The impact of airport competition on technical efficiency: A stochastic frontier analysis applied to Italian airport

Journal of Air Transport Management 22 (2012) 9e15

Contents lists available at SciVerse ScienceDirect

Journal of Air Transport Management journal homepage: www.elsevier.com/locate/jairtraman

The impact of airport competition on technical efficiency: A stochastic frontier analysis applied to Italian airport Davide Scotti, Paolo Malighetti*, Gianmaria Martini, Nicola Volta Department of Economics and Technology Management, University of Bergamo, Italy

a b s t r a c t Keywords: Airport efficiency Airport privatization Airport competition Stochastic distance function

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.
2012 Elsevier Ltd. All rights reserved.

1. Introduction

2. The Italian airport system

One effect of the liberalization process in the EU air transportation market has been the growth in the European network. European airlines can now provide intra-European connections (i.e., flights having an origin and a destination in airports within the EU 25) without restrictions provided there is slot availability. As a result, if we consider all the 460 airports of the 18 countries that belonged to the European Common Aviation Area (ECAA1) in 1997, the total number of connections among these airports rose from 3410 in 1997 to 4612 in 2008. This implies a compounded annual growth rate of 2.78%, with the number of connecting flights increasing from 4,102,484 to 5,228,688. The network expansion has increased the intensity of competition between airports, as they compete both directly for airlines and indirectly for passengers and freights, and as airline new business models have emerged, notably low cost carries (LCC). Further, travelers may now choose their travel suppliers from different airlines at the same airport (direct competition) or from ones operating at nearby ones (indirect competition). Here we investigate the impact of competition between airports and ownership on their technical efficiency. The latter impacts on both airport charges and services provided to passengers (e.g., shorter waiting times). For our empirical analysis we develop a potential demand approach and a multi-output stochastic frontier model. They are applied to 38 Italian airports between 2005 and 2008.

Before 1990, Italian airports were, as in many other European countries, controlled by the national government; although sometimes management was delegated to a public agency. The first important development was Act n. 537/93, which introduced changes in Italy’s airports’ ownership. First, it established that airports would no longer be under the control of the national government. Second, the management of airports was delegated to companies that may involve private agents, region or county governments, municipalities and chambers of commerce. Third, at least 20% of the shares in a company managing the airport had not to be in the hands of private agents. As a consequence, many local governments entered in the airports’ ownership, taking control in the vast majority of cases. In 1997, Act 521/97 eliminated the 20% minimum stake for local public and created a national public authority. Ente Nazionale per l’Aviazione Civile (ENAC) in charge of the sector’s regulation. ENAC directly manages Lampedusa and Pantelleria airports, facilities serving two small islands in the Mediterranean Sea. These reforms created the conditions for the gradual entry of private capitals into airport ownership. The first privatization took place in 1995 in Naples, where the British Airports Authority (BAA) got the majority of shares of the company managing the airport. Privatization occurred also in 2000 for ADR (that controls Rome Fiumicino and Rome Ciampino). Other airports with private ownership are Florence (year 2003), Venice (year 2005), Treviso (year 2007), Parma (second half 2008) and Olbia (since the beginning 1974). The majority of Italian airports are still, however, under the control of local public authorities or entail public or mixed ownership.

* Corresponding author. E-mail address: [email protected] (P. Malighetti). 1 15 EU members plus Iceland, Norway, and Switzerland. 0969-6997/$ e see front matter
2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.jairtraman.2012.01.003

10

D. Scotti et al. / Journal of Air Transport Management 22 (2012) 9e15

The Italian system consists of 45 airports open to commercial aviation. Rome Fiumicino and Milan Malpensa are the most important intercontinental airports, with further long haul European and domestic connections provided by 12 regional medium sized airports. The remaining 31 airports can be classified as regional with a limited number of European and domestic connections. All these airports have benefited from the EU liberalization of air transportation, with the average number of destinations served rising from 20 in 1997 to 37 in 2008, and with this has come increased competition between them.

lnðyMit Þ ¼ TLðxit ; yit =yMit ; a; b; zÞ  lnðDOit Þ

where, lnðDOit Þ is non-observable and can be interpreted as an error term in a regression. If we replace it with (vit  uit), we get the typical SFA composed error term: vit are random variables that are assumed to be iid as N(0, s2v) and independent of the uit; the latter are non-negative random variables distributed as N(mit, s2u). vit represent the random shocks, while the inefficiency scores are given by uit. Hence, we can now write the translog output oriented stochastic distance function for estimation as

lnðyMit Þ ¼ a0 þ

3. Methodology

M1 X

þ

K X

bk ln xkit þ

k¼1

We use stochastic frontier analysis to disentangle random shocks from on-going technical inefficiency (Aigner et al., 1977). Further, we incorporate exogenous variables, which are neither inputs to the production process nor outputs of it, but which nonetheless exert an influence on producers’ performance (Kumbhakar and Lovell, 2000). In terms of inputs, since our data set does not include monetary variables but only physical inputs and outputs, our aim is to measure technical efficiency e i.e., an airport management’s ability to achieve efficient input utilization. This means that we do not identify the input combination yielding the minimum cost. Moreover since airports are typically multi-product firms, an appropriate multioutput framework for estimating technical efficiency is required. As shown by Coelli and Perelman (2000), this implies the estimation of a stochastic distance function. Last we need to choose between input and output orientation. The former identifies the inputs’ reduction required to reach the efficient frontier. Given that in airport operation many inputs are indivisible, at least in the short run, an output oriented stochastic distance function seems to be appropriate, especially in a context where airports are in competition. In this framework we define P(x) as the airports’ production possibility set e i.e., the output vector y˛RM þ that can be obtained using the input vector x˛RKþ. That is: PðxÞ ¼ fy˛RM þ : x can produce yg. By assuming that P(x) satisfies the axioms listed in Fare et al. (1994), we introduce Shephard (1970) output oriented distance function:

DO ðx; yÞ ¼ minfq : ðy=qÞ˛PðxÞg; where q  1. Lovell et al. (1994) shows that the distance function is nondecreasing, positively linearly homogeneous, and convex in y, and decreasing in x. DO(x,y) ¼ 1 means that y is located on the outer boundary of the production possibility set e i.e., DO(x,y) ¼ 1 if y˛IsoqP(x) ¼ {y:y˛P(x), uy;P(x), u > 1}. If instead DO(x,y) < 1, y is located below the frontier; in this case, the distance represents the gap between the observed output and the maximum feasible output. Following Coelli and Perelman (2000), the translog distance function is given by:

lnðDOit =yMit Þ ¼ a0 þ

M1 X

am lny*mit þ

m¼1

þ

am ln y*mit þ

m¼1

3.1. The stochastic distance function econometric model

K X

bk lnxkit þ

k¼1



M1 X

X M1 X 1 M1 amn lny*mit lny*nit 2 m¼1 n¼1

K X K K 1X 1X bkl lnxkit lnxlit þ 2 2 k¼1 l¼1

zkm lnxkit lny*mit ;

k¼1

ð1Þ

(2)



M1 X

X X M1 1 M1 amn ln y*mit ln y*nit 2 m¼1 n¼1

K X K K 1X 1X bkl ln xkit ln xlit þ 2 2 k¼1 l¼1

zkm ln xkit ln y*mit þvit uit :

k¼1

(3)

m¼1

To investigate the determinants of inefficiency, we apply a single-stage estimation procedure following Coelli (1996) where the technical inefficiency effect, uit in Eq. (3) can be specified as:

uit ¼ dzit þ wit where the random variable wit is defined by the truncation of the normal distribution with zero mean and variance, s2, such that the point of truncation is dzit; i.e., wit  dzit. Further, zit is a px1 vector of exogenous variables that may influence the efficiency of a firm, and d is a 1xp column vector of parameters to be estimated. According to this time-varying specification of airports’ inefficiency, the technical efficiency of airport i at period t is defined as follows:

TEit ¼ euit : 3.2. Airport competition index One approach to defining markets assumes that an airport’s relevant geographic market consists roughly of a circle around its location. A fixed-radius technique is usually implemented to define the airport’s competitors (Malighetti et al., 2007). The fixed-radius technique, however, does not take into account the distribution of people living in the areas around the airport and neither does it consider the real access time to reach it nor determinants of the demand for airport services in the area (Gosling, 2003). To deal with these issues, we take into account that any measure based on the determinants of demand cannot be implemented using actual airport choices taken by users; their choices may be influenced by unobservable airport features (McClellan and Kessler, 2000). It is then necessary to compute predicted travelers choices based on exogenous factors. We consider traveling costs as exogenous factors affecting demand and build an airport geographic market (i.e., CA) based on this variable. The proxy we adopt is given by passenger traveling time to reach airports. Hence, we assume that individuals are potential passengers of any airport that they can reach in a reasonable time.2 Our technique, inspired by Propper et al. (2004, 2008), is composed of several steps. First, we draw a boundary around

m¼1

where M is the number of outputs, K is the number of inputs, DOit is the output distance from the frontier of firm i in period t and y*mit ¼ ymit/ yMit. Eq. (1) can be written as lnðDOit =yMit Þ ¼ TLðxit ;yit =yMit ; a; b; zÞ, where TL stands for the translog function. Hence:

2 As shown by Graham (2008), passengers’ demand for flights is function of their preferences regarding (1) the destination, (2) the type of flight (e.g., long/short haul, LCC/traditional, direct/connection flight, etc.) and (3) her/his “type” (e.g., business versus leisure). In this contribution we focus on a representative passenger, i.e., a passenger having an average of all the previous characteristics.

D. Scotti et al. / Journal of Air Transport Management 22 (2012) 9e15

airport i that defines all the zip codes within T min drive from that airport. We will consider the following specifications of the maximum traveling time: T ¼ {60,75,90,105,120}. Many national aviation authorities have shown that almost passengers choosing a given airport live at most within 90 min distance from it. We compute the traveling time from zip code j to airport i driving a car on three road types: urban roads, extra-urban roads, and motorways.3 All the zip codes falling within the T-minutes defined boundary are included in the catchment area of airport i; i.e., CAi. Second, we define hi as the set of population living in airport i’s catchment area. The latter is the population living in all zip code towns belonging to CAi. Similarly, hj is the set of population living in airport j’s catchment area, CAj. Third, since in air transportation each O-D route defines a separate market, airport i is subject to competition coming from airport j only if the same route is available at both airports. This means that airport i and airport j must have either the same airport destination, or a destination in different airports but located at a reasonable distance. We assume that flights have the same destination if the arrival airports are located at a maximum distance equal to 100 km. The application of different methodologies to estimating the potential demand at the origin and destination airports is due to the exogenous factors affecting them. Traveling costs are the main determinant of the origin airport’s potential demand, while the region where the travel is directed is instead the main factor influencing the destination airport’s potential demand. The intuition is the following: a traveler, when choosing a flight, considers first the region that needs to be reached (not necessarily the town but also the surrounding region), then she or he verifies whether, at a reasonable traveling distance, this region can be reached leaving from different origin airports. Hence, to consider all airports where route r is available, we define the following:

hij;r ¼ ¼

n

n

/

hi Xhj

hi Xhj Xhk

o

hk ; cksi; j hijk;r

/

o

hh ; chsi; j; k .;

where hij,r is the subset of population living in CAi, which has only the possibility to reach also airport j within T min traveling time for the route r; hijk,r is the subset of hi, which has only the possibility to reach also airport j and airport k within T min traveling time, always for the route r. Fourth, we denote as the potential demand of airport i on the route r:

h^i;r ¼ hi 

X1

X2

2

3

hij;r 

j

k

hijk;r 

X3 h

4

hijkh;r þ .

(4)

Fifth, the Competition Index for airport i on route r (CIi,r) is:

CIi;r ¼ 1 

h^i;r ; 0  CIi;r  1: hi

11

Fig. 1. An example of competition between airports.

CIi ¼

R X ASi;r r¼1

ASi

 CIi;r ;

(6)

where 0  CIi  1 and R is the number of routes available in airport i. This implies that the higher is CIi, the more airport i is subject to competition. Fig. 1 provides an example of the methodology. Suppose we want to compute CIA by applying Eq. (6). After having fixed a given level of T, the procedure draws the boundary of its catchment area, given by the gray area. Suppose that airport B is the unique nearby airport, and that people living in the dashed area represent the population that may, within T minutes, also reach airport B. Next we consider available routes involving the airports; airport A has routes A-C and A-D while airport B has only route B-E. Routes A-D and B-E belong to the same market for the population hAB since airport D is located at less than 100 km from airport E. Clearly, on route A-C, airport A is not subject to any competition coming from B. Hence, hAB,A-C ¼ 0, while hAB,A-D ¼ hAB. Consequently, from Eq. (4) ^A;AD ¼ hA  ð1=2ÞhAB . Then, from ^A;AC ¼ hA , while h we get that h Eq. (5) we get: CIA,A-C ¼ 0, while CIA,A-D ¼ 1  (hA  (1/2)hAB)/ hA ¼ hAB/2hA. Now, suppose that ASA,A-D ¼ 50 (i.e., during a year the number of available seats for the route A-D is equal to 50) and that ASA ¼ 100. Hence, from Eq. (6) we obtain CIA ¼ 0 þ (50/100)  (hAB/ hA) ¼ hAB/4hA, which is airport A’s competition index.

(5)

We need an aggregate index of competition for airport i e i.e., a measure that takes into account all of the routes available in that airport and also their relative importance. The latter is given, for route r, by the ratio between the number of available seats for route r in airport i (ASi,r) and the total number of available seats (ASi) in the same airport.4 Hence, the aggregate index of competition for airport i is defined as follows:

3 The driving times, influenced by the road types, are computed using Google Maps. 4 ASir and ASr are taken from the OAG database. The available seats are used to measure the flight capacity.

4. Data The multi-output/multi-input production frontier for Italian airports is estimated using annual data on 38 airports for 2005e2008. Our data set covers 84% of Italian airports and 99.97% of passenger movements. The data sources are ENAC for outputs (i.e., aircraft, passenger, and freight movements) and the technical information provided by the airports’ official documents for inputs. The latter have been integrated by a direct investigation with the managing boards of the airports. Information regarding exogenous variables have been collected from the Italian national institute for statistics (ISTAT) and from the airports’ balance sheets. We consider three outputs: yearly numbers of aircraft movements (ATM), passengers movements (APM) and freight (FRE). Regarding inputs we include in our data set a mixture of physical

12

D. Scotti et al. / Journal of Air Transport Management 22 (2012) 9e15

Table 1 Descriptive Statistics of Input (I) and Output (O) Variables.

ATM (O) (number) APM (O) (number) FRE (O) (tons) TERM (I) (sqm) CHECK (I) (number) FTE (I) (number) PARK (I) (number) CAP (I) (flights per hour) BAG (I) (number)

Table 2 Pearson Correlations of Input (I) and Output (O) Variables.

Average

Median

Std. Dev. Maximum Minimum

43,024 3,347,933 25,261 33,326 37 208 24 17 4

18,919 1,300,206 3569 11,600 17 74 16 12 3

63,881 6,048,541 74,169 69,630 62 387 25 17 3

346,650 35,226,351 486,666 350,000 358 2186 142 90 15

1748 7709 0 256 3 1 2 2 1

infrastructures (the runway capacity (CAP) measured as the maximum number of authorized flights per hour,5 the number of aircraft parking positions (PARK), the terminal surface area (TERM), the number of check-in desks (CHECK), the number of baggage claims (BAG)) and the number of employees measured in terms of full time equivalent units (FTE). The descriptive statistics regarding outputs and inputs are presented in Table 1. It is possible to check the validity of the chosen inputs and outputs by testing for their isotonicity e i.e., outputs should be significantly and positively correlated with inputs (Charnes et al., 1985). Pearson correlation coefficients are shown in Table 2. The correlation between all the inputs and the outputs is significant at a 1% level and positive. Moreover, the input correlation is positive, significant, and very high, as a confirmation that in managing airports, inputs are jointly dimensioned to avoid bottlenecks (Lozano and Gutiérrez, 2009). Further, we consider two types of exogenous variables. The first influences the production frontier, while the second one has an impact on the airports’ inefficiency scores. Hub (HUB) and Seasonality (SEASON) are the two variables influencing the frontier. HUB is a dummy variable equal to unity if the airport is a hub (an airport with a hub and spoke system employs different technologies, e.g., different BHS, than a non-hub airport). SEASON is a dummy variable equal to one if the airport belongs to a region whose monthly tourist flows are strongly seasonal and correlated with airports’ monthly passenger flows: tourist flows may have a high traffic variation across the months and this has an impact on airports’ production levels and not on their efficiency. Four variables are instead considered as determinants of airports’ inefficiency scores: the airport competition index (CIi), two dummies regarding ownership (PRIV for private ownership and MIX for mixed public-private ownership), and the degree of dominance of the main airline in a specific airport (DOM), which is a proxy of airline competition within the airport. The airport competition index (CIi) is computed from Eq. (6); Table 3 shows the distribution of the airport competition index as function of T. Fig. 2 confirms the positive correlation between the competition index and T, as well as the increase in its variance as the maximum traveling time grows. The latter implies that an enlargement of the airport’s catchment area does not have the same effect on all Italian airports. For some of them, this implies an increase in the competition index, while this is rather small for other airports. We consider two ownership dummies: PRIV, equal to one if the stake of private agents is higher than 50% of the capital stock and MIX, equal to one when the stake of private agents is greater than 25% but lower than 50% of the capital stock. Public airports are those where private agents have less than 25% of the shares. Twenty-eight airports were publically owned in 2005 and in 2008,

5 This variable takes into account both the runway length and the airport’s aviation technology level e e.g., some aviation infrastructure such as ground control radars and runway lighting systems.

ATM (O) APM (O) FRE (O) TERM (I) CHECK (I) FTE (I) PARK (I) CAP (I) BAG (I)

TERM (I)

CHECK (I)

FTE (I)

PARK (I)

CAP (I)

BAG (I)

0.936 0.968 0.808 1 0.979 0.895 0.927 0.920 0.875

0.969 0.958 0.642 0.979 1 0.928 0.923 0.943 0.903

0.958 0.856 0.695 0.895 0.928 1 0.859 0.932 0.836

0.890 0.874 0.802 0.927 0.923 0.859 1 0.904 0.858

0.944 0.946 0.738 0.920 0.943 0.932 0.904 1 0.875

0.878 0.936 0.860 0.875 0.903 0.836 0.858 0.875 1

private airports increased during the period, from five to seven, with the number of mixed ownership airports falling slightly. Finally, DOM is the percentage of AS offered by the airline with the largest market share at the airport; the greater the percentage, the lower is the competition among airlines. In terms of airports’ efficiency, this variable may also show the impact of incumbent carriers’ strategy to block entrance, which may limit the possibility to attract new airlines. This, in turn, may reduce the airport’s efficiency of asset utilization. 5. Results We estimate a multi-output stochastic distance function:

lnðAPMit Þ ¼ TLðATMit =APMit ; FREit =APMit ; TERMit ; CHECKit ; BAGit ; FTEit ; PARKit ; CAPit ; a; b; zÞ þl1 HUB þ l2 SEASON þ vit  uit ;

(7)

where APMit is the normalizing output (i.e., ATMit and FREit are expressed in APMit terms), a is a vector of coefficients for ATMit/ APMit and FREit/APMit, b is a vector of coefficients regarding inputs, and z is a vector of coefficients related to outputeinput interactions. The equation describing the impact of the exogenous variables on the inefficiency scores uit is the following:

mit ¼ d0 þ dCI CIit þ dPRIV PRIVit þ dMIX MIXit þ dDOM DOMit ;

(8)

where mit represents the mean of uit. Notice that not including an intercept parameter, d0, in Eq. (8) may imply the fact that the d parameters associated with the z variables are biased and that the shape of the inefficiency effects’ distributions are unnecessarily restricted (Battese and Coelli, 1995). Table 4 presents the results. Firstorder coefficients are all statistically significant with the exception of the number parking positions (PARK). Concerning second-order coefficients, terminal area (TERM), the number of check-in desks (CHECK) and the runway capacity (CAP) are statistically significant. Further, many interaction effects are statistically significant as a confirmation of the multi-output features of airport activity, with the exception of those coefficients regarding the interaction between freight movements and other inputs: a possible explanation is that many small regional airports are not involved in freight movements. Both the hub and seasonality dummies are not statistically significant. The likelihood function is expressed in terms of the variance parameters, s2 ¼ s2v þ s2u and g ¼ s2u =ðs2v þ s2u Þ. Table 4 shows Table 3 Distribution of Airport Competition Index as Function of T.

CI CI CI CI CI

(T (T (T (T (T

¼ ¼ ¼ ¼ ¼

60) 75) 90) 105) 120)

0

(0, 20] %

(20, 40] %

(40, 60] %

(60, 80] %

(80, 100] %

10 5 4 4 3

16 13 7 5 3

8 11 16 8 6

4 8 8 14 13

0 1 3 7 11

0 0 0 0 2

D. Scotti et al. / Journal of Air Transport Management 22 (2012) 9e15

13

Table 4 Results.

Fig. 2. The dispersion of airport competition as function of T.

that they are statistically significant at the 1% level, with the estimated g equal to 0.56; a relatively high value showing that a relevant part of the distance between the observed output levels and the maximum feasible ones is due to technical inefficiency. The hypothesis of normal error distribution is confirmed by the ShapiroeWilk normality test. Concerning the impact of airport competition on technical efficiency, since CIi is a function of T, we have performed a sensitivity analysis. Table 5 shows the estimated coefficients for different specifications of the maximum traveling time. They are always positive and statistically significant. Moreover, their magnitude is the largest among the determinants. This implies that airports with higher competitive pressure are less efficient. In contrast, in the Italian system, an airport that is closer to the local monopoly model (i.e., those airports with a competition index lower than 20% e Table 3) has an efficient utilization of its inputs. It would seem that airports confronted with higher levels of competition have lower technical efficiency because they still suffer from overcapacity. In these airports, the benefits coming from the traffic growth, induced by liberalization, have been distributed among many airports. As a result, those confronted with intense competition may have reduced their inefficiency levels in comparison with the pre-liberalization period, but not enough to get close to frontier. On the contrary, airports with local monopoly power did improve their performances, because they could fully exploit these benefits. This, in turn, has led to a more efficient assets’ utilization reducing spare capacity. Inefficient airports subject to intense competition may gain efficiency through the attraction of more passengers by increasing the number of routes they serve. In a competitive environment, however, this is not likely to be easy; active carriers incur switching costs when changing airports and, in a depressed market, new entry of any scale is unlikely. The coefficients of PRIV and MIX are both statistically significant and positive, and among them the coefficient of PRIV is the largest. This implies that public airports are more efficient than those with mixed ownership, whereas private airports have the lowest efficiency. This evidence confirms the results obtained by Curi et al. (2009) for Italian airports, but differs from Oum et al. (2008), who investigated the efficiency of the largest airports in the world, and from Chi-Lok and Zhang (2008), who studied the effects of privatization on Chinese airports.6

6 We rule out the possible endogeneity problem arising between inefficiency and privatization, since the anecdotical evidence we collected (mainly from newspapers) shows that the decision to privatize an airport has not been usually taken on the basis of efficiency reasons (but it was mainly based on political issues).

Parameter

Estimate

Constant ATM0 FRE0 TERM CHECK FTE PARK CAP BAG ATM0 ATM0  FRE ATM0  TERM ATM0  CHECK ATM0  FTE ATM0  PARK ATM0  CAP ATM0  BAG FRE0 2 FRE0  TERM FRE0  CHECK FRE0  FTE FRE0  PARK FRE0  CAP FRE0  BAG

5.1365* 2.0280*** 0.3386*** 1.7487** 2.3936** 3.7216*** 0.9748 7.6914*** 5.7138*** 0.2396*** 0.0002 0.33511*** 0.1341 0.1192** 0.3204** 0.4535*** 0.0129 0.0001 0.0379** 0.0179 0.0103 0.0029 0.0355 0.0376

TERM2 TERM  CHECK TERM  FTE TERM  PARK TERM  CAP TERM  BAG CHECK2 CHECK  FTE CHECK  PARK CHECK  CAP CHECK  BAG FTE2 FTE  PARK FTE  CAP FTE  BAG PARK2 PARK  CAP PARK  BAG CAP2 CAP  BAG BAG2 SEASON HUB

0.0236** 0.3697** 0.5148*** 0.2340* 0.9340*** 0.9591*** 1.9252*** 0.4880*** 0.0848 1.4290*** 1.7186*** 0.0825 0.2131* 0.5742*** 0.0093 0.0439 0.1836 0.4493*** 0.6835*** 0.2332 0.0777 0.0598 0.0851

ConstantZ CI (T ¼ 90) PRIV MIX DOM

2.2851*** 3.4519*** 1.0176*** 0.8134*** 0.6383*** 0.0309*** 0.5594*** 68.283 92.378

s2 g LR Log likelihood value

Note *,**,*** denote significance at 10%, 5% and 1% respectively. denotes normalized output (ATP/APM and FRE/APM)

0

Table 5 Regression Results. Parameter CI CI CI CI CI

(T (T (T (T (T

¼ ¼ ¼ ¼ ¼

60) 75) 90) 105) 120)

Estimate 2.1422*** 5.4519*** 3.4519*** 3.2369*** 0.4136**

Note *,**,*** denote significance at 10%, 5% and 1% respectively.

14

D. Scotti et al. / Journal of Air Transport Management 22 (2012) 9e15

Table 6 Airports’ Technical Efficiency Scores. Airport

2005

2006

2007

2008

CAGR

Alghero Ancona Bari Bergamo Bologna Bolzano Brescia Brindisi Cagliari Catania Crotone Cuneo Florence Foggia Forlì Genoa Lamezia Lampedusa Milan Linate Milan Malpensa Naples Olbia Palermo Pantelleria Parma Perugia Pescara Pisa Reggio Calabria Rimini Rome Ciampino Rome Fiumicino Turin Trapani Treviso Trieste Venice Verona

0.992 0.979 0.992 0.954 0.981 0.992 0.422 0.987 0.991 0.991 0.933 0.975 0.533 0.992 0.892 0.984 0.987 0.991 0.849 0.981 0.983 0.980 0.990 0.991 0.265 0.986 0.988 0.933 0.987 0.986 0.482 0.957 0.900 0.961 0.732 0.950 0.912 0.720

0.992 0.980 0.990 0.964 0.979 0.965 0.513 0.987 0.992 0.991 0.767 0.992 0.645 0.992 0.850 0.984 0.990 0.991 0.845 0.979 0.979 0.979 0.988 0.992 0.406 0.987 0.989 0.923 0.984 0.984 0.476 0.960 0.967 0.953 0.927 0.958 0.938 0.930

0.990 0.986 0.988 0.924 0.971 0.960 0.481 0.988 0.991 0.991 0.894 0.971 0.571 0.992 0.804 0.981 0.989 0.991 0.827 0.972 0.980 0.881 0.984 0.990 0.285 0.981 0.989 0.767 0.986 0.983 0.622 0.952 0.973 0.945 0.842 0.967 0.834 0.958

0.989 0.983 0.988 0.898 0.970 0.971 0.522 0.987 0.990 0.991 0.935 0.944 0.624 0.991 0.970 0.983 0.989 0.990 0.716 0.979 0.984 0.981 0.987 0.990 0.222 0.984 0.989 0.754 0.987 0.987 0.595 0.967 0.977 0.970 0.830 0.965 0.906 0.955

0.05% 0.11% 0.11% 1.49% 0.29% 0.54% 6.91% 0.01% 0.01% 0.00% 0.04% 0.81% 3.99% 0.02% 2.12% 0.03% 0.06% 0.04% 4.17% 0.05% 0.04% 0.04% 0.09% 0.02% 4.29% 0.07% 0.03% 5.19% 0.02% 0.04% 5.40% 0.27% 2.06% 0.23% 3.16% 0.38% 0.18% 7.32%

First, this may be because investments in indivisible inputs may have been greater in private airports; indeed many local governments have privatized airports taking into account the investment plans of the new owners. As a consequence, in the vast majority of cases privatization implied an increase in investment, especially in indivisible inputs. Given the difficulties involved in reaching the short-run volume of traffic required for an efficient utilization of these indivisible inputs, private airports have lower technical efficiency than the other types. Second, private airports maximize profit and thus they could turn out to be the most efficient ownership type if we estimated a cost rather than a production frontier. Profit maximization would embrace different criteria, including commercial revenues, that are not considered here, as well as the fact that the monetary costs vector across airports may differ from the physical because of local price variations. Further, many regional airports, controlled by local governments, increase their traffic by attracting new airlines, and especially LCCs, with subsidizes.7 As a result, they obtain higher utilization of their assets. DOM is statistically significant and positive indicating that airport efficiency is positively related to airline competition: when the latter is strong, an airport is more efficient. This negative dominance effect may be because of entry deterrence measures adopted by incumbent

7 The recent case of Ryanair and Alghero (a regional airport in Sardinia) is an example. In 2009, Ryanair received subsidies of V6.4 million (this is called “comarketing”), while the public company managing the airport incurred about V12 million of losses. The local government of Sardinian region, which is on the board of the company managing the airport covered this loss.

airlines. As a consequence, an airport’s capacity to attract new routes is limited, as, in turn, is utilization of assets. This factor may be particularly important in Italy where the main carrier is Alitalia, which has frequently acted to prevent new carrier entry. Table 6 presents the dynamics of efficiency over 2005e2008, and shows that the Italian system has raised its technical efficiency. 6. Conclusion This paper has investigated the impact of airport competition on the efficiency of 38 Italian airports by applying a stochastic distance function model with time-dependent inefficiency components to a panel data set covering 2005e2008. We find that airports confronted with more competition are less efficient than those benefiting from local monopoly power. Further, we show that public airports are more efficient, while private airports are even less efficient than those with mixed ownership. These results suggest, first, that there are two ways to deal with the negative relationship between airport indirect competition and technical efficiency: one is to induce airport specialization within the same territorial system (e.g., one airport may focus on LCCs and another on cargo). Since passengers living in these areas can choose among alternative airports, a further extreme possibility is to close down the highly inefficient airports. This option may be adopted when these airports persistently produce losses, that in Italy have been covered by public local taxation. Second, regulation should also monitor the efficient assets’ utilization even when private capitals are in control of the airport’s management. This may require adopting mechanisms providing incentives to improve the degree of technical efficiency, e.g., price cap regulation. This contribution has not considered airport cost efficiency, which may lead to different ownership rankings. Further, we did not take into account some negative effects in airport activities, such as noise and pollution produced in the surrounding area, which may overturn our results. These issues are left for future research. Acknowledgments The authors would like to thanks Kenneth Button and Anming Zhang for their helpful comments. We also thank participants to the 2009 Kuhmo Nectar Conference, 2010 ATRS Conference, 2010 WCTR Conference and 2010 JEI Conference. This research is supported financially by the University of Bergamo. References Aigner, D.J., Lovell, C.A.K., Schmidt, P., 1977. Formulation and estimation of stochastic frontier production function models. Journal of Econometrics 6, 21e37. Battese, G.E., Coelli, T.J., 1995. A model for technical inefficiency effects in a stochastic frontier production function for panel data. Empirical Economics 20, 325e332. Charnes, A., Cooper, W.W., Golany, B., Seiford, L., Stutz, S., 1985. Foundations of data envelopment analysis for Pareto-Koopmans efficient empirical production function. Journal of Econometrics 30, 91e107. Chi-Lok, A.Y., Zhang, A., 2008. Effects of competition and policy changes on Chinese airport productivity: an empirical investigation. Journal of Air Transport Management 15, 166e174. Coelli, T., 1996. A Guide to FRONTIER Version 4.1: A Computer Program for Stochastic Frontier Production and Cost Function Estimation. CEPA Working Paper 96/07. Centre for Efficiency and Productivity Analysis, University of New England, Armidale. Coelli, T., Perelman, S., 2000. Technical efficiency of European railways: a distance function approach. Applied Economics 32, 1967e1976. Curi, C., Gitto, S., Mancuso, P., 2009. The Italian airport industry in transition: a performance analysis. Journal of Air Transport Management 16, 218e221. Fare, R., Grosskopf, S., Lovell, C.A.K. (Eds.), 1994. Production Frontiers. Cambridge University Press, Cambridge. Gosling, G., 2003. SCAG Regional Airport Demand Model e Literature Review. Cambridge Systematics Inc, SH&E Inc, Berkeley.

D. Scotti et al. / Journal of Air Transport Management 22 (2012) 9e15 Graham, A. (Ed.), 2008. Managing Airports: An International Perspective. Elsevier, Oxford. Kumbhakar, S.C., Lovell, C.A.K. (Eds.), 2000. Stochastic Frontier Analysis. Cambridge University Press, Cambridge. Lozano, S., Gutiérrez, E., 2009. Efficiency analysis and target setting of Spanish airports. Networks and Spatial Economics 38, 131e139. Lovell, C.A.K., Richardson, S., Travers, P., Wood, L.L., 1994. Resources and functionings: a new view of inequality in Australia. In: Eichhorn, W. (Ed.), Model and Measurement of Welfare and Inequality. Springer-Verlag, Berlin. Malighetti, P., Martini, G., Paleari, S., Redondi, R., 2007. An empirical investigation on the efficiency, capacity and ownership of Italian airports. Rivista di Politica Economica 47, 157e188.

15

McClellan, M.B., Kessler, D.P., 2000. Is hospital competition socially wasteful? The Quarterly Journal of Economics 115, 577e615. Oum, T.H., Yan, J., Yu, C., 2008. Ownership forms matter for airport efficiency: a stochastic frontier investigation of worldwide airports. Journal of Urban Economics 64, 422e435. Propper, C., Burgess, S., Green, K., 2004. Does competition between hospitals improve the quality of care?: Hospital death rates and the NHS internal market. Journal of Public Economics 88, 1247e1272. Propper, C., Burgess, S., Gossage, D., 2008. Competition and quality: evidence from the NHS internal market 1991e1996. Economic Journal 118, 138e170. Shephard, R.W. (Ed.), 1970. Theory of Cost and Production Functions. Princeton University Press, Princeton.