Structural and operational management of Turkish airports: a bootstrap data envelopment analysis of efficiency
Introduction
Airports perform a vital function in globalization, connecting cities and countries, as well as the economic development of a region or the nation as a whole. They are also a crucial part of international commerce and tourism. Therefore, the performance of airport management can directly affect the competitiveness of a country. Performance evaluation of airports and policy suggestions for inefficient airports play a vital role in increasing this competitiveness.
In recent years, civil aeronautics is on a rising trend in Turkey. The number of aircraft raised by 180% and the number of airports approximately doubled since 2002. Moreover, the number of passengers increased by 343%, from 34 million to 150 million. In this way, the aviation sector produced USD 2.2 billion and USD 24 billion in income, in 2003 and 2014, respectively (Thomas, 2015). According to the World Bank data, the increase in air transport perfectly has illustrated in Fig. 1 over the last 30 years. It is seen that Turkish air traffic is higher than Belgium, France, Italy, Luxembourg and the Netherlands, the founder countries of the European Union except for Germany. Istanbul's Sabiha Gökcen Airport, ranked 18th with 23.5 million passengers, entered the list for the European Top 20 airports in 2014 via this rise. A new İstanbul airport constructed with six runways, four terminals, and 150 million passengers (per year). Thus, İstanbul international airport aims to become one of the greatest in the world between 2018 and 2020 and increase interest in the Turkish market. Antalya Airport from the major tourism regions in Turkey is also one of the top 20 European airports for passenger traffic with 28.3 million passengers in 2014 (Thomas, 2015).
The performance of airports is mainly based on the efficiency of their structural and operational. Structural efficiency measures whether or not immovable assets in an airport are used productively. Thus, the structural efficiency model incorporates inputs such as the number of runways, the capacity of car parks, the terminal area. Operational efficiency introduces to the capability of an airport to produce outputs in terms of its operating inputs. Therefore, the operational efficiency has involved movable inputs such as the number of full-time employees. For Turkey, the country that dedicates most of its total government budget to transport infrastructure, it is very essential to effectively evaluate and analyse the structural and operational efficiencies of the airports in Turkey. However, no studies have investigated their management performance with structural and operational efficiencies using a bootstrap approach of data envelopment analysis (DEA).
The two-stage framework to measure the efficiencies of the airports was proposed in this study. In the first stage, to separate the airports into homogeneous sub-groups, Classification and Regression tree algorithm (CART) as initiated by Breiman et al. (1984) was used. In the second stage, both conventional and bootstrap approaches in DEA were used to measure the airports' structural and operational efficiencies. It aims to suggest strategic plans and evaluate managerial performances of Turkish airports and offer suggestions for the inefficient airports. The data of this study were collected from the reports published by the Turkish General Directorate of State Airports (TGDSA, 2018). This study concentrates on the most important 43 airports in Turkey because of limited data of newly constructed airports in Turkey.
This study has three aspects: First, the study is divided into homogeneous sub-groups that can be compared. Second, this study investigates the structural efficiencies and operational efficiencies of Turkish airports to suggest performance improvement strategies. Finally, the airports were analysed using the bootstrap approach suggested by Simar and Wilson (2007).
The Turkish airport review did not pay attention to whether airports are national or international, private or public, and large-scale or small-scale. The performance of the airports was analysed based on the variables obtained, as in recent studies on Turkish airports (Örkcü et al., 2016). The first difference of this study, grouping 43 airports in Turkey with CART clustering methods, is to evaluate performance review for each cluster. There are no studies that group airports by using cluster analysis and then perform performance analysis. Comparable airports are in the same group, thanks to this cluster analysis method. Lo Storto (2018) clustered the Italian airports in three main typologies (only publicly owned, mostly owned by public partners, and mostly owned by private partners) without using any cluster analysis method. In this respect, our study is similar to lo Storto (2018) study.
CART method grouped many airports at the international level in the same cluster, and this group is called the big-scale group (BSG). The international airports such as Istanbul Ataturk, Ankara Esenboğa, Muğla Bodrum, Muğla Dalaman were included in this group. The middle-scale group (MSG) includes airports such as Gaziantep, Hatay, Erzurum, Elazığ, and Konya that serve mostly domestically but have international traffic. The small-scale group (SSG) consists mostly of local airports, such as those in Ağrı, Erzincan, Mardin, Muş, Siirt, Şırnak and Tokat.
In studies in the literature about the airports in Turkey, all airports or a specific number of airports were taken into account without any pre-analysis or information. For example, the handled number of airports in Ülkü (2015) and Örkcü et al. (2016) are limited when compared to all airports in Turkey.
The studies in the literature regarding Turkey airports are mostly based on conventional DEA and Malmquist analysis. As another difference, besides the conventional DEA model, analyses were performed with bootstrap DEA, which is accepted to have more consistent results. It is recommended to use bootstrap DEA methods since conventional DEA results are biased.
Another important difference is the structural and operational review of airport performances. Structural efficiency concerns land utilization and measures whether terminal land is used efficiently. A structural efficiency model includes unmovable inputs related to airport property, such as terminal area, the capacity of the car park, the number of the runway, and the number of all vehicles. Operational efficiency refers to an airport's ability to generate revenues in terms of its operating inputs such as labour, total expenditures. For example, Ennen and Batool (2018) analysed for 12 major airports in Pakistan to investigate performance levels such as runway system efficiency, passenger terminal efficiency, and staff efficiency. We similarly studied structural and operational efficiency by conducting a more detailed examination, for example, of any airport with a high performance score in terms of structure and low performance score in terms of operations. Such detailed results could not be observed if the variables were taken as a whole.
The remainder of the paper is organized into six sections. A short statement to both the transport sector in Turkey and airport efficiency measurement literature were presented in Section 2. The background of CART, the conventional and bootstrap DEA models, were presented in Section 3. The data set was introduced in Section 4. The results are presented in Section 5, and Section 6 summarizes the summary of this work and draws conclusions.
Section snippets
Literature review
One of the most important models used to measure the performance of airports is probably DEA models. The most common of these models are the CCR (Charnes et al., 1978) and the BCC (Banker et al., 1984) models. They were under the assumptions of the constant return to scale (CRS) and variable return to scale (VRS), respectively.
One of the earliest studies on airport efficiency indicators was performed by Doganis and Graham (1987). Their indicators determine the strengths and weaknesses of
Methodology
The suggested approach used in the paper is shown in Fig. 2. First, the CART method was used to divide all of the airports into several sub-groups to be both comparable and homogeneous. Then, to measure two efficiency type, the conventional and bootstrap approach in DEA have been analysed. These methods were briefly mentioned below due to beyond the principal scope of this paper.
First, CART, one of the machine learning approaches, is a useful tool to explore and discover meaningful data. CART
Airports sector in Turkey
The airport sector is developing rapidly in Turkey. Istanbul is becoming one of the prominent hub airports globally, along with the construction new airport, whereas Turkish Airlines is emerging as the most profitable among European firms recently (Thomas, 2015). The TGDSA controls 53 operational airports in Turkey. According to TGDSA data, Turkey is now one of the largest shareholders of the European air transportation network with the added 342 flights per daily. As a result of the driving
Empirical results
This study investigates both the structural efficiencies and operational efficiencies of the airports in Turkey to suggest performance improvement strategies. Structural and operational efficiency must be separately considered because they are different concepts. First, when considered for structural efficiency, it would be wrong to choose an input-oriented model so as not to damage the structural integrity that constitutes the basic structures of airports such as terminal areas and runways.
Summary and conclusion
The assessment of the efficiency of airports is a vital role in the aviation sector. In the present study, the managerial performances of 43 Turkish airports were evaluated to investigate by applying both conventional and bootstrap DEA. This study contributes to the literature by dividing airports into sub-groups to make a comprehensive assessment of operational and structural efficiency. This study also presented the bootstrap approach based on DEA, which succeeds in various shortcomings of
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
The author thanks the Editor-in-Chief, Professor Janice A. Beecher, and anonymous referees for their insightful comments and suggestions.
References (68)
Reform and efficiency of New Zealand's airports
Util. Pol.
(2015)- et al.
Measuring airport quality from the airlines' viewpoint: an application of data envelopment analysis
Transport Pol.
(2001) Airports in Argentina: technical efficiency in the context of an economic crisis
J. Air Transport. Manag.
(2008)- et al.
Measuring the economic efficiency of airports: a Simar–Wilson methodology analysis
Transport. Res. E Logist. Transport. Rev.
(2008) - et al.
Size versus efficiency: a case study of US commercial airports
J. Air Transport. Manag.
(2003) - et al.
The classification and regression tree approach to pump failure rate analysis
Reliab. Eng. Syst. Saf.
(2003) - et al.
Applied data envelopment analysis
Eur. J. Oper. Res.
(1991) - et al.
Measuring the efficiency of decision making units
Eur. J. Oper. Res.
(1978) - et al.
Estimating indices of airport productivity in Greater China
J. Air Transport. Manag.
(2012) - et al.
A comparison of parametric and non-parametric distance functions” with application to European railways
Eur. J. Oper. Res.
(1999)