In this issue:
- Three Keys to ATC Reform
- A Funding Crunch or Not?
- Cost of Billing (A Huge New Bureaucracy?)
- Governance—the Real Issue
- Why We Need Arms-Length Safety Regulation
- News Notes
- Quotable Quotes
For many years now I have been pointing out the success of overseas experiences with spinning off the ATC service provision portion of government aviation agencies, converting them into self-supporting air navigation service providers. Study after study shows that—freed from the constraints of government budgets and day-to-day political oversight—these new ANSPs do a much better job of defining and implementing new technology, avoiding boondoggles, and delivering value for their aviation customers' money. Nav Canada is a prime, nearby example.
Unfortunately, a critical mass of knowledge and support for this kind of transformation of the FAA's Air Traffic Organization (ATO) does not yet exist in Congress. And the clock is ticking, as the current funding authorization expires Sept. 30th and Congress needs to act on reauthorizing the FAA and continuing or revamping its funding sources. So I'm increasingly getting the question: If it's not possible to "corporatize" the ATO this time around, what are the minimum steps in the right direction that could be enacted by the current Congress?
Given the crucial importance of laying the basis for implementing the NextGen approach—which would use information technology to double or triple airspace capacity by shifting to a far less labor-intensive form of air traffic management—this Congress must take three key steps.
First, we need a way to fund a considerable portion of the estimated $20 billion NextGen capital cost via revenue bonding. Why? Because there are large, "lumpy" investments that are better made up-front, rather than dragged out over several decades, so that benefits for users can be realized sooner rather than later. But to issue revenue bonds requires a dedicated funding source, one that's not hostage to the vicissitudes of the federal budget process. Hence, we need some kind of user fee whose proceeds are used solely to support NextGen revenue bonds. It would remain in place at least until those bonds are paid off, untouchable by Congress. The $25 fee per IFR flight (for all turbine aircraft) in the Senate bill would support about $5 billion of such bonding—not enough, but at least a start.
Second, the entire NextGen program ought to be under the control of a stakeholder board, with its own staff (independent of FAA), able to assess whether the pieces of NextGen offer enough user benefits to be worth what they cost, and also able to set deadlines for aircraft to be suitably equipped. The Senate bill includes such a board, but it does not clearly provide for independent staff, and it does not include a broad enough set of aviation stakeholders.
Third, NextGen will make it possible to retire thousands of costly-to-maintain ground-based navigation aids, and to massively consolidate facilities. But since Congress is incapable of restraining itself from preventing the closure or consolidation of facilities in members' districts, we need to create a mechanism analogous to the base realignment and closure (BRAC) process that former Rep. Dick Armey invented years ago for military bases. Here, the Senate bill falls far short, in that it simply directs FAA to come up with a consolidation plan, but provides no mechanism to implement it. And we know how far that would get.
Those three measures would at least get NextGen moving in the right direction.
Is there or isn't there a looming budget shortfall that could impede timely implementation of the $20 billion NextGen system? Advocates of the status quo—both in Congress and among the general aviation alphabet groups—say there isn't. The FAA and others, such as your editor, maintain that there is. The most recent round in this back and forth was a letter from the Government Accountability Office, in response to a question from the House Space and Aeronautics Subcommittee (www.gao.gov/new.items/d07918r.pdf). GAO's Gerald Dillingham told the members that "the current FAA funding structure can provide sufficient funding for NextGen—with some caveats." Dillingham relied mostly on a projection made last fall by the Congressional Budget Office, which projected future aviation excise tax revenues through 2016.
That, unfortunately, is an incomplete and misleading picture. I wrote about that CBO projection last fall (issue #38), after talking with the CBO analysts who prepared it. As I'd suspected, they did a simple projection of the aviation tax revenues, assuming that they grow slightly faster than inflation and GDP, based on historic relations between air travel and economic growth. What that ignores is structural changes in air transportation, discussed in last fall's GAO report on the same subject (GAO-06-1114T) and in FAA's justification for its funding reform proposal. A fundamental disconnect exists between the drivers of aviation tax revenue (the number of passengers carried and the average ticket price) and the ATC system's annual cost (driven by workload, based on the growth in air traffic). As the same total number of people gets carried in more, smaller units (RJs instead of 737s, air taxis and fractionals instead of airliners, etc.), traffic grows faster than passengers, and therefore costs grow faster than revenue. It is this structural disconnect that threatens the ability to afford NextGen.
The Congressional Research Service pointed this out last fall in their background report, "Reauthorization of the Federal Aviation Administration: Background and Issues for Congress," Oct. 18, 2006. In the section on Airport and Airway Trust Fund Issues (p. 13), CRS points out that the "FAA sees little prospect of a major increase in revenue from the trust fund's existing tax and fee system," and that "The FAA position is supported by the Department of Treasury estimates that suggest that annual revenue increases to the trust fund in the years ahead will be modest." (U.S. Treasury, Office of Tax Analysis, "Airport and Airway Trust Fund: FY2007 Mid-Session Review, Current Law Baseline," Summer 2006).
Status-quo defenders also like to claim that the existing aviation excise tax structure has provided stable and predictable funding. Guess again. What's most relevant in looking at NextGen funding is FAA's capital budget, called "Facilities and Equipment." I went back and got F&E figures from FY1992 through 2006 and adjusted them for inflation. Over that time period, the real value has bounced around from a low of $2.4 billion (1998) to a high of $3.5 billion (1992). We're also told not to worry because Congress can always supplement FAA's budget by adding general funding. CRS looked at that, over the period FY1997-FY2006, finding that the general fund contribution varied enormously, from as high as 38% (1997) to as low as 0% (2000) and 8% (2002)—not exactly stable and predictable. The DOT Office of Inspector General has seconded this point. In a report last fall on FAA management questions, it said that it's "extremely difficult, if not impossible" to predict future government appropriations and general fund contributions.
Unfortunately, although both GAO and FAA have done a good job of explaining the "fundamental disconnect" between revenues and costs, neither has produced a budget projection based on that disconnect. That leaves the naïve CBO projection as the baseline for discussion—and a handy rack for defenders of the status quo to hang their hats on.
"A huge new bureaucracy." Again and again those words have been proclaimed by opponents of any form of ATC user fee. It's a snappy sound-bite, and it appeals to private pilots, who tend to be individualistic and critical of what they see as a bureaucratic and meddlesome FAA. And on first hearing, it sounds plausible, especially if you believe AOPA's scare talk that ATC user fees would be charged to hundreds of thousands of single-engine piston planes for every touch-and-go operation. What a nightmare, right?
What's actually in prospect is either the FAA proposal to charge only commercial turbine operators (airliners, air taxis, fractionals) or the Senate proposal to charge only turbine operators (including corporate jets). In either case, the charges would apply to IFR flights, where basic flight parameters are included in the flight plans filed electronically with the FAA. The basic data needed to create an ATC bill are all included in the flight plan, and the actual invoice can be created by software.
As I type this piece, I am looking directly at a copy of an FAA "Invoice for Air Traffic Control Services." After the cover sheet, with the portion to send in with payment (check, money order, or credit card), there are several pages itemizing the "Detail Bill of Current Charges" for the month in question (in this case, August 2005). Each flight made in U.S.-controlled airspace is listed on a single line, with the miles flown either in Oceanic or Enroute airspace listed, along with the charge for that flight. This billing has been going on for the past seven years for "overflights" of U.S.-controlled airspace. Airlines and corporate jets alike pay these bills, for every flight that crosses U.S. territory or the oceanic airspace that is under U.S. control.
I asked FAA what kind of huge bureaucracy they had to create, to carry out this billing. It takes all of 2.5 full-time equivalents, since most of the work is done by software. To be sure, if all domestic IFR turbine flights were to be billed for en-route and terminal ATC services, the volume would be much greater, but the unit cost would be even lower than the trivial amount it takes for the overflight billing (thanks to economies of scale). And if the job seemed too large to be handled internally, the Air Traffic Organization could easily outsource it to companies that do vastly larger amounts of monthly billing—such as American Express, MasterCard, or Visa.
The proposed NextGen system is the most fundamental change in how we separate and manage air traffic since the introduction of radar in the 1950s. Paying for it, planning its phase-in (and the simultaneous phase-out of the old system), ensuring that aircraft get equipped in a timely manner, keeping costs under control, and ensuring that there is a real business case for each of its wonderful new capabilities--these are not things the FAA has ever been good at.
There is growing recognition among some of those at the Joint Planning & Development Office (JPDO) that governance is the neglected step-child of their work. As Kevin Brown of Boeing ATM put it last winter, "[we] have miles to go in ensuring [that] investment decisions directly involve key stakeholders—the big B in business cases, not just the business case for the FAA." He went on to say that "A fully thought out governance model would also clarify the roles of industry, today's alphabet soup of advisory committees, working groups, labor, and stakeholder associations. I am talking about a governance process for making decisions and making them stick." And he noted that in the parallel ATC modernization effort well under way in Europe, called SESAR, "Everyone understands who votes and how many votes they hold." (Kevin Brown, "We All Fly Together," The Journal of Air Traffic Control, October-December 2006)
As I noted in the first article in this issue, the board proposed by the Senate reauthorization bill is a promising first step in this direction. But since NextGen really involves both capital spending and transforming ATC operations, more thorough-going governance reform is required. I've been discussing this with knowledgeable people, and here is how one sums it up: "The real issue is governance structure. To run the ATO like a business, it must be a business. It is not one now—it is a government organization, and that means it is in the political system. As we have discussed, it needs to be taken out of the annual appropriations process to really change and permit strategic planning, and they do not propose that in this legislation. . . .There are dozens of government corporations in the U.S. It is too bad we cannot have an Air Traffic Corporation. That would permit real innovation and improvement." Those words of wisdom come from Tony Broderick, who was Associate Administrator for Regulation & Certification, 1985 to 1996. Tony's been there; he knows the problems from the inside. We should take his recommendations seriously.
In their outstanding report on ATC reform for the IBM Center for the Business of Government (www.businessofgovernment.org/pdfs/OsterReport.pdf), Clint Oster and John Strong make a powerful case for separating air safety regulation from ATC operations. There will always be trade-offs between expanding capacity (e.g., decreasing separations) and air safety, they explain, but today those trade-offs are all made within the FAA. If the ATO was a separate organization, they wrote, "The same trade-offs between safety and capacity would remain and be just as technically difficult, but the regulatory tensions that are now internal to one organization would become external and between two different organizations." Moreover, these trade-offs would become more public and open to outside scrutiny.
Two recent events underscore these tensions. The first concerns landing operations at Memphis, the major hub for FedEx. For many years, during the very busy evening period when FedEx flights converge on Memphis, planes have been allowed to land simultaneously on east-west runway 27 while others are landing on intersecting north-south runways. Controllers there called the situation dangerous, especially when a plane has to abort a landing, as happened on Feb. 18th. In April, the FAA announced a rule change, requiring such landings to be staggered, prompting protests by FedEx that the change would lead to serious delays. I don't have enough information to assess this trade-off, but it's the kind of thing that would be more transparent with arm's length safety regulation.
Another example is the ongoing battle over consolidating TRACONs. Controllers union NATCA is opposing FAA's proposed consolidation of the Palm Springs TRACON into the Southern California TRACON (which, over the years, has consolidated much of the airspace in that very congested region, successfully consolidating other airport-specific TRACONs such as Burbank). They have recruited California Senators Boxer and Feinstein (both Democrats) as well as local Republican Rep. Mary Bono, in support of House and Senate bills to place a moratorium on further TRACON consolidations. NATCA argues that such efforts are "unsafe and unwise" unless controllers and other stakeholders are directly involved in the process.
Well, local knowledge as possessed by the controllers at Palm Springs could conceivably add value to the planning process. But controllers also have a vested interest, like the members of Congress, in preserving jobs where they are today. The case for TRACON consolidation is primarily one based on efficiency and capacity: there are economies of scale to be gained, and the opportunity to better match staffing to workload. Since controllers always play the safety card in situations like this one, this is another example of a case where the safety implications should be weighed, objectively, by a party external to the ATC service provider—namely, an arm's-length FAA.
Poole Interview on Avweb. This week's I-Pod interview on Avweb (an online aviation news service) deals with ATC user fees and corporatization, and features the editor of this newsletter. This is such a hot topic that they extended the interview to about twice the normal length, so if you download it, be prepared to spend up to 16 minutes listening. Go to www.avweb.com. (If they've taken it down by the time you get to this, it is also posted at www.avgroup.com, under the "Hangar Talk" section.)
CPDLC Moving Ahead in Europe. Both Eurocontrol and corporatized German ATC provider DFS are moving forward with controller-pilot data-link communications (CDPLC), a program put on ice in this country by the FAA several years ago, despite a promising trial by American Airlines. The team of SITA/Sofreavia has won a contract to provide front-end processors for both ATC providers; they should be up and running by next spring. The European Commission is expected to mandate that all air navigation service providers have CPDLC in place by 2010, with all aircraft equipped by 2014. About 350 European airliners are already using CPDLC in the Maastricht Upper Area Control Center airspace.
NextGen Concept of Operations Published. The Joint Planning & Development Office (JPDO) has released version 2.0 of the Concept of Operations for the Next Generation ATC system, dated June 13, 2007. You can download it from www.jpdo.gov/library/NextGen_v2.0.pdf.
Correction, re Controller Work Schedules. In issue #43, in the piece about the NTSB's findings on controller fatigue, I wrote that the 2-2-1 shift scheduling pattern that leads to fatigue problems had been demanded by NATCA, the controllers' union. A controller emailed to explain that this shift pattern pre-dates NATCA, and must have been negotiated by its predecessor, PATCO. He also pointed out that in many facilities, most controllers don't work a midnight shift as part of this pattern, because the workload doesn't require it; they work a third day shift instead. I appreciate the correction.
Two More ANSPs Separated from Larger Government Bodies. The Civil Air Navigation Services Organization newsletter reports that Iceland is the latest European country to formally separate ATC service provision from aviation safety regulation. As of the first of this year, the ATC portion of the Icelandic Civil Aviation Administration was spun off as Flugstodir ohf (Isavia). ICAA will continue as the safety regulator. Poland has enacted a law to separate airports and air traffic control. The new air navigation service provider is the Polish Air Navigation Services Agency.
"We can't have a system where commercial airline passengers are subsidizing travel on corporate jets. While user fees may not be the right solution, we need to find a way to confront this problem."
--Sen. Frank Lautenberg (D, NJ), Philadelphia Inquirer, April 16, 2007
"[A] single person working in today's [ATC] operational concept can handle at most about 15 airplanes at a time. There are about 15,000 controllers, and let's say that about 4,000 are working a day shift. It would be great if those people were actually controlling an average of even 10 airplanes each. That would equate to about 40,000 controlled aircraft. Today, a busy period in the NAS has only about 6 to 8 thousand aircraft under IFR control. That's less than 2 aircraft per controller. We've got plenty of room to go in creating an operational concept and controller automation support that significantly increases the overall productivity."
--Michael S. Lewis, Boeing ATM, Journal of Air Traffic Control, April-June 2006, p. 10.