Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

An empirical investigation of the effects of managerial and ownership structure on the efficiency of… Beaudoin, Justin 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2006-0372.pdf [ 6.19MB ]
Metadata
JSON: 831-1.0092972.json
JSON-LD: 831-1.0092972-ld.json
RDF/XML (Pretty): 831-1.0092972-rdf.xml
RDF/JSON: 831-1.0092972-rdf.json
Turtle: 831-1.0092972-turtle.txt
N-Triples: 831-1.0092972-rdf-ntriples.txt
Original Record: 831-1.0092972-source.json
Full Text
831-1.0092972-fulltext.txt
Citation
831-1.0092972.ris

Full Text

A N E M P I R I C A L INVESTIGATION O F T H E E F F E C T S O F M A N A G E R I A L A N D OWNERSHIP S T R U C T U R E O N T H E E F F I C I E N C Y O F N O R T H A M E R I C A ' S AIRPORTS by J U S T I N B E A U D O I N B . C o m . (Hons.), University of British Columbia, 2004 A T H E S I S S U B M I T T E D I N P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R O F S C I E N C E I N B U S I N E S S A D M I N I S T R A T I O N in T H E F A C U L T Y O F G R A D U A T E S T U D I E S (Transportation and Logistics) T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A August 2006 © Justin Beaudoin, 2006 Over the past several decades, the worldwide trend for airport ownership and management has been a gradual movement towards privatization and away from direct governmental management of airport operations. However, several factors have contributed to privatization not being adopted in North America up to this point. Instead, there has been a movement towards quasi-privatization in the form o f not-for-profit/non-share airport authorities. The principal objectives o f establishing the authorities are three-fold: 1. To increase operational efficiency 2. To increase the commercialization of airports and become more responsive to user needs 3. To ensure financial self-sustainability o f operations The objective o f this paper w i l l be to examine whether there is empirical evidence to support the hypothesis that the airport authority structure achieves these objectives. The airport industry in North America is characterized by four different managerial structures: Canadian airport authorities, U S airport authorities, U S city-run airports, and U S port-run airports. After discussing the nature o f the different managerial structures, 5 measures o f productivity and efficiency are employed: 1. Variable Factor Productivity 2. Data Envelopment Analysis 3. Stochastic Frontier Analysis 4. Uni t Cost Index 5. Operating Expense per Passenger The analysis is based on a set o f panel data covering 72 airports over the 10-year period from 1996-2005. The efficiency measures obtained are then adjusted for operational factors deemed to be beyond managerial control, in order to obtain an indication o f managerial efficiency. Multivariate regression analysis is then undertaken to assess whether efficiency varies according to managerial structure. This study found that there is strong evidence that the authority structure achieves higher operating efficiency, a greater degree of commercialization, and is characterized by more proactive management. It is highly l ikely that gains in efficiency in the United States could be achieved by a further movement away from city-managed airports towards the airport authority form. T A B L E O F C O N T E N T S ABSTRACT ii TABLE OF CONTENTS iii LIST OF TABLES v LIST OF FIGURES vi ACKNOWLEDGEMENTS vii 1 INTRODUCTION 1 1.1 Background 1 1.2 Purpose and Significance 3 1.3 Scope 4 1.4 Outline and Organization 5 2 LITERATURE REVIEW 6 3 DISCUSSION OF MANAGERIAL STRUCTURES ..11 3.1 Section Outline 11 3.2 An Outline of the Different Forms of North American Airport Management 11 3.2.1 Canadian Airport Authorities.... 12 3.2.2 US Airport Authorities 13 3.2.3 US City-Run Airports 15 3.2.4 US Port-Run Airports 16 3.3 A Cross-Structural Comparison of Operating Characteristics 16 4 EVALUATION OF EFFICIENCY AND THE EFFECTS OF INSTITUTIONAL FORM 29 4.1 Section Outline 29 4.2 The Data 30 4.2.1 Outputs 31 4.2.2 Inputs 33 4.2.3 Study Limitations 35 4.3 Multilateral Output/Input Index Number Analysis 37 4.3.1 Variable Factor Productivity Index Number Background and Derivation 37 4.3.2 Variable Factor Productivity Index Number Efficiency Results 38 4.4 Data Envelopment Analysis (DEA) 39 4.4.1 DEA Background and Derivation 39 4.4.2 DEA Technical Efficiency Results 43 4.5 Stochastic Frontier Analysis (SFA) 44 4.5.1 SFA Background and Derivation 44 4.5.2 SFA Technical Efficiency Results 45 4.6 Unit Cost Index Analysis 46 4.7 Comparison of Results between Methodologies 48 5 THE IMPACT OF MANAGERIAL STRUCTURE ON EFFICIENCY 51 5.1 The Relationship between Commercialization and Efficiency 51 5.2 The Effects of Managerial Structure on Efficiency 53 6 CONCLUSION 56 6.1 Summary of Key Findings 56 6.2 Suggestions for Further Research 56 7 BIBLIOGRAPHY 58 APPENDICES 61 Appendix A. 1 Canadian Airport Authorities Included in Study 61 Appendix A.2 US Airport Authorities Included in Study 62 Appendix A.3 US City-Run Airports Included in Study 63 Appendix A.4 US Port-Run Airports Included in Study 64 Appendix A. 5 VFP Rankings by Individual Airport 65 Appendix A.6 DEA Rankings by Individual Airport 67 Appendix A.7 SFA Rankings by Individual Airport 69 Appendix A.8 Unit Cost Rankings by Individual Airport 71 Appendix A.9 Operating Expense per Passenger Rankings by Individual Airport 73 LIST O F T A B L E S Table 1.1 Classifications o f North American Airports Included in the Study 4 Table 3.1 Cross-Structural Comparison: Operating Revenues 17 Table 3.2 Cross-Structural Comparison: Operating Costs 18 Table 3.3 Cross-Structural Comparison: Profitability 18 Table 3.4 Cross-Structural Comparison: Growth Rates 19 Table 3.5 Cross-Structural Comparison: Operating Characteristics 19 Table 3.6 Cross-Structural Comparison: Aeronautical Charges 19 Table 3.7 Cross-Structural Comparison: Traffic Output 20 Table 3.8 Cross-Structural Comparison: Physical Infrastructure 20 Table 4.1 The Marginal Revenue Contribution o f Traffic Output 33 Table 4.2 Comparative Rankings between Methodologies - Year 2005 48 Table 4.3 Spearman Rank Correlation Matrix - Year 2005 49 Table 5.1 Regression Results o f Factors Affecting Operating Efficiency 52 Table 5.2 Impact of Non-Aeronautical Revenue on Aeronautical Efficiency 53 Table 5.3 Efficiency Differences between Authorities and City-Run Airports 54 Table 5.4 The Effects o f Managerial Structure on Efficiency - Residual V F P 54 LIST O F FIGURES Figure 3.1 Organizational Chart: U S Airport Authority 14 Figure 3.2 Organizational Chart: U S City-Run Airport 15 Figure 3.3 Operating Expenses per Passenger at Canadian Airports 22 Figure 3.4 Percentage of Non-Aeronautical Revenue: 1996-2005 23 Figure 3.5 Non-Aeronautical Revenue per Passenger: 1996-2005 24 Figure 3.6 Aeronautical Revenue per Passenger: 1996-2005 24 Figure 3.7 Aeronautical Revenue per Aircraft Movement: 1996-2005 25 Figure 3.8 Passenger Facility Charges as a % o f Operating Revenue: 1996-2005 25 Figure 3.9 Passenger Facility Charges per Passenger: 1996-2005 26 Figure 3.10 Operating Expense per Aircraft Movement: 1996-2005 27 Figure 3.11 Operating Income per Passenger: 1996-2005 27 Figure 4.1 The Correlation Between Input and Output Price Levels 36 Figure 4.2 Mean V F P Results by Airport Managerial Structure 39 Figure 4.3 A n Illustration of D E A Input-Oriented Technical Efficiency 41 Figure 4.4 Mean D E A Results by Airport Managerial Structure 43 Figure 4.5 Graphical Representation o f a Stochastic Production Frontier 45 Figure 4.6 Mean S F A Results by Airport Managerial Structure 46 Figure 4.7 Mean Unit Cost Index Results by Airport Managerial Structure 47 Figure 4.8 Mean Operating Expense per Passenger by Airport Managerial Structure 47 Figure 5.1 Relationship between Managerial Efficiency and Aeronautical Charges 55 I would like to offer my sincerest gratitude to my supervisor, Professor Tae Oum, for all o f his support and guidance throughout my years at U B C and his assistance in helping me complete my thesis. I would like to express my appreciation to Professor David Gi l l en for serving on my committee and for all o f his invaluable advice along the way, and I would also like to thank Professor Anming Zhang for serving on my committee and providing very useful feedback. Special thanks also to Professor Yoss i Berechman for the great experience and academic advice over the past year, and to Dr . Chunyan Y u for her help and the many discussions over the years. I w i l l always be grateful for their willingness to help, but even more importantly I am grateful for the opportunity to learn from such a complement o f wel l -respected academics and their role in inspiring me to pursue the field of transportation economics. M y sincerest thank you to Professor K i n L o as well , for encouraging me to pursue graduate studies and for helping to provide me with the confidence in myself to do so. I would like to thank my family and friends for all that they have done over the years, as I couldn't have made it this far without them. Finally, I would like to thank my fiancee, El l iey, for her endless support and encouragement, and for making our goals in life worth striving for. I would also like to acknowledge my appreciation for the financial support for this work provided by the Social Sciences and Humanities Research Counci l (SSHRC) . 1 I N T R O D U C T I O N Over the last few decades, the North American aviation industry has undergone substantial changes in terms o f the prevailing market structure and operating environment. These changes have been primarily driven by the deregulation o f the airline industry, which has had wide-ranging ramifications. The financial environment facing airlines has been altered due to the changes in the competitive landscape between airlines and the evolving nature o f airline competition has resulted in changes for consumers in terms o f fare levels, service quality, and route offerings. While the changes in airport operations over this time period have not been as dramatic, they warrant consideration due to the important role that airports play in the industry and the public policy implications o f airport governance. 1.1 Background Throughout the world, many countries have privatized some o f their airports over the past 10-20 years, and this process is still on-going in many countries. Broadly speaking, changes in the management and ownership o f airports have occurred in two distinct phases (Kapur, 1995). In the early 1970s, a number o f countries began to create airport corporations under public ownership, with the intent of improving efficiency and initiating access to private capital markets. Then in the 1980s, the role o f the state in airport ownership evolved, as a number o f countries used the private sector to finance airport investments directly and gain further efficiency improvements. This second wave of private sector involvement introduced privatization to the ownership/management o f many airports. In stark contrast to this worldwide trend, however, North American airports have not adopted the privatization o f airports. In the United States, and Canada to a lesser extent, the contractual and operational relationship between airports and airlines is a limiting factor against privatization, as in many respects the privatized nature o f the airlines coalesces with the airport use agreements between the airlines and airports to serve as a de facto means of market discipline. Whi le thus far eschewing privatization, North American airports have, though, in recent years begun to be organized as quasi-privatized airport authorities, although differences remain between Canada and the United States. Prior to 1995, Transport Canada owned, operated, or subsidized 150 o f the 726 certified airports in Canada. While the majority o f airports in Canada were locally owned and operated, there was not a clearly defined role for the federal government in regards to the operations o f Canada's airports. The Minister o f Transport appointed a Task Force in 1985 to examine potential alternatives to the existing centrally-managed airport system, and the task force recommended the establishment o f "Loca l Airport Authorities", and the initial lease agreements were established in 1992. In 1995, the implementation o f Canada's National Airports Pol icy ( N A P ) began to be phased in over a period of 5 years. Under the terms o f the N A P , 137 airports were transferred to private or local concerns, including 26 o f Canada's primary airports that accounted for approximately 94% of Canada's passenger and cargo traffic. These 26 airports were classified as the National Airports System ( N A S ) and were leased to local airport authorities. Transport Canada retained ownership o f these airports, but the local 'not-for-profit/non-share capital' authorities have been responsible for the financial and operational management o f these 26 airports since that time. The intent was to create a commercially-oriented system o f airports and improve their efficiency and cost effectiveness by improving managerial and financial autonomy (Kapur, 1995). Fol lowing is a description of the existing structure o f the Canadian Airport Authorities: Each authority is a non-profit corporation headed by a board of directors. Members are nominated by local municipalities and other representative local groups, but cannot be elected politicians or civil servants. Profits generated by LAAs are plowed back into future airport improvements, while losses are offset by Transport Canada through a reduction in lease payments. The LAAs are responsible for management, operation, and maintenance, as well as capital investment projects of the airports they lease. This includes runways, terminal buildings, industrial properties, parking, ground transportation, emergency response services, and financial, personnel, and administrative functions. (Kapur, 1995) While the airport authorities in the N A S may have a different mandate than existed under the previous centrally-managed airport system, it is clear that the L A A structure is a far cry from privatization and the concomitant motivation o f profitability. In the United States, the nature o f airport ownership and management has not undergone a watershed transformation akin to that o f Canada, and changes have been much more gradual. A s opposed to the situation in Canada, the federal government has historically had little direct control over U S airports, as ownership has been at the regional or municipal level and cities and counties have typically been responsible for the operation o f airports. Over the years, however, airport authorities have become more commonplace, and the management o f U S commercial airports is currently comprised o f three alternate structures: airport authorities, port authorities and city-run departments. U S airport authorities are similar in nature to that o f Canadian airport authorities, insofar as they are not-for-profit/non-shareholder entities that re-invest retained earnings into future airport development programs and are by-in-large financially self-sustaining. The U S also has several airports run by local Port Authorities, whereby the Port Authority operates the local seaport(s) as wel l as the local airport(s), or the Port Authority has both a 'Port Div is ion ' and an 'Airport Div i s ion ' . Alternatively, many U S airports continue to be operated as a separate department within the city's or county's administrative organization. In examining the nature o f airport governance in North America, it is natural to consider the implications that these varying institutional structures have on the operational performance o f the airports. H o w does institutional structure affect the operating efficiency o f airports? What underlying factors affect operating efficiency, and what are the relationships between these factors and the different institutional structures? 1.2 Purpose and Significance The objective o f this research w i l l be to examine the effects o f managerial and governance form on the efficiency o f Canadian and American airports and the associated relationship with the extent of airport commercialization. The National Airports Pol icy in Canada had an explicit objective o f increasing airport efficiency and cost effectiveness, and to strive for a greater commercial orientation, as put forth by Transport Canada in 2001 J : Locally-owned and operated airports are able to function in a more commercial and cost-efficient manner, are more responsive to local needs and are better able to match levels of service to local demands. Recent experiences of the four existing airport authorities ... clearly demonstrate these realities. B y in large, American airport authorities purport to have similar objectives. N o w that the N A P has been fully implemented for five years, it is important to evaluate the success o f the policy in realizing its objectives. B y examining the recent operational performance o f North America 's airports, insights can be gleaned as to whether the airport authority structure actually achieves these goals relative to the traditional case o f government-run airports. The results o f such a cross-structural comparison could provide useful information in informing future policy decisions as to the direction that the ownership and management o f North America 's airports should take. A n assessment can be made as to whether the airport authority structure should be retained in Canada, and whether the U S should continue to divest government control o f airport operations to local airport authorities. 1 Retrieved from: http://www.tc.gcxa/prograrns/airports/policy/nap/NASImplementation.htm (Date accessed: August 16, 2006) 1.3 Scope The primary focus o f this research w i l l be a positive analysis o f the recent performance o f North America's major airports and an examination o f the factors affecting both the operating efficiency and the cost effectiveness o f the airports. After studying the relationship between institutional structures and operating performance, these results w i l l be considered in conjunction with the operating characteristics and strategic decisions o f the airports to determine which factors have the greatest effect on operating efficiency and to assess the differences in commercialization between the institutional structures, primarily focusing on the role o f non-aeronautical revenues. The secondary focus o f this research w i l l be a normative assessment as to the desired structure o f airport ownership and management, and potential ways in which the efficiency of North America's airports can be improved in the future. It should be noted that the issue o f privatization is beyond the scope o f this paper; the research here is more concerned with assessing the relative merits o f existing ownership/operating structures. For a detailed discussion o f the issue o f privatizing North American airports, with an in-depth focus on the United States, see Gesell (1999). This study involves panel data covering 72 North American airports over the period o f 1996-2005. A lack o f operational and financial data prior to the implementation o f the N A P precluded a direct time series analysis of the effects o f the policy in Canada. To compensate, U S airports were added to the study in order to ascertain more generally the differences in efficiency and commercialization between city-managed airports and airport authorities. Data collected include: operating revenues, operating expenses, traffic outputs, infrastructure inputs and various operating characteristics. The airports are classified according to four managerial structures, as shown in Table 1.1 and each structure is discussed in Section 3: Table 1.1 Classifications of North American Airports Included in the Study Number of Airports # of Observat ions (Airport-Years) Canadian Airport Authorities 17 140 U S Airport Authorities 19 189 U S City-Run Airports 28 2 7 8 US Port Authorities 8 80 Total 72 687 2 Note that throughout this paper the term "city-run" refers to US airports operated at both the city and county level 1.4 Outline and Organization Section 1 provides an introduction to the objectives o f the research and outlines the basis o f the study. Section 2 contains a review of the relevant literature related to airport managerial structure and efficiency. Section 3 provides a discussion o f the four types o f airport structures included in the study and contains a comparison o f the characteristics o f each o f the structures, with a primary focus on the degree o f commercialization o f the four categories o f airports and the role o f non-aeronautical revenues. Section 4 examines the operating efficiency, productivity, and cost effectiveness o f the airports in order to assess the relative performance o f the airports. To do so, five measures are utilized and introduced in turn: a multilateral output/input index number approach, data envelopment analysis, stochastic frontier analysis, and two indices examining operating costs. Section 5 then examines the efficiency results obtained in Section 4 to determine whether differences exist between the efficiency o f the four categories o f managerial structure. Section 6 concludes with a summary o f key findings and suggestions for further research. 2 L I T E R A T U R E R E V I E W Across the globe, the interest in the productivity and efficiency o f airports has increased in recent years alongside the continuing commercialization and deregulation that many airports and air transport systems have undergone, and this has been reflected in the amount o f research being undertaken. Indeed, many studies have been undertaken to assess airport efficiency, representing numerous countries, employing several different methodologies, and reflecting varying research objectives. This section summarizes the relevant literature relating to airport efficiency and managerial and ownership structure. B y analyzing the efficiency of airport operations, inferences can be made i n regards to the desirability o f various management and ownership structures. B y understanding which factors affect airport productivity and benchmarking the relative performance o f airports, steps can be taken towards improving future performance. There have been numerous theoretical and empirical papers discussing the measurement of efficiency in the transportation sector, and the airport industry specifically. Oum et al (1992) provide an overview o f the issues surrounding productivity measurement in transportation. Doganis (1992) contains a summary o f the traditional measures o f airport performance and efficiency, with a focus on partial factory productivity measures and "industry-oriented" performance measures. Forsyth (2000) discusses more complete measures o f airport performance, including a brief overview o f three o f the methods employed in Section 4 of this paper (variable factor productivity, data envelopment analysis, and stochastic frontier analysis). Total factor productivity (TFP) measures have been applied to airports in numerous studies. A s discussed in Section 4.3, T F P is an index number approach which aggregates the numerous outputs and inputs o f the airport into comparable output and input indices. Hooper and Hensher (1997) use the T F P approach to evaluate the efficiency o f 6 Australian airports over the period o f 1988-1992. Their approach is not equivalent to that employed in Section 4.3, however; their measure o f output was based strictly on deflated revenues and they also included capital inputs 3. The V F P procedure used in this paper is based on the research developed in A i r Transport Research Society (2005) and discussed in Oum et al (2003). These studies have included 50-70 major airports throughout the world, including many o f the North American airports in this paper. In addition to benchmarking the relative performance o f the 3 The inclusion of capital inputs results in a 'total' factor productivity approach, as opposed to the 'variable' factor productivity approach employed here. airports, several regression analyses were undertaken to determine the underlying factors affecting productivity. These studies differ from many others in that non-aeronautical revenue is specified as an output in addition to passengers and aircraft movements. The most prominent methodology applied in airport productivity studies has been that o f data envelopment analysis ( D E A ) , which is introduced in Section 4.4. Data envelopment analysis is a non-parametric approach which estimates an input (or output) frontier based on observed input and output levels, and efficiency is assessed by comparing the observed location o f each airport relative to the estimated frontier. Salazar de la Cruz (1999) studied airport efficiency by using panel data from 16 Spanish airports for the years 1993-1995. Outputs in this study were passengers and revenues (total revenue, infrastructure-related aviation revenue, non-infrastructure related aviation revenue, and non-aviation revenue), and total costs were considered the input. The primary focus o f this paper was on the level o f scale economies o f airport operations. Martin and Roman (2001) used D E A to examine the efficiency o f 37 Spanish airports in 1997. They used three outputs - passengers, air traffic movements and cargo volume - and three inputs - expenditures on labour, capital, and materials. Gi l len and L a l l (1997) took a unique approach to evaluating airport efficiency by classifying airport operations into airside (aircraft movements) and landside (passengers handled) functions and estimated the efficiency and productivity for each side via D E A . This study also contained a second-stage analysis in order to examine the performance changes over time and across airports. Data from 21 U S airports for the period of 1989-1993 were used, with the main objective being to separate airport operations into various components in order to identify sources o f efficiencies. Nyshadham and Rao (2000) studied European airport efficiency via total factor productivity (TFP) and explored the relationship between T F P and several partial factor productivity measures. Abbott and W u (2002) studied the efficiency o f 12 Australian airports over the period from 1990-2000 using a Malmquist T F P index and D E A , and Martin-Cejas (2002) utilized a translog cost function to evaluate Spanish airport efficiency. These studies provide an indication of the diversity o f methods available to study airport efficiency. Sarkis (2000) employed panel data from 44 major U S airports over the period o f 1990-1994 to explore operational efficiencies at airports. He constructed various complex D E A models with four inputs (operating costs, number o f employees, gates and runways), five outputs (operating revenues, number o f passengers, commercial and general aviation movements, and cargo volume), and explanatory variables such as the existence o f hub airlines, and multi- or single-airport systems. He found that, on average, efficiencies have increased over the years and that the presence o f hubbing and snowfalls strongly affected efficiencies at U . S . airports. In contrast, the type o f airport system was not a significant determinant o f efficiencies. Although he did not specifically examine the issue o f managerial structure effects on efficiency, he did refer to Inamete (1993), which lists a number o f factors that can affect airport performance, including: changes in public ownership structure through privatization; contracting out various functions o f airports to private organizations; combining government and private airport ownership; increasing autonomy for government-owned airport organizations; creating government holding corporations; commercializing the activities o f airport organizations; and creating competitive dynamics by having two or more public airport organizations. Kapur (1995) reached similar conclusions in discussing different aspects o f private and public ownership structures and the differences existing both across countries and within countries. Kapur states that: [p]ublicly-owned airports, with a few exceptions, generally have not performed at the same level of efficiency as compared to airports with private sector participation. Reasons contributing to the inefficiency of publicly-owned airports include: political interference in the appointment of management, uneven commercial structures, operational inefficiency resulting primarily from overstaffing and limited commercial orientation...the lack of responsiveness to user needs, and inadequate economic and environmental regulation. It is also noted that the principal objective for the privatization of airports has been to increase private investment given the scarcity of public funds. Since U S airports have access to tax-exempt revenue-backed bonds, this w i l l not be as important o f a pressure towards privatization as it is in other countries. Kapur found that worldwide, corporatized airport authorities achieved improved revenue diversification (via an increase in commercialization), increased efficiency and reduced costs via contracting non-essential services and reducing employment expenses. Whi le there have been numerous papers addressing the efficiency o f airports, few have directly addressed the issue o f ownership and managerial structures. Parker (1999) examined the efficiency o f the B A A airports and the effect that their privatization had upon their efficiency. Using D E A , he found that there was no noticeable impact on technical efficiency subsequent to the privatization o f the airports. Conversely, Y o k o m i (2005) used the Malmquist T F P methodology and found that almost all o f the airports under B A A Pic. achieved increased technical efficiency after privatization. Tretheway (2001) provides an overview o f the different managerial and ownership forms o f airports throughout the world, including a discussion o f the North American structures. However, there have been only two studies empirically considering differences in efficiency between North American city-run airports and airport authorities. Craig et al (2005) directly considered differences in efficiency between U S city-operated airports and U S airport authorities. Their study included unbalanced panel data for 52 airports over the period o f 1978-1992, and differed from this paper insofar as they did not include non-aeronautical revenue output and they included an (inexact) proxy for capital input. They employed cost function analysis and found that U S airport authorities had significantly higher technical efficiency than did the city-run airports. Oum et al (2006) used variable factor productivity analysis on a sample o f major North American, European, and Asia-Pacific airports for the years 2001-2003. They classified the airports according to five different categories o f ownership/governance, including North American airport authorities and city departments. Their measure of efficiency is equivalent to the V F P procedure employed in this paper and included non-aeronautical revenues as an output and excluded capital inputs. They found that there was no statistical significance in the difference in efficiency between the two categories of airports. Their study differed in that it used a shorter time frame than this study, and they did not distinguish between Canadian and U S airports and they did not isolate those airports operated by ports. O f note, they found significant evidence that airports that focus on commercial activities achieved significantly higher efficiency. Heaver and Oum (2001) studied the transition o f Canada's airports from the federal government to the local airport authorities. A s they state, [t]he National Airports Policy essentially shifts the cost of running Canada's airports from the federal government (taxpayers) to those who actually use the facilities. Its aim is to improve economic efficiency by applying market discipline to the development and operation of airports and making airports more responsive to the needs of their customers and local communities. They found that the Vancouver Airport Authority - the first airport to be transferred from Transport Canada - obtained very favourable early reviews, and their case study pointed to high passenger and cargo growth, the attraction o f several airlines, increased concession revenues, and proactive airport development. However, they also stated the following: However, while the experience in Vancouver (and elsewhere) has been favourable, it is not clear that the current system of accountability is sufficient to guarantee that the airport management will perform well in the long run. The self-regulating mechanism of the Board of Directors may not be sufficient to ensure the long-run success of Canada's commercialization model. The Boards lack shareholders to whom they are accountable. There is concern that Boards may be "captured" by the airport management, may lose sight of responsibilities to gain wide community input and may take advantage of an airport's market position. Some experts argue that the UK-style privatization with the efficient price-cap regulation to discourage abuse of monopoly power is a better solution in the long run than the current Canadian approach. The authors believed that the N A P would achieve short-run gains, but was not the optimal policy to avoid the exploitation o f market power and long-run efficiency gains. It should be noted that the study was not quantitative in nature and was limited in its ex post discussion of the results o f the N A P implementation. Wi l ey (1986) provides a theoretical look at the appropriateness o f the airport authority: In his 1953 landmark paper, Authorities as a Governmental Technique, presented at the height of this wave of mania for authorities, Austin Tobin provided the following concrete guidelines for determining the applicability of the authority form: (1) there is a task to be accomplished or a service to be performed, which in the judgement of the people as expressed through their government, either could not or should not be performed by private enterprise; (2) large amounts of capital are needed; (3) efficient management with initiative and business imagination is essential; (4) long-range planning must be in the hands of competent business, financial and professional technicians; (5) the task/service must be self-supporting; (6) free from political interference, bureaucracy and red tape; and (7) the scope of the task/service involves areas more extensive than the established boundaries of state and local government. This viewpoint posits that the authority structure should be advocated only i f the majority o f these seven conditions achieve a positive response. Tobin warns that "an authority should not be created simply to replace the normal functions of the established bureaus or divisions o f government; nor to lul l the public into belief that the activity is self supporting when in reality it is subsidized; nor solely as a device to avoid debt limitations." Overall however, the consensus seems to be that the movement to the airport authority structure should lead to increased efficiency relative to city management, and this sentiment is echoed by Doganis (1992): Some governments and municipalities, while maintaining ownership of their airports, have felt that they could be better operated and managed if those airports had greater autonomy. This has been achieved by setting up airport authorities with a specific brief to manage one or more airports.. .But its primary aim is generally to set up an administration with greater professional skills able to undertake and implement long-range plans while central or local political control is exercised only at the strategic policy level. The next section w i l l briefly summarize the different managerial structures before attention is turned to evaluating the efficiency o f airport operations. 3 DISCUSSION O F M A N A G E R I A L S T R U C T U R E S 3.1 Section Outline This section w i l l outline the characteristics o f each o f the four classifications o f airport management structures in North America and w i l l highlight several key differences between the different structures that could lead to differences in efficiency. Prior to examining the efficiency o f the airports it is important to understand the operating environment in which the airports exist and to examine potential factors that could affect the efficiency results obtained in Section 4. 3.2 An Outline of the Different Forms of North American Airport Management There are numerous differences and numerous similarities between the different management structures. The production process o f all o f the airports is relatively homogeneous; they all use physical capital inputs (runways, terminals, gates, etc) in conjunction with human capital in order to process passengers and facilitate the movement o f aircrafts. Differences arise in strategic decisions made by airport operators (for example, the extent to outsource services, the extent to focus on non-aeronautical services, and the amount o f marketing employed) and in exogenous factors largely beyond managerial control (the proportion o f international passengers, the average aircraft size at the airport, the total number o f passengers at the airport, and so forth). Differences in operational processes are l ikely to be generated by overarching differences in governance form and managerial incentives. Principal-agent theory postulates that people respond to incentives, and this should be no different in airport operations. Indeed, the profit motive is put forth as the driving force behind efficiency gains achievable with privatization. In the absence o f privatization, how effective is the not-for-profit/non-share airport authority structure in providing incentives for airport management to reduce inputs and/or increase outputs? On a spectrum o f the degree o f managerial incentive, it is difficult to presume where the authority structure would be located between the extremes of bureaucratic public-sector provision and unregulated private enterprise. Related to this, it is important to be cognizant o f the fact that the appropriateness o f efficiency measures is dependent upon the objectives of the "firms" being studied. The measures o f efficiency employed in this study implicitly assume that the airports strive to increase outputs and reduce inputs. This assumption is plausible, but it also potentially overlooks possible objectives such as providing some specified level o f service quality, generating regional economic benefits, and so forth. I f the objectives o f the different airport structures differ systematically, then the efficiency measures obtained may be biased towards one group. Wi th that said, the objectives of the airports studied appear to be relatively similar across airports: to increase passenger levels, to be cost efficient, to provide commercial services to airport visitors, and to provide high-quality services. The relative focus on each objective may differ somewhat between airports, and may be an underlying factor in differences in perceived efficiency. Each o f the managerial structures w i l l now be discussed briefly. 3.2.1 Canadian Airport Authorities A s mentioned in Section 1, all o f the Canadian airports were initially centrally controlled at the federal level by Transport Canada. While the majority o f airports in Canada were locally owned and operated, there was not a clearly defined role for the federal government in regards to the operations o f Canada's airports. In 1992, five Local Airport Authorities were created. The initial results were favourable, and provided the impetus for the National Airports Pol icy (NAP) . In 1995, the implementation o f Canada's National Airports Pol icy ( N A P ) began to be phased in over a period of 5 years. B y 2001, the 26 primary airports were classified as the National Airports System ( N A S ) and were leased to local airport authorities, and the structure o f operations has existed to this day; Transport Canada retains ownership o f these airports, but the local 'not-for-profit/non-share capital' authorities are responsible for the financial and operational management o f these airports. Prior to the devolution o f federal government control, airport performance was undermined by several factors, including "a large centralized administration and restrictive labour agreements that increased airports' labour requirements." (Canada Transportation Ac t Review, 2001) Wi th local control, the expectation was that airports would operate in a commercial and cost-effective manner and be more responsive to local needs. The following are some important characteristics of the Canadian airport authority structure: • Not-for-profit/non-share: all retained earnings are reinvested in the airport to cover operating expenses and to contribute towards capital investment • Airports lease the airport land under long-term leases with Transport Canada (rental payments are required under certain terms and conditions) • Board o f Directors represent local businesses and community interests: appointed by a standard procedure and chosen to have complementary skills in several areas (aviation, business, law, engineering, etc.) • H igh degree of transparency: Board o f Directors' backgrounds, compensation and appointment method must be divulged, financial statements must be made public, and competitive tendering o f contracts is required • Mandatory performance reviews every five years • Airport Improvement Fees (charges levied directly to airport passengers) comprise a substantial portion o f airport investment funds: Canadian airports do not have the ability to issue tax-exempt revenue bonds as do the U S airports Overall, there is a noticeable focus on efficiency and commercialization subsequent to the implementation o f the N A P . The main difference between the U S airports and the Canadian airports, apart from sources o f financing, is that there is a unified policy in Canada outlining the requirements and objectives of the Canadian airports, including explicit standards o f governance. Such a codified environment is absent in the United States and is manifested in the diversity o f governance conditions between U S airports. Appendix A . l contains the Canadian airport authorities included in the study. 3.2.2 US Airport Authorities Broadly speaking, the U S airport authorities are highly similar to the Canadian authorities. They are also not-for-profit/non-share entities that reinvest retained earnings back into the airport. The U S authorities also have a dedicated Board o f Governors; however, there are not explicit requirements regarding Board composition and most Board members are not compensated. The most notable differences between the U S and Canadian authorities are: • The U S authorities are not governed by a central policy/mandate • In some cases ownership o f the airports is transferred directly to the authority, while in other cases the airport is leased to the authority • U S authorities have the ability to issue revenue-guaranteed bonds to generate investment funds, and have more federal grant money available than do Canadian airports Overall, the similarities between U S and Canadian airport authorities outweigh the differences. A n example o f a U S airport authority's organizational structure is provided in Figure 3.1. O f note is the relationship between the executive director and the governing board, and the existence o f a director specifically charged with performance monitoring; this is consistent with the belief that the authority structure emphasises efficiency improvements. Appendix A . 2 contains the U S airport authorities included in the study. Figure 3.1 Organizational Chart: U S Airport Authority (Tampa International Airport) George Elbe Director of Air Service Development and International Commerce John Dei Executiv Wheat Mjty 8 Director T Don Welch Director of Human Resources 3.2.3 US City-Run Airports Unlike Canada, the U S continues to have several prominent commercial airports managed by local government departments. In this case, both ownership and management is retained by the government. Further, there are two scenarios. First, the airport may be operated as a component o f the city or county's overall budget. Second, the airport may be managed as an enterprise fund o f the city/county. In this case, the line between authority and government department is blurred somewhat, as all revenues are reinvested into the airport and it is treated as a self-sustaining operation. In this case, the main distinction is at the governance level; U S city-run airports typically report to the local Mayor/Governor and Ci ty Council/Board o f County Commissioners, for the case o f city- and county-run airports, respectively. While some city-run airports have advisory boards, the members are not compensated and have no involvement in the day-to-day operations o f the airport. Sources o f funding are similar for U S authorities and city-run airports. Figure 3.2 provides an illustration o f a typical organizational structure for a city-run airport, with the governance relationship highlighted. Appendix A . 3 contains the U S city-run airports included in the study. Figure 3.2 Organizational Chart: US City-Run Airport (Houston-Bush Intercontinental Airport) Deputy Director Finance ft Mmimstratirei Assistant Otrectcr Finance j AwiBsni Director ; Technical Services Assistant Director -* Airport Property Mgmt 4 Comntcrcnl Oevdopnwnt Assistant Director Human Resources ' Manage Business Services Assistant Manager Semer Manager Bush (AH Airport Assistant Drator Planning & Programming f Assistant Director "j J Operations i Manager Customer Sini ft Admin j Assistant CMieosr Mafotenanee Arport Manager -I NcAtiy HOU Airport ! Asiistsm rjcrvaor Design Assistant Ofrwclor Construction ElUngton FlcM Assistant Drvctor CWi Projects Chief Engineer POC AdmTiistraOon Senior Executive Media Rotations f Senior Qnsutfve •j Intern atkwid Relations Senior Executive -j Markethg Senior EXCCUUVD H Project AdmnistratnA Wanajer ~1 Airport Ssrvtces Assotant Drector Information Tochnorogy Manager Pub* Safety & Security t Airport Security Manager IAH Airport Security Airport Security Manager HOUCFD AifpCrt SCCVTty Airport Security Manager -j Program & Policy CoonHntUon Sr. Staff Analyst «-j Admfcilstratiw) Services 3.2.4 US Port-Run Airports The final, and least common, managerial form is that o f U S port-run airports. These airports were separated from the U S authorities and city-run airports in order to try to achieve the highest homogeneity o f groups as possible. The port-run airports combine characteristics o f both the authority and city-run structures; o f the 8 port-run airports in the sample, 5 are managed by port authorities and 3 are managed as government departments. In both cases separate seaport and airport divisions are created and have separate management directors but are overseen by the same governance structure. Since the sample of port-run airports in the study is relatively small, caution should be exercised in drawing any inferences from the results obtained. Appendix A . 4 contains the U S port-run airports included in the study. 3.3 A Cross-Structural Comparison of Operating Characteristics While the essential functions performed do not vary greatly between the four groups o f airports in the study, there are some differences that exist. These differences take the form o f strategic decisions, the nature o f the traffic served and other operating characteristics, and the relative input and output mixes employed. Several attributes o f the four managerial structures are compared in Tables 3.1-3.8, with the most notable differences being highlighted in turn. • Operating revenues: In terms of the distribution o f operating revenues, the most notable difference is in the relative percentage o f non-aeronautical revenue generated by the airport groups, as is shown in Table 3.1. The airport authorities, in both the U S and Canada, generate a noticeably more significant amount o f their operating revenue from non-aeronautical sources. This lends credence to the idea that the authority structure is indeed conducive to a higher degree o f commercialization. It is also clear that the U S city-run airports and the U S port authorities operate at a much higher level o f operating revenue, and the Canadian airports have a much greater reliance on levied passenger charges. • Operating costs: A s Table 3.2 shows, the scale o f operating expenses varies considerably. Somewhat surprisingly, the U S city-run airports have a higher proclivity to outsource services than do U S airport authorities. In a sense, this would seem to run counter to the intuition o f public-sector bureaucracy and private-sector commercialization. Operating Revenues - Average Values, 2005 Canadian Airport Authorities US Airport Authorities US City-Run Airports US Port Authorities Number 17 19 28 8 Distribution Landing Fee Revenue 39% (22%) 2 1 % 20% 25% Terminal Rental Revenue 24% (28%) 20% 27% 22% Concession Revenue 15% (25%) 19% 2 1 % 11% Parking Revenue 15% (15%) 24% 18% 15% Other Revenue 7% (15%) 16% 14% 27% Total Aeronautical Revenue 63% (51%) 47% 54% 62% Total Non-Aeronautical Revenue 37% (49%) 53% 46% 38% * numbers in parentheses indicate sample averages excluding Toronto (YYZ) Level (2005 $US) Landing Fee Revene 25,344,897 24,451,245 30,833,565 71,830,552 Terminal Rental Revenue 15,344,508 22,921,123 42,985,125 65,075,002 Total Aeronautical Revenue 40,689,405 53,476,143 84,929,067 180,588,732 Concession Revenue 9,938,674 21,545,501 33,451,185 30,542,769 Parking Revenue 9,962,244 27,032,540 28,344,057 44,478,643 Total Non-Aeronautical Revenue 23,960,993 60,804,886 72,658,059 109,059,897 Total Operating Revenue 64,650,398 114,281,029 157,587,126 289,648,629 A IF /PFC Revenue 20,115,688 26,280,563 33,938,820 30,588,623 A IF /PFC Revenue per Passenger 3.88 1.44 1.25 1.37 Operating Costs - Average Values, 2005 Canadian Airport Authorities US Airport Authorities US City-Run Airports US Port Authorities Number 17 19 28 8 Distribution % Labour Expense 4 1 % 4 2 % 4 0 % 2 8 % % Contractual Serv ices - 24% 3 1 % 19% % Soft Cost Expense 5 9 % 3 4 % 2 9 % 5 3 % Level (2005 $U S) Labour Expense 10,164,251 29,853,731 41,023,268 50,836,674 Contractual Service Expense - 17,356,833 31,847,381 34,221,642 Soft Cost Expense 21,044,011 23,819,648 29,727,089 95,059,088 Total Operating Expense 31,208,262 71,030,212 102,597,739 180,117,404 Table 3.3 Cross-Structural Comparison: Profitability Profitability • Average Values, 2005 Canadian Airport Authorities US Airport Authorities US City-Run Airports US Port Authorities Number 17 19 28 8 (2005 $US) Operating Income 40,130,564 53,631,013 67,578,803 135,818,720 R E V E X Ratio 1.78 2.06 1.94 2.08 Aeronautical Revenue per Passenger 6.3 4.0 4.2 8.0 Non-Aeronautical Revenue per Passenger 5.4 4.8 4.0 5.4 Total Operating Revenue per Passenger 11.7 8.8 8.2 13.4 Operating Expense per Passenger 8.6 5.4 5.6 8.2 Operating Income per Passenger 3.1 3.4 2.5 5.2 Growth - Average Values, 2005 Canadian Airport Authorities US Airport Authorities | US City-Run Airports | US Port Authorities Number 17 19 I 28 | 8 (1996-2005) Annual Operating Revenue Growth 6.5% 4.7% 3.6% 5.2% Annual Non-Aeronautical Revenue Growth 8.1% 7.0% 3.5% 5.6% Annual Passenger Growth 3.1% 2.3% 3.0% 2.9% Annual Aggregate Output Growth 3.1% 3.7% 2.7% 3.1% Annual Aggregate Input Growth 4.0% 2.4% 2.7% 2.7% Annual Variable Factor Productivity Growth -0.3% 1.1% 0.4% 0.5% Table 3.5 Cross-Structural Comparison: Operating Characteristics Operating Characteristics - Average Values, 2005 Canadian Airport Authorities US Airport Authorities US City-Run Airports US Port Authorities Number 17 19 28 8 % International Passengers 18% 4% 8% 14% % Transferring/Connecting Passengers 10% 2 5 % 26% 12% Passengers per Movement 30.4 60.2 72.3 80.2 Passenger Share - Dominant Airline - 37% 44% 33% Herfindahl-Hirschman Index - HHI (Top 5 Airlines) - 2156.8 2622.9 1828.9 # of Scheduled Airlines 15.5 22.0 25.2 34.6 # of Non-Stop Destinations 26.5 75.8 83.0 86.1 Table 3.6 Cross-Structural Comparison: Aeronautical Charges Aeronautical Charges - Average Values, 2005 Canadian Airport Authorities US Airport Authorities | US City-Run Airports US Port Authorities Number 17 19 | 28 8 (2005 $US) Residual Methodology (%) - 10 (53%) 18(64%) 2 (25%) Compensatory Methodology (%) - 4(21%) 6 (22%) 5 (62%) Hybrid Methodology (%) - 5 (26%) 4 (14%) 1 (13%) Aeronautical Revenue per Passenger 4.4 4.0 4.2 8.0 Landing Fee per Movement (per Passenger) 63.6 (2.39) 72.7(1.23) 88.2(1.22) 201.5(2.42) Terminal Rental per mz 194.0 193.8 211.7 422.4 Note: Canadian values exclude Toronto (YYZ) MO Traffic Output - Average Values, 2005 Canadian Airport Authorities US Airport Authorities US City-Run Airports US Port Authorities Number 17 19 28 8 International Passengers 1,845,019 1,118,384 3,022,129 4,657,492 Domest ic Passengers 2,754,106 17,105,055 22,640,503 19,296,137 Total Passengers 4,599,125 18,223,439 25,662,632 23,953,630 Cargo (tonnes) 60,710 201,157 393,245 535,089 Aircraft Movements 99,363 278,217 337,089 301,225 Table 3.8 Cross-Structural Comparison: Physical Infrastructure Physical Infrastructure - Average Values, 2005 Canadian Airport Authorities US Airport Authorities US City-Run Airports US Port Authorities Number 17 19 28 8 Runways 2.5 3.4 3.4 2.8 Runway Length (m) 5,969 9,508 9,620 7,712 Gates 20.1 72.2 73.4 70.5 Employees 139.9 475.2 688.7 401.0 Terminal S ize (m^) 59,245 134,568 209,863 185,932 O Profitability: A s Table 3.3 shows, each of the four groups averaged sizeable operating profits. In general, operating profits were roughly twice the size o f operating expenses (as shown by the R E V E X ratio). The Canadian airports and the U S port authorities had higher aeronautical revenue per passenger, and the authorities generate significantly more non-aeronautical revenue per passenger than do the U S city-run airports. The U S authorities had lower operating expenses per passenger and a higher operating margin than did the government-run airports. Growth Rates: Table 3.4 illustrates the growth of various factors over the 10-year period from 1996 to 2005. O f note, the airport authorities had a much higher growth rate in non-aeronautical revenues over this period, with Canadian airports obtaining an 8.1% rate o f growth and the U S airport authorities obtaining a 7% rate o f growth, both in real terms. This again supports the belief that the authority structure leads to a greater degree o f commercialization. Passenger growth rates were similar across the groups, and the U S airport authorities managed the highest overall growth in V F P . Operating Characteristics: Table 3.5 shows several operating characteristics o f the airport groups. The percentage o f international passengers varies noticeably between groups, as does the percentage o f connecting passengers and the average number o f passengers per movement. The degree o f airline concentration is the greatest at U S city-run airports; the average airport in this group has 44% of their passengers owing to the dominant airline, while this rate is only 37% for the U S airport authorities, and the Herfindahl-Hirschman Index o f the top 5 airlines at each airport reflects this trend as wel l . Finally, the Canadian airport system is much less connected than is the U S system; Canadian airports have a lower rate o f transferring passengers, fewer scheduled airlines on average, and a smaller non-stop route network. Aeronautical Charges: A s shown in Table 3.6, in the U S the type o f airline use agreement in place varies by group; the authorities are more l ikely to use a hybrid methodology than any o f the three, the city-run airports are more l ikely to use a residual methodology, and the port authorities are more l ikely to use the compensatory methodology. The U S airport authorities have the lowest aeronautical charges, while the port authorities have the highest aeronautical charges. Traffic Output: Table 3.7 summarizes the average traffic output o f each group. The scale o f output is demonstrably smaller for the Canadian airports, which limits the comparability with the U S airports. For the U S airports, the authorities have a smaller average number o f passengers, much less cargo activity, and fewer average aircraft movements per year. The scale o f output is an important factor to consider in assessing the relative efficiency o f the airports. Figure 3.3 plots the average operating costs o f the airports against the level o f passenger output. This figure shows significant economies o f cost savings as output increases up to the level o f 5 mil l ion passengers. This finding is consistent with Jeong (2005). 14 o f the 17 Canadian airports (82%) handle less than 5 mi l l ion passengers annually, while only 2 of 57 U S airports (4%) handle less than 5 mi l l ion passengers annually. This factor needs to be considered when calculating measures o f technical efficiency. Figure 3.3 Operating Expenses per Passenger at Canadian Airports a) c a> in (0 ra Q. a . °2 x C UJ c a. O 18 16 14 * ^ 12 10 8 0 0 10 15 Passengers 20 25 30 Millions • Physical Infrastructure: Finally, Table 3.8 provides information on the capital inputs o f the airports. The U S airports are generally similar in their degree o f capital inputs, and are much larger than the Canadian airports in this regard. Interestingly, in addition to having a higher proportion of outsourcing than do the U S airport authorities, the U S city-run airports also have a higher number of average employees. This is partly due to the larger average size o f U S government-run airports. The average number o f passengers per employee for the four groups in 2005 is as follows: o Canada Airport Authorities: 26,079 o U S Airport Authorities: 35,197 o U S City-Run Airports: 41,260 o U S Port-Run Airports: 68,301 This again supports the view that U S city-run airports rely more heavily on outsourced services. Figures 3.4-3.11 illustrate some interesting relationships as well . Figure 3.4 shows how both the U S and Canadian airport authorities have increased their focus on non-aeronautical revenues over the past decade, and how the U S city- and port-run airports have increasingly relied on aeronautical revenue sources and Figure 3.5 illustrates the growth in non-aeronautical revenues per passenger, in real terms. Not only do the authorities have a greater focus on non-aeronautical revenue sources, they have become more effective in exploiting non-aeronautical revenue sources over the years, whereas the city-run airports have stagnated in this regard. Figure 3.4 Percentage of Non-Aeronautical Revenue: 1996-2005 Canadian Airport Authorities 0.40 A 1 1 1 1 1 1 . . . 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Figure 3.5 Non-Aeronautical Revenue per Passenger: 1996-2005 Non-Aeronautical Revenue Per Passenger 5.00 -Canadian Airport Authorities -US City-Run Airports - US Port-Run Airports - US Airport Authorities 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 A s Figure 3.6 shows, aside from the port-run airports, the aeronautical charges per passenger have increased slightly in real terms, with no discernable difference between U S airport authorities and city-run airports. Figure 3.7 shows the amount o f aeronautical revenue per aircraft movement; the relative rankings differ from Figure 3.6 due to differences in average aircraft size operating at the airports (see Table 3.5). Figure 3.6 Aeronautical Revenue per Passenger: 1996-2005 2.00 -I 1 1 1 1 1 1 1 1 , 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Figure 3.7 Aeronautical Revenue per Aircraft Movement: 1996-2005 Aeronautical Revenue Per Aircraft Movement 500 .00 4 5 0 . 0 0 4 0 0 . 0 0 j« 3 5 0 . 0 0 300 .00 o Q CO 2 5 0 . 0 0 3 <S 200 .00 at 150.00 100 .00 50 .00 0.00 - Canadian Airport Authorities •US City-Run Airports - US Port-Run Airports • US Airport Authorities 1996 1997 1998 1999 2 0 0 0 2001 2 0 0 2 2 0 0 3 2 0 0 4 2 0 0 5 Figure 3.8 Passenger Facility Charges as a % of Operating Revenue: 1996-2005 P a s s e n g e r Facility C h a r g e s as a % of Operat ing R e v e n u e 1 9 9 6 1 9 9 7 1 9 9 8 1 9 9 9 2 0 0 0 2 0 0 1 2 0 0 2 2 0 0 3 2 0 0 4 2 0 0 5 Figure 3.9 Passenger Facility Charges per Passenger: 1996-2005 Figures 3.8 and 3.9 pertain to revenues obtained via passenger facility charges. The differences are negligible between the U S airports, but the Canadian airports have a much greater reliance on this source o f revenue. In 1999, passenger facility charges were roughly 20% the size of total operating revenues, and this figure exceeded 40% in 2005. These revenues are being generated specifically to fund investment projects at the airports; an intriguing area for future research is the desirability o f passenger facility charges and the implications for investment and the efficacy o f the governance mechanisms in place to facilitate optimal investment decisions. Figure 3.10 shows how operating expenses per aircraft movement changed over time. Interestingly, each airport group exhibited a decrease in cost efficiency, in real terms. Comparing Figure 3.10 with Table 3.3 provides an interesting result; Canadian airports are the least cost effective on a per passenger basis, but are the most cost effective on a per aircraft movement basis. This reinforces the fact that different measures can provide very different results. According to this measure though, U S airport authorities are again more efficient that U S city-run airports. Finally, Figure 3.11 compares the profitability per passenger o f the airports. There has been little change over the past years, and the U S airport authorities are slightly more profitable than are U S city-run airports. Figure 3.10 Operating Expense per Aircraft Movement: 1996-2005 Operating Expenses Per Aircraft Movement 0.00 -I 1 1 1 1 1 1 1 1 1 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Figure 3.11 Operating Income per Passenger: 1996-2005 Operating Income Per Passenger 4.00 - Canadian Airport Authorities - US City-Run Airports - US Port-Run Airports - U S Airport Authorities 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 This section has served to illustrate that there are several important differences in the operating characteristics o f the different airport groups. O f special interest is the strong evidence that the airport authority structure does indeed result in a higher degree o f commercialization. Graham (2003) discusses the recent emphasis on non-aeronautical revenue sources: There have been a number of factors which have contributed to the growth in dependence on non-aeronautical revenues. First, moves towards commercialization and privatization within the industry have given airports greater freedom to develop their commercial policies and diversify into new areas. A more business-oriented approach to running airports has also raised the priority given to commercial facilities. Such facilities were traditionally considered to be rather secondary to providing essential air transport infrastructure for airlines. Managers are now eager to adopt more creative and imaginative strategies and to exploit all possible aeronautical and non-aeronautical revenue generating opportunities. Greater attention began to be placed on the commercial aspects of running an airport such as financial management, non-aeronautical revenue generation and airport marketing. The operational aspects of the airport traditionally had overshadowed other areas and most airport directors and senior management were operational specialists. However, the commercial functions of an airport gradually were recognized as being equally important and, as a result, the resources and staff numbers employed in these areas were expanded. The airport authorities have a greater focus on non-aeronautical revenue sources, which is reflective o f a proactive approach to airport management. Oum et al (2006) found a strong relationship between airport commercialization and efficiency; Section 4 w i l l turn to measuring the efficiency o f the airports and Section 5 w i l l examine whether the airport authority also achieves its other primary objective - increased efficiency. 4 E V A L U A T I O N O F E F F I C I E N C Y A N D T H E E F F E C T S O F INSTITUTIONAL F O R M 4.1 Section Outline A s discussed in the literature review, there have been several studies related to the efficiency o f airports. The studies have utilized various techniques to measure efficiency and have had varying objectives in doing so. This section w i l l now turn to estimating the efficiency o f the major airports in North America. This is a worthwhile objective in and o f itself, as it is necessary to first ascertain which airports are efficient before it can be determined why these airports are efficient. This section begins with a discussion of the data used throughout the study. Five methodologies are subsequently introduced and applied in turn in order to estimate the efficiency o f the airports in the study. Mult iple measures o f efficiency are used in order to provide a more accurate assessment o f the underlying productivity o f the airports; it has been established that empirical efficiency results may vary between different methods o f analysis (Oum et al, 1999) and the section concludes with a brief comparison o f the results between the different methods. A s mentioned at the outset, a key impetus o f Canada's National Airports Pol icy was the belief that a shift to the Airport Authority structure would increase operating efficiency. A common notion is that government bureaucracy is associated with X-inefficiency, so it is worthwhile to examine whether this notion holds in the case o f airport operations. To test this hypothesis, several complementary methodologies w i l l be employed. The technical efficiency o f airport operations w i l l be measured by three different procedures: 1. multi-lateral index number approach 2. data envelopment analysis 3. stochastic frontier analysis, and an analysis o f the cost effectiveness of the airports w i l l also be employed: 4. unit cost index number approach 5. operating expenses per passenger In some respects, productivity analysis remains a nebulous concept, insofar as the very definition o f productivity is often the subject o f debate, in addition to the lack o f a universal agreement as to which methodologies provide the "correct" computation o f productivity, however it may be defined. Generally, productivity is defined as the ratio o f outputs to inputs, or the rate at which inputs are transformed into outputs. In the case of a firm with a single input and a single output, the calculation o f productivity is relatively trivial. But when a firm produces multiple outputs from multiple inputs, as is the case with the airport activities examined in this paper, the computation o f productivity is not nearly as precise and poses several computational and philosophical barriers. The literature concerning productivity analysis is expansive, and continually evolving. M a n y approaches to measuring productivity analysis have been developed, with a substantial variation in the underlying assumptions involved, the data required, and the transition from theory to practice. A s mentioned above, productivity can be thought o f as the ratio o f outputs to inputs or the productive output per unit o f input. When productivity improvement is considered, two viewpoints can be taken. First, productivity can be increased by increasing the output o f the firm relative to a constant (or decreased) level o f inputs. Second, productivity can be increased by decreasing the input o f a firm required to produce a given (or increased) level o f outputs. This is an important distinction that w i l l be brought to bear in a subsequent section o f the paper. Assessing the productivity o f airports is an important endeavour. Below are some motivating factors for assessing productivity: • Managerial performance can be evaluated • The "best practices" of the airports determined to be efficient can be replicated by less efficient airports • In an industry that exhibits public-sector involvement, such as the airport industry, productivity analysis can be used as a monitoring device • The sources of efficiency and the causes of productivity changes (both improvements and declines) can be investigated • The efficacy o f various operational and institutional policies can be evaluated The above factors are not an exhaustive list o f the reasons for undertaking productivity analysis, but are intended to convey the relevance o f the issues discussed throughout this paper. 4.2 The Data A s shown in Table 1.1, the study contains 72 airports that are categorized according to four ownership/management forms. The data form an unbalanced panel covering the years 1996-2005 and contain 687 observations in total. The following sources were utilized in obtaining the relevant data: • Airport websites and annual reports • U S financial data: F A A Airport Financial Reporting website • Statistics Canada • Bureau o f Transportation Statistics • Airports Counci l International - North America • Direct correspondence with airports This section w i l l briefly discuss the various outputs and inputs that are used throughout this study. 4.2.1 Outputs Traditionally, the outputs o f airports have been considered to be passengers, freight and aircraft movements. This viewpoint has focused on the role o f the airport as a node within the transportation network that facilitates the movement o f passengers and freight. Certainly, this can still be seen as the inherent reason for the existence of airport infrastructure. However, when the activities of an airport are considered from the airport operator's point o f view, this conception o f output is incomplete. In considering the activities o f airport operators in recent years, it is apparent, from both words and action, that non-aviation related activities have become an increasingly important component o f airport operations. Providing passengers and local residents with commercial services has been a central component o f airport marketing activities and development initiatives. The issue o f airport commercialization was addressed in greater detail in Section 3, but it suffices here to introduce the inclusion o f non-aeronautical output in the study. Three outputs o f airport operations are included in this study: • Passengers: The number of passengers utilizing an airport is generally the central output when considering airport operations. The figures in this paper include both enplaning and deplaning passengers on both domestic and international flights, calculated on an annual basis. • Aircraft Movements: Another output to be considered is the number o f commercial aircraft movements occurring at the airport. This measure includes both landings and takeoffs; general aviation and military operations were removed from the data. • Non-Aeronautical Revenue: The final output considered is non-aeronautical revenue, generally in the form o f food and beverage sales, retail activities, rental car services, other concession-type services, land and property revenues, and parking revenues, which are particularly significant for North American airports given the strong reliance o f most North American cities on the private automobile. Commercial activities at airports have taken on increasing importance over the years, as the contribution o f commercial activities towards financial success has been documented. Indeed, many airports have attempted to improve efficiency and reduce the fees charged to airlines in order to attract additional flights to the airport and spur commercial activities. A n explanation as to why cargo/freight handled at the airport is not included as an output is as follows. Traditionally, both passengers and freight have been viewed as the primary outputs o f an airport. However, the operations o f North American airports differ from most airports in regard to cargo. A t the majority o f U S airports, cargo is transported primarily in the belly o f passenger flights, and is by in large handled by the airlines, third-party cargo handling companies, and others that lease space and facilities from airports (Oum et al, 2006), so there is little overall participation from the airport operator. To examine the importance o f cargo output to the airports in the study, the operating revenue of the airports was regressed against the number o f international passengers, the number o f domestic passengers, and the volume o f cargo handled by the airports. The results are displayed in Table 4.1. The coefficients reported in Table 4.1 can be interpreted as the marginal revenue o f each o f the measures of traffic volume 4 . A s expected, international passengers have a higher marginal revenue value than do domestic passengers, at $18.62 and $2.22 respectively 5, and both results are strongly statistically significant. On the other hand, the marginal revenue o f cargo is not statistically different from zero, supporting the view that cargo volume is not an 4 Note that the model presented is likely not representative of the complete revenue function of the airports; it is intended to assess the relative importance of passengers and freight to the airports in the study 5 Figures are in 1996 $US important element o f North American airport operations and thus does not need to be considered as an output when analyzing the efficiency o f North American airports. Table 4.1 The Marginal Revenue Contribution of Traffic Output D e p e n d e n t Va r i ab le : Opera t ing R e v e n u e Coef f ic ient S tanda rd Error f-stat Intercept 2.481 E+07 2 .723E+06 9.111 International P a s s e n g e r s 18 .619 0 .8136 22 .890 D o m e s t i c P a s s e n g e r s 2 .234 0 .1572 14 .210 C a r g o 7 .998 8 .526 0 .938 R 2 Ad jus ted R 2 Obse rva t i ons (n) 0 .7933 0 .7924 687 4.2.2 Inputs Relatively speaking, the outputs of an airport are more easily identified and measured than are the inputs. Not only are there theoretical difficulties in determining the correct measure o f input usage, there are serious pragmatic limitations in obtaining economically meaningful input data. In the present study, three principal categories of inputs are identified and briefly discussed: • Labour input: Labour input can be measured in two ways. One, the number o f employees (full-time equivalent) employed directly by the airport operator, and two, the expenditures upon those employed directly by the airport and its management, including wages, salaries, benefits, and so forth. It is important to note that this does not include outsourced labour services. The amount o f outsourcing varies by airport, and the expenditures on such activities are included in the soft cost expenditure value. • Soft cost expenditures: This category is representative of all operating expenses (variable costs) exclusive o f labour expenditures, financing costs, and capital costs. It is a residual figure that, in the case o f airports, generally includes contractual services (outsourcing), materials purchases, utilities and maintenance expenditures, marketing costs, and so on. • Capital assets: Physical/capital assets are a significant input into the operations o f airports. Although airports have a service component, the production o f airport output is inherently infrastructure intensive. A n analysis o f airport productivity is incomplete without incorporating capital usage. However, there are many serious impediments to the comparison o f capital inputs between airports. Doganis (1992) contains a pertinent discussion of this issue. Ideally, a complex capital input index representing economic depreciation would be constructed to account for the capital input o f each airport (see Christensen and Jorgensen (1969) and Diewert (1980) for the theory underlying capital input measurement), but the data required is prohibitive for this study. Accounting depreciation is used as a proxy for capital input usage in many studies; however, the method o f computing accounting depreciation varies significantly across airports, and reliable measures were not readily available and thus would provide little probative value. Another approach to measuring capital asset inputs is to directly include the physical assets o f each airport. Available data for this study include: o the number of employees directly employed by each airport's operating authority, o the number o f gates, o the number o f runways and total runway length (measured in metres), and o the total terminal size (measured in square metres) o f each airport. While these figures are useful in understanding the operations at each airport, they are very crude indicators of capital usage, because they fail to provide information on the quality o f the assets, the age o f the assets, and the cost o f usage of these inputs. Further complications in measuring expenditures on capital infrastructure and facilities include the fact that airport infrastructure is discrete in nature (Oum and Zhang, 1990), the investment period is extended over many years, and the lead-time for new projects is also very long. In the United States, many airports also have capital assets that have been financed directly by the airlines 6, and the amount o f government subsidization, the sources o f financing/debt, and the tax rates facing airports are heterogeneous across the sample. A s a result, the focus on efficiency in this study w i l l primarily regard capital inputs as fixed, and efficiency w i l l be estimated in relation to the existing level o f capital inputs. It is important to be cognizant o f the fact that observed operating costs are a function o f the underlying capital inputs, and as such, the efficiency results in this study are incomplete. 6 For example, United Airlines has its own terminal at Washington Dulles International Airport (IAD), Continental Airlines financed a terminal at Houston-Bush International Airport (IAH), and Delta Airlines financed a terminal at John F. Kennedy International Airport (JFK). Ideally, the efficacy of the various operating structures would be assessed according to both operating efficiency as wel l as capital investment decisions. Given the substantive nature o f airport investment decisions, welfare gains from socially optimal investment decisions are l ikely a greater magnitude than are gains obtainable from improvements in operating efficiency. The issue of airport investment and the governance implications and incentives o f the various operating structures has significant implications for public policy, and is an intriguing area for further research. Finally, it should be noted that the financial data used in the study has undergone adjustments to facilitate comparisons. Canadian figures have been adjusted by the Wor ld Bank's Purchasing Power Parity (PPP) index to normalize financial values between the two countries. Then both the U S and Canadian data were adjusted by the Consumer Price Index (CPI) 7 to account for inflation over the study period. Thus, financial data is measured in 1996 $US, unless indicated otherwise, and changes observed over time can be regarded as real-value changes. After briefly addressing the potential limitations o f the study, the remainder o f Section 4 is devoted to developing the models used to estimate the efficiency o f the airports. 4.2.3 Study Limitations Before proceeding with the estimation o f efficiency, it is important to address the potential limitations of the study. The primary limitation is the failure to include capital inputs in estimating efficiency. The production function o f an airport can be considered as follows: Output = f(l,sc;d,k) where / = labour sc = soft costs o = vector of operating characteristics k = capital inputs Data regarding labour and soft costs is available, and efficiency results obtained can be adjusted i f necessary to account for differences in operating characteristics between airports. However, the present study does not account for capital inputs. This can present a problem in three instances: 1. It treats the level o f capital as fixed, k : this may not be an appropriate assumption over a 10-year period 7 The CPIs were obtained from the Bureau of Labour Statistics (BLS) and Statistics Canada, respectively. 2. It fails to determine whether the level o f capital employed is optimal, k*: as mentioned previously, an important factor in assessing the merits of the different managerial structures is determining how effective each structure is at investing in the optimal level o f infrastructure 3. It does not account for differences in the capital usage between airports, in terms o f quantity and/or quality, kl * kj: this may bias the efficiency results towards airports with higher levels of capital inputs i f there are economies of scale in capital inputs A s such, this study estimates variable operating efficiency, and in some respects reflects o f efficiently the airports are able to use their existing levels of capital. A more accurate assessment would include capital inputs. Another potential limitation of the study is that it does not control for variances in input prices facing the airports. A n adequate input price index was not readily available to normalize expenditures across airports. Ideally, a producer price index would be used to control for input prices and a consumer price index would be used to control for output prices facing the airports. However, this limitation may be minimal due to the fact that non-aeronautical revenues (an output) correlate strongly with soft cost expenses (an input), as shown in Figure 4.1. Therefore, any differences in the price levels between airports may largely cancel out. Figure 4.1 The Correlation Between Input and Output Price Levels in I 300 250 CO CO 0> at CD 3 C <D s a. 4) C O 200 150 100 50 Non-Aeronautical Revenues vs. Soft Cost Expenses 50 100 150 200 250 Soft Cost Expenses (1996 US$) 300 350 Millions 4.3 Multilateral Output/Input Index Number Analysis Given the multiplicity of inputs and outputs accompanying airport operations, aggregation is necessary in order to obtain a holistic view of airport productivity. Theoretically, however, this is not a trivial matter. For instance, how does one combine output in physical measurements (the number of passengers and the number o f aircraft movements) with output in financial measurements (non-aeronautical revenue) to obtain a valuation o f total output? There are a large number of approaches that vary in their method of aggregation to produce output and input index numbers. 4.3.1 Variable Factor Productivity Index Number Background and Derivation In this paper the (flexible) translog functional form is used in order to provide aggregate output and input indices. B y dividing the aggregate output index by the aggregate input index, one obtains a measure of the airport's productivity. Since capital inputs are not included in the calculations, it is a variable measure of productivity, as opposed to a measure o f total factor productivity incorporating capital inputs. The methodology used in this paper was first proposed by Caves, Christensen, and Diewert (1982). The translog multilateral output index, InSu, is defined by Caves, Christensen, and Diewert as: l n < 5 « = i n & - 5 i A = 7 Z ( ^ * + ^ ) ( l n y f - i r J / ) - \ I + R, ) ( l n - tayj) 1 i 1 i The index is formed by an exhaustive series of binary comparisons between the observations of each airport and the sample mean, with the result being a transitive set of comparisons across all observations. Yf represents the z'th output (/ = 1,2,3) for the kth airport (k = 1,2,3,...,72) andi?*representing the revenue share of the / t h output for the kth airport. i? ( and In Yi represent the arithmetic mean of the revenue share o f the * t h airport across the sample and the geometric mean of the output of the i t h airport across the sample, respectively. In words, the translog multilateral output index is computed by normalizing the logarithm of the three outputs (passengers, aircraft movements, non-aeronautical revenue) relative to the mean value of the airports in the study and then aggregating these relative outputs based on their respective share of total operating revenue in order to provide a measure o f total output relative to the other airports in the sample. Analogous to the output index above, Caves, Christensen, and Diewert also specified a translog multilateral input index, ln/?^ , as: ln/?J / = i n ^ - h i ^ = - i z {Wkn + Wn ) ( l n x * + ^ ) ( l n ^ -w h e r e b y r e p r e s e n t s the « t h input (n = 1,2) for the kth airport (k = 1,2,3, ...,72), wk„ is the cost share of the « t h input for the kth airport, Wn is the arithmetic mean of the cost share of the « t h input over the airports in the sample, and \nXn is the geometric mean of the n t h input over the airports in the sample. The intuition is equivalent to that of the output index above; in order to compute the translog multilateral input index, the two inputs (labour input and soft cost expenses) are aggregated based on their share of total operating costs, in accordance with the formula mentioned above, to provide an aggregate measure of variable inputs. Once the aggregate output and input indices are computed, the variable factor productivity (VFP) of the airports is tabulated by dividing the output index by the input index, such that In VFP = . In p w 4.3.2 Variable Factor Productivity Index Number Efficiency Results The V F P results are illustrated in Figure 4.2 where the mean V F P values for each operating structure are shown for the study period. A s can be seen, the U S airport authorities had the highest average V F P value throughout the 10-year period, generally followed by the U S city-run airports, although the results between the city-run airports, the port authorities, and the Canadian airports converged in recent years. O f note is the systematic decrease in productivity found in 2001/2002 due to the 9/11 attacks (there was an industry-wide decrease in passenger outputs and an increase in operating expenses), and the subsequent rebound in productivity that occurred in 2003-2005. Detailed analysis of the factors affecting V F P is contained in Section 5. The rankings of individual airports according to V F P are shown in Appendix A . 5. Figure 4.2 Mean V F P Results by Airport Managerial Structure Variable Factor Productivi ty 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 4.4 Data Envelopment Analysis The second method o f productivity analysis employed in this paper is that o f data envelopment analysis ( D E A ) . D E A is a well-developed procedure, and for the sake of brevity, the derivation w i l l not be provided here. This mathematical programming methodology and the exposition o f the two forms of efficiency it estimates were initially put forth by Farrell (1957). The methodology was refined, and labelled as data envelopment analysis, by Charnes, Cooper, and Rhodes (1978). A n excellent reference is Fare, Grosskopf, and Love l l (1994), and thorough reviews of the methodology are contained in Charnes et al. (1994), A l i and Seiford (1993) and Coel l i et al. (2005). 4.4.1 D E A Background and Derivation D E A is a non-parametric mathematical programming model that uses linear programming methods in order to determine a piece-wise linear production frontier. This frontier encapsulates all o f the observed data points (no points can lie beyond the production frontier), and then calculates the efficiency of each airport relative to this estimated frontier. D E A has the ability to incorporate multiple inputs and multiple outputs, and has become a widely-used method of analysis, including several studies on airports, as was discussed in the literature review section. The primary advantage of using D E A is that it does not require an exhaustive amount of data, particularly input prices which can be difficult to obtain for airports. D E A does have its limitations; notably, there is not an underlying economic rationale in the derivation of weights used. Also , the model here is not intended to convey changes in productivity over time. The efficiency results are re-calibrated each year, so the relative efficiency of the airports can only be determined on a year-to-year basis with the model employed here; the factors affecting the D E A efficiency results are analyzed in Section 5. In relation to this frontier, the analyst is able to determine the technical efficiency o f each unit (an indicator of the unit's ability to generate maximal output from a given set of inputs) as well as the allocative efficiency of each unit (an indicator of how efficiently each unit utilizes their inputs, given the prices of each input and the production technology available) (Coell i et al., 2005). These two aspects of efficiency must be considered in conjunction in order to ascertain the total efficiency o f a firm. It should be noted that we are concerned with technical efficiency in this study. A s such, we are referring to the operational performance o f the airports when we discuss technical efficiency. The computation o f allocative efficiency is intrinsically difficult, given the prohibitive level of data required on the input and output prices facing each airport. In the case of the North American airports examined in this paper, the disparate regions exhibit a large fluctuation in input and output prices, and as such the construction of accurate price indices and the related examination o f allocative efficiency are beyond the scope of this paper. D E A can take two different orientations - an output orientation and an input orientation. The output orientation examines the degree to which "quantities can be proportionally expanded without altering the input quantities used", while the input orientation examines the degree to which "input quantities can be proportionally reduced without changing the output quantities produced" (Coell i , 2005). The analysis in this paper utilizes the input orientation. The demand facing airports is by and large exogenous to the airport management's decision-making process, as air transportation is a derived demand; the demand for air transport is dependent upon the demand for business trips, vacation plans, and other traveller decisions that are beyond the control of airport management. Airport management does, however, have a degree of control over the inputs and processes used at the airport. For this reason, the input-oriented approach is favoured for this analysis. Graphically, the theory o f technical efficiency can be understood by examining Figure 4.3. The production frontier is constructed from the observed data of the airports in the sample. In the case of the input orientation, each observation w i l l be either on the frontier or above and to the right o f the frontier. The technical efficiency o f each observation can be determined by examining the radial distance from the origin to the observed point. The technical efficiency measure w i l l be between 0 and 1; airports C and D w i l l have a technical efficiency rating o f 1.00, as they are located on the frontier. Airports A and B , however, are located beyond the frontier, and are thus relatively inefficient (they could theoretically produce the same output while using less input). The technical efficiency rating of airport A w i l l be calculated as OA 70A, and the technical efficiency rating of airport B w i l l be calculated as 0 B 7 O B (with both ratings being less than 1.00). Clearly, the farther the firm is from the frontier, the lower is their technical efficiency. Figure 4.3 An Illustration of D E A Input-Oriented Technical Efficiency X 2 / Y A ' C / / / / B D Production Frontier X i / Y The D E A methodology can also be illustrated mathematically. In addition to having both input and output orientations, D E A analysis can be formed with either the assumption of constant returns to scale (CRS) or variable returns to scale (VRS) . First, consider the C R S model (first proposed by Charnes, Cooper and Rhodes (1978)) 8: This section again draws upon Coelli (1998, pp. 140-142). • There are K inputs and M outputs for N firms • The column vectors Xj and yj represent the inputs and outputs, respectively, of the z t h firm. • The data for all N firms are represented by the K x N input matrix, X , and the M x N output matrix, Y . • A ratio of all outputs over all inputs ( U ' V J / V ' X J ) is determined, where u is an M x l vector of output weights, and v is a K x l vector of input weights. The optimal input and output weights are then determined by the following linear programming model: M a x u > v (u'yj/v'xj), Subject to (u'yj/v'xj) < 1, j= l ,2 , . . . ,N , u, v > 0. A n equivalent linear programming model (the envelopment form) is as follows: M i n 6 ) x 6 , Subject to -yj + Y1 > 0, 6 X J - X ^ > 0 , X,>0, where 0 is a scalar value representing the efficiency score (0 < 9 < 1) o f each firm, and X. is a N x l vector of constants. This problem is then iterated for each firm in the sample. Banker, Charnes, and Cooper (1984) proposed a revision to the constant returns to scale model above that allows for the presence of variable returns to scale. If a firm is not operating at optimal scale, then the efficiency score that is calculated is biased by the impact o f scale inefficiencies. The envelopment form of the linear programming model shown above can be adjusted when it is assumed that variable returns to scale exist. A n additional constraint, that of convexity (NVX = 1, where N l is an N x l vector o f Is), is added to the model specification, resulting in the following linear programming problem: Min e , x0 , Subject to -yj + Y A , > 0, 9 X J - X X > 0 , N 1 ' X = 1 \>0. The frontier in this situation becomes a so-called "convex hul l " that provides a closer fit with the observed data points than does the frontier in the constant returns to scale specification. A s such, the technical efficiency score in the V R S scenario is always greater than or equal to the technical efficiency score indicated by the C R S model. The additional constraint imposed in the V R S case has the effect that, in the V R S case, each inefficient firm is only compared to firms of a similar size; this measure tries to extricate the inefficiency attributable to scale inefficiencies. In the C R S model, each and every firm is compared with one another, regardless o f the size of each firm. 4.4.2 D E A Technical Efficiency Results In order to perform the D E A analysis, T i m Coell i ' s D E A P software was used 9. This software solves the envelopment form of the linear programming model. A s mentioned above, an input orientation was selected. The model included balanced panel data for the years 2001-2005 1 0 . The results presented here are intended to be comparable with the index number results. A s such, the technical efficiency estimates here have excluded capital inputs. The model includes 3 outputs (passengers, aircraft movements, and non-aeronautical revenue) and 2 inputs (labour expenses and soft cost expenses). Figure 4.4 shows the results o f the D E A analysis assuming constant returns to scale 1 1 . It should again be emphasized that Figure 4.4 does not represent technical efficiency change over time; the focus is on the relative values on a year-to-year basis. The rankings of individual airports according to D E A are shown in Appendix A . 6 . Figure 4.4 Mean D E A Results by Airport Managerial Structure Data Envelopment Analysis (CRS) 0.4 - » - US Authority —A— US City-Run —•— US Port-Run - • — Canada 2001 2002 2003 2004 2005 9 The software can be found at the Centre for Efficiency and Productivity Analysis' website located at http://www.uq.edu.aii/econoimcs/cepa/software.htm 1 0 The full study period of 1996-2005 was not used due to the unbalanced nature of the panel data prior to 2001 1 1 Note that the concept of 'returns to scale' is incomplete due to the lack of capital inputs 4.5 Stochastic Frontier Analysis The third productivity analysis method used is stochastic frontier analysis (SFA) . The foundation of stochastic frontier analysis was constructed by Meeusen and van den Broeck (1977) and Aigner, Love l l , and Schmidt (1977). Since then, many different extensions o f S F A have been developed, in order to deal with varying data availability and differing assumptions about the statistical characteristics of this data. S F A can be used to estimate both production functions and cost functions. In this case, a stochastic production frontier is estimated, and technical efficiency is assessed relative to this frontier. 4.5.1 Stochastic Frontier Analysis (SFA) Background and Derivation: Stochastic frontier analysis specifies a production function, and, as its name implies, assumes an inherent randomness in this function. In other words, the production function is not deterministic; it is subject to an error term which is postulated to consist o f two components that must be separated. The production function is specified as follows: Yj = XJP + (V; - Uj) where i= l , . . . ,N (the number of firms in the sample) The notation is as follows: • Y ; is the production of the / t h firm • Xj is a K x l vector of the ith firm's inputs • p represents a vector of parameters that must be estimated • Vj are random variables that are generally assumed to be independent and identically distributed [N(0,a v 2 )] • Uj are non-negative random variables that represent the technical inefficiency in the z t h firm's production activities (also assumed to be independent and identically distributed [ N ( ( W ) ] • Uj and Vj are assumed to be independent of one another This is illustrated by Figure 4.5. A Z represents the degree to which the observed production o f the firm falls short of the maximum possible level of production, given the estimated production function frontier generated by the observed input and output levels. A Z is composed of the two random variables, Vj and Uj. S F A attempts to determine how much of A Z is attributable to these two components. O f interest is Uj, as the components of Vj are beyond the control of the airport's management and obfuscate the true underlying technical efficiency o f the airport. S F A attempts to extricate stochastic effects and measurement error to isolate Uj and more accurately predict technical efficiency levels. Output Production Frontier A (Vi-UO Input(s) 4.5.2 SFA Technical Efficiency Results In order to estimate the technical efficiency of the airports, the stochastic frontier analysis was performed using T i m Coel l i ' s F R O N T I E R software 1 2. The dependent variable (the observed output) was the logarithm of the translog multilateral output index developed in Section 3.3, which aggregated passengers, air traffic movements and non-aeronautical revenue into a single index number. The two independent variables (the observed inputs) were labour expenditures and soft cost expenses. The production function was specified as a translog production frontier using balanced panel data for the 72 airports for the period 2001-2005. A truncated normal distribution was assumed, and the following quadratic production function was estimated: ln(Qi) = p 0 + Piln(Si) + p 2 ln(Li) + p 3 ln(Si) 2 + p 4 (K; ) 2 + p 5ln(Si)ln(Li) + (V; - Uj), where Qj, Si, and Lj are the multilateral output index, soft cost expenses, and labour expenses, respectively. Vj is assumed to be normally distributed and Uj has a truncated normal distribution. Coel l i et al (2005) indicates that the technical efficiency estimate is defined as: EFFj = E(expYi*|Uj, Xj)/ E(expYj*|Ui=0, Xj) where Y j * is the production of the i t h f i rm 1 3 . A s a stochastic production function was estimated in this case (as opposed to a stochastic cost function), the efficiency measure is equal 1 2 As with the DEAP program, the FRONTIER program can be downloaded from The Centre for Efficiency and Productivity Analysis' website at: http://www.uq.edu.au/economics/cepa/software.htm 1 3 The production of the i* firm is denoted as expY,* in this case as the model used the logarithm form of the dependent variable. to exp(-Ui). The technical efficiency results of the S F A are summarized in Figure 4.6. The U S airport authority airports again achieve the highest estimated efficiency. There is a large discrepancy between the relative ranking of the Canadian airports according to S F A and according to V F P and D E A . The cause of this discrepancy w i l l be explored in Section 5 1 4 . The rankings o f individual airports according to S F A are shown in Appendix A . 7 . Figure 4.6 Mean SFA Results by Airport Managerial Structure 0.7 0.6 0.5 0.4 0.3 0.2 2001 Stochastic Frontier Analysis - US Authority - US City-Run - US Port-Run - Canada 2002 2003 2004 2005 4.6 Unit Cost Index Analysis The above procedures have all dealt with the productivity of airports - how effectively they are able to transform inputs into outputs. Another approach is to consider the cost effectiveness of airports; that is, how costly is it for the airports to produce some specified level o f output? To this end, two figures are computed. First, a unit cost index was computed. The unit cost index is defined as follows: UnitCost - T°talOperatingExpenses AggregateOutputlndex Where operating expenses represent total expenditures on labour and soft costs, and are measured in 1996 $US and are not adjusted for regional cost levels, and the aggregate output index is that computed in the V F P measurements. These results are shown in Figure 4.7 and the rankings o f individual airports are contained in Appendix A . 8 . 1 4 The main difference in the SFA results is due to differences of scale; as shown in Section 5, SFA indicates significant efficiency gains associated with increasing output scale, which is the cause of the low ranking for Canadian airports by this methodology A similar measure, which has more intuitive appeal, is that of Operating Expense per Passenger. This is defined as: „ . _ _ _ TotalOperatingExpenses OperatingExpensePerPassenger = -— TotalPassengers Operating expenses are again shown in constant 1996 $US, so any changes represent 'real ' changes over time. The results are shown in Figure 4.8 and a complete listing by individual airport is included in Appendix A . 9 . Figure 4.8 Mean Operating Expense per Passenger by Airport Managerial Structure Operating Expense Per Passenger -m- US Authority —A— US City-Run - •— US Port-Run -•—Canada B A ^ f^c 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 4.7 Discussion of Results and Comparison between Methodologies In total, five measures of efficiency have been calculated. They are each 'gross' measures of efficiency, insofar as they have not been adjusted for input prices and they have not controlled for systematic factors affecting efficiency that are beyond managerial control. These factors w i l l be discussed in Section 5. It is of interest to compare how the different measures of efficiency view each of the airports. Table 4.2 provides the relative ranking o f the airports according to each measure, including the mean ranking of each airport and its standard deviation. Overall, the rankings are largely consistent, although some disparities do exist in some cases. Table 4.3 provides a Spearman Rank Correlation Matrix, which indicates the similarity o f rankings between the measures. Most correlation coefficients are reasonably high, although the stochastic frontier results appear to vary the most from the other measures. Table 4.2 Comparative Rankings between Methodologies - Year 2005 VFP Unit Cost Expense/Pax DEA SFA Mean St. Dev. A T L Atlanta 1 1 1 1 1 1.0 0.0 C L T Charlotte-Douglas 2 2 2 1 4 2.2 1.1 T P A Tampa 7 5 9 11 5 7.4 2.6 L A S Las Vegas 9 6 6 19 2 8.4 6.4 R D U Raleigh-Durham 4 3 7 10 18 8.4 6.0 M S P Minneapolis-St. Paul 3 9 5 33 3 10.6 12.8 C V G Cincinnati/Kentucky 10 12 3 22 12 11.8 6.8 Y V R Vancouver 6 10 31 6 6 11.8 10.9 P H X Phoenix 17 15 8 18 7 13.0 5.1 BN'A Nashville 8 8 10 17 23 13.2 6.6 F L L Ft. Lauderdale 23 17 12 7 8 13.4 6.7 Y Y C Calgary 13 14 21 1 20 13.8 8.0 M C O Orlando 20 22 18 9 9 15.6 6.2 S L C Salt Lake City 16 11 4 41 11 16.6 14.3 HNL Honolulu 34 19 14 13 10 18.0 9.5 S N A Santa Ana 19 24 30 1 27 20.2 11.5 Y Y J Victoria 5 4 23 8 62 20.4 24.5 A B Q Albuquerque 14 13 15 26 40 21.6 11.5 M K E Milwaukee 12 16 26 21 36 22.2 9.3 RIC Richmond 18 7 36 14 53 25.6 18.7 IND Indianapolis 27 18 38 28 22 26.6 7.5 R N O Reno-Tahoe 21 21 24 20 47 26.6 11.5 MCI Kansas City 35 23 25 27 26 27.2 4.6 P D X Portland 25 31 35 29 16 27.2 7.2 S E A Seattle 32 35 28 30 13 27.6 8.6 IAD Washington (Dulles) 26 34 44 24 14 28.4 11.3 C M H Columbus 15 20 37 36 37 29.0 10.7 R S W Southwest Florida 24 26 19 37 41 29.4 9.2 VFP Unit Cost Expense/Pax DEA SFA Mean St. Dev. PBI Palm Beach 38 28 34 12 42 30.8 11.7 D F W Dallas/Fort Worth 29 36 11 55 25 31.2 16.1 D E N Denver 30 41 29 40 24 32.8 7.4 IAH Houston 36 44 13 70 17 36.0 23.0 DTW Detroit 42 47 22 54 21 37.2 15.0 Y X E Saskatoon 22 27 49 23 65 37.2 19.0 D C A Washington (Reagan) 43 42 42 45 15 37.4 12.6 C L E Cleveland 46 39 39 39 28 38.2 6.5 MDW Chicago (Midway) 49 54 20 35 33 38.2 13.6 S A N San Diego 51 53 27 31 29 38.2 12.7 Y E G Edmonton 28 29 43 43 49 38.4 9.4 Y Q T Thunder Bay 39 30 54 1 68 38.4 25.4 Y W G Winnipeg 11 25 55 50 52 38.6 19.5 J A X Jacksonvil le 40 33 45 34 44 39.2 5.5 S M F Sacramento 53 40 48 25 30 39.2 11.8 P H L Philadelphia 52 48 17 63 19 39.8 20.7 B O S Boston 33 52 56 32 31 40.8 12.2 O A K Oakland 58 55 47 16 32 41.6 17.5 A U S Austin 47 38 33 52 43 42.6 7.4 O R D Chicago (O'Hare) 48 49 16 53 51 43.4 15.4 Y O W Ottawa 31 43 57 46 54 46.2 10.2 Y Q R Regina 41 32 46 48 66 46.6 12.5 PIT Pittsburgh 55 50 51 51 35 48.4 7.7 S T L St. Louis 57 51 32 68 34 48.4 15.3 Y X U London 56 37 67 15 69 48.8 22.8 Y H Z Halifax 44 46 53 56 57 51.2 5.9 L G A New York (LaGuardia) 37 66 58 60 45 53.2 11.9 BWI Baltimore-Washington 64 62 40 64 38 53.6 13.4 S J C San Jose 63 61 61 49 39 54.6 10.3 Y Q M Moncton 54 45 65 44 67 55.0 10.8 E W R Newark 50 68 64 38 59 55.8 12.0 M S Y New Orleans 69 64 41 . 59 55 57.6 10.7 L A X Los Angeles 62 59 50 61 58 58.0 4.7 S A T San Antonio 68 60 52 62 48 58.0 8.0 Y Y T St. John's 45 56 59 67 63 58.0 8.4 Y Q B Quebec 60 57 69 42 64 58.4 10.2 A L B Albany 61 58 63 58 56 59.2 2.8 S F O San Francisco 59 63 60 65 50 59.4 5.8 ONT Ontario 65 65 62 66 46 60.8 8.4 Y Y Z Toronto 66 70 68 47 60 62.2 9.3 MIA Miami 67 71 66 72 61 67.4 4.4 J F K New York (Kennedy) 70 72 71 57 70 68.0 6.2 Y Y G Charlottetown 71 67 70 71 72 70.2 1.9 Y S J Saint John 72 69 72 69 71 70.6 1.5 V F P 1.00 - - - -Unit C o s t Index 0.93 1.00 - - -E x p e n s e / P A X 0.73 0 .75 1.00 - -D E A 0.71 0.79 0.51 1.00 -S F A 0.60 0.58 0.79 0.47 1.00 V F P Unit C o s t Index E x p e n s e / P A X D E A S F A 5 T H E I M P A C T O F M A N A G E R I A L S T R U C T U R E O N E F F I C I E N C Y The cross-structural comparison in Section 3 provided an indication that the U S airport authorities achieved the highest operating efficiency and cost effectiveness. However, this section w i l l perform regression analyses to econometrically determine whether there are meaningful differences in efficiency depending upon managerial structure. Table 5.1 provides the results of five separate regression analyses performed, whereby the dependent variables are each o f the efficiency measures obtained in Section 4. Several factors that could potentially affect the efficiency results obtained are included as independent variables. The first five independent variables characterize operating characteristics deemed to be beyond managerial control. The percentage of non-aeronautical revenue is used as an indicator o f the business strategy o f management. The Canada-US exchange rate variable is used to capture any effects differing between countries that are separate from the differences in managerial structure. Finally, several dummy variables were included to capture differences in efficiency according to managerial structure. The results are quite consistent across each of the methodologies. There are some variations in the estimated effects of the various operating characteristics. O f interest in this study, however, are the results concerning commercialization and managerial structure, and the results w i l l be discussed in turn. 5.1 The Relationship between Commercialization and Efficiency A s the literature review mentioned, Oum et al (2006) found very strong evidence that there is a high correlation between airport commercialization and efficiency. Their findings are corroborated in this study: Increasing the percentage of non-aeronautical revenue has a significant positive effect on variable factor productivity and D E A technical efficiency, and also significantly increases cost effectiveness1 5. Oum et al (2006) believe diversifying revenue sources into commercial and other non-aeronautical business allows airports to achieve higher operating efficiency and that "many airports aim to increase revenues from commercial services and other non-aeronautical activities in order to reduce aviation user charges, thus attracting more airlines. Such business diversification strategies...exploit the well-known demand complementarity between aeronautical services and commercial services". 1 5 Note that the negative coefficient in the Unit Cost Index and Operating Expense per Passenger regression analysis is indicative of lower costs and the positive coefficient in the VFP and DEA regression analyses indicates higher efficiency. Table 5.1 Regression Results of Factors Affecting Operating Efficiency Dependent Variable Regression Form VFP OLS (log-tog) Unit Cost OLS (log-log) Oper. Expense per PAX OLS (log-log) DEA Tobit (lin-log) SFA Tobit (lin-log) Coefficient f-stat Coefficient f-stat Coefficient f-stat Coefficient f-stat Coefficient f-stat Intercept -3.919 - 23.430 - 10.649 - -6.7273 - 3.3067 -Output Scale (Index) 0.035 1.219 -0.065 -2.073 * -0.010 -0.294 0.1593 0.869 1.3476 7.150 * Aircraft Size (Pax/ATM) -0.134 -3.420 * 0.093 2.202 * -0.313 -5.638 * -0.8476 -3.751 -0.5944 -2.659 Runway Utilization (ATM/runway) 0.172 5.488 * -0.258 -7.665 * -0.524 -11.800 * 0.7558 3.835 0.3488 1.800 % International Pax -0.009 -1.472 -0.003 -0.505 0.432 -0.517 -0.0075 -0.217 -0.1621 -4.630 % Transfer/Connecting Pax 0.072 6.253 * -0.070 -5.696 * -0.131 -8.054 * -0.0689 -1.066 0.0520 0.810 % Non-Aeronautical Revenue 0.597 14.450 * -0.772 -17.390 * -0.421 -7.195 * 1.3871 4.071 0.2759 0.826 Canada-US Exchange Rate 0.080 0.104 -0.558 -0.674 * -0.559 -0.512 * " " Dummv Variables 0.4842 1.881 Canadian Airport Authority 0.218 4.983 * -0.177 -3.966 * -0.202 -4.327 * 0.8122 3.131 US Airport Authority 0.085 3.396 * -0.050 -1.951 * -0.061 -2.398 0.0542 0.379 0.3914 2.730 US Port Authority 0.005 0.163 0.183 5.117 * 0.323 6.840 * 0.2504 1.211 0.0932 0.453 Multiple Airports -0.134 -4.745 * 0.245 8.057 * 0.296 7.399 * -0.4100 -2.479 * -0.8254 -4.934 * Year 0.013 2.085 * 0.001 0.183 0.016 1.918 0.1237 3.102 -0.0372 -0.945 R2 0.420 0.522 0.538 0.371 0.763 Adjusted R Log-likelihood value Observations (n) 0.4092 6.29 687 0.5134 -11971.00 687 0.5296 -1122.90 687 60.87 360 272.29 360 Note: An asterisk next to the r-statistic indicates statistical significance at the 0.05 level In addition to demand complementarities between commercial and aeronautical services, it is interesting to examine whether there are efficiency complementarities as wel l . To do so, the aggregate output index was re-calculated, removing the non-aeronautical revenue output. The V F P regression results were then carried out in order to isolate the impact of non-aeronautical revenues on the efficiency of aeronautical activities, with the results presented in Table 5.2. Table 5.2 Impact of Non-Aeronautical Revenue on Aeronautical Efficiency Dependent Variable Regress ion Form V F P (non-aeronautical revenue output removed) O L S (log-log) Coefficient f-stat Intercept % Non-Aeronautical Revenue Dummy Variables Canadian Airport Authority U S Airport Authority U S Port Authority  -0.587 0.150 0.097 0.123 -0.191 2.612 2.336 3.251 -4.019 Adjusted R Log-likelihood value Observat ions (n) 0.075 0.0699 -313.60 687 Note: * represents statistical signif icance at the 0.05 level, A at the 0.1 level This analysis provided interesting results; increasing the percentage o f non-aeronautical revenues by 10% increases aeronautical efficiency by 1.5%, over and above the direct benefits of increasing revenues. Further research could be beneficial in determining whether economies of scope exist between commercial and aeronautical activities, or whether this relationship is a reflection o f skilled management being concurrently more technically efficient and more proactive in generating commercial revenues. 5.2 The Effects of Managerial Structure on Efficiency Table 5.1 addresses the differences in efficiency between managerial structures. The explanatory variables allow for an extrication of efficiency effects attributable to differences in structure. The results are very consistent across all five measures o f efficiency: the airport authority structure achieves significant improvements in both productive efficiency and cost effectiveness relative to the government-run airports 1 6. A s Table 5.3 shows, after controlling for exogenous factors, Canadian airport authorities are between 12%-24% more efficient than U S city-run airports, and U S airport authorities are between 5%-12% more efficient than U S city-run airports. Table 5.3 Efficiency Differences between Authorities and City-Run Airports Methodology Canadian Airport Authorities US Airport Authorities V F P 2 4 % more eff icient 9 % more eff icient D E A 1 2 % more eff icient no d i f ference S F A 1 5 % more eff icient 1 2 % more eff icient Unit C o s t Index 1 6 % m o r e cos t ef fect ive 5 % more cos t ef fect ive E x p e n s e per P a x 1 8 % more cos t ef fect ive 6 % more cos t ef fect ive Note: R e s u l t s relat ive to U S city-run airports Next, the regression results in Table 5.1 can be used to create a residual measure of V F P that explicitly controls for factors beyond managerial control. To do so, the observed V F P is compared to the expected V F P , given the operating characteristics o f the airport. The residual (either positive or negative) is then attributed to managerial ski l l , and the impact o f output scale, aircraft size, runway utilization, the percentage o f international passengers, and the percentage o f transferring/connecting passengers is thus removed from the V F P measure. A one-way A N O V A analysis was then conducted to determine whether managerial efficiency was dependent upon airport structure. A s Table 5.4 shows, there is again strong evidence that both the Canadian and the U S airport authorities outperform the U S city-run airports. Table 5.4 The Effects of Managerial Structure on Efficiency - Residual V F P Management Structure Count Sum Average Variance C a n a d i a n Author i t ies 140 487 .44 3.48 0.51 U S Author i t ies 189 518 .23 2.74 0.45 U S C i t y - R u n 278 699 .06 2.51 0.49 U S P o r t - R u n 80 217 .37 2.72 0.20 Source of Variation S S df MS F B e t w e e n G r o u p s 88 .889 3 29 .63 66 .29 Wi th in G r o u p s 305 .274 6 8 3 0 .45 Tota l 3 9 4 . 1 6 3 686 The three dummy variables for Canadian authorities, US authorities, and US port-run airports are relative to the base case of US city-run airports Finally, what are the implications of managerial efficiency on the level o f user charges? Forsyth (2000) believes that since airports possess considerable monopoly power, they thus have the scope to operate inefficiently, and pass on the higher costs which result from this inefficiency to their customers (i.e. the airlines). There is no evidence o f this for the North American airports, however, as there is no correlation between the level of managerial efficiency (in the form of residual V F P ) and the level o f user charges (in the form of aeronautical revenues per passengers), as shown in Figure 5.1. Figure 5.1 Relationship between Managerial Efficiency and Aeronautical Charges 0> a . 0) 14 12 10 fl) >-> <D * S 8 3 ~ n 3 Q-<o c s < 4 2 V • • • • • Residual V F P 6 C O N C L U S I O N 6.1 Summary of Key Findings Prior to the implementation of Canada's National Airports Policy, several anticipated results were put forth. The main focus was on the expected increase in operating efficiency and commercialization of operations and the long-term goal of financial self-sustainability. N o w , five years subsequent to its implementation, there is strong evidence that such proclamations were more than "policy speak"; the benefits o f the airport authority structure are indeed borne out in the data, and robust across several different measures o f operating efficiency. Both U S and Canadian airports generate significantly more non-aeronautical revenue than do the U S city-run airports, and the authorities have also achieved much higher growth rates over the past decade. Additionally, both the authorities are more efficient and more cost effective than the city-run airports; on the order of 12-24% for Canadian airport authorities and 5-12% for U S airport authorities. When factors beyond managerial control are controlled for, the efficiency advantage o f the authority structure persists. Potential sources for the higher efficiency o f the airport authorities are: • Greater managerial autonomy: financially, operationally, and/or strategically • A more effective governance structure owing to a specialized Board of Governors • A reduction in X-inefficiency associated with public sector bureaucracy • Increased incentives due to the ability to re-invest retained earnings Inter-related to these findings, the study also found that airports that focus on generating non-aeronautical revenues are more technically efficient, regardless of whether non-aeronautical revenue is classified as an output. 6.2 Suggestions for Further Research There are several potential areas for further research. Work could be done to incorporate capital assets in order to get a more holistic view of airport efficiency and to assess how effective the governance structures are in determining levels o f capital investment. Further research into the linkage between efficiency, non-aeronautical revenues, and aeronautical charges is also warranted in order to obtain a better understanding of the causes and the effects since the three factors are in many ways tied together. Finally, while there is evidence that the N A P has been successful and that the U S should further embrace the airport authority structure, it remains to be seen whether the airport authority structure is indeed the optimal structure for the North American airport industry. If the benefits of the airport authority structure espoused are accurate, it is l ikely that these benefits would be even stronger under privatization. Would a move towards privatization, with an appropriate regulatory framework to control for market power, represent a further improvement? A t present, no Canadian airports are privatized (either fully or partially), so this thesis was not able to compare the performance of the current institutional forms with partial or fully privatized forms. Wi th the important public policy and industry implications inherent in the type o f airport ownership and managerial forms, this is an area that warrants further research. 7 B I B L I O G R A P H Y Abbott, M . , and S. W u (2002). "Total factor productivity and efficiency of Australian Airports," The Australian Economic Review 35, 244-260. Aigner, D . , C . A . K . Lovel l , and P. Schmidt (1997). "Formulation and Estimation o f Stochastic Frontier Production Function Models," Journal of Econometrics, 6, 21-37. A i r Transport Research Society (2005). Airport Benchmarking Report: Global Standards for Airport Excellence. A l i , A . I . , and L . M . Seiford (1993). "The Mathematical Programming Approach to Efficiency Analysis ," in Fried, H.O. , C . A . K . Love l l and S.S. Schmidt (eds.), The Measurement of Productive Efficiency: Techniques and Applications. Oxford University Press, N e w York , 120-159. Banker, R . D . , A . Charnes, and W . W . Cooper (1984). "Some Models for Estimating Technical and Scale Inefficiencies in Data Envelopment Analysis," Management Science, 30, 1078-1092. Canada Transportation Ac t Review: Vis ion and Balance, 2001 Caves, D . , Christensen, L . R . , and W . E . Diewert (1982). "Multilateral comparisons o f output, input and productivity using superlative index numbers," Economic Journal, 92, 73-86. Charnes, A . , W . W . Cooper, and E . Rhodes (1978). "Measuring the Efficiency o f Decision Making Units," European Journal of Operational Research, 2, 429-444. Charnes, A . , W . W . Cooper, A . Y . Lewin, and L . M . Seiford (1994). "Data Envelopment Analysis: Theory, Methodology and Applications, Kluwer Academic Publishers, Boston. Christensen, L . R . and D . Jorgensen (1969). "The Measurement o f U . S . Real Capital Input, 1929-1967," Review of Income and Wealth, 15(1), 293-320. Coel l i , T., D.S . Prasada Rao, C. O'Donnell , and G . E . Battese (2005). A n Introduction to Efficiency and Productivity Analysis, 2nd Edition. Boston: Kluwer Academic Publishers. Craig, S., Ai ro la , J., and M . Tipu (2005). "The effect of institutional form on airport governance efficiency," Unpublished manuscript, Department of Economics, University of Houston, Houston, Texas. Diewert, W . E . (1980). "Capital and Theory of Productivity Measurement," American Economic Review, 70 (2), 260-267. Doganis, R. (1992). The Airport Business. Routledge: London, U K and N e w York, N Y . Farrell, M . J . (1957). "The Measurement o f Productive Efficiency," Journal of the Royal Statistical Society, Series A (General), 120(3), 253-281. Fare, R., Grosskopf, S., and C . A . K . Lovel l (1994). Production Frontiers. Cambridge: Cambridge, University Press. Forsyth, P. (2000). "Models of Airport Performance," Handbook of Transport Modell ing, D . A . Hensher, K . J . Button, eds. Gesell, L . (1999). The Administration of Public Airports, 4th Edition. Coast Ai re Publications: Chandler, A Z . Gi l len , D . , and A . L a l l (1997). "Developing measures o f airport productivity and performance: an application of data envelopment analysis," Proceedings o f the A i r Transport Research Group Conference, Vancouver. Graham, A . (2003). Managing Airports: A n International Perspective, 2nd Edition. Elsevier. Heaver, T., and T. Oum (2001). "Commercialization of Airports and A i r Navigation Systems: The Canadian Experience," Journal of Aviat ion Management, 2000, 32-44. Hooper, P .G. , and D . G . Hensher (1997). "Measuring total factor productivity of airports: an index number approach," Transportation Research E 33, 249-259. Inamete, U . B . (1993). " K e y elements in managing airports: the policy environment and increasing efficiency," International Journal of Public Sector Management 6, 12-23. Jeong, J. (2005). " A n investigation of operating costs of airports: focus on the effects o f output scale," Master of Science Thesis, University of British Columbia, Vancouver, Canada. Kapur, A . (1995). Airport Infrastructure: The Emerging Role of the Private Sector. Wor ld Bank Technical Paper, Number 313. Martin, J .C., and C. Roman (2001). " A n application of D E A to measure the efficiency of Spanish airports prior to privatization," Journal of A i r Transport Management 7, 149-157. Martin-Cejas, R .R . (2002). " A n approximation to the productive efficiency of the Spanish airports network through a deterministic cost frontier," Journal of A i r Transport Management 8, 233-238. Meeusen, W . and J. van den Broeck (1977). "Efficiency Estimation from Cobb-Douglas Production Functions with Composed Error," International Economic Review, 8, 435-444. Nyshadham, E . A . , and V . K . Rao (2000). "Assessing efficiency of European airports: a total factor productivity approach," Public Works Management and Pol icy 5, 106-114. Oum, T . H . , and Y . Zhang (1990). "Airport Pricing: Congestion Tolls, Lumpy Investment, and Cost Recovery," Journal of Public Economics, 43, 353-374. Oum, T . H . , M . W . Tretheway, and W . G . Waters (1992). "Concepts, Methods, and Purposes of Productivity Measurement in Transportation," Transportation Research A , 26A (6), 493-505. Oum, T . H . , Waters, W . G . , and C . Y u (1999). "Survey of productivity and efficiency measurement in rail transport," Journal of Transport Economics and Pol icy 33, 9-42. Oum, T . H . , Y u , C , and X . Fu (2003). " A comparative analysis of productivity performance o f the world's major airports: summary report o f the A T R S global airport benchmarking research report - 2002," Journal of A i r Transport Management 9, 285-297. Oum, T., Adler, N . , and C . Y u (2006). "Privatization, corporatization, ownership forms and their effects on the performance of the world's major airports," Journal of A i r Transport Management, V o l . 12, 109-121. Parker, D . (1999). "The performance o f B A A before and after privatization," Journal of Transport Economics and Policy 33,133-145. Salazar de la Cruz, F. (1999). " A D E A Approach to the Airport Production Function," International Journal of Transport Economics, 26, 255-270 Sarkis, J. (2000). " A n analysis of the operational efficiency of major airports in the United States," Journal of Operations Management 18, 335-351. Tretheway, M . (2001). "Airport Ownership, Management, and Price Regulation," Research conducted for the Canada Transportation Ac t Review. Wiley, J.R. (1986). Airport Administration and Management. Eno Foundation For Transportation, Inc. Westport: C T . Y o k o m i , M . (2005). "Evaluation of technical efficiency at privatized airports: Case o f B A A P i c , " Paper presented at the A i r Transport Research Society ( A T R S ) Conference, July 3-6, 2005, Rio de Janiero, Brazi l . APPENDICES Appendix A.1 Canadian Airport Authorities Included in Study Airport Authorities - Canada Airport Name Owner Operator Date Transferred Y Y C Calgary International Airport Transport Canada Calgary Airport Authority July 1,1992 Y Y G Charlottetown Airport Transport Canada Charlottetown Airport Authority Inc. March 1,1999 Y E G Edmonton International Airport Transport Canada Edmonton Regional Airports Authority August 1,1992 Y Q M Greater Moncton International Airport Transport Canada Greater Moncton International Airport Authority September 1,1997 Y H Z Halifax Airport Transport Canada Halifax International Airport Authority February 1,2000 Y Q B Jean Lesage International Airport Transport Canada Aeroport de Quebec Inc November 1,2000 Y X U London International Airport Transport Canada Greater London International Airport Authority August 1,1998 Y O W Ottawa International Airport Transport Canada Ottawa Macdonald Cartier Intl. Airport Authority February 1,1997 Y Q R Regina Airport Transport Canada Regina Airport Authority May 1,199S Y S J Saint John Airport Transport Canada Saint John Airport Inc. June 1,1999 Y X E Saskatoon John G . Diefenbaker International Airport Transport Canada Saskatoon Airport Authority January 1,1999 Y Y T St. John's International Airport Transport Canada St. John's International Airport Authority December 1,1998 Y Q T Thunder Bay International Airport Transport Canada Thunder Bay International Airports Authority Inc. September 1,1997 Y Y Z Toronto Pearson International Airport Transport Canada Greater Toronto Airports Authority December 2,1996 Y V R Vancouver International Airport Transport Canada Vancouver International Airport Authority July 1,1992 Y Y J Victoria Airport Transport Canada Victoria Airport Authority April 1,1997 Y W G Winnipeg International Airport Transport Canada Winnipeg Airports Authority January 1,1997 Airport Authorities - United States I AT A Code Airport Name Owner Operator ALB Albany International Airport Albany County Airport Authority Albany County Airport Authority BNA Nashville International Airport Metropolitan Nashville Airport Authority Metropolitan Nashville Airport Authority C M H Port Columbus International Airport Columbus Regional Airport Authority Columbus Regional Airport Authority C V G Cincinnati/Northern Kentucky International Airport Kenton County Airport Board Kenton County Airport Board D C A Ronald Reagan Washington National Airport Metropolitan Washington Airports Authority Metropolitan Washington Airports Authority DFW Dallas/Fort Worth International Airport Cities of Dallas and Fort Worth DFW Airport Board DTW Detroit Metropolitan Wayne County Airport Wayne County Wanye County Airport Authority IAD Washington Dulles International Airport Metropolitan Washington Airports Authority Metropolitan Washington Airports Authority IND Indianapolis International Airport Indianapolis Airport Authority Indianapolis Airport Authority (BAA Indianapolis LLC) J A X Jacksonville International Airport Jacksonville Airport Authority Jacksonville Airport Authority M C O Orlando International Airport City of Orlando Greater Orlando Aviation Authority M S P Minneapolis/St. Paul International Airport Metropolitan Airports Commission Metropolitan Airports Commission PIT Pittsburgh International Airport Allegheny County Airport Authority Allegheny County Airport Authority RDU Raleigh-Durham International Airport Raleigh-Durham Airport Authority Raleigh-Durham Airport Authority RIC Richmond International Airport Capital Region Airport Commission Capital Region Airport Commission R N O Reno/Tahoe International Airport Reno-Tahoe Airport Authority Reno-Tahoe Airport Authority S A N San Diego International Airport San Diego County Regional Airport Authority San Diego County Regional Airport Authority STL St. Louis-Lambert International Airport City of St. Louis St. Louis Airport Authority T P A Tampa International Airport Hillsborough County Aviation Authority Hillsborough County Aviation Authority Appendix A.3 US City-Run Airports Included in Study City-Run Airports - United States IATA Code Airport Name Owner Operator A B Q Albuquerque International Sunport City of Albuquerque Aviation Department ATL Hartsfield-Jackson Atlanta International Airport City of Atlanta Department of Aviation AUS Austin-Bergstrom International Airport City of Austin Department of Aviation BWI Baltimore Washington International Airport State of Maryland Maryland Aviation Administration C L E Cleveland-Hopkins International Airport City of Cleveland City's Department of Port Control, Airport Division CLT Charlotte Douglas International Airport City of Charlotte Department of Aviation DEN Denver International Airport City and County of Denver Department of Aviation FLL Fort Lauderdale Hollywood International Airport Broward County Broward County Aviation Department HNL Honolulu International Airport State of Hawaii Airports Division, Department of Transportation IAH Houston-Bush Intercontinental Airport City of Houston Houston Airport System LAS Las Vegas McCarran International Airport Clark County Clark County Department of Aviation LAX Los Angeles International Airport City of Los Angeles Los Angeles World Airports (City Department) MCI Kansas City International Airport City of Kansas City Kansas City Aviation Department MDW Chicago Midway Airport City of Chicago Chicago Airport System - Department of Aviation MIA Miami International Airport Miami-Dade County Miami-Dade Aviation Department MKE General Mitchell International Airport Milwaukee County Milwaukee County - Department of Public Works MSY Louis Armstrong New Orleans International Airport City of New Orleans New Orleans Aviation Board ONT Ontario International Airport City of Los Angeles Los Angeles World Airports (City Department) ORD Chicago O'Hare International Airport City of Chicago Chicago Airport System - Department of Aviation PBI Palm Beach International Airport Palm Beach County Palm Beach County - Department of Airports PHL Philadelphia International Airport City of Philadelphia City of Philadelphia, Department of Commerce - Division of Aviation PHX Phoenix Sky Harbor International Airport City of Phoenix City of Phoenix - Aviation Department SAT San Antonio International Airport City of San Antonio Department of Aviation S F O San Francisco International Airport City and County of San Francisco Airport Commission (department of the City and County of San Francisco) S J C Norman Y. Mineta San Jose International Airport City of San Jose City of San Jose - Airport Department S L C Salt Lake City International Airport Salt Lake City Salt Lake City Department of Airports SMF Sacramento International Airport County of Sacramento Sacramento County Airport System - Department within County SNA John Wayne Airport Orange County Orange County - Department Appendix A.4 US Port-Run Airports Included in Study Port-Run Airports - United States Airport Name Owner Operator B O S Boston Logan International Airport Massachusetts Port Authority Massachusetts Port Authority - Aviation Department E W R Newark Liberty International Airport Port Authority of New York and New Jersey Port Authority of New York and New Jersey J F K New York-John F. Kennedy International Airport Port Authority of New York and New Jersey Port Authority of New York and New Jersey L G A LaGuardia International Airport Port Authority of New York and New Jersey Port Authority of New York and New Jersey O A K Oakland International Airport Port of Oakland Port of Oakland - Aviation Division P D X Portland International Airport Port of Portland Port of Portland - Aviation Division R S W Southwest Florida International Airport Lee County Port Authority Lee County Port Authority S E A Seattle-Tacoma International Airport Port of Seattle Port of Seattle - Aviation Division O N Canadian Airport Authorities Variable Factor Productivity 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 YYJ Victoria - 1.53 1.67 1.76 1.72 1.74 1.49 1.58 1.61 162 YVR Vancouver 1.51 1.57 1.59 1.64 1.75 1.61 1.47 1.40 1.48 1.48 YWG Winnipeg - 1.25 1.21 1.20 1.19 1.07 1.15 1.19 1.26 1.29 YYC Calgary 1.71 1.50 1.42 1.33 1.34 1.27 1.21 1.05 1.13 1.22 YXE Saskatoon - - - 1.08 1.27 1.24 1.10 1.04 1.12 1.13 YEG Edmonton 0.85 0.87 0.86 0.92 1.01 0.98 0.97 0.94 1.02 1.08 YOW Ottawa - 1.11 1.15 1.20 1.25 1.26 1.21 1.08 1.03 1.05 YQT Thunder Bay - - 0.89 0.90 0.97 0.99 1.03 1.19 1.07 1.00 YQR Regina - - - 0.92 0.91 0.81 0.79 0.81 0.92 0.96 YHZ Halifax - - - - 1.04 0.83 0.83 0.83 0.89 0.90 YYT St. John's - 0.87 0.92 0.88 0.85 0.83 0.77 0.81 0.82 0.89 YQM Moncton - - 0.63 0.74 0.82 0.82 0.74 0.74 0.76 0.81 YXU London - - - 0.89 0.93 0.78 0.76 0.79 0.80 0.79 YQB Quebec - - - - - 0.76 0.73 0.66 0.64 0.73 YYZ Toronto 1.10 1.13 1.00 0.94 0.87 0.75 0.68 0.52 0.57 0.54 YYG Charlottetown - 0.48 0.59 0.67 0.50 0.52 0.51 0.51 0.51 0.51 YSJ Saint John - - - 0.60 0.59 0.54 0.46 0.42 0.48 0.49 Mean 1.29 1.15 1.08 1.04 1.06 0.99 0.94 0.91 0.95 0.97 US Airport Authorities Variable Factor Productivity 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 MSP Minneapolis-St. Paul 1.47 1.63 1.49 1.55 1.58 1.42 1.58 2.23 1.68 1.71 RDU Raleigh-Durham - 1.73 1.70 1.95 2.08 2.00 2.04 1.71 1.59 1.64 TPA Tampa 1.04 1.10 1.17 0.76 1.31 1.29 1.33 1.27 1.34 1.41 BNA Nashville 1.24 1.22 1.35 1.34 1.36 1.40 1.32 1.26 1.36 1.39 C V G Cincinnati/Kentucky 1.25 1.31 1.35 1.41 1.38 1.20 1.32 1.32 1.35 1.34 CMH Columbus 1.10 1.08 1.15 1.09 1.18 1.34 1.26 1.20 1.19 1.20 RIC Richmond 1.02 1.19 1.19 1.23 1.25 1.27 1.18 1.20 1.08 1.16 MCO Orlando 1.20 1.25 1.20 1.18 1.12 1.06 1.15 1.04 1.09 1.14 RNO Reno-Tahoe 1.14 1.15 1.18 1.17 1.18 1.23 1.06 0.99 1.07 1.14 IAD Washington (Dulles) 0.78 0.82 0.90 0.85 0.81 0.75 0.68 0.80 1.04 1.10 IND Indianapolis 0.95 0.78 0.78 0.76 0.83 2.00 2.08 1.11 1.05 1.09 DFW Dallas-Fort Worth 1.04 1.12 1.62 1.35 1.30 1.17 1.19 1.11 1.32 1.06 JAX Jacksonville 0.90 0.99 1.04 1.02 1.01 1.26 0.94 0.91 0.88 0.96 DTW Detroit 1.00 1.07 1.04 1.03 1.07 1.07 0.90 0.84 0.83 0.94 DCA Washington (Reagan) 0.60 0.63 0.58 0.63 0.86 0.88 0.75 0.79 0.88 0.92 SAN San Diego 0.62 0.55 0.85 0.78 0.77 0.72 0.60 0.67 0.85 0.84 PIT Pittsburgh 0.67 0.70 0.71 0.68 0.64 1.05 1.00 0.91 0.86 0.80 STL St. Louis 1.32 1.26 1.28 1.31 1.39 1.26 1.18 0.98 0.83 0.79 ALB Albany 0.67 0.69 0.71 0.71 0.83 0.85 0.81 0.78 0.83 0.71 Mean 1.00 1.07 1.12 1.09 1.16 1.22 1.18 1.11 1.11 1.12 Variable Factor Productivity 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 ATL Atlanta 2.31 2.73 2.99 2.89 3.03 CLT Charlotte-Douglas - 1.80 1.84 1.78 1.85 LAS Las Vegas 1.23 1.28 1.25 1.26 1.36 MKE Milwaukee 1.17 0.97 0.97 1.03 1.09 ABQ Albuquerque 0.66 0.61 1.31 1.24 1.27 SLC Salt Lake City 1.27 1.16 1.23 1.17 1.16 PHX Phoenix 1.35 1.33 1.40 1.37 1,43 SNA Santa Ana 1.11 1.17 1.18 1.18 1.35 FLL Fort Lauderdale 0.79 0.85 1.05 1.10 1.20 DEN Denver 0.87 0.82 0.90 0.85 0.82 HNL Honolulu 1.85 1.72 1.64 1.64 1.63 MCI Kansas City 0.97 0.99 1.01 1.41 1.30 IAH Houston 0.95 0.99 1.04 1.05 1.18 PBI Palm Beach 0.85 0.91 1.01 0.95 1.07 CLE Cleveland 0.77 0.68 0.86 0.75 1.12 AUS Austin 0.92 0.86 0.76 0.83 0.96 ORD Chicago (O'Hare) 0.72 0.77 1.07 0.99 1.02 MDW Chicago (Midway) 1.14 1.02 1.32 1.37 1.46 PHL Philadelphia 0.70 0.77 0.80 0.72 0.79 SMF Sacramento 0.76 0.81 0.78 0.79 0.82 SFO San Francisco 0.91 0.92 0.86 0.82 0.89 LAX Los Angeles 0.79 0.79 0.94 0.84 0.76 SJC San Jose 0.81 0.89 0.93 0.82 0.76 BWI Baltimore-Washington 0.89 0.96 0.97 1.10 1.25 ONT Ontario 0.68 0.67 0.76 0.69 0.70 MIA Miami 0.88 0.87 0.82 0.85 0.80 SAT San Antonio 0.74 0.72 0.73 0.71 0.76 MSY New Orleans 0.60 0.58 0.61 0.57 0.56 Mean 0.99 1.02 1.11 1.10 1.16 Mean (excluding ATL) 0.94 0.96 1.04 1.03 1.09 2.74 1.85 1.31 1.24 1.21 1.16 1.45 1.33 1.19 0.80 1.55 1.20 1.07 1.01 0.89 0.88 0.87 1.21 0.79 0.83 0.73 0.69 0.84 1.11 0.60 0.63 0.86 0.71 1.10 1.04 2.49 1.84 1.23 1.12 1.24 1.15 1.16 1.01 1.00 0.80 0.93 0.92 0.94 0.86 0.76 0.86 0.83 0.91 0.76 0.71 0.63 0.64 0.75 0.88 0.68 0.68 0.90 0.66 0.98 0.92 2.79 1.72 1.22 1.20 1.18 1.09 1.20 1.06 1.02 0.92 0.88 1.00 0.94 0.92 0.82 0.78 0.85 0.95 0.64 0.75 0.57 0.65 0.71 0.67 0.61 0.54 0.86 0.69 0.97 0.90 2.81 1.71 1.43 1.26 1.16 1.08 1.18 1.12 1.09 1.04 0.99 1.00 1.00 0.97 0.94 0.86 0.92 0.97 0.70 0.73 0.77 0.67 0.70 0.67 0.60 0.55 0.86 0.66 1.02 0.95 1.78 1.38 1.28 1.20 1.18 1.17 1.15 1.12 1.06 1.02 1.02 1.01 1.00 0.87 0.87 0.86 0.85 0.84 0.84 0.74 0.68 0.66 0.63 0.60 0.53 0.53 0.53 0.94 0.94 US Port-Run Airports Variable Factor Productivity 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 RSW Southwest Florida 0.86 0.84 0.86 0.92 0.94 0.87 0.88 0.95 1.04 1.11 PDX Portland 0.80 1.22 1.09 1.23 1.22 1.16 1.12 1.07 1.13 1.10 SEA Seattle 1.25 1.19 1.05 1.08 1.00 0.92 0.88 0.92 0.99 1.04 BOS Boston 1.11 1.15 1.22 1.18 1.25 1.13 1.11 1.08 1.02 1.04 LGA New York (LaGuardia) 0.90 1.01 0.96 0.81 0.90 0.85 0.74 0.85 0.94 1.01 EWR Newark 0.89 0.97 0.95 0.78 0.77 0.81 0.66 0.73 0.82 0.85 OAK Oakland 0.81 0.86 0.81 0.81 0.79 0.85 0.77 0.77 0.81 0.77 JFK New York (Kennedy) 0.54 0.57 0.54 0.54 0.49 0.52 0.51 0.43 0.46 0.52 Mean 0.89 0.98 0.93 0.92 0.92 0.89 0.83 0.85 0.90 0.93 Canadian Airport Authorities DEA - Without Capital Inputs (CRS) 2001 2002 2003 2004 2005 Y Y C Calgary 0.813 0.827 0.965 0.890 1.000 YQT Thunder Bay 0.851 1.000 1.000 1.000 1.000 Y V R Vancouver 0.772 0.881 1.000 1.000 0.991 Y Y J Victoria 1.000 1.000 1.000 1.000 0.912 Y X U London 0.967 0.918 0.867 0.859 0.743 Y X E Saskatoon 0.673 0.641 0.722 0.661 0.622 YQB Quebec 0.629 0.689 0.635 0.555 0.509 Y E G Edmonton 0.358 0.387 0.421 0.464 0.508 Y Q M Moncton 0.580 0.647 0.575 0.542 0.508 YOW Ottawa 0.512 0.487 0.510 0.507 0.501 Y Y Z Toronto 0.419 0.458 0.479 0.511 0.497 Y Q R Regina 0.385 0.414 0.449 0.466 0.482 Y W G Winnipeg 0.473 0.481 0.494 0.467 0.470 YHZ Halifax 0.296 0.334 0.388 0.405 0.444 YYT St. John's 0.387 0.380 0.401 0.359 0.334 Y S J Saint John 0.322 0.333 0.313 0.318 0.319 Y Y G Charlottetown 0.286 0.289 0.296 0.331 0.301 Mean 0.572 0.598 0.619 0.608 0.597 US Airport Authorities DEA - Without Capital Inputs (CRS) 2001 2002 2003 2004 2005 MCO Orlando 0.757 0.865 0.902 0.909 0.878 RDU Raleigh-Durham 1.000 1.000 1.000 0.821 0.861 TPA Tampa 0.692 0.769 0.852 0.874 0.849 RIC Richmond 0.667 0.737 0.932 0.773 0.782 BNA Nashville 0.522 0.509 0.623 0.677 0.720 RNO Reno-Tahoe 0.823 0.630 0.622 0.650 0.682 C V G Cincinnati/Kentucky 0.561 0.763 0.733 0.698 0.643 IAD Washington (Dulles) 0.298 0.296 0.483 0.614 0.610 IND Indianapolis 0.721 0.698 0.650 0.572 0.593 SAN San Diego 1.000 1.000 1.000 0.579 0.568 MSP Minneapolis-St. Paul 0.470 0.540 0.998 0.599 0.565 JAX Jacksonville 0.602 0.440 0.514 0.549 0.562 CMH Columbus 0.578 0.578 0.556 0.545 0.561 DCA Washington (Reagan) 0.358 0.301 0.472 0.519 0.503 PIT Pittsburgh 0.460 0.443 0.438 0.510 0.469 DTW Detroit 0.354 0.322 0.413 0.425 0.454 DFW Dallas-Fort Worth 0.429 0.521 0.399 0.491 0.446 ALB Albany 0.379 0.373 0.431 0.480 0.426 STL St. Louis 0.493 0.479 0.372 0.377 0.327 Mean 0.588 0.593 0.652 0.614 0.605 DEA - Without Capital Inputs (CRS) 2001 2002 2003 2004 2005 ATL Atlanta 1.000 1.000 1.000 1.000 1.000 CLT Charlotte-Douglas 1.000 1.000 1.000 1.000 1.000 SNA Santa Ana 0.893 0.834 1.000 1.000 1.000 FLL Ft. Lauderdale 0.675 0.697 0.797 0.801 0.953 PBI Palm Beach 0.616 0.608 0.765 0.773 0.812 HNL Honolulu 1.000 0.649 0.839 0.828 0.804 PHX Phoenix 0.653 0.613 0.743 0.737 0.696 LAS Las Vegas 0.530 0.555 0.663 0.706 0.695 MKE Milwaukee 0.530 0.566 0.691 0.673 0.655 S M F Sacramento 0.459 0.465 0.559 0.547 0.607 A B Q Albuquerque 0.491 0.519 0.623 0.596 0.603 MCI Kansas City 0.552 0.476 0.551 0.603 0.601 MDW Chicago (Midway) 0.285 0.417 0.488 0.607 0.561 C L E Cleveland 0.414 0.381 0.417 0.514 0.518 DEN Denver 0.321 0.362 0.438 0.486 0.513 S L C Salt Lake City 0.468 0.527 0.469 0.478 0.512 S J C San, Jose 0.496 0.481 0.496 0.483 0.480 A U S Austin 0.431 0.439 0.445 0.453 0.463 O R D Chicago (O'Hare) 0.271 0.315 0.378 0.447 0.461 M S Y New Orleans 0.462 0.411 0.472 0.426 0.405 LAX Los Angeles 0.363 0.368 0.418 0.394 0.383 SAT San Antonio 0.470 0.654 0.502 0.464 0.375 PHL Philadelphia 0.353 0.351 0.314 0.314 0.371 BWI Baltimore-Washington 0.604 0.494 0.450 0.447 0.369 S F O San Francisco 0.270 0.253 0.284 0.350 0.369 ONT Ontario 0.279 0.375 0.325 0.311 0.338 IAH Houston 0.569 0.532 0.558 0.487 0.302 MIA Miami 0.251 0.286 0.301 0.309 0.250 Mean 0.525 0.522 0.571 0.580 0.575 US Port-Run Airports DEA - Without Capital Inputs (CRS) 2001 2002 2003 2004 2005 OAK Oakland 0.634 0.685 0.824 0.930 0.733 PDX Portland 0.537 0.507 0.572 0.595 0.589 S E A Seattle 0.363 0.405 0.444 0.501 0.573 BOS Boston 0.422 0.450 0.553 0.556 0.567 RSW Southwest Florida 0.365 0.396 0.525 0.513 0.523 E W R Newark 0.350 0.365 0.502 0.543 0.520 JFK New York (Kennedy) 0.311 0.312 0.403 0.445 0.429 LGA New York (LaGuardia) 0.266 0.226 0.368 0.408 0.401 Mean 0.406 0.418 0.524 0.561 0.542 Canadian Airport Authorities SFA - Without Capital Inputs 2001 2002 2003 2004 2005 YVR Vancouver 0.792 0.769 0.761 0.769 0.774 Y Y C Calgary 0.603 0.584 0.562 0.584 0.633 Y E G Edmonton 0.316 0.312 0.316 0.346 0.378 Y W G Winnipeg 0.321 0.322 0.332 0.336 0.351 YOW Ottawa 0.344 0.324 0.317 0.332 0.340 YHZ Halifax 0.247 0.240 0.256 0.271 0.282 YYZ Toronto 0.457 0.405 0.239 0.235 0.206 YYJ Victoria 0.145 0.139 0.143 0.151 0.163 YYT St. John's 0.102 0.101 0.107 0.110 0.120 YQB Quebec 0.107 0.101 0.099 0.103 0.108 Y X E Saskatoon 0.097 0.095 0.093 0.097 0.107 YQR Regina 0.083 0.077 0.080 0.080 0.090 YQM Moncton 0.072 0.071 0.079 0.081 0.089 YQT Thunder Bay 0.069 0.070 0.079 0.072 0.072 YXU London 0.057 0.061 0.065 0.064 0.069 Y S J Saint John 0.029 0.028 0.027 0.027 0.028 Y Y G Chariottetown 0.025 0.025 0.025 0.026 0.027 Mean 0.228 0.219 0.211 0.217 0.226 US Airport Authorities SFA • Without Capital Inputs 2001 2002 2003 2004 2005 M S P Minneapolis-St. Paul 0.775 0.798 0.840 0.813 0.805 TPA Tampa 0.770 0.768 0.758 0.774 0.785 MCO Orlando 0.748 0.774 0.738 0.743 0.746 C V G Cincinnati/Kentucky 0.658 0.700 0.705 0.716 0.716 IAD Washington (Dulles) 0.562 0.506 0.582 0.693 0.686 DCA Washington (Reagan) 0.625 0.549 0.603 0.646 0.654 RDU Raleigh-Durham 0.602 0.636 0.604 0.610 0.634 DTW Detroit 0.697 0.605 0.570 0.565 0.616 IND Indianapolis 0.799 0.795 0.574 0.593 0.614 BNA Nashville 0.595 0.563 0.556 0.584 0.604 DFW Dallas/Fort Worth 0.710 0.705 0.688 0.752 0.598 SAN San Diego 0.594 0.549 0.546 0.577 0.581 STL St. Louis 0.748 0.714 0.647 0.581 0.562 PIT Pittsburgh 0.697 0.667 0.609 0.602 0.551 CMH Columbus 0.562 0.550 0.514 0.513 0.519 JAX Jacksonville 0.532 0.367 0.370 0.393 0.439 RNO Reno-Tahoe 0.389 0.354 0.347 0.377 0.406 RIC Richmond 0.375 0.346 0.350 0.314 0.342 ALB Albany 0.308 0.296 0.299 0.325 0.298 Mean 0.618 0.592 0.574 0.588 0.587 SFA - Without Capital Inputs 2001 2002 2003 2004 2005 ATL Atlanta 0.908 0.900 0.905 0.908 0.908 LAS Las Vegas 0.813 0.796 0.799 0.832 0.821 CLT Charlotte-Douglas 0.790 0.796 0.780 0.779 0.790 P H X Phoenix 0.831 0.778 0.790 0.782 0.773 FLL Ft. Lauderdale 0.732 0.690 0.696 0.722 0.750 HNL Honolulu 0.842 0.709 0.689 0.728 0.736 S L C Salt Lake City 0.708 0.700 0.680 0.680 0.723 IAH Houston 0.786 0.746 0.734 0.736 0.638 PHL Philadelphia 0.641 0.624 0.543 0.564 0.633 DEN Denver 0.461 0.438 0.536 0.600 0.600 MCI Kansas City 0.671 0.578 0.577 0.590 0.594 S N A Santa Ana 0.580 0.516 0.545 0.574 0.589 C L E Cleveland 0.590 0.527 0.529 0.589 0.583 S M F Sacramento 0.544 0.512 0.539 0.533 0.581 MDW Chicago (Midway) 0.554 0.565 0.577 0.609 0.564 M K E Milwaukee 0.472 0.469 0.499 0.519 0.540 BWI Baltimore-Washington 0.748 0.651 0.545 0.549 0.503 S J C San Jose 0.614 0.541 0.511 0.501 0.489 A B Q Albuquerque 0.464 0.485 0.468 0.475 0.480 PBI Palm Beach 0.440 0.394 0.423 0.445 0.473 A U S Austin 0.496 0.465 0.433 0.452 0.465 ONT Ontario 0.416 0.443 0.394 0.394 0.406 SAT San Antonio 0.427 0.420 0.417 0.437 0.381 S F O San Francisco 0.300 0.270 0.226 0.398 0.370 ORD Chicago (O'Hare) 0.309 0.328 0.352 0.434 0.368 M S Y New Orleans 0.437 0.406 0.409 0.393 0.336 LAX Los Angeles 0.310 0.290 0.301 0.269 0.223 MIA Miami 0.245 0.326 0.212 0.217 0.200 Mean 0.576 0.549 0.540 0.561 0.554 US Port-Run Airports SFA - Without Capital Inputs 2001 2002 2003 2004 2005 S E A Seattle 0.635 0.623 0.640 0.685 0.692 P D X Portland 0.683 0.643 0.631 0.646 0.647 B O S Boston 0.653 0.649 0.617 0.588 0.572 O A K Oakland 0.613 0.586 0.578 0.610 0.571 R S W Southwest Florida 0.351 0.347 0.397 0.432 0.474 L G A New York (LaGuardia) 0.430 0.323 0.369 0.422 0.424 E W R Newark 0.278 0.151 0.190 0.234 0.216 J F K New York (Kennedy) 0.102 0.082 0.052 0.047 0.049 Mean 0.468 0.425 0.434 0.458 0.456 Canadian Airport Authorities Unit Cost Index 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 YYJ Victoria - 0.54 0.59 0.57 0.58 0.58 0.65 0.62 0.62 0.63 YVR Vancouver 0.61 0.61 0.64 0.65 0.54 0.68 0.69 0.71 0.71 0.72 YYC Calgary 0.52 0.58 0.63 0.69 0.67 0.72 0.77 0.89 0.85 0.78 YWG Winnipeg - 0.72 0.82 0.82 0.82 0.91 0.92 0.92 0.89 0.89 YXE Saskatoon - - - 0.91 0.75 0.80 0.90 0.93 0.89 0.90 YEG Edmonton 1.07 1.11 1.14 1.10 0.97 1.01 1.03 1.07 0.97 0.94 YQT Thunder Bay - - 0.98 0.99 0.90 0.95 0.89 0.78 0.87 0.94 YQR Regina - - - 0.82 1.00 1.15 1.16 1.11 0.99 0.96 YXU London - - - 0.90 0.80 0.95 1.01 1.01 0.96 1.00 YOW Ottawa - 0.84 0.88 0.84 0.79 0.81 0.87 0.99 1.07 1.07 YQM Moncton - - 1.39 1.21 1.04 1.06 1.17 1.19 1.16 1.09 YHZ Halifax - - - - 0.92 1.24 1.24 1.19 1.11 1.10 YYT St. John's - 0.97 0.86 0.90 1.11 1.20 1.34 1.22 1.27 1.25 YQB Quebec - - - - - 1.29 1.31 1.45 1.52 1.34 YYG Charlottetown - 1.92 1.53 1.31 1.88 1.74 1.83 1.82 1.76 1.71 YSJ Saint John - - - 1.23 1.61 1.79 2.09 2.20 1.94 1.89 YYZ Toronto 0.65 0.73 0.92 1.01 1.13 1.37 1.53 2.03 1.91 2.00 Mean 0.71 0.89 0.94 0.93 0.97 1.07 1.14 1.18 1.15 1.13 US Airport Authorities Unit Cost Index 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 RDU Raleigh-Durham - 0.37 0.40 0.35 0.33 0.35 0.36 0.46 0.54 0.52 TPA Tampa 0.68 0.65 0.62 1.03 0.58 0.60 0.60 0.66 0.64 0.64 RIC Richmond 0.73 0.63 0.66 0.67 0.65 0.63 0.64 0.62 0.69 0.65 BNA Nashville 0.70 0.72 0.66 0.69 0.68 0.69 0.73 0.75 0.68 0.67 MSP Minneapolis-St. Paul 0.61 0.59 0.66 0.64 0.66 0.75 0.67 0.51 0.65 0.69 CVG Cincinnati/Kentucky 0.61 0.60 0.61 0.60 0.62 0.74 0.70 0.71 0.71 0.73 IND Indianapolis 0.82 1.00 1.03 1.07 0.97 0.39 0.41 0.80 0.84 0.80 CMH Columbus 0.67 0.71 0.70 0.76 0.75 0.66 0.73 0.81 0.82 0.82 RNO Reno-Tahoe 0.70 0.72 0.72 0.76 0.73 0.72 0.82 0.92 0.87 0.83 MCO Orlando 0.77 0.73 0.76 0.79 0.79 0.84 0.78 0.87 0.86 0.84 JAX Jacksonville 0.82 0.75 0.74 0.77 0.81 0.69 0.87 0.94 0.98 0.97 IAD Washington (Dulles) 1.16 1.10 1.03 1.12 1.19 1.32 1.49 1.27 0.98 0.98 DFW Dallas-Fort Worth 0.86 0.82 0.59 0.69 0.76 0.88 0.90 0.94 0.77 0.99 DCA Washington (Reagan) 1.44 1.39 1.52 1.44 1.02 1.02 1.24 1.17 1.06 1.06 DTW Detroit 0.92 0.86 0.88 0.98 0.94 0.97 1.19 1.25 1.25 1.13 PIT Pittsburgh 1.42 1.32 1.38 1.40 1.38 0.86 0.91 1.02 1.05 1.16 STL St. Louis 0.65 0.69 0.71 0.72 0.66 0.74 0.84 1.03 1.13 1.17 SAN San Diego 1.05 1.24 0.78 0.88 0.89 0.96 1.20 1.23 1.15 1.18 ALB Albany 1.02 0.99 1.01 1.09 1.03 1.05 1.11 1.17 1.10 1.35 Mean 0.87 0.84 0.81 0.86 0.81 0.78 0.85 0.90 0.88 0.90 Unit Cost Index 1996 1997 1998 1999 2000 2001 ATL Atlanta 0.33 0.28 0.26 0.28 0.27 0.32 CLT Charlotte-Douglas - 0.37 0.36 0.38 0.37 0.38 LAS Las Vegas 0.58 0.59 0.63 0.67 0.63 0.67 SLC Salt Lake City 0.56 0.66 6.64 0.70 0.69 0.70 ABQ Albuquerque 1.25 1.35 0.61 0.65 0.66 0.72 PHX Phoenix 0.59 0.61 0.60 0.62 0.60 0.61 MKE Milwaukee 0.77 0.93 0.95 0.89 0.87 0.76 FLL Fort Lauderdale 0.94 0.91 0.74 0.72 0.67 0.70 HNL Honolulu 0.44 0.48 0.52 0.52 0.51 0.53 MCI Kansas City 0.80 0.82 0.79 0.60 0.65 0.71 SNA Santa Ana 0.77 0.73 0.76 0.77 0.68 0.70 PBI Palm Beach 1.03 0.96 0.88 0.93 0.82 0.87 AUS Austin 0.63 0.70 0.88 0.89 0.82 0.97 CLE Cleveland 1.04 1.23 0.95 1.10 0.78 0.99 SMF Sacramento 0.91 0.84 0.88 0.93 0.89 0.92 DEN Denver 1.10 1.21 1.12 1.21 1.26 1.33 IAH Houston 0.93 0.89 0.87 0.91 0.65 0.73 PHL Philadelphia 1.17 1.06 1.03 1.17 1.06 1.11 ORD Chicago (O'Hare) 1.52 1.46 1.10 1.20 1.19 1.29 MDW Chicago (Midway) 1.16 1.25 1.04 1.03 0.97 1.17 LAX Los Angeles 0.86 0.89 0.77 0.90 1.01 1.20 SAT San Antonio 0.97 0.97 0.98 1.03 0.94 0.82 SJC San Jose 0.99 0.93 0.91 1.07 1.19 1.14 BWI Baltimore-Washington 0.84 0.78 0.79 0.72 0.65 0.76 SFO San Francisco 1.01 1.02 1.14 1.27 1.22 1.60 MSY New Orleans 1.37 1.42 1.33 1.42 1.45 1.11 ONT Ontario 1.08 1.13 1.01 1.24 1.25 1.46 MIA Miami 1.23 1.26 1.32 1.27 1.34 1.73 Mean 0.92 0.92 0.85 0.90 0.86 0.93 Mean (excluding ATL) 0.94 0.94 0.87 0.92 0.88 0.95 2002 2003 2004 2005 0.35 0.39 0.72 0.74 0.71 0.77 0.83 0.84 0.89 0.96 0.94 1.06 1.00 1.17 1.13 1.36 0.84 1.15 1.27 1.19 1.32 0.77 1.31 1.04 1.86 1.24 1.20 1.58 0.33 0.43 0.71 0.78 0.74 0.74 0.79 0.85 0.98 0.90 0.93 0.99 1.12 1.10 1.10 1.19 0.88 1.38 1.22 1.17 1.32 0.82 1.39 1.34 2.08 1.22 1.52 1.99 0.32 0.43 0.61 0.80 0.77 0.76 0.78 0.82 0.82 0.86 0.89 0.96 1.00 0.96 1.15 1.05 0.88 1.32 1.07 1.07 1.34 0.83 1.43 1.32 1.51 1.28 1.58 1.95 0.42 0.65 0.73 0.74 0.78 0.79 0.80 0.81 0.85 0.88 0.93 1.01 1.03 1.04 1.04 1.09 1.13 1.13 1.21 1.39 1.47 1.47 1.50 1.53 1.55 1.58 2.11 1.02 1.05 1.07 1.10 1.02 1.05 1.10 1.10 US Port-Run Airports Unit Cost Index 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 RSW Southwest Florida 0.96 1.00 0.99 0.95 0.91 1.00 1.01 0.95 0.91 0.89 PDX Portland 1.09 0.72 0.83 0.76 0.80 0.85 0.91 0.96 0.93 0.95 SEA Seattle 0.71 0.72 0.91 0.95 1.02 1.11 1.15 1.11 1.00 0.98 BOS Boston 0.98 0.95 0.92 0.96 0.91 1.03 1.07 1.13 1.17 1.17 OAK Oakland 1.01 0.96 1.03 1.06 1.08 1.03 1.16 1.20 1.10 1.23 LGA New York (LaGuardia) 1.62 1.40 1.54 1.83 1.63 1.71 2.14 1.85 1.65 1.61 EWR Newark 1.62 1.42 1.50 1.78 1.81 1.76 2.30 2.06 1.82 1.83 JFK New York (Kennedy) 2.79 2.47 2.55 2.56 2.79 2.72 3.05 3.39 3.12 2.92 Mean 1.35 1.21 1.28 1.36 1.37 1.40 1.60 1.58 1.46 1.45 Canadian Airport Authorities Operating Expense per Passenger (1996 $US) 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 YYC Calgary 2.38 2.52 2.75 2.90 2.79 3.56 3.84 4.24 3.52 3.31 YYJ Victoria - 2.35 3.18 3.01 3.01 3.10 3.54 3.19 3.21 3.38 YVR Vancouver 3.25 3.32 3.86 4.00 3.34 4.51 3.81 3.88 3.75 3.59 YEG Edmonton 4.70 4.57 4.73 4.69 4.06 4.53 4.80 4.84 4.33 4.18 YQR Regina - - - 3.32 5.15 6.18 5.65 5.25 4.43 4.36 YXE Saskatoon - - - 4.59 3.78 4.53 5.20 5.17 4.55 4.53 YHZ Halifax - - - - 4.11 5.36 5.17 5.12 4.64 4.85 YQT Thunder Bay - - 4.71 4.75 4.17 4.39 4.51 4.28 4.36 4.88 YWG Winnipeg - 3.25 4.47 4.61 4.58 5.62 6.04 5.84 5.24 4.96 YOW Ottawa - 3.82 4.62 4.40 3.99 4.14 4.39 4.92 5.23 5.20 YYT St. John's - 4.48 4.35 5.20 5.48 6.70 8.38 6.70 6.70 5.58 YQM Moncton - - 12.75 11.39 8.68 8.63 11.50 9.07 8.54 7.60 YXU London - - - 7.43 6.01 8.48 8.32 8.14 7.76 7.65 YYZ Toronto 2.56 3.04 4.17 4.45 4.93 6.38 7.01 9.37 7.93 8.04 YQB Quebec - - - - - 9.37 9.36 10.17 9.52 8.20 YYG Charlottetown - 9.12 9.58 8.89 10.91 11.28 12.10 10.88 10.56 9.41 YSJ Sain John - - - 6.85 10.59 12.72 14.37 14.00 11.96 11.36 Mean 3.22 4.05 5.38 5.37 5.35 6.44 6.94 6.77 6.25 5.95 US Airport Authorities Operating Expense per Passenger (1996 $US) 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 CVG Cincinnati/Kentucky 1.40 1.35 1.39 1.41 1.46 1.95 1.73 1.77 1.79 1.84 MSP Minneapolis-St. Paul 1.54 1.49 1.73 1.64 1.67 2.01 1.87 1.46 1.88 2.00 RDU Raleigh-Durham - 1.21 1.33 1.09 0.95 1.14 1.48 1.89 2.15 2.05 TPA Tampa 2.71 2.59 2.58 4.08 2.31 2.48 2.50 2.78 2.57 2.66 BNA Nashville 2.75 2.94 2.64 2.71 2.55 2.73 3.03 3.09 2.83 2.75 DFW Dallas-Fort Worth 2.14 2.08 1.63 1.78 1.87 2.34 2.48 2.42 2.28 2.79 MCO Orlando 2.89 2.79 3.12 3.19 3.31 3.75 3.32 3.47 3.25 3.18 DTW Detroit 2.39 2.26 2.14 2.48 2.41 2.72 3.23 3.68 3.54 3.33 RNO Reno-Tahoe 2.28 2.21 2.47 2.78 2.72 2.86 3.26 3.59 3.37 3.46 SAN San Diego 2.74 3.29 2.10 2.42 2.35 2.68 3.51 3.64 3.39 3.51 STL St. Louis 1.48 1.62 1.66 1.67 1.53 1.93 2.20 3.03 4.01 3.61 RIC Richmond 4.19 4.02 4.47 4.62 4.52 4.85 4.59 4.36 4.15 3.75 CMH Columbus 2.51 2.69 2.81 3.02 3.42 3.16 3.44 3.85 3.90 3.77 IND Indianapolis 6.87 6.92 8.60 8.85 7.90 3.30 3.67 3.88 4.06 3.84 DCA Washington (Reagan) 4.17 4.69 5.44 6.05 3.44 4.03 4.47 4.91 4.29 4.13 IAD Washington (Dulles) 4.18 4.10 4.06 4.34 4.36 5.09 5.64 5.66 4.36 4.22 JAX Jacksonville 3.35 3.03 3.07 3.02 3.28 4.03 3.37 3.79 4.27 4.25 PIT Pittsburgh 4.06 3.81 4.07 4.37 4.15 2.63 2.84 3.46 3.78 4.69 ALB Albany 4.99 4.80 4.75 5.50 5.19 5.46 5.71 6.40 6.05 7.06 Mean 3.15 3.05 3.16 3.42 3.13 3.11 3.28 3.53 3.47 3.52 Operating Expense per Passenger (1996 $US) 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 ATL Atlanta 0.76 0.69 0.68 0.67 0.63 0.76 0.78 0.73 0.73 CLT Charlotte-Douglas - 0.79 0.82 0.94 0.90 0.95 0.99 1.08 0.99 0.89 SLC Salt Lake City 1.41 1.66 1.62 2.00 1.92 2.01 2.13 2.18 2.27 1.95 LAS Las Vegas 1.59 1.76 1.94 2.04 1.87 2.15 2.30 2.28 1.92 2.02 PHX Phoenix 1.72 1.85 1.89 2.01 1.94 2.01 2.36 2.19 2.27 2.31 FLL Fort Lauderdale 3.15 3.09 2.65 2.48 2.28 2.56 2.87 2.84 2.63 2.86 IAH Houston 2.86 2.67 2.65 2.83 1.78 2.01 2.27 2.41 2.37 2.90 HNL Honolulu 2.08 2.32 2.61 2.62 2.45 2.71 3.14 4.02 3.03 2.91 ABQ Albuquerque 4.47 4.91 2.30 2.39 2.46 2.88 3.05 3.11 3.17 3.01 ORD Chicago (O'Hare) 4.08 3.95 2.99 3.34 3.34 3.60 3.54 3.32 2.86 3.06 PHL Philadelphia 3.38 3.03 2.87 3.43 3.03 3.41 3.27 3.72 3.40 3.13 MDW Chicago (Midway) 3.55 4.12 3.02 2.94 2.70 3.24 3.31 3.11 2.81 3.29 MCI Kansas City 2.97 2.97 3.06 2.37 2.43 2.82 3.89 3.64 3.43 3.46 MKE Milwaukee 3.10 4.03 4.32 4.17 3.94 3.52 3.88 3.62 3.50 3.47 DEN Denver 3.56 3.97 3.78 4.12 4.34 4.62 4.85 4.11 3.56 3.55 SNA Santa Ana 3.01 2.87 3.01 3.07 2.82 3.25 3.81 3.81 3.64 3.58 AUS Austin 2.07 2.28 3.11 3.34 3.29 4.08 4.13 4.33 3.68 3.65 PBI Palm Beach 3.86 3.60 3.55 3.67 3.36 3.59 4.26 3.97 3.73 3.65 CLE Cleveland 3.24 3.67 3.00 3.55 2.57 3.57 4.18 3.81 3.54 3.89 BWI Baltimore-Washington 2.80 2.69 2.87 2.68 2.45 2.67 3.51 3.76 3.36 4.06 MSY New Orleans 4.07 4.17 4.03 4.17 4.10 3.06 3.33 3.25 3.13 4.07 SMF Sacramento 3.60 3.48 3.63 3.98 3.78 3.86 4.96 4.89 4.94 4.48 LAX Los Angeles 2.79 2.88 2.92 2.99 3.18 4.10 4.42 4.44 4.25 4.57 SAT San Antonio 2.99 3.00 3.05 3.20 3.03 2.70 2.55 2.79 2.83 4.84 SFO San Francisco 3.20 3.29 3.85 4.25 4.16 6.33 6.75 7.63 5.39 5.62 SJC San Jose 3.90 3.75 4.00 4.55 4.86 5.02 5.83 6.03 5.92 5.97 ONT Ontario 3.95 4.03 3.77 5.09 5.24 6.27 5.20 6.17 6.27 6.47 MIA Miami 7.68 7.67 7.85 7.24 7.61 7.45 6.73 8.01 7.86 7.65 Mean 3.18 3.19 3.07 3.22 3.09 3.40 3.65 3.76 3.48 3.75 US Port-Run Airports Operating Expense per Passenger (1996 $US) 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 RSW Southwest Florida 3.53 3.61 3.95 3.83 3.55 3.52 3.55 3.50 3.25 3.25 SEA Seattle 2.83 2.94 3.16 3.15 3.47 3.95 3.90 3.63 3.35 3.53 PDX Portland 4.09 2.69 3.05 2.99 3.36 3.74 3.70 3.89 3.70 3.68 OAK Oakland 3.81 3.95 4.23 4.34 4.12 4.47 4.67 4.58 4.24 4.45 BOS Boston 3.95 3.88 3.86 4.06 3.81 4.99 4.97 5.31 4.95 5.14 LGA New York (LaGuardia) 5.93 5.74 5.32 5.66 4.92 5.38 6.58 6.53 5.71 5.58 EWR Newark 5.61 5.24 5.58 6.82 6.55 6.81 9.17 8.47 7.38 7.46 JFK New York (Kennedy) 8.43 8.40 9.46 9.56 9.63 10.25 10.55 11.40 10.27 9.60 Mean 4.77 4.55 4.83 5.05 4.93 5.39 5.89 5.91 5.35 5.33 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0092972/manifest

Comment

Related Items