time performance of a single flight.
- Longer flight times limit the ability of airlines to get the most out of each aircraft. Airlines don’t overschedule how long it takes to fly, sandbagging means higher labor costs for a flight. It means less efficient utilization of aircraft. It also means that shorter trips are less competitive by air compared to driving or other modes of transportation.
On-time statistics matter, but only in the aggregate. Airlines don’t want to be the worst because of negative attention that grabs, but that matters only at certain margins. Very poor on-time performance may cost bookings — corporate contracts will tend to gravitate towards better performing carriers who also offer strong price concessions. (This could change in the future if Delta’s performance started slipping markedly, gaming the statistics could help them avoid payouts to their corporate customers.)
On the other hand airlines want to achieve strong performance because poor performance is costly. But artificially good performance would still be costly. It means crew overscheduling. It also means paying out performance bonuses to employees based only the airline’s own marketing rather than actual performance.
So then why does every flight LAX/OAK I take weekly arrive 10-15 minutes early?? Same with Vegas, NYC, etc.
Gary,
I’m not bashing you for the post, I promise. But as an industry insider, having this topic discussed in the mainstream press drives me crazy. I don’t know why they do it, or what they hope to accomplish.
So what if they airlines pad their flight schedules? As it is, 85% on time arrivals (and that’s within 15 minutes of scheduled arrival time, mind you) is considered good. So even with the “padding”, they’re in the aggregate not running a perfectly on time operation. If they didn’t pad, they’d run late more often, and with it, more misconnects. Is that the outcome the press is seeking?
Take that LA to San Francisco flight, whose shortest published block time on some random date in December is 75 minutes. If that block time has increased by 10% over the years, that means 6.8 minutes have been added to the flight time. Is that really newsworthy?
Besides, the increase in “flight time” (actually block time depending on what stats are being analyzed) can easily be attributed to scheduling practices on the ground. Take IAD for instances, where pushing out at 5pm in the middle of UA’s bank is going to cost 20 minutes in waiting time. Do it at 3pm, and you have no wait.
Dan gets at another important point: All airlines bank their departures from hubs because it’s convenient for connecting passengers, but that very practice contributes directly to longer scheduled flight times. United has about 35 departures between 9-9:15am on weekday mornings out of SFO. But we can only depart at a rate of one plane per minute (since arrivals land on intersecting runways). Merely spreading out the departure bank would alleviate the departure delays (and resulting padded schedules). Even with parallel runways (LAX, ORD, ATL), we can only push so much tin, since only one plane is allowed on a runway at a time. And since there is little appetite to invest in new runway construction, that means airlines need to get smarter and more creative about scheduling.
Increasing scheduled flight times is far more operational, but still partly marketing.
Take JFK-LHR and vice-versa as a good example. What’s changed over the years. First, taxiways are much more congested at both airports so it takes longer to get to the runway. Second, the queues to take off, particularly at LHR, are much longer than they used to be – it’s not uncommon to wait 15-20 minutes to take off. Likewise for landing. Third, airlines are often using slower aircraft than they used to. This was the province of the 747 – nowadays it’s not uncommon to see a 757 which will take 10 minutes longer to do the flight. These are all operational reasons why the block time will have increased by 30 minutes or so.
But then there are marketing reasons. Many companies will not pay for Business Class for flights less than 8 hours. When that became common, it was extraordinary how this flight started taking 8hr05 rather than 7hr55.
So, it’s a combination of the two. But I don’t know who would choose a flight merely because it’s blocked at 5 minutes less than another. They’ll choose on price, on flight times, on airline loyalty but not on a 5 minute shorter flight which anyone with any sense will know is illusory.
@Greg z
Do you really care that much? I bet you don’t.
@NB
In some sense, it goes “the other way” too. For longer flights, augmented crews are required. I don’t recall the regs, at but at a certain point, one relief pilot is required, and at another, two are required. I heard at one point one union was going at it with the scheduling department because flights were blocked at 7h55m and routinely taking over 8h. The thing is, an augmented crew is required at 8h, and the airline was trying to shortcut it.
In this business, it’s near impossible to be “on time” for every flight. The choice is “early” or “late”. As a consumer, I’d rather be early than miss my connection and have hours to kill as opposed to minutes to kill.
I cared four weeks ago as I was landing in Guam from HNL late…saw this UAL 737 taxiing towards the runway and said, “Gee I hope that isn’t my connector.” It was. Then the ground staff told me (as they were handing me hotel vouchers for the night since we were now stuck on Guam) that this one leg blows its arrival time about 30-40% of the time which, with a 45 min layover to fly on to Mania or Koror, means people get stranded quite a bit since Chicago never holds the planes.
30-40% blown arrival time seems high to me. I did that exact same leg two years ago without trouble. But if that number has any basis on reality, I find it unconscionable that UAL hasn’t adjusted the schedule to accommodate a route it apparently can’t maintain a reasonable on-time percentage on.
@Gary You prove the point you are trying to refute, namely airlines do in fact pad their schedules to improve their “on-time” performance. Period. End of story.
Gary, you admit as much when you write “Airlines adjust [pad] their schedules to reflect how long it will take to fly from one airport to another.” In other words scheduling to account for expected delays for airport congestion, ATC delays, and other delays factored in regardless of the actual flying time to cover the necessary distance. If airlines based their schedules on flying time without factoring in these delays, on time performance would suffer and other serious problems would result.
Correct me if I’m wrong, it just seems you disagree with the reason that schedules are adjusted (padded). For passengers, the reason(s) are irrelevant. Collectively, we spend hundreds of millions of hours of our valuable time each year on airplanes and in airports that we shouldn’t have to spend if the ATC system and airlines operated efficiently. Surely someone has calculated the economic impact of this wasted time.
No doubt improvements in ATC cost money that the airlines do not want to spend and pass on to their customers. Yet as you aptly point out, reducing delays will have a huge positive economic impact on the airlines from better utilization of aircraft and labor. Those savings and more time we could use for other purposes could be passed on to customers, too.
A route I fly often, CLT-ATL, is a good example of padded scheduling. I always shake my head at the fact that DL schedules this 34-minute flight for 85 or 90 minutes. From the perspective of passengers who value their time, “on-time” statistics are worthless. Your thoughts?
@john airlines adjust schedules largely to reflect how long it will take to travel from A to B, which is different than sandbagging to hit on time numbers.
Gary,
I think you list some very rational reasons why airlines would not want to stretch their schedules.
But I agree with other commentators that flights are being scheduled for longer than they were previously.
Since it costs them more, I’d assume they are not doing it to pad their on-time arrival stats. Rather, it’s because flights can often take longer, due to all the reasons listed above- slower planes, more congested airports, etc.
So why do flights sometimes take shorter amounts of time than scheduled? Faster boarding, less congestion than budgeted, tail winds, etc.
I think planes can go supersonic, but they just put speed limiters like they do on school busses.
— asdfasfdad
“Thought Leader in Blog Comments”