Learning Center
Plans & pricing Sign in
Sign Out



									ATO-P, Portals
800 Independence Ave, SW
Washington, DC 20591
Phone: 202-385-7308

Room 6401, Docket OST-1998-4043
Office of Aviation Analysis
US Department of Transportation
400 7th St, SW
Washington, DC 20590

20 July 2005

Dear Sir or Madam:

The director of our office asked me to submit my recent experience with the Origin and Destination
Survey of Airline Passenger Traffic Database (DB1B) as comments to Docket OST-1998-4043. I
will describe my recent use of the database to illustrate its current shortcomings and how an
improvement would improve our work.

The Federal Aviation Administration created the Air Traffic Organization in February 2004 to
improve the performance of Air Traffic Control services. In early 2005, our office asked me to
take a more detailed look at their revenues and costs for the upcoming Business Outlook.

Specifically, we wanted to determine how much aviation trust fund revenue was generated on a
per-flight basis and on a service-delivery-point basis. In other words, could we determine the total
taxes paid into the Aviation Trust Fund from any given flight 123 by Airline X? And then, how
much revenue is generated by a particular sector or terminal airspace, based on an apportionment of
the tax revenue from all flights that use the airspace?

We could then compare the unit revenues with the unit costs to determine which operations resulted
in a net gain or loss to the trust fund. We could aggregate these results by aircraft type,
geographical region, and other categories to identify risks in the current tax structure. This analysis
would be used to justify our operating budget or to support changes in the tax structure at the
upcoming trust fund reauthorization. Furthermore, the SDP revenue-cost comparison could be used
to prioritize capacity enhancement and other infrastructure improvements.

The O&D survey is our only source of disaggregated fare prices--from which we could determine
the taxes that individual passengers paid. We apportioned the total itinerary fare among its
component segments. Unfortunately, the O&D survey has no time or date fields, flight numbers, or
tail numbers. Therefore, we were unable to match the passengers onto the actual segments in the
T-100 segment data.

As a compromise, we determined the average fare for each origin-destination segment and air
carrier. The consequence of this assumption is that it assigned equal fare to every person on every
flight for a particular carrier and OD pair regardless of aircraft type. This may have tended to
understate revenues generated by larger aircraft types relative to smaller aircraft types.

If we had more information in the DB1B, we would be able to more accurately attribute revenue by
flight segment and by SDP. Furthermore, if it were a 100 percent survey rather than a 10 percent
sample, this would make our data completely definitive.

Our office therefore recommends the following improvements to the O&D survey:

            Move from a 10% sample to a 100% survey
            Include time fields--(dates and scheduled or actual segment flight time)
            Include aircraft tail number
            Include aircraft type
            Include flight number

Note that the last 4 items are included in the T-100 segment data. Another alternative is to provide
a unique identifier field in every record in the T-100 segment data, and match that with the same
identifier in the O&D survey.

An improvement of the O&D survey, as recommended above, would improve the ability of the Air
Traffic Organization to observe its own performance. And, in turn, it will allow the Air Traffic
Organization to provide improved services to its customers and owners—the air carriers and the
flying public.


Tony Dziepak
Business Planning & Development Office,
Operations Planning Services,
Air Traffic Organization,
Federal Aviation Administration

To top