Exploratory analysis on LCC potential to influence airport efficiency

Size: px
Start display at page:

Download "Exploratory analysis on LCC potential to influence airport efficiency"

Transcription

1 Exploratory analysis on LCC potential to influence airport efficiency António Sérgio de Azevedo Domingues Dissertação para obtenção do Grau de Mestre em Complex Transport Infrastructure Systems Júri Presidente: Prof. Dr. Luís Guilherme de Picado Santos Orientador: Prof. Dr. Maria do Rosário Maurício Ribeiro Macário Vogal: Prof. Dr. Paulo Manuel da Fonseca Teixeira July 2011

2 Abstract The aim of this study is to perform an evaluation of the Portuguese airport infrastructure s efficiency performance. It will be undertaken an extensive literature review of methodological approaches in airport performance research in order to understand which techniques are the most meaningful to be used in this case study. Then, airport efficiency will be object of study in the different areas that characterize it: airside, landside and within the airport-airline relationship. Finally, we will choose one of the methodologies to perform the analysis of the Portuguese airport efficiency performance in order to withdraw and analyse the results. Keywords: LCC, Airports, DEA, efficiency 1

3 Acknowledgments I would like to thank Professor Rosário Macário for supervising my thesis dissertation. Moreover, the development of this dissertation would not be as valuable without the contributions of Professor Richard deneufville s priceless advices and Professor Carlos Pestana Barros support regarding the DEA methodology. In addition, I would like to thank all CTIS faculty, lecturers and colleagues, responsible for some of most the interesting discussions I had in the last two years. I am also indebted to ANA s marketing directors Renata Tavares, Andreia Pavão and Jocelyn Ferreira for their availability in providing useful data. I want to address a special thank you note to Nuno Costa for his personal availability and explanations concerning Lisbon airport s operations. Last but not least, I want to thank my family and closest friends for their enthusiastic support on this stage of my academic experience. I am truly grateful to have you all in my life. 2

4 List of Contents 1 Introduction Deregulation of the air transport market The background of no-frills airlines Literature review First steps on efficiency research: Farrell s contribution Efficiency measurement: Simple Case The efficient production: Simple case The development of Farrell s ideas Estimation processes for parametric frontiers Linear programming methods Methodological framework on airport benchmarking Partial Measures Total factor productivity Malmquist index of productivity change Data Envelopment Analysis (DEA) Stochastic Frontier Analysis (SFA) Literature on airport efficiency research Main Drivers of Airport Efficiency Airside Airfield design Capacity and delays of airfields Demand management Landside Passenger buildings Security and check-in processes Low-Cost Airports Airport-Airline Relationship Airport privatization and management LCC s implication on airports revenues Regulatory environment Analysis of Portuguese airports efficiency Institutional setting

5 4.2 Data Collection Model formulation Selection of Inputs and Outputs Discussion of results Approach 1: Monthly analysis Approach 2: Each airport in each year as an individual firm Approach 3: Five year panel data Overall results Conclusions Bibliography Appendix 1 Airport efficiency performance research Appendix 2 Compilation of Data Appendix 3 DEA scores on Portuguese airports

6 List of Figures 1-1: European Seat Capacity by Service Type between 2002 and 2006 AND annual variation (Source: RDC, 2007) : Evolution of passenger traffic structure in portuguese airports (Source: ANA, 2009) : Farrell's Technical Efficiency (left) and piecewise Production Frontier (Right) (Source: (Farrell, 1957)) : Farrell's original diagrams on Increasing and Diminishing Returns to Scale (Source: (Farrell, 1957)) : ATRS TFP for world airports (Source: (Graham A., 2008)) : Scale efficiency in the BCC-dea MODEL (source: (Civil Aviation Authority, 2000)) : New York LaGuardia Airport Layout (Source: (FAA, 2003)) : Denver International Airport layout (Source: (FAA, 2003)) : Left-Conventional exit taxiway (source: (FAA, 2007) ) Right - high-speed exit (source: (ICAO, 2004) ) : Potential congestion points at boston Logan (Adapted from (de Neufville & Odoni, 2003) and (FAA, 2003)) : 2009'S MONTHLY PAX DISTRIBUTION ON THE TOP5 BUSIEST AIRPORTS (SOURCE: Adadpted from (ANA, 2009)) : Finger pier at Laguardia airport (left) and satellites at Tampa airport (right) (source: (FAA, 2003)) : Midfield linear at Denver airport (left) and X-shaped at Pittsburgh airport (source: (FAA, 2003) ) : Linear (Left) and Hybrid configuration (right) at Kansas City and Seattle Tacoma airports (source: (FAA, 2003)) : World distribution of airports using CUSS kiosks and percentage with BTC (source: (IATA, 2010)) : Portuguese airport operator s net profit between 2004 and 2009 (source: (ANA, 2009b)) : Structure of ANA s group of companies (Source: (ANA, 2009B)) : Annual variations of PAX and ATM between 2005 and 2009 (SOURCE: ANA) : VRS-DEA OUTPUT ORIENTED: MODELS 1 (LEFT) AND 2 (RIGHT)

7 4-4: VRS-DEA OUTPUT ORIENTED: MODEL : VRS-DEA OUTPUT ORIENTED: MODEL : VRS-DEA OUTPUT ORIENTED: MODELS 5 (LEFT) and 6 (RIGHT) List of Tables Table 2-1: Major methodologies in airport performance analysis (adapted from (Lai, Potter, & Beynon, 2010)) Table 2-2: Airport efficiency papers (adapted from (Graham A., 2008) and (Lai, Potter, & Beynon, 2010)) Table 3-1: Traffic at worlds' 25 busiest airports (source: (ICAO, 2009), ACI web site) Table 3-2: Utilization ratio of Portuguese busiest airports in 2009 (Source: ANA) Table 3-3: Airport regulatory method in EU countries in 2006 (Source: (Marques & Brochado, 2008)) Table 3-4: Single or dual-till approach in EU countries in 2006 (adapted from (Marques & Brochado, 2008) ) Table 3-5: Characteristics of the new Portuguese economic regulation on airports' charges Table 4-1: LCC operating in Portuguese airports between 2005 and Table 4-2: Yearly disaggregated data used in the VRS-DEA Output-Oriented models Table 4-3: Model characterization on the VRS-DEA efficiency benchmark Table 4-4: Global statistics for models 1 (disaggregated) and 2 (aggregated) Table 4-5: Global statistics for models 3 (disaggregated) and 4 (aggregated) Table 4-6: Global statistcs for models 5 (disaggregated) and 6 (aggregated)

8 List of Abbreviations ACI ANA ANAM ATM BSC BTC CAA CID CRS CUSS DBOT DEA ELFAA FAA IACTA ICAO LCC LoS LP MI MOPTC NIRS NLA OLS PCR RDC RoR SFA SMC STATFOR TBG TE TFP TPS TTA VRS WLU Airport Council International Aeroportos de Portugal, SA Aeroportos e Navegação Aérea da Madeira, S.A Air Transport Movement Baggage System Capacity Bag tag capability British Civil Aviation Authority Check-In Desks Constant Returns to Scale Common-use Self-service (Check-in) Design-Build-Operate-Transfer Data Envelopment Analysis European Low Fares Airlines Association Federal Aviation Administration International Air Transportation Competition Act International Civil Aviation Organization Low Cost Carrier Level of Service Linear programming Malmquist Index Ministry of Public Works, Transports and Telecommunications Non-Increasing Returns to Scale New Lisbon Airport Ordinary least squares Price Cap Regulation Runway Declared Capacity Rate of Return Regulation Stochastic Frontier Analysis Social Marginal Cost Eurocontrol s Air Traffic Statistics and Forecasts Total Boarding Gates Technical Efficiency Total Factor Productivity Total parking stands Total Terminal Area Variable Returns to Scale Workload unit 7

9 1 INTRODUCTION This dissertation aims to understand the influence of Low-Cost carriers traffic in Portuguese airports efficiency. Air travelling is now accessible to more people that could not afford to do so twenty years ago, and low-fares airlines are largely responsible for this democratization process. In one hand, the low-fares segment in air transportation has been registering a steady growth in the past decade. On the other hand, as a public monopoly, the Portuguese airport operator registers yearly profits that are becoming more attractive to private operators that coupled to State s need of budgetary balance, may result in privatization process to perform additional income in order to reduce national debt. The structure of this dissertation is as follows. Chapter 1 will introduce the historical framework of no-frills airlines. In chapter 2, we will conduct a literature review on efficiency research, from the first steps taken on this field to the research conducted on airport efficiency. Afterwards, chapter 3 will try to describe which are the drivers the influence the most airport operations, and thus their efficiency. Such drivers will be approached under three perspectives airside, landside and the airport-airline relationship. We then hope to have gathered all necessary information to conduct an analysis on Portuguese airports efficiency on chapter 4. This chapter will focus on the institutional setting, model formulation and ultimately discussion of results. Finally, conclusions will be drawn in chapter DEREGULATION OF THE AIR TRANSPORT MARKET The deregulation and liberalization of air transport, firstly in the United States (US) in the late 70 s and more recently in Europe and Asia in the late 90 s, transformed entirely the industry and travel patterns worldwide. This is the primary reason why we face today a paradigm shift in the management of airport infrastructures and the need for in-depth research of airport s performance. Although formal deregulation of air transport in the US only started in 1978 when President Carter signed the Airline Deregulation Act, there was already an history of continuous efforts to adopt competitive bilateral agreements that goes back to the 1940 s. This process culminated in the International Air Transportation Competition Act of 1979 (IATCA) that stated clearly the main purpose to allow designation of multiple carriers, liberal access to charter carriers, the elimination of capacity and fare restrictions, and common treatment of domestic and foreign carriers for airport facilities. (H. Good, Röller, & Sickles, 1995) Boosted by the enthusiasm of US deregulation, one of the first IACTA s implementations was the US-Netherlands bilateral, signed in March 1978, which allowed individual carriers to determine their own capacity, frequency and tariffs with reduced government intervention. It also provided multiple carrier designation and virtually unlimited access by charter operators. This was to become the trendsetter for future US bilateral. By the end of the year, Germany and Belgium, mainly due to geographic reasons, had 8

10 also reviewed their bilateral agreements with the US, as they could not be less liberal on either scheduled or charter rights than the Dutch had been, otherwise transatlantic air traffic would be diverted via Amsterdam. (Doganis R., 2001) In 1982, the US also succeeded in negotiating an agreement with the European Civil Aviation Conference for North Atlantic routes. This agreement stipulated that governments would automatically approve any fare that was in a 'zone of reasonableness that was as low as 50% of current fares. (H. Good, Röller, & Sickles, 1995). Deregulation through bilateral renegotiation was also being pursued by the US in other international markets. Following the same pattern as of the agreements established in Europe, the North and mid-pacific markets were target of several key bilateral agreements between 1978 and 1980, with Singapore, Thailand and Korea, and other states later (Doganis R., 2001). These bilateral agreements were undoubtedly more favourable to the United Stated rather than to Europe or Asia, since the number of departing points from the US was restricted. On the other hand, European air and sea transport was exempt from rules regarding competition policy by the article 84 of the 1957 s Treaty of Rome. This resulted in a majority of stated-owned airline companies, with three main sets of implications according to Good, Röller, & Sickles (1995). Firstly, managerial incentives for productive efficiency are compromised because ownership is not concentrated in the hands of individuals who have the lowest monitoring costs. Secondly, government ownership does not focus on simple profit maximization due to other sets of political agendas (usage of national airlines for employment policy, power limitation to avoid competition with other state-owned transportation companies such as railroads, etc). Finally, government ownership has brought with it access to subsidy which tends to restrain the pursuit of productive efficiency. Until the French Seamen case (European Court of Justice, 1974), European Member-States strongly opposed EU s interference in the air and maritime sectors. This case was crucial for the longterm application of completion rules to transportation sectors. However, the historical Nouvelles Frontiéres case (European Court of Justice, 1986) was the turning point in the Commission s attempts to introduce liberality into the air sector. In this case, the European Court of Justice definitively confirmed that EU Treaty s competition rules applied to the air transport sector. Although from 1984 onwards, Europe too moved away from the traditional bilateral agreements into more liberal and free-market policies pushed by the United Kingdom, the jurisprudence of the French Seaman case was of great importance as the driving force of the Commission s efforts to liberalize European air transport. Nevertheless, this process still took five years and three legislation packages (respectively applicable from January 1 st 1988, November 1 st 1990 and January 1 st 1993) to be finalized, and it only concerned Member-states. These gradual steps can only be understood in a wider political perspective of European integration. Signed in February 7 th 1992, Maastricht Treaty had as goal the creation of a single internal market, covering (by then) the twelve member-states, to come into existence in the beginning of 9

11 1993. This meant that by the end of 1992, the twelve member-states immigration and custom controls were to be abolished, so that the European Union would truly be single domestic market, open to the free movement of goods, services and people (Doganis R., 2001). In April 2004, a new package was enforced. Single European Sky (SES) is an example of EU s reregulation of aviation policy as it aims to rationalise the costs and emissions along with the improvement of air safety and it involves currently 38 countries. (Kawagoe, 2008) As Good, Röller, & Sickles (1995) had foreseen, full liberalization can hardly be expected within the next few years, air carriers are currently feeling the impact of a more competitive environment. Several carriers are going through strategic evaluations of their competitive position, not only vis-à-vis other European competitors, but also vis-a-vis the US carriers. Indeed, air transportation market in Europe has become very competitive with constant entry and exit of players, mergers and managerial revolutions. 1.2 THE BACKGROUND OF NO-FRILLS AIRLINES One would think that the impact of deregulation on air transport market and consequent evolution of airlines business models such as alliances or e-commerce, would also be the natural cause of the lowcost revolution. On the contrary, Southwest Airlines first introduced the low-cost, non-frills business model in 1967 in the United States. By 1978, when the US deregulation act came, Southwest was well placed to expand beyond its home base in Texas. Southwest, as many European low-cost airlines did later, concentrated its strategy on operating short-haul distances at low and unrestricted fares, with high point-to-point frequencies and excellent departure punctuality. It achieved so by offering high frequency, scheduled, point-to-point, short-haul services at low simple fares. In order to be able to offer low fares, it operated a single aircraft type with high-density seating and aim at high daily utilisation by reducing turnaround times to thirty minutes or less. Using less congested and secondary airports, reduced airport related costs were possible to achieve, and facilitated short turnarounds and higher punctuality. All traditional scheduled frills such as free in-flight meals, pre-assigned seats and connecting flights were cut back. Some airlines went even further by completely cutting out travel agents commissions and only selling directly to their customers. Furthermore, Southwest betted on an intensive marketing strategy where flying is fun, and to do so, the key factor was flexible and highly motivated staff. (Doganis R., 2001). Launched in 1985, Ryanair was the first low-cost, no frills European airline. However, it was not profitable and by 1991, it had accumulated big losses. In 1992, following the Southwest Model, Michael O Leary s Ryanair recorded a small pre-tax profit and was to become responsible for the radical change in European air transport. 10

12 1-1: EUROPEAN SEAT CAPACITY BY SERVICE TYPE BETWEEN 2002 AND 2006 AND ANNUAL VARIATION (SOURCE: RDC, 2007) Low Cost Carriers (LCC) have come a long way since Ryanair broke the mould of conventional European airlines in the early 1990s. Between 1994 and 2002, the European low-cost market has grown from around 3 million passengers to over 20 million annual passengers (RDC, 2002). According to the last RDC Low Cost Monitor report available at this time, by 2006 the European low-cost seat capacity was already over 178 million passengers. (RDC, 2007) Portugal is not an exception and it has been registering a continuous and steady growth of no-frills airlines traffic. The change in passenger traffic structure at Portuguese airports between 2004 and 2009 highlights the exponential growth of low-cost traffic as depicted in figure 1-2. Growing on average 35.1% per annum, the low cost segment has conquered 21,6 % of market share in 5 years to traditional companies and charter companies. (ANA, 2009b) 1-2: EVOLUTION OF PASSENGER TRAFFIC STRUCTURE IN PORTUGUESE AIRPORTS (SOURCE: ANA, 2009) 100% 80% 60% 74,6% 61,9% Traditional 40% LCC 20% 0% 8,6% 30,2% 16,9% 8,0% Charter 11

13 The increase of LCC operating in the European market has lead to complaints demanding lower airport fees. Nonetheless, the monopoly position of the Portuguese airports and the centralized management by the Ministry of transports, make them not open to lower the fees and attracting more traffic. Moreover, the government has a national airline, which is protected with preferential fees, by the main Portuguese airports. A complaint from a private airline at European Union resulted in the condemnation of the Portuguese airports (Barros C. P., 2008). Moreover, rigid masterplans with lifespan of 30 years or more are usually the cornerstone of Portuguese airport infrastructure s building and expansion policies, which frequently did not forecast the current trends in civil aviation. Therefore, the bigger question aimed to be answered is to which extent are Portuguese airports efficient in leading with the expansion of LCC. As the air transport market continues to develop, the efficiency evaluation of European airports is subject of many studies, and it is our intention to contribute to that research as thoroughly as possible by trying to understand how LCC are influencing Portuguese airports efficiency. 12

14 2 LITERATURE REVIEW The present chapter consists mostly in the literature review of the different methodologies used on efficiency research. Although some of the techniques used go back to the 50 s, the amount of literature in this field has grown enormously in the past years, mostly due to the severe changes that the worldwide aeronautical market has experienced in the last quarter of the 20 th century as mentioned previously. To perform the intended efficiency analysis, it is crucial to start by understanding what does efficiency mean and which are the possible resources to use. Therefore, this literature review will start by the first steps given towards efficiency measurement and followed by the contributions made by the scientific community. Afterwards, the methodological framework for airport efficiency benchmarking is presented, and finally, the literature on airport efficiency analysis is given. 2.1 FIRST STEPS ON EFFICIENCY RESEARCH: FARRELL S CONTRIBUTION Firstly, it is important to distinguish efficiency from productivity. This is important in the sense that the terms efficiency and productivity are often used interchangeably, even though the underlying meanings of these two terms are not identical. According to (Fried, Lovell, & Schmidt, 2008) producer s productivity is defined as the ratio of its output to its input. This ratio is easy to calculate if the producer uses a single input to produce a single output. In the more likely event that the producer uses several inputs to produce several outputs, the outputs in the numerator must be aggregated in some economically sensible fashion, as must the inputs in the denominator, so that productivity remains the ratio of two scalars. On the other hand, by the efficiency of a producer, we have in mind a comparison between observed and optimal values of its output and input. The exercise can involve comparing observed output to maximum potential output obtainable from the input, or comparing observed input to minimum potential input required to produce the output, or some combination of the two. In these two comparisons, the optimum is defined in terms of production possibilities, and efficiency is technical. It is also possible to define the optimum in terms of the behavioural goal of the producer. In this event, efficiency is measured by comparing observed and optimum cost, revenue, profit, or whatever goal the producer is assumed to pursue, subject, of course, to any appropriate constraints on quantities and prices. In these comparisons, the optimum is expressed in value terms, and efficiency is economic. Immediately after World War II, for obvious reasons, there was a general interest in growth and productivity. One of the most influential papers on these issues was by Robert M. Solow (Solow, 1957) within a macroeconomic setting. At the same time, Farrell laid the foundation for new approaches to efficiency and productivity studies at the micro level, involving new insights on two issues: how to define 13

15 efficiency and productivity, and how to calculate the benchmark technology and the efficiency measures. (Førsund & Sarafoglou, 2002) According to Farrell, the problem of measuring the productive efficiency of an industry is important to both the economic theorist and the economic policy maker. If the theoretical arguments as to the relative efficiency of different economic systems are to be subjected to empirical testing, it is essential to be able to make some actual measurements of efficiency. Equally, if economic planning is to concern itself with particular industries, it is important to know how far a given industry can be expected to increase its output by simply increasing its efficiency, without absorbing further resources. (Farrell, 1957) EFFICIENCY MEASUREMENT: SIMPLE CASE The fundamental assumption present in his paper was the possibility of inefficient operations. Opposing the mainstream economic literature of the time that suggested the average performance as the production function, Farrell s concept of frontier production function as benchmark was pioneer. To that respect, the new aspect was to decompose efficiency at the micro level of a firm in three levels: technical, price and overall efficiency. Even though Farrell himself refers to the existence of similarities between Debreu s coefficient of resource utilization (Debreu, 1951) and his own measure of technical efficiency (TE), the former was able to create a much more relevant with the creation of a benchmarking tool in respect to firms operating in the same market whereas the latter worked mostly on the resource cost side, defining his coefficient as the ratio between minimized resource costs of obtaining a given consumption bundle and actual costs, for given prices and a proportional contraction of resources. (Førsund & Sarafoglou, 2002). Farrell s original diagrams, depicted in figure 2-1, illustrate the case of a firm producing a single output by employing two factors of production (x and y), while, for simplicity reasons, assuming constant returns to scale (CRS). The diagram on the left assumes to know the efficient production function, by which a perfectly efficient firm obtains the maximum output from any given combination of inputs. P represents the inputs of the two factor of production and the assumption of CRS allows representing the isoquant SS, where the firm might produce unit output efficiently with the various combinations of the two production factors. Understandably, Q represents an efficient firm using production factors in the same ratio as P (since OP and OQ have the same slope). Given so, one can say that the efficient firm produces OP/OQ times as much output from the same inputs, or in other words, Q produces the same output as P using only a fraction OQ/OP as much of each factor. Farrell defines OQ/OP as the technical efficiency of the firm P. However, one also needs a measure of the extent to which a firm uses the various factors of production in the best proportions, in view of their prices. (Farrell, 1957). By defining AA s slope equal to the ratio of the two production factors prices, we are able to find Q as the optimal method for production, 14

16 since that the production costs at Q will only be a fraction of OR/OQ of those at Q. Following the same line of reasoning above mentioned, the ratio OR/OQ is the price efficiency of the firm Q. Furthermore, if the firm P were to change the inputs proportions until the same defined in Q, while keeping the same technical efficiency and assuming that factor prices did not change, its costs would also be reduced by the same factor as Q, hence, OR/OQ is the price efficiency of the firm P too. Nevertheless, Farrell does call out attention to the impossibility to foresee whether TE remains constant with changes in proportions of inputs or not. Finally, if P were to be truly efficient, both technically and in respect of prices, its costs would be a fraction OR/OP of what in fact they are to produce the same output. Farrell calls this ratio the overall efficiency of the firm, and notes its equivalence to the product of technical and price efficiencies. 2-1: FARRELL'S TECHNICAL EFFICIENCY (LEFT) AND PIECEWISE PRODUCTION FRONTIER (RIGHT) (SOURCE: (FARRELL, 1957)) THE EFFICIENT PRODUCTION: SIMPLE CASE Farrell s efficiency measures, as mentioned above, were defined based on the assumption that the efficient production function is known, that is, each firm s performance observed is compared to some standard of perfect efficiency. Hence, the definition of the efficient production function is in need in order to understand the significance of the efficiency measures. Farrell starts by refusing the postulate standard of perfect efficiency that represents what is theoretically attainable in despite of an empirical function based on the best results observed in practice. Although the former is a reasonable and perhaps the best concept for the efficiency of a single production process, there are considerable objections to its application to anything so complex as a typical manufacturing firm, let alone an industry. If the measures are to be used as some sort of yardstick 15

17 for judging the success of individual firms or industries, it is far better to compare performances with the best actually achieved than with some unattainable ideal. (Farrell, 1957) For instance, as we shall see later in this discussion, the variety of ways to address airport s inputs (outsourcing of labour, services, etc), is such that it would be a process too complex, and consequently, less likely to obtain an accurate theoretical production function. Furthermore, being the airport industry an economy both of scale and scope, it is also more exposed to the inherent human imperfection. Hence, the theoretical function is likely to be widely optimistic. The problem then becomes how to estimate an efficient production function from the observed firms. Based on the same assumptions above mentioned, Farrell suggested that each firm could be represented by a point on an isoquant diagram, therefore obtaining the scatter of points in left diagram of figure 2-1. His contribution was to introduce a piecewise linear envelopment of the data as the most pessimistic specification of the frontier, in the sense of the function being as close to the observations as possible. (Førsund & Sarafoglou, 2002). Moreover, Farrell has had also the foresight to address the generalization to the case of multiple inputs and outputs, and analyzed the issue of increasing and diminishing returns to scale. Although his method was perfectly valid for the case of diseconomies of scale, his primary assumption of convexity would not hold for the opposite situation. Meaning that the method will give an optimistic instead of a conservative estimate of S (some straight line like OR in figure 2-2) and, of course a pessimistic estimate of the efficiency of any point. (Farrell, 1957). Finally, he also approached, although very roughly, the parametric method, by approximating his methodology to the Cobb-Douglas production function. 2-2: FARRELL'S ORIGINAL DIAGRAMS ON INCREASING AND DIMINISHING RETURNS TO SCALE (SOURCE: (FARRELL, 1957)) 16

18 Nonetheless, the first step has been given, and it was indubitably a very important one. As Farrell himself suggested, this paper would be of interest to a wide range of economic statisticians, business men and civil servants, and therefore the object of research for many years to come. 2.2 THE DEVELOPMENT OF FARRELL S IDEAS Farrell s efficiency concepts are still the basic definitions in use today. The estimation methods for both the non-parametric and parametric frontier introduced by Farrell are the foundation for later contributions. (Førsund & Sarafoglou, 2005). Both importantly, the broad range of applications and limitations of the piecewise efficiency frontier proposed by Farrell, gave rise to different research fields to contribute with further investigation. In this section, estimation processes and linear programming (LP) developments will be addressed, given that, they turned out to be today s most influential efficiency measurement tools ESTIMATION PROCESSES FOR PARAMETRIC FRONTIERS Regarding the developments in the estimation processes for parametric frontiers, Førsund & Sarafoglou (2005) proposes three main methodologies. Chronologically, C. B. Winsten was the first when, in the 1957 s Discussion on Mr. Farrell, wondered if the efficient production function turned out to be parallel to the average production function, and whether it might not be possible to fit a line to the averages, and then to shift it parallel to itself to estimate the efficient production function. Nevertheless, only seventeen years later, J. Richmond s Estimating the Efficiency of Production defines this approach as Corrected Ordinary Least Squares with any references neither to Farrell nor to Farrell s Discussion. In 1968, the Cobb-Douglas production function was used by (Aigner & Chu, 1968) as a benchmark for the efficiency frontier estimation, using linear and quadratic programming to calculate the frontier, as a deterministic parametric approach. (Afriat, 1972) was the next milestone. He elaborated further ideas from Farrell and the later discussion of Farrell s paper. A statistical framework was formulated for finding maximum likelihood estimators for the parameters of frontier functions, leaving the pure programming format. Nonetheless, according to (Førsund & Sarafoglou, 2005), (Afriat, 1972) also contributed within the non-parametric framework of piecewise linear frontier functions by formulating the model with variable returns to scale (in the single output case). This was later referred to as the BCC model (Banker, Charnes, & Cooper, 1984) by the group of the operations research and management science community. Finally, and simultaneously in 1977, (Aigner, Lovell, & Schmidt, 1977) and (Meeusen & van den Broeck, 1977) opened the door for more rigorous econometric analysis of frontier functions by introducing the composed error term in parametric models. This methodology, named by the former as stochastic 17

19 frontier function, suggested that the composed error consisted of two parts, a stochastic component symmetrically distributed, catching white noise and a stochastic component with a one-sided distribution, representing inefficiency. Moreover, within the single-output model the connection between the one-sided term and the Farrell measure of (output-oriented) technical efficiency was direct. (Førsund & Sarafoglou, 2002). This methodology is today known as Stochastic Frontier Analysis (SFA). Once again, the tribute to Farrell is paid. It has only been since the pioneering work of Farrell (1957) that serious consideration has been given to the possibility of estimating so-called frontier production functions. (Aigner, Lovell, & Schmidt, 1977) LINEAR PROGRAMMING METHODS In 1967, J.N. Boles developed an explicit linear programming model used by Berkeley agricultural economists to improve Farrell estimation method. Their efforts, however, failed to become acknowledgeable. Nevertheless, this methodology was to become a centrepiece of later important work. It was only one decade later, when in 1978, Abraham Charnes, William. W. Cooper and Edwardo Rhodes published the highly influential paper Measuring the efficiency of decision making units, based on Farrell s concept of efficiency measurement. According to (Førsund & Sarafoglou, 2002), the model was readily computable, either using standard linear programming codes on mainframes or developing more efficient tailor-made software. However, the linear programming model is identical to one of the models in Boles (1971). A very important contribution by (Charnes, Cooper, & Rhodes, 1978) was the explicit connection made between a productivity index (in the form of weighted sum of outputs on a weighted sum of inputs) and Farrell s technical efficiency measure. Although some may dispute that Farrell s ingenious concept of efficiency measure was put in simple terms for better comprehension, and therefore omitted heavier economic terminology, CCR offered the bridged between the engineering concepts of micro productivity ratios and economists concept of efficiency. (Førsund & Sarafoglou, 2002) This methodology would become later known as Data Envelopment Analysis (DEA), and, it is nowadays, the most used methodology in airports efficiency measurement. The 70 s witnessed major developments in estimation methods, all based on Farrell s ideas, spreading worldwide his ideas. However, all methodological developments were related with statistical inference, hence parametric methods, and apart from Berkeley agricultural economists, economic statisticians did not follow the non-parametric method. Nonetheless, Charnes, Cooper & Rhodes s paper was the one that allowed contributions from operations research and management sciences for further developments. Whereas the parametric approach is usually associated with statistical inference, the nonparametric has been labelled of deterministic. On the turn to the 21 st century new research has been 18

20 conducted, trying to merge both methodologies, by making statistical inference possible for the later. See, (Banker R. D., 1993), (Simar & Wilson, 2000a) and (Simar & Wilson, 1998) 2.3 METHODOLOGICAL FRAMEWORK ON AIRPORT BENCHMARKING In this section, a description of the most relevant methodologies used in airport s performance benchmarking is undertaken, followed by an inter-comparison between them. Regarding the methodological framework classification of performance benchmarking, (Tovar & Martin-Cejas, 2010) for example, suggests the division of the different approaches into two groups. On one hand, a traditional group that ignores inefficiency, hence considering that the observed output is always the best output. On the other hand, when inefficiency exists, a frontier approach where a best practice frontier, by which each firm is to be compared, has to be estimated. In both groups, parametric and non-parametric approaches exist. In the frontier group, SFA and DEA are, respectively, examples of parametric and non-parametric approaches, whereas in the traditional group, index numbers and ordinary least squares (OLS) are, respectively, examples of non-parametric and parametric approaches. Historically, it is also worthwhile mentioning a third group, which involves the use performance indicators in a partial approach. Currently, numerous methodologies allow the comparison of different firms performance in order to depict the best practice of that industry. Nonetheless, each one of the available methodologies on performance benchmarking has its pros and cons, and much is still to be done in order to overcome those drawbacks PARTIAL MEASURES This method uses partial ratio data to carry out performance comparison of target sample in single dimension, such as on financial and cost performance. (Lai, Potter, & Beynon, 2010). For instance, (Francis, Humphreys, & Fry, 2002) reveals that, comparative performance of airports amounted to the collection and comparison of financial and output measures by Governments, who typically owned and operated the majority of airports. Profiting from the airline s definition of workload unit (WLU) as one passenger processed or 100 kg of freight handled, airports adopted this measure during the 1980 s to provide a single measure of output for passenger or freight service. Typical partial measures used included: total cost per WLU; operating cost per WLU; labour cost per WLU; WLU per employee; total revenue per WLU and aeronautical revenue per WLU. According to the (Civil Aviation Authority, 2000), this method can provide useful insights into particular areas of inefficiency. However, when taken alone, it can also provide a distorted picture of performance by ignoring the interaction between inputs used and outputs produced. These partial productivity measures have been criticized on the ground that they: 19

21 do not reflect differences in factor prices; do not take into account of possible factor substitution in production; fail to take account of the differences in operating environments between firms; are unable to handle multiple outputs. Alternatively, to obtain an overall picture of performance it may be used index number techniques, as discussed hereafter TOTAL FACTOR PRODUCTIVITY In essence, the total factor productivity (TFP) method compares an index of outputs to an index of inputs, allowing comparisons of the same airport in different periods or between different airports. The weights in constructing the indexes can be the revenue shares and cost shares as indicators of the importance of outputs and inputs in the production process. (Hooper & Hensher, 1997). TFP is then calculated as the ratio of aggregated output to aggregated input. Although this measure does not suffer from the shortcomings of partial measures, as an aggregate measure it has no sufficient information for evaluating management strategies when taken alone. Furthermore, it requires information on prices, which are used to aggregate inputs and outputs, which may not be readily available. (Civil Aviation Authority, 2000). It is important to emphasize that there are a number of different ways in which TFP can be assessed. The above mentioned, is also called of non-parametric as it requires only the aggregation of all outputs into a weighted outputs index and all inputs into a weighted input index with no assumptions or estimates of the parameters of the underlying production or cost function having to be made. According to (Graham A., 2008), one of the most comprehensive studies of TFP in the airport sector is contained in the Global Airport Benchmarking Report, produced annually by the Air Transport Research Society. The methodology uses revenue shares as weights for the outputs (aircraft movements, passengers, cargo) and cost shares as weights for the inputs (labour, runways, terminals, gates) where capital input was excluded due the difficulties in accurate and comparable data. Conversely, in the parametric approach, a production or cost function is estimated by using either regression analysis or a stochastic frontier method. These models can be used to investigate factor substitution, the impact of variations of input and output prices and to test for economies of scale. However, because this approach has detailed data requirements, it has not been used very often for airports. A final approach is the endogenous-weight TFP method, where detailed cost and revenue data is deferred to a multi-input/multi-output production function instead. Then, physical and financial input and output data is used to estimate the TFP. Yoshida Y. (2004) applied this method to Japanese airports. (Graham A., 2008) 20

22 2-3: ATRS TFP FOR WORLD AIRPORTS (SOURCE: (GRAHAM A., 2008)) MALMQUIST INDEX OF PRODUCTIVITY CHANGE The Malmquist Index (MI) is a bilateral index that can be used to compare the production technology of two economies. To calculate the MI of economy A with respect to economy B, we must substitute the labour and capital inputs of economy A into the production function of B, and vice versa. (Wikipedia, 2010) The CAA (Civil Aviation Authority, 2000) recognizes some advantages of MI over TFP in that: price data are not required; assumption of cost minimization or revenue maximization is not needed; the MI obtained may be further decomposed into: o technical efficiency change (firms getting closer to the frontier); o technical change (shifts in the frontier itself). Hence, MI uses distance functions to describe a multi-input, multi-output production technology without the need to specify a behavioural objective. It measures the productivity change between two data points by calculating the ratio of the distances of each data point relative to a common technology. In fact, MI is the geometric mean of two TFP indices: one is evaluated with respect to base period s technology and the other one with respect to next period t technology. A value of MI bigger than one implies a positive TFP growth from period s to period t. In practice, four input-oriented distance measures must be calculated for each firm in each pair of adjacent periods by solving four (CCR) DEA-like linear programming problems. Similar to DEA approach, this can be further extended by decomposing the 21

23 technical efficiency change into scale efficiency and pure technical efficiency components by adding the convexity restriction DATA ENVELOPMENT ANALYSIS (DEA) Non-parametric frontier methodologies use the panel data to establish best-practice frontiers. They are parameter-free because no assumptions on production or cost functions are made, and the efficiency measurements are based on the comparison among the different firms. Based on (Farrell, 1957), and coined by (Charnes, Cooper, & Rhodes, 1978), the DEA methodology is a non-parametric approach that uses linear programming to construct a piecewise linear efficient frontier that envelops the data based on information of inputs and outputs only. Efficiency measures are then calculated relative to this frontier. DEA measures relative efficiency by comparing the efficiency of a decision-making unit (DMU) with the efficiency of other DMUs (or peer groups) that have a similar mix of inputs. This ensures that like-with-like comparisons are made. The peer group is defined such that a linear combination of these DMUs can be shown to have at least as great an output as the target DMU given the similar input mix and operating environment. (Civil Aviation Authority, 2000). Farrell s efficiency frontier, as depicted previously in figure 2-2, provides a more easily understanding example of an input minimisation DEA model under the CRS assumption. This methodology can consider a variety of models, such as: Constant returns to scale (CRS), also known as the CCR-DEA model, after (Charnes, Cooper, & Rhodes, 1978), ; Variable returns to scale (VRS), also known as BCC-DEA model, after (Banker, Charnes, & Cooper, 1984); Application of Malmquist DEA method to panel data to calculate indices of total factors of productivity change. There is some resemblance with the non-parametric TFP as this method also produces a weighted output index relative to a weighted input index. The key advantage is that not only DEA does not involve the estimation of underlying production or cost functions, but also the weights for inputs and outputs are not predetermined, but instead the result of the programming procedure. According to (Graham A., 2008), the DEA is therefore a more attractive technique for dealing with multiple input and output activities than the index number TFP because it has less demanding data requirements. Furthermore, this method measures the efficiency of DMUs engaged in performing the same function, therefore with regard to the sample instead of absolute terms. The most efficient DMUs or firms, will be located on the efficiency frontier with relative index of 1,0. Although DEA produces relative rankings, it does not explain by itself the observations. This can be partially overcome with the application of the MI, which when used with DEA is a useful way of identifying the sources of productivity differences 22

24 over a certain period. This happens because this index allows productivity change to be decomposed into technical changes gained from adopting new technologies and efficiency changes (which in turn can be decomposed into gains from utilizing scale and reducing inefficiency). It is also worth mentioning the fact the DEA methodology can be either input or output oriented. In other words, the former is desired if the objective is to produce the existing level of output with at least one less unit of input, whilst in the later more output must be produced for the same existing inputs, so that efficiency gains are attained. As mentioned above, as DEA allows for both CRS and VRS, it can also be used for measuring scale effects on airports. In the BCC-DEA model, technical efficiency can be further decomposed into scale efficiency and pure technical efficiency. Scale efficiency being defined relatively to the ratio of TE in the CCR-DEA model for TE in the BCC-DEA model. Figure 2-4 depicts clearly this fact. 2-4: SCALE EFFICIENCY IN THE BCC-DEA MODEL (SOURCE: (CIVIL AVIATION AUTHORITY, 2000)) However, since it is not statistically based, any deviation from the frontier is entirely attributable to inefficiency, making DEA more prone to data measurement error and outlying data problem than parametric approaches. (Civil Aviation Authority, 2000). On the other hand, the paper recently published by (Simar & Wilson, 2007), result of continuous work on the last decade on methods for bootstrapping non-parametric approaches, has lead to a two-stage DEA, which allows for statistical inference. 23

25 Other criticisms pointed to DEA are related with the tendency to overstate performance when the combined number of outputs and inputs is large relative to the number of DMUs, the difficulty to analyse productive efficiency changes over time and any deviation, which is sensitive to the presence of outliers, is labelled as inefficient. An alternative to DEA, is the free disposable hull methodology (FDH). It is a special case of the DEA model, where the points on lines connecting the DEA vertices are not included in the frontier. Because the FDH is interior to the DEA frontier, the estimates of average efficiency are larger than in DEA. FDH, like in DEA, assumes no assumption regarding the function form and do not require random error. However, to the best of my knowledge, there is no literature in airport performance benchmarking using this last methodology, thus, should not be considered as a relevant methodology for our case study STOCHASTIC FRONTIER ANALYSIS (SFA) Contrarily to the non-parametric methodologies, parametric approaches not only specify functional form, but also take account of the residual term in the analysis. In other words, it provides not only a measurement of efficiency, but can also be used as an explanation for inefficiency. SFA originated with two papers, published nearly simultaneously by two teams on two continents. (Meeusen & van den Broeck, 1977) appeared in June, and (Aigner, Lovell, & Schmidt, 1977) appeared one month later. The two papers are themselves very similar, since both were three years in the making and appeared shortly before a third SFA paper. (Kumbhakar & Lovell, 2000) Stochastic frontiers differ from the parametric approach in that they can accommodate data noise and statistical tests. The SFA methodology specifies a function form of the cost, profit or production relationship among inputs, outputs and environmental factors and allows for random errors. Hence, it is also called econometric approach. It gives a composed error model where inefficiencies are assumed to follow an asymmetric distribution, while random errors follow a symmetric distribution. The logic behind is that the inefficiencies must have a truncated distribution because inefficiencies cannot be negative. If inefficiency is bigger than zero, than the composed error term is negatively skewed, and there is evidence of technical (in)efficiency (TE). In order to obtain an estimate of the TE of each producer, this requires strong distributional assumptions in order to decompose the two error components. However, according to (Civil Aviation Authority, 2000), in many practical cases, there may not be sufficient data available to support these assumptions. In despite of the strong distributional assumptions, the success of SFA also depends critically on the functional form used to describe the relationship between the inputs and outputs, the variables included in or excluded from the model, and potential problems of correlation between two or more DMUs (multicollinearity) or correlation between DMUs and 24

26 the error term (endogeneity). However, being a statistically based approach, some statistical diagnostics could be used to gauge the appropriateness of the SFA model specification and assumptions if sufficient data is available. Therefore, the major criticisms made to SFA methodology are related with the fact that it requires behavioural assumptions (such as cost minimisation or revenue maximisation), has the need for clear definition of inputs and outputs to avoid endogeneity problem, is prone to misspecification of the functional forms (of both technology and inefficiency) and is subject to aggregation and multicollinearity problems. TABLE 2-1: MAJOR METHODOLOGIES IN AIRPORT PERFORMANCE ANALYSIS (ADAPTED FROM (LAI, POTTER, & BEYNON, 2010)) Partial measures Index Numbers Frontier analysis Parametric approach Non- Parametric approach Methodology Description This method uses partial ratio data to carry out performance comparison of target sample in single dimension such as on financial and cost performance of an airport. Total Factor Productivity (TFP) allows for measuring cost efficiency and effectiveness and for distinguishing productivity differences in airport performance. This technique can also be used for investigating the impact of variations of input and output price on an airport s performance. Stochastic Frontier Analysis (SFA) Sometimes referred to as econometric frontier approach, is one of the main parametric approaches used by researchers to evaluate efficiency. Data Envelopment Analysis (DEA) Is a nonparametric approach, which requires no assumptions about the functional form and calculates a maximal performance measure for each firm relative to all other firms. Weakness This method only focuses on certain fields of airport performance. The evaluation result of this method would not be able to provide a more comprehensive evaluation of an airport s performance. TFP requires an aggregation of all outputs into a weighted output index and all inputs into a weighted input index using predefined weights, which can be biased. Although the parametric approaches take into account the effect error, the parametric methods still faces challenges on separating random error from efficiency. The key drawback of the technique is that it does not allow for random error in the data, assuming away measurement error and luck as factors affecting outcome, which implies that the measured inefficiency is likely to be overstated. Additional parametric approaches are the distribution free analysis and the thick frontier approach. Such as the free disposable hull methodology in the non-parametric approach, neither one of these parametric approaches are seldom used in airport performance analysis, and for such reason, they will be briefly addressed. The former specifies a function form for the frontier approach, but separates the inefficiencies from the random error in a different way. It assumes that the efficiency of each firm is stable over time, whereas random error tends to average out to zero over time. The estimate of efficiency for each firm is determined as the difference between its average residual and the average residual of the firm on the frontier. 25

27 The later provides efficiency measure for the overall firms and not for individual firms. It specifies a functional form and assumes that deviations from predicted performance values within the highest and the lowest quartiles of observations represent random error, while deviations in predicted performance between the highest and the lowest quartiles represent inefficiencies. This approach does not intrude assumptions regarding distributions on either inefficiencies or random error. Above in table 2-1, we can see the description of major methodologies used in airport performance analysis and their main weaknesses. Each one of the previously presented has their own advantages and drawbacks, covers various aspects of performance and requires different data and different assumptions. Hence, the chosen methodology must be subject to a careful analysis so that one takes the best option possible within the methodology framework. 2.4 LITERATURE ON AIRPORT EFFICIENCY RESEARCH It is consensual that management sciences and operational research have provided several useful methodologies necessary for the efficiency analysis of industries. In this section, it will be addressed the motivation of literature on airport efficiency research in order to understand the existing methodological framework. Airport performance measures are important for day-to-day business and operational management, regulatory bodies, Government and other stakeholders such as passengers and airlines. According to (Humphreys & Francis, 2002), complex and dynamic organisations such as international airports provide a challenge in establishing an appropriate performance measurement system. The many interacting parts of airports; passengers, airlines, handling agents and surface transport service providers, in addition to the interests of the regional and national economy, complicate the development of performance measurement systems. Performance measurement is a critical management activity, both at the operational level of the individual airport and at the wider system level. Doganis (1992) proposes that there are several reasons why airport managers and governments measure airport performance: to measure efficiency from a financial and an operational perspective, to evaluate alternative investment strategies, to monitor airport activity from a safety perspective and to monitor environmental impact. Hence, given the present presence of various stakeholders in the airport infrastructure, performance analysis serves necessarily many purposes. Thus, each stakeholder will measure airport performance for different reasons. Whereas governments may use this data for environmental and economic regulation, airport managers along with owners and/or shareholders will privilege data related with business performance and return on the investment. Also very importantly, airlines will use this data so they can compare cost performance across airports. 26

28 Humphreys & Francis (2002) emphasizes that it is also important to recognise that airlines are the key customers of airports and that the airlines act as an intermediary between the airport and passengers or freight shippers. In addition to this fact, if we underline the increasing role of LCCs in today s air transportation market, we can easily infer that in a very near future, they will inevitably be responsible for a large share of transported passengers. Thus, airports must prepare themselves to be also efficient in a low-cost perspective, in order to attract those same companies. Most studies made in airport efficiency research have often used the methodologies described in the previous section. For instance, (Gillen & Lall, 1997) examined the performance of 23 US airports using DEA, whereas (Hooper & Hensher, 1997) have used TFP to assess the economic performance of six Australian airports. Econometric approaches have also been used in airport performance, as did (Oum, Yan, & Yu, 2008) using SFA for assessing the influence of the ownership forms in airport s efficiency. Similarly, but using variable factor productivity indexes (Oum, Adler, & Yu, 2006) conducted research in that area. However, according to (Lai, Potter, & Beynon, 2010), most airport efficiency measurement research fails to consider other important factors that can influence an airport s performance evaluation, such as the characteristics of airport authorities and airport users (airline companies or passengers). To the best extent of my knowledge, no studies on airport performance have been carried out regarding the LCCs perspective. That is to say that most benchmarking studies made have traditionally introduced input and output variables that account for the overall airport performance, not discriminating the contribution of traditional airline operators from that of LCCs. Separating for instance, traditional passengers from LCC s passengers, check-in desks from automatic check-in kiosks, and finger-bridged gates from walk-by or bus driven boarding gates. Airport s economic performance has thus become an important task for those involved, directly or indirectly, with the airport industry. The liberalization of air transport market, and consequently, airport commercialization and privatization, has marked interest in performance comparisons. Furthermore, consumer satisfaction levels and increasing concern on environmental impacts requires benchmarking techniques that are important to assess the state of the art in airports management. Nonetheless, according to (Graham A., 2005), there are still a wide range of operational activities which need to be monitored by looking at measures related to airside delays, baggage delivery, terminal processing times, equipment availability and so on. A summary of airport performance studies was compiled below, in table 2-2. It is further detailed with the respective input and output variables in Appendix 1. It gathers most of the work conducted on airport performance, involving several aspects in respect to what may influence that same performance. Despite most of the existing work is based in the comparison of airports within the same country, international benchmarking has also been subject to research studies. 27

29 TABLE 2-2: AIRPORT EFFICIENCY PAPERS (ADAPTED FROM (GRAHAM A., 2008) AND (LAI, POTTER, & BEYNON, 2010)) Authors Year Methodology Coverage Tololari 1989 Parametric TFP BAA UK airports Prices Surveillance Authority 1993 Index number TFP 6 Australian airports Gillen and Lall 1997 DEA 23 US airports Hooper and Hensher 1997 Index number TFP 6 Australian airports Graham and Holvad 1997 DEA 25 European + 12 Australian airports Parker 1999 DEA BAA and 16 other UK airports Murillo-Melchor 1999 DEA/Malmquist index 33 Spanish airports Salazar de la Gruz 1999 DEA 16 Spanish airports Jessop 1999 DEA/Multi-attribute assessment 32 major international airports Nyshadham and Rao 2000 Index number TFP 25 European airports Sarkis 2000 DEA 44 US airports Pels et al DEA/Parametric TFP 34 European airports Gillen and Lall 2001 DEA/Malmquist index 22 US airports Martin and Roman 2001 DEA 37 Spanish airports Abbott and Wu 2002 DEA/Malmquist index 12 Australian airports Martin-Gejas 2002 Parametric TFP 40 Spanish airports Pacheco and Fernandes 2002 DEA 33 Brazilian airports Bazargan and Vasigh 2003 DEA 45 US airports Holvad and Graham 2003 DEA 21 UK airports Oum et al DEA 50 major airports around the world Oum and Yu 2004 VFP 76 major airports around the world Barros and Sampaio 2004 DEA 13 Portuguese airports Sarkis and Talluri 2004 DEA 44 US airports Yoshida 2004 Endogenous weight TFP 30 Japanese airports Yoshida 2004 DEA/Endogenous weight TFP 67 Japanese airports Yu 2004 DEA 14 Taiwan airports Hanaoka and Phomma 2004 DEA 12 Thai airports Kamp and Niemeier 2005 DEA/Malmquist index 17 European airports Vogel 2006 DEA 35 European airports Lin and Hong 2006 DEA 20 major airports around the world Martin and Roman 2006 DEA 34 Spanish airports Vasigh and Gorjidooz 2006 Index number TFP 22 US and European airports Barros and Dieke 2007 DEA 31 Italian airports Fung et al DEA/Malmquist index 25 Chinese airports Barros 2008 DEA 31 Argentina airports Barros 2008 SFA 27 UK airports Barros and Dieke 2008 Two-Stages DEA 31 Italian airports Yu et al DEA 4 Thai airports Oum et al SFA 109 major airports around the world Barros 2009 Random SPA model 27 UK airports Chi-Lok and Zhang 2009 DEA 25 Chinese airports Martin et al MC Monte Carlo Simulation/SFA 37 Spanish airports Lan et al DEA 11 major airports in Asia Pacific 28

30 Tovar et al SFA/Malmquist TFP index 26 Spanish airports Nonetheless, producing meaningful inter-airport performance indicators is fraught with difficulties of serious problems of comparability, particularly due to the varying range of activities undertaken by airport operators themselves. According to (Graham A., 2005), comparing indicators from raw data can give misleading impressions as airports involved with more activities would inevitable have higher cost and revenues levels and poorer labour productivity. The fundamental difficulties associated with inter-airport comparisons (particularly from different countries) and with dealing with problems of comparability, arising largely from the diversity of inputs and outputs, still remain and have yet to be resolved effectively. Relatively few benchmarking studies have made a truly international comparison of performance. This seems to be out of line with the fact that both the airport and airline industry are becoming increasingly international or global in nature. Further research is needed. Interest in this area will undoubtedly increase with more of the industry being expected to go through the commercialisation and privatisation stages in the evolutionary cycle of the airport industry. Other organisations, such as regulatory authorities, may also help to improve the current practices in this area. Similarly, comparability difficulties are also expected to be found in our case study. Scale differences between Portuguese airports are evident. For instance, Lisbon airport registered over 13 million passengers in 2009, Porto and Faro had 4,5 and 5 million passengers respectively, and Funchal a bit more than 2 million (ANA, 2009b). Additionally, Aeroportos de Portugal, SA (hereafter ANA), makes mainland airports and Azores islands airports financial reports whereas Aeroportos e Navegação Aérea da Madeira, S.A (ANAM) is the responsible public entity for the accounting of Madeira s airports, which may implicate different methodologies in the financial reports. Discussed in more detail later, ANA has a participation of 70% on ANAM, but accounting differences may still happen. Taking for instance, in terms of outsourcing labour instead of direct labour, or hypothetical revenues from concessions put in commercial revenues instead of aviation revenues. Although performance evaluations may be undermined, to some extent, by data availability, the performance assessment of a country s internal market will always be easier than international benchmarking. Thus, difficulties in data interpretation are expected. Nonetheless, we will also rely on the considerable amount of literature to help us perform as unbiased as possible the evaluation on how LCCs affect Portuguese airports efficiency. 29

31 3 MAIN DRIVERS OF AIRPORT EFFICIENCY Airport production of outputs (passengers and freight) relates with several inputs that are not as concrete as in other industries. Moreover, the production process in airports is segmented by different operations that further complicate the assessment of airport s efficiency. From the air transport movements (ATM) in the airside of the airport, passing by the processing stage of passengers, freight and luggage on the landside, to the intrinsic relationship between airports and airlines, there are many details that influence the overall performance of an airport. Furthermore, the technical design of any airport must be aligned with the political objective drawn by decision-makers regarding the role of that same airport. In other words, a regional airport aimed to boost local economy is not expected to have the same infrastructures as a major international hub. In addition, technological innovation and new business models tend to shape or alter travel patterns that were not initially foreseen. These factors combined can result in the misuse of resources and consequently the inoperability of an infrastructure so expensive such an airport. Despite the several and unpredictable difficulties that arise within the air transport market, it is the airport manager responsibility to explore its assets in the most efficient way. In this chapter, it will be addressed the factors that have most impact on airports efficiency, namely on the airside, landside, and within the airport-airline relationship. Moreover, LCCs are becoming significant factors in airport planning. As their requirements differ from those of traditional carriers, they drive the development of secondary airports and cheaper passenger buildings (de Neufville R., 2008). Hence, the recognition of the different stakeholders in airport performance is of vital importance in order to achieve higher efficiency. 3.1 AIRSIDE The airside of an airport is constituted by all areas accessible to aircraft, including runways, taxiways, aprons and parking places. Airfield efficiency relates to the core business of airports that is the transport of people and goods. Airfield design follows international standard practices, namely the Annex 14 of the International Civil Aviation Organization (ICAO) (ICAO, 2004) or the Advisory Circular 150/ of the US Federal Aviation Administration (FAA) (FAA, 1989). Nevertheless, designers often tend to be misinformed about the economic implications of extra minutes in an aircraft operation for each hundreds of thousands of movements per year. For instance, an efficient airport nowadays must be aware of how crucial quick turnaround times are for LCCs and how delays affect the airport s economic performance by reducing the number of aircrafts allowed per day. The importance of efficient airside operations is thus motive of study in the following sections. 30

32 3.1.1 AIRFIELD DESIGN According to (de Neufville & Odoni, 2003), airfields typically account for 80 to 95% of the total land area occupied by an airport and affect in critical ways every facet of airport operations. Main determinants of airfield size are number and orientation of runways, geometric configuration of the runway system, the most demanding type of aircraft the airport can serve (also defined as critical airplane ), land area for possible future expansion and or environmental mitigation. Moreover, airfield configuration also affects landside facilities and services, as the layout of runways has enormous influence in the geographical distribution of passenger and cargo terminals, hangars and other buildings. Finally, technical planners are also subject to environmental, political and economical constraints at each site. As mentioned above, there are international sets of standards regarding airfield design that explore thoroughly the wide range of characteristics in order to assure desired levels of safety, and should be consulted if curiosity arises. Nevertheless, factors such as wind coverage, runway orientation and/or geometry, or physical obstacles are not crucial in the assessment of airport s efficiency performance when using the methodologies presented in the previous chapter. In appendix A, it is stated clearly that most authors make use of inputs such as the number of runways, gates and parking spots, apron area among others to perform efficiency analysis of airport. More rarely, runway length is also used. Therefore, it is not the ambition of this section to describe exhaustibly the characteristics of the airfield design process. Instead, we aim to enhance those that contribute to the overall efficient performance of airside operations, namely by reviewing airport layouts and the influence of taxiways and apron areas. This review is preceded by the identification of major constraints of current practices in airfield design CURRENT APPROACH AND ALTERNATIVES The traditional approach of airport design bases on the mould of masterplans as defined by FAA s Advisory Circular 150/5070 or ICAO s Airport Planning Manual-Part 1, for instance. And despite the advice of these documents regarding the fact that planners should tailor an individual master plan to the unique conditions at the study airport (FAA, 2007), they are also quite rigid in the sense that one single forecast becomes the cornerstone of the entire master plan development. As all alternatives of airport development are based in one single forecast (typically a linear regression from historical data regarding the number of passengers), and forecasts are most likely to be wrong, a major infrastructure master plan is started without examination of further alternative scenarios. The problem with this approach lies on the failure to anticipate the risk of possible changes in market conditions and to provide insurance against those same risks. This is specially truth when we take for example the growth of LCCs and how they changed the modus-operandi in terms of passenger transport between aircrafts and the passenger building. Whereas traditional airlines use expensive bridges, LCCs prefer buses or having their costumers walking in order to cut costs. Master planning is 31

33 thus inflexible and unresponsive to risks and has potential serious consequences such as extra costs or losses of opportunities. Although there are alternatives to traditional master planning, old habits die hard. According to (de Neufville & Odoni, 2003), strategic planning refers to a disciplined process for analyzing the current situation of a business activity, and identifying the vision of how that entity should position itself with respect to its customers and competitors. SWOT (acronym for strength, weakness, opportunity and threats) analysis is a generic version of strategic planning. By properly positioning their organizations according to this analysis, airport operators become more flexible to respond to possible events. An analogy to chess is convenient, in the sense that airport operators should look ahead several moves, but deciding only one at a time. Airport strategic planning can be divided in three phases. Firstly, one should recognize risk and complexity. It is wiser to foresee a wider range of futures since forecasts tend to be always wrong. Furthermore, as the number of possible choices is so big, it is advisable to adopt hybrid (mixing single configurations) solutions that are able to adapt to different outcomes. Secondly, airport operators should analyse possible futures. Using SWOT analysis is easier to identify risks and possibilities of response to actual events. Thirdly, a dynamic strategic planning is in order. Different choices differ in their likely benefits and performance over a range of possible futures. Any choice is thus a portfolio of risk and there is need for a dynamic approach to real events. Moreover, strategic planning buys insurance by flexible building and committing only to immediate. At the same time, it maintains present the understanding of need for flexibility. Although strategic planning may sometimes sacrifice maximum performance, it achieves overall best performance since it prepares the infrastructure to adjust to actual situations in later periods, which tends to be very costly. Airfield designers should therefore be able to provide flexibility to the infrastructure. According to (de Neufville & Odoni, 2003) they should respond correctly to the following questions at any period: How much land should be acquired or reserved for a new airport? What should be the overall geometric layout of the airfield? What size of aircraft should the airfield be designed for? How should the construction of airside facilities be phased? The current trend is towards airport expansion or remodelling, since that there is a relatively small number of major new airports. This does not mean however that the planning task is simpler. On the contrary, it becomes even harder when it comes to conciliate scheduled construction activities with the regular operations that take place in an airport AIRPORT LAYOUTS There is a variety of solutions concerning the layout of airfields. Naturally, each solution has its own pros and cons, and must be tailored to the specificity of the site in study. As mentioned above, airfields occupies most of the land area of an airport (between 80 and 95%), with larger percentages 32

34 applying to airport with smaller land areas such as the 2,6 million m 2 of New York/LaGuardia (see figure 3-1) that contrast with the 136 million m 2 of Denver/International Airport (see figure 3-2). It is important to emphasize that the need for intersecting runways relates to the need for wind coverage. Both FAA and ICAO define a minimum threshold of 95% of airport usability with regard to crosswind coverage. Nevertheless, the number of runways is largely responsible for airport s capacity, and for how much land area is needed. In respect to traffic demand capacity, there are singular differences among different regions of the world. When analysing world s busiest airports, we realize little differences between 2000 and North America remains the region of the world where airports have the largest number of ATMs. Inversely, it has the smallest number of passengers per movement. Whereas in 2009, Charlotte-Douglas airport reported 68 passengers per movement and New York s JFK had 110, in 2000 values ranged between 56 to 96 passengers per movement. Asian 1 airports also maintained the largest number of passengers per movement, although decreasing on average from 178 in 2000 to 151 in European airports slightly increased the number of passengers per movement between 2000 and 2009, going on average from 105 to 116. The explanation for these discrepancies appears to remain constant, though. As (de Neufville & Odoni, 2003) explains, the differences are mostly due to the fact that Asia s busiest airports prefer widebody aircrafts in favour of the aircraft mix of narrow-body and regional jets as happens in North America and Europe. Furthermore, this fact reflects the importance for runway requirements at each region s principal airports. North America s busy air market, and to a lesser extent, European airports, need more runways than the Asian airports given the higher number of operations. Ranking by ATMs, within the first eighteen busiest airports, we observe that the top six is constituted by North American airports, and Beijing International is the only Asian airport in 10 th place. The most accentuated trend appears to be the increase of the Asian market, gaining two airports (Tokyo/Haneda and Beijing Capital International) in the top-5 and five airports in the top-25 busiest airports in terms of passengers. Airport TABLE 3-1: TRAFFIC AT WORLDS' 25 BUSIEST AIRPORTS (SOURCE: (ICAO, 2009), ACI WEB SITE) Passengers (millions) Movements (thousands) Passengers/Movement Atlanta/Hartsfield-Jakson London/Heathrow Beijing Capital International Chicago/ O'Hare Tokyo/Haneda Including Dubai International airport, not present in 2000 s top-30 busiest airports. 33

35 Paris/Charles de Gaulle Los Angeles/International Dallas/Fort Worth Frankfurt Denver/International Madrid Barajas New York/John F. Kennedy Hong Kong/International Amsterdam-Schiphol Dubai International Bangkok/Suvarnabhumi Las Vegas Maccarran Houston/George Bush Phoenix Sky Harbor San Francisco/International Singapore Changi Guangzhou/Baiyun Jakarta/Soekarno Hatta Charlotte-Douglas Miami/International Above, in table 3-1, it was summarized data obtained from Airport Council International s (ACI) web site (ACI, 2010) and (ICAO, 2009), similarly to (de Neufville & Odoni, 2003) for comparison purposes. It is also worthwhile mentioning that 2009 had all around negative variations with respect to 2008, whereas 2000 had shown much better performance in the overall airports. Without entering deeply into technical details, it is important to refer to some geometric characteristics of the runway systems, as they are also important when referring to land area requirements. The range of runways length may vary from 2000 to 4000m in major airports, depending on the site elevation among other factors. If parallel runways exist, the distance between them also critically affects total land requirements. Last, but not least, apron areas and taxiways affect greatly the size and cost of the infrastructure, being normally dependent of the reference code for design purposes. In one hand, urban expansion in the last century has limited largely the expansion of several major and secondary airports. Land unavailability is therefore a factor that makes unlikely for those airports to add new runways. On the other hand, major international airports have two or more runways that may be parallel or not. Such is the case of the above mentioned New York LaGuardia and Denver International airports, which are also good examples for this purpose. 34

36 3-1: NEW YORK LAGUARDIA AIRPORT LAYOUT (SOURCE: (FAA, 2003)) New York LaGuardia (figure 3-1) has high land restrictions, and its intersecting runways present bigger managing difficulties from the air traffic point of view. The intersection of runways implies the existence of conflict points. Thus, aircraft movements in one runway must be carefully coordinated with those happening in the other. In addition, the airport capacity may be strongly affected if strong winds in one direction force the shutdown of the correspondent runway. Hence, operational challenges in this kind of airports are enormous. On the other hand, airfields such as Denver International airport (figure 3-2), consist in two (or more) parallel runways. According to the distance of their centrelines, they are said to be close, medium-spaced or independent. In the first case, the second runway is typically used as taxiway whereas independent parallel runways allow for any pair of ATMs to happen simultaneously. The intermediary case allows for arrivals on one runway and departures on the other. Furthermore, close and medium-spaced runways tend to be associated to limited land availability, whereas independent runways allow for landside facilities development in between. 35

37 3-2: DENVER INTERNATIONAL AIRPORT LAYOUT (SOURCE: (FAA, 2003)) Independent runways, as in Denver International, offer a set of advantages in comparison to close and medium-spaced in the measure that they promote better airfield traffic circulation and bigger proximity to passenger and cargo buildings, thus improving the efficiency of aircraft operations. Furthermore, they also allow for a better control of landside s development as well as ground access to the airport. However, there are disadvantages as well. Independent runways not only require more extensive taxiways systems but also larger transportation links such as highways of railways to the airport. Additionally, it may also affect the flexibility of expansion on central placed landside facilities. (de Neufville & Odoni, 2003) The geometry of runways also influences the shape and geography of landside buildings. For this reason, further airport diagrams are presented later in the text TAXIWAYS AND APRONS Taxiway systems are of huge importance for efficient air traffic operations. By allowing quick entries and exits of the runway systems, it reduces delays and thus increasing the number of ATMs, which is the fundamental core business of airports. Nonetheless, these systems tend to be seen as complementary infrastructure. The positioning and configuration of taxiway systems often occurs after runway and landside facilities are set in place. 36

38 Taking back the example of Denver International airport, we can easily identify that taxiway systems can be of complex configuration, of extensive dimensions and involve high maintenance and operational costs. Such airports, with landside buildings positioned in the central area of the airfield are especially prone to problems of accessibility. For instance, they require a large number of expensive bridges so that road traffic arrives to its destiny. Furthermore, as the extension of these systems increases, so does the operating costs of airlines, that spend more time in tortuous paths between the runway and the apron stands. Also in this matter, FAA and ICAO s design standards explore thoroughly the safety recommendations, throughout geometric characteristics such as longitudinal and transversal slope, sight distance, width, taxiway curves, separation distances and other factors. For further details, see for instance chapter 3 of ICAO s Annex 14 or chapter 4 of FAA s AC 150/ However, it is worthwhile the discussion of exit taxiways. These segments provide the way out of the runway. The cheapest solution is a segment perpendicular both to the taxiway and runway. Additionally, due to the importance of fast clearance of the runway, 30-degree high-speed exit are also feasible solutions, despite the natural higher costs. Figure 4-3, depicts FAA s and ICAO s figures of both solutions. According to (de Neufville & Odoni, 2003), the location of taxiways plays a significant role in determining runway occupancy times, and to some extent, runway capacity. Although high-speed exits contribute to reduce runway occupancy, returns diminish as their number increases, due their higher cost. Thus, they are advisable for runways with more than 30 peak ATM per hour, and should rarely surpass three exits for each direction of operation when other conventional exits exist. Naturally, the introduction of high-speed exits in the direction of runways used only for departures has no increase of capacity of the runway system. It has significant capacity benefits on mixed operations mode and to a smaller extent, in runways used for arrivals only. 3-3: LEFT-CONVENTIONAL EXIT TAXIWAY (SOURCE: (FAA, 2007) ) RIGHT - HIGH-SPEED EXIT (SOURCE: (ICAO, 2004) ) Aprons, on the other hand, provide the necessary interface between airside and landside facilities, namely long-term parking, refuelling, boarding passengers or loading/unloading goods. 37

39 Passenger building stands can be further divided into two groups. They may be contact or remote depending on their location regarding the landside facility. Several configurations are possible, depending on the concept of the passenger building. Nonetheless, this subject will be explored further ahead. However, (de Neufville & Odoni, 2003) emphasizes the fact that different apron designs involve complex questions regarding the operations efficiency. Thus, trade-offs must be met between easiness of aircraft movement and passenger comfort. In the LCCs point of view, remote stands suit the best with lower fixed costs, since expensive bridges are not used. On the other hand, it implies high variable costs if road transport is necessary and the acceptance of passengers for earlier boarding CAPACITY AND DELAYS OF AIRFIELDS Ultimately, the runway system capacity defines the capacity of the airfield and of the airport. As seen previously, land limitations and environmental factor deny the increase of runway capacity on several airports. Furthermore, it is on the runway that flows of air traffic from various origins meet together in toll plazas, originating delays. Oppositely, taxiway systems, apron areas and landside facilities can be increased, in order to match the capacity of the runway RUNWAY SYSTEMS Being the production of aircraft movements the core business of airports, capacity of airfields is thus an important driver of airport efficiency. When demand approximates limit capacity, delays can represent huge capital costs for both airlines and airports. Traditionally, capacity of runways is defined as the maximum throughput capacity, which represents the expected number of movements that can performe on a runway system in one hour, without violating air traffic management rules, and assuming continuous aircraft demand (de Neufville & Odoni, 2003). This definition requires information regarding the conditions under which the runway is operating, namely, type or aircrafts, type of movements (departures or arrivals), allocation of movements if there is more than one runway, among other factors. On the other hand, it does not refer to any level of service. Whether delays occur or not, this measure only cares for the number of movements operated in peak hours, or under conditions of continuous aircraft demand. Other definitions considering levels of service exist, although rather ambiguous. Such is the case of practical hourly capacity, which is defined as the expected number of movements that can be performed in one hour with an average delay of four minutes per movement. Being the delay of four minutes considered the threshold for acceptable levels of service, the practical hourly capacity roughly represents 80 to 90% of the maximum throughput capacity, depending on specific conditions. Declared capacity is another measure based on the notion of sustained capacity, or in other words, the number of aircraft movements per hour that can be accommodated at a reasonable LOS. According to (de Neufville & Odoni, 2003), declared capacity seems to be roughly percent of the maximum throughput capacity. 38

40 Measures considering levels of service are important in delay measurement and more recently, in airport benchmarking 2. Furthermore, the data available on Portuguese airports uses different capacity designations, although declared capacity seems to be used more often. There is a very wide range of factors that affect the capacity of the runway system. The complexity of the relationships between factors such as those presented below is enormous: (de Neufville & Odoni, 2003) Number and geometric layout of the runway system; Separation requirements between aircrafts; Visibility and overall weather conditions (wind, precipitation, snow, etc); Mix of aircrafts; Mix and sequencing of movements on runways (departures only, arrivals only or mixed) Type and location of taxiway exits from runways; Performance of the air traffic management system; Environmental constraints (noise, land availability, etc) Although mathematical models exist and can obtain good results, the dynamic characteristics of airport delays are quite difficult to predict accurately. As optimum conditions for maximum throughput capacity occurs less often than desirably, queues of arrival and departing aircraft will most certainly happen. The reason for delays is rather obvious for peak traffic periods of the day, but less clear to periods when the demand rate is reasonably smaller than the capacity offered. The later are essentially due to the variability of time intervals between the continuous flows of requests for runway usage, as well as to the variability of the time necessary to process each movement of landing or departure. (de Neufville & Odoni, 2003) Regarding airport efficiency analysis, many studies used variables such as the number of runways and runway length. In the analysis of European airports, (Pels, Nijkamp, & Rietveld, 2003) found that the number of runways is not significantly relevant. Additionally, second order effects indicate that airports with a number of runways larger than the average do not lead to an increase of ATM, all things remaining equal. Similarly, (Gillen & Lall, 1997) found that the number of runways have no significant impact on the airside efficiency. Whereas the number of runways is widely used in the field literature, such does not happens with declared capacity, with no model having used this variable, to the best of my knowledge. Lisbon airport is the only Portuguese airport that has two operating runway. This fact may have strong influence on its performance, when taking into account the influence of LCCs in Portuguese airports. Additionally, most LCCs operating in Portugal use similar aircrafts such as Ryanairs Boeing or easyjets Airbus A319, and have similar runway length requirements in the order of magnitude of 2 See (Forsyth & Niemeier, 2010) for further details. 39

41 2100 meters. Thus, runway length can be a significant input variable if the type of aircraft traffic is taken into account, otherwise it may be negligible. Since all Portuguese airports operating LCC services have runways larger than 2400 meters, this aspect does not seem to be of high significance with respect to inter-aiport performance. IATA Code TABLE 3-2: UTILIZATION RATIO OF PORTUGUESE BUSIEST AIRPORTS IN 2009 (SOURCE: ANA) ATM PAX Runway Length Night Time Operations 1st Peak Hour ATM Declared Capacity Utilization Ratio LIS / ,8% ,3% OPO ,2% ,7% FAO ,5% ,1% FUN ,8% ,9% PDL ,8% ,6% In table 3-2, some characteristics of the five busiest Portuguese airports were compiled. Firstly, it is notorious the discrepancy between Ponta Delgada (the main airport in Azores archipelago) and Lisbon airports in terms of scale of operations. Secondly, according to ANA s annual reports, night time operations (between 21h and 05 h) represent in average 10,2% of total operations. Whereas in Lisbon and Porto airports there is a considerable amount of aircraft movements until midnight, in other airports often happens of no traffic movements at all in certain hours. Nonetheless, this percentage is similar in all five airports, indicating that not even Lisbon airport has the characteristics of a major international hub. Thirdly, it is also notorious the difference between declared capacity and the number of operations registered in the first peak hour at each airport, where more touristic or seasonal airports such as Faro, Funchal and Ponta Delgada suffer of more coincident arrivals and/or departures not only throughout the day but also all year around. Hence, if we consider that nearly 90% of air traffic movements in Portugal occur within a 16 hours period, we can say that potential annual capacity of airports will be in the range of 5840 (16x365) hours of declared capacity. The utilization ratio will then be the quotient of total ATM s and the potential annual capacity of a given airport, and here we find high variability too. Lisbon and Porto have the higher utilization ratios whereas Ponta Delgada airport has the lowest value. This index is particularly relevant in terms of operations delays. It tends to be higher in airports with hub characteristics such as Lisbon or to a smaller extent Porto, where the 30 th and 40 th peak hour are relatively close to the 1 st. On the other hand, delays themselves tend to be longer in airports with more touristic vocation such as Faro and Funchal where aircrafts tend to operate at similar times of the day, causing severe capacity constraints in small periods during the day. Although is commonly found in the existing research the use of length and number of runways as inputs as opposed to declared capacity, the previous exercise reveals how Porto and Faro airports with similar characteristics in terms of passengers and number of runways present very different utilization ratios, that will mostly certainly influence the airports overall performance. 40

42 TAXIWAYS AND APRONS Full-length taxiway is the designation for taxiways running parallel to runways, representing the principal element of taxiway systems. Aircraft traffic between runways and aprons (and vice-versa) use full-length taxiways as one-way lanes. By having smaller restrictions with regard to separation of successive aircrafts, flow capacity of taxiways usually exceeds that of runways. There are however, elements of the taxiway system that may form local bottlenecks. As mentioned above, runway exits to taxiways may impose limitations on the airfield capacity. Moreover, short taxiways segments, intersection points of taxi lanes with different speed characteristics, or runways intersecting taxiways may form potential critical points to the normal flow of aircraft movements. 3-4: POTENTIAL CONGESTION POINTS AT BOSTON LOGAN (ADAPTED FROM (DE NEUFVILLE & ODONI, 2003) AND (FAA, 2003)) Let us take the example of Boston/Logan Airport as depicted in figure 3-4. When runways 22L and 27 are used for arrivals (represented by the blue arrow), arriving aircrafts must cross runway 22R used for departures (represented by green arrow). The movement of aircrafts is represent by red arrows, and 41

43 when several are closely, it represents tendency for queues formation. Finally, yellow bolts represent potential collision points. Queuing happens more frequently in short taxiway segments and locations where high-speed exits merge with taxiways. When this configuration is used, air traffic controllers may be obliged to interrupt the flow of departures of runway 22R in order to arriving aircrafts from runways 22L and 27 have the opportunity to cross that runway and reach apron areas. According to (de Neufville & Odoni, 2003), such flow-constraining points typically exist in the taxiway systems of older, space-constrained airports. Hence, the majority of fully developed taxiways systems do not compromise the airfield capacity, and delays sustained at critical points are typically much smaller than delays originated by capacity limitations of the runway system. On the other hand, aprons may constrain airfields overall capacity. Aprons are defined as areas reserved for aircraft stands and the taxilanes that pass through those areas. As mentioned in section , these areas may be remote or contact, if they are adjacent or not to passenger buildings, respectively. Aprons capacity can be classified of static or dynamic. Whereas the former is defined by the maximum number of aircrafts that can occupy the apron at one moment in time, the later is more consistent with notion of runway capacity in the measure that specifies the number of aircrafts that is possible to accommodate within an hour. Additionally, aircraft stands may be of exclusive use of an airline or of shared use by members of an airline alliance. Airport operators and airlines normally prefer fixed locations instead of last minute changes because it is disruptive and costly. Furthermore, since delays happen frequently, in addition to the scheduled period, a tolerance time is given to that stand occupation. Periods of stand occupation by aircrafts may vary from 20 minutes to over one hour, depending if it is a LCC or an intercontinental flight. (de Neufville & Odoni, 2003) As mentioned earlier, LCCs prefer remote stands, thus avoiding the costs of costly bridges necessary in contact stands. Additionally, it reduces manoeuvring times of the aircraft to enter or exit the taxiway system, diminishing its turnaround time. From the airport efficiency point of view, researchers frequently use aprons area and the number of parking spots as input variables, but only (Pels, Nijkamp, & Rietveld, 2001) and (Pels, Nijkamp, & Rietveld, 2003) use the number of remote stands. Nevertheless, the efficiency analysis of 34 European airports using DEA/TFP methodology in (Pels, Nijkamp, & Rietveld, 2001) revealed that the number of contact parking and remote parking stands is clearly significant. Also (Gillen & Lall, 1997) found that the increase of boarding gates (and implicitly, the number of parking stands) has significant importance in the increase of airside efficiency. Since the efficiency analysis will be performed taking into account LCCs, 42

44 the ratio between the number of remote and total parking stands seems to be an important variable to consider DEMAND MANAGEMENT In the business world, the term demand management is used to describe the proactive management of increasing demand with business constraints (supply). According to (de Neufville & Odoni, 2003), airports demand management refers to any set of administrative or economic measures and regulations aimed at constraining the demand for access to a busy airfield and /or modifying the temporal characteristics of such demand. The main goal is to assist in maintaining efficient operations at airports threatened by congestion. Being demand the subject in case, no capital investments on capacity expansion (supply) are considered. On the contrary, the aim is to change demand patterns through regulation or economic measures that will either reduce overall demand for airfield operations or shift demand from certain critical time periods of the day to less critical ones. Furthermore, the author divides demand management in three sub-categories: purely administrative, purely economic and hybrids (the combination of the former two), depending on which measures are used to reach the desired objective. Regarding the administrative approach, slot, i.e., the reserved interval of time for the arrival or departure of a flight is a critical concept. In this sense, a set of criteria is defined in order to allow a proper allocation of slots among different users. The most widely administrative approach is IATA s schedule coordination. It considers criteria such as flight length and regularity (e.g. charter flights may be deferred to regular flights operated on daily or weekly basis), origin and destination of the flight (e.g. new markets or existing locations may be seen as particularly important) or even the characteristics of the airline requesting the slot (e.g. promotion of more competitive environment or management of other infrastructures such as terminal gates and parking stands). The main criticism to this approach however, is that whenever significant excess of demand over capacity exists, the lack of economic penalties or incentives may well lead to market distortions by, for instance, preventing the entrance of new players or by protecting any sort of perceived public services such as flag carriers. The European Commission has responded to such complaints through the third deregulation package in 1993, allowing a more market based distribution of slots. In practice however, some member States still protect their flag carriers. On the other hand, pure economic approaches use various sorts of congestion pricing in order to control airport access. By applying higher access fees on peak periods (hour, day, week and/or year) and lower and the rest of the time, airport managers aim to redistribute demand to less congested periods. In economic terms, congestion pricing relates with users willingness to pay for the infrastructure access. This congestion toll is equal to the marginal external cost of the airfield in peak periods, or in other words, the cost of delay that the last aircraft will impose to the rest of the users. When the supply 43

45 curve is left-shifted due to the internalization of congestion toll, there is a decrease in the number of airline carriers willing to pay for this additional cost. Hence, airlines with lower cost of delay time will be more willing to persist using airports at congested times, whilst high value of time operations such as LCC with short turnaround times will be more sensitive to worsening congestion. By adding the marginal private cost, i.e., the cost supported by user due to the delay that will incur, we obtain the social marginal cost (SMC). As in other transportation infrastructures, SMC pricing schemes with its welfare enhancing characteristics are increasingly reason of studies and debate. And although SMC may occur naturally in perfect markets, this is far from happening in transport infrastructures, where there still exists information asymmetries, the existence of public or semi-public goods (the airport itself) or increasing returns to scale. Despite its relevance on airside efficiency, SMC pricing analysis goes far beyond the above mentioned, and should therefore be object of study in more detail. In practice however, bigger difficulties arise in the implementation of marginal external costs. In one hand, it is rather difficult to estimate accurately marginal external costs for any given level of demand as it is to predict the effects on demand for the proposed congestion pricing scheme due to lack of information on demand elasticity. In the other hand, and probably to larger extent, the problem is political, where conflict of interests among different stakeholders is prone to slow down implementation of such pricing policies. For instance, whereas regional and general carriers may see it as discriminatory as they are the users that can least afford to compensate others for external costs, environmentalists or airport neighbours may favour this form of access restriction as a possible way of controlling airport expansion (de Neufville & Odoni, 2003). Finally, hybrid approaches to demand management are such that combine administrative and economic mechanisms. Hence, in addition to slot coordination, hybrid approaches will include economic measures such as congestion pricing, slot market or slot auctions to achieve the final allocation among users. Demand management is hence a critical factor that influences airports operational efficiency and has an important role from the LCC viewpoint. However, the difficulties associated with both data unavailability regarding managerial aspects and the introduction of such variables on the methodological frameworks for efficiency analysis, are enormous. To the best of our knowledge, demand management has never been used as input or output on airport s efficiency literature despite its huge importance. Best practices in demand management are of the interest of any airport manager and such is widely accepted. Thus, the issue is not whether to apply demand management, but how to apply it the best. 3.2 LANDSIDE According to (de Neufville R., 2008), airport planners and investors need to recognize the effect of low cost airlines. It implies a downward shift in standards and acceptance of these carriers volatility. 44

46 Moreover, although many airport operators find this reality difficult to accept, the trend runs against the practice of massive, multi-billion Euro investments of extravagant buildings designed by signature architects. Summarizing, policy makers and investors should focus more attention on the development of airport facilities serving LCC, both at legacy and low-cost airports. The discussion on landside operational efficiency however, goes far beyond the passenger buildings. For instance, with regard to the detailed design of passenger buildings, one could also bear in mind the existence of hot spots that undermine airport s level of performance or even ground access and parking in the airport and the necessary distribution of baggage, goods and services. Although such operations are remarkably important for the efficient economic performance of this complex transport infrastructure, we will not inter in deeper considerations due its natural complexity and, at the same time, because it does not fit the econometric model we intend to use. In the following section, the influence of passenger buildings configuration and the introduction of low-cost airports will be discussed as important factors in the efficiency of airports landside operations. Its importance, if in one hand is highly important, not only in the consideration of new investments but also in the management of existing buildings, in the other hand does not serve the benchmarking model we will use. Hence, the subsequent analysis will be shortened as it is reasonably possible to do PASSENGER BUILDINGS Historically, national governments tended to view airports as national entry gates and thus symbols of the nation grandiosity, leading into multi-billion euro projects that may not be economically viable in the long run. The industry deregulation has been contributing to a paradigm shift, where increasing efficiency is a must. Hence, the usual designation of terminals has shifted to passenger buildings in the sense that they are not the terminus of a journey, but instead the beginning of a new one. The configuration of passenger buildings has great importance on airports operational efficiency as it provides to the different users of the airport the necessary infrastructures for the several purposes happening in that space. From passengers check-in process to waiting lounges, passing by all the necessary security check-points and baggage handling systems, retail opportunities and the cost associated with this usually pharaonic buildings, interests among different stakeholders tend to collide and the need of conflict mediation is needed. Designers thus face critical issues when putting different stakeholders interests in a single building. Passengers, owners, managers, airlines, commercial retailers among others have a word to say in the financial efficiency of such complex infrastructure. The efficient performance of passenger buildings thus relates with the traffic it handles, and more precisely, with three derived characteristics. Firstly, the overall level of traffic. Secondly, traffic s seasonality and finally the percentage of transfer passengers (de Neufville & Odoni, 2003). 45

47 3-5: 2009'S MONTHLY PAX DISTRIBUTION ON THE TOP5 BUSIEST AIRPORTS (SOURCE: ADADPTED FROM (ANA, 2009)) 16,0% 14,0% 12,0% 10,0% 8,0% 6,0% 4,0% 2,0% Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec LIS OPO FAO FUN PDL Lisbon airport accounts for more than 50% of total traffic in Portuguese airports, with over 13 million passengers in Porto and Faro airports operated roughly 5 million passengers in the same year whereas Funchal has had over 2 million and Ponta Delgada less than 900 thousand. It is curious to note though, that in 2009 (a particular sensitive year for the industry), Lisbon airport registered an homologous drop of nearly 73% on transfers to passengers, far beyond Porto and Faro airports with and transfer passengers respectively, and even Ponta Delgada airport has had more transfer passengers with passengers (ANA, 2009). Regarding yearly seasonality (see figure 3-5), Faro airport is the infrastructure more sensitive to total passenger variation along the year, with almost five times more passengers in the busiest month in comparison with the less busiest. On the other hand, Lisbon, Porto and Funchal have similar monthly distributions, more flatten but also with the summer peak in August representing, in average, 50% more passengers than the traditionally less busy month of February. Such analysis is particularly important when new infrastructure investments are to be done, as is the case of the New Lisbon Airport (NLA). The proper characterization of travel patterns may prevent erroneous conceptions of what buildings suit the best each airport PASSENGER BUILDING REQUIREMENTS As stated above, among airports different users, it is interesting to understand four primary perspectives, namely passengers, airlines, owners and retailers. 46

48 Passengers present an enormous range of needs to be met, and depending on the type of passenger, the costs for operating an airport may differ a lot. Depending on whether domestic or international passengers dominate traffic, it influences the number of border and customs control points. Furthermore, the needs of business or leisure passengers will also impose determined requirements, as typically the former travels with less baggage and requires special amenities (e.g. lounges, duty-free shops, restaurants) whereas the later normally has much luggage and prefers inexpensive facilities. As mentioned above, transfer passengers may also influence severely airport s operational efficiency, especially at major international hubs, since they are potential non-aeronautical revenue sources. In our case study however, transfer passengers have a marginal effect as they represent less than 0,8% of total commercial passengers (ANA, 2009). From the airlines perspective, poor passenger buildings configuration may impose heavy burden on the operational level, given that even small savings in time, when cumulated over a day can have a major impact by allowing, for instance, an extra departure or arrival with the same crew and airplane. Airlines give particular attention to costs associated with ground operations. In fact, when considering peak delays and the way they spread throughout the day, small savings in operations time may represent enormous savings and airlines willingness to invest in new passenger buildings. (de Neufville & Odoni, 2003) The owner s perspective depends largely on the shareholders structure. Depending on whether it is state-owned (as it is the case of Portuguese airports) or fully privatized, the infrastructure can be seen in one hand as monumental entry gates in the country or political milestones by the former, or in the other hand as a financially efficient company by the later. Finally, the viewpoint of retail operators is based on persons wanting to spend their money and time at their shops. Visibility, access and a coherent environment are crucial factor for these stakeholders. Hence, any arrangement featuring central areas is more attractive to stores than buildings with many entries and exits. (de Neufville & Odoni, 2003) BUILDINGS CONFIGURATION AND EVALUATION The requirements above-mentioned present a critical problem for passenger buildings designers: they need both to concentrate and spread them out. (de Neufville & Odoni, 2003). In this section, a short description of passenger buildings configurations will be given from a functional perspective and its evaluation made according the stakeholders needs. It is important to note that some configurations are evolutions or variations of existing concepts of passenger buildings, and therefore, some difficulties may arise while theoretically distinguishing them. In practice however, the name given has little importance. To the stakeholders, the infrastructure s operational efficiency matters the most. 47

49 Furthermore, the following configurations are best suited for larger airports. For instance, Ponta Delgada airport s passenger building, in order to handle its annual passengers, does not have the need of a heavily elaborated building. On the other hand, other configurations may arise from the tailoring of basic configurations, resulting in hybrid configurations that may suit the best each case study. are: Hence, and according to (de Neufville & Odoni, 2003), the basic passenger building configurations Finger piers; Satellites, with or without finger piers; Midfield, either linear or X-shaped; Linear, with one side devoted to aircraft; Transporters; Airfield designers first introduced finger pier configuration in the 1950 s. Similar to the plan view of the palm of a hand, the fingers can hold aircrafts on both sides of the building, from the central core to the end of the pier. An alternative configuration to finger piers consists in creating hammerhead fingers, which allows serving a higher number of aircrafts and reduces the space needed for lounges as they are placed in a shared area (the crosspiece of the T). By providing steady flows of passengers to the central building, this configuration is likely to increase retail opportunities with the cost of longer walking distances from passengers that use gates more distant from the main building. The inconvenient hike that passengers suffer in such configuration has lead airport designers to shorten the finger piers extension or, alternatively, to replace them with people movers that serve independent satellites and/or midfield concourses. 3-6: FINGER PIER AT LAGUARDIA AIRPORT (LEFT) AND SATELLITES AT TAMPA AIRPORT (RIGHT) (SOURCE: (FAA, 2003)) 48

50 The satellite configuration, as depicted in figure 3-6, is the natural extension of T-shaped finger piers, by eliminating the gates along the fingers and concentrating them at the end. As in other configurations, some variations occur. In one hand, the connection between satellites and central passenger building may happen underground or at surface level. Whereas underground connections have higher sunk costs it also allows aircrafts to maneuver more easily around the satellite, which may represent savings both for airlines in the short run as for airport operators in the long run as it increases the possible number of operations. In the other hand, the use or not of people movers can also influence greatly airport s operational costs. The main difference between satellite and midfield concourses configurations lies on the size and distance of the passenger building to the groundside, despite this distinction is not firm. Typically, midfield concourses are defined as major independent passenger buildings located far from the groundside. Although often located between parallel runways, it also occurs to find midfield concourses of complex passenger buildings positioned at the edge of runways, as it happens at Chicago O Hare airport. (de Neufville & Odoni, 2003). 3-7: MIDFIELD LINEAR AT DENVER AIRPORT (LEFT) AND X-SHAPED AT PITTSBURGH AIRPORT (SOURCE: (FAA, 2003) ) Figure 3-7 depicts the two traditional shapes of midfield concourses: linear and X-shaped. Although both configurations support aircraft stands on both sides of the buildings and have central shopping areas, linear midfield concourses provide better services for transfer passengers. The main argument to choose the former configuration relates mostly with the limited available area for airport expansion. An alternative variation to the X-shaped configuration is the cross-shaped passenger building. The former has appeared as a solution to reduce passengers walking distances in opposition to the best use of the scarce land area. One way or the other, X-shaped (or cross-shaped) concourses imply more complicated aircraft operations due to buildings geometry in comparison to linear midfield concourses. 49

51 Linear buildings, devoting one side to aircrafts and the other to the groundside, are long and relatively thin structures. The original idea was that passengers arriving to the airport would be able to get to their flight gate just by walking though the thin building, hence minimizing walking distances. However, such concept is counterproductive to the extent that requires additional security and check-in facilities, does not provide steady passenger flows for retail opportunities, and passengers do not traditionally arrive at the exact point where their flight gate is located. Hence, designers now tend to give fewer access points, which increases passengers walking distances, as opposed to the initial concept. Although some airports have chosen curved layouts (as Kansas City airport depicted in figure 3-8) which provides more frontage on the airside, such layout complicates both the initial construction and landside traffic flows. (de Neufville & Odoni, 2003) 3-8: LINEAR (LEFT) AND HYBRID CONFIGURATION (RIGHT) AT KANSAS CITY AND SEATTLE TACOMA AIRPORTS (SOURCE: (FAA, 2003)) Finally, transporters consist in the transport of passengers between buildings and aircrafts, typically done by low-platform buses, which also usually require passengers to use stairways while carrying their bags. Alternatively, lift lounges are used to move passengers between the building and the remote aircraft stand, thus preventing passengers from suffering climate exposure and speeding up the (un)loading process, with the natural side effect of being particularly expensive infrastructures. Transporters allow dealing with strong seasonal variations in passenger traffic. Since the cost of transporters can be minimized when they are not needed, it provides a cost-effective solution to aircraft gates in buildings that require maintenance, cleaning and considerable depreciation costs relentless their use. (de Neufville & Odoni, 2003) Seattle Tacoma airport s diagram, as depicted in figure 3-8, uses an underground people mover to the satellite (or midfield, depending on the interpretation) concourses, has two linear one-side devoted 50

52 aircraft buildings and two finger piers. Such is an example of an hybrid solution, with the intent of more capably satisfy the airport s different needs. In our case study, transfer passengers represent a residual sample of total commercial traffic. As mentioned previously, in 2009, Ponta Delgada airport has registered more transfer passengers than Lisbon, when the later has 15 times more commercial passengers. Although it should be noted that Ponta Delgada airport operates as a regional hub, diverting traffic to the other 3 airports of the Azores archipelago, and that 2009 was a particular sensate year in the aviation industry, the percentage of transfer of transfer passengers to total commercial passengers on the top-5 busiest Portuguese airport varies between 0,15% and 2,4%. In one hand, Faro airport confirms its touristic and thus, seasonal vocation with 0,9% of transfers. The seasonality previously identified at Ponta Delgada airport balances with its hub characteristics at the regional level. On the other hand, Lisbon s airport image of international hub is demystified with its mere 0,15% transfers of total commercial traffic. (ANA, 2009) The high seasonal variation of Faro and Ponta Delgada airports comply with the use of transporters in such peak traffic periods, with particular emphasis at Faro airport given the volume of low-cost passenger traffic reaching nearly 70% of total commercial traffic and the low-cost ATM representing 64% of total commercial movements in (ANA, 2009) From the stated above, it becomes clear that issues such as passenger walking distances, average taxiing times around passenger buildings, transporter economics and flexibility in airport design are central concerns from the managerial point of view, and must be dealt taking into consideration the various stakeholders interests. Yet, there are no magical recipes for passenger buildings design. Instead, it is the detailed knowledge of the airport characteristics and needs that will allow the airport designer preparing it for a wide range of possible circumstances that will enable to best adapt when paradigm shifts occur SECURITY AND CHECK-IN PROCESSES Security is a major issue regarding landside operational efficiency. According to (Graham A., 2008), airport security relates with the prevention of illegal activities such as terrorism. In this respect, the 9/11 events have lead to a higher scrutiny of airport operations, with many additional security measures being introduced and, more importantly, international binding legislation was introduced. The introduction of such measures has overburden airport s operational costs, with particular incidence in the airlines, which saw alternative check-in processes as a possible solution to overcome airport s scarce space and the delays inflicted by the additional security measures. Self-service check-in desks became a complementary tool to process passenger and baggage flows in airports. In the first decade of the 21 st century, airlines first introduced for their own use proprietary kiosks, copying other industries cost-effective self-use technologies. Later, and following the logic of economies of scale of airlines alliances, common-use self-service (CUSS) check-in kiosks were 51

53 implemented, allowing costs to be shared between different airlines and less airport counter staff was required. The main disadvantage with most self-service check-in kiosks arises in the difficulty to cope with hold baggage. In order to solve this problem, some airports already incorporate a bag tag printing capability (BTC) in CUSS kiosks, along with common bag drop locations. The use of CUSS kiosks is one of the aims of IATA s Simplifying the Business initiative launched in Since then, the number of airports with CUSS kiosks has raisin from only 10 to 134. Within the same initiative, it was also given focus in other areas such as electronic ticketing, bar-coded boarding passes, radio-frequency baggage identification and paperless cargo movements. (Graham A., 2008) Figure 3-9 below, shows (on the left axis) the world distribution of airports with CUSS kiosks and (on the right axis) the percentage of airports that have incorporated BTC system. (IATA, 2010) 3-9: WORLD DISTRIBUTION OF AIRPORTS USING CUSS KIOSKS AND PERCENTAGE WITH BTC (SOURCE: (IATA, 2010)) 60 80% 50 67% 70% 60% 40 50% 30 43% 40% % 30% 25% 30% 20% 10% 0 0% 0% AFRICA ASIA PACIFIC CIS EUROPE MENA NORTH ASIA THE AMERICAS 3% US 0% Airports with CUSS Kioks % of Airports with BTC on CUSS More recently, personal computer and mobile check-in has increased the number of possible alternatives for passengers. Although such technologies are not always at the dispose of travellers, they are even more attractive to the airlines as they do not need to install CUSS kiosks and passengers print boarding passes with their own ink and paper. Nowadays, the debate around CUSS kiosks has evolved from discrete check-in services to a concept of a more efficient integrated model that shares the information with different stakeholders. According to (Graham A., 2005), the ideal process flow concept combines CUSS, new generation passports, biometrics, secure databases and other technological advances to facilitate automatic authentication of a passenger s identity whenever this is required at the different stages in the travel 52

54 process. Although concerns may arise from the individual data protection point of view, the integration of this kind of technologies in airport operations most certainly involves benefits for all stakeholders, passengers included LOW-COST AIRPORTS Low-cost airlines are becoming significant factors in airport planning. Their requirements differ from those of legacy carriers. They drive the development of secondary airports and cheaper airport terminals. They catalyze low-cost airports around the legacy airports built for the legacy airlines. (de Neufville R., 2008) In Low-cost airports for low-cost airlines, Professor de Neufville presents a particular interesting paper for our case study. By challenging the traditional airport masterplan design with a flexible design strategy that copes with the uncertainty of the aeronautical industry, he presents a comprehensive economic analysis through a simple cash-flow exercise that reveals the advantages of flexible design regarding the construction of the NLA, taking into account the important growth of LCCs. The increasing role of LCCs relates with the expansion of low-cost airports, either secondary airports in multi-airport systems or former military bases. Whilst LCC s market- share sky-rockets, the provision of more accessible facilities capable to attract these companies starts to grow. To this respect, a new form of competition not experienced before now threatens traditional airports. From the geostrategic point of view, such competition is especially strong at multi-airport systems, where low-cost airports offer interesting alternatives to major international hubs. To a smaller extent, these airports also allow to bypass those same hubs, as it happens at Faro airport, where European touristic traffic avoids Lisbon s busy airport. Furthermore, LCCs and low-cost airports also provide parallel routes against established legacy networks, as it happens at Porto airport, with Ryanair flights to Girona (Barcelona), Stansted (London), Charleroi (Brussels) or Beauvais (Paris) among others. Additionally, there are also big differences between traditional and low-cost business models. Firstly, just like LCCs, low-cost airports avoid expensive infrastructures. Secondly, passenger s level of service in terms of space is lowered, and when combined with faster turnaround times, it directly relates to smaller infrastructure costs per high annual passenger volumes. Finally, and in clear contrast with airport s trendy bet on commercial lounges (that increase airports non-aeronautical revenue), low-cost airports will avoid this spaces, as it implicates further construction, security and other operational costs, diverging from the no-frills concept. The combination of these factors results in lower charges for the airlines, which are one of LCC s most relevant fixed costs. On the other hand, being relatively uncongested, low-cost airports avoid ground and air delays and when combined with quick turnaround times, are the cornerstone for smaller variable costs. These are the main arguments for the paradigm shift, whereby airport planners and investors should recognize the volatility of LCCs and accept the necessary downgrade in service standards. 53

55 Demand has thus become more uncertain than ever. Additionally, traffic growth rate changing according to different economic cycles, industry s volatile regulation and the uncertainties of an industry passing through an era of change, important consequences for airport planning must be taken in consideration, as (de Neufville R., 2008) enumerates: Major airport users may disappear. Following the Unites States, extinction and mergers of established legacy carriers in Europe has already started with the bankruptcy of Sabena and Swissair, and merge of KLM with air France. The Portuguese flag carrier TAP is soon expected to be privatized, mostly to remove an important expense from the national payroll. Change of distribution and patterns of traffic. LCCs may divert traffic to secondary metropolitan airports, or nationwide, as it happened at Faro and Porto, with high volumes of LCC passengers. Different design standards may apply. The airline clientele may reject facilities provided, leaving them underutilized and possibly unprofitable, has Ryanair has done in Porto. Ultimately, airport developers and decision makers face great risks regarding new infrastructure investments. It is thus needed a design process that allows to recognize explicitly the uncertainties that threaten planned investments and developments. As mentioned previously in the airfield design process, flexibility is the key to successful development strategies, by preparing airport managers to adapt more efficiently to reality changes. 3.3 AIRPORT-AIRLINE RELATIONSHIP The relationship between the airport operator and the airline is clearly fundamental to the success of any airport business. (Graham A., 2008) As primary users of airports, airlines throughout the recent decades have forced airports to become increasingly more efficient. Their entrepreneurial spirit and competitive industry promotes a rigid cost control in a deregulated environment. Increasing oil prices in recent years and scarce capacity for increasing demand obliges airlines to seek cost reductions and efficiency gains at their primary supplier. Meanwhile, airports with traditional governmental management have started to pursuit more business-oriented strategies that affect inevitably the airlines. In order to understand this important relationship, it is crucial to understand how airports are organized and managed, which are the economic ties that connect both partners in this business and finally, to comprehend the regulatory environment that dictates how is the game played. Such topics will be briefly addressed in the subsequent sections AIRPORT PRIVATIZATION AND MANAGEMENT 54

56 Powered by the airlines industry deregulation, airport management began to change in the last quarter of the 20 th century, moving towards to a more commercial perspective. This shift was done in many different ways, but mostly through the establishment of independent authorities that, in some cases, opened the door to private cooperation and investment. In Europe for instance, the Polish Airport State Enterprise was created in 1987, Spain launched in 1991 Aeropuertos Espanoles y Navegacion Aerea, and in the Portuguese case, ANA was created in 1998 (Wikipedia, 2010). The innovatory commercial aspects of airport management included financial management benchmarking, non-aeronautical revenue generation and airport marketing were important contributes to the traditionally operational perspective. A business like approach to airport management, coupled with a more commercially driven and competitive airline industry, encouraged airports to take a much more active and proactive role. (Graham A., 2008) In the end of the 20 th century, airports privatization was subject of intensive debate. In one hand, privation would reduce the need for public sector investment (meeting the European Member-States goal of finding financing sources outside the public budget) and provide access to the commercial markets. On the other hand, private monopolies along with its intrinsic overcharging schemes, poor standards of service, inadequate investments or lack of consideration towards social equity could appear as fearful arguments regarding the privatization of public services. Hence, for many countries, transferring airports, which are considered national or regional assets, to the private sector remains a politically sensitive policy. The fear is that social welfare will not deserve the same attention as shareholders interests. As in other transportation modes such as railways or highways, airports have the monopolistic position of the infrastructure manager that does not suffer the same kind of competition of airlines. It is therefore understandable why views of privatization vary in different regions around the world and even between local and central government bodies in individual countries. (Graham A., 2008). One interesting example is given by (Hooper P., 2002) regarding privatization of Asian airports. The author shows that despite the fact that some governments say that efficiency is important to them, the most common and important motive in privatization in Asia is to mobilize a new source of finance and while governments are concerned about abuse of monopoly powers they want to cross-subsidize regional airports, but they lack the institutional strengths to regulate effectively. The crucial risk by retaining majority control however, is losing the efficiency benefits of privatization MODELS FOR AIRPORT PRIVATIZATION Although seldom used, airports privatization may be a misleading term for airports changes in ownership. Airport privatization does not actually involve the sale of property but, quite differently, the transfer of ownership rights such as profits and management control on short and long-term development issues. (de Neufville & Odoni, 2003) 55

57 Due its monopolistic position, airports are generally seen as attractive organizations to investors. Moreover, the industry has shown steady growth in the recent decades, and has very high barriers to entry, either as of large capital investments or as the lack of convenient locations where airport development is allowed. (Graham A., 2008) Notwithstanding, there are considerable risks. Airport regulation and development can be changed through political legislation and along with airlines industry volatility, poses significant risks to shareholders. The privatization model adopted by political decisionmakers is therefore a complex process depending on government s objectives. According to (Graham A., 2008), airport privatization can occur in different ways: Share floatation; Trade sale; Concession; Project finance privatization; Management contract; Share flotation is the mean by which airport s share capital is issued and subsequently traded on the stock market. British Airport Authority is a rare example of an 100% share floatation undertaken in 1987, whereas in many other airports, less than half of the airport s capital is put on the market, or in other words, partial share floatation. One way or another, this kind of privatization will reduce the need for state involvement in the financing of airport investment while transferring partial or effective control and the inherent economic risks to the new shareholders. Nevertheless, governments can maintain a certain degree of influence in case of preservation of national interests through golden shares. Trade sales usually involve a public tender whereby the airport is sold to a strategic partner or consortium of investors, rather than just passive investors. Often, the strategic partner is an established airport operator or the purchasing consortium will contain a member with airport management experience. Same cases have occurred of participating airports in the privatization process were not actually privatized themselves, which leads to further complications in the definition of private airport. (Graham A., 2008) Concession is the type of arrangement whereby public tendering, consortiums or airport management companies purchase a lease to operate the airport during a defined period. Concession agreements are most popular in developing countries, with particular incidence in South America as it is the case of ANA that, in a consortium lead by Ferrovial, was awarded with a 25-year concession of 12 regional airports in Peru (MOPTC, 2008). Such is typically related with the involvement of initial payments, guaranteed levels of investments and/or payment of annual fees, factors of great importance for nations with less access to high volumes of financing. (Graham A., 2008) Although more politically acceptable since the state will not actually sell the airport and still receive a regular income, this process tends to be more complex with high transaction costs and requires particular attention in the design and implementation in order to fully achieve the governmental political objectives. 56

58 Project finance privatization is the option preferred for relatively large investments in greenfield projects or even redevelopments. After public tendering, the awarded private consortium or private-public partnerships (PPP) will bear the initial investment costs and all operational costs for a predetermined period of time, after which ownership reverts to government. Many methods exist in this privatization model, with the most popular being build-operate-transfer. In the Portuguese case, the government has chosen to build and finance the NLA, under a Design-Build-Operate-Transfer (DBOT) in conjunction with the partial privatization of the concessionaire ANA. In the context of an international public tender, the selected private partner while acquiring a majority stake of its share capital will be entrusted with the obligation to build and operate the NLA for a concession period of 40 years. (MOPTC, 2008) Finally, management contracts are the least radical privatization model. This option consists in maintaining airport ownership under governmental control while the winning consortium takes responsibility for daily operational management of the airport, usually for periods of years (Graham A., 2008). Such model has more political acceptance although investments are done at expense of the public budget. On the other hand, the private partner may find this alternative more attractive in countries with higher financial exposure. Some issues arise regarding airport privatization. One regards how governmental control depends on the chosen privatization model. Project finance or concessions are more politically acceptable and governments maintain a considerable degree of control that can influence the efficient performance of the airport. On the other hand, trade sales or share floatation usually carries along fears of private monopolies and have thus lead to the introduction of economic regulation. Another concern arises when the privatization of a group of airports happens, the question being whether airports should be sold together or split up into different companies. If sold as a group and successful financial records exist, higher prices can be achieved due to the lack of perceived competition. However, if the operator s income is largely affected by loss-making airports, the sale price may suffer accordingly. Hence, unprofitable airports are removed from the government s payroll at expense of a smaller sale. Then again, if only profitable airports are privatized, one option would be to use concession fees (in a concession arrangement) to cross-subsidize smaller airports. Furthermore, airlines tend to suspect of airport groups, fearing that the fees paid will be used as cross-subsidization for loss-making airports they do not use. Airport operators normally respond to this criticism arguing that through economies of scale, higher efficiency can be achieved. (Graham A., 2008) 57

59 3-10: PORTUGUESE AIRPORT OPERATOR S NET PROFIT BETWEEN 2004 AND 2009 (SOURCE: (ANA, 2009B)) Net Profit in Million ANA Group ANA In the Portuguese case, the partial privatization of ANA relates to Azores and mainland airports. Notwithstanding, ANA s group of companies still comprises one of the two Portuguese state-owned handling companies, 70% of the ANAM and around 84% of the special purpose company established for the planning and development of the NLA. From figure 3-10 above, we can observe both the increasing net profits of the group along the recent years and an average 10,6 million difference between ANA and the rest of the group companies. Furthermore, peripheral Azores Islands airports have typically negative results and are hence cross-subsidized by mainland airports. We can thus infer that the concessionaire will finance loss-making airports with mainland airports revenues, while the government will use its share of revenues to pay for the losses of the remaining companies of the group. The question to be answered some years from now is whether efficiency gains expected from the private operator will pay off or not today s profits derived from the state-owned monopoly OWNERSHIP EFFICIENCY STUDIES Privatization not only affects competition between airports but also raises doubts on whether it will lead to efficiency gains or not. To date, there has been only limited and in some cases contradictory research in this area. (Graham A., 2008) One of the first studies by (Parker D., 1999) used DEA methodology to assess to what extent technical efficiency changed following BAA s privatisation. The study found that privatisation had no noticeable impact on efficiency. On the other hand (Vogel, 2006) revealed economically meaningful and statistically significant differences between publicly owned and privatised airports. Using DEA on a panel data of 35 European airports, the author found that major differences lie in operating efficiency, capital productivity and capital structure. Although partially and fully privatised airports operate more efficiently, it does not translate into significantly higher returns on shareholders funds. Furthermore, due to their at 58

60 least indirectly government-backed credit standing, publicly owned airport companies can assume more debt relative to their respective equity. Within an international context of airport ownership benchmarking, (Oum, Adler, & Yu, 2006) conducted a cross-sectional, time series of major European, North American and Asia-Pacific airports. The authors found strong evidences that airports with government majority ownership and those owned by multi-level of government are significantly less efficient than airports with a private majority ownership. In addition, no statistically significant evidence to suggest that airports owned and operated by US government branches, independent airport authorities in North America, or airports elsewhere operated by 100% government corporations have lower operating efficiency than airports with a private majority ownership. The commercialization of airports with a private majority ownership becomes evident with a much higher proportion of their total revenue from non-aviation services than any other category of airports while offering significantly lower aeronautical charges than airports in other ownership categories excluding US airports. It is also found that airports with private majority ownership achieve significantly higher operating profit margins than other airports; whereas airports with government majority ownership or multi-level government ownership have the lowest operating profit margin. Particularly relevant to our case study, the authors suggest that private public partnership with minority private sector participation and multi-level governments ownership should be avoided, supporting the majority private sector ownership and operation of airports. Using a different methodology but with a similar sample, (Oum, Yan, & Yu, 2008) report four key findings. Firstly, countries considering privatization of airports should transfer majority shares to the private sector. Secondly, mixed ownership of airport with a government majority should be avoided in favour of even 100% government owned public firm. Thirdly, US airports operated by port authorities should consider transferring ownership/management to independent airport authorities. Finally, privatization of one or more airports in cities with multiple airports would improve the efficiency of all airports. Regarding Latin American airports, (Vasigh, Erfani, & Miner, 2009) suggest that privatized airports outperform government owned airports. However, there is no conclusive evidence between privatized airports and selected North American airports. There are thus strong arguments for and against privatization. For our case study, it is not a relevant issue in the sense that the same operator manages the airports in question. To analyze such influence, other benchmarking study approaching similar size European airports should be considered. The reality however is that he airport sector has moved from and industry characterized by public sector ownership and national requirements into a changed era of airport management, which is beginning to be dominated by the private sector and global players. (Graham A., 2008) In the Portuguese case, as mentioned previously, only some years from now will we be able to assess to what extent was the privatization of the public monopoly successful. Notwithstanding, it is crucial for political decision makers 59

61 to follow international guidelines and make best use existing literature on airport benchmarking in order to potentially maximize social welfare LCC S IMPLICATION ON AIRPORTS REVENUES Traditionally, airports were dependent on a combination of governmental funding and revenue from airlines by charging them so-called aeronautical charges for the use of their services. Although there is a variety of practice worldwide, typically, airports will charge airlines a weight related fee to land their aircraft, a fee per passenger that passes through the terminal, aircraft parking charges and charges for office space. Additional charges relate to ground handling and fuel and these may be provided by the airline itself or by a third party company (or companies). (Humphreys, Ison, & Francis, 2006). Nonetheless, as airports become increasingly more commercially oriented, typical aeronautical revenues are now complemented with charging schemes more complex and market based. In recent years, airport charges have been of growing concern to airlines especially due to the increased competitiveness in the industry and falling yields, which have resulted in cost cutting activities of internal activities (staff, wages etc.) they can best control. However, airlines have also been looking at their external cost such as airport charges and demanding that airports adopt cost cutting and efficiency saving measures themselves, rather than raising their charges. To this respect, LCCs and charter airlines have particular strong reasons to worry then traditional carriers. Whereas the later typically operates longhaul services, the former operates short sectors, representing payments of airport charges more frequently. And although airport costs represent on average 4% of traditional airlines operating costs, it accounts up to 17% for LCCs, being the third most important cost for LCCs right after fuel and aircraft leasing cost. (Graham A., 2008) Furthermore, the existence of airport incentive schemes or discounts for the exploration of new routes makes the airport charging policy having its greatest impact. Such discounts have, in many cases, been critical factor when LCCs are selecting suitable airports for their operations in addition to sufficient demand and fast turnaround facilities. Nevertheless, airport managers also tend to fear LCCs flexibility to enter and exit routes. Although it gives LCCs bargaining power, such lack of commitment poses the threat of unsuccessful investments to national or regional governments. LCCs have thus reshaped the airport-airline relationship. Airports have responded to the potential opportunities that have arisen from LCCs growth. The low cost model has implications for the airlineairport relationship, forcing airports to negotiate contracts which significantly reduce aeronautical revenues whilst seeking to address this short fall by commercial revenues via increased passenger numbers. According to (Humphreys, Ison, & Francis, 2006), airports have sometimes found it difficult to turn increased passenger volume into additional revenue and traditional airports are challenged in terms of if and how they should accommodate LCCs. The author further enunciates eight issues airport managers need to consider when accommodating LCCs: 60

62 i. Continual market monitoring: keeping the low-cost airline market under continual review and reassessing whether to accommodate LCCs or not; ii. Volatile nature of the low-cost sector: consider volatility in both revenue streams, networks available, high number of route entries, operators going out of business or transferring operations to another airport, especially when the airport management should be considering investing into new capacity; iii. Significance of the non-aeronautical revenue: having sufficient retail and car parking capacity to create commercial revenue streams, carefully calculate the break-even, monitor revenue streams; iv. Capacity to cope with the LCCs: in terms of airport capacity (both terminal and runway), need for new low-cost facilities (devoid of air bridges etc.) v. Tensions between incoming and incumbent airlines: the threat of pressure to reduce charges to existing operators when accepting LCCs; vi. Need for transparency: especially applies to publicly owned airports where the incentives might be seen as a not permissible subsidy; vii. Benefits to local economy: again, especially applies to public owned airports with interest in bringing benefits to the local economy, management should beware that such effects can be difficult to predict and quantify; viii. Innovative/risk sharing contracts: some airports use contracts that contain clauses that relate charges to the number of passengers carried or the number of services operated and along with investments into software that enables passenger monitoring by flight number. With such detailed data, airport managers can build up a picture of what passengers spend, on what routes, at what times of day. As finding shows, particular routes at particular times can pay for themselves just by passenger spending. With increasing passenger traffic, LCCs are thus a new powerful player in the airport-airline relationship, and key shareholders in the development of new or existing airports REGULATORY ENVIRONMENT According to (Graham A., 2008), airports, are subject to a number of different regulations at both international and national level. Many of these regulations relates to the operational, safety and security aspects of airport management. Also increasingly important, airports are now facing more strict environmental regulations, restricting for instance, the number of hours of operation or the available landside for expansion. However, the regulation that influences the most the airport-airline relationship is the economical, with particular focus being on charge control. In some areas of the world, other economic aspects of operational activities such as handling and slot allocation are also regulated. International guidelines have also been issued both by ICAO (ICAO, 2004) and the EU (European Comission, 2009). In fact, they do not differ much in the main essence that is cost related charges. 61

63 Furthermore, these guidelines encourage principles of non-discrimination, users consultation and transparency. Along with the recommendation of setting up of an independent national regulator, it is suggested that airports supply of financial information to airlines regarding the charges calculation method whereas airlines would supply their traffic forecasts and airport requirements. One meaningful change of EU s regulation was the scope of airports obliged to comply with this regulation. The threshold increase from one to five million passengers would reduce the number of regulated airports from 180 to 70 in (Graham A., 2008), The airlines response to the EU directive was rather of disappointment, in the sense that prefinancing was not prohibited and the general lack of provisions to make airports more cost efficient. Similarly, LCCs said that the directive marked a missed opportunity to introduce targeted, robust and effective regulation of those relatively few dominant airports in Europe that actually need to be regulated. Furthermore, LCCs see no justification for allowing airports to pre-finance future facilities from charges on existing users and were disappointed that airports are allowed to choose a dual-till business model. (ELFAA, 2008). As (Graham A., 2008) points out, also the CAA feels that regulation can be burdensome, and should only apply to airports with substantial market power. In a nutshell, competition is preferable to regulation, and even where competition acts only as a weak discipline on behaviour, regulation should only be preferred if it can deliver a clear net benefit. (CAA, 2007) According to (Marques & Brochado, 2008), there are several economic regulation methods at work worldwide. It is not easy to find consensus in their classification, but they can be sorted into two main groups, according to the incentives they offer the regulated industries towards costs minimization. The first group, with a very low degree of incentive, includes the rate of return regulation (RoR), whereas the second, with a high degree, corresponds to the incentive regulation. The remaining economic regulation methods are variations or interactions between these two classes, such as the wellknown sliding scale approach, in which the costs and revenues (profits) are shared among stakeholders. Although the RoR is widely used, and ensures that prices are cost related, this system is highly criticized since it provides no incentives to reduce costs, and therefore, the operator will be guaranteed a certain RoR irrespective of efficiency. On the other group, incentive regulation promotes efficiency and innovation. Created on Tatcher s utilities privatization legislature, this kind of regulation comprises different methods such as price cap regulation (PCR), revenue cap regulation, hybrid and yardstick competition methods. As (Marques & Brochado, 2008) refers, PCR consists in the imposition of an average maximum threshold for the charges of the services provided. With the prices ceilings defined at the beginning of each regulatory period, the regulated services hold the earnings corresponding to the cost reduction which happens during that period. As PCR is not based on individual costs, it fosters appropriate price structures, maximizing the welfare. The price cap formula is composed of two parts (CPI X), one corresponding to the consumer price index (CPI), and the other (X) to the operator s productivity change expected. 62

64 TABLE 3-3: AIRPORT REGULATORY METHOD IN EU COUNTRIES IN 2006 (SOURCE: (MARQUES & BROCHADO, 2008)) Country Regulatory method Country Regulatory method Country Regulatory method Austria Non-pure price cap Greece No regulation Poland No regulation Belgium Yardstick competition Hungary Pure price cap Portugal Rate of return Czech R. No regulation Ireland Revenue cap Slovak R. No regulation Cyprus No regulation Italy No regulation Slovenia No regulation Estonia Rate of return Latvia No regulation Spain Rate of return Denmark Pure price cap Lithuania No regulation Sweden Pure price cap Finland No regulation Luxembourg Rate of return UK Pure price cap France Revenue cap Malta Pure price cap Germany Non-pure price and revenue cap and RoR The Netherlands Rate of return Table 3-3 above, summarizes the airport economic regulatory models in European countries in In this table, no regulation means that the charges of airports are determined directly and opaquely by the Government. However, it is probable that most of them employ less incentivizing RoR methods. Curiously, (Graham A., 2008) remarks that while the airline industry is passing through a deregulation phase, the airport industry is to a certain extent burdensome with regulation. Notwithstanding, economic regulation can have enormous impact in the relationship between airlines and airport operators. Such impact is even greater when the debate around the sources of revenue of airports takes place, as briefly discussed in the next section SINGLE TILL VS. DUAL TILL The airport facilities and services that are considered when the airport prices are being set is a major concern of all airlines. Two alternative approaches exist. The first is the single till, whereby all revenues generated by both aeronautical and commercial activities are taken into account. In the great majority of cases, non-aeronautical services and facilities, which are becoming increasingly important sources of revenue for airports, will help reduce charges of aeronautical services. Airlines are thus strong supporters of the single till approach. (de Neufville & Odoni, 2003) Probably, the most important drawback to this approach relates to traffic growth and increasing congestion, where bringing down charges in an environment of scarce supply of resources makes no economic sense. (Graham A., 2008) The other approach is the dual till. It treats aeronautical and non-aeronautical areas as separate financial entities and focuses on the monopoly aeronautical airport services. It is however, a difficult task due to the allocation of many fixed and joint costs between both areas. Airport managers often argue that 63

65 commercial revenues reflect the premium location as opposed to monopolistic pricing, and hence, should not be subject to economic regulation. Such debate has leaded the British Civil Aviation Authority (CAA) in late 2000 to conduct a consultation paper on which approach to follow regarding price regulation in British airports. In this paper, five arguments for and against the dual till approach implementation are discussed and present an enlightening summary of this debate. It is first argued that designated airports might earn high profits due to market power in relation to commercial activities. If in one hand, the possibility of market power may exist, the CAA is not convinced it is such that economic regulation is warranted. Moreover, the single till would not be the answer, as it merely transfers the profits from this to lower the regulated charges. A second argument relates to the fact that commercial revenues are derived from the airlines passengers, and should therefore benefit from profits generated by passengers buying. The counter argument here is that single till has the effect of an additional tax on profits which serves to dampen the airports incentives to develop it efficiently. The third argument often is that the single till ensures that the price of the airport total bundle of prices is kept to competitive levels, permitting only normal returns on capital. Again, this does not protect passengers from market abuse by airports in commercial activities. Ultimately, it would only serve airlines, by transferring those rents to reduce aeronautical charges. Fourthly, it is argued that the removing the single till would generate windfall gains for airport operators. Here, the CAA shows some caution, admitting that a regulatory framework needs to be sustained over time, so that both an increase of airport charges and the overall level does not happens. The final typical argument is that the single till is simple to administer and therefore reduces regulatory intervention, rather than increasing it. CAA recognizes such advantage of the single till being relatively straightforward. Nevertheless, CAA s view is that the high level arguments in favour of the single till are not compelling. The basic presumption is that economic regulation should apply only to the core monopoly functions that an airport provides and should not be extended to cover other activities. TABLE 3-4: SINGLE OR DUAL-TILL APPROACH IN EU COUNTRIES IN 2006 (ADAPTED FROM (MARQUES & BROCHADO, 2008) ) Country Single/dual till Country Single/dual till Austria Single till Italy Dual till Belgium Single till Malta Dual till Denmark Dual till The Netherlands Dual till France Single till Portugal Single till Germany Dual/single till Spain Single till Greece Dual till Sweden Single till Hungary Single till UK Single till Ireland Single till According to (Marques & Brochado, 2008), the dual-till approach has recently gained prominence in Europe. Indeed, despite producing higher airport charges for users and stand up cost allocation issues, 64

66 dual-till regulation makes charges reflect costs more closely and maximizes the airport value. The authors further point out several studies defending the dual till to the detriment of the single-till regulation. Such is the case of (Beesley, 1999) arguing that regulation should focus on activities characterized by a natural monopoly and (Starkie, 2001) that goes farther by neglecting the need for economic regulation for the non-congested airports. Starkie defends, however, that for congested airports the application of a dual-till scheme would lead to higher aeronautical charges which would have positive effects on the allocation of scarce slot capacity and on the investment incentives. On the other hand, (Lu & Pagliari, 2004) stand for the single-till approach as welfare maximizer when compared with the dual-till method at non-congested airports. Table 3-4 above summarizes the approaches followed by European countries in The approval of economical regulation (DL 16/2009) in late 2009 in Portugal was done in order to prepare the eminent privatization of ANA. According to the Ministry of Public Works, Transports and Telecommunications (MOPTC) this model will follow international guidelines by ensuring transparent, predictable and stable rules. Another goal is to warrant aeronautical charges compatible with passenger rights and that enhance airports efficiency, thus fostering the levels of service from the airlines and passengers point of view. (MOPTC, 2009). Table 3-5 below indicates the main changes for the intervening actors with the implementation of the economical regulation. TABLE 3-5: CHARACTERISTICS OF THE NEW PORTUGUESE ECONOMIC REGULATION ON AIRPORTS' CHARGES Characteristics New model on Airports Economic Regulation Before Incentives Management efficiency; Refusal to accept proved inefficient investments by the regulatory authority; None; Single till Stability Flexibility Level of Service Commercial revenues will be used to reduce aeronautical charges; When large investment occur the model enforces tariff stability through: - Profit transfer over time; - Decomposition of revenues in fixed and variable compounds; Regulator defines an average maximum revenue per passenger in quinquennial periods; Charges adjustment according to demand in different airports and/or different times of the year; Some charges are eliminated; For each regulatory period, LoS are defined for several services (check-in, passport control, baggage claim, etc); Settlement of objectives to be met by the airport operator for each service; Reduction of aeronautical charges is partial and casuistic; No stability. When large investments are done, charges sky rocket and fall abruptly after the investment amortization; Aeronautical charges are defined on an annual basis without any sort of flexibility No regulation on the LoS; No systematic measure of LoS; The economic regulation model used is undoubtedly important for any airport s overall performance and the current discussion around which model suits the best each airport says it all. Albeit the 65

67 importance of this variable, it is not so relevant for our case study, given that all Portuguese airports operate within the same economic framework. In a near future however, with the eventual privatization of mainland and Azores islands airports, an efficacy performance comparison between them and the Madeira Islands airports maybe relevant as the later will remain under total state ownership and will not be under the scope of the new economic regulation. 66

68 4 ANALYSIS OF PORTUGUESE AIRPORTS EFFICIENCY 4.1 INSTITUTIONAL SETTING There are 37 Portuguese airports, most of which are small regional airports operating in a discontinuous fashion. The main airports are run by public enterprises, regulated by a public body, Instituto Nacional de Aviação Civil (National Institute of Civil Aviation), and this same public organization comes under the direct control of the MOPTC. This public authority manages the airports, which are situated throughout mainland Portugal and the archipelagos of Madeira and the Azores. The airports are owned by public enterprises: ANA owns the airports of mainland Portugal and the Azores, whilst ANAM owns the Madeira airports. (Barros C. P., 2008) One of the major shareholders of ANA is Parpública, a state holding. Parpública has particular relevance in the Portuguese aeronautical services since it also detains 100% of the flag-carrier TAP. In one hand, Parpública calls out attention to the need of recapitalization of the Portuguese airline group of companies and to the excessive personnel costs that represented 25% of operational costs in On the other hand, ANA s group registered a 43% profit increase in 2009 to 43 million (affected negatively by ANAM and Portway s 5 million losses each), mainly through control of operational costs. Parpública states how the low-cost segment was critical on national airports with an average 35% increase per year and having inaugurated 15 new routes in Lisbon, 26 in Porto, and 14 in Faro between 2007 and Furthermore, it points out the need of increasing efficiency in order to retain this segment of companies and passengers. (PARPÚBLICA, SA, 2009) The financial background of both companies is of great importance when considering the construction of the NAL. Whereas the privatization of a solid asset like ANA is sure to attract many investors, the uncertainty regarding the future of TAP along with the planning of a new airport that is TAP s main hub, responsible for 39% and 44% of total Lisbon s passengers and movements respectively, brings many doubts to this project. ANA Aeroportos de Portugal S.A is the concessionaire of the 3 airports in mainland Portugal (Porto, Lisbon and Faro) and 4 airports in the Azores Islands. Additionally, ANA holds equity stakes of 70% in ANAM (the airport operator of Madeira Islands) 49% in ADA (the Macao airport operator) and 100% in the company Portway (one of the two ground handling companies operating in Portuguese airports). As part of a consortium, ANA is also directly involved in the management of 12 regional airports in Peru. The company also holds a share of approximately 84% in NAER, S.A., a special purpose organization responsible for the studies and procedures inherent to the launching of the tender for the privatization of ANA and the construction of the new airport. (MOPTC, 2008). ANA s shares are divided between the Ministry of Finance (31,44%) and a Parpública (68,56%) a state holding focused on managing financial assets and real estate, support to public investment and to 67

69 the program of non strategic assets alienation. (ANA, 2009b). ANAM was first created in 1991 as a subsidiary of ANA s group of companies. The remaining 30% shares are divided between the Madeira Regional Government (20%) and the Portuguese State (10%). Below, figure 4-1 represents the structure of the main Portuguese airport operator. 4-1: STRUCTURE OF ANA S GROUP OF COMPANIES (SOURCE: (ANA, 2009B)) 4.2 DATA COLLECTION Regarding the collection of data for our case study, some considerations are in order. Firstly, the numbers of ATMs and passengers used in our model relate to commercial flights. All information was drawn from ANA s yearly statistics report. In the five busiest airports, commercial ATMs represent in average 92,3% of total flights. The significance of commercial flights is higher in Lisbon and Porto airports with nearly 96% whereas in touristic oriented airports such as Faro and Funchal the ratio drops to 88%. Such drop is explained by the increase of private aircrafts, related with the aeronautical services provided to private clients. Ponta Delgada airport sticks to the average with slightly over 93%. On the other hand, commercial passengers represent in average 99,9% of total passengers. In Figure 4-2 we can observe the annual variations between 2005 and 2009 on both passenger and aircraft commercial movements. Porto airport registers the highest increases both on passenger and movements annual variations. Lisbon airport has similar trends to Porto despite being rather inferior. Faro airport also shows high annual increases but remains the only airport with two years in the third quadrant. Curiously, 2008 was the only year that Funchal airport registered positive variations both on passengers and movements, when the liberalization of routes to Funchal occurred. Moreover, Faro is the only airport that registered two annual decreases in passenger traffic between 2005 and Ponta Delgada airport is the only infrastructure with constant positive annual variations of aircraft movements despite two years with decrease in the number of passengers. The overall picture reflects the increasing effect on the demand side with an average increase of 1,7% on ATM and 3,4% on passengers traffic. The other side of coin is how economical downturns play a major role in the industry, as 2009 proved to be. 68

70 4-2: ANNUAL VARIATIONS OF PAX AND ATM BETWEEN 2005 AND 2009 (SOURCE: ANA) 20,0% 07/06 Annual Variation of Commercial PAX 15,0% 10,0% 5,0% 0,0% -5,0% 06/07 09/08 09/08 09/08 06/05 08/07 08/07 06/05 08/07 09/08 06/05 07/06 08/07 07/06 08/07 06/05 07/06 06/05 09/08-10,0% -10% -8% -6% -4% -2% 0% 2% 4% 6% 8% 10% 12% Annual Variation of Commercial ATM LIS OPO FAO FUN PDL Secondly, data regarding LCC s operations was obtained from ANA and ANAM, but with some nuances. The initial concept was to gather data on a monthly basis in order to obtain a wider data panel that could define a more accurate efficient frontier. Evidently, this would lead to peak summer months being more efficient, with seasonality distorting the analysis. Nonetheless, data collection followed the principle of gathering information with the highest level of disaggregation possible in order to reduce the number of eventual assumptions. All airports with exception to Porto collaborated by providing the monthly distribution of LCC s traffic. Notwithstanding, the yearly number of LCC s passengers and movements was found as well as 2009 s LCCs monthly traffic at Porto (ANA, 2009a). For Lisbon airport, we also had to estimate LCC s monthly traffic distribution between 2005 and 2006 (ANA, 2006). The procedure followed was the same as for Porto airport in the period between 2005 and Such forced to introduce an assumption regarding the monthly distribution of the remaining years. For this purpose, three options were considered. The first was to use data gathered from STATFOR (Eurocontrol s air traffic statistics) which identifies the different segments of all Instrument Flight Rules (IFR) flights departing from any European country. This option required further assumptions, such as doubling the traffic to obtain total traffic and the number of passengers per movement to obtain passenger traffic. While assuming distribution between departures and arrivals could be consensual, the same 69

71 did not hold for the second assumption. For instance, differences up to 12% were found in the annual average ATM when comparing Porto to Faro given that STATFOR takes into account typical charter companies (Monarch, Thomas Cook) for LCC traffic, that do not operate at Porto, resulting in lower numbers of passenger per movement. The other two options consisted in using LCC s yearly traffic between 2005 and 2008 from ANA. LCC s monthly distribution at Porto airport could then assumed to be identical either to the distribution of total commercial traffic at Porto or to Faro s monthly distribution of low-cost traffic. It was found that similarities between LCC and total traffic distribution along the year are stronger in Faro than in Porto. This is mostly due to the continuous increase of LCC relevance at Faro airport in comparison to other airports. In December 2009 at Faro airport, 72% of total movements and 83% of total passengers were derived from the LCC segment, whereas for the same period in Porto the shares were only 31% and 49%, respectively. In addition, Faro s high seasonality would also compromise such assumption. The ratio between the busiest and slowest month in terms of monthly passenger traffic in Faro is nearly five, whilst in Porto is two. Hence, between 2005 and 2008, the monthly distribution of LCC traffic in Porto airport followed the same monthly distribution of total traffic in that period. The same line of reasoning was applied for Lisbon s monthly traffic distribution in the years of 2005 and The third aspect relates to the availability of infrastructures between 2005 and For Faro, Ponta Delgada and Funchal airports no changes were registered, hence data is constant over time. Lisbon and Porto, on the other hand, have had several renewal and expansion works in that period, which obliged to a more detailed scrutiny. As in traffic data, infrastructure data was gathered on a monthly basis, being considered an average yearly value for our analysis. Lisbon airport s second passenger terminal was finished in July 2007, and therefore the landside increase of capacity counts thereafter, namely with the addition of 22 check-in desks and 12 boarding gates. Baggage belts capacity also suffered improvements in June 2008, by converting 3 out of 7 linear belts to carousel-shaped type belts, increasing capacity in the former by 75% from 400 to 700 baggage per hour. The extensive usage of the number of baggage belts as input in efficiency literature appears to be inaccurate to the extent that Lisbon remained with 7 belts while increasing its overall handling capacity from 2800 to 3700 baggage per hour. On the airside, the increase in runways declared capacity was done gradually, passing from 32 to 36 ATM/h, with an intercalary period in 2006 with 34 ATM/h. Furthermore, an increase of 12 remote parking stands occurred in mid 2008, from 44 to 56 remote stands. Porto airport also suffered structural works during the period in study. Unfortunately, there was not the same availability from airport managers to explain when the changes occurred as it happened in Lisbon. To that respect, contradictory information was found. According to (Tribunal de Contas, 2009), while the passenger capacity increase in the passenger building was achieved before the study period for purposes of the 2004 European Football Cup, works in the passenger building continued until January 2007 with the conclusion of the escalators. ANA replied to this report alleging that regardless the 70

72 conclusion date of all physical works, the spaces that suffered interventions were made available as soon they were ready to use. Nevertheless, the only infrastructures relevant for our case study that appear to have suffered interventions during our period are the air bridges (increasing from 9 to 11 in November 2005) and the rehabilitation of the runway (resulting in increased declared capacity from 16 to 20 ATM/h by May 2006). Finally, we have assumed no change regarding available areas in the terminal for the period in study due to lack of information. The final aspect relates to the operational characteristics of each airport, namely how LCC s traffic is counted. The marketing department of Funchal airport informed that the LCC segment is registered according to the ELFAA s member list. On the other hand, Faro s marketing department made available an incomplete list of the main LCC operating in ANA s airports by 2010, further explaining that some carriers (airberlin, Brussels airlines) recently asked to be classified as full-service carrier. That is the case of airberlin, which counts for LCC traffic in Ponta Delgada in 2009, but requested to change its classification in February The segmentation appears to be similar to the one used by STATFOR 3, counting typical charter companies that operate under scheduled flights as well as typically low-fares airlines, that despite having low price tickets, still include some services traditionally paid in LCCs. This point arises some issues when we intent to perform a benchmarking analysis. If we take for instance the year of 2009, the airlines Thomsonfly, Air Berlin and Thomas Cook had a combined ATMs market share of 9% in Funchal, whilst ELFAA s members easyjet and Transavia combined 12,4% of market share (ANAM, 2009). According to Faro s criteria, this leads to a reduction of Funchal s LCC market share to nearly half. On the other hand, in 2009 airlines non-members of ELFAA (Germanwings, NIKI, Aer Lingus, SkyEurope, bmibaby, Blue Air) operated regularly in Lisbon and were counted as nearly 21% of the low-cost traffic (or 3,15% of total traffic). Moreover, whereas airberlin was by the time the only LCC in Ponta Delgada airport, in Faro, that claims to have 80% of low-cost traffic, more than 30% of the movements was generated by non-elfaa members (ANA, 2009d). One cherry on the top of this problem is for instance Porto considering Tuifly as a LCC in 2009 whereas Faro did not. In practice, this means that Funchal airport s low-cost traffic is diminished in comparison to Azores and mainland airports. If in one hand, we could have considered all regular low-fares flights in Funchal as low-cost flights, on the hand, other airports do not even make available what criteria is used in the segmentation of markets. In addition, the disaggregation to monthly traffic would impose more assumptions regarding each airline s traffic distribution in each airport. Hence, for simplicity reasons, data used in our efficiency model will follow the same criteria as obtained from the original source. Table 4-1, summarizes the airlines operating in Portuguese airports counted in the low-cost segment. Companies are presented in a cumulative perspective. In other words, companies involved in

73 mergers (LTU bought by air berlin in 2007), bankruptcies (Sky Europe in 2009), or punctual routes are all present in the table. IATA CODE LIS OPO FAO FUN PDL TABLE 4-1: LCC OPERATING IN PORTUGUESE AIRPORTS BETWEEN 2005 AND 2009 LCCs easyjet, Vueling, Brussels Airlines, Clickair, Germanwings, Air Berlin/Flyniki, Aer Lingus, Monarch Scheduled, LTU, Thomsonfly, bmibaby, Centralwings 1, Blue Air, Skyeurope 4 Air Berlin, easyjet, Ryanair, Tuifly, Transavia France Ryanair, Aer Lingus, easyjet, easyjet Switzerland, bmibaby, flybe, Jet2.com, Norwegian, Jet air, Blue air, Monarch, Niki, Jetairfly, Wizz air (Hungary and Ukraine), SmartWings, Cimber Sterling, Transavia, Transavia.com (France and Denmark), Germanwings, Arkefly, Vueling easyjet, flybe, Jet2.com, Norwegian Air Shuttle, Ryanair, Sverige Flyg, transavia.com, Vueling, Wizz Air Air Berlin To conclude the considerations regarding data collection, there are two aspects discussed both with Lisbon s and Funchal s marketing responsible worth mentioning. The first concerns easyjet s operations in Lisbon. With an overall ATM s market share of 9,5% and around 64% in its segment in 2009 (ANA, 2009c), easyjet only uses the low-cost-purposed Terminal 2 built in 2007 for flights to Funchal, all other flights being operated in Terminal 1 with necessary use of air bridges. This is because remote parking stands in Terminal 2 are too far away, and as the boarding process takes longer thus increases the turnaround time. All added up, it becomes cheaper to pay for the expensive air bridges that allow shorter turnaround times, than to use the low-cost terminal. The second aspect relates to transit numbers in Funchal airport. Adverse weather conditions often oblige traffic intended to land in Funchal to divert to Porto Santo, also operated by ANAM. This solution is obviously preferred in comparison to diverting aircrafts to Canarias Islands, as passenger and landing fees stay at home. Nonetheless, it creates bogus traffic to and from Porto Santo airport. Albeit the small individual importance of each of these issues, these examples show how only with the proper understanding of each firm s reality it is possible to perform a detailed benchmark of an industry. 4.3 MODEL FORMULATION The efficiency analysis will follow the DEA methodology described in section For this purpose, we will use DEAP, a MS-DOS freeware. This software allows applying the CRS, VRS and Malmquist DEA models 4. In our assessment of Portuguese airports efficiency the VRS model will be used in order to examine the existence of scale efficiency. The goal is to understand which airports fully 1999) 4 For more details see (Charnes, Cooper, & Rhodes, 1978), (Banker, Charnes, & Cooper, 1984) and (Coelli, 72

74 explore their fixed quantity of resources (airside and landside infrastructures) while increasing their production (number of low-cost passengers and aircraft movements). Hence, the above-mentioned model will be output oriented. In order to explain the VRS output oriented approach, we will start by addressing Farrell s inputbased measure of technical efficiency under constant returns to scale. In our data, we have inputs and outputs on each airports (DMUs). For the i-th DMU, these are represented by the vectors and, respectively. The input matrix and the output matrix thus represent all airports. For each DMU we would like to obtain a measure of the ratio of all outputs over all inputs, such as, where is an 1 vector of output weights and is an 1 vector of input weights (Coelli, 1999). The mathematical LP problem that selects optimal weights is as follows:,,: st 1, = 1,2,.., (1), 0, One particular problem with this ratio formulation is that it has an infinite number of solutions since if, is a solution, then, is another solution. The multiplier form of the LP solves this problem by imposing the constraint =1, transforming (1) in:,,: st =1, 0, = 1,2,.., (2), 0, The equivalent envelopment form can be obtained using the duality property of LP, converting the maximization of optimal weights through the ratio of inputs and outputs into the following model:,,: st + 0, 0, (3) 0, Where is the efficiency score for the i-th DMU and is a 1 vector of constants. As mentioned in section , the VRS model proposed by (Banker, Charnes, & Cooper, 1984) will permit us the 73

75 calculation of technical efficiency devoid of scale efficiency effects such as imperfect competition or finance constraints. The CRS model can be easily modified to account for VRS by adding the convexity constraint 1 1 to (3) where 1 is an 1 vector of ones. The inequality in this constraint provides technical efficiency scores that are greater than or equal to those obtained using the CRS model. Furthermore, it plots the non-increasing returns to scale (NIRS) frontier (defined by OCDE in figure 3-4) that allows the identification of the DMUs nature of scale inefficiencies. In other words, if a DMU has a NIRS technical efficiency score equal to the VRS score, then decreasing returns to scale apply. If they are unequal (as will be the case for the point F in figure 3-4) then increasing returns to scale exist for that DMU (Coelli, 1999). The output oriented model is very similar to the input oriented. In order to measure technical inefficiency as a proportional increase in output with input quantities held constant, we must consider the following LP model:,,: st + 0, 0, 1 1 (4) 0, where 1, and 1 is the proportional increase in outputs that could be achieved by the i th DMU. Note that 1 defines a technical efficiency score which varies between zero and one. The VRS and CRS models estimate exactly the same efficiency frontier and only measures associated with inefficient DMUs may differ between the two models. By calculating both the CRS and VRS efficiency scores, DEAP further computes the ratio between the two scores that measures scale efficiency. Literature typically separates the VRS-DEA technical efficiency into two components scale efficiency and pure technical efficiency. (Coelli, 1999) SELECTION OF INPUTS AND OUTPUTS In chapter 3, several drivers for airport efficiency were identified. One problem persists though. According to (Lin & Hong, 2006), a good rule of thumb to the necessary number of DMUs points to no less than the double of the sum of inputs and outputs. Since our assessment comprises five airports only, it is likely to verify efficiency overstatements. This was the main reason to disaggregate annual traffic into monthly traffic. Instead of having five airports in five years (25 observations) we would have 25x12 observations. Since we have no information regarding whether infrastructures are shut down when not used or not, airports are expected to be more efficient on summer time, typical peak load periods with 74

76 more emphasis in Faro and Ponta Delgada. In theory, we will be assessing the efficiency of five airports between 2005 and In practice, either we can treat each airport in each year (25 DMUs) as an individual firm or as a panel data of five airports (5 DMUs) in five years or even in sixty months. Whereas the former will create a more detailed efficiency frontier, the later will compute the average efficiency score of each airport for the period in study in comparison to the average values of other airports. To work out this pitfall in our study, a restriction of the number of outputs and inputs is in order. From the output perspective, it is inevitable to consider the number of low-cost ATMs and passengers in addition to the complementary non-low-cost traffic. Alternately, we will also consider only total commercial traffic, thus reducing outputs by half. From the input viewpoint, the selection of which variables to use in our model was more difficult. For once, important aspects such as capital cost or labour cost were not considered due to lack of financial information on Portuguese airports, thus focusing our analysis on the operational efficiency. In the airside, remote parking stands were deferred to total parking stands (TPS) for two reasons. Whereas the first relates to the fact that all stands in Funchal and Ponta Delgada airports are remote, the second has to do with easyjet s operations in Lisbon happening in the same framework as full service carriers. The same line of reasoning was applied to the total number of boarding gates (TBG), since it was initialled intended to discriminate whether boarding gates were bridged or not. Furthermore, to consider the number of runways seemed rather doubtful. Only Lisbon has two operational runways, but the second has little use and more prone to military flights, all other airports having one runway. In order to better characterize airfield capacity we chose to use the runway declared capacity (RDC) of each airport. It is particularly relevant in Porto and Lisbon, where expansion works that did not comprise the construction of new runways have contributed to the airfield capacity. On the landside, and similarly to airfield capacity, the overall capacity of baggage collection systems (BSC) was considered as opposed to the number of baggage belts. In this matter, great discrepancies were found to the extent that Faro has more arrival baggage capacity than Lisbon airport. Moreover, it was also difficult to obtain precise data regarding the use of self-service check-in kiosks. To this respect, only Lisbon airport cared to inform that the common-use kiosks belong to the main airline alliances, namely Star Alliance and SkyTeam and that no LCC has implement CUSS kiosks by this time. Hence, only traditional check-in desks (CID) were considered. Finally, and despite the availability of data concerning the different areas of passenger terminals (check-in, boarding and total), only total terminal area (TTA) was considered as a probable key factor in the production of outputs. Table 4-2 below summarizes the yearly aggregated data for the five busiest airports in Portugal between 2005 and It includes all above mentioned inputs and outputs, albeit some variations in the will be done in order to better characterize the industry reality. For instance, the outputs column has the total commercial number of aircraft movements and passengers disaggregated in LCC and non-lcc traffic and our model will also run total commercial and passenger values as outputs. Furthermore, data is 75

77 presented in a yearly basis, but we will also assess airports efficiency based on monthly evolution. All collected data is presented in Appendix 2. DMU # TABLE 4-2: YEARLY DISAGGREGATED DATA USED IN THE VRS-DEA OUTPUT-ORIENTED MODELS DMU NAME OUTPUTS INPUTS ATM PAX TPS RDC BSC CID TBG TTA NON-LCC LCC NON-LCC LCC # ATM/h bag/h # # m 2 1 LIS_ LIS_ LIS_ LIS_ LIS_ OPO_ OPO_ OPO_ OPO_ OPO_ FAO_ FAO_ FAO_ FAO_ FAO_ FUN_ FUN_ FUN_ FUN_ FUN_ PDL_ PDL_ PDL_ PDL_ PDL_

78 4.4 DISCUSSION OF RESULTS Our analysis of Portuguese airport s efficiency was conducted bearing in mind all inherent difficulties with the application of this model to such small data panel. Nevertheless, our intent to analyse the influence of LCCs in Portuguese airports lead to the reformulation of our assessment according to the difficulties found along the way. For this purpose, only output data was disaggregated in order to understand whether LCCs have influence or not on Portuguese airports, by assessing either total commercial traffic or LCC plus non-lcc traffic. In practice, this means that models using disaggregated output data will have the four outputs (ATM LCC ; ATM Non-LCC ; PAX LCC, PAX Non-LCC ) whereas the others will have only two (ATM TOTAL, PAX TOTAL ), whilst all six models will use the same inputs. Furthermore, we considered three approaches that are only distinct in a temporal perspective, that were tested under the two hypotheses of output data (dis)aggregation, resulting in a total of six models. Table 4-3 resumes the characterization of the results obtained from the above mentioned models. Hence, in this section we will analyze the results of each of such models. The disaggregation of output data corresponds to models with an odd number (1, 3 and 5), whilst models 2, 4 and 6 correspond to each one of three approaches under the assumption of aggregated output data. It is important to notice that we have conducted our analysis maintaining the same six inputs in all models total-parking stands, declared capacity; baggage belts capacity, total boarding gates, check-in desks and total terminal area. All results are compiled in Annex 3. TABLE 4-3: MODEL CHARACTERIZATION ON THE VRS-DEA EFFICIENCY BENCHMARK Approach description 1) Analysis on seasonality influence (5 DMUs in 60 months) 2) Each airport in each year as an individual firm (25 DMUs) Disaggregated (LCC + Non-LCC) Output Data Aggregated (Total) Model 1 Model 2 Model 3 Model 4 3) Panel data (5 DMUs in 5 years) Model 5 Model APPROACH 1: MONTHLY ANALYSIS Firstly, models 1 and 2 consider the influence of seasonality on the analysis of airport s efficiency. In order to perform the evaluation of the five airports efficiency we considered the 60 periods that correspond to the total number of months between 2005 and The results are presented below in figure 4-3. Albeit our large sample of 300 observations, DEAP presents only the average efficiency score of each airport. 77

79 If we take for sake of comparison figure -5, that illustrates 2009 s monthly passenger distribution in all five airports, one may cross relate high efficiency scores obtained at Faro and Ponta Delgada airports with summer peak traffic. For this reason, airports operating under the optimum efficiency level are located in the region of decreasing returns to scale. That is the case of Lisbon, Porto and Funchal in model 1 and all airports except for Ponta Delgada in model 2. Moreover, airports that more than quintuple the passenger s flow in summer time are strongly advised to implement the share use of facilities or even shut down facilities that are not used in low-demand periods, thus increasing its efficiency. Scale efficiency scores are found to decrease in the busiest mainland airports when aggregated output data is considered, meaning that LCCs have a positive effect in the scale of airports production. The biggest drop in scale efficiency happens in Faro (from 100% to 62,7%) and is most likely related with the high share of LCC traffic in this airport. Porto airport registers the worst efficiency scores albeit its high annual increases in commercial traffic. The misuse of landside infrastructures at Porto airport from the LCC point of view can be an important explanation for such inefficiency. Overall, the average VRS TE score of Portuguese airports drops only 0,2% from to 94,6% in model 1 to 94,4% in model 2. On the other hand, it is found a great increase in global standard deviation of scale efficiency scores, passing from 12,1% in model 1 to 23,7% in model 2, which may be an important indicator of LCCs influence in airport s efficiency. Global statistics are for models 1 and 2 are presented below. 4-3: VRS-DEA OUTPUT ORIENTED: MODELS 1 (LEFT) AND 2 (RIGHT) 100% 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% LIS OPO FAO FUN PDL CRS TE VRS TE SCALE E. 0% LIS OPO FAO FUN PDL CRS TE VRS TE SCALE E. TABLE 4-4: GLOBAL STATISTICS FOR MODELS 1 (DISAGGREGATED) AND 2 (AGGREGATED) MODEL 1 MODEL 2 CRS TE VRS TE SCALE E. CRS TE VRS TE SCALE E. Average 86,0% 94,6% 90,8% 68,5% 94,4% 72,2% Standard Deviation 14,2% 6,5% 12,1% 24,4% 6,6% 23,7% 78

80 Ultimately, disaggregation of output data has shown to have an important influence on airport s scale efficiency without major disruptions on each airport s technical efficiency score APPROACH 2: EACH AIRPORT IN EACH YEAR AS AN INDIVIDUAL FIRM The second approach results of the consideration of 25 DMUs, or in other words, the treatment of each one of the five airports in all five years as an individual firm. The obtained efficiency scores are represented graphically in figures 4-4 and 4-5. Under this approach, Lisbon airport stands out as the most efficient airport in the national framework on both variations of ouput aggregation of data. Moreover, Lisbon airport in 2009 is the only airport operating under decreasing returns to scale in model 3, while all others are operating under increasing returns to scale. As Faro airport between 2007 and 2009 operates fully efficiently in model 3 it is operating under CRS. Model 4 however, indicates that all airtports apart from Lisbon between 2006 and 2009 are operating under increasing returns to scale. 4-4: VRS-DEA OUTPUT ORIENTED: MODEL 3 100% 80% 60% 40% 20% 0% LIS OPO FAO FUN PDL CRS TE VRS TE SCALE E. Consistent with models 1 and 2, Porto maintains the worst efficiency scores with an average VRS TE value of 94,8% in model 3 and 90,2% in model 4, while the remaining Portuguese airports attain up to 99,2% and 98,1% respectively is for Porto airport the best performing year, when LCCs represented merely 4,4% of commercial passengers and 9,3% of aircraft movements. The continuous growth of LCC traffic between 2005 and 2009 moving from passengers and nearly ATMs to almost 2 million passengers and around ATMs has surely helped to counterweight the loss of market power of non-lccs, that have only registered an increase of aircraft movements in This is parcially explained in model 3 (that uses disaggregated output data) where we can observe the impact of 79

81 LCC s growth in the steady increase of VRS TE, whereas in model 4 (using aggregated ouput data) VRS TE decreases from 100% in 2005 to around 85% in the following years, rising to 94% with growth of non- LCC traffic in The high VRS TE attained in the year 2005 is also explained by the existing declared capacity of 16 ATM/h in 2005, expanded to 19 and 20 ATM/h in the following years, all other inputs remaining equal. Another relevant aspect has to do with consistent efficiency growth at Faro and Porto in model 3, likely related to the increase of output production of the LCC segment as oposed to managerial practices. On the other hand, Funchal airport presents slightly unstable VRS TE results in model 4, appearing to be more sensible to output variation with 2009 being a particular bad year, while all inputs remained equal. Regarding scale efficiency one can point out the continous increase in all airports of model 4, where Funchal airport presents the highest inefficiency, followed by Porto and Ponta Delgada airports 4-5: VRS-DEA OUTPUT ORIENTED: MODEL 4 100% 80% 60% 40% 20% 0% LIS OPO FAO FUN PDL CRS TE VRS TE SCALE E.. An interesting case study in this approach is Funchal airport. In one hand the decrease of non- LCC ATMs from in 2005 to in 2009 while loosing nearly passengers to 2 million passengers, appears to suffer little influence from the liberalization of routes to Funchal in October On the other hand, LCC s passenger traffic rose from in late 2007 to almost in 2009 while the number of aircraft movements moved from 422 to 2.612, clearly supporting the low growth of Funchal airport in terms of passengers but not in in terms of ATM. Furthermore, as mentioned previously, the assumptions regarding the number of LCC s operations at Funchal airport are different as of other airports with the natural prejudice of counting only for ELFAA members. It is thus curious to note that scale efficiency in model 3 shifts down from 70% to 60%, whilst in model 4 raises from 59% to 62%. Since inputs remain constant between 2005 and 2009, scale efficiency in model 3 appears to be negatively affected by the decrease of the non-low-cost segment along with the inability of the low-cost 80

Exploratory analysis on LCC potential to influence airport efficiency Sérgio Domingues. AIRDEV Seminar Lisbon, October 20th 2011

Exploratory analysis on LCC potential to influence airport efficiency Sérgio Domingues. AIRDEV Seminar Lisbon, October 20th 2011 Exploratory analysis on LCC potential to influence airport efficiency Sérgio Domingues AIRDEV Seminar Lisbon, October 20th 2011 1 Contents Introduction Drivers of airport efficiency Data Collection Discussion

More information

Impact of Landing Fee Policy on Airlines Service Decisions, Financial Performance and Airport Congestion

Impact of Landing Fee Policy on Airlines Service Decisions, Financial Performance and Airport Congestion Wenbin Wei Impact of Landing Fee Policy on Airlines Service Decisions, Financial Performance and Airport Congestion Wenbin Wei Department of Aviation and Technology San Jose State University One Washington

More information

Performance and Efficiency Evaluation of Airports. The Balance Between DEA and MCDA Tools. J.Braz, E.Baltazar, J.Jardim, J.Silva, M.

Performance and Efficiency Evaluation of Airports. The Balance Between DEA and MCDA Tools. J.Braz, E.Baltazar, J.Jardim, J.Silva, M. Performance and Efficiency Evaluation of Airports. The Balance Between DEA and MCDA Tools. J.Braz, E.Baltazar, J.Jardim, J.Silva, M.Vaz Airdev 2012 Conference Lisbon, 19th-20th April 2012 1 Introduction

More information

Airport Monopoly and Regulation: Practice and Reform in China Jianwei Huang1, a

Airport Monopoly and Regulation: Practice and Reform in China Jianwei Huang1, a 2nd International Conference on Economics, Management Engineering and Education Technology (ICEMEET 2016) Airport Monopoly and Regulation: Practice and Reform in China Jianwei Huang1, a 1 Shanghai University

More information

AIRPORT OF THE FUTURE

AIRPORT OF THE FUTURE AIRPORT OF THE FUTURE Airport of the Future Which airport is ready for the future? IATA has launched a new activity, working with industry partners, to help define the way of the future for airports. There

More information

Measure 67: Intermodality for people First page:

Measure 67: Intermodality for people First page: Measure 67: Intermodality for people First page: Policy package: 5: Intermodal package Measure 69: Intermodality for people: the principle of subsidiarity notwithstanding, priority should be given in the

More information

CRITICAL FACTORS FOR THE DEVELOPMENT OF AIRPORT CITIES. Mauro Peneda, Prof. Rosário Macário AIRDEV Seminar IST, 20 October 2011

CRITICAL FACTORS FOR THE DEVELOPMENT OF AIRPORT CITIES. Mauro Peneda, Prof. Rosário Macário AIRDEV Seminar IST, 20 October 2011 CRITICAL FACTORS FOR THE DEVELOPMENT OF AIRPORT CITIES Mauro Peneda, Prof. Rosário Macário AIRDEV Seminar IST, 20 October 2011 Introduction Airports are becoming new dynamic centres of economic activity.

More information

REAUTHORISATION OF THE ALLIANCE BETWEEN AIR NEW ZEALAND AND CATHAY PACIFIC

REAUTHORISATION OF THE ALLIANCE BETWEEN AIR NEW ZEALAND AND CATHAY PACIFIC Chair Cabinet Economic Growth and Infrastructure Committee Office of the Minister of Transport REAUTHORISATION OF THE ALLIANCE BETWEEN AIR NEW ZEALAND AND CATHAY PACIFIC Proposal 1. I propose that the

More information

ACI EUROPE POSITION PAPER. Airport Slot Allocation

ACI EUROPE POSITION PAPER. Airport Slot Allocation ACI EUROPE POSITION PAPER Airport Slot Allocation June 2017 Cover / Photo: Madrid-Barajas Adolfo Suárez Airport (MAD) Introduction The European Union s regulatory framework for the allocation of slots

More information

THIRTEENTH AIR NAVIGATION CONFERENCE

THIRTEENTH AIR NAVIGATION CONFERENCE International Civil Aviation Organization AN-Conf/13-WP/22 14/6/18 WORKING PAPER THIRTEENTH AIR NAVIGATION CONFERENCE Agenda Item 1: Air navigation global strategy 1.4: Air navigation business cases Montréal,

More information

De luchtvaart in het EU-emissiehandelssysteem. Summary

De luchtvaart in het EU-emissiehandelssysteem. Summary Summary On 1 January 2012 the aviation industry was brought within the European Emissions Trading Scheme (EU ETS) and must now purchase emission allowances for some of its CO 2 emissions. At a price of

More information

Hotel Investment Strategies, LLC. Improving the Productivity, Efficiency and Profitability of Hotels Using Data Envelopment Analysis (DEA)

Hotel Investment Strategies, LLC. Improving the Productivity, Efficiency and Profitability of Hotels Using Data Envelopment Analysis (DEA) Improving the Productivity, Efficiency and Profitability of Hotels Using Ross Woods Principal 40 Park Avenue, 5 th Floor, #759 New York, NY 0022 Tel: 22-308-292, Cell: 973-723-0423 Email: ross.woods@hotelinvestmentstrategies.com

More information

Airport Slot Capacity: you only get what you give

Airport Slot Capacity: you only get what you give Airport Slot Capacity: you only get what you give Lara Maughan Head Worldwide Airport Slots 12 December 2018 Good afternoon everyone, I m Lara Maughan head of worldwide airports slots for IATA. Over the

More information

An Econometric Study of Flight Delay Causes at O Hare International Airport Nathan Daniel Boettcher, Dr. Don Thompson*

An Econometric Study of Flight Delay Causes at O Hare International Airport Nathan Daniel Boettcher, Dr. Don Thompson* An Econometric Study of Flight Delay Causes at O Hare International Airport Nathan Daniel Boettcher, Dr. Don Thompson* Abstract This study examined the relationship between sources of delay and the level

More information

Schedule Compression by Fair Allocation Methods

Schedule Compression by Fair Allocation Methods Schedule Compression by Fair Allocation Methods by Michael Ball Andrew Churchill David Lovell University of Maryland and NEXTOR, the National Center of Excellence for Aviation Operations Research November

More information

Low-Cost Carrier Passengers at Airports Knowing Their Needs and Expectations to Enhance the Passenger Experience

Low-Cost Carrier Passengers at Airports Knowing Their Needs and Expectations to Enhance the Passenger Experience Low-Cost Carrier Passengers at Airports Knowing Their Needs and Expectations to Enhance the Passenger Experience Summary of a Research Study Hamburg, Mai 2017 By Torsten Hentschel & Gesa Klingenberg Introduction

More information

AIR TRANSPORT MANAGEMENT Universidade Lusofona January 2008

AIR TRANSPORT MANAGEMENT Universidade Lusofona January 2008 AIR TRANSPORT MANAGEMENT Universidade Lusofona Introduction to airline network planning: John Strickland, Director JLS Consulting Contents 1. What kind of airlines? 2. Network Planning Data Generic / traditional

More information

MODAIR. Measure and development of intermodality at AIRport

MODAIR. Measure and development of intermodality at AIRport MODAIR Measure and development of intermodality at AIRport M3SYSTEM ANA ENAC GISMEDIA Eurocontrol CARE INO II programme Airports are, by nature, interchange nodes, with connections at least to the road

More information

Antitrust Law and Airline Mergers and Acquisitions

Antitrust Law and Airline Mergers and Acquisitions Antitrust Law and Airline Mergers and Acquisitions Module 22 Istanbul Technical University Air Transportation Management, M.Sc. Program Air Law, Regulation and Compliance Management 12 February 2015 Kate

More information

REGULATORY POLICY SEMINAR ON LIBERALIZATION POLICY AND IMPLEMENTATION PORT OF SPAIN, TRINIDAD AND TOBAGO, APRIL, 2004

REGULATORY POLICY SEMINAR ON LIBERALIZATION POLICY AND IMPLEMENTATION PORT OF SPAIN, TRINIDAD AND TOBAGO, APRIL, 2004 REGULATORY POLICY SEMINAR ON LIBERALIZATION POLICY AND IMPLEMENTATION PORT OF SPAIN, TRINIDAD AND TOBAGO, 27-29 APRIL, 2004 JAMAICA S EXPERIENCE WITH AIR TRANSPORT LIBERALIZATION INTRODUCTION Today, the

More information

Airport Systems: Planning, Design, and Management

Airport Systems: Planning, Design, and Management Airport Systems: Planning, Design, and Management Richard de Neufville AmedeoR. Odoni McGraw-Hill New York Chicago San Francisco Lisbon London Madrid Mexico City Milan New Delhi San Juan Seoul Singapore

More information

NOTES ON COST AND COST ESTIMATION by D. Gillen

NOTES ON COST AND COST ESTIMATION by D. Gillen NOTES ON COST AND COST ESTIMATION by D. Gillen The basic unit of the cost analysis is the flight segment. In describing the carrier s cost we distinguish costs which vary by segment and those which vary

More information

Recommendations on Consultation and Transparency

Recommendations on Consultation and Transparency Recommendations on Consultation and Transparency Background The goal of the Aviation Strategy is to strengthen the competitiveness and sustainability of the entire EU air transport value network. Tackling

More information

Empirical Studies on Strategic Alli Title Airline Industry.

Empirical Studies on Strategic Alli Title Airline Industry. Empirical Studies on Strategic Alli Title Airline Industry Author(s) JANGKRAJARNG, Varattaya Citation Issue 2011-10-31 Date Type Thesis or Dissertation Text Version publisher URL http://hdl.handle.net/10086/19405

More information

IATA ECONOMIC BRIEFING FEBRUARY 2007

IATA ECONOMIC BRIEFING FEBRUARY 2007 IATA ECONOMIC BRIEFING FEBRUARY 27 NEW AIRCRAFT ORDERS KEY POINTS New aircraft orders remained very high in 26. The total of 1,834 new orders for Boeing and Airbus commercial planes was down slightly from

More information

BUSINESS INTELLIGENCE Airport Retail Study May 2007

BUSINESS INTELLIGENCE Airport Retail Study May 2007 BUSINESS INTELLIGENCE Airport Retail Study May 2007 The pursuit of knowledge Last month Moodie International acquired the assets of acclaimed business intelligence title The Airport Retail Study from its

More information

Transforming Intra-African Air Connectivity:

Transforming Intra-African Air Connectivity: z Transforming Intra-African Air Connectivity: The Economic Benefits of Implementing the Yamoussoukro Decision PREPARED FOR IATA in partnership with AFCAC and AFRAA PREPARED BY InterVISTAS Consulting LTD

More information

ACI EUROPE POSITION. A level playing field for European airports the need for revised guidelines on State Aid

ACI EUROPE POSITION. A level playing field for European airports the need for revised guidelines on State Aid ACI EUROPE POSITION A level playing field for European airports the need for revised guidelines on State Aid 16 June 2010 1. INTRODUCTION Airports play a vital role in the European economy. They ensure

More information

ARRIVAL CHARACTERISTICS OF PASSENGERS INTENDING TO USE PUBLIC TRANSPORT

ARRIVAL CHARACTERISTICS OF PASSENGERS INTENDING TO USE PUBLIC TRANSPORT ARRIVAL CHARACTERISTICS OF PASSENGERS INTENDING TO USE PUBLIC TRANSPORT Tiffany Lester, Darren Walton Opus International Consultants, Central Laboratories, Lower Hutt, New Zealand ABSTRACT A public transport

More information

ACI EUROPE POSITION PAPER

ACI EUROPE POSITION PAPER ACI EUROPE POSITION PAPER November 2018 Cover / Photo: Stockholm Arlanda Airport (ARN) Introduction Air traffic growth in Europe has shown strong performance in recent years, but airspace capacity has

More information

Civil Aviation Policy and Privatisation in the Kingdom of Saudi Arabia. Abdullah Dhawi Al-Otaibi

Civil Aviation Policy and Privatisation in the Kingdom of Saudi Arabia. Abdullah Dhawi Al-Otaibi Civil Aviation Policy and Privatisation in the Kingdom of Saudi Arabia Abdullah Dhawi Al-Otaibi A thesis submitted to the University of Exeter for the degree of Doctor of Philosophy in Politics September

More information

Preparatory Course in Business (RMIT) SIM Global Education. Bachelor of Applied Science (Aviation) (Top-Up) RMIT University, Australia

Preparatory Course in Business (RMIT) SIM Global Education. Bachelor of Applied Science (Aviation) (Top-Up) RMIT University, Australia Preparatory Course in Business (RMIT) SIM Global Education Bachelor of Applied Science (Aviation) (Top-Up) RMIT University, Australia Brief Outline of Modules (Updated 18 September 2018) BUS005 MANAGING

More information

FACILITATION PANEL (FALP)

FACILITATION PANEL (FALP) International Civil Aviation Organization WORKING PAPER FALP/10-WP/19 Revised 29/8/18 FACILITATION PANEL (FALP) TENTH MEETING Montréal, 10-13 September 2018 Agenda Item 6: Other matters FACILITATION FOR

More information

Press Release March 1, Refer to: Anna Salgado

Press Release March 1, Refer to: Anna Salgado Press Release March 1, 2018 Refer to: Anna Salgado 63 917 5637629 asalgado@megawide.com.ph MEGAWIDE GMR offers an alternative solution to NAIA congestion Aligned with the government strategy to develop

More information

WORLDWIDE AIR TRANSPORT CONFERENCE: CHALLENGES AND OPPORTUNITIES OF LIBERALIZATION. Montreal, 24 to 29 March 2003

WORLDWIDE AIR TRANSPORT CONFERENCE: CHALLENGES AND OPPORTUNITIES OF LIBERALIZATION. Montreal, 24 to 29 March 2003 26/2/03 English only WORLDWIDE AIR TRANSPORT CONFERENCE: CHALLENGES AND OPPORTUNITIES OF LIBERALIZATION Montreal, 24 to 29 March 2003 Agenda Item 1: Preview 1.1: Background to and experience of liberalization

More information

AIRLINES MAINTENANCE COST ANALYSIS USING SYSTEM DYNAMICS MODELING

AIRLINES MAINTENANCE COST ANALYSIS USING SYSTEM DYNAMICS MODELING AIRLINES MAINTENANCE COST ANALYSIS USING SYSTEM DYNAMICS MODELING Elham Fouladi*, Farshad Farkhondeh*, Nastaran Khalili*, Ali Abedian* *Department of Aerospace Engineering, Sharif University of Technology,

More information

Quantile Regression Based Estimation of Statistical Contingency Fuel. Lei Kang, Mark Hansen June 29, 2017

Quantile Regression Based Estimation of Statistical Contingency Fuel. Lei Kang, Mark Hansen June 29, 2017 Quantile Regression Based Estimation of Statistical Contingency Fuel Lei Kang, Mark Hansen June 29, 2017 Agenda Background Industry practice Data Methodology Benefit assessment Conclusion 2 Agenda Background

More information

Abstract. Introduction

Abstract. Introduction COMPARISON OF EFFICIENCY OF SLOT ALLOCATION BY CONGESTION PRICING AND RATION BY SCHEDULE Saba Neyshaboury,Vivek Kumar, Lance Sherry, Karla Hoffman Center for Air Transportation Systems Research (CATSR)

More information

HOW TO IMPROVE HIGH-FREQUENCY BUS SERVICE RELIABILITY THROUGH SCHEDULING

HOW TO IMPROVE HIGH-FREQUENCY BUS SERVICE RELIABILITY THROUGH SCHEDULING HOW TO IMPROVE HIGH-FREQUENCY BUS SERVICE RELIABILITY THROUGH SCHEDULING Ms. Grace Fattouche Abstract This paper outlines a scheduling process for improving high-frequency bus service reliability based

More information

Prof. Tae H. Oum The Air Transport Research Society (ATRS)

Prof. Tae H. Oum The Air Transport Research Society (ATRS) Key Results of the 2010 ATRS Global Airport Performance Benchmarking Project Prof. Tae H. Oum The Air Transport Research Society (ATRS) www.atrsworld.org The ATRS Global Airport Benchmarking Task Force

More information

Case Study 2. Low-Cost Carriers

Case Study 2. Low-Cost Carriers Case Study 2 Low-Cost Carriers Introduction Low cost carriers are one of the most significant developments in air transport in recent years. With their innovative business model they have reduced both

More information

Paper presented to the 40 th European Congress of the Regional Science Association International, Barcelona, Spain, 30 August 2 September, 2000.

Paper presented to the 40 th European Congress of the Regional Science Association International, Barcelona, Spain, 30 August 2 September, 2000. Airline Strategies for Aircraft Size and Airline Frequency with changing Demand and Competition: A Two-Stage Least Squares Analysis for long haul traffic on the North Atlantic. D.E.Pitfield and R.E.Caves

More information

MONTEREY REGIONAL AIRPORT MASTER PLAN TOPICAL QUESTIONS FROM THE PLANNING ADVISORY COMMITTEE AND TOPICAL RESPONSES

MONTEREY REGIONAL AIRPORT MASTER PLAN TOPICAL QUESTIONS FROM THE PLANNING ADVISORY COMMITTEE AND TOPICAL RESPONSES MONTEREY REGIONAL AIRPORT MASTER PLAN TOPICAL QUESTIONS FROM THE PLANNING ADVISORY COMMITTEE AND TOPICAL RESPONSES Recurring topics emerged in some of the comments and questions raised by members of the

More information

Performance monitoring report for 2014/15

Performance monitoring report for 2014/15 Performance monitoring report for 20/15 Date of issue: August 2015 Gatwick Airport Limited Summary Gatwick Airport is performing well for passengers and airlines, and in many aspects is ahead of the performance

More information

PRAJWAL KHADGI Department of Industrial and Systems Engineering Northern Illinois University DeKalb, Illinois, USA

PRAJWAL KHADGI Department of Industrial and Systems Engineering Northern Illinois University DeKalb, Illinois, USA SIMULATION ANALYSIS OF PASSENGER CHECK IN AND BAGGAGE SCREENING AREA AT CHICAGO-ROCKFORD INTERNATIONAL AIRPORT PRAJWAL KHADGI Department of Industrial and Systems Engineering Northern Illinois University

More information

TWENTY-SECOND MEETING OF THE ASIA/PACIFIC AIR NAVIGATION PLANNING AND IMPLEMENTATION REGIONAL GROUP (APANPIRG/22)

TWENTY-SECOND MEETING OF THE ASIA/PACIFIC AIR NAVIGATION PLANNING AND IMPLEMENTATION REGIONAL GROUP (APANPIRG/22) INTERNATIONAL CIVIL AVIATION ORGANIZATION TWENTY-SECOND MEETING OF THE ASIA/PACIFIC AIR NAVIGATION PLANNING AND IMPLEMENTATION REGIONAL GROUP (APANPIRG/22) Bangkok, Thailand, 5-9 September 2011 Agenda

More information

Foregone Economic Benefits from Airport Capacity Constraints in EU 28 in 2035

Foregone Economic Benefits from Airport Capacity Constraints in EU 28 in 2035 Foregone Economic Benefits from Airport Capacity Constraints in EU 28 in 2035 Foregone Economic Benefits from Airport Capacity Constraints in EU 28 in 2035 George Anjaparidze IATA, February 2015 Version1.1

More information

ScienceDirect. Prediction of Commercial Aircraft Price using the COC & Aircraft Design Factors

ScienceDirect. Prediction of Commercial Aircraft Price using the COC & Aircraft Design Factors Available online at www.sciencedirect.com ScienceDirect Procedia Engineering 67 ( 2013 ) 70 77 7th Asian-Pacific Conference on Aerospace Technology and Science, 7th APCATS 2013 Prediction of Commercial

More information

ICAO Options for Allocating International Aviation CO2 Emissions between Countries an Assessment

ICAO Options for Allocating International Aviation CO2 Emissions between Countries an Assessment ICAO Options for Allocating International Aviation CO2 Emissions between Countries an Assessment 1. Background The issue of how to allocate responsibility for the CO 2 emissions generated by international

More information

THE ECONOMIC IMPACT OF NEW CONNECTIONS TO CHINA

THE ECONOMIC IMPACT OF NEW CONNECTIONS TO CHINA THE ECONOMIC IMPACT OF NEW CONNECTIONS TO CHINA A note prepared for Heathrow March 2018 Three Chinese airlines are currently in discussions with Heathrow about adding new direct connections between Heathrow

More information

Programme initiative.pt 2.0 Regulations

Programme initiative.pt 2.0 Regulations Programme initiative.pt 2.0 Regulations Article 1 Object 1. The object of the present Regulations is the definition of the terms of the support granting to projects aimed at attracting or developing air

More information

REPUBLIC OF BULGARIA MINISTRY OF TRANSPORT, INFORMATION TECHNOLOGY AND COMMUNICATIONS

REPUBLIC OF BULGARIA MINISTRY OF TRANSPORT, INFORMATION TECHNOLOGY AND COMMUNICATIONS REPUBLIC OF BULGARIA MINISTRY OF TRANSPORT, INFORMATION TECHNOLOGY AND COMMUNICATIONS TO ALL INTERESTED PERSONS IN THE PROCEDURE FOR GRANTING A CONCESSION FOR SERVICE FOR THE OBJECT CIVIL AIRPORT FOR PUBLIC

More information

PERFORMANCE MEASURES TO SUPPORT COMPETITIVE ADVANTAGE

PERFORMANCE MEASURES TO SUPPORT COMPETITIVE ADVANTAGE PERFORMANCE MEASURES TO SUPPORT COMPETITIVE ADVANTAGE by Graham Morgan 01 Aug 2005 The emergence in the 1990s of low-cost airlines and the expansion of the European travel market has shown how competition

More information

Performance Indicator Horizontal Flight Efficiency

Performance Indicator Horizontal Flight Efficiency Performance Indicator Horizontal Flight Efficiency Level 1 and 2 documentation of the Horizontal Flight Efficiency key performance indicators Overview This document is a template for a Level 1 & Level

More information

IATA ECONOMICS BRIEFING

IATA ECONOMICS BRIEFING IATA ECONOMICS BRIEFING NEW AIRCRAFT ORDERS A POSITIVE SIGN BUT WITH SOME RISKS FEBRUARY 26 KEY POINTS 25 saw a record number of new aircraft orders over 2, for Boeing and Airbus together even though the

More information

Civil and military integration in the same workspace

Civil and military integration in the same workspace Civil and military integration in the same workspace Presented by PLC 1 introduction Civilian and Military ATCOs work alongside each other in various countries and are employed in a number of different

More information

Aviation Competitiveness. James Wiltshire Head of Policy Analysis

Aviation Competitiveness. James Wiltshire Head of Policy Analysis Aviation Competitiveness James Wiltshire Head of Policy Analysis 1 Air Connectivity and Competitiveness Aviation is a major enabler of economic activity and social cohesion Air Connectivity drives economic

More information

Airport Evolution and Capacity Forecasting

Airport Evolution and Capacity Forecasting Internet: www.gap-projekt.de Contact: info@gap-projekt.de Airport Evolution and Capacity Forecasting Branko Bubalo GAP/B Research Project branko.bubalo@googlemail.com partner/sponsor: 8 th GARS Aviation

More information

Introduction: Airline Industry Overview Dr. Peter Belobaba Presented by: Alex Heiter & Ali Hajiyev

Introduction: Airline Industry Overview Dr. Peter Belobaba Presented by: Alex Heiter & Ali Hajiyev Introduction: Airline Industry Overview Dr. Peter Belobaba Presented by: Alex Heiter & Ali Hajiyev Istanbul Technical University Air Transportation Management M.Sc. Program Network, Fleet and Schedule

More information

An Exploration of LCC Competition in U.S. and Europe XINLONG TAN

An Exploration of LCC Competition in U.S. and Europe XINLONG TAN An Exploration of LCC Competition in U.S. and Europe CLIFFORD WINSTON JIA YAN XINLONG TAN BROOKINGS INSTITUTION WSU WSU Motivation Consolidation of airlines could lead to higher fares and service cuts.

More information

JOSLIN FIELD, MAGIC VALLEY REGIONAL AIRPORT DECEMBER 2012

JOSLIN FIELD, MAGIC VALLEY REGIONAL AIRPORT DECEMBER 2012 1. Introduction The Federal Aviation Administration (FAA) recommends that airport master plans be updated every 5 years or as necessary to keep them current. The Master Plan for Joslin Field, Magic Valley

More information

Forecast and Overview

Forecast and Overview Forecast and Overview DENVER INTERNATIONAL AIRPORT Overall goals of the (MPR): Work with DEN to refine the preferred airport development plan to guide the development over an approximate 25-year planning

More information

Airports Commission. Discussion Paper 04: Airport Operational Models. Response from the British Air Transport Association (BATA) June 2013

Airports Commission. Discussion Paper 04: Airport Operational Models. Response from the British Air Transport Association (BATA) June 2013 Airports Commission Discussion Paper 04: Airport Operational Models Response from the British Air Transport Association (BATA) June 2013 Introduction The British Air Transport Association (BATA) welcomes

More information

Competition in the aviation sector: the European Commission s approach

Competition in the aviation sector: the European Commission s approach SPEECH/06/247 Neelie Kroes European Commissioner for Competition Policy Competition in the aviation sector: the European Commission s approach Conference celebrating the twentieth Anniversary of the International

More information

DAA Response to Commission Notice CN2/2008

DAA Response to Commission Notice CN2/2008 22 nd September 2008 DAA Response to Commission Notice CN2/2008 1 DAA welcomes the opportunity to respond to the Commission notice CN2/2008 which discusses the interaction between the regulations governing

More information

1.2 Some of the figures included in this publication may be provisional and revised in later issues.

1.2 Some of the figures included in this publication may be provisional and revised in later issues. FOREWORD 1 CONTENT 1.1 "UK Airlines - Operating and Traffic Statistics" is published by the Civil Aviation Authority with the co-operation of the United Kingdom airline operators. 1.2 Some of the figures

More information

MEASURING ACCESSIBILITY TO PASSENGER FLIGHTS IN EUROPE: TOWARDS HARMONISED INDICATORS AT THE REGIONAL LEVEL. Regional Focus.

MEASURING ACCESSIBILITY TO PASSENGER FLIGHTS IN EUROPE: TOWARDS HARMONISED INDICATORS AT THE REGIONAL LEVEL. Regional Focus. Regional Focus A series of short papers on regional research and indicators produced by the Directorate-General for Regional and Urban Policy 01/2013 SEPTEMBER 2013 MEASURING ACCESSIBILITY TO PASSENGER

More information

NATIONAL AIRSPACE POLICY OF NEW ZEALAND

NATIONAL AIRSPACE POLICY OF NEW ZEALAND NATIONAL AIRSPACE POLICY OF NEW ZEALAND APRIL 2012 FOREWORD TO NATIONAL AIRSPACE POLICY STATEMENT When the government issued Connecting New Zealand, its policy direction for transport in August 2011, one

More information

Economic benefits of European airspace modernization

Economic benefits of European airspace modernization Economic benefits of European airspace modernization Amsterdam, February 2016 Commissioned by IATA Economic benefits of European airspace modernization Guillaume Burghouwt Rogier Lieshout Thijs Boonekamp

More information

Presentation Outline. Overview. Strategic Alliances in the Airline Industry. Environmental Factors. Environmental Factors

Presentation Outline. Overview. Strategic Alliances in the Airline Industry. Environmental Factors. Environmental Factors Presentation Outline Strategic Alliances in the Airline Industry Samantha Feinblum Ravit Koriat Overview Factors that influence Strategic Alliances Industry Factors Types of Alliances Simple Carrier Strong

More information

SHIP MANAGEMENT SURVEY. July December 2017

SHIP MANAGEMENT SURVEY. July December 2017 SHIP MANAGEMENT SURVEY July December 2017 INTRODUCTION The Ship Management Survey is conducted by the Statistics Department of the Central Bank of Cyprus and concentrates primarily on transactions between

More information

Australian Airport Association Stakeholder Dinner. 31 May 2018 Sydney, Australia. Speech by Angela Gittens

Australian Airport Association Stakeholder Dinner. 31 May 2018 Sydney, Australia. Speech by Angela Gittens Australian Airport Association Stakeholder Dinner 31 May 2018 Sydney, Australia Speech by Angela Gittens Good evening ladies and gentlemen. It is a pleasure to be with you tonight and thank you again Caroline

More information

THE EFFECTIVENESS OF DUTCH AIR TRANSPORT POLICY

THE EFFECTIVENESS OF DUTCH AIR TRANSPORT POLICY THE EFFECTIVENESS OF DUTCH AIR TRANSPORT POLICY STUDY PREPARED BY: THE BRATTLE GROUP BY JOHN HORN JAMES REITZES ADAM SCHUMACHER 2 December 22 6 th Floor 8 th Floor 15 Berners Street 1133 2 th Street, NW

More information

STANSTED AIRPORT LIMITED REGULATORY ACCOUNTS PERFORMANCE REPORT FOR THE YEAR ENDED 31 MARCH Financial Review...1. Performance Report...

STANSTED AIRPORT LIMITED REGULATORY ACCOUNTS PERFORMANCE REPORT FOR THE YEAR ENDED 31 MARCH Financial Review...1. Performance Report... PERFORMANCE REPORT CONTENTS Page Financial Review...1 Performance Report...3 Notes to the Performance Report...4 Stansted Regulatory Accounts PERFORMANCE REPORT Financial Review General overview Stansted

More information

The private financing of airport infrastructure expansions

The private financing of airport infrastructure expansions The private financing of airport infrastructure expansions Economic and financial challenges Aviation Insight Series, Singapore Aviation Academy 15 July 2015 Greg Houston Partner, HoustonKemp Australia

More information

2. Our response follows the structure of the consultation document and covers the following issues in turn:

2. Our response follows the structure of the consultation document and covers the following issues in turn: Virgin Atlantic Airways response to the CAA s consultation on Economic regulation of capacity expansion at Heathrow: policy update and consultation (CAP 1658) Introduction 1. Virgin Atlantic Airways (VAA)

More information

LCC Competition in the U.S. and EU: Implications for the Effect of Entry by Foreign Carriers on Fares in U.S. Domestic Markets

LCC Competition in the U.S. and EU: Implications for the Effect of Entry by Foreign Carriers on Fares in U.S. Domestic Markets LCC Competition in the U.S. and EU: Implications for the Effect of Entry by Foreign Carriers on Fares in U.S. Domestic Markets Xinlong Tan Clifford Winston Jia Yan Bayes Data Intelligence Inc. Brookings

More information

How can markets become more contestable?

How can markets become more contestable? How can markets become more contestable? By the end this lesson you will be able to Explain how markets can become more contestable? Differentiate the level of contestability between markets and what determines

More information

ELEVENTH AIR NAVIGATION CONFERENCE. Montreal, 22 September to 3 October 2003

ELEVENTH AIR NAVIGATION CONFERENCE. Montreal, 22 September to 3 October 2003 4/8/03 English, French, Russian and Spanish only * ELEVENTH AIR NAVIGATION CONFERENCE Montreal, 22 September to 3 October 2003 Agenda Item 3: 3.1 : Air traffic management (ATM) performance targets for

More information

NETWORK DEVELOPMENT AND DETERMINATION OF ALLIANCE AND JOINT VENTURE BENEFITS

NETWORK DEVELOPMENT AND DETERMINATION OF ALLIANCE AND JOINT VENTURE BENEFITS NETWORK DEVELOPMENT AND DETERMINATION OF ALLIANCE AND JOINT VENTURE BENEFITS Status of Alliances in Middle East Compared with other world regions, the Middle East is under represented in global alliances.

More information

Introduction to Airports and Their Characteristics Prof. Amedeo Odoni

Introduction to Airports and Their Characteristics Prof. Amedeo Odoni Introduction to Airports and Their Characteristics Prof. Amedeo Odoni Istanbul Technical University Air Transportation Management M.Sc. Program Air Transportation Systems and Infrastructure Module 3 May

More information

Case No IV/M DELTA AIR LINES / PAN AM. REGULATION (EEC) No 4064/89 MERGER PROCEDURE. Article 6(1)(b) NON-OPPOSITION Date:

Case No IV/M DELTA AIR LINES / PAN AM. REGULATION (EEC) No 4064/89 MERGER PROCEDURE. Article 6(1)(b) NON-OPPOSITION Date: EN Case No IV/M.130 - DELTA AIR LINES / PAN AM Only the English text is available and authentic. REGULATION (EEC) No 4064/89 MERGER PROCEDURE Article 6(1)(b) NON-OPPOSITION Date: 13.09.1991 Also available

More information

Case No IV/M KUONI / FIRST CHOICE. REGULATION (EEC) No 4064/89 MERGER PROCEDURE. Article 6(1)(b) NON-OPPOSITION Date: 06/05/1999

Case No IV/M KUONI / FIRST CHOICE. REGULATION (EEC) No 4064/89 MERGER PROCEDURE. Article 6(1)(b) NON-OPPOSITION Date: 06/05/1999 EN Case No IV/M.1502 - KUONI / FIRST CHOICE Only the English text is available and authentic. REGULATION (EEC) No 4064/89 MERGER PROCEDURE Article 6(1)(b) NON-OPPOSITION Date: 06/05/1999 Also available

More information

WORLD TOURISM ORGANIZATION (UNWTO) Final Report WORKSHOP ON TOURISM STATISTICS (SESRIC-TR MINISTRY OF CULTURE AND TOURISM-TURKSTAT-UNWTO)

WORLD TOURISM ORGANIZATION (UNWTO) Final Report WORKSHOP ON TOURISM STATISTICS (SESRIC-TR MINISTRY OF CULTURE AND TOURISM-TURKSTAT-UNWTO) WORLD TOURISM ORGANIZATION (UNWTO) Final Report WORKSHOP ON TOURISM STATISTICS (SESRIC-TR MINISTRY OF CULTURE AND TOURISM-TURKSTAT-UNWTO) 16-18 June, 2008 ANKARA, TURKEY Juan Falconi Morales, UNWTO Consultant

More information

Regulating Air Transport: Department for Transport consultation on proposals to update the regulatory framework for aviation

Regulating Air Transport: Department for Transport consultation on proposals to update the regulatory framework for aviation Regulating Air Transport: Department for Transport consultation on proposals to update the regulatory framework for aviation Response from the Aviation Environment Federation 18.3.10 The Aviation Environment

More information

Alternatives. Introduction. Range of Alternatives

Alternatives. Introduction. Range of Alternatives Alternatives Introduction Federal environmental regulations concerning the environmental review process require that all reasonable alternatives, which might accomplish the objectives of a proposed project,

More information

PREFERENCES FOR NIGERIAN DOMESTIC PASSENGER AIRLINE INDUSTRY: A CONJOINT ANALYSIS

PREFERENCES FOR NIGERIAN DOMESTIC PASSENGER AIRLINE INDUSTRY: A CONJOINT ANALYSIS PREFERENCES FOR NIGERIAN DOMESTIC PASSENGER AIRLINE INDUSTRY: A CONJOINT ANALYSIS Ayantoyinbo, Benedict Boye Faculty of Management Sciences, Department of Transport Management Ladoke Akintola University

More information

COMMISSION OF THE EUROPEAN COMMUNITIES COMMUNICATION FROM THE COMMISSION. Developing an EU civil aviation policy towards Brazil

COMMISSION OF THE EUROPEAN COMMUNITIES COMMUNICATION FROM THE COMMISSION. Developing an EU civil aviation policy towards Brazil COMMISSION OF THE EUROPEAN COMMUNITIES Brussels, 5.5.2010 COM(2010)210 final COMMUNICATION FROM THE COMMISSION Developing an EU civil aviation policy towards Brazil COMMUNICATION FROM THE COMMISSION Developing

More information

A COMPARITIVE ANALYSIS OF AIRLINES EFFICIENCY: EVIDENCE FROM MIDDLE EAST AND NORTH AFRICA

A COMPARITIVE ANALYSIS OF AIRLINES EFFICIENCY: EVIDENCE FROM MIDDLE EAST AND NORTH AFRICA A COMPARITIVE ANALYSIS OF AIRLINES EFFICIENCY: EVIDENCE FROM MIDDLE EAST AND NORTH AFRICA Mostafa Kazemi Associate Prof. Department of Management, Ferdowsi University of Mashhad Iran kazemi@um.ac.ir Amirhossein

More information

Measuring performance and profitability of regional European airports and implications for financial break even

Measuring performance and profitability of regional European airports and implications for financial break even Measuring performance and profitability of regional European airports and implications for financial break even Branko Bubalo, Volodymyr Bilotkatch, Juergen Mueller, Gordana Savic, Tolga Ülkü Measuring

More information

Sunshine Coast Airport Master Plan September 2007

Sunshine Coast Airport Master Plan September 2007 Sunshine Coast Airport Master Plan September 2007 Contents CONTENTS... I ACKNOWLEDGEMENT... II DISCLAIMER... III 1 EXECUTIVE SUMMARY...IV 1 INTRODUCTION... 1 2 AVIATION DEMAND FORECAST... 5 3 AIRCRAFT

More information

ICAO Forecasts for Effective Planning and Implementation. Sijia Chen Economic Development Air Transport Bureau, ICAO

ICAO Forecasts for Effective Planning and Implementation. Sijia Chen Economic Development Air Transport Bureau, ICAO ICAO Forecasts for Effective Planning and Implementation Sijia Chen Economic Development Air Transport Bureau, ICAO Appendix C : Forecasting, planning and economic analyses The Assembly: Requests the Council

More information

An Assessment on the Cost Structure of the UK Airport Industry: Ownership Outcomes and Long Run Cost Economies

An Assessment on the Cost Structure of the UK Airport Industry: Ownership Outcomes and Long Run Cost Economies An Assessment on the Cost Structure of the UK Airport Industry: Ownership Outcomes and Long Run Cost Economies Anna Bottasso & Maurizio Conti Università di Genova Milano- IEFE-Bocconi 19 March 2010 Plan

More information

Network of International Business Schools

Network of International Business Schools Network of International Business Schools WORLDWIDE CASE COMPETITION Sample Case Analysis #1 Qualification Round submission from the 2015 NIBS Worldwide Case Competition, Ottawa, Canada Case: Ethiopian

More information

JUNE 2016 GLOBAL SUMMARY

JUNE 2016 GLOBAL SUMMARY JUNE 2016 GLOBAL SUMMARY FAST FACTS The world of air transport, 2014 All figures are for 2014, unless otherwise stated, to give a single set of data for one year. Where available, the latest figures are

More information

ACI EUROPE POSITION. on the revision of. EU DIRECTIVE 2002/30 (noise-related operating restrictions at community airports)

ACI EUROPE POSITION. on the revision of. EU DIRECTIVE 2002/30 (noise-related operating restrictions at community airports) ACI EUROPE POSITION on the revision of EU DIRECTIVE 2002/30 (noise-related operating restrictions at community airports) 6 SEPTEMBER 2011 EU Directive 2002/30 Introduction 1. European airports have a long

More information

ACI EUROPE ECONOMICS REPORT This report is sponsored by

ACI EUROPE ECONOMICS REPORT This report is sponsored by ACI EUROPE ECONOMICS REPORT 2009 This report is sponsored by Copyright ACI EUROPE 2010 This document is published by ACI EUROPE for information purposes. It may copied in whole or in part, provided that

More information

Aircraft Arrival Sequencing: Creating order from disorder

Aircraft Arrival Sequencing: Creating order from disorder Aircraft Arrival Sequencing: Creating order from disorder Sponsor Dr. John Shortle Assistant Professor SEOR Dept, GMU Mentor Dr. Lance Sherry Executive Director CATSR, GMU Group members Vivek Kumar David

More information

Methodology and coverage of the survey. Background

Methodology and coverage of the survey. Background Methodology and coverage of the survey Background The International Passenger Survey (IPS) is a large multi-purpose survey that collects information from passengers as they enter or leave the United Kingdom.

More information

OPEN SKIES TREATY Last Updated 2/18/10 Compiled by Dave Harris

OPEN SKIES TREATY Last Updated 2/18/10 Compiled by Dave Harris OPEN SKIES TREATY Last Updated 2/18/10 Compiled by Dave Harris mothflyer@gmail.com The following was excerpted from Wikipedia. The Legislative Committee does not necessarily endorse or agree with some

More information