The Urgent Need to Reform the FAA's Air Traffic Control System

The U.S. economy depends on safe, reliable, and affordable air transportation. Beginning in 1978, airline deregulation transformed commercial aviation from a luxury for the few to a service available to essentially all Americans. U.S. companies depend on the airlines to transport their employees, and a growing number of all sizes make use of business aviation: corporate jets and turboprops, air taxi services, and fractional-ownership programs.

This ubiquitous and affordable air transportation depends critically on the U.S. aviation infrastructure, which consists of airports owned and operated by local government agencies and the air traffic control (ATC) system owned and operated by the Federal Avi­ation Administration (FAA).

The steady growth in aviation activity was inter­rupted in 2001 by the 9/11 attacks and by a recession, but flight activity returned to pre-recession levels by 2005, and significant summertime congestion has reemerged. The FAA and other aviation experts pre­dict serious trouble over the next two decades, driven by continued aviation growth.

First, a growing number of air travelers are flying in planes of smaller average size as narrow-body planes replace wide-body planes, regional jets replace nar­row-body planes, and business jets replace regional jets. This is increasing the number of planes that the ATC system needs to control significantly faster than the number of air travelers is growing, exacerbating the FAA's funding problem.

Second, this increased volume of air traffic will soon bump up against inherent limits of the current air traffic control system.

In 2003, Congress acknowledged the serious­ness of the ATC problem in the Vision 100 re­authorization of the FAA[1] by creating the Joint Planning and Development Office (JPDO) to plan for and coordinate the transition to a Next Gener­ation Air Transportation System (NGATS). NGATS would be a major redesign of aviation in­frastructure, aimed at replacing the traditional methods of separating planes "by hand" with a far more automated, technology-intensive system that could triple air traffic capacity by 2025. The JPDO effort is chaired by the Secretary of Trans­portation, with active participation of the Depart­ment of Commerce, the Department of Defense, the Department of Homeland Security, the FAA, the National Aeronautics and Space Administration (NASA), and the White House Office of Science and Technology Policy, as well as considerable in­volvement by the private sector.

As part of its early efforts, the JPDO modeled future demand for air travel and compared it to a business-as-usual scenario of continued modest annual increases of airport and ATC capacity. It concluded that "the demand scenarios quickly out­strip current and anticipated [ATC] capacities…. At higher levels of demand, system delays quickly rise over the course of a simulated day to untenable levels." The JPDO white paper summarized the sit­uation: "These extreme delays indicated that antic­ipated 'baseline' levels of current and future capacity will be inadequate for providing even min­imally acceptable levels of service quality for National Airspace System users."[2]

The JPDO has estimated that not expanding the system's capacity by 2020 will cost the U.S. econ­omy $40 billion per year. The white paper noted that, when demand exceeds the system's capacity, preventing untenable delays will require "a proce­dure or algorithm for 'trimming' flights from the ini­tial demand scenario until [existing capacity] is able to serve the 'trimmed' demand while meeting the agreed-upon service quality standard."

In other words, failure to expand ATC capacity will force significant rationing of airline and busi­ness aviation flights. This will significantly increase the average price of the restricted flights, and some valuable trips could not be made at all.

The leaders of the U.S. Chamber of Commerce have said that, unless the United States acts soon to address this fundamental problem with aviation infrastructure capacity, the consequences could be "devastating." To avoid this crisis, they have called for designing and setting up an ATC system that can safely and efficiently handle this heavier demand.[3]

Tripling Capacity

The JPDO is well along in fleshing out a techno­logical and operational concept for the Agile Air Traffic System, the ATC component of the overall Next Generation Air Transportation System.[4] Its basic premise is that system capacity is not a law of nature, but rather a function of the technology and operational concept employed.

The purpose of air traffic control is to keep planes from running into each other-more technically, to provide safe separation between planes in all phases of flight, including on the ground. Before radar was used to separate aircraft, controllers on the ground used procedural separation methods: rules that specified how far apart planes must stay along a given flight path (in-trail separation) and between different altitudes. When planes and controllers can only approximate aircraft latitude, longitude, and altitude, the rules call for huge separation margins to allow for large errors.

The introduction of radar over the land area of the United States in the 1950s and 1960s allowed reduction of lateral and in-trail spacing, since con­trollers were able to determine approximately where each plane was. Within the past few years, more precise altimeters have allowed reductions in the required vertical separation of jet cruising alti­tudes, thereby increasing the number of "flight lev­els" for the en route portion of flights. The increasing availability of GPS units on both airliners and business jets means that pilots themselves have much more accurate information on their locations, although current ATC practices make very little use of this capability.

While the accuracy of locational information has increased over the past several decades, the funda­mental concept of ATC is still the manual model developed prior to World War II. Thus, before every significant action, a pilot must receive permission from an air traffic controller on the ground, who watches a traffic display and tells the pilot what to do and when. Even though a great deal of "intelli­gence" is built into most airliners' flight manage­ment system computers, pilots are not allowed to use it unless authorized by a controller.

Furthermore, planes are still controlled largely "by hand" because, while controllers' displays have been modernized for the most part, controllers have been given very few automation tools to predict conflicts or to manage large amounts of information in short periods of time. Because of the understand­able limits on how much information a controller can manage at a time, the system must retain very large separation margins fore and aft, left and right, and above and below each plane to ensure safe operations.

The premise of the next-generation system is that many routine functions can be automated by obtaining and sharing precision real-time informa­tion about planes in flight and the weather and that the separation responsibilities can be shared among control centers on the ground and cockpits in the air. Some have termed this model network-centric air traffic management (ATM) as opposed to the tra­ditional human-centric air traffic control. Planes can safely fly much closer together with far more pre­cise, real-time information on:

• Each plane's exact location and heading (its intention);
  
  
• Weather conditions throughout the system; and
  
  
• The extent and duration of the vortices that spin off a plane's wings. These vortices can be hazardous to following aircraft during landings and takeoffs.

This reduction in separation distances could double or triple system capacity.

The network-centric model could greatly improve the quality of air service, both airline and non-airline. The major breakthrough is to let auto­mation on the ground and in the aircraft perform routine functions and separate aircraft based on their known flight profiles. The controller's role would change dramatically: Automation would manage the flights and monitor conformance with clearances and planned trajectories, and the con­troller would manage exceptions.

In the aircraft, synthetic vision techniques now in field-testing will enable planes to land in low-visi­bility conditions that today frequently cut airport arrival rates in half. Other new technologies will provide "precision approaches" to thousands of smaller airports at far less cost than traditional instrument landing systems. More precise informa­tion about planes' positions and their tip vortices will allow planes to use closely spaced parallel run­ways simultaneously. Some airports could even add an additional parallel runway without having to enlarge their physical land areas.

Shifting to the new model should also produce large cost savings. One premise of the network-cen­tric approach is that control of aircraft will be possi­ble from anywhere to anywhere. Historically, ATC facilities have been located adjacent to the airspace that they control. A tower is physically at the air­port, and each TRACON (Terminal Radar Approach Control) and each of the 21 en route centers is located within the regional airspace it controls. However, with satellites, dispersed sensors, and high-speed data links, facilities can be located virtu­ally anywhere and be sized to do an economically efficient amount of work. Some tentative plans call for replacing the FAA's 21 centers and 171 TRACONs-most of them fairly old-with 35 new service hubs.[5] Thousands of costly-to-maintain ground radars and other navigation aids (navaids) could be retired once planes are equipped for net­work-centric operations.

While all of the details are not yet finalized, experts from the federal agencies sponsoring the JPDO (especially NASA and the Departments of Defense and Transportation) agree that the net­work-centric ATM model can double or triple the system's capacity and cost no more-and perhaps less-to operate than the current system. This means major productivity gains in a field in which productivity has not fundamentally changed since radar was introduced. This prospect contrasts starkly with the dismal vision of congestion and rationing if aviation continues business as usual.

Obstacles to Implementing the
Next-Generation System

This vision of a network-centric ATM system faces several serious obstacles. Some are the normal kinds of resistance to change from those who are comfortable with the status quo. For example, the air traffic controllers union has resisted early moves toward automation technologies and has clearly expressed its preference for retaining a human-cen­tered ATC system over the next several decades. Some aircraft operators (including some airlines and many private pilots) are resistant to any man­dates to install new onboard avionics equipment, even though full benefits for all system users (such as large cost savings from retiring costly ground-based navaids) can be realized only after all planes in the system are properly equipped.

Three more fundamental obstacles pose even more serious threats: lack of funding, high modern­ization risks, and political constraints.

Lack of Funding. As FAA Administrator Marion Blakey and then-Transportation Secretary Norman Mineta said repeatedly in 2005 and 2006, the changes in aviation over the past decade have devas­tated the FAA's funding base. A large majority of the FAA budget-the ATC system accounts for nearly two-thirds-comes from aviation excise taxes, and the lion's share of the tax revenue comes from the 7.5 percent tax on the price of airline tickets.

The long-term trends of declining ticket prices due to increased market share for low-cost carriers and increasing air traffic due to increased use of smaller planes have put a serious squeeze on ATC funding. Payroll costs of the labor-intensive human-centric ATC system consume most of the available budget, leaving little for capital investment. In fiscal year (FY) 2005 and FY 2006, the FAA budget for facilities and equipment was reduced by 20 percent ($500 million) below the authorized levels.

The transition to NGATS will require major capi­tal investments over the next two decades to install new technologies and to replace numerous obsoles­cent facilities with a much smaller number of new ones. The cost estimate produced by the FAA's Research, Engineering, and Development Advisory Committee-the only estimate available so far-is an extra $1 billion per year over the next 20 years.[6] The FAA's current capital spending budget is focused on patching up the existing system, replac­ing antiquated display consoles with newer ones, and replacing the host computer system. While nec­essary in the short term, these investments will add little capacity to the system, but they are all that the FAA can afford under the current funding system.

Some, especially in the general aviation commu­nity, argue that Congress could solve the problem by appropriating a larger amount of general federal revenue each year, such as 25 percent to 30 percent of the FAA's budget instead of the recent level of about 18 percent to 21 percent. Yet given the federal budget deficit and numerous other claims on gen­eral-fund monies, this alternative appears extremely unlikely, especially for a program that has the potential to raise revenue from its users. This is why Blakey and Mineta have called funding reform essential for ATC modernization.

Technology Implementation Risks. The FAA has been attempting to modernize the National Air­space System (NAS), expanding its capacity and increasing its productivity, since it launched the NAS Plan in 1982. During the next 25 years, scores (if not hundreds) of reports from the Government Accountability Office (GAO)[7] and the Office of Inspector General (OIG) in the U.S. Department of Transportation (DOT) faulted the agency for bad management that had led to projects that were chronically late and seriously over budget.

In 2005, two OIG researchers presented an over­view of this failed modernization experience, trying to assess what went wrong.[8] They concluded that FAA modernization efforts had neither reduced costs nor increased productivity:

NAS modernization architecture and project designs have been consistently subverted by requirements growth, development delays, cost escalations, and inadequate benefits management. But all these things were symptomatic of the fact that FAA didn't think it needed to reduce operating costs.[9]

Thus, many observers are greatly concerned that the FAA's institutional culture is poorly suited to implementing anything as dramatic as the shift from human-centric ATC to network-centric ATM. In late 2004, the National Academy of Sciences convened an expert panel to assist the GAO in understanding the cultural and technical factors that have impeded previous ATC modernization efforts.[10] It found that "the key cultural factor impeding modernization has been resistance to change…[which is] characteristic of FAA personnel at all levels" and that "the key technical factor affecting modernization…has been a shortfall in the technical expertise needed to design, develop, or manage complex air traffic systems."[11]

The FAA is not designed to take risks, make investments, manage people to produce results, reward excellence, or punish incompetence. It is therefore not equipped to effect fundamental reform of the ATC system. Thus, major institutional change is probably a prerequisite for implementing the pro­posed network-centric ATM system.

Political Constraints. The third impediment to implementing a fundamentally different approach is political. The network-centric model can deliver major cost savings, ultimately providing two to three times the ATC capacity with the same number of-or even fewer-people because the changed paradigm makes the operations dramatically less labor-intensive. However, realizing these gains requires relatively swift retirement of huge numbers of costly radars and other ground-based navaids and consolidation of numerous ATC facilities. One current proposal would replace 21 en route centers and 171 TRACONs with 35 air traffic service hubs while redesigning all U.S. airspace.[12] Physical con­trol towers located at each airport would gradually be phased out as "virtual tower" functions are built into the new super-hubs.

As with the closing of military bases, Congress has a history of resisting the closure and consolida­tion of ATC facilities. The original 1982 NAS Plan included plans for facility consolidation, which were quietly dropped after it became clear that get­ting them through Congress would be very difficult. Congress came extremely close to forbidding the FAA's recent success in outsourcing its Flight Service Station system, which involved consolidating from 58 facilities to 20 facilities. The prohibition was ulti­mately defeated due to a credible veto threat from the White House. Many observers expect that, if left to the annual appropriations process, a facility con­solidation of the magnitude being considered for the next-generation system would suffer the same fate as the consolidations proposed in the NAS Plan.

An Institutional Alternative: A Self-Supporting Air Traffic Organization

One approach to addressing all three obstacles is to take the ATC system out of the federal budget process and make it a self-supporting entity, funded directly by its customers, analogous to the Tennes­see Valley Authority (TVA) or the U.S. Postal Service (USPS). Variants of this approach have been recom­mended by a series of federal studies and commis­sions over the past 15 years, including:

• The Aviation Safety Commission in 1988,
  
  
• The National Commission to Ensure a Strong Competitive Airline Industry in 1993,
  
  
• The National Performance Review in 1993,
  
  
• The Secretary of Transportation's Executive Oversight Group in 1994, and
  
  
• The National Civil Aviation Review Commis­sion (Mineta Commission) in 1997.

Funding Problem. This approach would address the funding problem by shifting from aviation excise taxes that are paid to the Treasury and appro­priated annually by Congress to fees for ATC ser­vices that are paid directly by customers to the new self-supporting Air Traffic Organization (ATO). Thus, fees would grow in proportion to the growth of flight activity rather than being tied to something much less relevant, such as airline ticket prices. Moreover, a predictable revenue stream, not subject to the federal budget process, would provide the basis for issuing long-term revenue bonds to fund modernization, in particular the transition to the network-centric system.

Cultural and Technical Obstacles. The com­mercialization approach would address the cultural and technical obstacles by enabling the ATO to attract and retain private-sector managers and engi­neers who are skilled at implementing complex technology projects. The ATO, like the TVA, would operate completely outside the federal civil service system and could hire, fire, and compensate its employees as any other high-tech business does. It would be governed by a board of directors largely representing the aviation customers.

Under this system, the overall NGATS approach, individual projects, and their implementation schedules would have to pass muster as delivering real value for the investment. That kind of vetting process is largely absent from the FAA.

Political Obstacles. A self-supporting ATO would also address the political obstacles to retiring navaids and consolidating facilities. By passing the enabling legislation, Congress would delegate these contentious issues to the customer-governed ATO.

Of course, persuading Congress to pass the enabling legislation is no small challenge. It is anal­ogous to the problem of military base closings. Members of Congress found it almost impossible to allow the Pentagon to close bases in their individual districts, with the result that obsolete bases were almost never shut down. To resolve the problem, Congress created the Base Realignment and Closure (BRAC) Commission in 1988 to make recommen­dations on which bases to close, shrink, or expand. Congress must then either accept or reject the entire package without amendment. The success of the first BRAC round has prompted three more rounds, which serve the national interest by making the mil­itary more cost-effective.

Congress could similarly vote one time to autho­rize a self-supporting ATO, thereby delegating the contentious tasks of consolidating numerous ATC facilities and shutting down thousands of navaids. By constraining itself from future temptations to intervene, Congress would safeguard the process of modernization (and avoid threatening the invest­ment-grade ratings of the revenue bonds issued to finance the modernization process).

Foreign Models. At first blush, this reform may sound like a utopian prospect, but over the past 20 years, several dozen countries have reformed their ATC systems along these lines, transforming them from tax-supported government depart­ments to self-supporting air navigation service providers (ANSPs).

Commercialized ANSPs in Other Countries

New Zealand converted its ATC operation from a division of its Ministry of Transport to a self-sup­porting government corporation in 1987; since then, more than 40 countries have done likewise. All of these commercialized ANSPs belong to the Civil Air Navigation Services Organization (CANSO), which has become an active participant in international aviation discussions and policy debates. Table 1 provides a brief overview of 10 of the leading commercialized ANSPs, which range from government departments (e.g., France) to a private, nonprofit corporation (Canada).

Experts consider financial autonomy the basic prerequisite for a commercialized ANSP, regardless of ownership or placement within or outside of gov­ernment departments, such as transport ministries. Financial autonomy requires that the ANSP:

• Fund its operations by means of direct charges to its customers and
  
  
• Have access to the capital markets to finance major technology and facility projects.

Because ANSPs continue to have a monopoly on provision of ATC services within their airspace, all of them are subject to some form of economic regulation. Further, because their countries are sig­natories to the International Civil Aviation Organi­zation (ICAO), they are also subject to arm's-length safety regulation.

Only three of the 10 ANSPs in Table 1 can be considered "privatized" in the sense of being outside of government (the governments of the United Kingdom and Switzerland continue to own major blocks of shares of NATS and Skyguide, respec­tively). The large majority of CANSO member ANSPs are government corporations. Yet, by definition, all are financially self-supporting and free to make business decisions on an ongoing basis, subject only to some kind of economic regulation and, of course, air safety regulation.

The GAO Study. In 2005, the Government Accountability Office conducted the first of several large-scale assessments of the performance of com­mercialized ANSPs.[13] The GAO selected five major ANSPs-Australia, Canada, Germany, New Zealand, and the United Kingdom (U.K.)-and GAO staff collected data and made site visits to all five. Among the findings were the following:

• Safety. "Available data from the five ANSPs indicate that since commercialization, the safety of air navigation services has remained the same or improved."
  
  
• Cost Control. "All five ANSPs have taken steps to control their operating costs, whether by eliminating some administrative positions or by consolidating facilities."
  
  
• Modernization. "All five ANSPs have also invested in new technologies and equipment, which the ANSPs say have lowered their costs by increasing controllers' productivity and pro­duced operating efficiencies, such as fewer or shorter delays."[14]

The GAO's report notes that the reasons for com­mercialization were similar in all five countries:

Before commercialization…[m]any were un­derfunded, as evidenced by freezes on air traffic controllers' wages and insufficient funds to replace aging technologies. Technol­ogy replacement programs often cost more, took longer, and delivered less than prom­ised, and stakeholders complained about performance and customer service. In some cases, the country as a whole faced wide­spread fiscal problems, and the commercial­ization of air navigation services was simply part of a larger movement to reform govern­ment enterprises.[15]

The same problems still beset the FAA's Air Traffic Organization today.

The report goes into some detail to explain what it means for an ANSP to operate as a commercial entity, albeit usually within government. "Each ANSP makes and carries out its own strategic, oper­ating, and financial decisions," reports the GAO. "A supervisory board oversees policy making and operations, and, when applicable, has fiduciary responsibility to stakeholders." As in a corporation:

An executive officer implements the ANSP board's policies and is, in turn, accountable to the board. Individual business units within the ANSP report to the CEO [chief executive officer] and are directly responsible for various aspects of the ANSP's day-to-day operations.[16]

The ANSPs establish performance measures and report annually on operational and financial perfor­mance. Their financial statements are subject to independent third-party audits.

Financial autonomy means that the ANSPs "rely on user charges as their primary source of revenue and on capital markets for additional funding." The charges are based on the ICAO's cost-recovery prin­ciples and cover oceanic, en route, and terminal-area services. General aviation users may be charged a flat fee for access to the system rather than the per-transaction charges that apply to air­lines. ICAO policies allow for the costs of some ser­vices not to be recovered via charges "in recognition of local, regional, or national benefits." The GAO notes that, in Canada, flights for search and rescue, air ambulance, and firefighting are exempt from user charges. ANSPs may also gener­ate revenues from miscellaneous business services, such as consulting, training, and the provision of ATC services in other countries.[17]

In accordance with ICAO principles, all five ANSPs are subject to economic regulation, prima­rily to protect users who must obtain their ATC ser­vices from the provider in question. The report goes into some detail on the different ways in which this regulation is carried out in the five countries.

The GAO also concluded that "[c]ommercializa­tion has allowed the ANSPs to implement modern­ization projects more efficiently." The uncertainty of annual appropriations had made it difficult for them to plan multi-year projects, but access to their own cash flow, along with the proceeds from issuing rev­enue bonds, "has allowed them to plan and execute projects more efficiently and has improved their ability to deliver projects on time, within budget, and to specifications."

The ANSPs have also found that "involving stake­holders in efforts to design, acquire, and deploy new technologies can be beneficial." This involve­ment continues through the design and implemen­tation process. In addition to involving customers, the ANSPs have found it valuable to involve air traf­fic controllers throughout the process.[18]

The MBS Ottawa Study. MBS Ottawa, a con­sulting firm, in conjunction with George Mason University, Syracuse University, and McGill Uni­versity, conducted a larger and more detailed study of the performance of commercialized ANSPs, which was published in January 2006.[19] The project was overseen by a 15-member inter­national advisory committee. The project team made site visits to and collected detailed quantita­tive data from the 10 commercialized ANSPs listed in Table 1, including all five that were stud­ied by the GAO. In addition to documenting the performance of the ANSPs since they were com­mercialized, the study and its appendix provide considerable detail on the governance structure and institutional framework of each.

Overall, the MBS study judged ATC commercial­ization to be a success. With respect to commercial­ization's impact on key performance measures, trend analysis found that:

• Safety was neutral or enhanced;
  
  
• Modernization was greatly improved;
  
  
• Service quality was improved;
  
  
• Costs were generally reduced, significantly in some cases;
  
  
• Financial stability was maintained; and
  
  
• Public interest was neutral or positive in most areas.

The MBS's "major finding is that commercialization models that provide the right balance of incentives have resulted in significant cost reductions, dramatic improvements in modernization, and major improve­ments in service quality, while improving safety." One of the most interesting findings is that:

Providing more autonomy for the ANSP has tended to cause a reorientation from treating government as the primary client to respond­ing to the needs of the aviation community. There is no longer any doubt as to who the customer is. Commercial ANSPs have dem­onstrated enhanced ability to respond quickly to customer needs.[20]

In addition, the report notes:

[T]here has been a…clarification of the government's role. Governments have en­sured the public interest through effective safety and economic oversight, financial regulations, environmental laws, protection of consumer rights, and recourse through the legal system.[21]

Summarizing the findings, the report notes that:

Commercialized ANSPs exhibit three main strengths-sensitivity to customer needs, agility in reaching a decision, and ability to carry it through. These characteristics have led to continuous improvements in efficiency, business discipline that delivers projects on schedule and on budget, and rapid deploy­ment of modern technology to enhance ser­vice quality.[22]

The IBM Study. Clinton Oster and John Strong researched and wrote the most recent study, which was published in mid-2006 by the IBM Center for the Business of Government.[23] Their report focused on applying lessons learned from the ATC commer­cializations in Canada and the U.K. to reforming the U.S. system. Thus, the report provides three parallel studies of ATC reform: the long history of efforts to reform the FAA and case studies of the commercial­izations of NAV CANADA and NATS. (FAA reform is discussed in the next section of this paper.)

The IBM study provides considerably more detail on how both NAV CANADA and NATS were cre­ated in an effort to solve long-standing problems in the provision of ATC services. Although their orga­nization models are quite different and both had to cope with the serious downturn in North Atlantic airline traffic for several years after 9/11, the authors conclude:

[B]oth NAV CANADA and NATS have emerged from the 2001-2004 period as fi­nancially solid organizations that are both well positioned not only to modernize to meet the growing needs of their own air­space, but also to extend their provision of various air traffic management services to other parts of the world.[24]

Several other lessons specific to NAV CANADA are worth highlighting. Unlike most European and Asia-Pacific countries, where ATC user charges have been in effect at least since the end of World War II, Canada was one of the few remaining coun­tries that funded ATC by taxing airline tickets prior to ATC commercialization. Reflecting on that transi­tion, the authors conclude that:

The adoption of a user charge system in principle increased the desire for users to play a role in governance. The not-for-profit structure with board representation by stakeholders creates good incentives for cost control and improved capital program management, and reduces the need for eco­nomic regulation.…

The customer orientation appears to extend to a capital program and planning approach that has been much better at both modern­ization and the development of new tech­nology, with respect to cost, delay, and performance.[25]

Turning to the need for NAV CANADA to cope with the economic downturn following 9/11, the authors conclude that:

Nav Canada's organization structure turned out to be an asset…. The stakeholder model in effect required all parties to make contri­butions and sacrifices. The nonprofit status established a clear financial objective dur­ing the period, while the rate stabilization fund allowed the company to manage the consequences of the downturn over a longer period.[26]