ArticlePDF Available

An estimation of returns to scale of airport airsides under multiple optimal solutions in DEA

Authors:

Abstract and Figures

The returns to scale (RTS) nature of 37 Chinese airport airsides are investigated in this paper. Multiple optimal solutions in DEA (Data Envelopment Analysis) models may lead to error RTS estimation. To address this problem, we use the (Zhu and Shen, 1995) RTS method. The empirical study shows that all those airsides with two runways operate under decreasing RTS and those airsides with only one runway either operate in the area of increasing RTS or in the area of constant RTS.
Content may be subject to copyright.
An estimation of returns to scale of airport airsides under multiple
optimal solutions in DEA
Baocheng Zhang
*
, Hao Wu, Xinsheng Yang, Wenpeng Zhai, Qingjun Xia, Yafei Li
School of Air Trafc Management, Civil Aviation University of China, Tianjin 300300, PR China
article info
Article history:
Available online
Keywords:
Airport airside
Data Envelopment Analysis (DEA)
Returns to scale (RTS)
Runway declared distance
Multiple optimal solutions
abstract
The returns to scale (RTS) nature of 37 Chinese airport airsides are investigated in this paper. Multiple
optimal solutions in DEA (Data Envelopment Analysis) models may lead to error RTS estimation. To
address this problem, we use the (Zhu and Shen, 1995) RTS method. The empirical study shows that all
those airsides with two runways operate under decreasing RTS and those airsides with only one runway
either operate in the area of increasing RTS or in the area of constant RTS.
©2014 Elsevier Ltd. All rights reserved.
1. Introduction
Rapid economic growth has signicantly increased the demands
for air services in China in the past few decades; the number of
aircraft movements grew at an average rate of 13.8% per annum
between 2000 and 20011 (Wang, 2012). This increasing air demand
has placed enormous pressure on Chinese airport infrastructure,
especially on the airside area.
As a result, in addition to the infrastructure investment, there is
an urgent need to improve the efciency of Chinese airports in
order to relieve the pressure. Airport operations can be separated
into airside and landside operations. The former are concerned
with aircraft operations and include runways, taxiways and aprons,
and any other activities required to prepare the aircraft for ight.
The latter, comprising the airport terminal access roads, parking
areas and passenger terminals, are an essential element of the
airport landside infrastructure. These two components of the
airport, although conditionally related, have different production
technologies and certainly different strategies available to them.
Accordingly, airport congestion can be separated into airside
and landside congestions. Air Trafc congestions include airport
airside congestion, terminal (approach) congestion and en-route
congestion. Therefore, airport airside congestion is one of air
trafc congestions while airport landside congestion is not. Airside
congestion is a major cause for the large delays that currently affect
the ATM (Air Trafc Management) system. Solutions for airside
congestion include airside efciency improvement and airside
capacity enhancement. Solutions for airside capacity enhancement
have extensively been discussed (Le et al., 2008; Janic, 2004;
Fernandes and Pacheco, 2002; Zhang and Zhang, 2003, 2006,
2010; Madas and Zografos, 2008). The focus of the paper is on
airside efciency improvement.
Airside technical efciency, landside technical efciency, and
overall (both airside and landside) technical efciency are evalu-
ated separately in Zhang et al. (2012), Gillen and Lall (1997), Yu
(2010), Curi et al. (2011), Scotti et al. (2012). Also, there are some
studies on returns to scale (RTS) of airport airside. RTS is important
to airport airside operation as well as technical efciency.
To the best of our knowledge, airside RTS, landside RTS and
overall (both airside and landside) RTS are assessed separately in
Pels et al. (2003) while other studies do not (Martín and Roman,
2001; Martín et al., 2011; Perelman and Serebrisky, 2012; Lozano
and Guti
errez, 2011a; Ablanedo-Rosas and Gemoets, 2010; Chow
and Fung, 2009).
Most of the previous studies concerning airport airsides focused
on the technical efciency evaluation while less attention has been
paid to the RTS of airport airside. Furthermore, none of the existing
studies on Chinese airport airsides, including Fan et al. (2014),Fung
et al. (2008), Cui et al. (2013), Chi-Lok and Zhang (2009), Chow and
Fung (2009, 2012), Chang et al. (2013), and Huber (2010), therefore,
considers the RTS of the airsides in the use of their physical infra-
structure, which is the main aim of this study.
2. Airport airside activities
Air trafc control (ATC) is a service provided by ground-based
controllers who direct aircraft on airport airside and through
*Corresponding author.
E-mail address: bczhang@cauc.edu.cn (B. Zhang).
Contents lists available at ScienceDirect
Journal of Air Transport Management
journal homepage: www.elsevier.com/locate/jairtraman
http://dx.doi.org/10.1016/j.jairtraman.2014.07.001
0969-6997/©2014 Elsevier Ltd. All rights reserved.
Journal of Air Transport Management 40 (2014) 149e156
controlled airspace, and can provide advisory services to aircraft in
non-controlled airspace. The primary purpose of ATC worldwide is to
prevent collisions, organize and expedite the ow of trafc, and
provide information and other support for pilots. To prevent colli-
sions, ATCenforces trafc separation rules, which ensure each aircraft
maintains a minimum amount of empty space around it at all times.
To organize and expedite the ow of trafc, ATC takes strategic and
tactical measurers to mitigate delays and congestion. Airport control,
Terminal (approach) control and En-route control (Area control) are
the main three parts of air trafc control (ICAO, 2006b).
Airport control.
The primary method of controlling the immediate airport air-
side environment is visual observation from the aerodrome control
tower (TWR). Airport or Tower controllers are responsible for the
separation and efcient movement of aircraft operating on the
parking position, taxiways and runways of the airport itself, and
aircraft in the air near the airport, generally 5 to 10 nautical miles
(9e18 km) depending on the airport procedures.
Surveillance displays are also available to controllers at larger
airports to assist with controlling air trafc. Controllers may use a
radar system called Secondary Surveillance Radar for airborne
trafc approaching and departing. These displays include a map of
the area, the position of various aircraft, and data tags that include
aircraft identication, speed, altitude, and other information
described in local procedures.
Terminal (approach) control.
Many airports have a radar control facility that is associated with
the airport. In most countries, this is referred to as Terminal or
Approach Control; in the U.S., it is referred to as a TRACON (Ter-
minal Radar Approach Control). While every airport varies, termi-
nal controllers usually handle trafc in a 30-to-50-nautical-mile
(56e93 km) radius from the airport. Where there are many busy
airports close together, one consolidated Terminal Control Center
may service all the airports. The airspace boundaries and altitudes
assigned to a Terminal Control Center, which vary widely from
airport to airport, are based on factors such as trafcows,
neighboring airports and terrain.
A large and complex example is the London Terminal Control
Centre which controls trafc for ve main London airports up to
20,000 feet (6100 m) and out to 100 nautical miles (190 km). Ter-
minal controllers are responsible for providing all ATC services
within their airspace. Trafcow is broadly divided into de-
partures, arrivals, and overights. As aircraft move in and out of the
terminal airspace, they are handed off to the next appropriate
control facility (a control tower, an en-route control facility, or a
bordering terminal or approach control). Terminal control is
responsible for ensuring that aircraft are at an appropriate altitude
when they are handed off, and that aircraft arrive at a suitable rate
for landing.
En-route control.
En-route air trafc controllers issue clearances and instructions
for airborne aircraft, and pilots are required to comply with these
instructions. En-route controllers also provide air trafc control
services to many smaller airports around the country, including
clearance off of the ground and clearance for approach to an airport.
En-route controllers adhere to a set of separation standards that
dene the minimum distance allowed between aircraft. These
distances vary depending on the equipment and procedures used in
providing ATC services.
It should be noted that Airport or Tower controllers are
responsible for aircraft movement in airside. The airside activities
are of Airport control domain and the landside ones are not. Many
factors affect airside activities, which include:
runway
temperature
wind
presence of obstructions in the vicinity of the airport
locally imposed noise abatement restrictions
conguration of Terminal
radar ability.
3. Methodology issues of multiple optimal solutions
3.1. Multiple optimal solutions
When estimating RTS by the CCR (Charnes et al., 1978) RTS
method or BCC (Banker et al., 1984) RTS in DEA (Banker and Thrall,
1992), there exist multiple optimal solutions which lead to error
RTS estimation (Banker et al., 1996; Zhang, 2008). Specically,
given the existence of an optimal solution with Pl
*
s1 in the
DEAeCCR envelopment model (u
0
s0 in the DEAeBCC multiplier
model), we cannot reach the conclusion that there is no existence
of an optimal solution with Pl
¼1ðu
0
¼0Þbecause of multiple
optimal DEA solutions. In real world applications, the examination
of alternative optima is a laborious task. According to Seiford and
Zhu (1999), there are two methods to avoid multiple solutions
effect. They are the (Zhu and Shen, 1995) method and the (F
are
et al., 1994) method.
3.2. The (Zhu and Shen, 1995) methods
Suppose, that we have nDecision Making Units (DMUs) where
each j,j¼1,2,,n, produces the same soutputs in (if possibly)
different amounts, y
rj
(r¼1,2,,s), using the same minputs, x
ij
(i¼1,2,,m), also in (if possibly) different amounts. The technical
efciency of a specic DMU
0
can be evaluated by the output-ori-
ented BCC modelof DEA in envelopment formas follows.
max f
BCC
þε X
m
i¼1
s
i
þX
s
r¼1
s
þ
r
!
s:t:
8
>
>
>
>
>
>
>
>
>
>
<
>
>
>
>
>
>
>
>
>
>
:
P
n
j¼1
x
ij
l
j
þs
i
¼x
i0
;i¼1;2; :::; m;
P
n
j¼1
y
rj
l
j
s
þ
r
¼f
BCC
y
r0
;r¼1;2; :::; s;
P
n
j¼1
l
j
¼1;
l
j
;s
i
;s
þ
r
0ci;r;j:
(1)
where εis a non-Archimedean innite small constant, which en-
sures that no input or output is allocated zero weight;s
i
and s
þ
r
are
the slacks of the inputs and outputs, respectively.
f
BCC
is the scalar
variable that represents the possible radial increase to be applied to
all outputs so as to obtain the projected output values.
l
j
(j¼1,2,...,n) is the variable whose optimal values form a linear
combination of units that make up the performance of the airside
under analysis, and establishes a direction in which to identify the
sources of inefciency at the airside.
Since the inputs of airport airside represent existing facilities, it
does not make much sense to reduce them. Therefore, an output
orientation DEA model is more appropriate to be used than the
input-orientated one.
B. Zhang et al. / Journal of Air Transport Management 40 (2014) 149e156150
Denition 1. A DMU is technically efcient if and only if both (i)
f
BCC
¼1; (ii) s
i
¼s
þ
r
¼0;ci;r.
RTS generally has an unambiguous meaning only if DMU is
technically efcient since it is only in this state that a tradeoff be-
tween inputs and outputs is required to improve one or the other of
these elements (Banker et al., 2004). If a DMU is not technically
efcient, we can use optimal values in Eq (1) to project this DMU
onto the technical efciency frontier via the following formulas.
b
X
0
;b
Y
0
¼8
>
>
>
<
>
>
>
:
P
n
j¼1
x
ij
l
j
¼x
i0
s

i
;i¼1;2; :::; m;
P
n
j¼1
y
rj
l
j
¼f
BCC
y
r0
þs
þ
r
;r¼1;2; :::; s:
(2)
Turn to the CCR model which takes the following envelopment
form.
max f
CCR
þε X
m
i¼1
s
i
þX
s
r¼1
s
þ
r
!
s:t:
8
>
>
>
>
>
<
>
>
>
>
>
:
P
n
j¼1
x
ij
l
j
þs
i
¼x
i0
;i¼1;2; :::; m;
P
n
j¼1
y
rj
l
j
s
þ
r
¼f
CCR
y
r0
;r¼1;2; :::; s;
l
j
;s
i
;s
þ
r
0ci;r;j:
(3)
Theorem 1.(Zhu and Shen, 1995) (i) f
BCC
¼f
CCR
if and only if
constant RTS prevail on DMU
0
; otherwise, if f
BCC
sf
CCR
, then (ii)
P
n
j¼1
l
j
<1 if and only if increasing RTS prevail on DMU
0
; (iii)
P
n
j¼1
l
j
>1 if and only if decreasing RTS prevail on DMU
0
.
Thus, in empirical applications, we can explore RTS in two steps.
First, select all the airsides that have the same CCR and BCC ef-
ciency scores regardless of the value of P
n
j¼1
l
j
. These airsides show
constant RTS. Next, use the value of P
n
j¼1
l
j
to determine the RTS
for the remaining airsides. We observe that in this process we can
safely ignore possible multiple optimal solutions of l
j
.
3.3. The (F
are et al., 1994) method
If we impose P
n
j¼1
l
j
1 in the CCR model (Eq (3)), then we
obtain the following DEA model
max f
f
þε X
m
i¼1
s
i
þX
s
r¼1
s
þ
r
!
s:t:
8
>
>
>
>
>
>
>
>
>
>
<
>
>
>
>
>
>
>
>
>
>
:
P
n
j¼1
x
ij
l
j
þs
i
¼x
i0
;i¼1;2; :::; m;
P
n
j¼1
y
rj
l
j
s
þ
r
¼f
f
y
r0
;r¼1;2; :::; s;
P
n
j¼1
l
j
1;
l
j
;s
i
;s
þ
r
0ci;r;j:
(4)
Obviously, f
BCC
f
f
f
CCR
.
Theorem 2.(F
are et al., 1994) (i) f
BCC
¼f
CCR
if and only if con-
stant RTS prevail on DMU
0
; otherwise, if f
BCC
sf
CCR
, then
(ii)f
BCC
<f
f
if and only if increasing RTS prevail on DMU
0
; (iii)
f
BCC
¼f
f
if and only if decreasing RTS prevail on DMU
0
.
3.4. Comparison of the (Zhu and Shen, 1995) method and the (F
are
et al., 1994) method
The above two RTS methods, in fact, are equivalent but different
presentations. The proof of the equivalence of the above two RTS
methods can be found in Seiford and Zhu (1999) or Banker et al.
(2004). In terms of computational efforts, the RTS method pro-
posed by Zhu and Shen (1995) is relatively easy to apply. Therefore,
the (Zhu and Shen, 1995) method is adopted in this paper while the
(F
are et al., 1994) method is adopted in Lozano and Guti
errez
(2011a).
4. The inputeoutput selection and the data
To measure the overall operational performance of an airport
airside, we must rst identify the outputs that an airport airside
produces and the inputs that it uses in producing these outputs.
The selected airports have common airside service as well as
common inputs. On the input side, our specication includes
three quasi-xed inputs, namely, take-off distance available (in
meter), landing distance available (in meter) and aircraft parking
position.
The introduction of stopways and clearways and the use of
displaced thresholds on runways have created a need for accurate
information regarding the various physical distances available and
suitable for the landing and take-off of airplanes.
4.1. Clearway
In aviation, clearway is a term related to the dimension of some
runways and it is abbreviated with CWY. Clearway is an area
beyond the paved runway, free of obstructions and under the
control of the airport authorities. The length of the clearway may be
included in the length of the takeoff distance available. For
example, if a paved runway is 2000 m long and there are 400 m of
clearway beyond the end of the runway, the takeoff distance
available is 2400 m long. The clearway extends from the end of the
runway with an upward slope not exceeding 1.25 percent, above
which no object nor any terrain protrudes. Clearways are not
physical structures as shown in Fig. 1; they are simply an area of
dened width and length which are free of obstacles (Fig. 2). An
aircraft may make a portion of its initial climb to a specied height
over clearway. When the runway is to be used for takeoff of a large
airplane, the maximum permissible takeoff weight of the airplane
can be based on the takeoff distance available, including clearway.
Clearway allows large airplanes to takeoff at a heavier weight than
would be allowed if only the length of the paved runway is taken
into account. Thus, a clearway increases the allowable airplane
operating takeoff weight without increasing runway length.
4.2. Stopway
Stopways are an area beyond the take-off runway, no less wide
than the runway and centered upon the extended centerline of the
runway, able to support the airplane during an abortive take-off,
without causing structural damage to the airplane, and desig-
nated by the airport authorities for use in decelerating the airplane
during an abortive take-off. Stopways are physical structures (Fig.1)
and are usually paved. However, stopways are not as strong as the
main length of runway and therefore are only used to help bring an
airplane to a stop in the event of an abandoned take-off. Stopways
are identied by large yellow chevrons on either end of the main
runway (Fig. 2).
4.3. Displaced thresholds
A displaced threshold is a runway threshold located at a point
other than the physical beginning or end of the runway (Fig. 2).
There are at least a couple of reasons why a runway threshold
would be displaced:
B. Zhang et al. / Journal of Air Transport Management 40 (2014) 149e156 151
1. Most often the offset threshold is in place to give arriving aircraft
clearance over an obstruction while still allowing departing
aircraft the maximum amount of runway available.
2. A displaced threshold may also be introduced if a beginning
section (the touchdown zone) of the runway is no longer able to
sustain the continuous impact from landing aircraft. In such a
case, aircraft are expected to land beyond the displaced
threshold.
3. Increases obstacle clearance reduces the noise footprint below
any approaching aircraft. A comparison between normal
threshold approach and displaced threshold approach is illus-
trated in Fig. 3.
Departing aircraft are permitted to use the displaced section of
the runway for takeoffs or landing rollouts because those aircraft
are not impacting the runway with the force of a landing aircraft.
Thresholds are counted as part of the runway, and are included in
the runway size. When viewing a runway's size with displaced
thresholds, one must nd out how long the displaced thresholds
are in order to calculate the available landing distance. The portion
of the runway so displaced may be used for takeoff but not for
landing. Landing aircraft may use the displaced area on the oppo-
site end for roll out.
4.4. Runway declared distances
For these purposes, the term declared distancesis used with
the following four distances associated with a particular runway.
The following declared distances shall be calculated to the nearest
meter for a runway intended for use by international commercial
air transport:
1. Take-off Run Available (TORA). The length of runway declared
available and suitable for the ground run of an airplane taking
off. The take-off run available is the distance from the point on
the surface of the aerodrome at which the airplane can
commence its take-off run to the nearest point in the direction
of takeoff at which the surface of the aerodrome is incapable of
bearing the weight of the airplane under normal operating
conditions. At most aerodromes the take-off run available is the
length of the runway from threshold to threshold.
2. Take-off Distance Available (TODA). The length of the takeoff run
available plus the length of the clearway (CWY), if clearway is
provided. The take-off distance available must be compared to
the airplane's actual take-off distance. The requirements for
take-off state that the airplane must be able to complete the
take-off within the take-off distance available. Although clear-
ways can be of any length, there is a limit to the amount of
clearway that can be used when calculating the TODA. The
maximum length of clearway in this case cannot be more than
half the length of TORA.
3. AccelerateeStop Distance Available (ASDA) or emergency dis-
tance available. The length of the takeoff run available plus the
length of the stopway (SWY), if stopway is provided. The
accelerateestop distance available must be compared to the
airplanes actual accelerateestop distance. The requirements for
take-off state that the airplanes accelerateestop distance must
not exceed the accelerateestop distance available.
4. Landing Distance Available (LDA). The length of runway which is
declared available and suitable for the ground run of an airplane
landing.
Fig. 1. Illustration of clearway and stopway
Fig. 2. Illustration of runway declared distances. Fig. 3. Displaced threshold approach.
B. Zhang et al. / Journal of Air Transport Management 40 (2014) 149e156152
Some principles concerning the calculation of the above
declared distances are illustrated as follows (ICAO, 2006a).
Where a runway is not provided with a stopway or clearway and
the threshold is located at the extremity of the runway, the four
declared distances should normally be equal to the length of the
runway.
Where a runway is provided with a clearway (CWY), then the
TODA will include the length of clearway, as shown in Fig. 2.
Where a runway is provided with a stopway (SWY), then the
ASDA will include the length of stopway, as shown in Fig. 2.
Where a runway has a displaced threshold, then the LDA will be
reduced by the distance the threshold is displaced, as shown in
Fig. 2. A displaced threshold affects only the LDA for approaches
made to that threshold; all declared distances for operations in
the reciprocal direction are unaffected.
Where more than one of these features exist, then more than
one of the declared distances will be modied ebut the
modication will follow the same principle illustrated. An
example showing a situation where all these features exist is
shown in Fig. 2.
In addition, a runway is a different runway from its reciprocal
runway even though they have the same declared distance. The
slope and obstructions in the departure area may be different. Also,
the approaching capability may be quite different.
4.5. Inputeoutput selection
Clearway is important to an aircraft's take-off and displaced
threshold is essential to an aircraft's landing. Hence, TODA and LDA
illustrate the operation status of the airside of an airport better than
distance of runway. Similarly, aircraft parking position illustrates
the operation status of the airside of an airport better than apron
area because the number of parking position is crucial to aircraft
parking. These inputs cannot be changed for two reasons. Firstly,
the three inputs are xed in capacity. Secondly, a minimum TODA is
required for an aircraft to takeoff for safety reasons; similarly, a
minimum LDA is required. Thus, the output-oriented DEAeBCC
model is more appropriate to be used than the input-oriented
model that requires a proportional reduction in all input usage.
The most commonly used output measures for airports are the
number of passengers, the volume of air cargo, and the number of
aircraft movements. The number of passengers andtons of cargo are,
however, landside outputs. Thus, the only genuine airside output is
the number of aircraft movements; it is measured in terms of the
number of movements for aircraft landing and taking-off per year.
In addition, some inputs including radar capacity and ngers
(passenger boarding bridge) are not considered in our study due to
data availability. The control of aircraft on the ground is typically
provided visually and audibly between the aircraft control tower,
pilots and ground staff on the taxiway, in bad weather such as fog,
heavy rain and even snow, visual control of the distance between
aircraft and other vehicles on the ground is seriously hampered and
can close an airport. Use of a SharpEye radar has been proven to
provide extremely clear and sharp target detection with the ability
to remove clutter at an airport. Such an approach to ground man-
agement enables radar to be utilized for aircraft control and sup-
port vehicles on the ground as well as in the air, improving safety
signicantly and increasing airport operating periods.
4.6. The data
In our analysis, the data used in this paper (Table 1) were
collected for a homogeneous set of 37 Chinese airport airsides
during the year 2009. They provide service to both international
ights and domestic ights. Table 1 is the data set.
5. Empirical study
5.1. Illustration of runway declared distances
Table 2 is the data of four airport airsides in China. First, we
present Beijing airside's runway 01/19 in Fig. 4. TODA of runway 01
is 4300 m (runway length 3800 m plus clearway length 500 m)
while TODA of runway 19 of is 3800 m due to no clearway.
Second, we present Shanghai (Hongqiao) airside's runway 18L/
36R and runway 18R/36L in Fig. 5. Threshold and end of runway
18L/36R are both displaced 100 m inwards. Thus TODA and LDA are
both 3300 m though runway length is 3400 m. Threshold of runway
18R/36L is displaced 300 m inwards. Thus, LDA is 3000 m while
TODA is not affected.
Runways in Shanghai Pudong airside have no clearway or
stopway, or a displaced threshold; Thus, all declared distance are
equal to runway length.
Jiuzhaigou airport lies in a valley. On one side, Runway 20 can be
used for take-off while it cannot be used for landing because of
obstructions (mountain) around the airport. On the other side,
Runway 02 can be used for landing while it cannot be used for take-
off because of obstructions (mountain) around the airport. Unfor-
tunately, we cannot take Jiuzhaigou airport airside into efciency
evaluation because of data availability (lack of number of parking
Table 1
The data set.
Airport Take-off distance
available (m)
Landing distance
available (m)
Parking
position
Movements
Beijing 22500 21600 320 487918
Guangzhou 14800 14600 133 308863
Pudong 22400 22400 196 287916
Shenzhen 6800 6800 96 202627
Chengdu 14400 14400 90 190094
Hongqiao 6600 6400 75 189071
Kunming 6920 6800 48 172572
Xian 6000 6000 34 146272
Hangzhou 7200 7200 64 134058
Chongqing 6400 6400 47 132619
Wuhan 7200 6800 85 113332
Changsha 6400 6400 30 110023
Nanjing 7200 7200 36 106142
Xiamen 6450 6100 58 105939
Qingdao 6800 6800 45 98033
Dalian 6600 6300 25 85390
Zhengzhou 7200 6800 31 75743
Tianjin 14000 13200 214 75116
Haikou 7600 7200 33 69114
Shenyang 6800 6400 34 67027
Urumchi 7600 7200 71 65511
Jinan 7200 7200 44 63602
Sanya 6800 6800 42 59811
Harbin 6400 6400 24 57440
Guiyang 6840 6400 27 57354
Taiyuan 7500 7200 45 52236
Fuzhou 7200 7200 22 51575
Guilin 6800 6400 20 49525
Nanning 6400 6400 14 44597
Ningbo 6400 6250 16 37512
Changchun 6400 6400 19 35909
Hefei 6000 6000 19 35814
Huhhot 7800 7200 72 33190
Lanzhou 7200 7200 21 28353
Yantai 6400 6400 12 25830
Yinchuan 6400 6400 25 21393
Xichang 7700 7200 8 4134
B. Zhang et al. / Journal of Air Transport Management 40 (2014) 149e156 15 3
position and movements). Otherwise, we can illustrate the affec-
tion of the inclusion of LDA in efciency evaluation better.
5.2. RTS estimation
In this section, we apply the (Zhu and Shen, 1995) method to the
airside sectors of 37 Chinese airports. The f
BCC
,f
CCR
,and P
n
j¼1
l
j
of
37 Chinese airport airsides are calculated using DEA Excel Solver
(Zhu, 2003) and the results are presented in Table 3. A number of
points emerge from the present study.
In the rst step, f
BCC
¼f
CCR
is found in Shenzhen, Shanghai
(Hongqiao), Xian, and Shenyang, respectively. Thus constant RTS
prevail on these four airsides. Dimension makes a difference. Those
with constant RTS have the adequate dimension. All the 4 airsides
only have one runway.
For the rest airsides, P
n
j¼1
l
j
<1 is found in Huhhot, Changsha,
Xiamen, Dalian, Zhengzhou, Haikou, Harbin, Guiyang, Fuzhou,
Guilin, Nanning, Ningbo, Changchun, Hefei, Lanzhou, Yantai, Yin-
chuan, and Xichang, respectively. Thus increasing RTS prevail on
these 17 airsides. These airsides are still in the increasing RTS zone
of the production function. All the 17 airsides only have one
runway. Those airsides are too small in dimension and scale
dimension should be increased.
Similarly, P
n
j¼1
l
j
>1 is found in Beijing, Guangzhou, Shanghai
(Pudong), Chengdu, Kunming, Hangzhou, Chongqing, Wuhan,
Nanjing, Qingdao, Tianjin, Urumchi, Jinan, Sanya, and Taiyuan,
respectively. Thus decreasing RTS prevail on these 16 airsides.
These airsides are still in the decreasing RTS zone of the production
function. Those airsides are too large in dimension and scale
dimension should be decreased. All those airsides with two run-
ways operate under decreasing RTS.
According to Table 1, Beijing, Guangzhou, and Shanghai
(Pudong) rank the rst three in terms of the output (aircraft
movements). However, they all show decreasing RTS.
In addition, if only TODA is included in the analysis, both the BCC
efciency scores and CCR efciency score are different in some
Table 2
Runway details of 4 Chinese airport airsides.
Airside Runway
designator
Runway
length
TORA TODA ASDA LDA
Beijing 01 3800 3800 4300 3920 3800
19 3800 3800 3800 3920 3800
18L/36R 3800 3800 4000 3860 3800
18R/36L 3200 3200 3200 3260 3200
Guangzhou 02L/20R 3600 3600 3600 3600 3600
02R 3800 3800 3800 3800 3800
20L 3800 3800 3800 3800 3600
Shanghai
(Hongqiao)
18L/36R 3400 3300 3300 3300 3200
18R/36L 3300 3300 3300 3300 3000
Shanghai
(Pudong)
17R/35L 3400 3400 3400 3400 3400
17L/35R 4000 4000 4000 4000 4000
16/34 3800 3800 3800 3800 3800
Jiuzhaigou 20 3200 3200 3200 3300 NU
02 3200 NU NU NU 3000
Fig. 4. Beijing airside: TODA and clearway.
Fig. 5. Shanghai Hongqiao airside: LDA and displaced threshold.
Table 3
The RTS report.
No. Airside f
BCC
f
CCR
P
n
j¼1
l
j
RTS
1 Beijing 1 1.319 3.176 Decreasing
2 Guangzhou 1 1.287 2.336 Decreasing
3 Pudong 1.282 2.076 3.550 Decreasing
4 Shenzhen 1 1 1 Constant
5 Chengdu 1.262 1.880 2.378 Decreasing
6 Hongqiao 1 1 1 Constant
7 Kunming 1 1.007 1.117 Decreasing
8 Xian 1 1 1 Constant
9 Hangzhou 1.393 1.439 1.138 Decreasing
10 Chongqing 1.231 1.237 1.038 Decreasing
11 Wuhan 1.750 1.778 1.042 Decreasing
12 Changsha 1.145 1.173 0.882 Increasing
13 Nanjing 1.413 1.459 1.059 Decreasing
14 Xiamen 1.482 1.592 0.976 Increasing
15 Qingdao 1.703 1.740 1.116 Decreasing
16 Dalian 1.177 1.260 0.735 Increasing
17 Zhengzhou 1.730 1.761 0.912 Increasing
18 Tianjin 4.340 5.236 1.941 Decreasing
19 Haikou 2.043 2.054 0.971 Increasing
20 Shenyang 2.182 2.182 1 Constant
21 Urumchi 2.954 3.073 1.148 Decreasing
22 Jinan 2.595 2.797 1.191 Decreasing
23 Sanya 2.697 2.815 1.124 Decreasing
24 Harbin 1.661 1.798 0.706 Increasing
25 Guiyang 1.930 2.025 0.794 Increasing
26 Taiyuan 3.196 3.429 1.193 Decreasing
27 Fuzhou 1.653 1.835 0.647 Increasing
28 Guilin 1.516 1.737 0.588 Increasing
29 Nanning 1 1.351 0.412 Increasing
30 Ningbo 1.126 1.835 0.471 Increasing
31 Changchun 1.950 2.276 0.559 Increasing
32 Hefei 1 2.282 0.559 Increasing
33 Huhhot 5.853 6.091 1.146 Decreasing
34 Lanzhou 2.828 3.186 0.618 Increasing
35 Yantai 1 1.999 0.353 Increasing
36 Yinchuan 4.699 5.027 0.735 Increasing
37 Xichang 1 8.325 0.235 Increasing
B. Zhang et al. / Journal of Air Transport Management 40 (2014) 149e156154
airside, which are shown in Table 4. Wuhan (N0.11), Xiamen
(No.14), Urumchi (No.21), Ningbo (No.30), and Huhhot (No.33) have
different BCC efciency scores in the two different input selection.
Beijing (No.1), Kunming (No.7), Wuhan (N0.11), Xiamen (No.14),
Tianjin (No.18), Taiyuan (No.26), Urumchi (No.21), Ningbo (No.30),
and Huhhot (No.33) have different CCR efciency scores in the two
different input selection. Xiamen shows decreasing RTS if only
TODA is included in the analysis (Xiamen shows increasing RTS if
both TODA and LDA are included in the analysis).
6. Conclusions
This paper has investigated the RTS nature of 37 Chinese airport
airsides. Rather than runway length, we take TODA and LDA as the
inputs. In most cases these two inputs are not equal to runway
length and are more important for aircraft movements than runway
length. The empirical study indicates only 4 airsides operate in
constant RTS while 16 in decreasing RTS and 17 in increasing RTS.
To address multiple optimal solutions when estimating RTS in
DEA, we adopted the (Zhu and Shen, 1995) RTS method. The
possible effect of multiple optimal DEA solutions on the determi-
nation of RTS can be avoided.
In addition, we focus on airport airside performance in pro-
ducing aircraft movement. Airport performance in producing pas-
senger and cargo is not considered in this paper mainly due to data
availability. Also we do not consider congestion levels and aircraft
sizes which might bias performance signicantly. Several
researchers have also taken airport congestion into consideration in
airport benchmarking analysis by incorporating ight delays.
Considering ight delays may better reect the operational status
of airports (Lozano and Guti
errez, 2011b; Lozano et al., 2013; Fan
et al., 2014). Without considering ight delays, congested airports
might be found to be efcient (Pathomsiri et al., 2008). A possible
way to overcome these limitation could be to take total number of
available seats offered and ight delays as outputs into RTS esti-
mation. The inclusion of these outputs into the proposed model is
worth investigating in the future. But we cannot conduct such
research in this paper mainly due to data availability.
Acknowledgments
The authors thank two anonymous reviewers for their valuable
comments and suggestions. This work is supported by the United
Fund Foundation of the National Natural Science Foundation of
China and the Civil Aviation Grant (U1333116), the research pro-
jects of the social science and humanity on Young Fund of the
ministry of Education of China (14YJC630185).
References
Ablanedo-Rosas, J., Gemoets, L., 2010. Measuring the efciency of Mexican airports.
Journal of Air Transport Management 16 (6), 343e345.
Banker, R., Chang, H., Cooper, W., 1996. Equivalence and implementation of alter-
native methods for determining returns to scale in data envelopment analysis.
European Journal of Operational Research 89 (3), 473e481.
Banker, R., Charnes, A., Cooper, W., 1984. Some models for estimating technical and
scale inefciencies in data envelopment analysis. Management Science 30 (9),
1078e1092.
Banker, R., Cooper, W., Seiford, L., Thrall, R., Zhu, J., 2004. Returns to scale in
different DEA models. European Journal of Operational Research 154 (2),
345e362.
Banker, R.D., Thrall, R.M., 1992. Estimation of returns to scale using data envelop-
ment analysis. European Journal of Operational Research 62 (1), 74e84.
Chang, Y.-C., Yu, M.-M., Chen, P.-C., 2013. Evaluating the performance of Chinese
airports. Journal of Air Transport Management 31, 19e21.
Charnes, A., Cooper, W., Rhodes, E., 1978. Measuring the efciency of decision
making units. European Journal of Operational Research 2 (6), 429e444.
Chi-Lok, A., Zhang, A., 20 09. Effects of competition and policy changes on Chinese
airport productivity: an empirical investigation. Journal of Air Transport Man-
agement 15 (4), 166e174.
Chow, C., Fung, M., 2009. Efciencies and scope economies of Chinese airports in
moving passengers and cargo. Journal of Air Transport Management 15 (6),
324e329.
Chow, C.K.W., Fung, M.K.Y., 2012. Estimating indices of airport productivity in
greater china. Journal of Air Transport Management 24, 12e17.
Cui, Q., Kuang, H.-b., Wu, C.-y., Li, Y., 2013. Dynamic formation mechanism of airport
competitiveness: the case of china. Transportation Research Part A: Policy and
Practice 47, 10e18.
Curi, W., Gitto, S., Mancuso, P., 2011. New evidence on the efciency of Italian air-
ports: a bootstrapped DEA analysis. Socio-Economic Planning Sciences 45 (2),
84e93.
Fan, L., Wu, F., Zhou, P., 2014. Efciency measurement of Chinese airports with ight
delays by directional distance function. Journal of Air Transport Management
34, 140e145.
F
are, R., Grosskopf, S., Lovell, C., 1994. Production Frontiers. Cambridge University
Press.
Fernandes, E., Pacheco, R., 2002. Efcient use of airport capacity. Transportation
Research Part A: Policy and Practice 36 (3), 225e238.
Fung, M.K.Y., Wan, K.K.H., Hui, Y.V., Law, J.S., 2008. Productivity changes in Chinese
airports 1995e2004. Transportation Research Part E: Logistics and Trans-
portation Review 44 (3), 521e542.
Gillen, D., Lall, A., 1997. Developing measures of airport productivity and perfor-
mance: an application of data envelopment analysis. Transportation Research
Part E: Logistics and Transportation Review 33 (4), 261e273.
Huber, H., 2010. Planning for balanced growth in Chinese air trafc: a case for
statistical mechanics. Journal of Air Transport Management 16 (4), 178e184.
ICAO, 2006a. Aerodrome Design Manual (ICAO Doc 9157). Part 1-Runways. Inter-
national Civil Aviation Organization (ICAO), Montreal.
ICAO, 2006b. Air Trafc Service (ICAO ANNEX 11). International Civil Aviation Or-
ganization (ICAO), Montreal.
Janic, M., 2004. Expansion of airport capacity at London Heathrow airport. Trans-
portation Research Record: Journal of the Transportation Research Board 1888
(1), 7e14.
Table 4
Efciency score under different input selection.
No. Airside BCC efciency CCR efciency
TODA LDA TODA TODA LDA TODA
1 Beijing 1.0000 1.0000 1.3192 1.3741
2 Guangzhou 1.0000 1.0000 1.2874 1.2874
3 Pudong 1.2823 1.2823 2.0764 2.0764
4 Shenzhen 1.0000 1.0000 1.0000 1.0000
5 Chengdu 1.2621 1.2621 1.8798 1.8798
6 Hongqiao 1.0000 1.0000 1.0000 1.0000
7 Kunming 1.0000 1.0000 1.0074 1.0157
8 Xian 1.0000 1.0000 1.0000 1.0000
9 Hangzhou 1.3935 1.3935 1.4390 1.4390
10 Chongqing 1.2311 1.2311 1.2371 1.2371
11 Wuhan 1.7499 1.7732 1.7776 1.8317
12 Changsha 1.1446 1.1446 1.1731 1.1731
13 Nanjing 1.4135 1.4135 1.4591 1.4591
14 Xiamen 1.4817 1.6324 1.5925 1.6360
15 Qingdao 1.7029 1.7029 1.7404 1.7404
16 Dalian 1.1772 1.1772 1.2595 1.2595
17 Zhengzhou 1.7298 1.7298 1.7608 1.7608
18 Tianjin 4.3399 4.4393 5.2364 5.5537
19 Haikou 2.0428 2.0428 2.0541 2.0541
20 Shenyang 2.1823 2.1823 2.1823 2.1823
21 Urumchi 2.9541 2.9868 3.0727 3.1477
22 Jinan 2.5952 2.5952 2.7975 2.7975
23 Sanya 2.6968 2.6968 2.8151 2.8151
24 Harbin 1.6615 1.6615 1.7975 1.7975
25 Guiyang 1.9299 1.9299 2.0253 2.0253
26 Taiyuan 3.1958 3.1958 3.4289 3.5361
27 Fuzhou 1.6533 1.6533 1.8351 1.8351
28 Guilin 1.5164 1.5164 1.7374 1.7374
29 Nanning 1.0000 1.0000 1.3505 1.3505
30 Ningbo 1.1256 1.4599 1.8350 1.8350
31 Changchun 1.9498 1.9498 2.2763 2.2763
32 Hefei 1.0000 1.0000 2.2824 2.2824
33 Huhhot 5.8533 5.9744 6.0906 6.3568
34 Lanzhou 2.8280 2.8280 3.1864 3.1864
35 Yantai 1.0000 1.0000 1.9987 1.9987
36 Yinchuan 4.6987 4.6987 5.0275 5.0275
37 Xichang 1.0000 1.0000 8.3253 8.3253
B. Zhang et al. / Journal of Air Transport Management 40 (2014) 149e156 15 5
Le, L., Donohue, G., Hoffman, K., Chen, C.-H., 2008. Optimum airport capacity uti-
lization under congestion management: a case study of New York Laguardia
airport. Transportation Planning and Technology 31 (1), 93e112.
Lozano, S., Guti
errez, E., 2011a. Efciency analysis and target setting of Spanish
airports. Networks and Spatial Economics 11 (1), 139e157.
Lozano, S., Guti
errez, E., 2011b. Slacks-based measure of efciency of airports with
airplanes delays as undesirable outputs. Computers &Operations Research 38
(1), 131e139.
Lozano, S., Guti
errez, E., Moreno, P., 2013. Network DEA approach to airports per-
formance assessment considering undesirable outputs. Applied Mathematical
Modelling 37 (4), 1665e1676.
Madas, M.A., Zografos, K.G., 2008. Airport capacity vs. demand: mismatch or
mismanagement? Transportation Research Part A: Policy and Practice 42 (1),
203e226.
Martín, J.C., Roman, C., 2001. An application of DEA to measure the efciency of
Spanish airports prior to privatization. Journal of Air Transport Management 7
(3), 149e157.
Martín, J.C., Rom
an, C., Voltes-Dorta, A., 2011. Scale economies and marginal costs in
Spanish airports. Transportation Research Part E: Logistics and Transportation
Review 47 (2), 238e248.
Pathomsiri, S., Haghani, A., Dresner, M., Windle, R.J., 2008. Impact of undesirable
outputs on the productivity of us airports. Transportation Research Part E:
Logistics and Transportation Review 44 (2), 235e259.
Pels, E., Nijkamp, P., Rietveld, P., 2003. Inefciencies and scale economies of Euro-
pean airport operations. Transportation Research Part E: Logistics and Trans-
portation Review 39 (5), 341e361.
Perelman, S., Serebrisky, T., 2012. Measuring the technical efciency of airports in
Latin America. Utilities Policy 22 (1), 1e7.
Scotti, D., Malighetti, P., Martini, G., Volta, N., 2012. The impact of airport compe-
tition on technical efciency: a stochastic frontier analysis applied to Italian
airport. Journal of Air Transport Management 22 (1), 9e15.
Seiford, L., Zhu, J., 1999. An investigation of returns to scale in data envelopment
analysis. Omega 27 (1), 1e11.
Wang, Z., 2012. Statistical Data on Civil Aviation of China. Civil Aviation of China
Press.
Yu, M., 2010. Assessment of airport performance using the SBM-NDEA model.
Omega 38 (6), 440e452.
Zhang, A., Zhang, Y., 2003. Airport charges and capacity expansion: effects of con-
cessions and privatization. Journal of Urban Economics 53 (1), 54e
75.
Zhang, A., Zhang, Y., 2006. Airport capacity and congestion when carriers have
market power. Journal of Urban Economics 60 (2), 229e247.
Zhang, A., Zhang, Y., 2010. Airport capacity and congestion pricing with both
aeronautical and commercial operations. Transportation Research Part B:
Methodological 44 (3), 404e413.
Zhang, B., 2008. A note on Fukuyama (2000). European Journal of Operational
Research 184 (3), 1182e1185 .
Zhang, B., Wang, J., Liu, C., Zhao, Y., 2012. Evaluating the technical efciency of
Chinese airport airside activities. Journal of Air Transport Management 20,
23e27.
Zhu, J., 2003. Quantitative Models for Performance Evaluation and Benchmarking:
Data Envelopment Analysis with Spreadsheets and DEA Excel Solver. Kluwer
Academic Pub.
Zhu, J., Shen, Z., 1995. A discussion of testing DMUs' returns to scale. European
Journal of Operational Research 81 (3), 590e596.
B. Zhang et al. / Journal of Air Transport Management 40 (2014) 149e156156
... Also, there are some studies on RTS of airport airside (Ablanedo-Rosas and Gemoets 2010; Barros and Dieke 2007;Wing Chow and Fung 2009;Lozano and Gutiérrez 2011a;Malighetti et al. 2007;Martín and Roman 2001). To the best of our knowledge, airside RTS, landside RTS and overall RTS (both airside and landside) are assessed separately in Pels, Nijkamp, and Rietveld (2003), Zhang et al. (2014a) and Zhang et al. (2014b) while in other studies they are not. ...
... Moreover, landside operations can be further decomposed into passenger-terminal operations and cargo-warehouse operations. In addition, airside operation is one part of air traffic control (ATC) (Zhang et al. 2014b). ATC is a service provided by ground-based controllers who direct aircraft on airport airside and through controlled airspace, and can provide advisory services to aircraft in non-controlled airspace. ...
... Many papers take runway length or number of runways as the principal index of a multiple runway system (Fung et al. (2008), Merkert et al. (2012) and Assaf, Gillen, and Tsionas (2014)). Zhang et al. (2012), Zhang et al. (2014a), and Zhang et al. (2014b) introduce take-off distance available (TODA) and landing distance available (LDA) as the index of capability of a multiple runway system in providing service for aircraft. However, whether several runways can be used simultaneously is not taken into account in the above index. ...
Article
Full-text available
This paper evaluates the operational performance of airside and landside at Chinese airports with two novel inputs. Furthermore, the airport landside operation is decomposed into passenger-terminal operations and cargo-warehouse operations. One novel input is the capability of the runway system which is introduced into airside performance evaluation. The other novel input is cargo warehousing which is introduced into landside performance evaluation. To address multiple optimal solutions when estimating Returns to Scale in Data Envelopment Analysis, we adopted the Zhu and Shen method. This empirical study shows that neither the number of runways nor total runway length is a genuine index of runway system capability in the evaluation of airside performance. Only four airports achieved full efficiency in all eight measures while eight airports did not achieve any full efficiency. In addition, one airport did not perform well in the benchmark analysis.
... By applying their approach to a real-world setting, they verified their approximations concerning the observed delays. Zhang et al. [23] investigated the impact of airside efficiency improvements with respect to capacity expansions. They especially focused on the take-off and landing distances of runways. ...
Article
Full-text available
The majority of the ground vehicles operating on the airside parts of commercial airports are currently powered by diesel engines. These include vehicles such as apron buses, fuel trucks, and aircraft tractors. Hence, these vehicles contribute to the overall CO 2 emissions of the aviation transport system and thus negatively influence its environmental footprint. To reduce this damaging environmental impact, these vehicles could potentially be electrified with on-board batteries as their energy sources. However, the conductive charging of such vehicles via stationary cable connections is rather time-consuming. A dynamic wireless charging system to supply public transportation passenger buses with electric energy while in motion has recently been installed on the Korea Advanced Institute of Science and Technology (KAIST) campus and in the Korean city of Gumi. In this paper, we study configuration problems related to the use of this technology to make airport operations more environmentally sustainable. We concentrate on the power supply for apron buses and analyze the location planning problems related to the distribution of the required power supply and the wireless charging units in the apron road system. To this end, we develop a formal optimization model and discuss the first numerical results.
Article
Most of the previous works ignore the fact that equilibrium flights in self-profit maximization scenario are totally different from that in joint profit (social welfare) maximization scenario and take price difference (flight fare difference) between the two scenarios as congestion price, which is a passenger-based method. Most of all, the function of congestion pricing is to alleviate congestion by making airlines reduce flights at peak time. Therefore, the equilibrium flights under self-profit maximization should be the same as the ones under joint profit maximization after congestion prices are tolled. Flight-based congestion pricing method is provided in our paper. The analysis suggests no role for congestion pricing when total real flight production of all airlines is less than the equilibrium flights under joint profit maximization scenario. Otherwise, congestion tolls should be levied to all airlines. Furthermore, congestion price can be determined by solving the corresponding equations system.
Article
Full-text available
The aim of this research is to investigate the operational efficiency and potential output increase scenarios of 21 small and medium sized airports (SMA) located in 10 different European countries. A key feature of the proposed Data Envelopment Analysis (DEA) models is the consideration of two novel inputs related to the number of airlines and scheduled routes operating at each airport, in addition to the more common capital non-discretionary inputs (runways, boarding gates and apron). These novel inputs are a way of considering the effects of the air transport market on airports' performance and growth prospects. In particular, a model to estimate the growth in passenger numbers, aircraft movements and cargo that can be achieved by an airport through increasing its number of airlines and routes, and assuming an efficient operation, is presented. Also, a multi-objective DEA model is proposed to explore the possible trade-offs in the output Pareto efficient frontier of a given airport. In addition, a second stage fractional regression model specific for explaining DEA efficiency scores is conducted and the results suggest that ownership structure and hub airport status are significant factors for explaining variations in the operational efficiency of European SMA.
Article
Full-text available
In the United States, most airports do not place any limitations on airline schedules. At a few major airports, the current scheduling restrictions (mostly administrative measures) have not been sufficiently strict to avoid consistent delays and have raised debates about both the efficiency and the fairness of the allocations. With a forecast of 1.1 billion yearly air travelers within the US by 2015, airport expansion and technology enhancement alone are not enough to cope with the competition-driven scheduling practices of the airline industry. The policy legacy needs to change to be consistent with airport capacities.Flights on US airlines arrived late more often in the first four months of 2007 than in any other year since the government began tracking delays, and flight cancellations increased 91% over 2006. With a forecast of 1.1 billion yearly air travelers within the US by 2015, airport expansion and technology enhancement alone are not enough to cope with the competition-driven scheduling practices of the airline industry. Our research studies how flight schedules might change if airlines were required to restrict their schedules to runway capacity. To obtain these schedules, we model a profit-seeking, single benevolent airline whose goal is to maintain current competitive prices and service as many current passengers as possible, while remaining profitable. Our case study demonstrates that at Instrument Meteorological Conditions (IMC) runway rates, the market can find profitable flight schedules that reduce substantially the average flight delay to less than 6 minutes while simultaneously satisfying virtually all of the current demand with average prices remaining unchanged. This is accomplished through significant upgauging to high-demand markets.
Article
Full-text available
This paper evaluates the technical efficiency of 37 Chinese airport airsides. We introduce take-off distance available, landing distance available, and aircraft-parking position as inputs into the efficiency evaluation. This study also estimates the input excess and output shortfalls. Benchmarks are identified for those technically inefficient airsides.
Article
We investigate how the intensity of competition among airports affects their technical efficiency by computing airports' markets on the basis of a potential demand approach. We find that the intensity of competition has a negative impact on airports' efficiency in Italy from 2005 to 2008. This implies that airports belonging to a local air transportation system where competition is strong exploit their inputs less intensively than do airports with local monopoly power. Further, we find that public airports are more efficient than private and mixed ones. Hence, policy makers should provide incentives to implement airports' specialization in local systems where competition is strong and monitor the inputs' utilization rate even when private investors are involved.
Article
This paper employs directional distance function to evaluate the technical efficiency of twenty major Chinese airports from 2006 to 2009 within a joint production framework of desirable and undesirable output (i.e. flight delays). The results indicate that the overall average efficiencies of Chinese airports increased over the period of time. The international hub airports are found to operate at higher efficiency level, which indicates that these airports run well in producing more desirable outputs and controlling flight delays. Although the average efficiency score of listed airports is higher than non-listed ones, the difference between the two group is statistically insignificant. A comparison between the results without and with flight delays show that several airports experienced significant changes in their efficiency scores after considering undesirable output.
Article
This paper assesses whether geographical characteristics and service strategies influence the performance of Chinese airports. It examines the technical efficiency of 41 Chinese airports in 2008; with data envelopment analysis then regressed on environmental factors. The results show that airports located in cities with populations of more than two million are more efficient than those for other city sizes. Further, airports that are able to accommodate A380 or Boeing 747 aircraft are operationally more efficient than other categories, and airports' efficiency improves with the number of airlines using them. The distance between an airport and the local central business district, however, does not significantly affect its performance.
Article
In this paper, we measure the productivity changes of 30 airports in Greater China during 2000–2006 by computing their Malmquist productivity indices using parametric output distance functions. On average, the productivity of airports grew at 10.3% annually over our period. In addition, decomposing the Malmquist index did not provide evidence supporting any “catching-up” in technical efficiency between regional airports and major hubs.
Article
In this paper, a directional distance approach is proposed to deal with network DEA problems in which the processes may generate not only desirable final outputs but also undesirable outputs. The proposed approach is applied to the problem of modelling and benchmarking airport operations. The corresponding network DEA model considers two process (Aircraft Movement and Aircraft Loading) with two final outputs (Annual Passenger Movement and Annual Cargo handled), one intermediate product (Aircraft Traffic Movements) and two undesirable outputs (Number of Delayed Flights and Accumulated Flight Delays). The proposed approach has been applied to Spanish airports data for year 2008 comparing the computed directional distance efficiency scores with those obtained using a conventional, single-process directional distance function approach. From this comparison, it can be concluded that the proposed network DEA approach has more discriminatory power than its single-process counterpart, uncovering more inefficiencies and providing more valid results.