Reason Foundation

Reason Foundation

Air Traffic Control Reform Newsletter #87

Airline ticket tax broken, increased controller errors, little progress on RNP, fully integrated aviation weather

Robert Poole
October 28, 2011

In this issue:

The Airline Ticket Tax Is Broken—and Needs Replacing

The last few months have witnessed story after story about what many airline passengers have come to view as ridiculous ancillary fees and charges made by airlines. Spirit now charges $5 for those who don’t print out their boarding passes in advance. Numerous airlines charge $150 to $250 if you change from one flight to another—even if doing so uses fills a seat that would otherwise be empty on the earlier flight and frees up a seat they can re-sell on the flight you changed from. Fees for checked bags are nearly universal, giving people incentives to carry on bags they would otherwise have checked.

The U.S. DOT has proposed requiring airlines to report on 19 different categories of ancillary revenue, instead of just the two on which they currently report (baggage fees and change fees), which airlines have denounced as unwarranted and intrusive. At the same time, of course, they joined forces with just about all other aviation groups in lobbying, successfully, to kill the Administration’s proposed $100 per flight ATC fee, which would have been deposited into the Aviation Trust Fund, to help pay for ATC and airport investments.

Yet for all the airlines’ complaints about being the most highly-taxed industry in the United States, nowhere in all that rhetoric is it made clear that, apart from their normal corporate income taxes (in years when they have net income), practically none of these “taxes” goes to the federal government’s general fund. They are all some form of user tax, mostly going into the Aviation Trust Fund, with another small portion going to pay for part of the TSA’s aviation security program. If airlines were not paying for airports and ATC via those user taxes, they’d be paying for them via ATC user fees and various airport fees and charges, as airlines do all over the world.

What really frosts me about ancillary fees, however, is that they enable airlines to make a partial end run around the 7.5% ticket tax—the largest single source of user-fee revenue for the Trust Fund. The ticket tax applies only to the actual fare charged for the flight. By shifting more and more of what used to be included in the fare (meals, headsets, boarding passes, luggage, etc.) to a la carte pricing, airlines have managed to reduce the base against which the 7.5% is calculated, thereby starving the Trust Fund of funds that would otherwise be going into NextGen and airport improvements.

Last week Amadeus and IdeaWorks released their jointly produced Worldwide Estimate of Ancillary Revenue for 2011. Their global total for this year is $32.5 billion, a 43.8% increase over their 2010 figure of $22.6 billion. To estimate how much the U.S. Aviation Trust Fund has lost due to this phenomenon, we first need to parse the Amadeus numbers.  Their subtotal for U.S. Major Airlines is $12.5 billion—but that includes international flights on which the ticket tax is not charged. And a pie chart breaking down that category estimates that 50% comes from the sale of frequent flyer miles (and those are already taxed). “Major US Airlines” does not include low-cost carriers like Southwest or “Ancillary Revenue Champs” like Allegiant and Spirit. For the sake of argument, let’s assume the net US domestic ancillary revenue (exclusive of frequent flyer miles sales) is $7 billion. Were that sum included in the ticket price, the Trust Fund would have had over half a billion dollars more for aviation infrastructure investment this year.

I don’t understand why the airlines think starving aviation infrastructure investment is a good thing. And I especially don’t understand why the major airlines don’t see an unfair competition issue in the very heavy reliance on ancillary revenues by up-and-coming competitors like Allegiant and Spirit. Aviation Daily (Sept. 9, 2011) reports that Spirit claims it can break even on an average fare of $64 per flight segment—because it averages another $43/segment from ancillary revenue.

The United States is the last developed country in the world to use a hodge-podge of excise taxes (like the ticket tax) to pay for its ATC system. Every single one of our OECD colleagues pays for the capital and operating costs of ATC by charging directly for en-route and terminal services. There is zero logical relationship between the price charged for a ticket and the ATC services an airline receives. The major airlines are massively cross-subsidizing the ATC services received by their low-fare and high-ancillary-revenue competitors, even when the majors start playing the same ancillary revenue game.

And this makes a difference. Congress always talks a good line about “fully funding NextGen,” but everyone knows that what the FAA requests is not necessarily what its top management thinks makes the most sense for timely investments in NextGen and keeping the system running smoothly. The budget request is whatever OMB approves, which is dictated by overall budget constraints and Administration policy. And the actual budget is whatever Congress decrees. If that were not the case, we would not witness Deputy Administrator Michael Huerta warning FAA managers that “the planning we have done and the tools that we employ, such as NextGen’s Segment Implementation Plan and the NAS Enterprise Architecture, will assist us in making appropriate decisions about where to take cuts,” should the budget they’ve requested not be approved. (“FAAMA Q&A with Michael P. Huerta,” Managing the Skies, Sept./Oct. 2011)

The Air Traffic Organization cannot plan and manage the ATC system like a business, because it has no control over its revenue stream, which is totally disconnected from the services it provides. Its largest funding source, the ticket tax, is an anachronism. It should be replaced by fees and charges sufficient to pay for the capital and operating costs of the ATC system. And those revenues should not have to pass through the Treasury and the congressional budget process before reaching the ATO.

A Large Increase in Controller Errors?

Earlier this month the Government Accountability Office released a report documenting large increases in reported operational errors by controllers (as well as significant increases in the rate of runway incursions). The big question is whether these increases are the result of improved data reporting or of some underlying problems. (GAO-12-24, available at

My reading of the GAO report leads me to conclude that much of the increase in operational errors is real. Certainly the reported numbers in both cases look alarming. Comparing time periods in 2008 and 2011, the data show a 166% increase in operational errors (OEs) in the TRACON environment, as well as a 53% increase in OEs in the tower environment. Two factors are erroneously believed to account for much of these increases. One is the relatively new voluntary anonymous reporting program called ATSAP, but the FAA correctly points out that incidents reported to ATSAP do not get entered into its ATQA database of OEs (but are used to identify and take action on safety problems). Some overlap between the two does exist, but the GAO reports that only 35% of the incidents reported to ATSAP are separately entered into ATQA by a supervisor or manager.

The other alleged cause of higher reported OE numbers is a new automated error-reporting system for towers and TRACONs called TARP. It automatically detects losses of separation in those portions of the airspace, similar to a system that has been in use for many years at en-route centers. TARP has been deployed at 150 towers and TRACONs (out of 673) thus far, but is not in full-time use at those locations; indeed, GAO reports that it is used as little as two hours per month at some facilities. FAA says it “will fully implement TARP at selected facilities and phase in additional facilities every 30 days until full deployment is reached.” It’s unclear if this means two-hours-per-month at all towers and TRACONs or 24/7 use, as with the en-route system. Incidentally, based on data from the latter, the OE rate in the en-route environment was 13% higher in 2011 than in 2008.

Thus, given that ATSAP data are not included in the official numbers, that TARP is contributing only a small amount of its data potential, and that OEs are trending upward in the more reliably reported en-route environment, there is a real increase in errors. To figure out why, we must look into what else has changed during this time period. The best answer I’ve seen was that provided by the DOT Inspector General’s Office in several recent reports: the large influx of poorly trained new-hire controllers, especially at some of the busiest and most-complex ATC facilities.

Last month the FAA released the report of its Independent Review Panel (IRP) on the Selection, Assignment, and Training of Air Traffic Control Specialists. ( Though it seems to have received no media coverage, it is well worth reading. Based on interview with managers at towers, TRACONs, and centers, the IRP found frustration with “the lack of basic knowledge possessed by the new controllers assigned to their facility.” The FAA Academy in Oklahoma City “does not provide new controller candidates with sufficient instruction in the fundamental air traffic control knowledge and skills necessary to become certified controllers.”  ATC facilities “are struggling because a record number of inadequately prepared Academy graduates are being assigned to their facilities.” The tests they must pass to graduate from the Academy must be pretty perfunctory. Every single candidate who takes the knowledge-based portion passes, and 95% pass the skills-assessment portion. The tests, incidentally, are graded on a pass-fail basis. Consequently, managers at the facilities they are assigned to are left “without insight as to the relative strengths and weaknesses of their incoming controllers”; moreover, the Academy “is left without metrics to monitor the quality of their programs.” Worst of all, Academy graduates are assigned to facilities before Academy training; hence, no information about their relative knowledge and skill levels is used to determine which facility the person is assigned to for on-the-job training (OJT). As a result, 20 to 30% of those in OJT wash out—a considerable waste of FAA money. And many of the busiest and most complex facilities are left with large numbers of poorly trained controller aspirants. No wonder there are more errors than there used to be.

The IRP therefore makes a number of sensible recommendations for a major revamp of how the agency trains and assigns new controllers. They should all be required to take and pass an “advanced” course at the Academy before being assigned to a facility, and that assignment decision should take place at the end of the process, not the outset, and be based on demonstrated abilities and performance at the Academy; no more pass/fail grading. Given the ongoing retirement of the generation of controllers hired after the 1981 strike, the agency has no choice but to hire and train many thousands of new controllers. The current training and assignment process is a fiasco and needs a thorough overhaul, as recommended by the Inspector General and now the IRP.

Why Such Little Progress on RNP Implementation?

One of the key concepts in the paradigm shift from 20th-century air traffic control to 21st-century air traffic management is performance based navigation (PBN). This is the general term for procedures such as area navigation (R-NAV) and required navigation performance (RNP) that rely primarily on GPS and on-board computers (Flight Management Systems—FMS) to fly far more precise tracks in all phases of flight. That’s in stark contrast to the 20th-century model in which planes fly straight-line segments from one ground-based navigation aid to another. PBN offers the potential of significant savings in time, fuel, and emissions, but it’s proving to be a major culture change for air navigation service providers (ANSPs) and their safety regulators.

On a global scale, the International Civil Aviation Organization (ICAO) some years ago set ambitious goals for PBN-based approaches to airports. They were to be deployed to 30% of airports worldwide by 2010 (the actual number is about 15%), 70% by 2014, and 100% by 2016. Meeting those targets should be less difficult than many other aspects of NextGen, because the equipage cost for airliners is modest. In this country, GE Aviation estimates that half the commercial fleet is already capable of flying RNP-AR approaches today—including most of the fleets of Alaska, American, and Southwest and much of Delta and United-Continental. All Airbus and Boeing planes now being delivered are equipped for RNP-0.3 (meaning they can maintain a track within plus or minus 0.3 nautical miles), a typical RNP-AR requirement. Southwest has retrofitted most of its 737 fleet. Between equipment and crew training, the carrier spent $90 million on 370 planes, for an average all-in per-plane cost of $250,000.

But despite its ambitious efforts, Southwest has run up against two problems in making use of RNP. First, time-saving/fuel-saving RNP approaches exist at only 16 of the more than 100 airports it serves. Second, controllers frequently don’t give its equipped planes the clearance needed to fly the RNP approach, on the grounds that it’s too difficult to do when large fractions of an airport’s approaching flights are flying non-RNP instrument approaches. As a result, in the nine months since Southwest completed RNP installation and training, it has used the procedure on only about one percent of all its landings.

That means the ball is squarely in the FAA’s court. Why aren’t there more public-use RNP procedures, such as the curved approaches that save the most time and fuel? And why don’t controllers have the tools to accommodate both RNP and non-RNP approaches? It’s not like the aviation community hasn’t pressed this point with the agency. The RTCA NextGen Advisory Committee (NAC) has made this one of its priorities for action, along with Data Link, as key near-term goals. Its September 29th meeting included recommendations for the FAA to “develop capabilities to allow the large percentage of currently equipped users to routinely perform RNP 0.3 with RF [curved, radius-to-fix] flight leg procedures.” Since about half the airline fleet is already equipped, the NAC estimates the remaining equipage cost at $2.6 billion between now and 2020.

But will the FAA follow through? I reported in this newsletter’s May issue that the Air Traffic Organization’s FY2011 Business Plan included an initiative for the ATO to “work with third-party vendors . . . to complete [the] qualification process for end-to-end procedure development for RNP SAAAR approach procedures, with a due date of Sept. 30, 2011.” At press time, my query to FAA about the status of this had not been answered. And even if the ATO completes this action item, there is still the question of cost. Third-party developers (e.g., Boeing’s Jeppeson and GE Aviation’s Naverus) are not charities. For them to develop large numbers of public-use RNP procedures (as opposed to proprietary ones, paid for and used by a single airline), they must have a customer. Only on rare occasion will a single airline pay the costs for a procedure that all other airlines can use without paying for. So the only plausible customer for public-use procedures is the ATO. But we have yet to see a line item in the FAA-ATO budget for the development of third-party RNP approaches.

It’s fine that Boeing and Alaska Airlines are putting their own money into the Greener Skies RNP pilot program at SEA-TAC in Seattle. The FAA has also put some money (via its NextGen advisory board) into Greener Skies, but it’s been agonizingly slow to develop, with certification expected maybe by 2013, but regular use still a question mark, depending on controller training and other internal FAA factors. Boeing has also received a $3.1 million Greener Skies 2 contract to analyze new procedures for such approaches. But the NAC is trying to send a message that the time for pilot programs is winding down; it’s time for serious implementation.

There is a lot of progress with RNP implementation in other countries, including Australia, Indonesia, New Zealand, and Peru, among others. At Brisbane, Australia’s third largest airport, RNP approaches and departures were developed starting in 2007 in an effort by GE-Naverus, Qantas, and the ANSP, Airservices Australia. Although only 18% of the aircraft were initially fully equipped, the project developed procedures so that controllers could handle a synchronized flow of differently equipped airliners. ICAO offers airports a “PBN Go-Team” which makes an on-site educational visit and offers guidance to the ANSP to develop a three-year PBN blueprint within six months after the visit. First operational implementation is expected one year after finalization of the three-year plan. That kind of can-do approach is sorely needed in this country.
Toward Fully Integrated Aviation Weather Information

One of the underlying principles of Next Gen is to make better use of weather information in flight planning and operations. As a September 2010 report from the Joint Planning & Development Office (“ATM-Weather Integration Plan,” Sept. 24, 2010) put it, “Integration, as used in this plan, refers to the inclusion of weather information into the logic of a decision process or as a decision aid, such that weather constraints are taken into account when the decision is made or recommended. The goal of weather integration is to minimize the need for humans to gauge NAS [national airspace system] weather constraints and to determine the optimum mitigation of these constraints.”

That report explains that weather information is not currently integrated into all ATC decision systems and processes. It also explains the numerous different weather data products, most produced by the National Weather Service (part of the Department of Commerce) but some produced in-house by FAA people. The NextGen paradigm calls for developing two integrated systems called NextGen Weather Processor (NWP) and NextGen Network Enabled Weather (NNEW). A key point, consistent with the new air traffic management paradigm of being able to manage traffic “anywhere from anywhere” is that the totality of near-real-time weather information would be available to all participants (air traffic managers, pilots, dispatchers, etc.) nationwide on a 24/7 basis.

Like other aspects of the NextGen model, these weather concepts conflict with the established ways of doing business in our ground-based ATC system. A mini-scale version of this clash came about in 2005 when the FAA outsourced and consolidated its network of Flight Service Stations. When this set of facilities was originally created, there were more than a hundred, spread around the country, to assist (mostly) private pilots with weather and other pre-flight briefings, flight plan filing, etc. Pilots could call a local number (long-distance calls were expensive, and there was no internet) and talk to a local briefer about weather conditions, any local or nearby airport constraints, etc. This system was very labor-intensive and facility-intensive, and was provided at no charge to private pilots. When the cost began exceeding $500 million a year, even the general aviation organizations got behind the outsourcing and consolidation plan. The program has reduced drastically reduced the number of facilities, down to just three hubs and three satellite facilities. Thanks to the internet and some nifty automation, a briefer anywhere can respond with accurate information to pilots from anywhere. While there are still occasional grumbles from some private pilots, the outsourcing and consolidation is generally considered to be a success, including by the DOT Inspector General.

Perhaps inspired by that consolidation, the FAA has been trying for a number of years to streamline one aspect of the weather services provided to it by the NWS. For decades, the weather agency has provided on-site meteorologists at all 21 en-route centers. According to a GAO report on aviation weather from September 2010 (GAO-10-843), the NWS typically provides three meteorologists and a supervisor at each center, two shifts per day, seven days a week (with none on duty during the 8-hour night shift). As of 2010, this was costing FAA $13 million per year. Starting in 2003, the FAA has attempted to standardize the weather products and services provided at the 21 centers and to reduce the cost. In 2005, it asked NWS to consolidate these services to a smaller number of sites, with fewer meteorologists but operating 24/7. After several years of proposals and counter-proposals, NWS in June 2009 finally submitted a proposal FAA liked: it would consolidate the program to two sites and provide services on a 24/7 basis. But as NWS was finalizing the plan, due in December 2009, the NWS labor union objected to the consolidation, and NWS backed off. Instead, in March 2010 it submitted a new proposal that would not only retain all 21 existing units at the centers but to add new staffing at the Aviation Weather Center to provide consolidated services for the night shift—adding about $3 million in additional costs. An interim settlement provided for continuation of the status quo through Sept. 30, 2011, which date has now passed. As of the date of GAO’s report, the future of the program beyond that date was unknown.

This sad tale suggests that the vision of centralized, integrated, near-real-time NextGen weather may be harder to realize than many expect. About the only good news I’ve seen on the weather front is last week’s announcement of three firms having teamed up to bid for the NWP and NNEW procurements, for which an RFP is expected next spring. They are AirDat, Metron, and Raytheon. I was especially pleased to see AirDat in the picture, since I’ve been impressed with their approach to providing near-real-time aviation weather information. I described this in the lead article in the June 2010 issue of this newsletter. Basically, they use hundreds of airliners as data probes, which produces vastly more, and more-current, data than NWS’s twice a day weather balloons. AirDat’s software uses these data elements to generate real-time maps of icing, turbulence, winds aloft, etc. That sounds to me like the kind of weather information NextGen is going to need.

News Notes

Clarification re “Commercial” ANSPs
In my article last issue about the large majority of CANSO member air navigation service providers that are “commercialized,” I did not precisely define the term. That brought a useful response from CANSO Director General Graham Lake, suggesting a different interpretation of the meaning of the term than what I meant.  Lake prefers the term “corporatized” for CANSO’s Category 1 members, meaning that they 1) are financially independent of government, 2) are able to raise capital from the financial markets, 3) charge for their services, 4) have a monopoly on providing en-route traffic management, 5) have non-civil-service employees. As it happens, that is exactly what I meant by “commercialized.” However, Lake and CANSO would reserve that term for companies like SERCO that are actually private-sector business that must compete for all their business. And he notes that SERCO is one of the world’s largest ANSPs as measured by IFR flight hours.

Benefits from JFK Departure Metering
MIT Lincoln Labs and PASSUR Aerospace last month announced the results of a study they jointly carried out to quantify the benefits of six months worth of departure metering at New York’s JFK International Airport. Compared with the period prior to implementation of departure metering, the program produced annual fuel savings of $10-15 million for the airlines, eliminated 48,000 metric tons of CO2, and reduced departure taxi hours by 14,800 per year. The system holds flights in a virtual queue with engines off until it’s time for them to taxi to the departure runway. Briefing slides are available at

JPDO Revising 2025 NextGen Targets
In the February issue of this newsletter, I noted that according to the FAA’s 2011 Aerospace Forecast, ATC activity will not double by 2025, which was one of the premises for the creation of the Joint Planning & Development Office, which is doing the long-term planning for NextGen. In fact, the 2011 forecast projects an increase in en-route IFR activity of 58% by 2031, and a 42% in TRACON operations by that date. Consequently, I was not surprised by the JPDO’s announcement last month that it has drafted a revised 2025 “end state” for NextGen and is circulating it among ATC working groups.  JPDO Director Karlin Toner says that revising the 2025 end-state will let R&D resources be refocused on the technologies that will need to be in place by that date.

ARINC to Assist with Controllers in Cockpits
FAA’s recently revived Flight Deck Training program, under which controllers will get jump-seat rides in cockpits to observe ATC procedures from the customer’s perspective, will get an assist. ARINC will modify its CASS (Cockpit Access Security System), which airlines use to verify the status of pilots flying jump-seat, to handle controllers as well as pilots.

NASA Developing Large ADS-B Demonstration Project

In FY 2015, NASA and the FAA will operate a large-scale Airspace Technology Demonstration 1, under which airliners equipped with ADS-B/In will be able to use merging-and-spacing tools for coordinated approaches at a high-traffic airport (which is yet to be selected). Previous demonstrations of ADS-B aided approaches have been mostly for a small number of planes making occasional approaches, except for UPS’s nightly operations at Louisville. A subsequent Airspace Technology Demonstration 2 will focus on airport surface operations.

CANSO Up to 64 Full Members
The Civil Air Navigation Services Organization in August welcomed Aeroportos de Mozambique as its 64th full member. AdM operates the country’s principal airports and its air traffic control facilities. Full members are air navigation service providers (ANSPs). CANSO also has 65 associate members, who are suppliers to the industry.

New York TRACON Equipped for ADS-B
Lockheed Martin last month delivered new software that will enable the New York TRACON to use inputs from ADS-B/Out on equipped aircraft to monitor their positions in real time. The software modifies the Common ARTS operating platform to combine surveillance data from radar and ADS-B, displaying it as a single track for each aircraft. Similar upgrades have already been made in the Anchorage, Houston, and Louisville TRACONs (which also use Common ARTS) as well as at Philadelphia TRACON, whose platform is Raytheon’s STARS.

Europe-Wide Controller Licensing Debuts
Effective August 31, 2011 the European Commission’s new Regulation on the Licensing of Air Traffic Controllers became effective. It provides common EU-wide rules on controller licensing and training, aimed at making it easier for controllers to relocate to areas where trained controllers are in short supply. It is the first EU-wide regulation developed by the European Aviation Safety Agency relating to air traffic controllers.

Multilateration Deployed at More Airports

In recent weeks, Saab Sensis has been selected to provide multilateration systems for surveillance at two key airports. In Scotland, the company’s wide angle multilateration (WAM) will replace the aging monopulse secondary surveillance (MSSR) radar at Edinburgh, the country’s busiest airport. The WAM system will provide approach control surveillance of all flights within 60 nautical miles of the airport, giving NATS (the ANSP for the airport) a higher update rate and more-precise surveillance at a lower life-cycle cost. Nav Canada will deploy Saab Sensis multilateration for surface surveillance at Calgary International Airport, providing all-weather surveillance of all vehicles on the airport’s surface. All ground vehicles will be equipped with VeeLo NextGen vehicle locators. In a related development, NATS received the 2011 Air Traffic Control Association Earl F. Ward Memorial Award for deploying a Saab Sensis WAM system in the North Sea, primarily for helicopter surveillance in Scotland’s offshore oil industry.

Quotable Quotes

“These days a $15,000 car comes with GPS satellite navigation fitted as standard. But a $150 million airliner still has to be guided through the skies by spoken instructions from a bloke with a radio in a control tower. That is because air traffic management (ATM) systems are stuck in the 1950s.”
--“Flight to the Future,” The Economist, Oct. 1, 2011

“Air traffic controller (ATCO) mobility . . . is of paramount importance in the current situation where market demand for ATCOs outweighs supply. From my point of view, the main hindrance to controller mobility is the lack of harmonization in training, accreditation, and implementation, both at the global and regional levels. . . . In Europe, there have been timid advances on these issues. The European Commission has recognized the importance of a European student/controller license, for example, but so far mobility has not become a reality. Despite the existence of a common core course, the common syllabus, a basis for unified training is needed, and activities such as ‘ab-initio training’ must be harmonized.”
--Carmen Librero, former CEO Director General of Air Navigation, AENA, “The CEO Column,” Airspace, Quarter 3, 2011.

Robert Poole is Searle Freedom Trust Transportation Fellow and Director of Transportation Policy

Print This