In this issue:
- Passenger trip delays-bad and getting worse
- New competition in Registered Traveler program
- More slot auctions proposed
- Will Austin be the next privatization candidate?
- High-speed rail-complement or competitor?
- News Notes
For many years the U.S. DOT has been reporting flight delays-the fraction of flights by route or airline that depart or arrive more than 15 minutes later than scheduled. Besides the fact that widespread schedule padding has made this yardstick rather elastic, it's a poor measure of actual passenger experience, since it ignores cancellations, diversions, and denied boardings due to over-booking.
The researchers at George Mason University's Center for Air Transportation Systems Research (CATSR) have come up with a better measure. "Passenger trip delay" includes trip disruptions due to delays, rebookings due to cancellations, rebookings after denied boardings, and diverted flights. The average passenger trip delay is a weighted average-the probability of each of these things occurring times the average duration of the delay due to each cause.
CATSR researchers crunched the numbers for all scheduled single-segment flights in the United States for 2007 and 2006. The results may be aggregated by airport, by airline, and by route. Their report, "U.S. Airline Passenger Trip Delay Report (2007)," published in April 2008, provides the data by airline and airport (http://catsr.ite.gmu.edu). Data for individual flights and routes is available on a special CATSR website: www.GreenFlights.info.
The results are eye-opening. Although cancellations, rebookings, and diversions are still relatively infrequent, when they do occur the delay times per passenger are quite large. While the average for a simple flight delay last year was 57 minutes, if your flight was diverted, your average delay was 3.75 hours, and if you had to be rebooked due to an oversold flight, your average delay was 3.0 hours. But if your flight was cancelled (as happened to me four times last year), your average delay was 11 hours. Putting all this together, 26% of all passengers experienced a delay last year (compared with 23% in 2006) and the average passenger trip delay was 1 hour and 54 minutes-a 24-minute increase from the previous year.
The biggest reason for the increase was the increase in the number of cancelled flights, combined with higher load factors which meant there was less slack in the system to accommodate stranded passengers.
And in case you are curious, the airlines with the lowest average passenger trip delay were (in order) Hawaiian, Aloha, Southwest, Frontier, and Airtran (averaging from 4.3 to 11.5 minutes). The five worst, averaging from 24.4 to 29.5 minutes, were Colgan, JetBlue, Pinnacle, Mesa, and (in last place) American. The most improved performance was by Alaska (in 7th place) and the biggest decline was by JetBlue.
How did airports fare? Of the 35 biggest airports, the five with the lowest average passenger trip delays were Honolulu, Chicago Midway, Salt Lake City, San Diego, and Portland (OR), all of whose figures were between 7.6 and 10.0 minutes. The five worst were Dallas-Ft. Worth, Newark, LaGuardia, Chicago-O'Hare, and in last place, JFK. Average delays for these airports ranged from 26.4 to 34.3 minutes. The biggest declines in performance last year were at JFK, Minneapolis/St. Paul, DFW, and LaGuardia. Most-improved were Denver and Houston-Bush.
In general, the airports with the greatest delays were-not surprisingly-congested hub airports. For smaller airports, the delay experience "reflects more about how that airport is connected to the network than how the airport or the airlines at the airport operate. Flights to/from larger, congested airports will propagate delays to these smaller airports."
This new, more comprehensive approach to calculating passenger delays is an important step forward, not only for aviation policy analysts and policymakers but also for individual air travelers.
Delay at TSA screening checkpoints is not included in the average passenger trip delays studied by CATSR (above), but it's yet another hassle of flying. According to the latest TSA figures (analyzed by Oliver Wyman and reported in Aviation Daily last month), average screening wait times across all categories of airports are in the four to six-minute range. But those averages conceal a great deal of variation. At the 38 largest airports, for example, the morning (6-10 AM period, generally the worst of the day) wait time average goes from a low of 3.0 minutes at Dulles to a high of 11.8 minutes at Tampa. But the maximum reported delay-which is what air travelers have to be prepared to face-ranges from 5.7 minutes at Baltimore to a whopping 29.0 minutes at Atlanta. In fact, for nearly all these large airports, the maximum delays during morning, mid-day, and afternoon periods were in the double-digit range. In the medium-size airport category, some of the highest average and maximum waits were at Anchorage (evening), Cleveland (morning), Orange County/John Wayne (mid-day), and West Palm Beach (mid-day and afternoon).
So I was cheered by the May 12 announcement that Atlanta was proceeding to implement "Clear" Registered Traveler lanes after all, reversing a previous decision to hold off until after this summer. Unfortunately, none of the other worst-case airports noted above has announced any plans to do likewise.
The RT industry is still evolving, with the two smaller players (Vigilant and FLO) testing new packages of service in order to compete with market leader Clear/Verified Identity Pass. FLO, which recently bought out partner Unisys, still has just a single contract, at Reno. But president Glenn Argenbright told Aviation Daily last month that he believes 40 to 45% of the eventual market is still up for grabs. FLO has unveiled two new membership levels, gold and platinum, offering a variety of other travel-related benefits (such as worldwide travel assistance and global concierge services). This is similar to the approach taken by Vigilant, which also has a single airport contract, with Jacksonville. Still, since all RT systems must be inter-operable to receive TSA certification, FLO and Vigilant can build a business based primarily on memberships rather than airport contracts.
But a possible X factor is being injected into the appeal of separate RT lanes without long lines. Under the rubric "Black Diamond," the TSA itself is experimenting (at large airports) with separate lanes for "expert travelers" who know the drill, "casual travelers" who fly infrequently, and family/special assistance lanes. Thus far, Black Diamond is operational at eight airports, four of which (Cincinnati, Denver, Oakland, and Orlando) already have RT in operation. Some travelers have expressed confusion at seeing both Expert Traveler and RT lanes at the same airport, but both TSA and Clear founder Steven Brill maintain that the programs are not competitors. It's true that RT thus far offers a "no-waiting" experience (at least that's been my personal experience thus far, at several airports), as opposed to merely a probably shorter Expert Traveler line. But I worry that the faction within TSA that has never been comfortable with RT may be behind this n ew approach.
I'm still betting on Registered Traveler as the better solution for frequent flyers, but it will be interesting to see how this plays out.
The U.S. DOT is certainly not behaving like a lame-duck agency. Only a month after announcing plans to auction off a portion of slots at congested LaGuardia airport, on May 16th Secretary Mary Peters announced plans for similar slot auctions at Kennedy and Newark. The announcement came at the same time as DOT released the final number for the cap on hourly operations at Newark: 83 flights per hour, until at least October 2009. Peters acknowledged that flight caps by themselves-by limiting supply-would tend to increase prices while also keeping out new entrants (which would also mean higher prices). But she said that adding slot auctions to the mix "strikes a balance" between incumbents and new entrants, facilitating competition that should keep prices in check.
The structure of the auction proposal is similar to the one already announced for LaGuardia. Incumbents would be guaranteed a minimum of 20 slots per day, with a small fraction of all slots above that number to be auctioned off over a period of years. The Newark and JFK plans differ in detail, reflecting the very large presence of Continental at Newark versus the diversity of airlines serving JFK.
To no one's surprise, the Air Transport Association and the incumbent airlines at the New York airports denounced the proposal. The Port Authority said it would "work with the airlines to examine our options for prohibiting the federal government from implementing this auction plan." There are two principal avenues for doing that: litigation and legislation. On the former, DOT general counsel D.J. Gribbin stressed the long legal history under which DOT has always maintained that by creating slots in the first place (via the original High Density Rule back in 1968), DOT was allocating airspace but not creating any property rights for slot users. I'm no attorney, but the fact that no one claimed that a "taking" was going on when Congress abolished the HDR in the FAA reauthorization act of 2000 suggests that DOT has the better legal case on this point.
That leaves legislation, which Kevin Mitchell of the Business Travel Coalition wasted no time in calling for. Although the New York delegation has a lot of clout, it's not clear that preventing limited slot auctions at the three New York metro airports would rouse many other members to action. However, without an aggressive DOT to champion the proposal next year, a legislative veto is certainly possible.
Personally, I would prefer to see runway pricing (or second-best, slot auctions) applied at an airport that has both a serious congestion problem and an open-mindedness to market measures. That airport could be Chicago O'Hare. It has at least one important new-entrant hoping to gain access-Virgin America, for service to Los Angeles and San Francisco. And it has JetBlue, which most likely would welcome the opportunity to expand its limited footprint there. A runway congestion pricing program could generate net new revenue for the not-fully-funded O'Hare Modernization Program, addressing airline concerns about whether pricing proceeds would be used to expand capacity.
Early last year, the Austin American-Statesman reported that both the mayor and the city manager were sounding out city council members on the possible long-term lease of Austin-Bergstrom International Airport. The inspiration was Chicago Mayor Richard Daley's then-new plan to lease Midway Airport under the federal Airport Privatization Pilot Program. Apparently, those discussions have continued over the past 14 months, because the May 2, 2008 issue of the Austin Business Journal now reports that Macquarie Group Ltd., one of the bidders on Midway, has recently had discussions on the subject with various council members and business leaders.
The point Mayor Will Wynn and City Manager Toby Futrell made last year has begun to sink in. Unlike other city-owned enterprises, such as Austin Energy and the convention center, the airport contributes no net revenue to the city budget. Indeed, under the terms of its federal grant agreements, it must not divert any airport revenue to non-airport purposes. As the mayor put it last year to reporter Ben Wear of the American-Statesman, "There's no financial reason whatsoever for the city to own the airport, as by law we can make not one penny from its operations or even property rentals. So why not explore a sale or a long-term lease that could net us hundreds of millions of dollars up front that we could put towards any number of community needs such as transportation?"
And that seems to be the nature of the discussions going on today in Austin, as well. The 2008 Business Journal article quotes council member Brewster McCracken as being strongly opposed to privatizing revenue-producing enterprises such as Austin Energy but open to leasing the airport, since it's the only legal way for the city to make a return on its investment.
Assuming the Midway lease gains Chicago City Council, TSA, and FAA approval, the odds of other cities applying for one of the four remaining slots in the federal pilot program will increase. The terms of the pilot program permit only one "large" airport to be leased, and Midway counts as large. But Austin is a "medium" airport as FAA measures these things, so it would have clear sailing. But in order for a "large" Tampa or Cleveland or San Diego airport to be leased, Congress would have to liberalize the terms of the pilot program. My guess is that the next Congress will not be keen on things with "privatization" in their name. But if America's mayors put this issue on their agenda, I think they could get the pilot program expanded.
They don't quite phrase it that way, of course, but leaders of the House Transportation and Infrastructure Committee from both parties have introduced legislation that would devote more than $14 billion of taxpayers' money to expanding Amtrak and funding high-speed rail corridors to compete with airlines. The Passenger Rail Investment and Improvement Act of 2008 was introduced May 8th. Over the next five years, it would give Amtrak $9.7 billion in capital and operating grants, provide $2.5 billion for grants to states to develop new or improved passenger rail corridors, and another $1.75 billion for grants to states and/or Amtrak to develop 11 "authorized high-speed rail corridors."
In his announcement of the bill, ranking Republican member John Mica (R, FL) explained that the bill envisions "true high-speed rail" beginning with a new high-speed link for the New York to Washington, DC route, with express service on this route taking no more than two hours. Mica explained that "three-fourths of chronic aviation delays in this nation are a result of congestion in New York area airspace," and since "air traffic control modernization is still years away" and "will only improve aviation congestion at the margins," what we need is "true high-speed rail in the Northeast Corridor" and then in other corridors around the country.
Well, it's true that in Europe we now see high-speed rail taking the majority share of trips between London and Paris, Cologne and Frankfurt, Paris and Brussels, and perhaps a few other markets. Rail works fairly well in those markets, since those European cities are far more centralized than most American city-pairs. Hence, where very large fractions of people want to go is "downtown to downtown." That's less and less true in the United States, as every census study on commuting patterns makes clear. And Europe's high-speed rail service has been achieved at very high taxpayer cost, which is what the T&I Committee leadership is now proposing to emulate.
I can already hear the arguments defenders of this proposal will make. After all, they will say, we subsidize air travel and highway travel; why not bite the bullet and subsidize high-speed rail, too? This argument was thrown at me in a debate on a National Public Radio program about Amtrak and high-speed rail several years ago by someone who must have known better. But just in case you haven't seen the figures, here is what the U.S. DOT's Bureau of Transportation Statistics figured out when they examined the extent of federal transportation subsidies by mode. BTS's December 2004 report, "Federal Subsidies to Passenger Transportation," first defined federal subsidy: it is not the total federal money spent on a mode. Rather, it is the difference between the revenue from user taxes paid by users of a mode and the amount of federal spending on that mode. Putting all the numbers on a common basis of net federal subsidy per thousand passenger miles, the aver ages (over the 1990-2002 period) were as follows:
- Amtrak: $186
- Urban transit: $118
- Airline travel: $6
- Highway travel: -$2
The highway figure is negative, because highway users pay slightly more in federal highway user taxes than they get back in actual federal spending on highways. And the air travel figure is so small because nearly all air-carrier airports are self-supporting, as is the air traffic control system (from ticket taxes and other user fees); only about $1.9 billion per year for FAA's air safety regulatory functions comes from federal general-fund revenues.
So in other words, what the T&I leaders (and high-speed rail boosters) want to do is to spend billions of general federal tax dollars to create heavily subsidized competition for self-supporting short/medium-haul air service. If I were an airline CEO or an airport authority chief executive, I wouldn't be too happy about that.
And how effective would a two-hour train between New York and DC be at decongesting New York's airspace? My current Official Airline Guide, pocket edition, shows 259 destinations reachable from the New York Airports, only one of which is Washington, DC. I have no doubt that a two-hour express train from Union Station to Penn Station would increase rail's market share in this one market, 10 or 15 years from now. But alleviating air congestion in the New York airspace? As ABC's John Stossel would say, give me a break!
Remaining Aussie Airports Going Private. In the late 1990s, Australia went through a major airport privatization, in which 50-year leases (with 49-year renewal options) were offered for nearly all the major airports, including Sydney and Melbourne, with results that the Centre for Asia Pacific Aviation calls "generally outstanding." But for various local reasons, a few important airports were left out. Last year saw the privatization of Hobart Airport, and in April the state of Queensland announced the upcoming privatization of Cairns and Mackay airports, along with a further 12% stake in part-privatized Brisbane. Cairns is the country's fifth-busiest international gateway.
New TRB Report on Airport Ground Access. The Transportation Research Board has announced the publication of Synthesis Report #5 in the new Airport Cooperative Research Program. "Airport Ground Access Mode Choice Models" reviews various mode choice models and how they can be used in airport planning. (http://onlinepubs.trb.org/onlinepubs/acrp/acrp_syn_005.pdf)
Bill Would Prohibit Tipping Off Screeners. CQ Weekly reports that the House Homeland Security Subcommittee on Transportation Security and Infrastructure Protection approved HR 5909, which would prohibit employees of TSA, DHS, or any other federal agency from sharing information about any upcoming undercover tests of airport security. According to Chairwoman Rep. Sheila Jackson Lee, screeners have been tipped off in advance to a number of undercover tests of airport screening.
Enrollment Starts for International Registered Traveler. Customs & Border Protection has begun processing applications from frequent international air travelers for its Global Entry registered traveler pilot program (discussed in Issue No. 33). The program will begin in June at JFK, Houston-Bush, and Washington Dulles airports. Applications can be made at www.cbp.gov/travel. The fee is $100.
New Brookings Book Compares Aviation Policy Across Countries. Cliff Winston and Gines de Rus have edited an important new book called Aviation Infrastructure Performance: A Study in Comparative Political Economy. It contrasts the relative lack of reform in the United States with "privatization, regulation, and slot reform" in Europe, with other chapters on airport and ATC policy in Canada, the U.K., Australia and New Zealand, China, and developing countries. (www.brookings.edu/press/books/2008/aviationinfrastructureperformance.aspx).