Airline on-time performance is a game of expectations, not absolutes. (Image: Jim Glab)
When you’re planning a trip, do you ever rely on airline on-time performance statistics to choose one carrier over another? Maybe you shouldn’t.
Consider how on-time performance is defined: A flight is considered on time if it arrives within 15 minutes of its scheduled arrival time. But here’s the catch: Who determines just how long a given flight is supposed to take? The airlines do.
We’ve all seen stories in the past about how airlines that show a lot of late arrivals relative to their competitors can resort to a very simple solution — just adjust the schedule so that the flight can take longer to get there but still be on time. This is known as schedule padding.
And now an interesting analysis by Travel Weekly sheds more light on the inaccuracy of on-time statistics. It notes that in setting their expected gate-to-gate travel times, different airlines have different priorities that can affect those estimates. So even if two airlines achieve the very same travel time, one flight could be on time while the other could be considered late.
-Travel Weekly
Using data from FlightStats, the publication noted, for example, that on the Atlanta-Chicago O’Hare route, American Airlines and Spirit Airlines flights both managed exactly the same average gate-to-gate time of 129 minutes. And yet American showed a 78 percent on-time performance rate, while Spirit’s was only 40 percent.
Recent: The #1 Boeing 787 Dreamliner hub in the U.S.
Low-cost carriers face more pressures to keep estimated flight times short, the article noted; for one thing, their labor costs are based on estimated rather than actual travel times. For another, scheduling shorter travel times lets them squeeze in more flights per aircraft.
Just one more reason why on-time performance numbers can be misleading at best.
Do you consider on-time performance stats when making airline choices? Please leave your comments below.
NOTE: Be sure to click here to see all recent TravelSkills posts about: New Two brand new United Clubs + Jennifer Aniston needs a shower + Best Megahub? + Big Hilton/SkyMiles bonus
Do you follow us on Twitter? It’s a great way to keep up with the latest news!
Please join the 100,000+ people who read TravelSkills every month! Sign up here for one email-per-day updates!
I’d rather see stats for on time departures. This would be so much more indicative of how much an airline has their act together.
To carry that thought on further, in ATL the rule of thumb is into and out of ATL by 2:00 from April to mid September. I sometimes look at the airports airlines have as hubs. UA has the worst delay hubs to me. SFO, DEN, ORD, EWR are nightmares. I guess the finally ditched JFK to lessen the beating.
Schedule padding to help on-time performance is better than just being late, though. It helps prevent you from missing a connection, since minimum connection times are based on the scheduled arrival (if both flights actually take two hours, the one that’s been padded to 2 hours will make sure you make the connection rather than the one that didn’t pad and thought it would be there in 1.5 hours). It also helps customers plan their schedule better. I’d rather be 15 minutes early to a meeting than 15 minutes late, right? If a plane tells me it’s arriving at 2:15 and it arrives by then, that’s way more useful than one that departs at the same time, claims to arrive at 2, but doesn’t arrive to 2:15. It should also help prevent delays on later flights, because the airlines that pad the schedule aren’t trying to force an unreasonably short turn around (hopefully).
Basically, the flights may take the same amount of time, but I’d much rather use the airline that accurately predicts how long the flight would take than one that thinks it’s planes are 10% faster than they are. I think accurate flight time estimation is a useful metric for being reliable, it means the airline isn’t trying to force things unrealistically. Although yeah, not as good as an airline that can consistently actually hit a faster goal than all the other airlines.
Excellent! Thanks, Eric! Leave it to Nate Silver to crunch the numbers. Good on Virgin America 🙂
Related, see: http://fivethirtyeight.com/datalab/virgin-america-is-the-michael-jordan-of-airlines/
Thanks, Jedi! Unfortunately, government stats are the only source we have here… and like all stats, there are flaws and easy manipulations. I think the most reliable source is our “gut”… and the experience we have over time with a specific carrier at a specific airport. My gut tells me that Delta is tops in reliability, which is reflected in the current stats… but that could always change!
This is an area pre-merger US always did a good job with. They kept block times as short as possible, yet still managed to usually be near the top of on time arrival ratings. Also, these things come and go through the years. In 13 yrs at United, I saw them add block time three times and cut it three times. It’s really dependent on what the theme of the week is.
Thanks, Bill. I agree… ontime performance seems more a function of airport than airline to me
This is another reason why the on-time performance ranking of UA relative to AA and DL always seems so poor comparatively…even though my experiences with UA have always been reasonably good and not much different from AA or DL. The biggest difference is when I fly through EWR or ORD, but UA delays there are matched by those with AA at ORD and AA and DL at LGA or JFK, respectively. SFO has fog delays, so that hampers UA being their hub. Otherwise, I find them all comparable despite the “rankings” that suggest otherwise.
Yet another dumb article. Chris, what’s going over there? If an airline cares about its customers, it will pad the schedule so that everyone makes their connections and arrives on time for their meetings. If an airline is more interested in profit, they chose not to pad their schedules so that they can make more money off of the aircraft, at the expense of missed connections and missed meetings for their customers. Which one would you rather fly?