oversight

Slot-Controlled Airports: FAA's Rules Could Be Improved to Enhance Competition and Use of Available Capacity

Published by the Government Accountability Office on 2012-09-13.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

                 United States Government Accountability Office

GAO              Report to the Committee on Commerce,
                 Science, and Transportation,
                 U.S. Senate


September 2012
                 SLOT-CONTROLLED
                 AIRPORTS
                 FAA’s Rules Could be
                 Improved to Enhance
                 Competition and Use
                 of Available Capacity




GAO-12-902
                                             September 2012

                                             SLOT-CONTROLLED AIRPORTS
                                             FAA’s Rules Could Be Improved to Enhance
                                             Competition and Use of Available Capacity
Highlights of GAO-12-902, a report to the
Committee on Commerce, Science, and
Transportation, U.S. Senate




Why GAO Did This Study                       What GAO Found
To help manage airport congestion,           The 16 new beyond-perimeter flights that were authorized in 2012 for Reagan
airlines operating at four U.S.              National Airport are likely to have a limited effect on the airports in the
airports—Washington’s Reagan                 Washington, D.C., area. Reagan National has sufficient runway capacity to
National and the three major New York        accommodate the new beyond-perimeter flights and, with some improvements to
City area airports—must obtain               baggage handling and security screening facilities, will have sufficient terminal
operating authorizations called slots        capacity. Reagan National is routinely operating below 67 hourly takeoffs and
from FAA to take off or land. Airlines       landings (“slots”)—the maximum number authorized in any one hour—mostly
operating out of Reagan National also        because general aviation or other unscheduled aircraft operations decreased
may not operate flights beyond a             substantially after new security restrictions were imposed following the
1,250-mile perimeter without                 September 11, 2001, terrorist attacks. GAO’s analysis found that the new
congressional approval. In addition to       beyond-perimeter flights could add a maximum of about 5 percent of 2011
the 24 flights (12 round trips)              passenger levels at Reagan National, and if all of that increase came from
previously approved, Congress                passengers moving from Dulles International or Baltimore-Washington
recently authorized 16 more beyond-
                                             International Thurgood Marshall, these airports could experience a maximum
perimeter flights (8 round-trips) at
                                             decline of 4.1 percent of 2011 passenger levels. As a result, because Reagan
Reagan National—flights that the
airport authority fears will adversely
                                             National and Dulles generally do not share what are largely fixed costs,
affect Reagan National and the               passengers shifting from Dulles to the new beyond-perimeter flights at Reagan
authority’s ability to continue servicing    National would, in the worst-case scenario, increase average airline costs to use
its debt. Some airlines seeking to           Dulles by about 4 percent. Similarly, the new beyond-perimeter flights are not
serve slot-controlled airports assert        likely to affect the Washington Metropolitan Airports Authority’s (MWAA) ability to
that slot control rules cause the            service its $5.2 billion debt for the two airports. Finally, because half of the new
airports to be underutilized or used         beyond-perimeter flights were awarded to low cost airlines, thereby increasing
inefficiently. GAO was asked to review       competition, these new flights could have a positive effect on airfares on routes
(1) the effects of adding more beyond-       where new flights were added.
perimeter flights from Reagan National
and (2) how well slot control rules are      Slot control rules help the Federal Aviation Administration (FAA) manage
working to reduce congestion, while          congestion at these four airports by limiting the number of takeoffs and landings
maximizing capacity and encouraging          that airlines may make per hour. The rules, however, in effect, allow some
competition. GAO analyzed slot               existing airport capacity to go unused, and this capacity is therefore unavailable
allocation and airline schedule data,        to other airlines because airlines are not required to schedule a flight for each of
developed a statistical model, and           their slots, but instead are only required to use their slots 80 percent of the time.
interviewed FAA officials and others.        Moreover, FAA’s recordkeeping and its process for reviewing airlines’ self-
                                             reported slot utilization data do not provide sufficient assurance that FAA can
What GAO Recommends                          identify instances when airlines do not meet the 80 percent utilization
Among other things, GAO                      requirement or determine how much capacity is going unused. In addition, FAA’s
recommends that FAA improve its              allowing airlines to apply the requirement collectively to their pool of slots, rather
administration of the slot control rules,    than individual slots, further contributes to slots going unused and provides an
including applying the utilization           advantage to airlines with large slot holdings. In addition to some existing
requirement to individual slots.             capacity going unused, flights at the slot-controlled airports, even when operated,
Department of Transportation (DOT)           tend to be scheduled in such a way that available capacity is used more
and FAA provided technical comments,         inefficiently than at like-sized airports that are not slot-controlled, thereby limiting
which GAO incorporated as                    passenger growth and access by new-entrant airlines that could offer new
appropriate, and agreed to consider          service or lower fares. For example, GAO found that flights operated at slot-
the recommendations.                         controlled airports tend to be scheduled with smaller aircraft. Using statistical
                                             analyses, GAO found that scheduled passenger flights at slot-controlled airports
View GAO-12-902. For more information,       are 75 percent more likely to be scheduled by airlines using an aircraft with fewer
contact Susan Fleming at (202) 512-2834 or   than 100 seats than flights at other like-sized airports that are not slot-controlled.
flemings@gao.gov                             Slot-controlled airports also tend to have certain routes that are flown at higher
                                             daily rates and aircraft that are less full.
                                                                                        United States Government Accountability Office
Contents


Letter                                                                                      1
               Background                                                                   4
               Reagan National Can Accommodate New Beyond-Perimeter
                 Flights, Which Are Likely to Have a Limited Effect on Other
                 Washington, D.C., Area Airports                                          16
               Slot Control Rules Help Manage Congestion, but Allowing Airlines
                  Not to Use Some Slots May Hinder Competition                            29
               Conclusions                                                                53
               Recommendations for Executive Action                                       54
               Agency Comments and Our Evaluation                                         55

Appendix I     Objectives, Scope, and Methodology                                         58



Appendix II    Comparative Information on Major Airports in the New York City and
               Washington, D.C., Area                                                     64



Appendix III   History of Slot Control Rules and Related Actions                          65



Appendix IV    Trends in Average Annual Passenger Traffic and Airfares at Selected
               Washington, D.C., Area Airports                                            75



Appendix V     Logistic Regression Analysis of Slot-Controlled Airports and
               Aircraft Size                                                              99



Appendix VI    Comments from the Metropolitan Washington Airports Authority              107



Appendix VII   Comments from the Port Authority of New York and New Jersey               111




               Page i                                         GAO-12-902 Airport Slot Controls
Appendix VIII   GAO Contact and Staff Acknowledgments                                      112



Tables
                Table 1: Selected Slot Control Rule Provisions for the Four U.S.
                         Slot-Controlled Airports                                           10
                Table 2: Hourly Aircraft Operations (Takeoffs and Landings)
                         Allowed at Reagan National                                         17
                Table 3: Impact of New Beyond-Perimeter Exemptions on Average
                         Peak and Non-Peak Hour Slot Allocations at Reagan
                         National for the Summer 2012 Season                                18
                Table 4: Projected Impact of the New Beyond-Perimeter Flights at
                         Reagan National on the Cost per Enplaned Passenger in
                         2013 and 2014                                                      25
                Table 5: Percentage Difference in Flights at Slot-Controlled
                         Airports That Use Small Aircraft Compared with Flights at
                         Non-Slot-Controlled Airports, 2011                                 42
                Table 6: Percentage Difference in Likelihood of a Flight’s Using
                         Small Aircraft at Slot-Controlled Airports Compared with
                         Non-Slot-Controlled Airports                                       44
                Table 7: Logistic Regression Results for Slot-Controlled-Airport and
                         Aircraft-Size Model, Small Aircraft with 100 Seats or Fewer       104
                Table 8: Logistic Regression Results for Alternative Slot-
                         Controlled-Airport Variable Definitions, Small Aircraft
                         with 100 Seats or Fewer                                           104
                Table 9: Logistic Regression Results for Alternative Slot-
                         Controlled-Airport Variable Definitions, Small Aircraft
                         with 124 Seats or Fewer                                           105
                Table 10: Logistic Regression Results for Alternative Slot-
                         Controlled-Airport Variable Definitions, Small Aircraft
                         with 80 Seats or Fewer                                            106


Figures
                Figure 1: Annual Percentage of Late Arrivals and Flight
                         Cancellations at the Four Slot-Controlled Airports
                         Compared with All Airports from 2002 through 2011                    6
                Figure 2: Nonstop Beyond-Perimeter Slot Exemption Destinations
                         from Reagan National from 2000 through 2012                        15
                Figure 3: Areas of Unused Capacity at Slot-Controlled Airports              32



                Page ii                                         GAO-12-902 Airport Slot Controls
Figure 4: Airports’ Share of Scheduled Flights to or from the Same
         Destination by the Same Airline within 15 Minutes or 30
         Minutes of One Another, February 2011                             47
Figure 5: Average Load Factors at Large Hub Airports for
         Scheduled Flights during 2011                                     48
Figure 6: Proportion of Legacy and Other Airline Passenger Traffic
         at Slot-Controlled and Non-Slot-Controlled Large Hub
         Airports, 2011                                                    50
Figure 7: Timeline of Major Actions Taken Regarding Slot Controls
         and Perimeter Rules at the Major New York City Area
         Airports and Reagan National                                      66
Figure 8: Indexed Average Passenger Traffic for All BWI Domestic
         Flights and Flights between BWI and Denver International
         Airport, 2000 to 2011                                             75
Figure 9: Indexed Average Annual Airfare for All BWI Domestic
         Flights and Flights between BWI and Denver International
         Airport, 2000 to 2011                                             76
Figure 10: Indexed Average Annual Passenger Traffic for All BWI
         Domestic Flights and Flights between BWI and Las Vegas
         McCarran International Airport, 2000 to 2011                      77
Figure 11: Indexed Average Annual Airfares for All BWI Domestic
         Flights and Flights between BWI and Las Vegas McCarran
         International Airport, 2000 to 2011                               78
Figure 12: Indexed Average Passenger Traffic for All BWI Domestic
         Flights and Flights between BWI and Los Angeles
         International Airport, 2000 to 2011                               79
Figure 13: Indexed Average Annual Airfares for All BWI Domestic
         Flights and Flights between BWI and Los Angeles
         International Airport, 2000 to 2011                               80
Figure 14: Indexed Average Passenger Traffic for All BWI Domestic
         Flights and Flights between BWI and Phoenix Sky Harbor
         International Airport, 2000 to 2011                               81
Figure 15: Indexed Average Airfares for All BWI Domestic Flights
         and Flights between BWI and Phoenix Sky Harbor
         International Airport, 2000 to 2011                               82
Figure 16: Indexed Average Passenger Traffic for All BWI Domestic
         Flights and Flights between BWI and Salt Lake City
         International Airport, 2000 to 2011                               83
Figure 17: Indexed Average Airfares for All BWI Domestic Flights
         and Flights between BWI and Salt Lake City International
         Airport, 2000 to 2011                                             84




Page iii                                       GAO-12-902 Airport Slot Controls
Figure 18: Indexed Average Passenger Traffic for All BWI Domestic
         Flights and Flights between BWI and Seattle-Tacoma
         International Airport, 2000 to 2011                               85
Figure 19: Indexed Average Airfares for All BWI Domestic Flights
         and Flights between BWI and Seattle-Tacoma
         International Airport, 2000 to 2011                               86
Figure 20: Indexed Average Passenger Traffic for All Dulles
         Domestic Flights and Flights between Dulles and Denver
         International Airport, 2000 to 2011                               87
Figure 21: Indexed Average Airfares for All Dulles Domestic Flights
         and Flights between Dulles and Denver International
         Airport, 2000 to 2011                                             88
Figure 22: Indexed Average Annual Passenger Traffic for All Dulles
         Domestic Flights and Flights between Dulles and Las
         Vegas McCarran International Airport, 2000 to 2011                89
Figure 23: Indexed Average Annual Airfares for All Dulles
         Domestic Flights and Flights between Dulles and Las
         Vegas McCarran International Airport, 2000 to 2011                90
Figure 24: Indexed Average Annual Passenger Traffic for All Dulles
         Domestic Flights and Flights between Dulles and Los
         Angeles International Airport, 2000 to 2011                       91
Figure 25: Indexed Average Annual Airfares for All Dulles
         Domestic Flights and Flights between Dulles and Los
         Angeles International Airport, 2000 to 2011                       92
Figure 26: Indexed Average Annual Passenger Traffic for All Dulles
         Domestic Flights and Flights between Dulles and Phoenix
         Sky Harbor International Airport, 2000 to 2011                    93
Figure 27: Indexed Average Annual Airfares for All Dulles
         Domestic Flights and Flights between Dulles and Phoenix
         Sky Harbor International Airport, 2000 to 2011                    94
Figure 28: Indexed Average Annual Passenger Traffic for All Dulles
         Domestic Flights and Flights between Dulles and Salt
         Lake City International Airport, 2000 to 2011                     95
Figure 29: Indexed Average Airfares for All Dulles Domestic Flights
         and Flights between Dulles and Salt Lake City
         International Airport, 2000 to 2011                               96
Figure 30: Indexed Average Annual Passenger Traffic for All Dulles
         Domestic Flights and Flights between Dulles and Seattle-
         Tacoma International Airport, 2000 to 2011                        97
Figure 31: Indexed Average Annual Airfares for All Dulles
         Domestic Flights and Flights between Dulles and Seattle-
         Tacoma International Airport, 2000 to 2011                        98



Page iv                                        GAO-12-902 Airport Slot Controls
Abbreviations

AIR-21            Wendell H. Ford Aviation Investment and Reform Act for
                  the 21st Century
BWI               Baltimore Washington International Thurgood Marshall
                  Airport
BTS               Bureau of Transportation Statistics
DOJ               Department of Justice
DOT               Department of Transportation
IATA              International Air Transport Association
FAA               Federal Aviation Administration
JFK               John F. Kennedy International Airport
MWAA              Metropolitan Washington Airports Authority
TSA               Transportation Security Administration




This is a work of the U.S. government and is not subject to copyright protection in the
United States. The published product may be reproduced and distributed in its entirety
without further permission from GAO. However, because this work may contain
copyrighted images or other material, permission from the copyright holder may be
necessary if you wish to reproduce this material separately.




Page v                                                    GAO-12-902 Airport Slot Controls
United States Government Accountability Office
Washington, DC 20548




                                   September 13, 2012

                                   The Honorable John D. Rockefeller IV
                                   Chairman
                                   The Honorable Kay Bailey Hutchison
                                   Ranking Member
                                   Committee on Commerce, Science, and Transportation
                                   United States Senate

                                   For each takeoff or landing at Ronald Reagan Washington National
                                   Airport (Reagan National) and the three major New York City area
                                   airports—John F. Kennedy International (JFK), LaGuardia International
                                   (LaGuardia), and Newark Liberty International (Newark)—airlines must
                                   obtain an operating authorization, also known as a “slot,” 1 from the
                                   Federal Aviation Administration (FAA). Although more than 150 airports
                                   are slot-controlled throughout the world, including all major European
                                   airports, only these four airports are currently subject to slot controls in
                                   the United States. FAA started limiting the number of takeoffs and
                                   landings per hour at these airports more than 40 years ago because of
                                   congestion and delay. Reagan National is also subject to a federally-
                                   imposed 1,250-mile limit on the distance of nonstop flights to and from the
                                   airport, called the perimeter rule. However, on three occasions since
                                   2000, federal statutes have been enacted requiring a total of 40 slot
                                   exemptions (20 round trips) to the perimeter rule, allowing nonstop flights
                                   to destinations more than 1,250 miles from the airport. This includes 16
                                   more beyond-perimeter flights (8 round trips) required as part of the
                                   reauthorization of FAA in February 2012. 2 Although we reported in 1999
                                   and 2007 3 that Reagan National could accommodate additional flights,
                                   the airport authority that operates Reagan National—the Metropolitan


                                   1
                                    FAA regulations at 14 C.F.R. Part 93, subparts K and S applicable to Reagan National
                                   Airport use the term “slots,” whereas FAA orders applicable to JFK, LaGuardia, and
                                   Newark use the term “operating authorization.” For the purposes of this report we refer to
                                   both as slots.
                                   2
                                    FAA Modernization and Reform Act of 2012, Pub. L. No. 112-95, § 414, 126 Stat. 11
                                   (2012).
                                   3
                                    GAO, Reagan National Airport: Capacity to Handle Additional Flights and Impact on
                                   Other Airports, GAO/RCED-99-234 (Washington, D.C.: Sept. 17, 1999) and Reagan
                                   National Airport: Update on Capacity to Handle Flights and Impact on Other Airports,
                                   GAO-07-352 (Washington, D.C.: Feb. 28, 2007).




                                   Page 1                                                    GAO-12-902 Airport Slot Controls
Washington Airports Authority (MWAA)—raised concerns about the
airport’s ability to accommodate these new beyond-perimeter flights as
well as their effect on the other Washington, D.C., area airports and the
ability of MWAA to continue servicing its $5.2 billion debt.

The New York City area airports are operating under temporary FAA
orders, which constitute the current slot control rules that limit the number
of takeoffs and landings for flights out of those airports. However, despite
operating under slot control rules, these airports continue to be among
the most delayed in the country, which in turn affects the efficiency of the
nation’s entire airspace system. Moreover, although the rules were
designed to control congestion, while also maximizing the available
airport capacity, some airlines seeking entry into these airports assert that
the current rules do not adequately address inefficient uses of the
constrained airspace and airport capacity, and that not enough of existing
capacity is allocated to new entrants, restricting competition and access
to the primary commercial airports in the region. They also assert that
airlines currently operating out of these airports use their slots in ways
that keep out competitors. To address these problems at the New York
City area airports, FAA, which is responsible for managing congestion at
airports, tried to implement rules in 2006 (that were revised in 2008) that,
among other things, would have allowed FAA to auction some slots held
by incumbents and, under one of the two options, keep the proceeds. In
October 2009, before implementing any changes, FAA rescinded the final
congestion management rules for the New York City area airports, in
general, citing congressional and airline opposition to auctioning slots,
among other things, and a federal court ruling. 4 In 2010, the Department
of Transportation (DOT) began developing a “Slot Management and
Transparency” rule for JFK, LaGuardia, and Newark to replace the
temporary slot control rules under which they have been operating since
2006 for LaGuardia and since 2008 for JFK and Newark. These
temporary rules have been extended twice, creating a level of uncertainty
in the market about when the final rule will be issued and what it will
contain.




4
 74 Fed. Reg. 52132 (Oct. 9, 2009); 74 Fed. Reg. 52134 (Oct. 9, 2009). The federal court
ruling was an order by the United States Court of Appeals for the District of Columbia
Circuit. Port Authority of New York and New Jersey v. Federal Aviation Administration,
Court of Appeals for the District of Columbia, Order No. 08-1329 (Dec. 8, 2008).




Page 2                                                   GAO-12-902 Airport Slot Controls
In response to your request, we reviewed the potential effects of
increasing the number of beyond-perimeter slot exemptions at Reagan
National and whether slot control rules at the four U.S. slot-controlled
airports are working to reduce congestion, while maximizing available
capacity and encouraging competition.

To review the potential effects of increasing the number of beyond-
perimeter slot exemptions at Reagan National, we

•   compared the airport’s airside capacity assessment with actual usage;
•   reviewed terminal space needs;
•   assessed the effects of the additional beyond-perimeter flights on
    security screening wait times, passenger traffic, capital financing, and
    aircraft noise; and
•   interviewed officials from FAA, MWAA—which operates Reagan
    National and Dulles International Airport (Dulles)—and Baltimore-
    Washington International Thurgood Marshall Airport (BWI) as well as
    representatives from airlines and others.
To review whether slot controls at U.S. slot-controlled airports are working
to reduce congestion while maximizing available capacity and
encouraging competition, we compared and contrasted DOT data on
flight delays, as well as airline schedule data on aircraft size, flight
frequency, and other factors at slot-controlled airports with airports that
are not slot-controlled. We also constructed a statistical model to assess
whether flights at the slot-controlled airports are maximizing passenger
throughput as compared with other large hub airports 5 that are not slot-
controlled by examining aircraft size. We reviewed the reliability of the
data used in our analyses and concluded that they were sufficiently
reliable for our purposes. We also interviewed officials from DOT, FAA,
the Department of Justice (DOJ), MWAA, and the Port Authority of New
York and New Jersey, which operates the New York City area airports,
and interviewed representatives from airlines and others. We conducted
this performance audit from October 2011 to September 2012 in
accordance with generally accepted government auditing standards.
Those standards require that we plan and perform the audits to obtain
sufficient, appropriate evidence to provide a reasonable basis for our
findings and conclusions based on our audit objective. We believe that



5
 Federal law defines large hub airports as those commercial service airports that have at
least 1 percent of the passenger boardings. 49 U.S.C. § 40102(29).




Page 3                                                    GAO-12-902 Airport Slot Controls
             the evidence obtained provides a reasonable basis for our findings and
             conclusions based on our audit objectives. More details about our scope
             and methodology of our work are presented in appendix I.


             The Washington, D.C., region and New York City area airports vary with
Background   regard to their size, proximity to downtown areas, and type of service
             provided. In the Washington, D.C., region, Reagan National is located in
             Arlington County, Virginia, near the nation’s capital, while the two other
             major airports—Dulles and BWI—that are part of the Washington, D.C.,
             area air passenger market are larger and located farther from city
             centers. While Reagan National and BWI primarily provide domestic
             service, about one-fourth of enplanements at Dulles in 2011 were
             international. Reagan National and Dulles are owned by the federal
             government and are operated by MWAA under a 50-year lease that has
             been extended 30 years (1987-2067). Reagan National is subject to slot
             controls, but Dulles and BWI are not. In the New York City area,
             LaGuardia is located in Queens, near Manhattan, and primarily provides
             domestic air service, while JFK and Newark are larger airports that are
             located farther from Manhattan and provide more international service.
             Appendix II provides more detailed comparative information about the
             four slot-controlled airports, plus Dulles and BWI.

             On-time flight performance at Reagan National is comparatively better
             than the other slot-controlled airports and cancellations and delays at the
             New York City area airports have been among the worst in the nation.
             According to on-time arrival performance data that airlines reported to the
             Bureau of Transportation Statistics (BTS) for 2011, among 29 major
             airports, 6 Reagan National ranked 22nd, JFK ranked 25th, LaGuardia
             ranked 27th, and Newark ranked 29th. Because, according to FAA, one-
             third of the approximately 50,000 aircraft that FAA guides through the
             national airspace system every day move through the New York area at
             some point during a typical day, delays in this region can have a
             disproportionate impact on airspace and airport delays experienced
             throughout the rest of the system. According to 2012 research conducted
             by the MITRE Corporation for FAA, one-third of the flights in the country


             6
              BTS defines these airports as those within the continental United States that enplaned 1
             percent or more of the domestic scheduled-service passengers as of June 2010. Federal
             law defines large hub airports as those commercial service airports that have at least 1
             percent of the passenger boardings. 49 U.S.C. § 40102(29).




             Page 4                                                   GAO-12-902 Airport Slot Controls
are directly affected by delays in the New York City and Philadelphia
areas. Furthermore, the rate of late arrivals and flight cancellations—
another indicator of on-time performance—at the three New York City
area airports was worse than the system-wide average during the last 10
years. 7 The rate of late arrivals and cancellations at Reagan National,
however, was better than the system-wide average during 4 of the 10
years (see fig 1).




7
 The data indicated that the cancellation rate at LaGuardia was generally twice the
system-wide average during this period. See also GAO, National Airspace System: DOT
and FAA Actions Will Likely Have a Limited Effect on Reducing Delays during Summer
2008 Travel Season, GAO-08-934T (Washington, D.C.: July 15, 2008).




Page 5                                                GAO-12-902 Airport Slot Controls
Figure 1: Annual Percentage of Late Arrivals and Flight Cancellations at the Four Slot-Controlled Airports Compared with All
Airports from 2002 through 2011




                                         Note: In this figure, the percentage of delays and cancellations has been combined. The system-wide
                                         data include the four slot-controlled airports. Flights that arrived 15 minutes or more late compared
                                         with the scheduled time are counted as late arrivals. BTS data do not include international flights,
                                         which represented about half of the enplanements at JFK and about one-third of the enplanements at
                                         Newark in 2011.


                                         FAA previously relied on its statutory authority to manage airspace usage
                                         to implement slot controls through an FAA regulation called the High
                                         Density Rule, which capped the number of hourly arrivals and departures
                                         permitted at five designated “high density traffic airports”—LaGuardia,
                                         JFK, Newark, Washington National, 8 and Chicago O’Hare International—


                                         8
                                          In 1998, Washington National Airport was renamed Ronald Reagan Washington National
                                         Airport.




                                         Page 6                                                           GAO-12-902 Airport Slot Controls
and, among other things, required airlines to obtain slots for these
operations. The High Density Rule was implemented at Reagan National
and the three New York City area airports in the late 1960s to reduce
congestion and delay. Reagan National has remained slot-controlled
under the High Density Rule since 1969, although there have been
modifications to the operating limits and provisions of the rule throughout
the years. On the other hand, rules at the New York City area airports
have at times been relaxed or suspended as a result of statutory changes
and FAA actions. For example, because of lessened concerns about
congestion at Newark, FAA suspended slot controls at that airport from
1970 through 2008. In addition, in the 1990s, there was some perception
that slot controls were a barrier to improved service, in part, because new
airlines were unable to establish service at the slot-controlled airports.
Subsequently, in 2000, the Wendell H. Ford Aviation Investment and
Reform Act for the 21st Century required the High Density Rule to be
phased out at JFK and LaGuardia by January 1, 2007. These 2000
statutory provisions also directed DOT, in the interim, to grant exemptions
to the High Density Rule to allow for a limited number of flights operated
by new entrant airlines and for flights serving small communities. As a
result, airlines scheduled more flights at LaGuardia than the airport could
handle without unreasonable delays. FAA subsequently issued temporary
orders limiting scheduled operations at LaGuardia to 75 per hour in 2006
and at JFK and Newark to 81 per hour in 2008. 9 The limits at LaGuardia
were based on the limits prior to the 2007 phase-out. In setting the limits
on scheduled operations at JFK and Newark, FAA undertook an analysis
that included modeling the estimated level of delay associated with
various levels of operations. (As shown in fig. 1, from 2002 through 2011,
the rate of late arrivals and flight cancellations at the slot-controlled
airports and system-wide was greatest in 2007 when slot controls were
not in effect at JFK and Newark.)

Before FAA adopted slot allocation and transfer rules, in 1985, airlines
organized scheduling committees through which slots were allocated. In
1986, FAA, believing a secondary market in slots would address
concerns by new entrant airlines and smaller carriers about access to and
competition at High Density Rule-controlled airports, amended the rule to


9
 In 2009, FAA reduced the operating limits at LaGuardia to 71 scheduled operations per
hour to improve operational efficiency. The current slot allocations were kept, but any slots
returned to FAA, withdrawn for non-use, or unallocated during each affected hour would
be retired until the number of slots in that hour reaches 71.




Page 7                                                     GAO-12-902 Airport Slot Controls
allow airlines to buy and sell slots. Under the temporary orders at the New
York City area airports, imposed after the statutory expiration of the High
Density Rule at JFK and LaGuardia, FAA allocated slots to airlines based
on their current operating schedules at those airports in conjunction with
voluntary airline schedule-reduction agreements or, in the case of
LaGuardia, based on their current slot or slot exemptions holdings. To
encourage competition at these airports under the High Density Rule,
Congress also has periodically directed DOT to allocate slots exemptions
to airlines operating new service at slot-controlled airports. Generally,
once airlines have been allocated slots or slot exemptions, so long as
they comply with the rules, they may keep them indefinitely 10 and
consider them to be valuable assets. 11

Although slot controls were designed to manage congestion by limiting
the number of scheduled flight operations per hour at each of these
airports, the rules also contain a provision designed to encourage use of
airport capacity by requiring airlines to use their slots a minimum
percentage of the time. Beginning in 1986, FAA’s High Density Rule
required airlines at specified airports to use their slots at least 65 percent
of the time, but starting in 1992, they were required to use them at least
80 percent of the time. Although the 80 percent usage requirement is
applicable to all four slot-controlled airports, other provisions of the rules
differ between the airports, including

•    the hours when airlines are required to have a slot to operate a flight,
•    the time periods for which slots usage is measured and reported to
     FAA, and
•    whether airlines can buy or sell slots (see table 1).
In general, at Reagan National and LaGuardia, slots are allocated for
every day of the week and must be used at least 80 percent of time in a
fixed, 2-month reporting period (e.g., January-February, March-April,



10
  Slots are subject to FAA withdrawal, for example, if a slot is not used at least 80 percent
of the time over the applicable time period. In addition, slots are subject to FAA withdrawal
at any time to meet operational needs such as reduced airport capacity.
11
  Although airlines consider slots to be valuable assets, FAA slot regulations at 14 C.F.R.
§ 93.223(a), for example, provide that slots do not represent a property right but represent
an operating privilege subject to absolute FAA control. The FAA slot regulations applicable
to Reagan National also provide, in general, that slots may be bought, sold, leased, or
traded for slots at other slot controlled airports. 14 C.F.R. § 93.221(a).




Page 8                                                     GAO-12-902 Airport Slot Controls
etc.). 12 At JFK and Newark, slots are allocated for a specific day of the
week and must be used at least 80 percent of the time for that day during
a scheduling season. 13 Slots not used in accordance with the 80 percent
rule are subject to withdrawal by FAA. Additionally, because the slot rules
at the New York City area airports are contained in temporary FAA orders
and are not considered by FAA to be a long-term solution, slots may not
be bought or sold at those airports. 14 Slots may be leased and traded,
however, at any of the slot-controlled airports.




12
 Under FAA’s temporary order applicable to LaGuardia, slots are not allocated on
Saturdays.
13
  The summer season runs from the fourth Sunday in March to the fourth Saturday in
October. The winter season runs from the fourth Sunday in October to the fourth Saturday
in March.
14
  See, e.g., 73 Fed. Reg. 3510, 3513-14 (Jan. 18, 2008). In comparison to the temporary
FAA orders in place at the New York City area airports, the FAA High Density Rule
regulations applicable to slots at Reagan National provide, in part, that except as
otherwise provided, slots may be bought, sold, or leased. 14 C.F.R. § 93.221. The major
exception to this regulation at Reagan National is that slot exemptions may not be bought,
sold, leased, or otherwise transferred, except through an air carrier merger or acquisition.
49 U.S.C. 41714(j).




Page 9                                                     GAO-12-902 Airport Slot Controls
Table 1: Selected Slot Control Rule Provisions for the Four U.S. Slot-Controlled
Airports

                          Reagan
                          National       LaGuardia           JFK                  Newark
    Minimum slot          80 percent     80 percent          80 percent           80 percent
    usage
    requirement
    Hours that            All hours      6:00 a.m. to        6:00 a.m. to         6:00 a.m. to
    airport is slot-                     9:59 p.m.,          10:59 p.m.,          10:59 p.m.,
    controlled                           Eastern Time,       Eastern Time,        Eastern Time,
                                         Monday –            daily                daily
                                         Friday and
                                         12:00 p.m. to
                                         9:59 p.m.,
                                         Eastern Time,
                                         Sunday
    Usage         All days over 2- All days over 2- Each day of the               Each day of
    measured for… month reporting month reporting week over a                     the week over
                  period           period           scheduling                    a scheduling
                                                          a                              a
                                                    season                        season
    Authorization         Allowed        Not authorized      Not authorized       Not authorized
    to buy or sell
    slots
    Authorization         Allowed        Allowed             Allowed              Allowed
    to lease or
    trade slots
Source: FAA and FAA regulations
a
 Slots that are not operated because the airline returned them to FAA prior to that season’s return
deadline are not counted as unused for purposes of measuring slot usage. Airlines can return to FAA
slots during periods of a season that the airline does not intend to operate.


According to FAA officials, the differences in the rules for each airport are
largely a product of history. For example, when FAA replaced the High
Density Rule with temporary limitations on flight operations at LaGuardia
in 2007, it adopted most of the rule’s provisions including, most
importantly, the slot caps. Long-term changes were expected through a
proposed rule issued by FAA in the summer of 2006. Additionally,
according to an FAA official and International Air Transport Association
(IATA) representatives, because JFK and Newark support a large number
of international operations, the rules for those airports are more
consistent with IATA’s Worldwide Slot Guidelines. These guidelines
contain slot processes and policies for slot-controlled airports to facilitate
international operations to or between other slot-controlled airports
around the world.




Page 10                                                         GAO-12-902 Airport Slot Controls
FAA and DOT’s Office of the Secretary share responsibility for managing
slot control issues. FAA is responsible for administering the slot rules,
which primarily includes slot allocation, transfers, and monitoring and
enforcing airlines’ compliance with the 80 percent slot usage requirement.
DOT’s Office of the Secretary is responsible for awarding within-perimeter
and beyond-perimeter slot exemptions at Reagan National and
monitoring competitive aspects of the airlines’ slot holdings.

In 2006, FAA proposed a congestion management rule for LaGuardia that
would limit operations, encourage the use of larger aircraft, and assign
rolling expiration dates for slots. That notice of proposed rulemaking was
withdrawn and superseded in 2008 by congestion management rules at
JFK, LaGuardia, and Newark that would have created a market by
annually auctioning a limited number of slots in each of the first 5 years of
the rule, which had a 10-year term. In October 2009, before the terms of
the 2008 rules became effective, FAA rescinded the rules citing, among
other things, that the rulemakings had been highly controversial, and a
court challenge that resulted in a December 2008 stay of the rules by the
United States Court of Appeals for the District of Columbia Circuit. In
addition, because of the complexity of the issues, the impact of an
Omnibus Appropriations Act, 2009 provision, 15 and the possible impact of
the significantly changed economic circumstances on the slot auction
program, FAA believed it was better to rescind the rule rather than
propose to extend it. According to DOT’s Office of the Secretary and
FAA, their Slot and Management Transparency notice of proposed
rulemaking currently being drafted will propose to replace the current
temporary orders limiting scheduled operations at JFK, LaGuardia, and
Newark with a permanent rule to address congestion and delay, while
also promoting fair access and competition, including establishing a




15
   Pub. L. No. 111-8, Division I, Title I, §115, 123 Stat. 524, 921 (2009). The Omnibus
Appropriations Act, 2009 contained a provision prohibiting the Secretary of Transportation
from using such fiscal year 2009 appropriations to promulgate regulations or take any
action regarding the scheduling of airline operations at any commercial airport in the U.S.
if such regulation or action involved, among other things, auctioning or rights or
permission to conduct airline operations, the implementation of peak-period pricing or
other forms of congestion pricing, withdrawing a right or permission to conduct operations
at such an airport, or requiring a carrier to transfer involuntarily any such right or
permission to another.




Page 11                                                   GAO-12-902 Airport Slot Controls
secondary market that would allow airlines to buy and sell slots at the
New York City area airports. 16

In 2009, because airlines are prohibited from selling or purchasing slots
under the temporary order at LaGuardia, Delta Air Lines and US Airways
sought a waiver from DOT’s Office of the Secretary and FAA from this
prohibition to exchange slots at LaGuardia and Reagan National because
Delta wanted to establish a domestic hub at LaGuardia and US Airways
wanted to increase service at Reagan National. In 2011, DOT’s Office of
the Secretary and FAA together granted the joint waiver request that
permitted the transfer and sale of more than 300 slots between the two
airlines, provided that they divest a total of 48 slots (24 slot pairs) at
Reagan National and LaGuardia to new airline entrants and airlines
holding less than 5 percent of the slots at those airports. After the slot
swap, Delta Air Lines increased its share of slot holdings at LaGuardia
from 24 percent in April 2011 to 44 percent in May 2012 and US Airways
increased its share of slot holdings at Reagan National from 31 percent in
April 2011 to 54 percent in May 2012. The divestiture effort, conducted
through a blind sale open to new entrants and limited incumbents only,
resulted in JetBlue attaining 16 slots at LaGuardia for $32 million and 16
slots at Reagan National for $40 million, and WestJet attaining 16 slots at
LaGuardia for $17.6 million. Under the terms of the slot swap, US Airways
and Delta were to receive the proceeds from the sale of their divested
slots.

DOJ is responsible for enforcing antitrust laws for most sectors of the
economy, including the airline industry. According to DOJ officials, this
includes reviewing competitive aspects of airlines’ holdings at slot-
controlled airports. For example, transactions that increase concentration
of slot holdings may substantially lessen competition in violation of the
antitrust laws. In 2010, DOJ offered formal comments in support of the
tentative decision by DOT’s Office of the Secretary and FAA to condition
approval of the slots transfer between Delta Air Lines and US Airways
with regard to the airlines’ slot holdings at LaGuardia and Reagan



16
  From 1986 to 2006, when FAA managed congestion at JFK and LaGuardia under the
High Density Rule, airlines were allowed to buy and sell slots at these airports. However,
in comments to FAA on a proposed congestion management rule for LaGuardia, some
airlines indicated that a secondary market did not develop because the Wendell H. Ford
Aviation Investment and Reform Act for the 21st Century (AIR-21) granted exemptions
from the High Density Rule’s flight restrictions to some airlines for free.




Page 12                                                   GAO-12-902 Airport Slot Controls
National on the airlines divesting a certain number of slots. While DOJ did
not take action to block the slot swap, DOJ officials told us in June 2012
that they are still reviewing competitive aspects of the airlines’ slot
holdings at Reagan National following completion of the revised slot
swap.

In addition to slot controls, Reagan National and LaGuardia are subject to
perimeter rules, which limit nonstop flights serving these airports to a
distance of 1,250 and 1,500 miles, 17 respectively. While LaGuardia’s
perimeter rule is imposed by the airport’s operator, the perimeter rule at
Reagan National is statutory. 18 Two federal laws—the Wendell H. Ford
Aviation Investment and Reform Act for the 21st Century (AIR-21) 19 in
2000 and the Vision 100-Century of Aviation Reauthorization Act (Vision
100) 20 in 2003—increased the number of slot exemptions at Reagan
National and for the first time required exemptions to be granted to the
Reagan National perimeter rule known as “beyond-perimeter” slot
exemptions. These laws required the Secretary of Transportation to grant
exemptions to the existing rules, adding 44 slot exemption operations (22
round trips) per day during a 15-hour period and requiring 24 (12 round
trips) of these slot exemptions to be granted for beyond-perimeter
flights. 21 In February 2012, the FAA Modernization and Reform Act of
2012 22 required the Secretary of Transportation to grant 8 additional
beyond-perimeter slots exemptions (4 round trips) at Reagan National.
The 2012 act, similar to AIR-21 in 2000 and Vision 100 in 2003, required
DOT, in awarding the slot exemptions, to consider specified criteria in
granting beyond-perimeter exemptions. The 2012 act required the
consideration of issues such as their impact on providing network benefits
(connecting flights) outside of the perimeter and passenger service to
small communities. In May 2012, to implement the requirements of the


17
  The perimeter rule for LaGuardia provides an exception for nonstop flights to and from
Denver because there was existing service to Denver when the rule went into effect.
18
   Pub. L. No. 99-500, 100 Stat. 1783, 1783-375 (1986); Pub. L. No. 99-591, 100 Stat.
3341, 3341-376 (1986).
19
 Pub. L. No. 106-181, 114 Stat. 61 (2000).
20
 Pub. L. No. 108-176, 117 Stat. 2490 (2003).
21
  The statutes also authorized 20 additional slot exemptions within the 1,250-mile
perimeter.
22
 Pub. L. No. 112-95, 126 Stat. 11 (2012).




Page 13                                                   GAO-12-902 Airport Slot Controls
FAA Modernization and Reform Act, DOT awarded the beyond-perimeter
slot exemptions to airlines proposing to provide nonstop service between
Reagan National and Austin, Texas; Portland, Oregon; San Francisco,
California; and San Juan, Puerto Rico. The act also authorized the four
incumbent carriers at Reagan National to each convert two slots that
were being used for flights to large hubs within the perimeter, to two slot
exemptions for nonstop flights beyond the perimeter. Accordingly, four
round trips were suspended to large hubs and four new round trips were
initiated instead to Los Angeles, San Diego, and San Francisco,
California, and Salt Lake City, Utah (using a total of eight slot exemptions)
(see fig. 2).




Page 14                                          GAO-12-902 Airport Slot Controls
Figure 2: Nonstop Beyond-Perimeter Slot Exemption Destinations from Reagan National from 2000 through 2012




                                       Page 15                                            GAO-12-902 Airport Slot Controls
                      Appendix III provides more detailed information about the implementation
                      of slot controls and perimeter rules, the proposed Slot Management and
                      Transparency rule, and the slots transfer between Delta Air Lines and US
                      Airways.


                      Reagan National has sufficient airside (runway) capacity to accommodate
Reagan National Can   the new nonstop beyond-perimeter flights. The airport’s terminal and
Accommodate New       landside facilities are more constrained, but with some improvements and
                      adjustments, it will also have sufficient terminal capacity. The new flights
Beyond-Perimeter      are also likely to have limited effects on passenger security screening
Flights, Which Are    wait times, passenger traffic at the other Washington, D.C., area airports,
Likely to Have a      MWAA’s ability to service its debt, and aircraft noise. In addition, the low
                      cost airlines that will be operating half of the new beyond-perimeter flights
Limited Effect on     at Reagan National will increase airline competition on those routes.
Other Washington,
D.C., Area Airports
Airside Capacity      Reagan National is routinely operating below the maximum authorized 67
                      hourly takeoffs and landings (slots) and the 67.6 hourly takeoffs and
                      landings that a 2011 FAA capacity analysis indicated could be
                      accommodated based only on average runway capacity. 23 While 67 is the
                      maximum number of hourly takeoffs and landings authorized in any one
                      hour, the number of slots and slot exemptions authorized on a daily basis
                      is generally less than 67 per hour, on average. Table 2 shows the
                      maximum number of hourly takeoffs and landings allowed at Reagan
                      National, broken down by the type of aircraft operations.




                      23
                        MITRE, an FAA contractor, conducted this analysis in August 2011. For information
                      about how MITRE analyzes airport capacity, see Appendix V of GAO, National Airspace
                      System: Setting On-Time Performance Targets at Congested Airports Could Help Focus
                      FAA’s Actions, GAO-10-542 (Washington, D.C.: May 26, 2010).




                      Page 16                                               GAO-12-902 Airport Slot Controls
Table 2: Hourly Aircraft Operations (Takeoffs and Landings) Allowed at Reagan
National

                                                                               Maximum number of takeoffs and
    Type of aircraft operations                                                landings (slots) allowed per hour
    Air carriera                                                                                                 37
                                            b
    Commuter aircraft operations                                                                                 11
    General aviation and unscheduled flights                                                                     12
    Slot exemptionsc                                                                                               5
           d
    Slides (allowing slots to be used in different time                                                            2
    periods)
    Totale                                                                                                       67
Source: 14 C.F.R. § 93.123, 49 U.S.C. § 41718(c)(2)(A)(ii), 49 U.S.C. § 41714(d)(1) and FAA.
a
 Under FAA slot control regulations applicable to Reagan National, “air carrier slots” are intended for
operations with turboprop and reciprocating engine aircraft with 75 or more seats or turbojet aircraft
with 56 or more seats. However, air carrier slots may be used for operations using aircraft of any size.
FAA officials estimated that 30 to 40 percent of air carrier slots are operated in commuter aircraft slots
based on a review of published schedules and the number of commuter aircraft operated above the
commuter slot quota.
b
 Under FAA slot control regulations applicable to Reagan National, “commuter slots” may be used
only for operations with turboprop and reciprocating engine aircraft with no more than 76 seats or
turbojet aircraft with fewer than 56 seats.
c
    These are exemptions to the hourly restrictions at Reagan National that are authorized by statute.
d
 Under 49 U.S.C. § 41714 (d)(1), the Secretary of Transportation is authorized to grant certain
exemptions from specified slot rules applicable to Reagan National. For example, under this authority
the Secretary may grant exemptions from hourly limitations on slots at Reagan National but may not
increase the number of aircraft operations in any 1-hour period by more than 2 operations. FAA may
not grant more than 2 exemptions for a 1-hour period. FAA uses this authority to consider requests
for what are referred to as a grant “slot-slides,” which would allow slots allocated for certain hours to
be reassigned and used in a different slot periods.
e
 Regulatory limits are 60 per hour. However, slot exemptions are authorized by statute to increase
hourly operations by up to 5 in an hour, but only between 7:00 a.m. and 9:59 p.m. In addition, slot
slides may also increase operations but under statute (1) may not result in an increase in the total
number of slots per day, (2) may not result in an increase in the total number of slots from 7:00 a.m.
to 9:00 p.m., and (3) may not increase operations in any 1-hour period by more than two operations.
FAA has not modeled the delay impact of 67 operations in each hour.


Our analysis of FAA data found that a maximum of about 53 slots was
allocated and used per hour at Reagan National, compared with the 67
hourly takeoffs and landings that are generally permitted. Specifically,
FAA reported that a maximum of 52 slots in one hour was allocated for
airline and commuter operations in April 2012 and an average of about 1
aircraft operation per hour was used for general aviation or unscheduled




Page 17                                                                              GAO-12-902 Airport Slot Controls
                                          aircraft operations from May through August 2011. 24 The new beyond-
                                          perimeter flights will cause the largest increase in the maximum number
                                          of slot allocations for airline and commuter aircraft operations from 4:00
                                          p.m. to 6:59 p.m., when the number of slot allocations will increase by 2
                                          per hour. 25 Table 3 shows the impact of the new beyond-perimeter flights
                                          on slot allocations during peak and nonpeak hours.

Table 3: Impact of New Beyond-Perimeter Exemptions on Average Peak and Non-Peak Hour Slot Allocations at Reagan
National for the Summer 2012 Season

                                                          Average number of
                                                     slots allocated per hour                               Average number of slots
                                                                             a
                                                          during peak hours            Average number of allocated per hour during
                            Average number of              including the new            slots allocated per nonpeak hours including
Type of aircraft       slots allocated per hour            beyond-perimeter           hour during nonpeak the new beyond-perimeter
                                              a
operations                  during peak hours                         flightsb                       hours                  flights
                                                                                                                                   b


Airline and commuter                      51.3                                52.5                        45.6                              45.7
operationsc
                                          Source: GAO analysis of FAA data.
                                          a
                                              Peak hours used in this analysis are from 7:00 a.m. to 10:00 a.m. and from 4:00 p.m. to 7:00 p.m.
                                          b
                                              New beyond-perimeter slots that Congress authorized in February 2012.
                                          c
                                           Based on FAA data regarding slot holders for Thursdays for the summer 2012 season as of April 5,
                                          2012.


                                          Much of the difference between the maximum number of hourly aircraft
                                          operations authorized and the average number of slots allocated and
                                          used can be attributed to a substantial decrease in general aviation and
                                          unscheduled aircraft operations at Reagan National after the September
                                          11, 2001, terrorist attacks. After the 2001 terrorist attacks, TSA imposed
                                          security restrictions on general aviation operations operating at Reagan
                                          National, such as requiring an armed security officer to be on board all



                                          24
                                            Based on an FAA analysis of unscheduled operations from May through August 2011
                                          indicating that there was an average of 17.3 daily general aviation operations weekdays
                                          from 6:00 a.m. to 10:59 p.m. We reported in 2007 that there were about 200 general
                                          aviation and unscheduled flights at Reagan National in 2006, or an average of one-half of
                                          a flight operation per day. The number of general aviation and unscheduled flights has
                                          increased at Reagan National since 2006 in part because TSA updated security
                                          procedures that made more general aviation flights possible, for example, by expanding
                                          the number of airports from which flights can emanate.
                                          25
                                            From 4 p.m. to 4:59 p.m., the number of slot allocations will increase from 52 to 54, and
                                          from 5 p.m. to 5:59 p.m. and from 6 p.m. to 6:59 p.m., the number of slot allocations will
                                          increase from 51 to 53.




                                          Page 18                                                             GAO-12-902 Airport Slot Controls
                    flights. Before the attacks, there were about 210 general aviation and
                    unscheduled flights per day at Reagan National. By comparison, FAA
                    indicated that there were about 17 general aviation and unscheduled
                    flights per day at Reagan National in 2011. 26 Of the 60 authorized hourly
                    slots at Reagan National, excluding any slot exemptions, 12 are allocated
                    for general aviation aircraft operations or other aircraft that do not operate
                    on a scheduled basis. However, from May through August 2011, an
                    average of about 1 of the 12 allocated slots was used per hour. Although
                    airlines with large commercial aircraft could not necessarily use the slots
                    reserved for smaller general aviation aircraft and unscheduled flights,
                    which may use the airport’s shorter runways and do not generally use
                    airport gates and terminal infrastructure, the reduced amount of overall
                    general aviation traffic leaves some airport capacity unused. In addition,
                    some slots are available at Reagan National during nonpeak hours
                    because airlines are primarily interested in obtaining slots at peak hours.
                    FAA indicated that, for July and August 2012, 34 airline and 12 commuter
                    slots were available, all at early morning (6:00 a.m.) or late evening
                    (10:00 p.m. or 11:00 p.m.) hours. 27


Terminal Capacity   Although Reagan National has sufficient airside capacity to accommodate
                    the new beyond-perimeter flights, the airport has more limited terminal
                    capacity, which includes space for gates, ticket counters, baggage
                    handling, and security screening. As of May 2012, two of the new
                    beyond-perimeter round trips (those operated by Jet Blue and Southwest
                    Airlines) are expected to be operated out of Terminal A, the airport’s
                    original terminal that opened in 1941, and two of the new beyond-
                    perimeter round trips (those operated by Virgin America, which is the
                    airport’s only new airline entrant among the operators of the new beyond-
                    perimeter flights, and Alaska Airlines) are expected to be operated out of
                    the newer Terminal B/C, which opened in 1997. To help accommodate as
                    many as 1,245 additional daily passengers who could fly on the new
                    beyond-perimeter flights, MWAA is planning to expand baggage and


                    26
                      Based on an FAA analysis of general aviation and unscheduled weekday operations
                    from May through August 2011.
                    27
                      Three slots each were available at 6:00 a.m. and 10:00 p.m. and 40 were available at
                    11:00 p.m. These numbers exclude other slots that are allocated on an ad hoc day-of
                    basis following a request from the operator and one commuter slot that is available at 9:00
                    p.m. and another at 10:00 p.m. that are allocated on a temporary basis and subject to
                    recall by FAA.




                    Page 19                                                   GAO-12-902 Airport Slot Controls
                     security screening space in Terminal A and complete those
                     improvements by the summer of 2013. An MWAA official said these
                     improvements were already needed to address increased passenger
                     volume and complexity of operations in Terminal A. For example, seven
                     airlines, each with its own needs for space in Terminal A, 28 are now
                     operating separate, uncoordinated schedules, compared with a few years
                     ago when one dominant airline and a few smaller airlines were operating
                     there. No improvements will be needed to accommodate the new beyond-
                     perimeter flights in Terminal B/C. MWAA estimated that improvements in
                     Terminal A will cost about $45.6 million, which will be financed through
                     bond funds and airline leases. However, an MWAA official added that
                     these improvements do not allow for any additional growth in Terminal A,
                     which could, if needed, require replacing the terminal at a cost of several
                     hundred million dollars.


Gate Usage           Based on current gate usage rates at Reagan National, on average,
                     gates will be fully used. MWAA data indicated that airlines operating at
                     Reagan National will average 7.7 turns per day for each gate. This is in
                     the range of six to eight daily turns per gate that was defined as full gate
                     utilization in a 2010 Transportation Research Board study 29 on common-
                     use gates. That report also indicated that full gate utilization may be
                     affected by noise restrictions limiting the hours of operations or other
                     factors that may vary according to the airline using the gate. Exceeding
                     full gate utilization could lead to flight delays.


Passenger Security   We asked TSA to estimate the impact of the new beyond-perimeter flights
Screening            on passenger security screening wait times at Reagan National using the
                     agency’s security screening throughput model. TSA’s model suggested
                     that the impact of the new beyond-perimeter flights on passenger security




                     28
                       The official said that although these airlines need the same type of space to function,
                     e.g., ticket counters, gates, and baggage systems, each airline needs a different amount
                     of space, so economies of scale cannot always be maximized.
                     29
                      Transportation Research Board of the National Academies, Reference Guide on
                     Understanding Common Use at Airports, (Washington, D.C.: 2010).




                     Page 20                                                   GAO-12-902 Airport Slot Controls
                     screening wait times would be minimal. 30 However, an MWAA official said
                     that until improvements to the security screening space in Terminal A are
                     completed next summer, passengers may need to wait longer to be
                     screened at that checkpoint.

Airport Access and   The additional beyond-perimeter flights are not expected to affect access
Parking              to the airport or parking, considering that the maximum number of
                     additional passengers that the new flights could generate represents
                     about 5 percent of the airport’s total daily enplanements. According to a
                     survey of air passengers in the Washington, D.C., and Baltimore area, 16
                     percent of Reagan National passengers arrive by Metrorail, which is
                     among the highest public transportation rates in the nation for airport
                     access, 74 percent of passengers arrive by car (private and rental) or taxi,
                     and the remaining 10 percent arrive by shuttle bus or Metrobus. 31 MWAA
                     provided data showing that since parking facilities were expanded since
                     2009, the number of times that lots have been filled to capacity has
                     dropped considerably. The data indicated that in 2011, the airport’s
                     parking lots were filled to capacity approximately 38 times, compared with
                     2009, when they were filled to capacity a total of 353 times. According to
                     MWAA, from January through May 2012, the parking lots were filled to
                     capacity 13 times. However, an MWAA official said that the impact of the
                     new beyond-perimeter slot exemptions should not be considered in
                     isolation and that the sale of slots to JetBlue under the Delta-US Airways
                     slots swap earlier in 2011 has had a more pronounced effect. For
                     example, the official said that since more JetBlue flights began operating
                     out of Terminal A in June 2012, the economy parking lot has been filled to
                     capacity an average of 4 days a week. In providing comments on a draft
                     of this report, MWAA indicated that during the first 8 months of 2012, the
                     economy parking lot has been filled to capacity more than 60 times, which



                     30
                       We asked TSA to assess the impact of the new beyond-perimeter flights on wait times
                     using its security screening throughput model. As inputs to the model, we provided TSA
                     with the flight times for the new beyond-perimeter flights, data on the maximum number of
                     passengers on those flights, and the screening checkpoints that the passengers would
                     use. Although the results from this model suggested that any increase in wait times at
                     Reagan National would be minimal, we are not reporting the precise estimates of wait time
                     changes because they would imply a greater level of precision than the model could
                     reasonably provide.
                     31
                       National Capital Region Transportation Planning Board and Metropolitan Washington
                     Council of Governments, Draft 2011 Washington-Baltimore Regional Air Passenger
                     Survey, (Washington, D.C.: May 2011).




                     Page 21                                                  GAO-12-902 Airport Slot Controls
                           was more than double the number of times that the lot was forced to
                           close in 2011.

Impact on Other Airports   Our analysis found that the beyond-perimeter flights authorized in
in the Washington, D.C.,   February 2012 could add between zero (assuming that no new
Area                       passengers shift from using other airports) and a maximum of 1,245 daily
                           enplanements (departing passengers) at Reagan National. 32
                           Longstanding policy designed to guide the operation of Reagan National
                           and Dulles, as well as agency rules designed to implement such policy,
                           have included the roles of Reagan National in providing short-haul
                           domestic service and Dulles in providing all types of service. 33 In addition,
                           under the policy, MWAA is to promote the “better utilization” of Dulles,
                           which was underutilized for many years after opening in 1962. During our
                           review, MWAA and BWI officials expressed concern about the possible
                           impact of the new beyond-perimeter flights at Reagan National on
                           passenger traffic at Dulles and BWI. We found, on the basis of 2011
                           enplanement data, that if all of the new beyond-perimeter flights are 100
                           percent full, the number of enplanements on those flights would represent
                           an increase of about 5 percent of annual enplanements at Reagan
                           National. Correspondingly, Dulles or BWI would lose a maximum of 4.1
                           percent of their enplanements, respectively, if all the passengers changed
                           from using either of these two airports to using Reagan National.
                           Therefore, if all the passengers changed from using Dulles or BWI, the
                           new beyond-perimeter flights are likely to have some impact on
                           passenger traffic at these airports. However, it is extremely unlikely that



                           32
                             In calculating the maximum number of enplanements, we included the eight new daily
                           beyond-perimeter departures and assumed that each daily flight used the aircraft specified
                           in the airline’s application for the slot exemption and that those aircraft were 100 percent
                           full. We did not include the passenger impact of the within-perimeter flights that were
                           converted to beyond-perimeter slot exemptions because there were too many unknown
                           factors, such as whether the passengers on those flights were connecting or originating at
                           Reagan National or the load factors on them.
                           33
                             In 1981, FAA adopted rules to implement FAA’s Metropolitan Washington Airports’
                           policy (45 Fed. Reg. 62398 (Sept. 18, 1980)) to guide the future operations and
                           development of Reagan National and Dulles. This implementing Metropolitan Washington
                           Airport rule provided, among other things, that the perimeter rule would maintain the long-
                           haul nonstop service at Dulles and be consistent with the roles proposed for Reagan
                           National as a short/medium-haul facility and for Dulles as an unrestricted facility available
                           for all types of operations. 48 Fed. Reg. 58036 (Nov. 27, 1981). In 1986, the federal law
                           creating MWAA provided that the regulations of MWAA’s predecessor agency shall
                           become the regulations of MWAA. Pub. L. No. 99-500, 100 Stat. 1783, 1783-375 (1986);
                           Pub. L. No. 99-591, 100 Stat. 3341, 3341-376 (1986).




                           Page 22                                                     GAO-12-902 Airport Slot Controls
all of these passengers would shift from using one airport; instead, any
loss likely would be shared between Dulles and BWI. In addition, the
airports are not perfect substitutes for Reagan National, so it is unlikely
that all the passenger traffic would shift from Dulles or BWI. But, even in
the worst case, a 4.1 percent change is less than the growth expected at
these airports over the next 2 or 3 years. The actual impact of the new
beyond-perimeter flights at Reagan National on passenger traffic at the
other Washington, D.C., area airports cannot be quantified because all of
the new flights did not start until August 2012 and because of unknown
factors such as how competing flights will be priced, how the new direct
flights at Reagan National will draw passengers from the area airports,
how other airlines may adjust their schedules, and the extent to which the
new flights may generate new passenger demand on those routes.

We found no evidence that the previously added beyond-perimeter flights
at Reagan National had affected passenger traffic or airfares at the
surrounding airports. We found two cases—one in 2005 and another in
2012 34—when airlines discontinued or planned to discontinue service
between Dulles and beyond-perimeter destinations after they were
awarded new beyond-perimeter slot exemptions at Reagan National, but
any shift in passenger traffic, as noted above, represented about 4
percent of overall passenger traffic. (Figures 8 through 31 in appendix IV
show trends in average annual passenger traffic and airfares at BWI and
Dulles for destinations granted beyond-perimeter slot exemptions at
Reagan National under AIR-21 and Vision 100 between 2000 and 2004).
Moreover, any deviation in airfares after the initiation of the slot
exemptions at Reagan National could be a result of many factors, such as
airline bankruptcies and mergers. Additionally, the presence of low cost
airlines on the newly-awarded beyond-perimeter flights will increase
competition on those routes and could help reduce airfares to those
destinations. However, of the four new nonstop beyond-perimeter
destinations, only one low cost airline (Virgin America) will be competing
with a legacy airline (United Airlines) in providing service between
Reagan National and San Francisco.




34
  The 2005 example involved Frontier Airlines and the 2012 example involved Delta Air
Lines.




Page 23                                                 GAO-12-902 Airport Slot Controls
Financial Impact on   If the new beyond-perimeter flights cause passengers to shift from using
MWAA                  Dulles to using Reagan National, it could increase per passenger airline
                      costs of using Dulles, which, if substantial, could cause the airlines to
                      reduce service at the airport. To better understand the potential financial
                      impact of possible changes in passenger traffic resulting from the new
                      beyond-perimeter exemptions, we asked MWAA to estimate the flights’
                      impact on the cost per enplaned passenger, 35 which is a generalized
                      measure of airline’s airport charges, assuming a worst case scenario.
                      Assuming all of the passengers on the new beyond-perimeter flights
                      shifted from using Dulles to using Reagan National—the worst case
                      scenario—and all other factors stayed the same, MWAA’s analysis found
                      that the cost per enplaned passenger would increase at Dulles in 2013 by
                      $1.10, or 3.9 percent, and increase in 2014 by $1.51 or 5.7 percent, and
                      decrease at Reagan National in 2013 by $0.58, or 4.4 percent, and
                      decrease in 2014 by $0.89, or 6.5 percent (see table 4). These results
                      show that because the same costs, most of which are fixed costs, would
                      be spread among fewer passengers at Dulles and more at Reagan
                      National so that the cost per enplaned passenger would change
                      accordingly. However, even in the worst case, the changes would be
                      relatively small.




                      35
                         Cost per enplaned passenger is defined as all landing fees, airside usage charges, fuel
                      flowage fees, terminal rents, and other terminal payments to an airport, divided by the
                      number of enplaned passengers. We did not assess the possible impact of the new
                      beyond-perimeter flights on the cost per enplaned passenger at BWI because of
                      differences in how airport authorities calculate those numbers. Airlines use cost per
                      enplaned passenger as a standardized measure of relative unit costs to compare costs
                      among airports.




                      Page 24                                                   GAO-12-902 Airport Slot Controls
Table 4: Projected Impact of the New Beyond-Perimeter Flights at Reagan National on the Cost per Enplaned Passenger in
2013 and 2014

                                             Projected 2013 cost per                                     Projected 2014 cost per
                                            enplaned passenger with                                     enplaned passenger with
                    Projected 2013 cost per   new beyond-perimeter              Projected 2014 cost per   new beyond-perimeter
Airport                enplaned passenger                     flights              enplaned passenger                     flights
Reagan National                    $13.23a                           $12.65                    $13.70b                      $12.81
                                             c
Dulles                             $27.99                            $29.09                    $26.61d                      $28.12
                                        Source: MWAA
                                        a
                                            Based on projected enplanements of 9,765,000.
                                        b
                                            Based on projected enplanements of 9,850,000.
                                        c
                                            Based on projected enplanements of 12,095,000.
                                        d
                                            Based on projected enplanements of 12,415,000.


                                        In recent years, the cost per enplaned passenger has increased
                                        substantially at Dulles, owing to a major capital development project at
                                        the airport, while it has decreased slightly at Reagan National. From 2010
                                        to the first half of 2012, the cost per enplaned passenger at Dulles
                                        increased from $17.16 to $26.03 or 51.7 percent, but decreased at
                                        Reagan National from $12.84 to $12.74, or 0.8 percent. Airline
                                        representatives generally told us that an airport’s cost as measured by
                                        cost per enplaned passenger is among many factors that they consider
                                        when deciding whether to provide service there. However,
                                        representatives from one large airline said that if the cost per enplaned
                                        passenger at Dulles and other costs such as fuel continue to increase, it
                                        would consider reducing or cancelling its service there. MWAA officials
                                        also said that, given the recent substantial increases in the cost per
                                        enplaned passenger at Dulles, some airlines may stop serving the airport,
                                        particularly in providing domestic service, and that an increase of even
                                        one dollar in the cost per enplaned passenger is significant. 36 In a recent
                                        credit-rating report on MWAA prepared in September 2011, before the
                                        new beyond-perimeter flights were awarded, one agency had a negative
                                        outlook on the authority’s rating because of expected significant increases
                                        in the cost per enplaned passenger and lower debt service coverage over
                                        the next 5 years. However, two other credit-rating agencies reported in



                                        36
                                          For comparison purposes, MWAA provided data showing that the costs per enplaned
                                        passenger was substantially higher at Dulles than at most other large hub airports.
                                        However, these data were not directly comparable because airports vary in terms of what
                                        airlines and airports pay for, which affects the costs per enplanement.




                                        Page 25                                                     GAO-12-902 Airport Slot Controls
September 2011 their ratings outlooks for MWAA were stable, citing the
strong air passenger market in the Washington, D.C., area. Moreover,
recent increases in the cost per enplaned passenger at Dulles were not
due to significant shifts in passenger traffic. From 2010 through 2011,
enplanements at Dulles decreased by about 2 percent, while they
increased at Reagan National by about 3.6 percent. However, from 2010
and 2011, international enplanements—which generate higher airport
revenues because airline fees are based in part on aircraft weight, and
aircraft used for international flights are generally larger and heavier—
increased at Dulles by more than 4 percent and represented 28 percent
of all enplanements in 2011. In addition, in a forecast of airport activity
from 2011 to 2040, FAA ranked Dulles second among 30 major U.S.
airports in terms of projected increases in enplanements, with a projected
annual growth rate of 3.9 percent. FAA projected that from 2011 to 2040,
enplanements at Dulles will increase by 215 percent, compared with 34
percent at Reagan National and 110 percent at BWI. 37

Under current lease agreements between MWAA and its incumbent
airlines, debt costs are not shared between the two airports which could
make servicing the much larger debt at Dulles difficult if revenue declines
at Dulles were substantial. MWAA currently maintains $5.2 billion in long-
term debt. 38 Much of MWAA’s debt was incurred for recent major capital
improvements at Dulles, such as an automated train system 39 connecting
the main terminal to gates and construction of a fourth runway. According
to an MWAA official, about $3.8 billion of the $5.2 billion in total debt is
allocated to Dulles and about $1.4 billion is allocated to Reagan
National. 40 MWAA officials are concerned about the financial impact of
the new beyond-perimeter flights on the authority’s ability to service its


37
  Federal Aviation Administration, Terminal Area Forecast Summary, Fiscal Years 2011-
2040.
38
  According to Moody’s Investors Service, Inc., MWAA is among five U.S. airport
authorities with the most debt. In 2011, MWAA incurred $222 million in interest expense.
39
   The automated passenger train system at Dulles, called Aerotrain, is separate from the
project to extend Metrorail to Dulles. MWAA aviation funds are responsible for 4.1 percent
of the total cost of the Metrorail extension to Dulles. However, according to MWAA, all
bonds for the corridor project are backed by a separate MWAA toll road credit that is
firewalled from the MWAA’s aviation credit.
40
  According to an MWAA official, debt is not issued on an airport-by-airport basis; but
rather as airport-system revenue-bond debt for both airports. However, MWAA tracks
where the debt proceeds are applied by airport.




Page 26                                                    GAO-12-902 Airport Slot Controls
                 debt at Dulles, because revenues and debt service charges for Reagan
                 National and Dulles are maintained separately. 41 According to MWAA
                 officials, in the event that one of the two airports could not pay its debt,
                 the lease agreements contain an “extraordinary coverage protection
                 payments” provision designed to allow for the sharing of debt service
                 between the two airports. More specifically, according to MWAA officials,
                 if Dulles could not continue paying its debt service, the use of the
                 extraordinary coverage provision by the authority could have the impact
                 of mitigating airline rates and charges at Dulles and, in turn, could
                 mitigate a cost per enplaned passenger increase at Dulles and cause a
                 corresponding increase in the cost per enplaned passenger at Reagan
                 National. However, an MWAA official said that the airports authority does
                 not expect that this provision will be triggered before the current lease
                 agreements expire in the fall of 2014. In addition, the MWAA official said
                 the inclusion of this provision in the new leases is a matter that will be
                 negotiated with the airlines.

                 With respect to MWAA’s authority to generate additional revenues that
                 could help mitigate the cost per enplaned passenger in the future, the
                 FAA Modernization and Reform Act of 2012 amended the definition of the
                 “airport purposes” that MWAA may engage in to include “a business or
                 activity not inconsistent with the needs of aviation that has been approved
                 by the Secretary.” 42 An MWAA official said that, pursuant to this authority,
                 the airport authority plans to develop land near Dulles for warehouses,
                 cargo-related integrators, freight forwarders, and general aviation.
                 However, the official added that the property first needs to undergo an
                 environmental impact review, which could take 2 to 3 years, and that new
                 revenues from the development are not expected to have an impact on
                 airline lease rates in the next 3 to 7 years.


Aircraft Noise   Although some members of the community surrounding Reagan National
                 expressed limited concerns about the additional aircraft noise that the
                 new beyond-perimeter flights could generate, FAA and airport officials



                 41
                   MWAA receives revenues from landing fees, which are based on aircraft weight; airport
                 leases with airlines; and concessions. Although non-aeronautical revenues, such as
                 concessions, are not maintained separately by airport, MWAA officials said revenues are
                 not sufficient to cover much debt.
                 42
                  Pub. L. No. 112-95, 126 Stat. 11, 92 (2012).




                 Page 27                                                 GAO-12-902 Airport Slot Controls
indicated that the relatively small number of flights being added will not
have a significant overall effect on aircraft noise. Concerns about noise
affecting nearby residential areas were a primary reason why Reagan
National became subject to the perimeter rule in the 1960s, when aircraft
were noisier than they are today. We interviewed individuals from
community organizations concerned about aircraft noise at Reagan
National as well as officials from the Washington Council of
Governments, the Arlington County (Va.) Council, MWAA, and FAA about
the possible impact on aircraft noise from the new beyond-perimeter
flights. Individuals from community organizations said they were
concerned about aircraft noise at the airport as a whole and that
continuing to add more beyond-perimeter flights has a cumulative
negative effect. The Washington Council of Governments opposed
changing the slot or perimeter rules at Reagan National, which the
organization indicated are important tools in helping to manage the
volume of air traffic at the airport and the impact on aircraft noise on the
surrounding community. In addition, the Arlington County Council official
said that people would notice noise from the new flights if they were
added in the early morning or late evening hours. However, MWAA and
FAA officials said that the additional beyond-perimeter flights are not
expected to cause a noticeable increase in aircraft noise because they
constitute only 8 of about 800 daily flights at Reagan National, or about 1
percent, of the total daily aircraft operations. They also noted that aircraft
make less noise than they did in the past and that the types of aircraft
being used or planned to be used for the new beyond-perimeter flights
(Airbus 319s and 320s and Boeing 737s) tend to be newer and quieter.
An MWAA noise abatement official also said and that in February 2012, a
new advanced navigational procedure was implemented for northbound
departures at Reagan National, which is designed for aircraft to follow the
river and fly less directly over homes. FAA and MWAA officials also said
that because of the technological advances, noise contours, which
highlight areas of significant aircraft noise exposure surrounding the
airport, have generally shrunk and that while noise from the new beyond-
perimeter flights would be noticeable individually, the relatively small
number of new beyond-perimeter flights being added as a whole is not
significant enough to require updating the airport’s noise compatibility
study. 43




43
  FAA supports airports’ efforts to mitigate aviation noise mainly through its voluntary
noise compatibility program, known as the Part 150 Noise Compatibility Program, which




Page 28                                                   GAO-12-902 Airport Slot Controls
                          MWAA, which monitors noise at Reagan National, provided data showing
                          that in 2011, it received 505 complaints from 117 callers, 255 of which
                          came from 5 callers, compared with 197 complaints in 2010 and 99
                          complaints in 2009. An MWAA official attributed the increased number of
                          complaints to construction on the main runway last year, which is now
                          complete. During construction, pilots took a different flight path over
                          residential areas at night, rather than following the normal path along the
                          Potomac River.



Slot Control Rules
Help Manage
Congestion, but
Allowing Airlines Not
to Use Some Slots
May Hinder
Competition
Slot Control Rules Help   Even though the New York City area airports still experience some of the
FAA Manage Congestion     worst delays in the country, the delay situation at these airports would be
                          far worse in the absence of slot controls. 44 The best example of what
                          would occur without slot controls at New York City area airports occurred
                          in 2000 when a then-new federal law required the Secretary of
                          Transportation to grant exemptions to the High Density Rule at JFK and
                          LaGuardia to new entrant airlines and for service to small communities
                          with small regional jets. 45 Soon after, qualifying airlines requested more

                          provides guidance to airports on how to develop noise compatibility plans to mitigate noise
                          on and around airports. MWAA last conducted a noise compatibility study on Reagan
                          National in 2004. See also the forthcoming GAO report, Airport Noise Grants: FAA Needs
                          to Better Ensure Project Eligibility and Improve Strategic Goal and Performance
                          Measures, GAO-12-890.
                          44
                            Slot controls are one type of demand management measure used to help manage
                          congestion. Other demand management measures include congestion pricing and slot
                          auctions.
                          45
                            Wendell H. Ford Aviation Investment and Reform Act for the 21st Century (AIR-21), Pub.
                          L. No. 106-181 (2000). AIR-21 mandated that the High Density Rule slot controls would no
                          longer apply to JFK and LaGuardia after January 2007. The exemptions for new entrant
                          airlines and for service to small communities using regional jets at JFK and LaGuardia
                          were interim measures until the AIR-21 slot related phase-out became effective.




                          Page 29                                                   GAO-12-902 Airport Slot Controls
                            than 600 new slot exemptions at LaGuardia and by September 2000, 25
                            percent of the flight delays in the United States were attributed to
                            congestion at LaGuardia. FAA reports that after it again capped the
                            number of scheduled takeoffs and landings at LaGuardia in 2001, aircraft
                            delays of 15 minutes or more fell from 330 per day in October 2000 to 98
                            per day in April 2001. 46 In setting operating limits at JFK and Newark in
                            2008, FAA conducted analyses with the goal of reducing average
                            departure delays at JFK by about 15 percent compared with their 2007
                            level, and to keep delays at Newark from increasing above their 2007
                            level. 47 Airline representatives with whom we spoke mostly agreed that
                            some form of slot controls are needed at the four airports to manage
                            congestion. However, representatives from two airlines believed that if
                            slot controls were lifted, congestion initially would spike, but airlines would
                            eventually make the adjustments needed to reach a level of manageable
                            congestion.


Despite Demand for          Existing slot control rules require airlines to operate their slots at least 80
Additional Slots at Slot-   percent of the time, which allows airlines to not schedule some of their
Controlled Airports, Some   allocated slots. FAA has interpreted the rules to allow airlines to apply the
                            usage requirement across a pool of slots they hold within a slot period—
Slots Go Unused While       an allowance that provides an advantage to airlines with large slot
Others Are Underutilized    holdings in meeting the usage requirement. Further, FAA’s current
                            process for overseeing airlines’ compliance of the 80 percent slot usage
                            requirement, including its recordkeeping, does not provide sufficient
                            assurance that it can adequately identify instances when airlines do not
                            meet the 80 percent slot usage requirement. We also found that airlines
                            tend to underutilize the slots they do operate at slot-controlled airports as



                            46
                              In announcing its intent to limit the number of slot exemptions allowed to operate at
                            LaGuardia, FAA noted language in section 231 of AIR-21 providing, in general, that AIR-
                            21’s slot phase out provisions shall not be construed to affect FAA’s authority for safety
                            and the movement of air traffic. 65 Fed. Reg. 69126 (Nov. 15, 2000). FAA explained that
                            under its authority for safety and the movement of air traffic, and pending the development
                            of a longer term solution, it sought to temporarily cap AIR-21 slot exemptions at
                            LaGuardia. This limit, according to FAA, still allowed a significant increase in operations at
                            the airport above the regulatory limits, thus serving congressional objectives while
                            stretching capacity to its practical limits. 71 Fed. Reg. 51360 (Aug. 29, 2006).
                            47
                              FAA and its federally funded research and development center, the MITRE
                            Corporation’s Center for Advanced Aviation System Development, used a model that
                            estimated the level of delay associated with various levels of operations at both JFK and
                            Newark airports.




                            Page 30                                                     GAO-12-902 Airport Slot Controls
                  compared with like-sized airports that are not slot-controlled. 48
                  Specifically, flights operated at slot-controlled airports tend to use smaller
                  aircraft and are on average less full, which means that fewer passengers
                  are served then would be otherwise.


Unused Capacity   Available airport capacity at slot controlled airports can go unused in
                  three distinct ways. The first of these are slots that are unallocated (i.e.,
                  slots that are not currently held by any airline). These slots, however, are
                  generally at times of day that airlines find undesirable, such as early
                  morning or late evening slots. Although unallocated slots go unused, they
                  are available for any airline to request and operate. 49 Second, some
                  additional amount of available capacity goes unused because airlines do
                  not schedule all of the slots that they are allocated. Third, more capacity
                  goes unused because airlines do not operate all of the flights that they
                  schedule, although this is in part because of cancellations resulting from
                  unforeseeable circumstances. 50 As discussed below, we could not
                  calculate precisely how much capacity is unused among these three
                  levels or even overall because current airline reporting and FAA
                  recordkeeping do not allow these calculations. Figure 3 illustrates the
                  three areas in which capacity can go unused at slot-controlled airports.
                  The amount that goes unused varies depending on the airport, airline,
                  time of year, and demand for air travel.




                  48
                    Because our analysis did not assess the behavior of individual airlines, we could not
                  determine the prevalence of this practice among airlines. See appendix V for a detailed
                  discussion of our analysis and the results.
                  49
                    FAA officials told us that some additional operations have been added in less desirable
                  hours.
                  50
                     According to DOT and enforcement case precedent, airlines may not routinely cancel
                  flights for economic reasons. This practice, according to DOT, would constitute an unfair
                  method of competition to engage in “unrealistic scheduling of flights” identified in 14
                  C.F.R. § 399.81 as an unfair or deceptive practice. See, DOT Order 97-1-6, Bellair
                  Incorporated (Jan. 13, 1997).




                  Page 31                                                   GAO-12-902 Airport Slot Controls
Figure 3: Areas of Unused Capacity at Slot-Controlled Airports




Unscheduled Slots                        Although slot control rules require airlines to operate their slots at least 80
                                         percent of the time, or the slots are subject to withdrawal by FAA, the
                                         rules do not require a minimum level for which the airlines must schedule
                                         their slots. Therefore, the usage requirement allows airlines to not
                                         schedule or to suspend flights for operational reasons (such as
                                         cancellations because of bad weather or ground delays) up to 20 percent
                                         of the time. 51 According to FAA officials, it is not the agency’s intention for
                                         airlines to operate at the minimum usage rate of 80 percent because
                                         unforeseen circumstances would result in the airline losing their slots.


                                         51
                                           Under specified circumstances, FAA may waive the 80 percent usage requirement. For
                                         example, with respect to JFK, LaGuardia, and Newark, FAA may waive the usage
                                         requirement “in the event of a highly unusual and unpredictable condition which is beyond
                                         the control of the carrier and which affects carrier operations for a period of five
                                         consecutive days or more.” Waiver provisions applicable to Reagan National require such
                                         a condition to exist for a period of 9 or more days. 76 Fed. Reg. 18616 (Apr. 4, 2011); 76
                                         Fed. Reg. 18618 (Apr. 4, 2011); 76 Fed. Reg. 18620 (Apr. 4, 2011); 14 C.F.R. § 93.227.




                                         Page 32                                                   GAO-12-902 Airport Slot Controls
Generally, airlines consider slots that they hold to be valuable assets, in
part, because slots are a scarce resource that allows airlines to operate
flights at airports with a high demand for air travel. 52 However, airlines
may not need all of their allocated slots at a point in time. Even so,
airlines still have an incentive to hold onto their slots, even if that means
leasing them, because if they return them to FAA the slots could be
reallocated to a competitor. During our review, we found that airlines at
U.S. slot-controlled airports are not scheduling all of their allocated slots.
Specifically, our analysis of aggregated airline schedules for certain days
in 2011, indicated that generally airlines in aggregate scheduled between
80 percent and 100 percent of their slot allocations (depending on the
airport, day of week, and season), but on specific days, one or more
airlines scheduled as little as 75 percent of all daily allocated slots. 53
Because slot usage is measured across multiple days of a reporting
period—all days within a 2-month period at Reagan National and
LaGuardia and each day of the week within a season for JFK and
Newark—airlines can schedule less than 80 percent of their allocated
slots on a single day and still meet the 80 percent usage requirement.

Internationally, where slot controls are more prevalent than in the United
States, airlines generally schedule all of their slots. IATA representatives
told us that airlines operating at slot-controlled airports outside of the
United States generally schedule 100 percent of their allocated slots and
that it may be seen as abuse if they don’t schedule all of their slots. Under
the Worldwide Slot Guidelines, airlines may only hold slots that they
intend to operate, transfer, exchange or use in a shared operation, and
violations are subject to sanction under local regulation or national law.
FAA, however, in implementing U.S. slot control rules, generally does not
calculate or know airlines’ schedule rates. Requiring airlines to schedule
all of their slots could encourage airlines to sell or lease slots to airlines
that desire them or return slots to FAA that they do not intend to use, but
it could also cause airlines to operate more flights than necessary to meet
demand for routes within their business model as a means of meeting the
80 percent usage requirement. Airlines would only operate additional


52
  Although airlines might consider slots to be valuable assets, FAA slot regulations at 14
C.F.R. § 93.223(a), for example, provide that slots do not represent a property right but
represent an operating privilege subject to absolute FAA control.
53
  To conduct this analysis, we compared the total number of scheduled flights for each of
7 days in February 2011 and each of 7 days in August 2011 with daily slot allocation
numbers provided by FAA for each slot-controlled airport.




Page 33                                                    GAO-12-902 Airport Slot Controls
                             flights, though, if the perceived cost of relinquishing their slot—such as
                             the risk of a competitor operating the slot—is higher than the cost they
                             would incur from operating the flight.

Slot Usage Rate Compliance   FAA’s current process for overseeing airlines’ compliance of the 80
                             percent slot usage requirement, including its recordkeeping, does not
                             provide sufficient assurance that it can adequately identify instances
                             when airlines do not meet the 80 percent slot usage requirement. FAA
                             uses a slot management system to compile usage data reported by
                             airlines and validate that data against FAA’s slot allocation records. FAA
                             then reviews instances of potential noncompliance on an exception
                             basis—that is, FAA reviews only those instances that its system identifies
                             as below the 80 percent slot usage rate. Our review of FAA’s process for
                             overseeing airlines’ slot usage, combined with FAA’s inability to provide
                             documentation verifying airlines’ compliance with the usage requirement
                             and our limited analysis of FAA aggregated data on airlines’ flight
                             schedules and slot allocation records, raised questions regarding FAA’s
                             ability to effectively identify instances of noncompliance. Issues we
                             identified include the following.

                             •   Usage rates are generally not calculated: At the end of each reporting
                                 period, airlines submit reports to FAA that include the minimum
                                 required data for each slot, including whether the slot was operated,
                                 the flight number, and the scheduled time of departure, but do not
                                 include actual usage rates. FAA officials said that they do not
                                 calculate the actual slot usage rate for each airline’s slot holdings, in
                                 part, because the format in which airlines report the data makes it
                                 difficult to correlate the slots to the flights that were operated. For
                                 example, flights operated by Airline A using slots leased or exchanged
                                 from Airline B may be reported by Airline A, but would need to be
                                 considered in Airline B’s usage rate but not in Airline A’s rate. The
                                 airline reports generally do not make this distinction. Furthermore,
                                 FAA does not require airlines to report the same flight in a specific slot
                                 throughout the reporting period, which further complicates FAA’s
                                 ability to correlate a slot to the flight that was operated. Without
                                 calculating airlines’ actual usage rate, FAA does not know precisely
                                 how much available capacity goes unused, which hinders its ability to
                                 manage slots and the airspace.
                             •   Reliance on self-reported data in a variety of formats: Instead of
                                 calculating the actual usage rate, FAA compares airlines’ reported
                                 data with FAA slot allocation records and in so doing will correct
                                 discrepancies or errors. However, FAA does not fully verify the
                                 accuracy of airlines’ self-reported data. Furthermore, FAA does not



                             Page 34                                           GAO-12-902 Airport Slot Controls
    require a specific format for airlines’ usage reports. Therefore, some
    reports are received in a database format that can be entered into
    FAA’s slot management system through an automated interface.
    Others require FAA to manually enter data from a spreadsheet into its
    system. In some cases, FAA officials told us that if extensive reporting
    errors or formatting or other administrative changes are needed, they
    may require the airline to submit a supplemental or corrected usage
    report.
•   Limitations in FAA’s database: Because of the difficulties in
    calculating usage rates by FAA, we conducted a limited analysis of
    airline schedule data and FAA data to assess the extent to which
    airlines were using their slots. Using slot allocation records for
    scheduled passenger flights provided by FAA for select days and
    matching them against airline flight schedules for those days, we
    examined the extent to which airlines were scheduling their slots.
    Based on our initial analysis, we determined that in one instance at
    least one airline had not met the 80 percent usage requirement at
    Newark. 54 Although our method of analysis is not the method that
    FAA uses to oversee individual airlines’ compliance with the usage
    requirement, FAA officials agreed that our method could be used to
    assess whether some of the slots were, at times, operated less than
    80 percent of the time. Our method would not gauge individual
    airline’s compliance, but would provide an aggregated check of usage
    by all airlines. FAA officials told us that they were not aware of the
    problem at Newark so they rechecked their slot allocation records and
    realized that the total slot allocation numbers in their data for Newark
    and JFK were inaccurate. Specifically, their records did not account
    for slots that airlines “handback” (i.e., return) to FAA prior to the slot’s
    handback deadline. Airlines at Newark and JFK are allowed to and
    routinely do return to FAA blocks of slots (e.g., specific days of the
    week or short periods of time) that they initially requested and were



54
  Specifically, we found that one or more airlines at Newark scheduled less than 80
percent of their allocated slots for selected Saturdays in the 2011 summer season. We
calculated schedule data for the third Saturday of each month of the summer season
(April, May, June, July, August, September, and October), which represented 7 of the 30
Saturdays during the 2011 summer season. FAA officials said that the third week of each
month is most representative of the rest of the month. Of the 7 Saturdays that we
examined, schedule rates were between 66 percent and 77 percent. Given average
cancellation rates for Newark and the typical scheduling rates observed for the days that
we examined, it is practically impossible (though theoretically possible) for all airlines to
have met the 80 percent usage requirement for all slots on Saturdays in the 2011 summer
season at Newark.




Page 35                                                    GAO-12-902 Airport Slot Controls
     allocated by FAA but do not intend to operate during the season. 55
     FAA officials said that they do not always update their slot allocation
     database with the lists of returned slots because it would be too
     resource-intensive and because of limitations in the computer
     software. While removing the handbacks from the allocation numbers
     provided to us increased the airlines’ presumed usage rate above 80
     percent at Newark airport, FAA did not calculate the usage rates with
     the corrected data and thus, could not provide us verification that the
     airlines usage exceeded 80 percent. And, it does not remove our
     concerns regarding the accuracy of FAA’s data or process for
     verifying compliance with the 80 percent rule.
•    Slot trades and leases: FAA’s review is further complicated by the
     number and frequency of slot trades and leases between airlines.
     FAA officials estimate that it records more than 5,000 airline trades or
     leases per season, which are manually entered into its records.
     Furthermore, slot usage is often reported by the slot operator, but the
     slot holder is responsible for complying with the slot usage
     requirement. Therefore, FAA often has to analyze and correct
     differences in data submitted by a slot holder and the slot operator
     (i.e., the lessee).
FAA officials acknowledged some inaccuracies in their data and that their
process is somewhat inefficient, but told us that they have identified
instances when airlines have not met the 80 percent usage rate and
required airlines to return the underutilized slots to FAA. In these cases,
FAA has issued letters to airlines indicating that they had lost rights to
their slot. Furthermore, FAA officials explained that they are replacing the
existing software system, which they said will address some of the
problems we identified in maintaining complete and accurate records and
make their oversight of airlines’ compliance with the usage requirement
more efficient. In particular, the new system should reduce the amount of
data that requires manual entry into their records system. FAA anticipates
implementing the new system by the end of 2012. However, because the
system is not yet operational and its design is not finalized, we do not
know to what extent the new system will improve the accuracy of FAA’s
records or improve its ability to check airlines’ compliance with the usage
requirement.


55
  This practice is also common at slot-controlled outside the United States and required
under the Worldwide Slot Guidelines. The returned slots do not count against the airline’s
usage rate, so long as they are returned before the handback deadline, which is about 10
weeks before the start of the season.




Page 36                                                   GAO-12-902 Airport Slot Controls
FAA’s administration of the slot rules provides airlines with large slot
holdings an advantage over those airlines that hold few slots because
they have more flexibility in meeting the 80 percent usage requirement.
Specifically, FAA does not require that airlines report the same flight in a
specific slot over the course of a reporting period, which essentially allows
airlines that hold multiple slots within a slot period—30 minutes at the
New York City airports and 1 hour at Reagan National—to apply the 80
percent usage requirement collectively to their pool of slots within that slot
period, as opposed to each individual slot. For example, an airline that
holds 5 slots within one slot period could hold but not operate one of its 5
slots (20 percent) and still meet the usage requirement if the other 4 were
used at 100 percent by “rotating” 4 flights across its 5 allocated slots
during the reporting period. FAA officials told us that they recognize that
this method of applying the usage requirement provides more flexibility to
airlines with multiple slots in a slot period, and that although this method
is not specifically addressed in the slot control rules, it is not prohibited
either. FAA officials told us that this method is a long-standing custom
that has evolved within FAA over time. FAA has no written guidance for
airlines describing the method. Most slot coordinators outside of the
United States oversee compliance with the usage requirement based on
each slot’s individual usage. Port Authority officials said that measuring
usage for each individual slot would reduce the incidence of airlines
holding onto slots that they do not need, making some existing capacity
available for redistribution. Representatives from four smaller slot-holding
airlines told us they would favor FAA requiring airlines to apply the usage
requirement to each individual slot, while representatives from five
airlines—four of which have larger slot holdings—said they were opposed
to it because it would reduce their scheduling flexibility. Using the current
method for measuring usage of an airline’s slot pool within a slot period,
airlines have an additional incentive to obtain as many slots as possible to
increase their operating flexibility without necessarily increasing their
number of operations—which would be an inefficient use of limited airport
capacity. Port Authority officials told us that airlines in the past have
announced new air service without obtaining any new slots. However, the
only way for an airline to add more flights without obtaining more slots is if
that airline begins using a slot that it was not fully utilizing before. In
addition, DOJ raised this as a concern in its comments on the US Airways
and Delta Air Lines proposed slot swap, referring to the practice of




Page 37                                           GAO-12-902 Airport Slot Controls
airlines using flights on slots that exceed the 80 percent threshold to
cover for an unused slot in the same slot period as “babysitting”. 56

Increasing the minimum usage requirement above 80 percent could also
result in less capacity going unused. For example, according to Airports
Council International in Europe, an airport organization that supports the
European Commission proposal to increase the minimum usage
requirement in Europe to 85 percent, increasing the slot usage
requirement would encourage airlines to optimize their use of slots and
allow slots that are not being used effectively to be returned and
reallocated. Because airlines operating outside the United States
generally schedule all of their slots, they tend to operate a higher
percentage of their slots than in the United States. Even so,
representatives from the Association of European Airlines told us that
increasing the required usage rate could result in an unintended
consequence of having airlines fly more aircraft than consumer demand
requires on specific routes, and with lower load factors (percent of
available seats filled with passengers), to meet the higher usage
requirement. Although most U.S. airline representatives with whom we
spoke said they opposed increasing the slot usage requirement,
representatives from two airlines with small slot holdings said they would
support increasing it and a representative from another airline supported
increasing it if the new rate did not penalize for cancellations because of
bad weather and other unforeseeable circumstances. Representatives at
an airline with a large number of slots said that increasing the slot usage
requirement above 80 percent would not increase efficiency because it
would reduce airlines’ flexibility to cope with unforeseeable events such
as weather, air traffic control delays, or technical problems. On average,
however, flight cancellations account for less than 2 percent of the total
scheduled flights (the cancellation rates at the New York City area
airports are generally higher than the system-wide average) and FAA




56
  See, DOJ, Comments of the United States Department of Justice on Notice of Petition
for Waiver of the Terms of the Order Limiting Scheduled Operations at LaGuardia Airport
and Solicitation of Comments on Grant of Petition with Conditions, (Washington, D.C.:
Mar. 24, 2010).




Page 38                                                 GAO-12-902 Airport Slot Controls
                         may grant waivers to the minimum usage requirement under other
                         unusual or unpredictable circumstances. 57

Underutilized Capacity   Despite limits on airport capacity, flights operated at slot-controlled
                         airports—as compared with flights at other like-sized airports that are not
                         slot-controlled—

                         •    tend to use smaller aircraft;
                         •    are, in some instances, flown to the same destination at higher rates
                              per day; and
                         •    have lower overall load factors (percentage of available seats filled
                              with passengers).
                         As a result, the level of passenger traffic at these airports is not as great
                         as the number of slots could allow and would-be-new entrant airlines that
                         could potentially offer lower fares or service to other destinations may be
                         shut out of these markets. The potential inefficient use of slots by
                         incumbent airlines at the New York City area airports as a way to block
                         competitors’ entry to the New York market has been examined before.
                         Under FAA’s 2006 proposed, but later withdrawn, congestion-
                         management rule for LaGuardia, FAA proposed requiring airlines to use
                         an average minimum aircraft size. FAA—in explaining its proposal—
                         indicated that some inefficiencies at LaGuardia were related to airlines
                         using smaller aircraft and that increasing the overall number of
                         passengers without increasing the number of flights would result in a
                         more efficient use of the national airspace system. Although FAA
                         ultimately withdrew the proposed average minimum aircraft size
                         requirement in 2008, the agency indicated that the concept behind the
                         proposed minimum aircraft-size requirement remained valid because
                         capacity cannot be considered merely on the basis of the number of
                         aircraft being handled by FAA’s Air Traffic Control system. 58 In particular,


                         57
                           At the New York City area airports, FAA may waive the 80 percent requirement in the
                         event of a highly unusual and unpredictable condition which is beyond the control of the
                         carrier and which affects carrier operations for a period of 5 consecutive days or more.
                         The provision applicable to Reagan National provides that such period must be for 9 or
                         more days. This provision further provides examples of conditions which could justify
                         waiver – weather conditions which result in the restricted operation of an airport for an
                         extended period of time or the grounding of an aircraft type.
                         58
                           In April 2008, FAA published a supplemental congestion-management rule proposal for
                         LaGuardia without the minimum aircraft size requirement. 75 Fed. Reg. 20846 (Apr. 17,
                         2008). The final rule for the Congestion Management Rule for LaGuardia, based on the
                         supplemental proposal, was issued in October 2008 and rescinded in 2009.




                         Page 39                                                   GAO-12-902 Airport Slot Controls
                FAA cited certain market patterns involving multiple daily flights on small
                aircraft that were not related to the size of the community served, which
                indicated an inefficient use of slots or behavior that stifled competition.
                Similarly, in 2010, when DOJ commented on the competitive aspects of
                the proposed slot swap between Delta Air Lines and US Airways at slot-
                controlled Reagan National and LaGuardia, it found that airlines engaged
                in what DOJ referred to as “slot hoarding” by flying excessive frequencies
                or using small planes. 59 While the use of smaller aircraft, high
                frequencies, and low load factors may be indications that airlines are
                hoarding slots as a way to keep competitors out of the slot-controlled
                airports, there are also possible market-based explanations for these
                tendencies. Without other evidence of airlines’ motivations, all that can be
                determined is that flights at the slot controlled airports tend to use the
                available capacity more inefficiently than like-sized airports that are not
                slot-controlled. But, given that slots are a scarce resource and interest
                from new entrant airlines is known, this provides sufficient reason to
                improve slot utilization.

Aircraft Size   To examine the extent to which flights at slot-controlled airports use
                smaller aircraft (100 seats or fewer 60), we conducted statistical analyses
                using data on aircraft size of scheduled passenger flights between
                domestic, large hub airports. 61 We analyzed airline schedule data that
                included, among other things, the scheduled flight date, origin and
                destination of the scheduled flight, and the size of the aircraft. We used
                data from one Thursday each in February 2011 and in August 2011. 62



                59
                  See, Department of Justice, Comments of the United States Department of Justice on
                Notice of Petition for Waiver of the Terms of the Order Limiting Scheduled Operations at
                LaGuardia Airport and Solicitation of Comments on Grant of Petition with Conditions,
                (Washington, D.C.: Mar. 24, 2010). In an analysis of aircraft size, DOJ controlled for the
                fact that commuter slots at Reagan National come with a restriction as to the size of the
                aircraft that can be flown.
                60
                  For the purpose of this report, we defined “small” aircraft as one that contains 100 seats
                or fewer. We also examined flights for which a “small” aircraft was smaller than or equal in
                size to the median-sized aircraft (124 seats) in the sample of flights used for the model, as
                well as fewer than or equal to 80 seats.
                61
                  For the purposes of this report, we define large hub airports as all FAA-identified large
                hub airports in the United States for calendar year 2010 (the most recent year available for
                this list), excluding Honolulu. Federal law defines a large hub airports as those that have
                at least 1 percent of the passenger boardings. 49 U.S.C. § 40102(29).
                62
                  The specific dates we selected were Feb. 10, 2011, and Aug. 11, 2011.




                Page 40                                                    GAO-12-902 Airport Slot Controls
Thursdays are generally considered to be a high-demand day, and
February and August are representative months for the winter and
summer seasons, respectively.

Our analysis indicates that the proportion of flights using small aircraft is
significantly greater at the slot-controlled airports than at other large hub
airports that are not slot-controlled. The proportion of flights using small
aircraft at slot-controlled airports ranged from 36 percent at LaGuardia to
25.8 percent at Newark, whereas the proportion of flights using small
aircraft to and from large hub airports that are not slot-controlled was 16.4
percent. That is, the percent of flights using small aircraft at any of the
slot-controlled airports was between 9.4 percent and 19.6 percent
higher—depending on the airport—than the percent of flights using small
aircraft at other non-slot-controlled large hub airports. We found that
these differences were all statistically significant (see table 5). To some
extent, the perimeter rules at Reagan National and LaGuardia contribute
to the use of smaller aircraft at these airports because shorter haul flights
are generally served by smaller aircraft, but also limit service to larger
long-distance markets that use larger aircraft. Also, 11 out of the 60
hourly slots allowed under the High Density Rule slots at Reagan National
are reserved for and require the use of smaller aircraft used for commuter
flights. And, under LaGuardia’s temporary order, FAA assigned some
slots to airlines that held slot exemptions granted under AIR-21, which
included slot exemptions for flights serving small communities using
aircraft with fewer than 71 seats. See appendix V for a detailed
discussion of our analysis and the results.




Page 41                                          GAO-12-902 Airport Slot Controls
Table 5: Percentage Difference in Flights at Slot-Controlled Airports That Use Small
Aircraft Compared with Flights at Non-Slot-Controlled Airports, 2011

                                                                     Percentage differences in small
                                       Percentage of flights       aircraft at slot-controlled airports
                                             that use small           compared with other large hub
    Airport(s)                                      aircraft    airports that are not slot-controlleda
    Newark                                       25.8 percent                        9.4 percent more
    JFK                                          26.6 percent                       10.2 percent more
    Reagan National                              34.0 percent                       17.6 percent more
    LaGuardia                                    36.0 percent                       19.6 percent more
    All 4 U.S. slot-                             29.3 percent                       12.9 percent more
    controlled airports
    All 3 New York City                          30.2 percent                       13.8 percent more
    area slot-controlled
    airports
    Large hub airports that                      16.4 percent                                      N/A
    are not slot controlled
Source: GAO analysis of airline schedule data.

N/A = not applicable.
a
 The percentage differences are all statistically significant.


To better understand the extent to which small aircraft are used at slot-
controlled airports, we also developed a logistic regression model that
controlled for other factors that are likely associated with aircraft size. 63
The model assessed whether scheduled passenger flights to or from slot-
controlled airports tend to have a greater likelihood of being scheduled
with small aircraft compared with flights to and from other large hub
airports that are not slot-controlled. In addition, the model controlled for
other factors that include (1) whether the origin or destination of the flight
is a high-tourism airport, 64 (2) whether the airline marketing the flight is a


63
  A logistic regression (or logit model) provides an indication of which independent
variables are correlated with the dependent variable—which in this case, is a flight using a
small aircraft. Moreover, it provides this measure of correlation independent of the effects
of the other independent variables included in the model. It is important to note, however,
that this type of statistical method only suggests correlations between variables and not
causation. That is, our findings do not provide an indication that any of the independent
variables we included in the model actually caused (or did not cause) an airline to use a
small aircraft for a flight, but only how those variables are correlated with the use of a
small aircraft.
64
  We defined Fort Lauderdale and Orlando, Florida, and Las Vegas, Nevada, as high-
tourism airports.




Page 42                                                               GAO-12-902 Airport Slot Controls
“legacy” airline, 65 (3) the daily frequency of flights to the same destination
by the same airline, and (4) the flight distance. Although the model
controls for distance, which helps control the effect of the perimeter rules
at Reagan National and LaGuardia, the model does not control for the
designated commuter slots at Reagan National that are limited to
commuter-sized aircraft. 66 However, FAA officials told us that based on
their analysis, about 30 to 40 percent of slots that do not require the use
of commuter aircraft also use commuter-sized aircraft. Also, as with any
regression model, this analysis does not prove a causal relationship
between different factors, such as slot control rules and an airlines’ use of
a small aircraft. Thus, the extent to which the slot-control rules, the
perimeter rules at Reagan National and LaGuardia, and the commuter-
slot designation at Reagan National contribute to underutilization at these
airports is unknown. Other factors related to the use of small aircraft may
not be fully controlled for, such as the type of passenger traffic. See
appendix V for a detailed discussion of the model.

Similar to our analysis discussed above, the model results also suggest
that airlines are more likely to use small aircraft for flights at slot-
controlled airports than they are for flights at other large hub airports that
are not slot-controlled. Specifically, when controlling for other factors that
may be associated with using a small aircraft, flights to or from the four
slot-controlled airports are between 22 percent and 108 percent more
likely—depending on the airport—to use a small aircraft than flights that
arrive and depart large hub airports that are not slot-controlled (see table
6). See appendix V for a detailed discussion of the model results.




65
  The airlines classified as “legacy” included Delta Air Lines, United Airlines, American
Airlines, US Airways, and Continental Airlines. Legacy (sometimes called network) airlines
support large, complex hub-and-spoke operations with thousands of employees and
hundreds of aircraft (of various types), with flights to domestic communities of all sizes as
well as to international destinations. The modifier “legacy” derives from the fact that these
airlines were operating when the industry deregulated in 1978.
66
  Under federal statute and FAA slot control regulations applicable to Reagan National,
“commuter slots” may be used only for operations with turboprop and reciprocating engine
aircraft with no more than 76 seats or turbojet aircraft with fewer than 56 seats.




Page 43                                                    GAO-12-902 Airport Slot Controls
Table 6: Percentage Difference in Likelihood of a Flight’s Using Small Aircraft at
Slot-Controlled Airports Compared with Non-Slot-Controlled Airports

                                                           Likelihood (as measured by odds ratio)a
                                                            that a flight to or from a slot-controlled
                                                       airport used a small aircraft compared with
                                                          other large hub airports that are not slot-
    Slot-controlled airport(s)                                                              controlled
    Newark                                                                    22 percent more likelyb
    JFK                                                                       63 percent more likelyc
    Reagan National                                                           72 percent more likelyc
    LaGuardia                                                                108 percent more likelyc
    All 4 U.S. slot-controlled airports                                       75 percent more likelyc
    All 3 New York City area slot-controlled                                  68 percent more likelyc
    airports
Source: GAO analysis of airline schedule data.

Notes: The results of our analyses for 124 seat and 80 seat cutoffs also found odds ratio values
greater than 1, and in some instances the results were more significant.
The percentage differences in the likelihood that a flight used a small aircraft presented here are
based on odds ratios calculated from our model. Specifically, the likelihood equals the odds ratio
minus 1, then multiplied by 100.
a
    See appendix V for a description of odds ratios.
b
    Statistically significant at p-value <0.1.
c
    Statistically significant at p-value <0.01.


According to representatives from some airlines with smaller slot holdings
and some airport representatives, airlines with large slot portfolios have
an incentive and the ability to fly smaller and more frequent flights at slot-
controlled airports in a desire to hold on to slots they would otherwise not
use because they do not want to lose them to a competitor. Of the 11
airlines that we interviewed for this review, representatives from 7 non-
legacy airlines favored establishing minimum aircraft size requirements at
the slot-controlled airports, while representatives from 3 legacy airlines
opposed it. 67 Representatives from two airlines also said they would
support a minimum aircraft size requirement if small aircraft could still be
used to serve small communities. Representatives from two airlines who
favored establishing a minimum aircraft size said that using small aircraft
between major markets is a particularly inappropriate use of constrained


67
  Representatives from the remaining airline did not oppose establishing a minimum
aircraft size, but did not believe that it was the best method to incent airlines to use larger
aircraft.




Page 44                                                              GAO-12-902 Airport Slot Controls
                   airspace. However, representatives from legacy airlines that we spoke
                   with said establishing minimum aircraft size requirements would
                   negatively affect service to small- and medium-sized markets and that it
                   would limit their ability to use aircraft sizes that are most appropriate to
                   meet customer needs, for example, regarding schedule frequency and
                   market conditions.

Flight Frequency   In analyzing airline schedule data, we found instances when airlines
                   scheduled flights to or from slot-controlled airports for the same
                   destination at a more frequent rate than other large hub airports that are
                   not slot-controlled. For example, airlines might fly smaller aircraft at
                   higher rates to provide frequent shuttle service for business travelers. We
                   found, however, that on February 11, 2011, one airline scheduled more
                   than 20 daily flights from LaGuardia to Philadelphia—a distance of 96
                   miles by air—and that some of those flights were scheduled 10 minutes
                   or less apart. As of June 21, 2012, the airline had reduced the number of
                   daily nonstop flights on this route to 15. Representatives from the airline
                   operating these flights said the high frequency of flights between
                   LaGuardia to Philadelphia is driven by consumer demand and provides
                   more choice for customers. They also said that almost all of the
                   passengers on these flights are connecting to other flights and that the
                   airline is not competing with other airlines in this market, but with cars and
                   trains. To better understand the extent to which airlines schedule flights to
                   or from slot-controlled airports for the same destination at short intervals,
                   we examined the frequency of scheduled flights to the same destination
                   by the same airline between all large hub airports for one day in February
                   2011. 68 February is a low demand month in which airlines at slot-
                   controlled airports might not reduce the number of scheduled flights as
                   they might at other airports that are not slot-controlled. Rather, the airline
                   might operate smaller, more frequent flights to meet the 80 percent usage
                   requirement. 69 Our analysis found that of the scheduled flights to or from
                   the same destination by the same airline within 15 minutes of one
                   another, LaGuardia accounted for 17 of the 50 total scheduled flights
                   identified (34 percent). The airport with the next highest proportion was
                   Philadelphia International with 11 of the 50 identified flights (22 percent) –



                   68
                     We examined airline schedule data for scheduled domestic passenger flights for Feb.
                   10, 2011.
                   69
                     FAA’s 80 percent minimum usage requirements pertain to how often a slot is used and
                   not size or capacity of the aircraft using the slot.




                   Page 45                                                 GAO-12-902 Airport Slot Controls
all of which were scheduled to fly to or from LaGuardia. Moreover, the 11
scheduled flights that we identified within 15 minutes of one another
operated by the same airline between LaGuardia and Philadelphia had an
average aircraft size of 38 seats. When we examined scheduled flights
within 30 minutes of one another to or from the same destination by the
same airline, LaGuardia still had the highest share with 70 of the 264
identified flights (27 percent), followed by Philadelphia International with a
share of 29 of the 264 identified flights (see fig. 4). While current slot
control rules do not prohibit airlines from this type of scheduling practice,
FAA officials noted that the slot swap between Delta Air Lines and US
Airways have enabled these airlines to use their resources more
efficiently. Port Authority officials told us that they believe that any airline
that schedules a second flight to the same destination within a slot period
(30 minutes at the New York City area airports) should have to
demonstrate that there is demand for that second flight that could not be
met by a single larger aircraft.




Page 46                                            GAO-12-902 Airport Slot Controls
Figure 4: Airports’ Share of Scheduled Flights to or from the Same Destination by the Same Airline within 15 Minutes or 30
Minutes of One Another, February 2011




                                         Notes: “Other” airports in the 15 minute chart include Charlotte/Douglas International, Chicago
                                         O’Hare International, Dallas-Fort Worth International, Las Vegas McCarran International, Newark,
                                         and San Diego International.
                                         “Other” airports in the 30 minute chart include Hartsfield-Jackson Atlanta International, Boston Logan
                                         International, BWI, Charlotte/Douglas International, Dallas-Fort Worth International, Denver
                                         International, Detroit Metro Wayne County, Dulles, Las Vegas McCarran International, Los Angeles
                                         International, Orlando International, Miami International, Newark, Phoenix Sky Harbor International,
                                         Reagan National, San Diego International, San Francisco International, and Tampa International.




                                         Page 47                                                           GAO-12-902 Airport Slot Controls
Load Factors                            We also examined FAA data on average load factors (percentage of
                                        seats occupied) at large hub airports for all domestic scheduled
                                        passenger flights in calendar year 2011 and found that flights at slot-
                                        controlled airports were overall less full than flights at other large hub
                                        airports that are not slot-controlled. As shown in figure 5, we found that
                                        Reagan National (70 percent) and LaGuardia (72 percent) had the lowest
                                        average load factors for 2011 among the large hub airports, with Newark
                                        (77 percent) tied with Boston Logan International and Philadelphia
                                        International for the third lowest load factor.

Figure 5: Average Load Factors at Large Hub Airports for Scheduled Flights during 2011




                                        Page 48                                          GAO-12-902 Airport Slot Controls
                           According to Port Authority officials and some airline representatives,
                           some airlines at slot-controlled airports might operate flights with lower
                           load factors to help meet the 80 percent usage requirement and maintain
                           control of their slots, whereas flights with similarly low load factors at a
                           non-slot-controlled airport might be discontinued. However, another factor
                           to explain the comparatively low load factors at New York airports is the
                           relatively high airline yields (a measurement of airline revenue per mile).
                           Yields at Newark, LaGuardia, and Reagan National are some of the
                           highest among the large hub airports, which means that a flight with a low
                           load factor may still be profitable for these airlines. If the objective is to
                           maximize the number of passengers at these airports, however, this
                           represents an inefficient use of airport capacity.


Slot Control Rules Limit   Representatives from low cost and other airlines that have sought to
New Entry and Hinder       obtain slots told us that they have had difficulty gaining access to slot-
Competition                controlled airports. Airline representatives seeking slots at these airports
                           told us that they have spoken to FAA, but generally not enough slots are
                           available or the slots are at undesirable times of day (e.g., early morning
                           or late evening) for these airlines to operate profitable service. 70 DOT
                           officials are also aware of the access issue, and it was a focus of the
                           previous congestion-management rulemakings and an advisory Aviation
                           Rulemaking Committee report for the New York City area airports. 71 DOT
                           also took steps to mitigate potential competitive harm from the Delta Air
                           Lines and US Airways proposed slot swap by requiring divestiture of
                           some slots at the airports to airlines with small or no presence at the
                           airports. DOJ, in commenting on competition at LaGuardia and Reagan
                           National during the Delta Air Lines and US Airways proposed slot swap in
                           2010, indicated that the lack of availability of slots is a substantial barrier
                           to entry at those airports, especially for low cost airlines. Based on
                           literature that we reviewed, airfares can drop up to 30 percent on a flight
                           route as a result of a low cost airline entering the airport and beginning




                           70
                             One airline representative told us that the airline would need enough slots for three
                           transcontinental flights per day to be profitable.
                           71
                             New York Aviation Rulemaking Committee Report, December 13, 2007.




                           Page 49                                                    GAO-12-902 Airport Slot Controls
                                         service on that route. 72 Because average airfares at the slot-controlled
                                         airports are some of the highest among domestic large hub airports, the
                                         entrance of low cost competition could be helpful in bringing down
                                         airfares at those airports. While in some cases Congress has directed
                                         that new airport capacity in the form of slot exemptions be allocated to
                                         new entrants or “limited incumbents,” which are generally low cost
                                         airlines, those new slot exemptions are a small percentage of the total
                                         slots at the airport. Figure 6 shows that the proportion of passenger traffic
                                         handled by the nonlegacy airlines in 2011 generally is smaller at slot-
                                         controlled airports compared with the other 24 large hub airports.

Figure 6: Proportion of Legacy and Other Airline Passenger Traffic at Slot-Controlled and Non-Slot-Controlled Large Hub
Airports, 2011




                                         DOJ also indicated in 2010 that increasing the number of slots held by
                                         two large legacy carriers at the slot-controlled LaGuardia and Reagan
                                         National airports would make it more difficult for new entrants to obtain
                                         slots because large legacy carriers would have greater incentives to



                                         72
                                           See Steven A. Morrison, “Actual, Adjacent, and Potential Competition: Estimating the
                                         Full Effects of Southwest Airlines,” Journal of Transport Economics and Policy, Vol. 35,
                                         part 2, May 2001, and Austan Goolsbee and Chad Syverson, “How Do Incumbents
                                         Respond to the Threat of Entry? Evidence from the Major Airlines,” The Quarterly Journal
                                         of Economics (2008).




                                         Page 50                                                  GAO-12-902 Airport Slot Controls
forestall entry by refusing to sell or lease their slots. Representatives of
some airlines with few slot-holdings told us that they believe that the
airlines that hold the majority of slots at the airports are hesitant or
unwilling to lease to them. In general, airlines holding a slot under FAA’s
slot rules may lease or trade slots to other airlines and can select which
airlines they lease or trade slots with. 73 Airlines may have legitimate
business reasons for leasing slots to code-share partners, for example, to
feed into the lessor airlines’ network. In addition, representatives from
large slot-holding airlines said that they may hold onto slots that they do
not currently need by leasing to other airlines in case they want them
back later. In general, airlines only return slots to FAA that airlines are
unable to find other airlines interested in leasing or buying. Airlines with
large slot holdings also noted that they have made substantial financial
investments in terminal facilities, such as terminals at JFK, at the slot-
controlled airports. While the slots that are returned can be reallocated to
other airlines, including new entrants or limited incumbents, the returned
slots generally are the least desirable—such as early morning or late
evening or weekend-only—and may only be for a couple of months,
which makes them undesirable for airlines trying to build year-round
service.

According to Port Authority officials and some airline representatives, if
FAA were to limit the length of time that an airline could hold on to a slot
when that airline was not using or leasing it to another airline—after which
the slots would be withdrawn by FAA for reallocation—this could provide
opportunities for new entrants or small slot-holders to obtain slots.
Representatives from most of the airlines we contacted, except the legacy
airlines, said they favored limiting how long airlines can hold onto slots
they are not using. However, representatives from a legacy airline said
that imposing such a holding limit is unnecessary because slots do not go
unused. These representatives told us that although there are some
instances that on paper it appears that airlines are not using their slots,
the airlines are actually holding the slots by leasing them to other airlines
for a future time when they want to use them again. Also, the



73
  Under FAA’s slot control orders for JFK, LaGuardia, and Newark, slots may not be
bought or sold but may be leased or traded through the effective date of the orders, or
relinquished to FAA. Under FAA slot regulations for Reagan National, slots may be
bought, sold, leased, or traded, with the exception of slot exemptions, which may not be
bought, sold, leased, or otherwise transferred, except through an air carrier merger or
acquisition. 49 U.S.C. 41714(j).




Page 51                                                   GAO-12-902 Airport Slot Controls
representatives noted that airlines may be holding onto slots for long
periods of time (e.g., over multiple seasons) because the temporary slot
orders for the New York City area airports do not allow airlines to buy or
sell slots.

FAA’s administration of its slot rules can hinder the ability of airlines to
add service at slot-controlled airports because it does not provide
complete transparency about what slots are available. FAA officials said
that, upon request, FAA provides airlines with lists of slot holders and
operators as well as available slots, but some airline representatives and
Port Authority officials told us that the format of this information does not
make clear what slots are available. Furthermore, as discussed earlier,
because of discrepancies in FAA’s slot allocation records, it is unclear
whether the information provided to airlines is accurate. Port Authority
officials recommend public reporting of slot leasing and slot usage,
including the reasons why the slots are not being operated (e.g., flights
were not scheduled or were cancelled) to provide transparency of
information. In addition, although FAA requires the submission of certain
records of all trades and leases between airlines, it does not require the
submission of the information—such as the origin and destination that a
lessee intends to use for the slot—that would allow FAA or DOT to
monitor the extent to which leases occur between nonpartner airlines. 74
While DOT officials questioned the efficacy of reviewing data currently
collected on airline slot leases in isolation, additional information relating
to the transaction could be helpful. But, officials noted that even with
additional data, the department faces a very high burden of proof in
alleging that an airline is engaged in unfair or deceptive competitive
practices. 75 DOT officials told us that they believe that under the current


74
  Airlines partner with other airlines on some routes and not others. Under current slot
control rules, airlines are not required to provide FAA copies of slot leases (or any
information regarding consideration for slots transferred). FAA slot control orders for JFK,
LaGuardia, and Newark require airlines to submit notice of a slot trade or lease to FAA in
writing. Such trades or leases are subject to the confirmation and approval of FAA. The
FAA slot-control regulations for Reagan National provide, in part, that requests for FAA
confirmation of transfers, including leases, shall include the names of the transferor and
recipient; whether the slot is to be used for an arrival or departure; the date the slot was
acquired by the transferor; whether the slot has been used by the transferor for
international or essential air service operations; and whether the slot will be used by the
recipient for international or essential air operations, among other things. A recipient may
not use a transferred slot without written confirmation from FAA.
75
  Under 49 U.S.C .§ 41712, DOT may investigate whether airlines have engaged in
“unfair and deceptive practices and unfair methods of competition.”




Page 52                                                    GAO-12-902 Airport Slot Controls
              temporary orders, they are unable to react to certain market failures at
              slot-controlled airports, such as inefficient use of constrained airspace.
              DOT officials could not comment further on this because of the pending
              Slot Management and Transparency rulemaking for the New York City
              area airports. DOJ officials told us that, unless they have an open
              investigation, they do not monitor ongoing day-to-day activities such as
              short-term leases at slot-controlled airports. An investigation of
              competition-related issues at these airports is generally “triggered” by a
              transaction such as an airline merger or slot swap, although DOJ may
              also investigate in response to complaints, such as complaints by airlines
              unable to obtain slots. Additional information on leases and trades of
              slots, particularly the relationship between the lessor and the lessee,
              would allow FAA to provide more complete information to airlines on the
              availability of slots, which increases transparency.


              Because the New York City airports are both capacity constrained and
Conclusions   among the nation’s most sought after by airlines to serve, FAA has had to
              institute slot controls to manage congestion and resultant delays. It is
              therefore important that slots allocated to airlines are efficiently utilized,
              thereby maximizing the airports’ available capacity, or reallocated to other
              airlines that will use them efficiently, which could, in turn, provide access
              to new entrant airlines that offer new service destinations and lower fares.
              Despite these goals, we found that considerable existing capacity is not
              being used or is used inefficiently at these airports. Certain aspects of the
              slot control rules, as well as FAA’s administration of them, contribute to
              this situation and, thus, may hinder the ability of some new entrant
              airlines to obtain slots at the four U.S. slot-controlled airports. In
              particular, although the current slot control rules require slot holders to
              operate their slots 80 percent of the time, they do not require slot holders
              to schedule a flight for each of their slots. And, as opposed to airlines
              operating at slot-controlled airports outside the United States that
              generally schedule flights for all of their slots, domestic airlines operating
              at the U.S. slot-controlled airports do not always schedule all of their
              slots, which contributes to capacity going unused. Further, FAA’s current
              recordkeeping and process for reviewing airline slot usage data do not
              provide sufficient assurance that it can adequately identify when airlines
              do not meet the 80 percent usage requirement. Not requiring airlines to
              report usage data in a standard format or having management systems
              that can adequately compile data received in different formats, as well as
              its reliance on airlines’ self-reported data and discrepancies in its slot
              allocation record, hinders FAA’s ability to check compliance with the slot
              usage requirement. Moreover, because FAA does not calculate the actual


              Page 53                                           GAO-12-902 Airport Slot Controls
                      slot usage rates for airlines, the amount of capacity that goes unused is
                      unknown, which limits FAA’s ability to manage slots or set policy for the
                      airspace. FAA’s planned upgrades to its slot management software later
                      this year will hopefully improve its oversight capabilities, but not without
                      better and more standardized reporting by airlines. Other practices that
                      affect utilization, however, will not be fixed with the new software. For
                      example, FAA allows an airline to apply the usage requirement across its
                      entire pool of slots within a slot time period, rather than applying the
                      requirement to its individual slot—a practice that provides advantages to
                      airlines with large slot holdings over those with small slot holdings in
                      meeting the requirement. This practice can lead to underutilization of
                      individual slots and hinder other airlines’ ability to obtain slots at these
                      airports. Limitations in FAA’s recordkeeping, particularly the
                      discrepancies in its slot allocation records, as well as its lack of
                      information on the relationship between slot lessors and lessees and
                      public disclosure on the availability of slots and lease information, also
                      hinder access to new entrant airlines at slot-controlled airports.

                      To help maximize the use of available capacity at slot-controlled airports,
Recommendations for   enhance competition through greater airline access to slots, and enhance
Executive Action      transparency of slot information, we recommend that the Secretary of
                      Transportation direct the FAA Administrator to take the following five
                      actions:

                      •   require airlines to report usage data in a standard format at the end of
                          each reporting period or ensure that future slot management systems
                          have the capability to compile data in different formats;
                      •   periodically calculate the rates at which airlines are scheduling and
                          using their slots;
                      •   apply slot rules to individual slots, as opposed to pools of slots within
                          a slot period;
                      •   periodically disclose information, which may include current slot
                          holders and operators, on currently available slots; and
                      •   collect and disclose data, including the relationship between lessors
                          and lessees, on slot leases.
                      Furthermore, with respect to possible future regulatory action, we
                      recommend that DOT Secretary consider requiring airlines to schedule a
                      certain percentage, or all, of their slot allocations, similar to practices
                      maintained elsewhere in the world.




                      Page 54                                          GAO-12-902 Airport Slot Controls
                     We provided DOT, MWAA, the Port Authority of New York and New
Agency Comments      Jersey, DOJ, and the Department of Homeland Security (DHS) with a
and Our Evaluation   draft of this report for their review and comment. DOT’s Office of the
                     Secretary and FAA provided technical comments that we incorporated as
                     appropriate. In addition, in comments e-mailed to us, FAA reiterated that
                     the slot control rules at the three New York City area airports and Reagan
                     National were adopted primarily to manage congestion and delay at these
                     airports by limiting planned operations during peak hours. Officials
                     highlighted that complex issues arise with the allocation and use of a
                     scarce resource like airport slots and that congestion-management policy
                     objectives must be balanced with realistic competition and market access
                     goals. FAA indicated that economic and compliance benefits and costs to
                     stakeholders and regulatory and statutory guidelines must also be
                     considered. We agree that the slot control rules can manage congestion
                     and delay and that issues may arise when allocating a scarce resource.
                     We also believe that the rules can be improved to make the use of current
                     capacity more efficient, as well as enhance competition. Officials also
                     noted that DOT’s Office of the Secretary and FAA are developing rules to
                     replace the temporary slot control Orders at the New York City area
                     airports with more permanent rules. As stated in our report, these
                     temporary rules have been extended twice, creating a level of uncertainty
                     in the market about when the final rule will be issued and what it will
                     contain. While we acknowledge DOT’s and FAA’s efforts to develop a
                     permanent rule, we also believe that FAA could take separate action now
                     to help maximize the use of available capacity at slot-controlled airports,
                     enhance competition through greater airline access to slots, and enhance
                     transparency of slot information.

                     MWAA provided written comments, which are reprinted in appendix VI.
                     MWAA agreed with our recommendations to improve FAA’s oversight of
                     slot utilization and encouraged DOT to adopt these recommendations to
                     better manage available airport capacity and provide a more transparent
                     process. However, MWAA disagreed with our conclusion that the
                     additional beyond-perimeter slot exemptions at Reagan National are likely
                     to have a limited effect on the Washington, D.C., area airports and can be
                     supported by terminal capacity at Reagan National. In the comments,
                     MWAA indicated that the impact of the beyond-perimeter flights should be
                     viewed in the broader context of other changes under way at Reagan
                     National, for example, those resulting from legacy carrier mergers, the
                     Delta Air Lines and US Airways slot-swap transaction, and other recent
                     commercial slot transactions. MWAA believes that these factors are
                     changing the operating character of Reagan National, change that may
                     require considerable additional capital investment for extensive facility
                     modifications. However, our review was expressly limited to the impact of


                     Page 55                                         GAO-12-902 Airport Slot Controls
the additional slot exemptions authorized in 2012, and not the cumulative
effect of previously authorized beyond-perimeter slot exemptions and
other slot-related and commercial changes. The report includes the views
of MWAA management with regard to the specific challenges faced by
the expansion of beyond-perimeter slot exemptions throughout, but our
overall assessment remains that the effect of the 2012 FAA
Reauthorization slot exemptions is manageable and that the impact on
other Washington, D.C., area airports is relatively small. On this point,
MWAA disagrees with our report in regard to the financial impact of
potential passenger erosion on Dulles International. Our report does
recognize the potential effect on per-passenger airline costs at Dulles and
Reagan National, but also states that this estimated effect is a worst case
scenario and should be considered in light of (1) the recent substantial
increases in costs from capital investments at Dulles that are unrelated to
the new slot exemptions; (2) the passenger growth forecast by FAA that
could ameliorate that erosion; (3) the views of airlines and credit-rating
agencies that airline service is also dependent on market demand, which
is strong in this region; and (4) the “extraordinary coverage protection
payments” provision in the airline lease agreements that may allow
MWAA to share debt service between the two airports should they need
to and the possibility of revising this provision in the new lease. Finally, as
we indicate in the report, we concur with MWAA that because some of the
new beyond-perimeter flights only started in August 2012, the actual
impact of the new beyond-perimeter flights cannot yet be fully
determined. MWAA also provided some specific updated information,
which we incorporated as appropriate in the report.

The Port Authority of New York and New Jersey provided written
comments, reprinted in appendix VII, agreeing with the report’s
conclusions and recommendations. The Port Authority of New York and
New Jersey provided some clarifications that we incorporated as
appropriate. DHS, on behalf of TSA, also provide some clarifications that
we incorporated. DOJ provided no comments.


We are sending copies of this report to interested congressional
committees and the Secretary of Transportation. In addition, the report is
available at no charge on the GAO website at http://www.gao.gov.

If you or your staff have any questions about this report, please contact
me at (202) 512-2834 or flemings@gao.gov. Contact points for our
Offices of Congressional Relations and Public Affairs may be found on




Page 56                                           GAO-12-902 Airport Slot Controls
the last page of this report. GAO staff who made major contributions to
this report are listed in appendix VIII.




Susan A. Fleming
Director, Physical Infrastructure Issues




Page 57                                        GAO-12-902 Airport Slot Controls
Appendix I: Objectives, Scope, and
              Appendix I: Objectives, Scope, and
              Methodology



Methodology

              In this report, we reviewed (1) the potential effects of increasing the
              number of beyond-perimeter slot exemptions at Reagan National and
              (2) whether slot control rules at the four U.S. slot-controlled airports are
              working to reduce congestion, while maximizing available capacity and
              encouraging competition.

              To review the potential effects of increasing the number of beyond-
              perimeter slot exemptions at Reagan National, we

              •   compared airside capacity assessments with actual usage;
              •   reviewed terminal space needs;
              •   assessed the impact on the additional beyond-perimeter flights on
                  passenger traffic, capital financing, security screening wait times, and
                  noise; and
              •   interviewed officials from the Federal Aviation Administration (FAA),
                  the Metropolitan Washington Airports Authority (MWAA), and the
                  Baltimore-Washington International Thurgood Marshall Airport (BWI),
                  as well as representatives from airlines 1 and others.
              Regarding airside capacity, we analyzed data on the volume of aircraft
              operations at Reagan National, regulations and FAA actions regarding
              slot allocations at the airport, and a capacity assessment conducted by an
              FAA contractor. With respect to terminal capacity, we observed the
              airport’s baggage handling equipment, gates, ticket counter space, and
              security screening facilities; reviewed usage data on gates and parking
              lots; and discussed plans to expand terminal and security screening
              space with MWAA officials.

              To assess the maximum impact that the new beyond-perimeter flights
              would have on passenger traffic, we assumed that the new flights would
              be 100-percent full and that all passengers would be shifting from Dulles
              to Reagan National. This provided the highest possible estimate of the
              loss of passengers and the associated rise in cost per enplaned
              passenger at Dulles because of the new beyond-perimeter flights at
              Reagan National. However, for a variety of reasons, it is unlikely that the
              number of passengers shifting from Dulles to Reagan National would
              actually be this high. First, the new flights are unlikely to draw travelers


              1
               To obtain views from airlines on these issues, we contacted 12 airlines that serve
              Reagan National, had applied for beyond-perimeter slot exemptions there, or had
              expressed interest to FAA in serving that airport. Representatives from 11 of the 12
              airlines agreed to talk with us.




              Page 58                                                   GAO-12-902 Airport Slot Controls
Appendix I: Objectives, Scope, and
Methodology




solely from Dulles. For example, consumers who have a strong
preference for flying out of Reagan National might have been choosing to
fly from Reagan on a connecting basis but will now have a direct choice
and, despite the likelihood of a substantially higher fare, some of these
passengers will opt for the more expensive direct flight. Similarly, some
travelers might shift to the new flights at Reagan National from BWI.
Second, the availability of the new flights may induce some pricing
impacts that could influence travelers’ decisions. For example, airlines
that fly the same routes from Dulles may attempt to maintain their traffic
by lowering prices on those routes to better compete for passengers who
now have a nonstop choice from Reagan National. This would likely
mitigate the flow of passengers from Dulles to Reagan National, and may
even induce some new travelers on these routes. And finally, although
the new routes will likely be very popular, it is possible that not all seats
will be sold indicating that not all seats are filled with passengers who
would have previously chosen Dulles. For these various reasons, it is
likely that our “worst case” estimate of the number of passengers shifting
from Dulles to Reagan National is likely overstated. We did not include
the within-perimeter flights that were converted to beyond-perimeter slot
exemptions in our analysis because there were too many unknown
factors, such as whether the passengers those flights were connecting or
originating at Reagan National and the load factors on those flights.

Regarding the new beyond-perimeter flights’ impact on capacity
financing, using our estimate of the number of passengers who could shift
from using Dulles to Reagan National, we asked MWAA to estimate the
impact on the cost per enplaned passenger at the two airports. We also
discussed the methodology that MWAA used and determined that its
methods were sufficiently reliable for our purposes. We also reviewed
credit-rating reports on MWAA and discussed the reports with rating
agency officials. Moreover, we reviewed passenger traffic and airfare and
Dulles and BWI data relating to the beyond-perimeter flight destinations
that were added between 2000 and 2004.

With respect to the impact that the new beyond-perimeter flights could
have on aircraft noise, we reviewed the most recent noise compatibility
report on Reagan National and noise complaints since 2008 and
interviewed individuals from community groups concerned about aircraft
noise at Reagan National as well as officials from the Washington Council
of Governments, the Arlington County (Va.) Council, MWAA, and FAA.

To assess the impact that the new beyond-perimeter flights could have on
passenger security screening wait times, we asked the Transportation


Page 59                                          GAO-12-902 Airport Slot Controls
Appendix I: Objectives, Scope, and
Methodology




Security Administration (TSA) to estimate, using its security screening
throughput model, how the wait time would be affected. As inputs to the
model, we provided TSA with the flight times for the new beyond-
perimeter flights, data on the maximum number of passengers on those
flights, and the screening checkpoints that the passengers would use. We
did not report the precise estimates of wait time changes because they
would imply a greater level of precision than the model could reasonably
provide.

To review whether slot controls at U.S. slot-controlled airports are working
to reduce congestion, while maximizing available capacity and
encouraging competition, we analyzed airline schedule data and FAA slot
allocation data to determine the rate at which slots at slot-controlled
airports are scheduled. We also analyzed and compared and contrasted
airline schedule data on aircraft size and flight frequency and Department
of Transportation (DOT) data on average annual load factors and
passenger traffic at the four slot-controlled airports with the remaining
large hub airports 2 that are not slot-controlled. We interviewed officials
from DOT; FAA; Department of Justice (DOJ); MWAA; and the Port
Authority of New York and New Jersey, and interviewed representatives
from 11 airlines that currently serve or have interest in serving one or
more U.S. slot-controlled airport. We also contacted other stakeholders,
including representatives from Airlines For America; the International Air
Transport Association; Airports Council International, North America; and
others, about slot control issues. We also reviewed slot control
regulations for the four U.S. slot-controlled airports and obtained and
analyzed applicable federal laws, regulations, and agency orders. In



2
 For the purposes of this review, we defined large hub airports as all FAA-identified large
hub airports in the United States for calendar year 2010 (the most recent year available for
this list), excluding Honolulu. Federal law defines large hub airports as those commercial
service airports that have at least 1 percent of the passenger boardings. 49 U.S.C. §
40102(29). The large hub airports included in our study are: Hartsfield-Jackson Atlanta
International, Boston Logan International, Baltimore-Washington International Thurgood
Marshall, Charlotte/Douglas International, Ronald Reagan Washington National, Denver
International, Dallas-Fort Worth International, Detroit Metro Wayne County, Newark
Liberty International, Fort Lauderdale-Hollywood International, Dulles International,
George Bush Intercontinental, John F. Kennedy International, Las Vegas McCarran
International, Los Angeles International, LaGuardia International, Orlando International,
Chicago Midway, Miami International, Minneapolis-St. Paul International, Chicago O’Hare
International, Philadelphia International, Phoenix Sky Harbor International, San Diego
International, Seattle-Tacoma International, San Francisco International, Salt Lake City
International, and Tampa International.




Page 60                                                    GAO-12-902 Airport Slot Controls
Appendix I: Objectives, Scope, and
Methodology




addition, we reviewed studies on airline competition, airport capacity, and
demand management measures, including slot controls, as well as the
slot reform proposal for the European Union and the Worldwide Slot
Guidelines.

To calculate the rate at which slot allocations for scheduled passenger
flights are scheduled at slot-controlled airports, we divided the total daily
scheduled passenger flights within slot-controlled hours for a specific day
and airport by the total number of daily slot allocations (provided by FAA)
for that day and airport. We obtained airline schedule data from Innovata
for scheduled passenger flights for the third week of February and August
2011 for JFK, Newark, LaGuardia, and Reagan National. FAA provided
the daily total slot allocation numbers for scheduled passenger flights
during slot-controlled hours for the same days for JFK, Newark,
LaGuardia, and Reagan National. 3 FAA officials told us that the third
week of the month is most representative of the month. To further
examine schedule rates at Newark, we also obtained and analyzed daily
slot allocations and airline schedule data for the third Saturday of April,
May, June, July, September, and October 2011 for Newark. During our
review FAA reported discrepancies in the slot allocation numbers
provided to us. Specifically, FAA does not systematically update their slot
allocation records with the lists of returned slots received from airlines
each season for JFK and Newark. As a result, the data on slot allocations
used for our analysis for JFK and Newark generally represents the
number of flights initially allocated to each airline for the season rather
than the final number they held.

To examine whether flights at slot-controlled airports use smaller aircraft,
we conducted statistical analyses, including the construction of a logistic
regression model to assess whether flights at slot-controlled airports use
smaller aircraft than flights at other large hub airports that are not slot-
controlled, while controlling for other factors that are likely associated with
aircraft size, such as the flight distance (see app. V for more details on
our model and results). We analyzed Innovata airline schedule data for


3
 From the total slot allocations numbers, FAA removed all UPS and FedEx cargo
operations from the totals. Additionally, FAA only provided slot totals for those assigned
during slot-controlled hours. At JFK and Newark, slot-controlled hours are 6:00 a.m. to
10:59 p.m. Eastern Time. At LaGuardia, slot-controlled hours are 6:00 a.m. to 9:59 p.m.
Eastern Time, Monday through Friday, and 12:00 p.m.to 9:59 p.m. Eastern Time on
Sunday. At Reagan National, the airport is slot-controlled 24 hours a day, but we
examined high-demand hours, which are from 7:00 a.m. to 9:59 p.m. Eastern Time.




Page 61                                                    GAO-12-902 Airport Slot Controls
Appendix I: Objectives, Scope, and
Methodology




one Thursday each in February 2011 and in August 2011. 4 Thursdays are
generally considered to be a high-demand day, and February and August
are representative months for the winter and summer seasons,
respectively. We calculated the proportion of scheduled passenger flights
arriving or departing each large hub airport that use aircraft with 100
seats or fewer. We also compared and contrasted the proportion of flights
using aircraft with 100 seats or fewer departing or arriving slot-controlled
airports with other large hub airports that are not slot-controlled. We
tested that the difference was statistically significant. Since a variety of
factors, in addition to slot controls, may be correlated with the use of a
small aircraft, we developed a logistic regression model. The model
assessed whether flights to or from slot-controlled airports used an
aircraft with 100 seats or fewer compared with flights that arrive and
depart other large hub airports that are not slot-controlled, while
controlling for other factors that include (1) whether the origin or
destination of the flight is a high-tourism airport, 5 (2) whether the airline
marketing the flight is a legacy airline, 6 (3) the daily frequency of flights to
the same destination by the same airline, and (4) the flight distance. We
also conducted alternative scenarios examining aircraft with 124 seats or
fewer (the median-sized aircraft in our sample) and 80 seats or fewer.
See appendix V for more details on the logistic regression model and
results.

To examine closely-scheduled flights to the same destination by the same
airline, we analyzed airline schedule data, obtained from Innovata, for all
flights arriving or departing FAA-identified large hub airports on Thursday
February 10, 2011. For this analysis we calculated the number of minutes
between each flight and the flight immediately prior to and following on
the same route and by the same airline. We then identified those flights
with another flight less than or equal to15 or 30 minutes apart, and
calculated the proportion of such flights from each airport.


4
 The specific dates we selected were Feb.10, 2011, and Aug. 11, 2011.
5
 We defined Fort Lauderdale and Orlando, Florida, and Las Vegas, Nevada, as high-
tourism airports.
6
 The airlines classified as legacy included Delta Air Lines, United Airlines, American
Airlines, US Airways, and Continental Airlines. Legacy (sometimes called network) airlines
support large, complex hub-and-spoke operations with thousands of employees and
hundreds of aircraft (of various types), with flights to domestic communities of all sizes as
well as to international destinations. The modifier “legacy” derives from the fact that these
airlines were operating when the industry deregulated in 1978.




Page 62                                                    GAO-12-902 Airport Slot Controls
Appendix I: Objectives, Scope, and
Methodology




Average annual load factor data and average annual airline yield data for
large hub airports was obtained from DOT’s T-100 database, which
includes operational data collected from airlines and DOT’s origin and
destination survey, respectively. To calculate airlines’ proportion of
passenger traffic on domestic scheduled passenger flights, we obtained
and analyzed data from DOT’s origin and destination survey for 2011 at
each of the large hub airports.

To assess the reliability of airline schedule data, and DOT’s T-100 and
origin and destination survey, we (1) reviewed existing documentation
related to the data sources and (2) interviewed knowledgeable agency
officials and representatives about the data. We determined that the data
were sufficiently reliable for the purposes of this review.




Page 63                                        GAO-12-902 Airport Slot Controls
Appendix II: Comparative Information on
                                              Appendix II: Comparative Information on Major
                                              Airports in the New York City and Washington,
                                              D.C., Area


Major Airports in the New York City and
Washington, D.C., Area

                                                                                            Reagan
Airport             JFK                 LaGuardia                 Newark                    National                    Dulles              BWI
Location            Queens County,      Borough of                Newark and                Arlington County,           Chantilly, VA,      Anne Arundel
                    NY, 15 miles        Queens, New               Elizabeth, NJ,            VA, 3 miles from            26 miles from       County, MD, 10
                    from midtown        York City,                14 miles from             downtown                    downtown            miles south of
                    Manhattan           8 miles from              Manhattan                 Washington,                 Washington,         Baltimore and 32
                                        midtown                                             D.C.                        D.C.                miles north of
                                        Manhattan                                                                                           Washington,
                                                                                                                                            D.C.
Operator            Port Authority of   Port Authority of         Port Authority of         Metropolitan       Metropolitan       Maryland
                    New York and        New York and              New York and              Washington         Washington         Department of
                    New Jersey          New Jersey                New Jersey                Airports Authority Airports Authority Transportation’s
                                                                                                                                  Aviation
                                                                                                                                  Administration
Acres of land       4,930               680                       2,027                     733                         11,830              3,596
Number of           4                   2                         3                         3                           4                   4
runways             14,572 feet         both are 7,000            11,000 feet               6,869 feet 5,204            2 are 11,500 feet   10,502 feet
                    11,351 feet         feet                      9,980 feet 6,800          feet 4,911 feet             10,500 feet         9,501 feet 6,000
                    10,000 feet                                   feet                                                  9,400 feet          feet 5,000 feet
                    8,400 feet
Number of gatesa    124                 76                        115                       44                          139                 68
Number of           77                  17                        28                        11                          29                  11
airlinesa
Airline holding the Delta Air Lines     Delta Air Lines           United Airlines           US Airways                  United Airlines     Southwest
           b
most slots          (33 percent)        (44 percent)              (75 percent)              (54 percent)                (73 percent)c       Airlines
                                                                                                                                            (57 percent)c
2010                22,892,372          11,988,578                16,542,619                8,726,183                   11,251,041          10,770,973
enplanements
2011                23,585,722          11,983,322                16,786,357                9,039,072                   11,020,479          10,985,559
enplanements
Change in           3 percent           0 percent                 1.5 percent               3.6 percent                 (2 percent)         2 percent
enplanements
from 2010 to 2011
Domestic share      50 percent          96 percent                66 percent                98 percent                  72 percent          98 percent
of 2011
enplanements
                                              Sources: MWAA, Port Authority of New York and New Jersey, BWI, and BTS.
                                              a
                                                  As of June 2012.
                                              b
                                                  As of May 15, 2012.
                                              c
                                               Because Dulles and BWI are not slot-controlled airports, the data reflect the airlines’ percentage of
                                              total scheduled departures on May 15, 2012, rather than the percentage of slots held.




                                              Page 64                                                                         GAO-12-902 Airport Slot Controls
Appendix III: History of Slot Control Rules
               Appendix III: History of Slot Control Rules and
               Related Actions



and Related Actions

               Slot controls at Reagan National and the New York City area airports
               were first implemented in the late 1960s. Since then, the number of
               takeoffs and landings permitted at these airports has been revised
               periodically as well as the slot control procedures that airlines should
               follow. In addition, since the 1950s at LaGuardia and since the 1960s at
               Reagan National, airlines have been subject to limitations on the distance
               of their nonstop flights known as “perimeter rules.” Figure 7 shows when
               major events related to slot controls and perimeter rules occurred at the
               four airports.




               Page 65                                           GAO-12-902 Airport Slot Controls
                                        Appendix III: History of Slot Control Rules and
                                        Related Actions




Figure 7: Timeline of Major Actions Taken Regarding Slot Controls and Perimeter Rules at the Major New York City Area
Airports and Reagan National




                                        Slot controls were implemented in 1969 when the Federal Aviation
                                        Administration (FAA) designated JFK, LaGuardia, Newark, Chicago


                                        Page 66                                              GAO-12-902 Airport Slot Controls
Appendix III: History of Slot Control Rules and
Related Actions




O’Hare, and Washington National 1 as high density airports. In
implementing what became known as the High Density Rule, FAA
indicated that slot controls were needed to provide relief from excessive
delays at these airports. 2 Under the rule, FAA required airlines to obtain
slots to operate during certain hours at those airports and initially imposed
limits of 60 takeoffs and landings per hour at Newark, LaGuardia, and
Washington National, and 80 per hour at JFK. In 1970, FAA lifted the
High Density Rule at Newark because of a lower level of usage of the
slots at that airport as evidenced by, among other things, that the average
number of aircraft operations at peak hours was 18 less than the 60 slots
allowed. 3 In general, Newark was not slot-controlled from late October
1970 until 2008. 4 FAA implemented the High Density Rule in 1969 and,
because of its effectiveness in reducing congestion and delays, extended
it for an indefinite period of time in 1973. Airline scheduling committees,
operating under then-authorized antitrust immunity, conferred by the Civil
Aeronautics Board, initially allocated the slots. However, according to
FAA officials, after the Airline Deregulation Act of 1978 was enacted into
law, 5 increased airline competition made it more difficult for the
scheduling committees to agree on slot allocations and antitrust immunity
for the committees’ operations expired under the terms of the Airline
Deregulation Act. In 1986, as a part of its High Density Rule, FAA
replaced the scheduling committees with slot procedures such as
providing for the withdrawal of slots not used at least 65 percent of the
time, the voluntary return of slots, the reallocation of withdrawn and
returned slots, and allowing airlines to buy, sell, or lease their slots. 6 In
implementing this 1986 minimum slot usage rule, FAA cited the
Department of Justice’s (DOJ) concerns that a use or lose provision “may
be necessary to prevent large carriers or several large carriers from



1
 Slot controls at Chicago O’Hare Airport expired in 2002. In 1998, Washington National
Airport was renamed Ronald Reagan Washington National Airport.
2
 33 Fed. Reg. 17896 (Dec. 3, 1968).
3
 35 Fed. Reg. 16591 (Oct. 24, 1970).
4
 73 Fed. Reg. 29550 (May 21, 2008).
5
 Pub. L. No. 95-504, 92 Stat. 1705 (1978).
6
 50 Fed. Reg. 52180 (Dec. 20, 1985). The 1986 allocation and withdrawal rules at the
High Density Traffic Airports rules did not apply to Newark because the High Density Rule
was not then in effect at that airport.




Page 67                                                  GAO-12-902 Airport Slot Controls
Appendix III: History of Slot Control Rules and
Related Actions




hoarding slots in an attempt to restrict service to drive up fares or keep
smaller competitors from entering into or expanding in certain markets.”

In the 1990s, we, 7 Transportation Research Board, 8 Department of
Transportation (DOT) 9, and new entrant carriers, respectively, studied the
High Density Rule and presented reports and congressional testimony
that the High Density Rule was a barrier to competition and to improved
service, in part because new airlines were unable to establish service at
the slot-controlled airports because of the lack of slot availability. 10
Subsequently, in April 2000, Congress required that the application of the
High Density Rule be phased out at JFK and LaGuardia airports by
January 1, 2007, and directed DOT, in the interim to grant exemptions to
the High Density Rule to allow for a certain number of flights operated by
new entrant airlines and flights serving small hub and nonhub airports if
the aircraft had fewer than 71 seats. 11 By the fall of 2000, DOT had
granted the exemptions required by statute, and airlines had added 300
scheduled flights at LaGuardia, resulting in a 144 percent increase in the
average minutes of delay for arriving flights, from 15.52 minutes in March
2000 (a month before the law was enacted) to 37.86 minutes in
September 2000. Furthermore, FAA reported that by September 2000,
flight delays at LaGuardia accounted for 25 percent of the nation’s delays,
compared with 10 percent during the previous year. To address
congestion at LaGuardia, FAA reduced the number of daily slot
exemptions and distributed the exemptions through a lottery. 12




7
 GAO, Airline Competition: Industry Operating and Marketing Practices Limit Market
Entry, GAO/RCED-90-147 (Washington, D.C.: Aug. 29, 1990).
8
 National Research Council Transportation Research Board, Entry and Competition in the
U.S. Airline Industry: Issues and Opportunities, Special Report 255 (Washington, D.C.:
1999).
9
 Department of Transportation, Study of the High Density Rule: Report to Congress
(Washington, D.C.: May 1995).
10
    See, e.g., H. R. Rep. No. 106-67, at 77 (1999).
11
  Wendell H. Ford Aviation Investment and Reform Act for the 21st Century (AIR-21), Pub.
L. No. 106- 181, 114 Stat. 61, 108-109 (Apr. 5, 2000). AIR-21 also required the phase-out
of the application of the High Density Rule at O’Hare by July 1, 2002.
12
    66 Fed. Reg. 48157 (2001).




Page 68                                                  GAO-12-902 Airport Slot Controls
Appendix III: History of Slot Control Rules and
Related Actions




In 2006, FAA proposed a congestion management rule for LaGuardia that
would have replaced the High Density Rule when it expired on January 1,
2007. 13 The proposed rule would have encouraged the use of larger
aircraft by implementing an average aircraft size requirement (with
proposed exemptions for service to small communities). However,
because FAA did not expect to complete the congestion management
rule before the High Density Rule expired, the agency imposed temporary
slot controls at LaGuardia effective on January 1, 2007 (a maximum of 75
slots per hour) and required airlines to use slots at least 80 percent of the
time. 14 In addition, in 2007, the New York Aviation Rulemaking
Committee 15 explored ways to reduce congestion and efficiently allocate
scare capacity at the New York City area airports, including making
operational and infrastructure improvements; using congestion pricing or
auctions; requiring airlines to use larger aircraft; modifying gate
allocations; amending or eliminating the perimeter rule; implementing
priority aviation traffic preferences; or following International Air Transport
Association’s Worldwide Slot Guidelines.

After JFK and Newark airports experienced flight delays during the
summer of 2007, in October 2007, FAA announced operational targets of
81 flights per hour from 3:00 p.m. to 7:59 p.m. at JFK. In March 2008, by
order, FAA imposed temporary slot controls at JFK (a maximum of 81
slots per hour) 16 and in June 2008, by order, imposed temporary slot
controls at Newark (a maximum of 81 slots per hour) 17 and required
airlines at those airports to use slots at least 80 percent of the time. 18
Since 2008, FAA has extended the duration of the orders imposing slot
controls at JFK, LaGuardia, and Newark; they are now in effect until
October 2013 or until DOT’s Office of the Secretary and FAA issue a




13
  71 Fed. Reg. 51360 (Aug. 29, 2006).
14
  71 Fed. Reg. 77854 (Dec. 27, 2006).
15
  Committee members included officials from DOT’s Office of the Secretary and FAA, the
Port Authority of New York and New Jersey, and the State of New York as well as
representatives from airlines and consumer groups.
16
  73 Fed. Reg. 3510 (Jan. 18, 2008).
17
  73 Fed. Reg. 29550 (May 21, 2008).
18
  73 Fed. Reg. 3510 (Jan. 18, 2008); 73 Fed. Reg. 29550 (May 21, 2008).




Page 69                                                GAO-12-902 Airport Slot Controls
Appendix III: History of Slot Control Rules and
Related Actions




congestion management rule. 19 In April 2008, FAA proposed that the
congestion management rule for LaGuardia include two different options
for auctioning new or returned slots at LaGuardia whereby under one
option, the agency would have kept the proceeds and under the second
option, the proceeds would have gone to the incumbent carrier holding
the slot. 20 In May 2008, FAA proposed applying such congestion
management rules to JFK and Newark. 21 In October 2008, FAA published
final congestion management rules for JFK, LaGuardia, and Newark
assigning to existing operators the majority of slots at the airports and
creating a market by annually auctioning off a limited number of slots in
each of the first 5 years of the 10-year rule. 22 Before going into effect, on
October 9, 2009, FAA rescinded 23 the final rules for the congestion
management rules for JFK, LaGuardia, and Newark citing, among other
things, that the rulemakings had been highly controversial, a court
challenge, and a resulting December 2008 United States Court of
Appeals for the District of Columbia Circuit ruling 24 that put on hold the
implementation of the congestion management rules and their slot
auction provisions. 25 In its rescission of the congestion management
rules, FAA additionally cited fiscal year 2009 appropriations act
legislation 26 providing, among other things, a statutory prohibition on



19
  74 Fed. Reg. 845 (Jan. 8, 2009); 74 Fed. Reg. 51648 (Oct. 7, 2009); 74 Fed. Reg.
51650 (Oct. 7, 2009); 74 Fed. Reg. 51653 (Oct. 7, 2009); 76 Fed. Reg. 18616 (Apr. 4,
2011); 76 Fed. Reg. 18618 (Apr. 4, 2011); and 76 Fed. Reg. 18620 (Apr. 4, 2011).
20
  73 Fed. Reg. 20846 (Apr. 17, 2008).
21
  73 Fed. Reg. 29626 (May 21, 2008).
22
  73 Fed. Reg. 60544, 60574 (Oct. 19, 2008). 73 Fed. Reg. 60544 (Oct. 10, 2008).
23
  74 Fed. Reg. 52132 (Oct. 9, 2009); 74 Fed. Reg. 52134 (Oct. 9, 2009).
24
  See, Dec. 8, 2008 Order, Port Authority of N.Y. & N.J. v. FAA, No. 08-1329.
25
  In addition, in a legal opinion issued to multiple congressional requesters, we concluded
that FAA lacked the authority to auction slots and therefore also lacked authority to retain
and use auction proceeds. GAO Legal Opinion on Federal Aviation Administration—
Authority to Auction Airport Arrival and Departure Slots and to Retain and Use Auction
Proceeds, B-316796 (Sept. 30, 2008). The United States Department of Justice, Office of
Legal Counsel, opined that FAA “would not violate the [Anti-Deficiency Act] by issuing and
implementing the slot auction regulation.” U.S. Dep’t of Justice, Office of Legal Counsel,
Whether the Federal Aviation Administration’s Finalizing and Implementing of Slot Auction
Regulation Would Violate the Anti-Deficiency Act (Oct. 7, 2008).
26
  Omnibus Appropriations Act, 2009, Pub. L. No. 111-8, 123 Stat. 524, 921 (2009)




Page 70                                                    GAO-12-902 Airport Slot Controls
Appendix III: History of Slot Control Rules and
Related Actions




using fiscal year 2009 funds to take any action regarding the auctioning of
rights or permission to conduct airlines operations at U.S. commercial
airports. 27 Subsequent to the rescission of the slot auction rules, FAA
sought a dismissal of the case that was granted on October 14, 2009. 28

In 2010, DOT’s Office of the Secretary and FAA began drafting a “Slot
Management and Transparency” rule for the New York City area
airports. 29 According to DOT, this rulemaking would replace the current
temporary orders limiting scheduled aircraft operations at JFK,
LaGuardia, and Newark with a more permanent rule to address
congestion and delay while also promoting fair access and competition.
DOT’s rulemaking is also aimed at ensuring that congestion and delay
are minimized by limiting scheduled and unscheduled aircraft operations.
Although the temporary orders prohibit the buying and selling of slots at
the New York City area airports, according to DOT, the rulemaking would
establish a secondary market for U.S. and foreign air carriers to buy, sell,
trade, and lease slots at the three airports and “allow carriers serving or
seeking to serve the New York area airports to exchange slots as their
business models and strategic goals require.” Under DOT’s rulemaking
schedule, the department was to provide the draft to the Office of
Management and Budget for review by April 25, 2011, but as of July
2012, that had not yet occurred. At present, the High Density Rule applies
only to Reagan National.

In 2009, Delta Air Lines and US Airways proposed trading slots that they
held at LaGuardia and Reagan National, a trade that would allow Delta to
establish a domestic hub at LaGuardia and US Airways to enhance its
network at Reagan National. The airlines sought FAA’s permission
because the transaction was considered a purchase of slots that, under
the order limiting operations at LaGuardia, was not allowed. Under the
proposal, Delta would transfer 42 slot pairs it held at Reagan National in
exchange for 125 slot pairs at LaGuardia, among other provisions. DOT’s
Office of the Secretary and FAA granted the waiver request with
conditions to counteract a substantial increase in market concentration by
the two airlines that would result at those airports and to provide an


27
  74 Fed. Reg. 52132 (Oct. 9, 2009); 74 Fed. Reg. 52134 (Oct. 9, 2009).
28
  See Oct. 14, 2009 Order, Port Authority of N.Y. & N.J. v. FAA, No 08-1329.
29
  Reagan National is not included in this rulemaking because it is covered under FAA’s
High Density Rule regulations.




Page 71                                                  GAO-12-902 Airport Slot Controls
                              Appendix III: History of Slot Control Rules and
                              Related Actions




                              opportunity for low cost airlines to compete, which required them to divest
                              14 pairs of slots at Reagan National and 20 pairs of slots at LaGuardia to
                              new entrant and limited incumbent airlines (those holding less than 5
                              percent of slots at Reagan National and LaGuardia). 30 However, Delta
                              and US Airways decided not to proceed with the transaction under these
                              conditions. In 2011, the two airlines reapplied for permission to exchange
                              slots. Under the revised proposal, Delta would acquire 132 slot pairs (265
                              slots) at LaGuardia from US Airways and US Airways would acquire 42
                              slot pairs at LaGuardia from Delta, among other provisions. Delta would
                              also pay US Airways $65 million. In their application, the two airlines said
                              consumers would benefit from the transaction through, among other
                              things, improved connectivity at LaGuardia and Reagan National and by
                              using larger aircraft. DOT’s Office of the Secretary and FAA approved the
                              transaction and granted the waiver request from the temporary order’s
                              prohibition on purchasing slots, with conditions, citing expected consumer
                              benefits and the increased presence of low cost airlines at LaGuardia and
                              Reagan National since the previous application. 31 The conditions required
                              Delta and US Airways to divest to new entrants and limited incumbents
                              fewer slots—8 slot pairs at Reagan National and 16 slot pairs at
                              LaGuardia—than DOT’s Office of the Secretary and FAA had required
                              when they approved the first application. DOT’s Office of the Secretary
                              and FAA directed that the divested slots be sold through a bidding
                              process, permitting Delta and US Airways to keep the proceeds. In 2011,
                              Jet Blue purchased 8 slot pairs at LaGuardia for $32 million and 8 slot
                              pairs at Reagan National for $40 million, and West Jet purchased 8 slot
                              pairs at LaGuardia for $17.6 million.


Perimeter Rules Subject       Since the 1950s, the Port Authority of New York and New Jersey, which
Airlines to Flight Distance   operates the New York City area airports, has had a rule limiting the
Limitations to and from       distance of nonstop flights into and out of LaGuardia known as the
                              perimeter rule. Until 1984, the perimeter rule was informal and prohibited
Reagan National and           nonstop flights into or out of LaGuardia to or from destinations more than
LaGuardia                     2,000 miles from the airport. In 1984, the Port Authority formally instituted
                              a 1,500-mile perimeter rule but exempted service to Denver, which is
                              more than 1,600 miles from LaGuardia, and flights on Saturdays.



                              30
                                75 Fed. Reg. 26322 (May 11, 2010).
                              31
                                76 Fed. Reg. 63702 (Oct. 13, 2011).




                              Page 72                                           GAO-12-902 Airport Slot Controls
Appendix III: History of Slot Control Rules and
Related Actions




In 1966, airlines at Reagan National voluntarily agreed to limit nonstop
flights from the airport to destinations less than 650 miles away, with
some exceptions. The Metropolitan Washington Airports Authority Act of
1986, enacted into law in October 1986, authorized the transfer of
authority over Reagan National and Dulles from the federal government
to the Metropolitan Washington Airports Authority, set Reagan National’s
perimeter rule at 1,250 miles, and prohibited the Airports Authority from
increasing slot limits at Reagan National set in FAA’s High Density Rule. 32
Two federal laws respectively enacted in 2000 and 2003 (the Wendell H.
Ford Aviation Investment and Reform Act for the 21st Century 33 (AIR-21)
and the Vision 100-Century of Aviation Reauthorization Act 34 (Vision
100)) enacted statutory modifications relating to the slot and perimeter
rules at Reagan National. Among other things, AIR-21 required the
Secretary of Transportation to grant 12 beyond perimeter and 12 within
perimeter exemptions at Reagan National. Vision 100 amendments
increased these numbers to 24 beyond perimeter exemptions and 20
within perimeter exemptions. 35 In effect, AIR-21 and Vision 100,
combined, required the Secretary of Transportation to award 22 round
trips, 12 with service to airports beyond the perimeter and 10 with service
to airports within the perimeter. 36 In 2012, the FAA Modernization and
Reform Act of 2012 37 was enacted into law, which reauthorized FAA,
authorized 16 additional beyond perimeter exemptions at Reagan
National comprised of eight within-perimeter flights to large hub airports to
be converted to nonstop beyond-perimeter flights (four round-trips) and



32
  Pub. L. No. 99-500, 100 Stat. 1783, 1783-375 (1986); Pub. L. No. 99-591, 100 Stat.
3341, 3341-376 (1986).
33
  Pub. L. No. 106-181, 114 Stat. 61 (2000).
34
  Pub. L. No. 108-176, 117 Stat. 2490 (2003).
35
  In addition, these two laws together mandated that the slot exemptions at Reagan
National (1) could not be used for operations between the hours of 10:00 p.m. and 7:00
a.m. and (2) could not increase the number of operations in any 1-hour period during the
hours between 7:00 a.m. and 9:59 p.m. by more than 3 operations.
36
  More specifically, these two laws together required the Secretary of Transportation to
grant 24 beyond perimeter exemptions and 20 slot exemptions within the 1,250-mile
perimeter.
37
  Pub. L. No. 112-95, 126 Stat. 11 (2012). In addition the FAA Modernization and Reform
Act of 2012 amended the hourly limitations provision whereby such slot exemptions may
not increase the number of operations in any 1-hour period during the hours between 7:00
a.m. and 9:59 p.m. by more than 5 operations




Page 73                                                   GAO-12-902 Airport Slot Controls
Appendix III: History of Slot Control Rules and
Related Actions




eight new additional nonstop beyond-perimeter flights (four round trips).
With respect to the converted flights, under the 2012 act, each incumbent
airline with 40 or more slots at Reagan National could convert one within-
perimeter slot pair to one beyond-perimeter slot pair, 38 while the new
beyond-perimeter slots were to be awarded to new entrants or airlines
holding fewer than 40 slots. As the converted nonstop beyond-perimeter
destinations, the incumbent airlines chose Los Angeles (American
Airlines); Salt Lake City (Delta Air Lines); San Diego (US Airways); and
San Francisco (United Airlines). Seven airlines applied for the four new
beyond-perimeter slot pair exemptions, which, on May 14, 2012, DOT
awarded to Alaska Airlines for Portland, Ore.; Jet Blue for San Juan,
Puerto Rico; Southwest Airlines for Austin, Tex.; and Virgin America for
San Francisco.




38
  Four airlines each qualified to convert one slot pair.




Page 74                                                    GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
               Appendix IV: Trends in Average Annual
               Passenger Traffic and Airfares at Selected
               Washington, D.C., Area Airports


Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports
               Figures 8 through 31 below show the trends in average annual passenger
               traffic and airfares at Baltimore-Washington International Thurgood
               Marshall Airport (BWI) and Dulles International Airport (Dulles) for
               destinations granted beyond-perimeter slot exemptions at Reagan
               National Airport (Reagan National) under AIR-21 and Vision 100 between
               2000 and 2004.

               Figure 8: Indexed Average Passenger Traffic for All BWI Domestic Flights and
               Flights between BWI and Denver International Airport, 2000 to 2011




               Page 75                                             GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 9: Indexed Average Annual Airfare for All BWI Domestic Flights and Flights
between BWI and Denver International Airport, 2000 to 2011




Page 76                                              GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 10: Indexed Average Annual Passenger Traffic for All BWI Domestic Flights
and Flights between BWI and Las Vegas McCarran International Airport, 2000 to
2011




Page 77                                             GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 11: Indexed Average Annual Airfares for All BWI Domestic Flights and
Flights between BWI and Las Vegas McCarran International Airport, 2000 to 2011




Page 78                                             GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 12: Indexed Average Passenger Traffic for All BWI Domestic Flights and
Flights between BWI and Los Angeles International Airport, 2000 to 2011




Page 79                                             GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 13: Indexed Average Annual Airfares for All BWI Domestic Flights and
Flights between BWI and Los Angeles International Airport, 2000 to 2011




Page 80                                             GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 14: Indexed Average Passenger Traffic for All BWI Domestic Flights and
Flights between BWI and Phoenix Sky Harbor International Airport, 2000 to 2011




Page 81                                             GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 15: Indexed Average Airfares for All BWI Domestic Flights and Flights
between BWI and Phoenix Sky Harbor International Airport, 2000 to 2011




Page 82                                              GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 16: Indexed Average Passenger Traffic for All BWI Domestic Flights and
Flights between BWI and Salt Lake City International Airport, 2000 to 2011




Page 83                                             GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 17: Indexed Average Airfares for All BWI Domestic Flights and Flights
between BWI and Salt Lake City International Airport, 2000 to 2011




Page 84                                              GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 18: Indexed Average Passenger Traffic for All BWI Domestic Flights and
Flights between BWI and Seattle-Tacoma International Airport, 2000 to 2011




Page 85                                             GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 19: Indexed Average Airfares for All BWI Domestic Flights and Flights
between BWI and Seattle-Tacoma International Airport, 2000 to 2011




Page 86                                              GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 20: Indexed Average Passenger Traffic for All Dulles Domestic Flights and
Flights between Dulles and Denver International Airport, 2000 to 2011




Page 87                                              GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 21: Indexed Average Airfares for All Dulles Domestic Flights and Flights
between Dulles and Denver International Airport, 2000 to 2011




Page 88                                              GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 22: Indexed Average Annual Passenger Traffic for All Dulles Domestic
Flights and Flights between Dulles and Las Vegas McCarran International Airport,
2000 to 2011




Page 89                                             GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 23: Indexed Average Annual Airfares for All Dulles Domestic Flights and
Flights between Dulles and Las Vegas McCarran International Airport, 2000 to 2011




Page 90                                             GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 24: Indexed Average Annual Passenger Traffic for All Dulles Domestic
Flights and Flights between Dulles and Los Angeles International Airport, 2000 to
2011




Page 91                                              GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 25: Indexed Average Annual Airfares for All Dulles Domestic Flights and
Flights between Dulles and Los Angeles International Airport, 2000 to 2011




Page 92                                              GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 26: Indexed Average Annual Passenger Traffic for All Dulles Domestic
Flights and Flights between Dulles and Phoenix Sky Harbor International Airport,
2000 to 2011




Page 93                                              GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 27: Indexed Average Annual Airfares for All Dulles Domestic Flights and
Flights between Dulles and Phoenix Sky Harbor International Airport, 2000 to 2011




Page 94                                              GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 28: Indexed Average Annual Passenger Traffic for All Dulles Domestic
Flights and Flights between Dulles and Salt Lake City International Airport, 2000 to
2011




Page 95                                               GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 29: Indexed Average Airfares for All Dulles Domestic Flights and Flights
between Dulles and Salt Lake City International Airport, 2000 to 2011




Page 96                                              GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 30: Indexed Average Annual Passenger Traffic for All Dulles Domestic
Flights and Flights between Dulles and Seattle-Tacoma International Airport, 2000
to 2011




Page 97                                              GAO-12-902 Airport Slot Controls
Appendix IV: Trends in Average Annual
Passenger Traffic and Airfares at Selected
Washington, D.C., Area Airports




Figure 31: Indexed Average Annual Airfares for All Dulles Domestic Flights and
Flights between Dulles and Seattle-Tacoma International Airport, 2000 to 2011




Page 98                                              GAO-12-902 Airport Slot Controls
Appendix V: Logistic Regression Analysis of
                          Appendix V: Logistic Regression Analysis of
                          Slot-Controlled Airports and Aircraft Size



Slot-Controlled Airports and Aircraft Size

                          The regression analysis explained in this appendix is designed to assess
                          whether airlines tend to use smaller aircraft on flights to or from slot-
                          controlled airports. The analysis is based on data for all scheduled
                          passenger airline flights between large hub domestic airports for two
                          different Thursdays (one in winter and one in summer) in 2011, and
                          assesses whether and how several factors—including whether the flight is
                          scheduled to arrive or depart a slot-controlled airport—are correlated with
                          the use of smaller aircraft. For the base-case analysis we defined a
                          smaller aircraft as ones with 100 seats or fewer. Specifically, this
                          appendix discusses (1) the model hypothesis, (2) the conceptual
                          framework of the model, (3) data source and variable definitions, and (4)
                          base-case model results and sensitivity analyses.


Model Hypothesis          As discussed in the report, a variety of parties—airlines, the Port Authority
                          of New York and New Jersey, and Department of Justice—assert that
                          airlines currently operating out of slot-controlled airports use some of their
                          slots in ways to keep out their competitors. One of these ways is using
                          smaller aircraft, on average, at the four slot-controlled airports in an effort
                          to meet the 80 percent usage rule.


Conceptual Framework of   As discussed in our report, we examined the number and proportion of
Model                     scheduled passenger flights using smaller aircraft departing or arriving
                          slot-controlled airports (which includes four airports) and the remaining 24
                          large hub airports. 1 In comparing flights arriving and departing the slot-
                          controlled airports with other large hub airports that are not slot-
                          controlled, we found that the four slot-controlled airports had a higher
                          proportion of scheduled flights with smaller aircraft than flights to or from
                          the remaining 24 large hub airports. We found that these differences are
                          statistically significant.

                          Many factors may contribute to an airline’s decision to use smaller
                          aircraft, such as location and proximity to other major markets, restrictions
                          on distance flown, the type of passengers being served, and the airline’s
                          route structure (e.g., a hub-and-spoke structure verses point-to-point). For
                          example, all four of the slot-controlled airports are on the Eastern



                          1
                           We included all large hub airports in the United States identified by FAA for calendar year
                          2010 (the most recent year available for this list), excluding Honolulu International.




                          Page 99                                                    GAO-12-902 Airport Slot Controls
                            Appendix V: Logistic Regression Analysis of
                            Slot-Controlled Airports and Aircraft Size




                            Seaboard of the United States with many destinations—large and small—
                            within reasonably close proximity. Because shorter distance flights are
                            generally served by smaller aircraft, flights to and from the slot-controlled
                            airports may have a higher proportion of flights with small aircraft. Also,
                            LaGuardia and Reagan National are subject to perimeter rules of 1,500
                            and 1,250 miles, respectively, that limits the distance of flights from these
                            airports. A restriction on flight distance such as this would likely skew
                            aircraft use toward smaller sized craft. Whether an airport being served is
                            a business versus leisure (tourist) market may also affect the size of
                            aircraft—for example, flights serving leisure markets may tend to use
                            larger aircraft less frequently than a business market. And finally, an
                            airline’s route structure could influence the distance of flights or aircraft
                            size—for example, an airport that serves as a hub for an airline will
                            require more short distance flights than a point-to-point business model.

                            Since a variety of factors in addition to slot controls may be correlated
                            with the use of a small aircraft, we developed a logistic regression model
                            that controls for other factors that may be correlated with aircraft size. To
                            isolate the correlation between flights arriving or departing slot-controlled
                            airports and aircraft size, we defined aircraft as small if its capacity was
                            100 seats or fewer. 2 We applied a logistic regression technique (or logit
                            model) in which the scheduled flights are defined as using small aircraft
                            (100 seats or fewer) or large (more than 100 seats), and assessed how
                            each of a set of independent factors (including whether the flights arrives
                            or departs a slot-controlled airport) correlates with the odds of a small
                            aircraft used on that flight. All factors related to aircraft size may not be
                            fully controlled for by this model. Also, as with any regression model, this
                            analysis does not prove a causal relationship between aircraft size and
                            slot-controlled airport, only the correlation between the two.


Data Source and Variables   The primary data source for the model is airlines’ schedule data from
                            Innovata, which provides information on all airlines scheduled flights. The
                            level of observation for this analysis is each individually scheduled




                            2
                             One hundred seats is generally the maximum seat capacity of most regional jet aircraft.
                            See below for a discussion of alterative cutoffs for the definition of a small aircraft.




                            Page 100                                                 GAO-12-902 Airport Slot Controls
                          Appendix V: Logistic Regression Analysis of
                          Slot-Controlled Airports and Aircraft Size




                          passenger flight. We used all flights between the 28 large hub airports 3 on
                          two Thursdays in 2011—February 10, 2011, and August 11, 2011—for a
                          total of 13,901 flights. Thursdays are generally considered to be a high-
                          demand day, and February and August are representative months for the
                          winter and summer seasons, respectively. The schedule data provide
                          relevant information for each flight, including the size of aircraft, the date
                          and time of the scheduled flight, the origin and destination airport, and the
                          marketing and operating airline.

Dependent Variable        The dependent variable for the analysis is a dummy (or indicator)
                          variable—that is, a variable that takes a value of one or zero depending
                          on the presence or absence of some characteristic. In this case, the
                          dependent variable takes the value of one if the scheduled flight uses a
                          small aircraft size (100 or fewer seats) and the value of zero if the flight
                          uses a large aircraft (greater than 100 seats).

Independent Variables 4   Slot-controlled airport (variable of interest): A dummy variable that
                          takes the value of one if either the origin or destination of the flight is a
                          slot-controlled airport. We used several different slot-controlled airport
                          dummy variables, including one for all four slot-controlled airports as a
                          group (slot4), one for only the three New York City area slot-controlled
                          airports (slot3) and one for each individual slot-controlled airport alone.

                          Legacy airline: A dummy variable that takes the value of one if the flight
                          is marketed by one of the legacy airlines we defined (Delta Air Lines,
                          United Airlines, American Airlines, US Airways, and Continental Airlines)
                          and takes the value of zero for all other marketing airlines. Airlines vary in
                          their fleet mix. Some airlines—either on their own or through partnerships


                          3
                           The large hub airports included in our analysis are: Atlanta Hartsfield International,
                          Boston Logan International, Baltimore-Washington International, Charlotte/Douglas
                          International, Reagan National, Denver International, Dallas-Fort Worth International,
                          Detroit Metro Wayne County, Newark, Fort Lauderdale-Hollywood International,
                          Washington Dulles International, George Bush Intercontinental, JFK, Las Vegas McCarran
                          International, Los Angeles International, LaGuardia, Orlando International, Chicago
                          Midway, Miami International, Minneapolis-St. Paul International, Chicago O’Hare
                          International, Philadelphia International, Phoenix Sky Harbor International, San Diego
                          International Lindbergh, Seattle-Tacoma International, San Francisco International, Salt
                          Lake City International, and Tampa International.
                          4
                            In addition to the independent variables listed here, we tested whether scheduled flights
                          in February (winter) had any greater likelihood of using a small aircraft than scheduled
                          flights in August (summer). We did not find any significant difference, so we did not
                          include this variable in our analysis.




                          Page 101                                                   GAO-12-902 Airport Slot Controls
Appendix V: Logistic Regression Analysis of
Slot-Controlled Airports and Aircraft Size




with other airlines—operate a varied fleet mix with both large and small
aircraft, while other airlines tend to fly only a few aircraft types that may
have little variation in the size of aircraft. We expect that when the
marketing airline is a legacy airline, all else equal, there is a greater
propensity to use smaller aircraft because these airlines and their
partners tend to deploy a wide range of aircraft size, including many
commuter-sized aircraft, across a broad range of types of routes.

Flight distance: A categorical variable that is derived from distance
quartiles of the flights analyzed. The four distance categories are less
than 325 miles, 325 to 602 miles, 603 to 998 miles, and greater than 998
miles. Flight distance is a key factor in the decision underlying the
assignment of aircraft to flights. We used the flights that fell into the
second quartile (325 miles to 602 miles in distance) as the reference
category that the other flight distance dummy variables are compared to.
We expect that shorter flights would tend to use smaller aircraft.

Flight frequency for an airline route: A categorical variable indicating
the importance of a particular route to an airline’s operations. We define
flight frequency on an airline’s route as the total number of flights
operated by the same airline on a given day from the same origin to the
same destination. Few frequencies on an airline’s route indicate that the
route may not be highly connected to the airline’s network, while
numerous frequencies indicate a greater degree of connection, possibly
involving one of the airline’s hub airports. 5 We divided the number of
flights on an airline’s route into quartiles and used the two lower quartiles
(which together included all airlines’ routes with fewer than 6 flights per
day) as reference categories that the remaining two quartiles are
compared to. The third quartile includes airlines’ routes with 6 to 8 flights
and day and the fourth quartile includes routes with 9 or more flights per
day. We expect that greater dedication to serving a route (higher flight
frequency) will be correlated with reduced odds of using a small aircraft.

Tourism: A dummy variable that takes the value of one if either the origin
or destination of a flight is one of three cities that have substantial tourist
travel (Las Vegas, Nevada; Orlando, Florida; and Fort Lauderdale,
Florida) based on hotel revenues relative to population and our judgment



5
 Flight frequency also serves as a proxy for an airport hub variable that would consider
whether the airport serves as a network hub for an airline.




Page 102                                                  GAO-12-902 Airport Slot Controls
                           Appendix V: Logistic Regression Analysis of
                           Slot-Controlled Airports and Aircraft Size




                           about which cities have substantial tourist travel. We defined Las Vegas
                           McCarran International, Orlando International, and Fort Lauderdale-
                           Hollywood International airports as the three tourism-oriented airports in
                           the analysis. We expect that smaller aircraft will be less likely to be used
                           on flights to and from these airports.


Base-Case Model Results    Table 7 provides results of our base-case model that defines a small
and Sensitivity Analysis   aircraft as one with 100 seats or fewer. The results are expressed as
                           odds ratios, which represents the odds that a flight uses a small aircraft
                           (100 seats or fewer) divided by the odds that a flight uses a large aircraft
                           (more than 100 seats). The odds ratio for the variable of interest (slot4) is
                           greater than 1, indicating that a flight to or from any of the four slot-
                           controlled airports is more likely to use a small aircraft than a flight to or
                           from the other 24 large hub airports that are not slot-controlled. 6
                           Specifically, the odds that a flight to or from a slot-controlled airport uses
                           a smaller aircraft is 75 percent higher than the odds for a flight to and
                           from the other 24 large hub airports that are not slot-controlled. This result
                           and the odds ratios for all other variable in the base-case scenario are
                           statistically significant.




                           6
                            Odds are a ratio of the probability that an event will occur versus the probability that the
                           event will not occur, or probability / (1-probability). Odds ratios, therefore, are simply a
                           ratio of odds; in general they refer to the ratio of the odds of an event (small aircraft)
                           occurring in the exposed group (slot-controlled airports) versus the unexposed group
                           (other large hubs).




                           Page 103                                                     GAO-12-902 Airport Slot Controls
Appendix V: Logistic Regression Analysis of
Slot-Controlled Airports and Aircraft Size




Table 7: Logistic Regression Results for Slot-Controlled-Airport and Aircraft-Size
Model, Small Aircraft with 100 Seats or Fewer

    Variable                                                                                Odds ratioa
    To or from any of the four slot-controlled airports                                              1.75
    To or from tourism airport                                                                       0.16
    Marketed by a legacy airline                                                                     8.71
    6 to 8 flights per day by same airline from same origin to same destination                      0.32
    9 or more flights per day by same airline from same origin to same                               0.11
    destination
    Flight distance less than 325 miles                                                              3.99
    Flight distance between 603 and 998 miles                                                        0.40
    Flight distance more than 998 miles                                                              0.06
Source: GAO analysis of airline schedule data
a
 All odds ratios are significant at the p-value <0.01 (1 percent level). The significance of the
coefficients in the models was evaluated using a simple Wald test statistic, which is asymptotically
equivalent to the likelihood-ratio chi-square statistic. The percentage of flights correctly predicted by
the model to be on small versus larger aircraft was 85-percent. The measure is known as the
concordance rate.


In addition to the model discussed above, we also ran with the model with
alternative definitions of the slot-controlled-airport dummy variable. Table
8 shows the odds ratios from these 5 addition model runs for the variously
defined slot-controlled-airport variable. The results for the independent
variables in these model runs were reasonably stable both in the direction
and magnitude of the odds ratios.

Table 8: Logistic Regression Results for Alternative Slot-Controlled-Airport Variable
Definitions, Small Aircraft with 100 Seats or Fewer

    Slot-controlled-airport variable                                                        Odds Ratio
    To or from any of the three New York City area slot-controlled airports                         1.68a
    To or from LaGuardia                                                                            2.08a
    To or from JFK                                                                                  1.63a
    To or from Newark                                                                               1.22b
    To or from Reagan National                                                                      1.72a
Source: GAO analysis of airline schedule data
a
 Indicates odds ratio is significant at a p-value > 0.01.
b
 Indicates odds ratio is significant at a p-value < 0.1.


We also conducted sensitivity analyses by testing two alternative
definitions for small aircraft. We ran the same models described above,



Page 104                                                            GAO-12-902 Airport Slot Controls
Appendix V: Logistic Regression Analysis of
Slot-Controlled Airports and Aircraft Size




but defined a small aircraft as (1) aircraft with 124 seats or fewer (which is
the median aircraft size for this data set) and (2) aircraft with 80 seats or
fewer (which excludes the largest regional jets). Tables 9 and 10 show
the results of these model runs, including the results for each of the six
different slot-controlled airport dummy variable definitions. All other
independent variables included in the models were robust—both in terms
of the direction, magnitude, and statistical significance of odds ratios—
across the three definitions for small aircraft.

In addition, the odds ratios for all definitions of the slot-controlled-airport
dummy variable are greater than 1—indicating a greater odds of a flight
using a small aircraft to or from the slot-controlled airport, however the
slot variable was defined. The magnitude of the odds ratio, however,
varies based on the definition of a small aircraft (that is, a cutoff of 80,
100, or 124 seats). Also, the degree of the statistical significance of the
odds ratios for Newark and JFK vary. The odds ratio for Newark becomes
more significant under the alternative aircraft definitions (i.e., a cutoff of
80 or 124 seats). And, while odds ratio for the JFK dummy variable was
significant in both the base case (100-seat cutoff) and the 124-seat cutoff,
the odds ratio for JFK is not statistically significant in the 80 seat cutoff.

Table 9: Logistic Regression Results for Alternative Slot-Controlled-Airport Variable
Definitions, Small Aircraft with 124 Seats or Fewer

    Slot-controlled airport variable                                             Odds Ratio
    To or from any of the four slot-controlled airports                                 2.49a
    To or from any of the three New York City area slot-controlled airports             2.18a
    To or from LaGuardia                                                                3.68a
                                                                                            a
    To or from JFK                                                                      1.53
    To or from Newark                                                                   1.48a
    To or from Reagan National                                                          4.29a
Source: GAO analysis of airline schedule data
a
 Indicates odds ratio is significant at p-value < 0.01.




Page 105                                                      GAO-12-902 Airport Slot Controls
Appendix V: Logistic Regression Analysis of
Slot-Controlled Airports and Aircraft Size




Table 10: Logistic Regression Results for Alternative Slot-Controlled-Airport
Variable Definitions, Small Aircraft with 80 Seats or Fewer

    Slot-controlled airport variable                                             Odds Ratio
    To or from any of the four slot-controlled airports                                 1.55a
    To or from any of the three New York City area slot-controlled airports             1.61a
    To or from LaGuardia                                                                2.17a
    To or from JFK                                                                       1.22
    To or from Newark                                                                   1.26b
    To or from Reagan National                                                          1.41a
Source: GAO analysis of airline schedule data
a
 Indicates odds ratio is significant at p-value < 0.01.
b
 Indicates odds ratio is significant at p-value < 0.05.




Page 106                                                      GAO-12-902 Airport Slot Controls
Appendix VI: Comments from the
              Appendix VI: Comments from the Metropolitan
              Washington Airports Authority



Metropolitan Washington Airports Authority




              Page 107                                      GAO-12-902 Airport Slot Controls
Appendix VI: Comments from the Metropolitan
Washington Airports Authority




Page 108                                      GAO-12-902 Airport Slot Controls
Appendix VI: Comments from the Metropolitan
Washington Airports Authority




Page 109                                      GAO-12-902 Airport Slot Controls
Appendix VI: Comments from the Metropolitan
Washington Airports Authority




Page 110                                      GAO-12-902 Airport Slot Controls
Appendix VII: Comments from the Port
              Appendix VII: Comments from the Port
              Authority of New York and New Jersey



Authority of New York and New Jersey




              Page 111                               GAO-12-902 Airport Slot Controls
Appendix VIII: GAO Contact and Staff
                  Appendix VIII: GAO Contact and Staff
                  Acknowledgments



Acknowledgments

                  Susan Fleming (202) 512-2834 or flemings@gao.gov
GAO Contact
                  In addition to the individual named above, Amy Abramowitz, Paul
Staff             Aussendorf, Geoff Hamilton, Bob Homan, David Hooper, Delwen Jones,
Acknowledgments   Kirsten Lauber, Josh Ormond, Dae Park, and Gretchen Snoey made
                  significant contributions to this report.




(541086)
                  Page 112                                   GAO-12-902 Airport Slot Controls
GAO’s Mission         The Government Accountability Office, the audit, evaluation, and
                      investigative arm of Congress, exists to support Congress in meeting its
                      constitutional responsibilities and to help improve the performance and
                      accountability of the federal government for the American people. GAO
                      examines the use of public funds; evaluates federal programs and
                      policies; and provides analyses, recommendations, and other assistance
                      to help Congress make informed oversight, policy, and funding decisions.
                      GAO’s commitment to good government is reflected in its core values of
                      accountability, integrity, and reliability.

                      The fastest and easiest way to obtain copies of GAO documents at no
Obtaining Copies of   cost is through GAO’s website (www.gao.gov). Each weekday afternoon,
GAO Reports and       GAO posts on its website newly released reports, testimony, and
                      correspondence. To have GAO e-mail you a list of newly posted products,
Testimony             go to www.gao.gov and select “E-mail Updates.”

Order by Phone        The price of each GAO publication reflects GAO’s actual cost of
                      production and distribution and depends on the number of pages in the
                      publication and whether the publication is printed in color or black and
                      white. Pricing and ordering information is posted on GAO’s website,
                      http://www.gao.gov/ordering.htm.
                      Place orders by calling (202) 512-6000, toll free (866) 801-7077, or
                      TDD (202) 512-2537.
                      Orders may be paid for using American Express, Discover Card,
                      MasterCard, Visa, check, or money order. Call for additional information.
                      Connect with GAO on Facebook, Flickr, Twitter, and YouTube.
Connect with GAO      Subscribe to our RSS Feeds or E-mail Updates. Listen to our Podcasts.
                      Visit GAO on the web at www.gao.gov.
                      Contact:
To Report Fraud,
Waste, and Abuse in   Website: www.gao.gov/fraudnet/fraudnet.htm
                      E-mail: fraudnet@gao.gov
Federal Programs      Automated answering system: (800) 424-5454 or (202) 512-7470

                      Katherine Siggerud, Managing Director, siggerudk@gao.gov, (202) 512-
Congressional         4400, U.S. Government Accountability Office, 441 G Street NW, Room
Relations             7125, Washington, DC 20548

                      Chuck Young, Managing Director, youngc1@gao.gov, (202) 512-4800
Public Affairs        U.S. Government Accountability Office, 441 G Street NW, Room 7149
                      Washington, DC 20548




                        Please Print on Recycled Paper.