One thing that’s relatively unique about the US airline industry is that there is good access to data thanks to the government’s filing requirements. This has recently gotten even better as new regulations from the Biden/Buttigieg era have gone into effect. My primary data source, Cirium, has recently loaded the first round of the data which makes me very happy, but it’s also available directly from the government. I thought it might be interesting for you to understand what has changed.

Specifically today, I’m talking about the Origin & Destination Survey which has historically taken a 10 percent sample of all domestic tickets (and some international which is highly restricted, so I never use it in posts) and then provided the fare and passenger detail quarterly. Tools like those offered by Cirium then gross up the data and clean it to present a respectable estimate of the total market. (If you’re doing it from the raw data, be very careful how you use it.) I use this data all the time, though you may have heard me refer to it as DB1B.

As of July 2025, DB1B has become DB1C, and the data is much better. You can read the lengthy and exhausting file in the federal register if you’d like. Cirium also has a summary.

So, let’s go through some of the most important changes.

Sample Size Grows from 10 to 40 Percent

The biggest change is that we are no longer going to have to rely on a small 10 percent sample. Airlines now have to submit 40 percent of tickets to the feds which will make for more accurate reporting, especially in small markets where there just isn’t much data to be filed.

Frequency Goes From Quarterly to Monthly

As far as I’m concerned, the best change is that we will no longer be beholden to quarterly releases. The data will now flow monthly. This is huge since it has required some awfully uncomfortable groupings in the past. Does it make sense that if I want to look at June, I also need to look at April and May? Or if I want to see July, why do I also need August? Now we will be able to better isolate performance by actual seasons.

This also means data will flow more freely. Normally, we have to wait about 80 days beyond the end of a quarter for that data to be released. Now the target has been reduced to 75 days, and it’ll be every month. So July data will show up by mid-October instead of mid-December. DOT hopes to reduce that turnaround time even more. Oh, and for T100 traffic data (no fares), the wait will drop from six months to three months.

Advance Purchase Data

Maybe the most exciting change is the inclusion of how far in advance each ticket was purchased. This isn’t an exact number, but it will be lumped into three buckets: within three weeks of travel, three weeks to 90 days, and beyond 90 days. Obviously more granular would be useful, but getting a sense of when people booked will be a very welcome addition.

No More Making Up Taxes

It used to be that the fare had to be provided, but taxes were not in there. Some models would estimates taxes. Now, the taxes are included so there’s no estimation required.

Reporting Carrier Clarity

In the old system, the airline that “lifted” the first coupon on the ticket at time of travel would be the reporting carrier. This led to some confusion and probably bad data. For example, if United issued a ticket where the first flight was on JetBlue, JetBlue would have reported previously but it wouldn’t surprise me if it got reported twice or incorrectly. Now, the airline that issues the ticket has to report it regardless of the first airline, and that makes it very clear who does what.

The only place where it gets iffy is when a foreign airline that doesn’t report issues the ticket with travel partly on a US carrier. In that case, the first US airline still has to report.

All Airlines Reporting

Before this change, smaller commuters — those with aircraft under 60 seats only — didn’t need to report at all. This didn’t impact regional partners of big airlines, but it did exempt little guys. Now they all have to report.

---

There are a few other minor changes, but as you can see, this is going to be a big upgrade for the data. I’m writing about it now, because Cirium first loaded the data into its tool last week. I can finally see July 2025 all the way through October 2025 data. There are going to be some issues with year-over-year comparisons for a bit, but it’s already exciting to see this stuff.

You didn’t think I’d leave you without any data at all, did you? Here’s a look at some fare data which shows that Southwest really took a big hit after peak summer on its short-haul fares in the LA area. Others did too, but look at how October saw fares creep up.

Average Fare for Trips Under 500 Miles Departing the LA Basin by Month

^{Data via Cirium}

Just some random food for thought. Stay tuned for more.