Dear Politicians: Steal This Idea

The airline industry might be a favourite whipping boy for all kinds of reasons, sometimes for understandable ones, but now they’re even taking a beating for getting people to their destinations on-time.

A Nov. 29 Los Angeles Times report cites the example of a semi-retired electrical engineer who used to be able to fly from Palm Springs, Calif. to San Francisco in 55 minutes. Today, that same trip is scheduled to take 90 minutes, leading to accusations that airlines “pad” their schedules to improve their on-time performance.

“It tells me that the on-time statistics are worthless,” the Times reports Joe Nolan, who lives in Palm Desert, a Palm Springs suburb, as saying.

The writer concedes that Nolan “might have a point”. Indeed, he does; but in a different way than one might expect.

Two commonly used measures of airline on-time performance are the average delay, and the percentage of flights arriving within 15 minutes of the scheduled time.

In September, the most recent available month, the U.S. Bureau of Transportation Statistics’s numbers for the two U.S. carriers serving Winnipeg looked pretty good: the average United flight system-wide arrived two minutes early, and the average Delta flight pulled up to the gate four minutes ahead of schedule. No such information was available for Air Canada and WestJet on this side of the border.

According to, which shares each airline’s on-time performance according to the 15-minute rule, all four carriers saw more than 80 percent of their flights arrive on time during the two months from Sept. 15 to Nov. 15.

But what use is that to you, the consumer? Not much; in fact, it might give you a false sense of security.

As anyone routinely traveling from Winnipeg will know, it’s often necessary to make at least one flight connection to reach the desired destination. But when the consumer books a flight, they’re not assuming a flight will arrive at the gate at the scheduled time with fifty-fifty-ish probability — they’re assuming something closer to 100-percent probability.

After all, if the average flight arrives early, and the vast majority of flights arrive within 15 minutes of the scheduled arrival time, isn’t it just going to be alright?

This week, I looked up the Nov. 24-Dec. 1 stats for 40 randomly chosen aircraft — 10 each flying in Air Canada, WestJet, Delta and United colours — and tracked their on-time performance during that period. This included 235 Air Canada-branded flights, 309 WestJet-branded flights, 259 Delta-branded flights, and 241 United-branded flights. (Since data reports landing times, I assumed an average eight-minute taxi-in time, consistent with U.S. Bureau of Transportation Statistics information.)

The average estimated deviation from schedule was, at worst, insignificant: three minutes late on Air Canada, four minutes early on United, five minutes early on WestJet, and 12 minutes early on Delta — which is not unusual for that airline, being the most conservative scheduler over the past couple of years.

Most of each airline’s branded flights were also on-time according to the 15-minute rule: 74 percent for Air Canada, 81 percent for WestJet, 60 percent for United and 61 percent for Delta. (The busy U.S. Thanksgiving holiday, which was expected to be more of a problem than usual this year due to tighter security, likely accounts for much of the difference.)

But if one were to look at the cut-off separating the worst five percent of each airline’s flights from the other 95 percent, that is where one starts to get a sense of how much of a delay should be assumed if the goal is a leisurely walk through the terminal from the arriving flight to the departure gate, and not a panicked sprint.

In this case, WestJet was the best performer, with 95 percent of flights arriving no more than 22 minutes late.

Delta and Air Canada were essentially tied, at 35 and 37 minutes respectively. United was the straggler, with the best 95 percent of flights arriving no more than 45 minutes late.

That is information that consumers can actually use, because those are the margins they should build into their schedules if they want to avoid a high-stress situation.

Given that governments like to play “the consumer’s best friend” when it comes to air travel, they could actually do something useful for a change by providing the public with not just average delays, but with an idea of what to expect when their flight turns out to be anything but average in a negative way.

Publicizing information on the cut-off between the “best 95 percent” and “worst 5 percent” of flights would not only save travellers the stress of running through airports, or arriving at the destination and finding their bags didn’t make the connecting flight; it would give a far more vivid illustration of each airline’s network performance than conventional on-time statistics ever could.