The FAA's Management Advisory Council has generally been content to fly below the radar, but the looming crisis in air traffic control funding and governance has moved it to take a much higher profile. In a hard-hitting report dated May 12, 2005, the MAC called for sweeping reforms aimed at reducing costs, increasing productivity, and modernizing the antiquated ATC system.
To set the stage, the report notes that the FAA's current 10-year plan, which (heroically) assumes full funding of operations and capital investment, would increase the ATC system's capacity by less than the likely growth in air transportation. They cite a study by DRI-WEFA which projects that under this (unlikely) fully funded scenario, passengers would experience 63% more delay hours in 2012 than they did in 2000, the previous record year. They also note that as things look now, during FYs 2004 and 2005, the FAA will have suffered a net loss of over 1,000 controllers (nearly 7%) at a time of rapid air traffic growth.
Although the report makes the predictable calls for "full funding" from Congress, which would mean a large increase in general fund support (highly unlikely in the current deficit-reduction environment), it also suggests the need for innovative financing such as issuing revenue bonds (as proposed in Reason Foundation's recent report on the ATC funding crunch, at www.reason.org/ps332.pdf), to accelerate needed investment in advanced technology. And that's all to the good.
But where the report really breaks new ground is in mentioning the previously unmentionable: the need for, and feasibility of, major structural cuts in current FAA spending. The needed major gains in productivity, it says, "can only be achieved by fundamentally rethinking the facilities FAA currently operates and by closing or consolidating some of those facilities." In particular, it suggests:
- Consolidating over 150 terminal radar approach control centers (TRACONs) into about 50 new, high-tech facilities;
- Replacing FAA's nine regional offices with three;
- Outsourcing the remaining 71 non-radar towers (226 are already run safely and efficiently by private contractors, at less than half the cost of comparable FAA-run towers); and
- Longer-term, consolidating the 20 en-route centers into a smaller number of new, high-tech centers.
None of this is rocket science. In fact, similar facility consolidation is under way (or completed) in many of the countries that have commercialized their ATC systems over the past 15 years, such as Australia , Canada , Germany , and the U.K. And although various studies have documented the existence of economies of scale in ATC facilities, never before has an official body dared to recommend such sensible changes out in the open. So we owe the MAC a debt of gratitude for speaking up.
But of course, the MAC is merely an advisory body. It is not the ATC system's board of directors. It has no power to make any of these changes, which a real board, charged with running the system, would do. That's why governance reform�not mentioned in the MAC's report—is an even more essential ingredient in fixing what is wrong.
The old clich�, "Be careful what you ask for, because you may get it," came true for the National Air Traffic Controllers Association (NATCA) early this month, when an in-depth investigation blew the lid off a festering situation at the New York TRACON. For years, insiders have viewed this facility as the most out-of-control in the whole ATC system�meaning that FAA management had abdicated control to the union. But nobody ever took the trouble to document what was going on, until now. And it all happened because NATCA tried to draw attention to the facility.
Here's what happened. After several DOT Inspector General reports on abuse of overtime and sick leave at ATC facilities in general, management at the New York TRACON decided, in January, to crack down on overtime abuse. From now on, they announced, all such requests must be approved in advance by management. What happened next was well-orchestrated. Anonymous calls started coming in on the FAA hotline in February, reporting an alarming number of controller errors at the facility. A gullible press reported them at face value, quoting union officials to the effect that because the TRACON was way under-staffed, and because management was no longer approving much overtime, air safety was now being jeopardized, with errors being made by overworked controllers. Union allies, Sens. Hilary Clinton and Charles Schumer, both sang from the same hymn book.
Instead of caving in, FAA headquarters decided to fight back. In March, they sent in a Tiger Team of 25 safety experts--current and former controllers, field supervisors, human resources experts, and management professionals--to do a detailed, on-site assessment of the TRACON. Their report was released on June 2, and remains posted on the FAA web-site. I urge you to down-load it and read for yourself what I can't really do justice to in these few paragraphs (www.faa.gov/library/reports/media/N90_Report_2.pdf).
First, the TRACON was not understaffed. The staffing level referred to by the union (270) was not based on FAA standards or operational requirements; instead, it was negotiated as part of the union contract in 1992 (and modified by subsequent memoranda of understanding). The Air Traffic Organization's current standards call for 170 controllers there, with proper scheduling. Staffing at present is far above that, at 225. But you'd never know that by looking at how many positions were staffed on a typical shift. It turns out that controllers at the New York TRACON averaged only 3 hours and 39 minutes "on-position" out of each eight-hour shift. This is well below the level of other large TRACONs, such as Atlanta (4 hours and 56 minutes) or Southern California (5 hours and 6 minutes). Nobody expects controllers in these high-activity facilities to work eight hours on position without breaks, but spending more than half the shift on break is ridiculous.
And that brings me to one of the most fascinating tables in the report. Figure 23 (p. 36) provides comparative numbers on five large TRACONs: New York , Southern California, Chicago, Atlanta, and Dallas/Ft. Worth. These are all high-traffic facilities, with FY04 staffing ranging from 69 (Atlanta) to 210 (New York). What is truly amazing is the differences in controller productivity. New York averaged 9,841 IFR operations per controller in FY04. But Chicago and Atlanta each averaged more than 20,000�i.e., twice the productivity. Yet the bill for overtime in Atlanta was just $148K and in Chicago $700K that year, compared with $4.1 million in New York.
Here's how the overtime-abuse game was played, according to the report. A controller calls in sick, so management brings in an off-duty controller on overtime. Then the "sick" worker comes in and gets paid for working (or at least for being there on-break). The same game is played by workers on vacation coming in unannounced after a substitute has already been brought in on overtime. Last year 21 controllers ended up earning over $200,000, and that number is on the way to 50 controllers this year, according to FAA spokespersons.
And, by the way, the report goes into great detail analyzing the "surge" in operational errors in February, concluding that none jeopardized safety. The majority were what the team calls "compression errors" on final approach, in which the 3-mile spacing between planes may slip briefly to 2.9 or 2.8 miles. "The team believes that this phenomenon occurs at every major airport across the country," the report says. Indeed, the NATCA-endorsed novel TRACON (by reporter Paul McElroy), presents this as routine practice in the busy Chicago TRACON. MIT aeronautics professor John Hansman, a noted aviation safety expert, told Newsday that the three-mile standard was set in the 1950s, when radar was much less reliable. "I'm not particularly worried if airplanes are 2.7 miles apart," he said.
What is most chilling about the report, however, is its finding number 8: "A culture of insubordination and intimidation exists at the New York TRACON that requires management attention to prevent derogation of safety." The team found "evidence that local NATCA officials engaged in physical intimidation and harassment of non-bargaining unit employees," and created a working environment that "could be inconsistent with safe and efficient air traffic control." Read pages 53-55 for all the gory details.
What to make of all this? First, the New York TRACON appears to be an extreme case of what can go wrong when management cedes control to the union, and some union people take extreme advantage of the situation. But many of these problems exist to a lesser degree at other facilities, all part of the legacy of the previous Administrator's approach to dealing with FAA unions. Second, some management heads should roll for letting this ( New York ) situation get so out of control for so long. The fact that FAA headquarters created the Tiger Team and has widely publicized its findings certainly sends the message that things are changing, but a few firings would underscore this message. And third, there's work to be done re-asserting management control at a lot of other facilities. According to ATW Daily News, the Inspector General found that at 26 of 30 facilities, negotiated staffing levels were higher than FAA's own staffing standards. In working out the new NATCA contract this summer, FAA should scrap those negotiated staffing levels and stick to what its own analysis shows to be required.
Footnote: After writing this assessment, I was given an internal NATCA response to the FAA report. It takes issue with just about everything, and makes a few good points about data and methodology. But overall it did not change my assessment. One other point worth noting is that sick leave is something of a special case for the FAA's ATC workforce. Today's controllers are mostly in their late 40s and 50s--and ailments increase with age. Moreover, controllers can't, while working, take any medications that may cause drowsiness, so if they need to use them, they have to stay home. So we should expect controller sick leave to be higher than that of the average workforce.
In my recent Aviation Week guest editorial (May 9, 2005), I argued that the crisis facing air traffic control is so severe, and the window of opportunity for change is so unprecedented, that the whole aviation community needs to come together in support of funding and governance reform. Unfortunately, the National Business Aviation Association has decided on an all-out defense of the status quo, demonizing any consideration of shifting to a fee-for-service model that would facilitate bonding major modernization investments.
To make its case, NBAA is engaging in numbers games. On a macro level, it is stirring up fears that a user-pays system would shift $3 billion in costs from airlines to business aviation. That number apparently comes from the fact that aviation excise taxes total nearly $10 billion and the airlines have been loudly noting that it's unfair that they (and their passengers) pay 90% of that total while general aviation (GA) uses 30% of ATC services. Thirty percent of $10 billion is $3 billion, and there you have it! NBAA, incidentally, estimates that business aviation (a subset of GA that includes air taxis) pays $600 million per year. And since no one has proposed or is proposing that small piston planes pay user fees, the difference between $3 billion and $600 million would still be a whopping $2.4 billion in new costs on business aviation. Is anything like that actually in prospect?
Before getting to that, let's try to get a better handle on how much high-end GA�jets and turboprops used by air-taxi companies, fractional-ownership services, corporations, and some individuals�actually uses ATC services. At the 49th annual convention of the Air Traffic Control Association, NBAA's Bob Lamond told the audience that business aviation constitutes 20-25% of en-route traffic. Data from the latest FAA Aerospace Forecast shows that in 2001, GA, air taxis, and commuters constituted 21% of IFR "overs" and 43% of IFR departures, while representing overall 36% of IFR aircraft handled. So the airline claim of about one-third of ATC services is in the right ballpark. And all these proportions will grow larger by 2016. This same subset of all air traffic will account for 25% of IFR overs, 48% of IFR departures, and 41% of overall IFR aircraft handled by 2016, according to the Forecast.
Against all of this, the NBAA still maintains, with a straight face, that business aviation is a "marginal user" of ATC, and that the system would be just as large and complex if business aviation disappeared. That argument may have been credible 20 or 30 years ago, but it's at odds with reality today, and will be completely irrelevant in the national airspace of two decades from now�which is what the next-generation ATC system needs to be planned (and funded) for.
Now despite the fact that non-airline jets and turboprops are increasingly important ATC users, I don't know of anyone who thinks that a fee-for-service system would be put into place that would charge a Learjet as much as a 777--even if it turns out that it costs the ATO just as much to provide services to the one as to the other. No country in the world does this, even though some economists would like to see them do so. What all the commercialized ATC providers use is some formula that takes aircraft gross weight into account (usually along with distance flown).
Back in 2001, Reason published an ATC policy paper by Viggo Butler and me, in which we used a weight-distance formula similar to Nav Canada's to illustrate what several turbine-powered GA planes might pay, in comparison with today's fuel tax. For a King Air, it turned out to be about $278/year more�i.e., $6,000 in fees instead of $5,722 in fuel taxes. For the Lear, the increase would be nearly $18,000 a year. That worked out to a fee averaging $79/hour versus a fuel tax of $47/hour. But we also pointed out that if the shift to a commercialized ATC system led to more capacity and reduced delays, it would take only a five percent reduction in flight hours (thanks to reduced delays) to entirely offset the increased cost of paying fees instead of taxes.
NBAA claims that a user-fee system, by increasing the cost of flying, would have the unintended consequence of reducing the amount of flying. In other words, they think that business aviation is highly price-sensitive. We can test that proposition by looking at what's happened in the past two years as fuel prices have increased. Jet-A cost $3.53 a gallon in March 2005, compared with $2.88 in March 2003�a 65 cents/gallon increase. For a Lear 35 consuming 215 gal./hour, that's an increase in fuel cost from $619/hour in 2003 to $759/hour in 2005. What's happened to GA flying over this two-year period? Flight hours are up, and both new and used GA aircraft sales are up�despite an increase in Learjet fuel cost of $140/hour! If a user-fee system substituted a $79/hour fee for a $47/hour tax, that would be a difference of just $32/hour. I just don't see that tiny difference reducing the use (or purchase) of Learjets.
Instead of standing athwart history yelling no, NBAA should start working actively to shape proposals for a new ATC system governed by (as well as paid for directly by) its customers.
One of the concerns of people in general aviation is that the board of directors of a user-funded Air Traffic Organization would be airline-dominated. Even if business aviation accounts for 30 to 40% of ATC services, if airlines had 60 to 70% of the board seats, in principle they could make all the key decisions in the interests of airlines—and potentially against the interests of business aviation that competes, to some extent, for high-end customers.
The country whose aviation sector looks the most like ours, especially in having a large general aviation component, is Canada. And Canada offers us a very useful model of governance reform in air traffic control. With the creation of Nav Canada in 1996 came the world's first stakeholder board of directors. The legislation that shifted Canada's ATC system to this new model specified the structure of its board, which is designed to balance the interests of airlines, business aviation, employees, the Canadian government, and the traveling public. The 15-member board has four airline seats, one for business aviation, two for Nav Canada 's unions, and three for the government. Each of these "stakeholders" appoints people to those designated seats. Those 10 select the CEO, who becomes the 11th board member. And those 11 select four more, from the traveling public, as independent directors. That's not exactly the composition I would recommend for an ATO board, since our aviation sector is more diversified, but it should give you the general idea.
To see how business aviation fares under this structure, in effect for nine years now, I interviewed Rich Gage, the head of the Canadian Business Aviation Association. I asked him straight out whether he thinks business aviation gets a fair shake, given that there are four airline-appointed board members and only one for business aviation. He noted that airlines are four out of 15 total seats, which means they can't dominate things. And he pointed out that all board members have a fiduciary duty to Nav Canada itself, not to the stakeholder group that appoints them. I asked if he thought business aviation is getting good value for the money they pay in fees. Although they've had some concerns, he said, on balance they get good service and a continually upgraded system. In response to my request that he speculate—if it were possible to go back to the old system, whether his members would prefer that--he said he doubted it, saying that the new system is pretty well accepted.
Because of concerns being raised by the GA community here, I asked Gage about Nav Canada's billing system. He said he's unaware of any problems with it, and he would be likely to hear from his members if bills were wrong or sent out late. What about the concern that user fees would increase the cost of owning and operating a business aircraft? "I think it's had no effect on business aviation," he said. "I just don't see nor do I hear people making aircraft decisions on the basis of Nav Canada fees."
Gage also explained some other aspects of Nav Canada's governance of which I was not aware. In addition to being represented on the board of directors, business aviation is also one of four "members" of Nav Canada (the other three being the airline community, labor, and the government). These four have the responsibility for things like bylaws changes and voting on new directors. And CBAA is also part of a large Nav Canada Advisory Committee, which includes representatives of airports, provincial aviation organizations, U.S. airlines, the Canadian pilots' union, etc.
In short, our neighbors to the north have provided one model of how the business aviation community, in view of its significant usage of ATC services, can become part of the system's governance. The U.S. business aviation community should see which aspects of this model could be applied in a user-governed Air Traffic Organization.
User Fees Pay for Albania Modernization. A Dow Jones report June 13 tells us of Albania getting a $48 million loan from the Export-Import Bank to buy new ATC technology from Lockheed Martin. The loan will be repaid largely from overflight fees. Yes, even little Albania charges planes that fly over its territory, using its ATC services. Yet the United States still lets hundreds of flights a day use our ATC services for free. Go figure.
Report on ATC Governance. The International Public Management Journal (Vol. 7, No. 3) carries a thoughtful article reviewing in some detail six different models for ATC governance. It concludes that the Nav Canada model is the one most applicable to the FAA. The article is by Ira Lewis of the Naval Postgraduate School in Monterey. You should be able to find it on the journal's web site: www.inpuma.net/editors.htm.
All but Four in Europe Now Corporatized. The latest newsletter of CANSO, the membership organization for commercialized ATC providers, allowed me to check on the progress of ATC reform in Europe . Comparing a map of Europe with the 40 full members of CANSO revealed that nearly every European country has made the transition to a self-supporting ATC system regulated for safety by the country's transport ministry. The only countries not included in this overwhelming trend are: France (are you surprised?), Finland, Greece, and the former Yugoslavia's components.