Escolar Documentos
Profissional Documentos
Cultura Documentos
67
introduction allocation approaches for LGA, since the HDR rule was
legislated to be removed by January 2007. (In fact, LGA
R e a s o n F o u n d a t i o n • w w w . r e a s o n . o r g
driven mechanisms such as auctions date back to 1979
with the work of Grether, Issac, and Plot [13] and
Rassenti, Smith and Bulfin [14]. European research-
ers, DotEcon Ltd [15] and National Economic Research
Associates (NERA) [16], conducted macro-economic
analysis to conclude that proper implementation of
auctions will result in higher passenger volumes,
higher load factors, reallocation of flights to off-peak
times or to less congested airports, and lower fares on
average. Ball, Donohue and Hoffman [17] put forward
the need for three types of market mechanisms: an
auction of long-term leases of arrival and/or departure
from being scheduled in peak periods. In addition, slots, a secondary market that supports inter-airline
increasing per-flight cost is expected to encourage exchange of long-term leases, and a near-real-time
airlines to up-gauge (substitute larger-capacity planes market that allows for the exchange of slots on a par-
for some flights), and therefore increase passenger ticular day of operation.
throughput.
Congestion pricing as applied to runway allocation
would result in the price of an arrival or departure time
The Strategic Game
slot varying by time of day and day of week, and the
prices would dynamically change as the demand for The first simulation was held at GMU in November
operations changes over time. Congestion pricing of of 2004. It was principally focused on evaluating and
transport networks has been common in road traffic. comparing administrative measures and congestion
Examples include traditional tolling as well as more pricing. There were six major game players consisting
dynamic electronic-charges to users such as those used of teams from four airlines, the federal government
in London [4], in Trondheim, Norway [5], Singapore and the Port Authority of New York and New Jersey
[6], Toronto’s Highway 407, and SR 91 and I-15 HOT (PANYNJ), which operates LGA. Other participants
lanes in California. The airlines would find congestion included representatives of other airlines and airports,
prices that were set at day-of-operations (as in road the Air Transport Association, and various experts
pricing) difficult to manage, since their schedules are from academia, industry and government. The game
announced 90 days in advance. Thus, unlike individ- projected the participants to a hypothetical setting in
ual drivers, airlines should not be encouraged to cancel November 2007. The baseline scenario was an LGA
flights at the last minute due to high arrival costs. For schedule involving approximately 1,400 total daily
this reason, we considered a congestion management operations (arrivals and departures), a number that
approach whereby the prices are announced 120 days exceeds recommended operational levels. The airline
in advance, and the airlines base their schedules on teams adjusted their schedules in response to vari-
these announced prices (see Daniel [7], Pels [8], Fan ous government policies put in place. These policies
[9], and Schank [10], and Berardino[11] for more on involved federal regulations, administrative restric-
runway congestion pricing). tions, and congestion-based fees (substituting for cur-
An alternative to congestion pricing is the alloca- rent weight-based landing fees). For each alternative
tion of slots by auction for a much longer period of presented in these exercises, the resulting aggregated
time. The buyer has, in essence, leased the right to schedule was fed to two independently-developed
a given takeoff or landing and can use or re-sell that simulation models to calculate the levels of delay and
right for any portion of the lease period. The airlines cancellations that would have resulted from an attempt
bid for the right to land and/or depart at a given time. to operate that schedule (see Lovell, et al. 2003 and
Proposals to allocate airport time slots using market- Donohue and Le, 2004).
35 $1,400
30 Difference congestion Price
$1,200
Difference in Operations
25
20 $1,000
Congestion Price
15
10 $800
5 $600
0
-5 $400
-10
$200
-15
-20 $0
515 715 915 1115 1315 1515 1715 1915 2115 2315
seats. Figure 1 shows how the schedule was modi- times; the airlines provide new schedules, and the
fied from an administratively dictated schedule (e.g. process continues until the capacities and schedules
a schedule very similar to the LGA summer of 2004 are in balance. This process of determining the con-
schedule). The left axis shows the difference in sched- gestion prices before schedules are announced to the
uled flights in 15 minute aggregated time bins. The public results in a pricing approach that is, in essence,
right axis shows the operational price for each period, a short-term ascending auction for rights to announce
set to encourage efficient use of the landing opportuni-
schedules at LGA.
ties and to reduce congestion back to 2005 levels. One
More generally the results of this strategic game
thing the game highlighted is that the schedules were
support economic arguments that market-based alloca-
very sensitive to the times when prices either increased
tion mechanisms, e.g. congestion pricing or slot auc-
or decreased substantially, i.e. the airlines concen-
trated flights just prior to price increases or following tions, are likely to lead to better use of the scarce airport
price decreases. Further pricing changes could miti- resources than the present administrative measures.
gate these steep step increases. We note that the administrative actions could
Figure 2 shows the effect of congestion pricing on have led to more significant reduction in delays had
aggregate airline gauge choice throughout the 24-hour the administration been willing to set the capacity at a
schedule. Overall gauge is increased significantly at lower level. Thus, one important component to man-
almost all times of the day. Figures 3 and 4 show how aging delay is the determination of a proper capacity
the schedule was modified by flight distance, aircraft limit. The planned capacity (operations rate) is the
gauge and schedule frequency. Notice that the conges- most influential control available to determine the
tion pricing options produced a complex response by level of delays contributed to the NAS by each airport,
both flight distance and aircraft gauge. and the game revealed that there is little policy or
We note that the game invented a hypothetical consensus providing guidance for trading off delay/
Pricing Board with authority to set prices dynamically unpredictability against unused capacity. The level
based on schedules submitted (in principle) every 90 of schedule predictability is a major public policy
to 120 days. The process would work as follows: The issue. Recent over-scheduling at ORD (2005) and JFK
airlines submit schedules based on initial announced (2007) have demonstrated that significant delays at
prices. The Board evaluates the schedules provided, one airport propagate throughout the network. Thus,
and returns prices to reduce demand in oversubscribed mis-specification of capacity can lead to significant
180
Average Gauge(seats/opera tion)
160
The Second Game
140
120 A second strategic exercise took place on
100 February 24-25, 2005 at The University of
80 Maryland. At the end of the first game, the
60 industry indicated that they did not under-
40 Admin. Meas. Round 1 stand auctions and believed them to be too
20 Conjestion Price Round 2 complex. The exercise in February had the
0
industry use combinatorial clock auction soft-
45 300 515 730 945 1200 1415 1630 1845 2100 2315
ware where only price and aggregate demand
Time (45 min. rolling Avg.)
information was provided to the industry in
each round of the auction. Given prices based
Figure 3: Percent Change in Number of Operations on time of day, the industry was asked to pro-
by Flight Distance
vide schedules. The industry learned that the
30.0%
auction was not dissimilar from the conges-
20.0%
tion pricing exercise.
10.0%
Since the purpose of the game was to
0.0%
illustrate how auctions would work, we only
-10.0%
had the industry participate in a few rounds.
-20.0%
Therefore, we do not provide any results from
-30.0%
this game other than to say that the auction
-40.0% <500
[500,1000) resulted in frequencies and up-gauging simi-
-50.0%
[1000,1500) lar to those seen in the first game.
-60.0%
>=1500
-70.0%
baseline PbR Admin1 Admin2 cP1 cP2
Response Conclusions of the
Figure 4: Percent Change in Number of Seats by Flight Distance Strategic Exercises
30.0% We are not suggesting that the charts
20.0% and tables provided reflect the final prices or
10.0% schedules that would occur if these policies
0.0%
were implemented. In each case, only a few
-10.0%
rounds were employed. And, as prices got
higher, the airlines indicated that they needed
-20.0%
more time and their sophisticated scheduling
-30.0%
<500 technology to determine their next moves. In
-40.0% [500,1000)
addition, they worried that their responses
[1000,1500)
-50.0%
>=1500
might provide strategic decisions that they
-60.0% were not at liberty to reveal.
baseline PbR Admin1 Admin2 cP1 cP2
Response One result that we do believe is true: the
200
FAA based on the capacity of the
runway(s) when instrument landings
are required. A second question was: 150
LaGuardia will always be a popular airport with [6] C. Keong, “Road pricing : Singapore’s experi-
limited capacity. LGA, however, is not the only airport ence,” in 3rd Seminar of the IMPRINTEUROPE The-
facing congestion caused by scheduling that exceeds matic Network: “Implementing Reform on Transport
runway capacity. There are at least 10 U.S. airports Pricing: Constraints and solutions: learning from best
with current schedules greater than their runway practice”, Brussels, October 23-24, 2001.
capacity, and the number is likely to grow given the [7] J. Daniel, “Peak-load-congestion pricing and
costs, long lead-time, and politics involved in airport optimal capacity of large hub airports: With applica-
expansions. tion to the Minneapolis St.. Paul airport,” Ph.D. disser-
To overcome current and future delays, one must tation, University of Minnesota,1992.
address two issues: (a) How should one set the capac-
[8] E. Pels and E. Verhoef, “The economics of
ity restriction? (b) How should one allocate that
airport congestion pricing,” Tinbergen Institute Dis-
capacity? We believe that DOT should consider policy
cussion Paper No. 03-083/3, Amsterdam, The Nether-
decisions that treat all congested airports uniformly. A
lands, Tech. Rep., October 10, 2003.
workable solution to this problem is to set capacity at
each airport to its IMC rate and to use market-clearing [9] T. P. Fan and A. R. Odoni, “The potential of
mechanisms (congestion pricing or auctions) to allo- demand management as a short-term means of reliev-
cate the capacity. By so doing, one would provide pas- ing airport congestion,” in Proceedings of EURO-
sengers with predictable travel, reduce airline fuel and CONTROL-FAA Air Traffic Management R&D Review
repositioning costs, improve the overall safety of the Seminar, Santa Fe, NM,, 2001.
airspace, and improve U.S. economic productivity. [10] J. Schank, “Solving airside airport congestion: