Solving the Mysterious Case of United’s High Denied Boarding Numbers

United has been in the news lately for a lot of reasons, many of them bad. But one thing that the airline has been rightfully crowing about is its dramatically reduced denied boarding numbers over the last few months. So imagine my surprise when I saw that, in fact, United’s most recent second quarter numbers filed with the Department of Transportation (DOT) were actually HIGHER than last year. An eagle-eyed reader brought this to my attention, and I was able to get to the bottom of it.

Back on April 9, United had its infamous incident where Dr. Dao was dragged off a plane against his will. Even though it really had nothing to do with the flight being oversold (it was due to a crew needing seats at the last minute), in the public eye, overselling was the culprit. United came out with a 193,429 point plan to show how seriously it took the incident. One of those points was that it would reduce overbooking and pay volunteers more. The hope was that this would result in very few involuntary denied boardings, the bad kind of bump where there aren’t enough volunteers and people have to be kicked off.

According to United, it had been making great progress. In May, it said involuntary denied boardings (IDBs) were down 79 percent from the previous year. And United spokesperson Megan McCarthy told me that in June IDBs declined 88 percent from the year before. Naturally, United was spreading this news far and wide. It was losing revenue opportunities by reducing overbooking, so hopefully it would at least get the PR benefit.

When the DOT report came out, however, it told a different story. In the second quarter of 2016, United involuntarily denied boarding to 871 people, a rate of 0.40 per 10,000 passengers. But in the second quarter of 2017, that number went up to 1,064, and the rate increased to 0.44 per 10,000 passengers. Or for the visually-inclined:

So, uh, how is that possible?

The DOT numbers are just for domestic flights, so maybe there was a huge reduction internationally to offset a domestic increase? Nah, that’s pretty unlikely. Domestic flights tend to see higher levels of bumping anyway just because there’s more frequency to allow for reaccommodating passengers.

DOT numbers also only include mainline whereas United was talking about everything, including regional operations. That sounded like a more plausible reason, but the numbers still didn’t add up. United flew 24.4 million passengers in the second quarter while regionals under the United Express banner flew 13.8 million. The only regional carriers that report to the DOT are SkyWest and ExpressJet, and both saw big decreases year over year. But SkyWest and ExpressJet fly for a lot of airlines, and DOT doesn’t break it out by partner. These certainly weren’t enough to offset the mainline numbers. Maybe there was a culprit that didn’t report. Could Mesa or Colgan have bumped a ton of people last year and fixed it this year? In theory, but that still seemed pretty suspect. (Don’t even get me started on how obnoxious it is that DOT still reports numbers by operating airline.)

Ultimately, all of these theories had a glimmer of potential, but it still wouldn’t have smelled right. United has changed its overbooking policies and it is offering people more money to volunteer. It just seems strange that it would have seen an increase in year-over-year IDBs on the mainline with all the work it had been doing. So that left me with one more thought. United had said how great May and June were, but could it be that April was just really terrible?

Yep.

After getting all wound up and thinking of these complex theories that would explain the problem, it turned out to be that the simplest explanation was the right one. Megan at United confirmed it was a downright awful month in April. She even gave me numbers.

Remember how I said that last year mainline United had 871 domestic IDBs in the second quarter per the DOT report? Yeah, well Megan told me that in April alone this year, United had a whopping 957 IDBs (still using mainline domestic to keep it consistent). That’s more than 1 IDB per 10,000 passengers, far worse than any other airline except for Spirit. (And when you’re in Spirit territory, you know you’ve done something wrong.)

By May, the number had plummeted down to a more than respectable 61 IDBs. June saw the number drop even further to 46. So yes, United really has cut its IDBs significantly. It’s just that April was one of the more awful months it’s had.

I was satisfied to finally have an answer, but I still can’t figure out why that would have been. Megan gave the usual excuses of bad weather and a high load factor, and indeed, it was a high load factor month. United posted an 87.8 percent mainline domestic load factor in April. But May was 86.7 percent, June was 88.1 percent, and July was 87.9 percent. This isn’t abnormal.

April’s operation wasn’t bad either. In all, 81 percent of flights arrived within 14 minutes of schedule with a completion factor of 98.6 percent. Those numbers were better than March and about the same as May.

My guess? Remember, Scott Kirby detailed changes to the revenue management system when he arrived. I’m sure there’s been strategic tweaking as well as they continue to work on improving that system. It could be that they just got too aggressive and ran into trouble. Or then again, maybe I should stop theorizing and again just accept the simplest answer. Even a bad month is still less than 1,000 people out of millions of travelers. A few messed up flights have the ability to sway it one way or the other.

Either way, now we know why the numbers look the way they do.