Attention passengers: your next flight will likely arrive early. Here?s why

“Every­body is smil­ing,” says Van Mieghem.

But are these smiles due to more than good luck? Van Mieghem, the Stu­art pro­fes­sor of man­age­r­i­al eco­nom­ics and oper­a­tions, and Salant, an asso­ciate pro­fes­sor of man­age­r­i­al eco­nom­ics and deci­sion sci­ences, began to won­der whether air­lines were strate­gi­cal­ly adjust­ing their sched­ules to make ear­ly arrivals more like­ly.

In a new study with Assis­tant Pro­fes­sor Den­nis Zhang of Wash­ing­ton Uni­ver­si­ty in St. Louis, the researchers ana­lyze two decades of data on 43 mil­lion domes­tic U.S. flights. They dis­cov­er that pub­lished flight times — the flight dura­tion that con­sumers see when they shop for plane tick­ets — increased 8.1 per­cent between 1997 and 2017, amount­ing to an addi­tion­al 341 mil­lion pas­sen­ger hours.

But when the researchers break down what caused that increase, they find that planes are not fly­ing slow­er than they used to.

Rather, near­ly half of the addi­tion­al time comes from air­lines strate­gi­cal­ly padding their sched­ules. Late planes are bad for busi­ness — so, by adding time to their pre­dict­ed flight lengths, air­lines can increase the odds that their planes will arrive on time. This sched­ule padding is espe­cial­ly com­mon on flight routes with less com­pe­ti­tion, the researchers find.

In Search of Lost Time

The team looked at his­tor­i­cal data from the Bureau of Trans­porta­tion Sta­tis­tics con­tain­ing detailed data on 43 mil­lion flights over the last 21 years. They ana­lyzed changes in flight dura­tion for the same route oper­at­ed by the same car­ri­er year over year.

On paper, it appeared that the same flights now take sig­nif­i­cant­ly longer than they did in the  ‘90s.

The researchers knew that there were sev­er­al pos­si­ble expla­na­tions for this. Planes may be spend­ing more time cir­cling in the air wait­ing for an open land­ing strip, or more time on the ground wait­ing for a gate to become avail­able. Anoth­er pos­si­bil­i­ty is that air trav­el has become less pre­dictable — if so, air­lines could be adding more buffer time to guard against increas­ing­ly fre­quent delays from bad weath­er or oth­er unplanned events.

A final pos­si­bil­i­ty was  “strate­gic padding”: air­lines might extend their sched­uled flight times not for any logis­ti­cal rea­son, but as a mat­ter of sheer busi­ness strat­e­gy.

?“When flights arrive on time or ahead of sched­ule, you get hap­py cus­tomers,” says Van Mieghem.

How­ev­er, Salant notes,  “there are costs asso­ci­at­ed with that.” For one, if an air­line pads too much, a com­peti­tor can under­cut them by offer­ing a short­er flight. Fur­ther­more, flight crew mem­bers are often paid for the full sched­uled flight dura­tion, regard­less of how long the flight actu­al­ly takes.

The researchers cre­at­ed a math­e­mat­i­cal mod­el to tease out how much each fac­tor — air time, ground time, unpre­dictabil­i­ty, and strate­gic padding — con­tributed to the increase in pub­lished flight times.

They found that planes spend rough­ly the same amount of time in the air as they did 21 years ago. And unpre­dictabil­i­ty did not seem to be play­ing much of a role in the increase, either.

Ground time was a dif­fer­ent sto­ry.  “Pas­sen­gers were spend­ing more time on the plane wait­ing for take-off, or wait­ing for a gate after land­ing, pos­si­bly because of increased air traf­fic,” Salant says.

But that increase in ground time explained about half of the increase in pub­lished flight times, leav­ing more than 150 mil­lion pas­sen­ger hours unac­count­ed for. The researchers con­clud­ed that this remain­ing time was the result of strate­gic padding.

Less Com­pe­ti­tion Means Longer Flight Times

Why has strate­gic padding become more com­mon? The researchers sus­pect­ed that part of the rea­son is dwin­dling com­pe­ti­tion.

?“If you have plen­ty of com­peti­tors, log­ic dic­tates that you will seek to offer cus­tomers the most effi­cient route from A to B. You’ll cut back your sched­uled flight time,” Van Mieghem explains.

But due to bank­rupt­cies and merg­ers, there are few­er large U.S. air­lines today than in 1997. So air­lines today may feel less com­pet­i­tive pres­sure to offer short­er flight times.

To test this the­o­ry, the researchers looked at how pub­lished flight times changed after com­peti­tors either start­ed or stopped offer­ing flights on a par­tic­u­lar route.

As pre­dict­ed, when com­pe­ti­tion grew stiffer, air­lines cut down trav­el time. Con­verse­ly, Van Mieghem says,  “as the play­ing field thins out, less com­pe­ti­tion makes it eas­i­er for air­lines to do the opposite.

A Sil­ver Lin­ing to Sched­ule Padding?

Walk around O’Hare air­port, for exam­ple, and you are like­ly to come across a bill­board adver­tis­ing the  “reli­a­bil­i­ty” of Unit­ed Air­lines’ flights.

?“But it’s not real­ly about reli­a­bil­i­ty,” Van Mieghem says.  “If air­lines real­ly want­ed to improve reli­a­bil­i­ty, they’d focus their efforts on enhanc­ing their oper­a­tions. What our study shows is that this is not the case. What they’re going for instead is the eas­i­er — and cheap­er — option of sched­ule padding.”

Nonethe­less, the researchers stress that sched­ule padding has its ben­e­fits. After all, a flight that usu­al­ly takes 90 min­utes may indeed take 120 on an unlucky day. By pub­lish­ing a time clos­er to two hours, the researchers explain, air­lines are sim­ply erring more on the side of cau­tion.

In fact, sched­ule padding may actu­al­ly help trav­el go more smooth­ly. Pas­sen­gers are not only more like­ly to expe­ri­ence the joy of an on-time arrival, but also get more buffer time to make a con­nect­ing flight.

And the paper’s rev­e­la­tions sug­gest that there is an oppor­tu­ni­ty for pas­sen­gers to be strate­gic, too, says Van Mieghem.

?“We tend to allow a lit­tle extra time for delays or hold-ups when we’re trav­el­ling,” he notes. But if con­sumers know that their flight is like­ly to get in on time, they might save time by reduc­ing that buffer in their sched­ule.  “Of course,” he adds,  “this is a ques­tion of our indi­vid­ual tol­er­ance for risk.”