The main data set was obtained from the Bureau of Transportation Statistics. A subset of the T-100 Segment dataset is used. Departures, Total Capacity, and Total Passengers are taken from the DepPerformed, Seats, and Passengers columns respectively. The Load Factor is calculated as the Total Passengers divided by the Total Capacity.

In many cases mainline carriers (Delta, United, American, etc.) will have a regional carrier operate a particular route under the Mainline carrier sub-banner (such as Delta Connections, United Express, or American Eagle). The T-100 Segment table shows data at the carrier level and does not note if the carrier was operating the flight for a mainline carrier. I instead use a data from the DB1B Origin and Demand survey coupon data which has a 10% survey of all domestic tickets. This data contains information about the both the operating carrier for a coupon (a single flight portion of a ticket) and the ticketing carrier. So if someone has a United ticket on a flight that ExpressJet is operating, we can map how many ExpressJet fliers on a route are on a United branded flight.

Here's a simple example. United may operate 2 flights per day on mainline aircraft between STL and ORD. But 4 more flights per day might be operated ExpressJet Airlines under the United Express banner. When selecting "Break out by regionals" on the STL-ORD route, both United and ExpressJet would be considered separately, but if Roll up to Mainline were selected, all the statistics would be rolled up to United.

Because the mapping uses a survey of 10% of all tickets and the ticketing carrier is not guaranteed to be the mainline brand of the flight, these rollups will not be 100% perfect, but will give a very good approximation of the mainline carriers on these routes. If you know of a better way to source this regional to mainline mapping or notice any mistakes please let me know!

Matthew Lawder (mtlawder@gmail.com)