The Air Transportation Industry: Economic Conflict and Competition 0323915221, 9780323915229

The aviation sector consists of various actors such as airlines, ground handling companies, and others all with conflict

577 134 5MB

English Pages 474 [476] Year 2021

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

The Air Transportation Industry: Economic Conflict and Competition
 0323915221, 9780323915229

Table of contents :
Front Cover
The Air Transportation Industry
Contemporary Issues in Air Transport
The Air Transportation Industry: Economic Conflict and Competition
Copyright
Contents
List of contributors
Preface
1 - Economic structure of the air transport business
1. Introduction
2. The market structure: highly competitive and heterogeneous
2.1 The market structure
2.2 Size order of market parties: turnover, money, and power
2.3 The interdependence and market power
3. Evolution toward new business models
4. Possible conflict situations within and between actors
4.1 Airlines
4.2 Aircraft manufacturers
4.3 Airport operators
4.4 Handling companies
4.5 Relations with the government
5. Conclusions
References
2 - The burden of a ton CO2! Emission trading systems and the air transport business
1. Introduction
2. The global challenge of climate change and CO2 emissions
3. Overview of global policies to address climate change
4. Air transport industry traffic and CO2 emissions
4.1 Historical development
4.2 Projections
5. Policies to address CO2 emissions from international air transport
6. Measures available to reduce air transport CO2 emissions
6.1 Technological developments
6.2 New standards
6.3 Airline operations
6.4 Air navigation service provider' operations
6.5 Sustainable aviation fuels
6.6 Market-based measures
6.7 Comparison of measures
7. Carbon Offsetting and Reduction Scheme for International Civil Aviation (CORSIA)
8. Analysis of the supply and demand for carbon offsets for CORSIA
9. The impact of CORSIA on air traffic and airline financial results
9.1 Assumptions and brief model description
9.2 Results
9.3 Analysis
10. Conclusion
References
Further reading
3 - Labor in the aviation industry: wages, disputes, and shocks
1. Introduction
2. Employment in the aviation industry
3. Wage determination
4. Monopsony (the power of the employers)
5. Monopoly (the power of the unions)
6. Bargaining power
6.1 Elasticity of demand
6.2 Elasticity of supply
6.3 Competition in the aviation market itself
7. Industrial action
8. Economic shocks
9. Conclusions
References
4 - The air transportation vertical channel, the global value added, and the role played by private versus public c ...
1. Introduction
2. The air transportation vertical channel
3. The aircraft manufacturers
4. The engine manufacturers
5. The leasing companies
6. The handlers
7. The distribution: GDS and others
8. Airports
9. Airlines
10. Conclusions
References
5 - Exogenous shocks on the air transport business: the effects of a global emergency
1. Introduction
2. Impact of COVID-19 on the airline business—the worst crisis since ever
2.1 Air transport demand hugely impacted
2.2 Deep impact on operating model: resources' productivity and aircraft capacity
2.3 Immediate reactions from the vast majority of the operators: fleet and capacity reductions
3. A new era of nationalization?
3.1 The financial support to traditional flag carriers
3.2 A new model of nationalized airlines
3.3 How the new market environment impacted—reactions from low-cost carriers
3.3.1 Southwest Airlines
3.3.2 Ryanair
3.3.3 EasyJet
3.3.4 Wizz Air
4. How the industry is going to face the crisis
4.1 How major airlines have been considering to face the mid-term issues
4.2 The reactions of low-cost carriers
5. Conclusions
References
Further reading
6 - The impact of regulation on the airport industry: the Italian case
1. Introduction
2. Airport regulation: a literature review
3. Airport regulation in Italy
4. The empirical analysis
4.1 The average cost function
4.2 The identification approach
5. Data
6. Empirical results
6.1 Baseline
6.2 Robustness checks
7. Conclusions
Appendix
References
7 - Airline pricing, incumbents, and new entrants
1. Introduction
2. Pricing principles: theory and literature
2.1 A fault line in aviation history
2.2 Theoretical foundations of pricing in aviation
2.3 Recent trends in airline pricing
3. Deviant behavior: pricing as a barrier to entry
4. A possible generalization and alternative strategies
5. Conclusions
References
Further reading
8 - The fight for airport slots: the case of Amsterdam Airport Schiphol11The authors have written this chapter on p ...
1. Introduction
2. The EU Slot Regulation
3. The changing context for slot allocation: COVID-19 and the airport capacity crunch
3.1 COVID-19
3.2 Demand growth and the airport capacity crunch
4. The implications for growing excess demand for slots: theory and research findings
5. The implications for growing excess demand for slots: the case of Amsterdam Airport Schiphol
5.1 Stabilizing connectivity
5.2 Seat capacity and load factor
5.3 Transfer traffic
5.4 Full freighter traffic
5.5 The role of slots in retaliatory actions in the aeropolitical arena
5.6 Use of remedies at Amsterdam Airport Schiphol
5.7 Deficiencies of grandfather rights: slot hoarding and babysitting
5.8 Gaming of the new entrant rule
6. Conclusions
6.1 Fixing the current system
6.2 The concept of the super-congested airport
6.3 Market-based measures
6.4 Concluding remarks
References
9 - Different approaches to airport slots. Same results, same winners?
1. Introduction
2. Airport slot allocation approaches in the world and the problems emerging
3. Discussion of the solution alternatives to the problems emerged from allocation approaches
4. Proposal of a new and untraditional auction mechanism for airport slot allocation
4.1 ASAM model architecture
The Institution
The Environment
The Agents' Behavior
The Slot Coordinator
The Airlines
4.2 Adaption of ASAM auction mechanism to internet keyword search auctions
4.3 The conceptual structure of the model and the auction mechanism
5. Analysis of airline agents' bidding behavior in ASAM
5.1 Components of bid price determination
5.2 Determination of bidding strategies and introducing learning mechanisms to airline agents in ASAM
6. Case study: application of ASAM to a synthetic auction market of Heathrow Airport
6.1 Specifications of the auction market
6.2 Experiments and results
7. Discussion of the proposed model ASAM and results
Network effects
8. Conclusions
Acknowledgments
References
10 - Black swans or gray rhinos on the runway? The role of uncertainty in airport strategic planning
1. Introduction and research questions
2. Increasing year-to-year traffic volatility at airports
2.1 Increased traffic volatility at airports in more competitive air transport markets
2.2 High-impact shock events from outside the air transport system
2.3 Shock events and economic characteristics of airports
3. High-impact shock events and deep uncertainty
3.1 Black Swans or fat tails
3.2 Color blindness regarding swans
4. Absorbing rare, high-impact shock events in airport strategic planning
4.1 Scenarios in aviation
4.2 Scenarios to be integrated in a responsive strategic plan
4.3 Robust airport strategic planning and flexible master planning
4.4 Increasing master plan flexibility
5. Final observations and conclusions
References
11 - Making sense of airport security in small and medium-sized airports
1. Introduction
2. A brief history of air transportation security
3. Regulatory framework
3.1 ICAO
3.2 European Framework
4. Air transport security costs
5. Proportionality of security in airports
5.1 Criticisms to aviation security framework
5.2 Risk based security
5.3 SeMS—security management system
6. A new approach for security in a network of airports
7. Conclusion
Acknowledgments
References
12 - How can airports influence airline behavior to reduce carbon footprints?
1. Introduction
2. Evolution in air transport traffic and impacts worldwide
3. Airports environmental practice and carbon reduction initiatives
4. Challenges in environmental sustainability practice at airports and ways forward
5. Negotiation
6. Good practice recommendations and opportunities
7. Conclusions
References
13 - The measurement of accessibility and connectivity in air transport networks
1. Introduction
2. An overview of air transport accessibility
2.1 The potential indicator
2.2 The daily accessibility indicator
2.3 The location indicator
2.4 The relative network efficiency indicator
3. Air transport accessibility and related concepts
3.1 Accessibility and connectivity
3.2 Accessibility, resilience, criticality, and vulnerability
4. A tentative future research agenda
4.1 Short distances and greener modes
4.2 Intermodality
4.3 ICT
4.4 Equity
4.5 Big data and open sources
5. Conclusions
References
14 - Fighting for market power: the case of Norwegian Airlines
1. Introduction
2. Why did EU deregulation initially not affect the Norwegian domestic airline market?
3. Phases in airline strategic behavior following the deregulation
3.1 The initial period 1994–1998
3.2 Second phase 1998–2002
3.3 Third phase 2002 onward
4. Airport competition
5. The low-cost carrier Norwegian's continued growth strategy
6. Concluding remarks
References
15 - Is privatization of ATC an economic game-changer? Who gains and who loses?
1. Introduction
2. Definitions
3. Literature review
3.1 Commercialization and competition
3.2 Effects of commercialization
3.2.1 Effects on safety and security
3.2.2 Effects on costs
3.2.3 Effects on ANS prices
3.2.4 Effects on service quality
3.2.5 Effects on customer relationships
3.2.6 Effects on government relationship
3.2.7 Effects on labor and capital
4. The emergence of the ANSP business model and its impact on ATM/CNS profits
4.1 Methodology
4.2 Variables influencing the ANSP business model
4.2.1 Strategic choices
4.2.1.1 Operational scope
4.2.1.2 Collaboration strategy
4.2.1.3 Innovation strategy
4.2.2 Asset choices
4.2.3 Governance choices
4.2.4 Strategy outcomes
4.2.4.1 Cost structure
4.2.4.2 Revenue structure
4.3 Business model constructs
4.4 The impact on ATM/CNS profits
5. Conclusions
References
16 - The forwarders' power play effect on competition in the air cargo industry
1. Introduction
2. Freight forwarders in a literature review
3. The business model of the air freight forwarder
4. Concentration in the air freight forwarding industry
5. Freight forwarders at major European cargo airports
6. Conclusions
References
17 - Fuel hedging: how many games can we play?
1. Fuel costs' relevance in aviation
2. Fuel hedging fundamentals
2.1 Instruments
2.1.1 Futures
2.1.2 Derivatives
2.1.3 Long versus short hedging
3. Hedging in reality
3.1 Introduction
3.2 Fuel hedging drivers
3.2.1 Company valuation
3.2.2 Mitigate volatility
3.2.3 Management competence
3.3 Challenges
3.3.1 Contractual costs
3.3.2 Accounting standards
3.3.3 Predictive modeling
3.4 Regional perspective
3.5 Cases
3.5.1 Southwest Airlines
3.5.2 Cathay Pacific Airways
4. Recent developments
5. Conclusions and future outlook
References
18 - The effect of accidents on aircraft manufacturers' competition
1. Introduction
2. Aircraft accidents, aircraft safety, and airline stock prices: a literature review
3. The aircraft manufacturers market: the story of a continuous consolidation
4. Aircraft accidents: a historic overview of air travel from safe to safest way of travel
5. The impact of accidents on aircraft manufacturers' competition
6. Why are airlines so faithful to their chosen aircraft manufacturer?
7. Conclusion
References
19 - How strategy can influence the market: recommendations and conclusions
1. Introduction
2. Market structures
3. Current market structure
4. What will the future bring?
5. Conclusions
References
Further reading
Index
A
B
C
D
E
F
G
H
I
L
M
N
O
P
Q
R
S
T
U
W
Back Cover

Citation preview

THE AIR TRANSPORTATION INDUSTRY

Contemporary Issues in Air Transport

Series Editors

STEPHEN ISON LUCY BUDD

Contemporary Issues in Air Transport

THE AIR TRANSPORTATION INDUSTRY Economic Conflict and Competition

Edited by

ROSÁRIO MACÁRIO EDDY VAN DE VOORDE

Elsevier Radarweg 29, PO Box 211, 1000 AE Amsterdam, Netherlands The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom 50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States Copyright © 2022 Elsevier Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein). Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. Library of Congress Cataloging-in-Publication Data A catalog record for this book is available from the Library of Congress British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library ISBN: 978-0-323-91522-9 For information on all Elsevier publications visit our website at https://www.elsevier.com/books-and-journals

Publisher: Joseph Hayton Acquisitions Editor: Brian Romer Editorial Project Manager: Michelle Fisher Production Project Manager: Maria Bernard Cover Designer: Matthew Limbert Typeset by TNQ Technologies

Contents List of contributors Preface

1. Economic structure of the air transport business

xiii xv

1

Rosário Macário and Eddy Van de Voorde 1. Introduction 2. The market structure: highly competitive and heterogeneous 3. Evolution toward new business models 4. Possible conflict situations within and between actors 5. Conclusions References

1 2 15 18 23 24

2. The burden of a ton CO2! Emission trading systems and the air transport business

27

Chaouki Mustapha 1. 2. 3. 4. 5. 6. 7.

Introduction The global challenge of climate change and CO2 emissions Overview of global policies to address climate change Air transport industry traffic and CO2 emissions Policies to address CO2 emissions from international air transport Measures available to reduce air transport CO2 emissions Carbon Offsetting and Reduction Scheme for International Civil Aviation (CORSIA) 8. Analysis of the supply and demand for carbon offsets for CORSIA 9. The impact of CORSIA on air traffic and airline financial results 10. Conclusion References Further reading

3. Labor in the aviation industry: wages, disputes, and shocks

27 28 29 30 34 35 39 42 45 52 52 53

55

Heather McLaughlin and Colm Fearon 1. Introduction 2. Employment in the aviation industry 3. Wage determination

55 56 58

v

vi

Contents

4. Monopsony (the power of the employers) 5. Monopoly (the power of the unions) 6. Bargaining power 7. Industrial action 8. Economic shocks 9. Conclusions References

4. The air transportation vertical channel, the global value added, and the role played by private versus public control

59 61 63 68 71 73 74

77

Gianmaria Martini 1. Introduction 2. The air transportation vertical channel 3. The aircraft manufacturers 4. The engine manufacturers 5. The leasing companies 6. The handlers 7. The distribution: GDS and others 8. Airports 9. Airlines 10. Conclusions References

5. Exogenous shocks on the air transport business: the effects of a global emergency

77 79 82 84 84 86 88 91 93 94 95

99

Cristiana Piccioni, Andrea Stolfa and Antonio Musso 1. Introduction 2. Impact of COVID-19 on the airline businessdthe worst crisis since ever 3. A new era of nationalization? 4. How the industry is going to face the crisis 5. Conclusions References Further reading

6. The impact of regulation on the airport industry: the Italian case

99 100 112 117 121 123 123

125

Carlo Cambini and Raffaele Congiu 1. 2. 3. 4.

Introduction Airport regulation: a literature review Airport regulation in Italy The empirical analysis

125 126 129 133

Contents

5. Data 6. Empirical results 7. Conclusions Appendix References

7. Airline pricing, incumbents, and new entrants

vii 137 138 145 146 148

151

Rosário Macário and Eddy Van de Voorde 1. Introduction 2. Pricing principles: theory and literature 3. Deviant behavior: pricing as a barrier to entry 4. A possible generalization and alternative strategies 5. Conclusions References Further reading

8. The fight for airport slots: the case of Amsterdam Airport Schiphol

151 152 161 165 167 168 169

171

Lisanne van Houten and Guillaume Burghouwt 1. Introduction 2. The EU Slot Regulation 3. The changing context for slot allocation: COVID-19 and the airport capacity crunch 4. The implications for growing excess demand for slots: theory and research findings 5. The implications for growing excess demand for slots: the case of Amsterdam Airport Schiphol 6. Conclusions References

9. Different approaches to airport slots. Same results, same winners?

171 172 172 176 178 187 192

195

S. Sera Cavusoglu 1. Introduction 2. Airport slot allocation approaches in the world and the problems emerging 3. Discussion of the solution alternatives to the problems emerged from allocation approaches 4. Proposal of a new and untraditional auction mechanism for airport slot allocation

195 196 200 202

viii

Contents

5. Analysis of airline agents’ bidding behavior in ASAM 6. Case study: application of ASAM to a synthetic auction market of Heathrow Airport 7. Discussion of the proposed model ASAM and results 8. Conclusions Acknowledgments References

10. Black swans or gray rhinos on the runway? The role of uncertainty in airport strategic planning

210 215 220 222 223 223

225

Jaap de Wit 1. Introduction and research questions 2. Increasing year-to-year traffic volatility at airports 3. High-impact shock events and deep uncertainty 4. Absorbing rare, high-impact shock events in airport strategic planning 5. Final observations and conclusions References

11. Making sense of airport security in small and medium-sized airports

225 226 233 236 242 242

247

Duarte Cunha 1. Introduction 2. A brief history of air transportation security 3. Regulatory framework 4. Air transport security costs 5. Proportionality of security in airports 6. A new approach for security in a network of airports 7. Conclusion Acknowledgments References

12. How can airports influence airline behavior to reduce carbon footprints?

247 248 251 257 258 263 269 270 270

273

Vasco Reis and Laura Khammash 1. 2. 3. 4.

Introduction Evolution in air transport traffic and impacts worldwide Airports environmental practice and carbon reduction initiatives Challenges in environmental sustainability practice at airports and ways forward 5. Negotiation

273 274 278 281 286

Contents

6. Good practice recommendations and opportunities 7. Conclusions References

13. The measurement of accessibility and connectivity in air transport networks

ix 290 292 293

295

Augusto Voltes-Dorta and Juan Carlos Martín 1. Introduction 2. An overview of air transport accessibility 3. Air transport accessibility and related concepts 4. A tentative future research agenda 5. Conclusions References

14. Fighting for market power: the case of Norwegian Airlines

295 296 300 305 310 311

315

Siri P. Strandenes 1. Introduction 2. Why did EU deregulation initially not affect the Norwegian domestic airline market? 3. Phases in airline strategic behavior following the deregulation 4. Airport competition 5. The low-cost carrier Norwegian’s continued growth strategy 6. Concluding remarks References

15. Is privatization of ATC an economic game-changer? Who gains and who loses?

315 316 318 324 327 330 332

335

Sven Buyle 1. 2. 3. 4.

Introduction Definitions Literature review The emergence of the ANSP business model and its impact on ATM/CNS profits 5. Conclusions References

335 336 337 345 357 358

x

Contents

16. The forwarders’ power play effect on competition in the air cargo industry

361

Thomas Van Asch 1. Introduction 2. Freight forwarders in a literature review 3. The business model of the air freight forwarder 4. Concentration in the air freight forwarding industry 5. Freight forwarders at major European cargo airports 6. Conclusions References

361 362 364 369 374 377 379

17. Fuel hedging: how many games can we play?

383

Carlos Filipe Marques 1. Fuel costs’ relevance in aviation 2. Fuel hedging fundamentals 3. Hedging in reality 4. Recent developments 5. Conclusions and future outlook References

18. The effect of accidents on aircraft manufacturers’ competition

383 384 388 402 404 406

411

Wouter Dewulf, Silke Forbes and Yufei Li 1. Introduction 2. Aircraft accidents, aircraft safety, and airline stock prices: a literature review 3. The aircraft manufacturers market: the story of a continuous consolidation 4. Aircraft accidents: a historic overview of air travel from safe to safest way of travel 5. The impact of accidents on aircraft manufacturers’ competition 6. Why are airlines so faithful to their chosen aircraft manufacturer? 7. Conclusion References

411 413 415 418 421 426 429 430

Contents

19. How strategy can influence the market: recommendations and conclusions

xi

433

Rosário Macário and Eddy Van de Voorde 1. Introduction 2. Market structures 3. Current market structure 4. What will the future bring? 5. Conclusions References Further reading Index

433 434 437 442 445 447 447 449

This page intentionally left blank

List of contributors Guillaume Burghouwt Royal Schiphol Group, Schiphol, Netherlands Sven Buyle Department of Transport and Regional Economics, University of Antwerp, Antwerp, Belgium Carlo Cambini Department of Management, Politecnico di Torino, Turin, Italy S. Sera Cavusoglu CERIS, Instituto Superior Técnico, DECivil, Transportation Systems, Lisboa, Portugal Raffaele Congiu Department of Management, Politecnico di Torino, Turin, Italy Duarte Cunha CERIS, Instituto Superior Técnico - Universidade de Lisboa, Lisbon, Portugal Jaap de Wit Emeritus Professor, University of Amsterdam, Amsterdam, The Netherlands Wouter Dewulf Department of Transport and Regional Economics, University of Antwerp, Antwerp, Belgium Colm Fearon Business School, University of Birmingham, Birmingham, England Silke Forbes Department of Economics, Tufts University, Medford, MA, United States Laura Khammash CERIS, Instituto Superior Técnico, Lisbon, Portugal Yufei Li Department of Economics, Tufts University, Medford, MA, United States Rosário Macário CERIS, Instituto Superior Técnico, Universidade de Lisboa, Lisbon, Portugal; Department of Transport and Regional Economics, University of Antwerp, Antwerp, Belgium Carlos Filipe Marques Faculty of Business and Economics, Antwerp, Belgium Gianmaria Martini Università degli studi di Bergamo, Department of Economics, Bergamo, Italy

xiii

xiv

List of contributors

Juan Carlos Martín Institute of Tourism and Sustainable Economic Development, University of Las Palmas de Gran Canaria, Las Palmas de Gran Canaria, Spain Heather McLaughlin De Montfort University, Leicester, England Antonio Musso DICEA, Department of Civil, Building, and Environmental Engineering, “Sapienza” University of Rome, Rome, Italy Chaouki Mustapha Air Transport, ICAO, Montreal, QC, Canada Cristiana Piccioni DICEA, Department of Civil, Building, and Environmental Engineering, “Sapienza” University of Rome, Rome, Italy Vasco Reis CERIS, Instituto Superior Técnico, Lisbon, Portugal Andrea Stolfa DICEA, Department of Civil, Building, and Environmental Engineering, “Sapienza” University of Rome, Rome, Italy Siri P. Strandenes Department of Economics, Norwegian School of Economics, Bergen, Norway Thomas Van Asch Department of Transport and Regional Economics, University of Antwerp, Antwerp, Belgium Eddy Van de Voorde Department of Transport and Regional Economics, University of Antwerp, Antwerp, Belgium Lisanne van Houten Royal Schiphol Group, Schiphol, Netherlands Augusto Voltes-Dorta The University of Edinburgh Business School, Edinburgh, United Kingdom

Preface If a far-sighted capitalist had been present at Kitty Hawk, he would have done his successors a huge favor by shooting Orville down.1

The year is 2021 and the aviation industry is in trouble. The COVID-19 pandemic has created tremendous stress in the industry, with aviation traffic slashed due to border closures and lockdown orders. Governments have pumped billions of dollars into aviation to save airlines from bankruptcy, protect jobs, and help airports survive a cash-flow crunch. But the current pandemic is far from the first time the aviation industry has been saved through government action. Previous rescues, although, perhaps, not quite as widespread as the current initiatives, have occurred with regularity over the past hundred years. Recessions, pandemics, wars, and terrorist incidents have all broadsided the aviation industry leading to requests for regulatory fixes and government bailouts. Aviation liberalization, the open skies movement, and the privatization of industry players were supposed to have led to a marketbased industry, where efficient, well-run firms survived and other, less efficient firms, exited the industry. Yet, the pro-cyclical nature of aviation, booming during good times and busting during bad, along with the fixed capital expenditures for aircraft and airports, the highly cyclical price for fuel, and the seemingly reckless capacity expansions by new and existing airlines, make the industry vulnerable to cash flow crunches. Government subsidies have been a feature of the aviation industry since the beginning, with inflated airmail contracts providing funds to keep the early airlines afloat. When governments chose not to subsidize private sector operators, they, instead, invested public money to take ownership of airlines, airports, and air navigation providers. While governments have been willing to let other industries die (not much apparel manufactured in the United States anymore), the aviation industry has been deemed too vital to fail. Every country, seemingly, must have a flag carrier. Airports require new terminals, additional runways, and the latest in passenger amenities.

1

Quote attributed to billionaire investor, Warren Buffett, https://nymag.com/intelligencer/ 2020/05/warren-buffett-should-have-listened-to-himself-on-airlines.html, accessed April 25, 2021.

xv

xvi

Preface

If the private sector is unwilling to invest capital to fund aviation, governments step in to provide funds. Aviation investments are deemed so important that they take priority over expenditures for education and poverty reduction. Why are aviation markets so fragile? Why have governments been so willing to intervene in these markets? The air transportation industry: Economic conflict and competition attempts to help us understand the functioning of markets in the aviation industry. How do airline pricing strategies impact competition? What is the impact of government regulatory policies on airlines? How do aviation labor markets function? How should scarce slots at airports be efficiently allocated? How should risks be considered in implementing airport security? What is the value of connecting cities to aviation networks? Is there value to privatizing air traffic control? Does fuel hedging pay off for airlines? How do aviation accidents impact aircraft sales? These are all questions addressed in this interesting look at aviation-related markets. Take the recent example of two fatal crashes of Boeing 737 Max aircraft. A casual observer might have thought that these tragic accidents would have dealt a death blow to Boeing’s 737 Max. But this does not seem to have been the case. Although there were some order cancellations for the aircraft, other airlines doubled down on their purchases, reconfirming orders or placing new orders. Airlines and leasing companies placed these orders even after investigations showed major flaws in the processes Boeing employed leading to the regulatory approval of the aircraft. The Boeing 737 Max was not the first aircraft for which Boeing’s development processes have been shown to be less than adequate. For example, the Boeing 787 was grounded in 2013 shortly after its service was inaugurated. In their interesting analysis of “The effect of accidents on aircraft manufacturers competition,” Dewulf, Forbes, and Li (Chapter 18) show that, over the years, accidents have appeared to have had little impact on aircraft manufacturer sales. Does this finding indicate a market imperfection? Perhaps, yes. Given the duopoly structure of the industry, buyers may see little choice in continuing to buy Boeing aircraft, despite the potential problems with the purchased aircraft. Moreover, many airlines are heavily invested in Boeing products, especially low-cost carriers with standardized Boeing fleets. So, switching to another manufacturer will require significant investments in aircraft and training. Finally, is Airbus any better? What good would there be cutting ties with Boeing if an Airbus fleet is similarly vulnerable?

Preface

xvii

It is not clear how to best fix the imperfections in the market for aircraft. Fixed costs and other barriers to entering the industry are high. Perhaps, government investments may be needed. This is, perhaps, why the Chinese government has invested heavily in the development of aircraft through COMAC. But the Chinese could take heed from Canada’s failed experiment with Bombardier, sinking hundreds of millions of dollars into an unsuccessful bid to maintain a locally owned and controlled commercial aircraft manufacturer. Other chapters in the book provide equally interesting examinations of the functioning of aviation markets. How privatization might affect the operations of industry participants has been analyzed extensively over the years. Researchers have compared the efficiency of privatized airlines to the efficiency of publicly owned airlines and the operations of privatized airports to the operations of publicly managed airports. Buyle (Chapter 15) examines the impact of privatization or “commercialization” on the operations of air navigation service providers. Until the COVID-19-related decline in air traffic, the air transport network had become increasingly crowded, contributing to flight delays, excess flight costs, and additional carbon emissions as aircraft circled, waiting for permission to land. Does the privatization of air navigation service providers increase the efficiency of our aviation networks leading to lower costs for system users? Unfortunately, Buyle’s analysis does not provide a positive indication that privatization contributes to lower user costs, even if efficiency improves. He concludes his chapter with the following statement: Overall the privatization and commercialization of [air traffic control] have not been the economic game-changer that governments hoped for . The winners are the shareholders, who achieve better returns and generate enough cashflows to make the necessary investments in new technologies and infrastructure. Airspace users who had hoped for lower air navigation charges often find themselves disappointed. The total user cost did not significantly decrease, as reductions in charges (if they exist) go hand in hand with higher delay costs.

Unfortunately, the conclusions are hardly a ringing endorsement of the air navigation privatization effort. Cambini and Congiu (Chapter 7) provide us with more positive news. The authors examine the impact of a change in Italian regulatory policy on the costs of airport operations. In 2014, following a European Union directive, a newly formulated “dual-till,” price-cap regulatory policy was

xviii

Preface

instituted, but only for some of Italy’s airports. The new approach allowed airports to share in profits generated by productivity improvements, thus encouraging the airports operating under the new regulatory formula to increase efficiency. The implementation of the regulatory measure for select airports afforded the researchers the opportunity to examine the impact of the regulatory change using a different-in-difference methodological approach. Analyzing data from twenty-two Italian airports over the period, 2008e18, the authors find that the imposition of the regulatory approach did lead to lower costs at the airports. Given the goals of regulation to keep costs down and improve efficiency, Cambini and Congiu have been able to uncover a successful regulatory intervention. In addition to examining how government initiatives, such as privatization and regulation, impact markets, chapters in the book also examine how strategic behavior can impact market outcomes. The book’s editors, Macário and Van de Voorde, contribute a chapter (Chapter 7) on how airline incumbents use strategic pricing behavior as a barrier to new entry. While reading this chapter, I was reminded of an article published in the Wall Street Journal about 30 years ago describing how airlines use coded information in computer reservation systems to strongly discourage rivals from competing too strongly in important markets.2 I was also reminded of the demise of Laker Airways, a pioneer low-cost carrier, driven from the North Atlantic market by fierce competition from entrenched rivals. Macário and Van de Voorde describe a case study of an incumbent network airline using a combination of low prices and increased capacity to fend off the entry by two low-cost airlines in their major market. The incumbent carrier is Brussels Airline. Vueling, a Spanish-based carrier, was the first of the low-cost carriers to enter multiple routes from Brussels. Its larger rival, Ryanair, joined the competitive onslaught with several new services from the same airport. The decision to provide services from Brussels represented a departure from Ryanair’s traditional strategy of flying routes from secondary airports, with a direct attack on the incumbent network carrier and the smaller low-cost rival at a first-tier airport. As a result of the expansion of services at Brussels airport, a price war ensued involving the two low-cost carriers and Brussels Airline. It was only after both Vueling and Ryanair retreated from their Brussels airport services that

2

The most famous of these was the “FU” fare code.

Preface

xix

the price war ended. Brussels Airline had been able to use its pricing and capacity strategy to successfully compete with the low-cost carriers, although at a severe cost to its finances. van Houten and Burghouwt (Chapter 8) also describe how airlines use strategic behavior to gain competitive advantage. In this case, the strategic behavior is concerned with gaining access to scarce slots at congested airports. Slot rules generally allow airlines to maintain slots (“grandfather” rules) if they are being actively used (“use it or lose it” rules). With limited capacity and growing demand at busy airports, one would expect airlines to increasingly use larger aircraft at these airports. Moreover, load factors at congested airports should rise as demand increases. However, in their study of congested Schipol Airport in Amsterdam, the authors find that, in fact, both aircraft size and load factors may be falling. Although some of these changes may be due to airport policy changes (e.g., restrictions on widebodied aircraft during certain operating periods), airline operating changes may also be partially due to strategic behaviors. van Houten and Burghouwt note that current grandfather rules provide strong incentives for airlines to deter entry by rivals by hoarding slots. In addition, airlines may downsize their aircraft and spread passenger traffic over greater frequencies to maintain control over scarce slots. However, new entrants can also use strategic behavior when competing for slots. European Union rules allow new entrants special access to slots at airports. These new entrants are defined as airlines that hold fewer than 5% of the slots at a particular airport or 4% of the slots at the airport system level. Airlines can evade these restrictions by flying under multiple operating authorities. Using this loophole, airlines can appear as new entrants even if they have already established operations at an airport, thus gaining access to airport slots. Perhaps, the European Union should consider adopting a fairer and more rationale system for slot allocation, along the lines outlined by Cavusoglu (Chapter 9). Although standard Econ 101 still teaches the functioning of perfectly competitive markets along the lines espoused by Adam Smith, we know from experience that most markets are imperfect. Competition is not perfectly competitive and is subject to manipulation by strategic behavior of market participants. Governments attempt to regulate this behavior with mixed success. The air transportation industry: Economic conflict and competition very intelligently describes the workings of the many aviation-related markets. The chapter authors assess the efficiency of the markets and offer proscriptions for ways to improve efficiency.

xx

Preface

The pause in the growth of air transport due to the COVID-19 pandemic has resulted in a rare opportunity to reassess the functioning of aviation markets. Hopefully, our policymakers and regulators will make good use of the analyses presented in this book. Martin Dresner R.H. Smith School of Business University of Maryland College Park, MD, USA

CHAPTER 1

Economic structure of the air transport business Rosário Macário1, 2 and Eddy Van de Voorde2 1

CERIS, Instituto Superior Técnico, Universidade de Lisboa, Lisbon, Portugal; 2Department of Transport and Regional Economics, University of Antwerp, Antwerp, Belgium

1. Introduction In recent decades, the aviation sector has proven extremely sensitive to the economic situation and the related cyclical economic movements. Periods of strong economic growth gave rise to the market entry of new carriers, expanding business and, for several consecutive years, generating significant profit. In contrast, economic crises automatically translate into aviation crises, which subsequently translate into waves of takeovers and mergers of companies, bankruptcies, and market exits. At the same time, the aviation sector appears to be a proactive indicator: developments in the aviation sector are often at the forefront of evolutions in the rest of the economy. The sector thus provides an economic laboratory in which new developments and trends become indicators of what will happen later in the global economy. Actors in the aviation sector approach their industry from the perspective of air transport chains. Potential travelers and shippers of goods do not select carriers and airports solely on their own characteristics and merits, but because they belong to specific aviation chains that optimally meet their own preferences and correspond to their willingness to pay and the attractiveness of the destinations. The ultimate success of airlines and airports is therefore largely dependent on whether they are or are not part of successful chains. Along with developments in the economy and industry, however, the market structure and market forces have evolved as well (Gillen and Niemeier, 2008). At one time, carriers at an airport controlled all operations, extending even to tourism (e.g., hotels, travel agencies). At one point, a trend developed in which all non-core activities (e.g., catering, baggage handling) were shifted toward other companies. A similar shift occurred in the field of freight transport. While freight was long considered a The Air Transportation Industry ISBN 978-0-323-91522-9 https://doi.org/10.1016/B978-0-323-91522-9.00001-4

© 2022 Elsevier Inc. All rights reserved.

1

2

The Air Transportation Industry

by-product of passenger transport, the freight market is currently the exclusive focus of a number of carriers. The aviation market is thus situated within an environment of constant evolution combined with permanent uncertainty. The dynamics of economic interaction have made the aviation sector exceptionally resilient, as was particularly evident during the period from 1970 to 2015 (ICAO, 2015). Nevertheless, many questions remain with regard to the future. How will aviation recover from the COVID-19 crisis of 2020? Will this crisis challenge the air transport industry’s capacity for resilience? Which existing aviation activities will be willing/able to expand, and which will disappear in the short to medium term? Where are new market opportunities emerging? Which exogenous and endogenous factors drive change? The answers to these questions have consequences that transcend the aviation sector as such. Any decision made by any actor has direct and indirect consequences for a variety of aspects, including employment, investments, the added value to be realized and the prevailing funding requirements. From another perspective, increasing uncertainty caused by pandemics is likely to increase the risk associated with investments, thereby affecting the expansion and management of airports. As demonstrated by these developments described, there is a need to identify any new developments in the aviation sector as quickly as possible and to be able to estimate their consequences. Only then will the various actors involved be able to take the proper action. In addition, the actors within the system compete, negotiate, and even come into conflict with each other. The relative power relations between aviation actors are discussed in this chapter, as well as throughout the book. In this chapter, the discussion of power relation is followed by further details on the evolution toward new business models and possible conflicts between aviation actors.

2. The market structure: highly competitive and heterogeneous Over the years, the aviation sector has undergone a relatively rapid evolution toward new market forms. The sector initially faced pressure to shift from highly regulated markets toward liberalized, deregulated markets (Odoni, 2016a; Gudmundsson, 2019). One consequence of this shift was that it made it relatively easy for newcomers to enter the market, while increasing competition for the traditional “flag carriers.” This has given rise to strong, rapid development within the aviation market, characterized by continuous entries and exits (see, e.g., Green, 2016).

Economic structure of the air transport business

3

Airlines have also increasingly moved away from the notion that they must control the entire aviation product. One prominent new trend has involved focusing on the “core business”dactual flight operationsdand phasing out all “non-core activities.” This trend has resulted in changes to the composition of the aviation product (i.e., movement from origin to destination) and its specific characteristics (e.g., price, comfort, frequency, timing). The field has increasingly come to involve a combination of several different market actors, from carriers and airport managers to all kinds of service providers, and the changes apply to both passenger and freight transport. The interaction between the strategies of the various actors is abundantly clear. The advance of chain thinking means acknowledging that every decision that an individual actor makes will have both direct and indirect consequences for the other actors within the chain, with implications for the competitive positions of all actors involved (Starkie, 2008). 2.1 The market structure Although the structure of a market may evolve over time, the basic economic rules remain the same. For example, the aviation industry continues to be driven mainly by the global economy. This world economy is subject to rapid changes in international trade, the international redistribution of labor and capital, and the far-reaching integration and globalization of markets (see, e.g., Kupfer et al., 2017). The core of an aviation chain initially consisted of airlines and airports (Merkert et al., 2017), and it would seem obvious to regard airlines as the largest and strategically most important airport customers. Over time, airlines have undergone a significant increase in scale, largely due to horizontal cooperation (cf. the three major alliances) and/or mergers and acquisitions (cf. Lufthansa’s control over other airlines, including Swiss International Airlines, Austrian Airlines, and Brussels Airlines). One important question concerns whether the trend toward scale increases among the carriers can also be expected among airport operators and the providers of various services, including handling, catering, cleaning, and hinterland transport (Morrison et al., 2008). The constant changes taking place within the airport sector have continued in recent decades. The field of airport management and operations has undergone an evolution from government-led management toward the increasing contribution of private capital, in some cases accompanied by interests in other transport modes (e.g., capital investments by VINCI Airports in the airports of Portugal). Traditional ground handling companies are shifting into more complex

4

The Air Transportation Industry

handling holdings, with a process of mergers and acquisitions initiated under pressure from the need for capital, combined with externally funded expansion projects. As chain thinking has increased, transport services from origin to destination have come to be considered in terms of complex logistics chains, with each link expected to contribute to a process of continuous optimization throughout the chain as a whole. In response to the detection of potential bottlenecks, carriers are likely to move toward vertical acquisitions and vertical integration. Such shifts affect the competitive balance, with specific carriers acquiring greater market power as the largest and most actively growing actors at specific airports. Estimating the future market structure requires insight into a number of possible future developments (Belobaba, 2016) Perhaps the most important questions concern future economic growth. Any economic growth that might take place will almost automatically translate into a growing demand for airport services across the chain. Before the COVID-19 pandemic, the only question in aviation circles concerned the extent to which long-term economic growth would continue and whether and to what extent potential short-term crises were likely to cause structural shifts. The pandemic will inevitably result in changes to several processes in our society. For example, teleworking will undoubtedly be much more common, potentially resulting in a significant reduction of business trips. This raises several important questions. In the future, will the demand for air transport follow a path of growth that depends on services rather than on industrial output? Will growth in passenger/freight transport continue to be related to GDP? Will the trend toward scaling-up carriers and service providers through horizontal and vertical forms of integration continue? Given the apparent end to the trend toward ever-larger devices (cf. the A380 story), which devices are likely to be used in the future? Even more important than estimating the future structure of the market is the analysis of the strategic steps of the carriers, as they are the initiators of connections and the frequency and capacity on offer (Kupfer et al., 2016). Important questions arise in this regard as well. Which time frames will those carriers follow in their search for new partnerships? How will the non-carriers within that transport chain respond to this? To what extent will the carriers become so powerful that they are able to impose their will on other actors (e.g., airport authorities and ground handling companies)? Will the aviation sector enter into new partnerships with other modes of transport (cf. political pressure to abolish short aviation connections)? Each of the crucial questions presented above is surrounded by uncertainty. This is partly due to the fact that the aviation market is a highly

Economic structure of the air transport business

5

dynamic environment. It is thus reasonable to assume that the various actors try to anticipate each other’s strategic decisions. This is competition in its purest form. Prior to the COVID-19 pandemic, the most likely future scenarios deserving of in-depth scientific investigation were largely known. Only the concrete timeline remained uncertain. Although the pandemic clearly changed this situation, it is not the only factor involved. The speed with which the various market actors within the aviation chain will implement their adapted strategies will also depend on a wide range of exogenous and endogenous variables (see e.g., Bringman et al., 2018). 2.2 Size order of market parties: turnover, money, and power To provide an idea of the relative market power of the various aviation actors, we initially limit our focus to an analysis of aggregated figures for carriers and airports. In a subsequent section, we examine the mutual ties between actors. Table 1.1 provides an overview of the order of magnitude of the scheduled passenger and cargo traffic for the world as a whole, which involved 4.5 billion passengers and 60.9 million tons in 2019. It is interesting to note that 58.6% of all passenger transport was domestic, with 41.4% international. Airlines thus introduced 4.5 billion customers into the aviation system. These figures can be used to derive the effects for airports and airport-dependent actors. Table 1.1 Worldwide scheduled passenger and cargo traffic (2019). World

International

Domestic

System-wide

Passengers carried (thousands) Freight tonnes carried (thousands) Passenger-kilometres flown (millions) Available seat-kilometres (millions) Passenger load factor Freight and mail tonne-kilometres (millions) Available freight tonne-kilometres (millions) Freight load factor Revenue tonne-kilometres (millions) Available tonne-kilometres (millions) Weight load factor

1,890,594 40,919 5,548,819

2,652,500 20,045 3,130,802

4,543,094 60,965 8,679,621

6,766,522 82.0% 220,689

3,752,652 83.4% 33,293

10,519,174 82.5% 253,982

423,906

119,910

543,717

52.1% 750,097

27.8% 318,413

46.6% 1,068,510

1,071,337

465,937

1,537,274

70.0%

68.3%

69.5%

Source: Reproduced from IATA, 2000e2020. World Air Transport Statistics (WATS), Montreal.

6

The Air Transportation Industry

The number of passengers carried is an important variable, but the actual transport performance is also a function of the distances flown (see, e.g., Balliauw et al., 2018). It is derived from the number of passengerkilometres flown (an indicator of demand) and the available seatkilometres (an indicator of supply). These two figures can be combined to yield a passenger load factor of 82.5% for system-wide transport. For freight transport, the freight load factor is equal to 46.7%, largely due to the low freight load factor for domestic transport (only 27.8%). Despite the fact that 2019 was a relatively good year for the aviation sector, this did not automatically translate into brilliant performance. As shown in Table 1.2, the total industry achieved a net profit of US$25.9 billion. The figures reveal several important geographic differences, with profits in North America, Europe, and Asia-Pacific, and with negative results in the Middle East, Latin America, and Africa. The analysis of specific negotiating positions among the various actors also calls for looking at the individual carriers. Table 1.3 provides an overview of the top 15 airlines in terms of scheduled passengers carried for 2019. Note that the first and fourth places are occupied by low-cost carriers (Southwest Airlines and Ryanair, respectively). Each of these carriers reach their markets primarily on their own continents. Furthermore, the three major American legacy carriers are ranked in the top five positions. These three carriers are core partners within their respective strategic alliances. The absence of large European carriers (e.g., Air France and British Airways) in the top 15 positions is remarkable. The number of passengers transported indicates the number of people who are brought to an airport anywhere in the world, although it does not provide a complete view of the total transport performance delivered by the carriers. Table 1.2 Regional financial performance (2019). Net profit, in US$ billions System-wide global commercial airlines

Global

2017

2018

2019

37.6

27.3

25.9

17.8 8.9 10.5 0.1 0.5 0.2

14.5 9.1 6.1 1.5 0.8 0.1

16.9 6.2 4.9 1.5 0.4 0.2

Regions:

North America Europe Asia-Pacific Middle East Latin America Africa

Source: IATA, 2000e2020. World Air Transport Statistics (WATS), Montreal.

Economic structure of the air transport business

7

Table 1.3 Top 15 airlines: scheduled passengers carried (thousands, 2019). Rank Airline International Domestic Total

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Southwest Airlines Delta Air Lines American Airlines Ryanair United Airlines China Southern Airlines China Eastern Airlines EasyJet IndiGo Air China LATAM Turkish Airlines Lufthansa Emirates All Nippon Airways

4263 27,387 28,672 146,299 28,783 16,784 15,903 85,017 7166 13,889 15,818 42,428 52,579 58,668 10,184

158,419 135,208 127,113 e 87,488 90,527 88,869 8341 67,173 59,576 57,366 30,281 12,592 e 40,565

162,681 162,595 155,785 146,299 116,271 107,312 104,772 93,358 74,338 73,465 73,184 72,709 65,171 58,668 50,749

Source: IATA, 2000e2020. World Air Transport Statistics (WATS), Montreal.

The top 15 carriers based on the number of scheduled passengerkilometres flown are presented in Table 1.4. These rankings are headed by the three major American legacy carriers, followed by Emirates. Lowcost carriers (e.g., Southwest Airlines and Ryanair) fly on their own continents, with shorter average distances per flight, relative to other carriers. The three largest European legacy carriers are represented in the top 15 positions in this ranking. The analysis of the order of magnitude of carriers, as presented above, is important in the sense that a company’s purchasing and/or bargaining power with regard to its suppliers is linked to the strength and/or size of that company. For carriers, this power can be increased by working with strategic alliances.1 In this sense, it is important to know how strong the position of possible opponents is. In the following paragraphs, we provide a more detailed view of the airport operators.

1

A large number of these airlines belong to one of the three alliances: STAR Alliance (founded in 1997), oneworld (founded in 1999) and SkyTeam (founded in 2000). Alliances, which are established for the purpose of pooling resources, aim to increase their own competitive strength relative to third parties, including by offering codeshare flights. Of equal importance, however, is the greater bargaining power that such alliances can have in other matters, including the purchase of aircraft and negotiations at airports.

8

The Air Transportation Industry

Table 1.4 Top 15 airlines: scheduled passenger-kilometres flown (millions, 2019). Rank Airline International Domestic Total

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Delta Air Lines United Airlines American Airlines Emirates China Southern Airlines Southwest Airlines China Eastern Airlines Ryanair Qatar Airways Air China Lufthansa British Airways Air France Turkish Airlines LATAM

141,353 162,962 126,533 299,496 71,788 7082 69,426 185,405 172,591 79,293 160,731 151,935 143,347 131,106 67,108

208,792 179,972 215,972 e 141,785 204,297 117,218 e e 89,737 5214 3317 7523 19,134 54,976

350,145 342,935 342,510 299,496 213,573 211,379 186,644 185,405 172,591 169,030 165,945 155,252 150,870 150,240 122,083

Source: IATA, 2000e2020. World Air Transport Statistics (WATS), Montreal.

A carrier initially negotiates with a set of potential airports to which it will fly. A possible agreement depends on the potential demand that can be served through a specific airport, as well as on a number of other conditions, including landing rights, available slots and available free capacity, as well as the way in which slots are allocated and the quality of the service providers present at that airport. The ultimate choice of a carrier for a specific airport is obviously the result of negotiations, and relative market power plays a major role in this process. At airports with little or no free capacity (e.g., London Heathrow), the airport operator’s market power is extremely high, insofar as the regulation allow it to use that market power effectively. Airports that are almost completely dependent on a single carrier as a customer have little leeway. This translates into lower landing fees, amongst other conditions. Consider the following illustrative example. In 2012, the cost per departing passenger in Brussels Zaventem, the national airport, was V31.36. At Charleroi airport, also known as Brussels South, and with Ryanair as the dominant carrier, the same costs amounted to a mere V2.39 (Flemish Airport Commission, 2012). To provide an idea of the order of magnitude of a number of airports, Table 1.5 presents an overview of the top 15, based on enplaning and deplaning passengers in 2019.

Economic structure of the air transport business

9

Table 1.5 Top 15 airports: total passengers enplaned and deplaned (2019). Enplaning and deplaning Rank City (Airport) (thousands)a

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Atlanta (United States) Beijing (China) Los Angeles (United States) Dubai (Emirates) Tokyo ( Japan) Chicago (United States) London Heathrow (Great Britain) Shanghai (China) Paris (France) Dallas/Fort Worth (United States) Guangzhou (China) Amsterdam (Netherlands) Hong Kong (China) Incheon (South Korea) Frankfurt (Germany)

110,531 100,011 88,068 86,397 85,505 84,649 80,888 76,153 76,150 75,067 73,395 71,107 71,415 71,204 70,556

a passengers in transit counted once. Source: Airport Council International (ACI), Passenger Traffic 2019.

A large volume of passengers is involved for each of these airports. Airports with an excellent geographic location and great market potential attract many carriers. In addition, a number of those airports are also used as hubs by certain carriers.2 Each passenger generates significant nonaeronautical revenue streams as well. The aggregate revenues for Gatwick Airport (London), both aeronautical and non-aeronautical, are presented in Table 1.6. For the accounting year ending March 31, 2019, total revenues amounted to more than £810 million. Taking into account total operating costs of £539.4 million, the operating profit amounted to £271.4 million. With the addition of £170.0 million in depreciation and amortization, this results in EBITDA3 of £441.4 million. The measurement of airport financial performance and the interpretation of the economic indicators derived from them should always take into

2

In the United States, United Airlines has seven hubs, with nine hubs for American Airlines, 10 hubs for Delta Air Lines, six hubs for JetBlue and five hubs for Alaska Airlines. Although Southwest Airlines does not use the “hub” concept, they do have 18 “focus airports.” 3 Earnings before interest, tax, depreciation, and amortization.

10

The Air Transportation Industry

Table 1.6 Gatwick Airport: revenue (year ending March 31, 2019). Revenue item £ million

Share

Aeronautical income Retail income Car parking income Property income Operational facilities and utilities income Other income Total revenue

52.8 23.6 10.9 3.9 4.0 4.8 100.0

427.8 191.3 88.3 31.9 32.5 39.0 810.8

Source: IVY HOLDCO LIMITED, Annual report (2019).

account the institutional contexts within which specific airports operate (Odoni, 2016b). In addition, explicit attention should be directed to ownership structures. Due to differences between these environments, some airports aim to maximize returns for investors and shareholders, while others strive merely to recoup the costs incurred in providing airport services and infrastructure. Airport operators and other aviation stakeholders use various forms of equity and debt financing. An airport’s ownership model has a direct impact on its capital structure. It is precisely the diversity of capital structures that makes the process of benchmarking the financial performance of airport managers so complex (Mahoney et al., 2012). No analysis of airport revenues and profitability can be complete without considering the role played by economic regulation. The possibilities that airports have to generate traffic and income depend in part on actual traffic volumes and market characteristics, as well as on the jurisdiction and regulations within which they operate. For this reason, caution is required when interpreting the various profitability indicators and benchmarks. Thus far, we have limited our focus to the two largest players: carriers and airport managers. Due to their size, these two actors often have significant bargaining power. Extensive concentration has been observed among carriers, including through the formation of alliances (Fageda et al., 2019). At the same time, however, two interesting developments have taken place with regard to airport operators. The first involves capital diversification, including with regard to pension funds and investment funds. One illustration is provided by the consortium that controlled the

Economic structure of the air transport business

11

shareholding of Gatwick Airport on March 31, 2019.4 In the course of 2019, the VINCI Airports group also entered the capital. The addition of VINCI is an example of an increase in scale on the part of the airport operators.5 Based on the aforementioned developments, we can conclude that, similar to the container business in shipping, aviation is evolving toward a situation in which large contracts are negotiated by a limited number of large groups. Negotiations for specific carriers and/or their alliances may thus involve approaching multiple airports. The extent to which such scaling-up will continue for the two actors and the consequences of this process in terms of negotiations remain unclear. One relevant question concerns whether this evolution is occurring at the top of the market toward negotiations within the framework of a bilateral oligopoly. The other actors in the aviation sector are of a smaller scale. They include handlers, parking facilities and other service providers. Some of these actors (e.g., fuel suppliers, caterers, and maintenance services) have direct contact and contracts with carriers. Others (e.g., shop operators) have agreements only with the airport manager. Even though some of these service providers are active at multiple airports and even on different continents, they have less turnover and, correspondingly, less bargaining power. This is despite an increase in scale among handlers (e.g., Menzies Aviation, Aviapartner Holding, and Dnata). Another striking development in recent years is that a number of handlers have experienced financial difficulties, with some even going bankrupt. The shops at airports form a separate category. This has to do with the fact that, in the field of regulations, the profits linked to non-aeronautical revenues and other aspects tend to be the subject of sharp discussions. A typical example concerns the discussion of single-till versus dual-till regimes.

4

5

Gatwick’s shareholding as of March 31, 2019, was composed as follows: Global Infrastructure Partners (41.95%); The Abu Dhabi Investment Authority (15.90%); California Public Employees’ Retirement System (12.78%); National Pension Service of Korea (12.14%); Future Fund Board of Guardians (17.23%). In 2019, VINCI Airports was present in 45 hubs of different sizes, spread across 12 countries, eight capitals and four continents, with a total of 255 million passengers handled in 2019. VINCI is also an investor in land transport (e.g., roads and bridges).

12

The Air Transportation Industry

2.3 The interdependence and market power Large numbers of actors of various sizes are active at an airport, each with a unique commercial approach and set of objectives (Templin, 2010). This in itself results in considerable heterogeneity within the aviation landscape (Meersman et al., 2011). Fig. 1.1 provides an overview of these actors in their mutual relationship, based on pricing and invoices. The airport manager plays a central role, while carriers bring in the main clientele, with passengers and cargo. These carriers do not necessarily have a great deal of relative market power, however, given the limited capacity assigned through “slot allocation” at airports with many competing carriers. In most cases, the government is responsible for regulations. Licences for handling aircraft are awarded through auction by the airport operator and/or the regulator. Some service providers are licensed according to agreements with specific airport providers and conclude contracts directly with carriers. Shops and parking operators conclude agreements with airport operators. In the ideal situation, we would be able to quantify the interrelationships between the various actors, thereby identifying who pays which amount to which actor, and at which time. Although such figures could provide valuable input for bilateral or multilateral negotiations, they are unfortunately not available at this time. It is therefore interesting to consider

FREIGHT

PAX

freight

taxes

AIRPORTS

CARRIERS

Ticket + taxes +sales on board fees

Handling fees

Government

landing fee + taxes pax concession

HANDLERS SERVICE PROVIDERS F U E L

CATE RING

MAIN TEN ANCE

PARKING



concession

RENT-A-CAR

… .

sales

SHOPS

Figure 1.1 Pricing and payment of airport bills. (Source: Own composition.)

Economic structure of the air transport business

13

Table 1.7 Brussels Airport: value added (2015, in millions, at current prices). Cluster and sector Value added Share (in %)

Air transport cluster Air transport Travel agencies and tour operators Airport operator Airport handling Building and repairing of aircraft Other air transport supporting activities Other airport-related activities Passenger transport over land Freight transport over land Cargo handling and storage Courier and post activities Security and industrial cleaning Trade Hotels, restaurants, and catering Other services Other industries Public services Direct effects Indirect effects Total

1129.8 409.2 13.3 376.5 102.5 64.5 163.8 614.1 14.0 11.3 163.8 141.2 50.7 39.3 58.7 31.8 11.7 91.7 1743.9 1672.8 3416.7

64.8 23.5 0.8 21.6 5.9 3.7 9.4 35.2 0.8 0.7 9.4 8.1 2.9 2.3 3.4 1.8 0.7 5.3 100.0 (95.9)

Source: Vennix, S., 2017. Economic Importance of Air Transport and Airport Activities in Belgium, Working Paper 324. National Bank of Belgium, Brussels, based on National Bank of Belgium, 2017. Central Balance Sheet Office, Brussels.

a study by the National Bank of Belgium with regard to the economic importance of aviation and aviation activities in Belgium.6 For the period under consideration (2013e2015), the aggregate passenger traffic kept pace with the growth at world level. Table 1.7 provides an overview of the added value realized for the period considered. The clusters defined in the analysis are interesting in themselves, having been divided into two large blocks: air transport and other airport-related activities. The study considers the economic importance of aviation and aviation activities in Brussels in terms of added value, employment and investments. In addition to activities that are directly connected with air transport, the study considers those that take place on the grounds of the airport. The direct and indirect effects of the sector are calculated based on 6

The latest available report covers the period 2013e2015.

14

The Air Transportation Industry

microeconomic data (obtained primarily from the Central Balance Sheet Office) and meso-economic data (obtained from the National Account Institute). The order of magnitude for each actor is important, as are the mutual relationships between the actors. Slightly more than half (51.0%) of the total direct effects of added value are generated by the air transport cluster, with 49.0% being generated by other airport-related activities. Within the air transport cluster, air transport is responsible for 23.5% of the direct effects, with the airport operator contributing 21.6%, airport handling 5.9% and other air transport supporting activities 9.4%. For the other airport-related activities, cargo handling and storage account for 9.4% of the direct effects, with courier and post activities accounting for 8.1% and public services accounting for 5.3%. Direct added value evolves in accordance with the traffic volumes. The added value of indirect effects contributes a nearly equal amount (in 2015, 95.9%). Table 1.8 presents a number of indicators calculated for Brussels Airport, as well as for three important actors within the airport context, based on available figures on added value, employment, and direct investment. Several interesting conclusions can be drawn from a comparison of the calculated indicators: added value per employee and direct investment per employee. Each of these indicators, as calculated for the three categories of actors considered, reinforces the image of an extremely heterogeneous aviation sector. Both are many times greater for the airport operator than they are for the two other actors. For airport handling, the figures are relatively low. For the carriers, it is important to note that an important part of the added value is realized by actual flight operations between two airports. Table 1.8 Brussels Airport: several important indicators (2015(IATA, 2000-2020)). Direct Value Direct Value Air investment added investment added (V Employment transport per FTE per FTE (V million) million) (FTEs) cluster

Air transport Airport operator Airport handling

409.2

3974

20.8

102,969

5234

376.5

771

121.3

488,327

157,328

102.5

1920

53,385

1823

3.5

Based on National Bank of Belgium, 2017. Central Balance Sheet Office, Brussels.

Economic structure of the air transport business

15

Despite these interesting conclusions, questions remain with regard to the extent to which extrapolation is possible. The hypothesis is that the broad outlines apply to most airports, although the final amounts may differ. This may be due to differences in the positioning of particular airports (e.g., whether it is or is not used as a hub by a specific carrier). Additional research on this point is desirable, if not essential.

3. Evolution toward new business models Within the aviation business, technological and organizational developments follow each other at a rapid pace. An entire battery of industrial-economic developments interact within the sector: from new market entry through mergers and acquisitions to bankruptcies. In addition, as the most important customers, airports and carriers are increasingly confronted with issues relating to ecology and capacity. This results in very strong competition, sometimes extending to the regional level. There is also an influx of new and innovative products. Air freight provides a recent example, with freighters filled with all kinds of medical products. Many questions are relevant in this regard. In which direction will this future aviation market evolve? What will this evolution depend on, and which effects will all of these developments have on future business models? In order to find answers to these questions, several additional questions must be analyzed in detail. To what extent will trends from the past continue? What are the potential effects of all possible types of crises? To what extent will long-term growth continue in the future? Which longterm effects will be caused by economic and health crises (e.g., the COVID-19 pandemic)? How will these effects translate into changes in investments in additional capacity, in terms of both airports and carriers? How is political policy evolving, including in terms of bottlenecks and final allocations? The answers to each of these questions will have crucial effects on market forces and business models. For too long, the aviation sector has been regarded as homogeneous. Nothing could be further from the truth. The aviation sector comprises a multitude of heterogeneous actors. Some remain subject to a form of state control, while others are fully privatized and yet others operate under mixed regimes. Most private actors work under an objective of profit maximization. Non-privately controlled actors usually have other objectives, including the maximization of employment and/or added value or, more generally, the maximization of socioeconomic surplus.

16

The Air Transportation Industry

A thorough analysis of the strategic behavior of a number of carriers reveals that, despite the fact that each carrier positions itself in its own specific way, the instruments that are available to them are invariably the same. Virtually every carrier takes an approach that combines unit cost control and the optimization of capacity utilization and load factor with an aim to maximize yield. Despite these similarities, the business model continues to undergo a strong process of evolution. In addition to emphasizing a sufficiently high yield, efforts are being made to generate additional income with non-flight activities. As a result, certain carriers (e.g., Ryanair) are realizing more than half of their operational results through sales that have little or nothing to do with actual flight operations (e.g., car rentals and hotel reservations). In recent years, many different forms of cooperation have arisen, both within and between the various actors. Table 1.9 provides an overview of the possible forms of cooperation, along with examples. The focus is limited to airport operators, carriers, and ground handling companies. Table 1.9 Control and cooperation between aviation actors. Actors Airport operators Carriers Handling companies

Airport operators Carriers

Handling companies

Mergers and acquisitions (e.g., VINCI Airports) Financial participation (e.g., Lufthansa in Munich) Cooperation between airport and airline (e.g., Charleroi and Ryanair)

Concessions and licences Integration

Mergers and acquisitions (e.g., Air France and KLM) Alliances Code share agreements Joint ventures Financial participation (e.g., Ryanair in AerLingus) Integration versus outsourcing Multi-airport contracts

Take-overs and increasing concentration (e.g., Dnata, Menzies)

Based on Meersman, H., Van de Voorde, E., Vanelslander, T., 2011. The future air transport sector. A modified market and ownership structure, In: Macario, R., Van de Voorde, E. (Eds.), Critical Issues in Air Transport Economics and Business. Routledge, Abingdon, pp. 10e28.

Economic structure of the air transport business

17

Airport operators have increasingly become the subject of privatization (Graham, 2010). Perceiving opportunities to make money from partial or full airport sales, governments have observed the arrival of investment funds and even pension funds. The reason is obvious: under normal circumstances, airports generate relatively stable results and cash flow. The next phase will involve an evolution toward scaling-up through mergers and acquisitions. A typical example is VINCI Airports, whose portfolio included 45 airports in 2020. Various forms of cooperation also exist between airport operators and carriers. Particularly in cases involving the restructuring of airlines that are facing difficulties and that serve an important hub function, attention is often paid to the possibility of support from airports in order to access the needed capital. Although the converse is less common, there are notable examples (e.g., Lufthansa and Munich Airport). There are also examples of extensive cooperation, as in the case of Ryanair and Charleroi Airport. Relations between handling companies and airport operators are usually limited to the granting of concessions and licences, usually after an auction process. In this case as well, in times of crisis (e.g., when a handling company goes bankrupt), actors are likely to turn to airport operators for possible support. The most visible forms of cooperation can be observed with regard to carriers. They all begin with mergers and acquisitions. It can be safely said that this has been an ongoing process since the Deregulation Act in the United States (Odoni, 2016a). The process has undeniably also been influenced by the evolution toward three major alliances, each having a strong European and American partner, and each being further supplemented by other players from all continents. It is interesting to note that some pillars of a given alliance also control a number of other companies from the same alliance through equity participations. One typical example is within the Star Alliance Lufthansa, which controls other European carriers (e.g., Swiss, Austrian, Brussels Airlines, and Eurowings). Other forms of cooperation will also remain possible across alliances, including through code share agreements and joint ventures. Unique capital holdings are emerging as well, as has been the case with Ryanair and AerLingus, two airlines based in Ireland that have been engaged in fierce competition for years. In the past, carriers were responsible for arranging their own handling. This was followed by a period in which carriers concentrated more on their core business (i.e., flying), thereby outsourcing their handling operations to specialized third-party companies. In times of crisis (e.g., strikes or

18

The Air Transportation Industry

bankruptcies), questions often arise with regard to whether carriers should or should not engage in full or partial self-handling. Takeovers and mergers are resulting in an increase in scale, and thus in concentration within the handling business, as exemplified by such companies as Dnata and Menzies. For the time being, this process is not enhancing the market power of the handling sector, and it is thus not increasing prices or improving financial results. Nevertheless, these developments are generating the possibility of concluding more multi-airport contracts in the long term, which will result in handling companies at different airports working for the same carrier. The analysis of new business models can be used to develop an exercise at the airport level, involving the quantification of the relationships existing between the various actors. The most important questions concern who is providing which services to whom, and the extent to which all actors are dependent on specific suppliers and customers.

4. Possible conflict situations within and between actors The preceding analysis provides an image of the industrial-economic structure of the aviation sector. This structure can be summarized as a heterogeneous market, with many actors of different sizes and with different economic power. This structure leads to competition within the various specializations, as well as to mergers, acquisitions, and financial participations, both horizontal and vertical. The future evolution of the market is related to an important three-part research question: to what extent will further scaling-up be pursued within the various actors? To what extent will this trend be driven by the search for economies of scale, scope or network benefits? How far will the future pursuit of verticalization go in order to gain control over a larger part of the transport chain? Within a continuously evolving market, it can be assumed that conflict situations will arise. Such conflicts are not limited to competition between airlines for passengers and freight on particular routes, to battles between airport operators to bring in carriers as customers, or to the competition between handling companies and other service companies to be allowed to work for the same airlines. Each of these market parties works with its own business model. These models evolve, often under the pressure of increasing competition and conflicts of interest. The specific issue in this regard has to do with the conflicts that can/will occur between different actors, partly as a result of exogenous and

Economic structure of the air transport business

19

endogenous developments. In the following paragraphs, we list and structure some of these potential conflicts. Several triggers stemming from the COVID-19 pandemic in 2020 are likely to persist beyond the pandemic, given the massive impact that the consequences of the pandemic has had on the aviation industry. The crisis affected the balance that had existed prior to the pandemic, and not every actor was affected in the same way. One serious and general effect that has serious consequences for the aviation actor is the fact that the COVID-19 pandemic brought world trade to a standstill. This is logical, given that potential market constitutes the first decision variable that carriers consider in the process of making airport choices. If the traffic volume on a specific route is not large enough to guarantee viable capacity utilization, there will be a risk of a major cash drain. With a sharply reduced supply and the elimination of passengers and cargo, the derived nature of demand threatens to cause a severe decline in the work volume of airports. This will subsequently have negative consequences for all actors established at specific airports, ranging from handling companies to other service providers, and even to the operators of car parks and shops. Several potentially crucial problems are addressed in the following section. 4.1 Airlines The sharp drop in demand caused by the COVID-19 pandemic occurred very quickly, and adjustments to supply are taking longer. In the short term, this is translating into overcapacity, low market prices and insufficient yield. In the slightly longer term, the situation will offer opportunities for reducing costs and making changes in positioning. Bankruptcies cause gaps in the market, planes are becoming less expensive, the scarcity of pilots has transformed into a surplus, and airports have slots to spare. Important disputes can arise between airlines and aircraft manufacturers and/or leasing companies. In response to overcapacity, airlines are currently seeking to remove aircraft from their fleets and are phasing out older aircraft. Even very low fuel prices cannot help to reduce this trend. At the same time, however, airlines are contractually obligated to accept delivery of the newly built aircraft that they have ordered, if only to safeguard their advance payments. Additional negotiations are largely related to possible delays in deliveries. Several carriers are likely to consider radical revisions to their financial structures, with a strong focus on the core business. For example, if airlines adopt a strategy of not owning their devices, strong growth could be expected in the leasing market. The COVID-19 pandemic has also resulted in far-reaching changes in the air cargo market. The crisis caused the cancellation of 90% of all

20

The Air Transportation Industry

passenger flights, with belly space accounting for 50% of global cargo capacity. As a result, air freight became a playing field for a limited number of specialized freight companies, such as Atlas and AirBridgeCargo, the express carriers (UPS, DHL and FedEx) and a few airlines with large freight fleets (e.g., Lufthansa).7 The sharply increased freight rates and the shortage of capacity have rendered these large international freight forwarders interesting to venture investors. Aging cargo aircraft are being taken out of storage and put back into service (including the American cargo company National Airlines). Not so long ago, those cargo planes were not expected to have much future potential. Carriers have responded to these developments in highly differing ways. Some have converted their stationary passenger planes into cargo-only flights. Others have adopted strategies of diversification, including the introduction of maritime companies into the capital of airlines (e.g., CMA CGM joined the French airline Groupe Dubreuil Aero in 2020). The management of the Lufthansa Group is considering a complete merger between the independent freight subsidiary Lufthansa Cargo and its passenger division. China Southern Airlines has submitted a request to the Chinese aviation authority (CAAC) to be allowed to convert its cargo division into a separate company. As a new player, Amazon is building its own cargo fleet. 4.2 Aircraft manufacturers The decreased demand for new aircraft and delays in the delivery schedule have combined to force aircraft manufacturers to reduce their production levels.8 This is also reflected throughout the rest of the production chain, as in the decreased purchase of parts from suppliers, including engine suppliers (e.g., Rolls-Royce and GE) and the manufacturers of fuselages and other systems. The situation will require renegotiations with customers and/or lawsuits. Moreover, it will inevitably translate into even stronger 7

United Airlines Cargo saw its third quarter 2020 freight revenue increase by 50% to $422 million. Freight accounted for one fourth of the company’s total turnover, in contrast to the usual 3%. The American freight forwarder Atlas Air saw its operating result (EBITDA) for the second quarter of 2020 increase by almost 200% to $247 million, with net profit amounting to $80 million. 8 In this case as well, the increasing importance of freight transport is at play. During the COVID-19 pandemic, Eva Air partially changed its existing order for seven B787-10s passenger aircraft from the US aircraft manufacturer Boeing to an order for three B777-200 cargo aircraft.

Economic structure of the air transport business

21

competition within the de facto duopoly of Airbus and Boeing, as well as with producers of smaller aircraft (e.g., Bombardier), given the risk that fewer wide-body jets will be used for long-haul flights. 4.3 Airport operators As a result of the COVID-19 pandemic, airports are being confronted with fewer flight movements, fewer passengers and cargo. As a result, they are generating less revenue. Most of their income from retail and car parks is disappearing as well. This means that almost all airports will incur losses in 2020.9 During the first 6 months of the 2020 financial year, the Schiphol Group suffered a record loss of V246 million, and Frankfurt Airport suffered a significant loss of V182 million in the second quarter of 2020. It is striking that, to date, airport operators have not been offered any support from the authorities. This situation has resulted in drastic cost reductions and savings programmes. All of these developments have had an effect on relations with shareholders. In the short term, the major problems resulting from such losses will translate into problems for some shareholders. This is especially the case for pension funds, which invested in the aviation industry largely because of the expected profits and, even more importantly, the expected stability of cash flow, which was intended to guarantee the ability to pay out pensions to their own members. It will be interesting to see how these shareholders will react to the radical changes in the pattern of cash flow? 4.4 Handling companies In many airports, access to the handling market is regulated, and the number of allowed players is therefore limited. For example, in Brussels, the number of actors after the auction process is currently limited to two. Although this essentially results in a duopoly, no duopoly power is exercised. On the contrary, devastating competition has resulted in the bankruptcy of one of the two playersdthe Belgian division of Swissportdin the course of 2020. The general character of this problem is also evident in clean-up operations. In the course of 2020, the Chinese owner of Swissport (HNA) decided 9

This also applies to the regional airports, many of which will certainly face major payment difficulties in 2020 if passenger numbers do not recover quickly (source: ACI). For example, in Germany, Paderborn Airport applied for a postponement of payment in 2020, following the loss of 85% of the passengers.

22

The Air Transportation Industry

to transfer the majority of shares in its subsidiary to British and American lenders through a sizable debt-restructuring process amounting to V1.9 billion. Debts are being exchanged for shares through a “debt-for-equity” scheme. A second part of the financing agreement involves a long-term loan of V500 million. Other types of changes are taking place in the shareholder structure as well. For example, Lufthansa Cargo is largely withdrawing from cargo handling activities at its home base of Frankfurt Airport (Fraport) and its outstations. As of July 1, 2020, these activities were assumed by the service provider Fiege, a company that is thus experiencing its first large-scale involvement in the area of air cargo handling. 4.5 Relations with the government Possible conflicts with the government are arising with regard to state aid and regulation. For example, the European Union has allowed various European national governments to provide state support to their own legacy carriers. The arguments for these arrangements boiled down to the highly important role played by these carriers with regard to the national economy in terms of employment and air connections, combined with the extreme impact that the crisis has had on aviation activities. Such assistance has also been granted to carriers in other continents. These developments have sparked complaints from carriers that have not been able to benefit from such government support, and they have raised questions concerning the existence of a level playing field within the market. At the same time, however, current situations are requiring governments to engage in perpetual monitoring for possible abuse. For example, the four major American cargo carriers (Atlas Air, Kalitta Air, Western Global Airlines, and Amerijet International) were shown to have wrongly received around $630 million in state aid.10 Under pressure from the climate debate, the entire aviation industry is having to rethink its ecological footprint. In exchange for some of the state aid, airlines must use the crisis to achieve innovation and rejuvenation in their fleets, while making them more ecologically friendly. Aircraft 10

The grant was provided through the “Payroll Support Program” (PSP) to provide financial assistance to ailing airlines throughout the first months of the COVID-19 crisis.

Economic structure of the air transport business

23

manufacturers are expected to achieve innovations that will ultimately result in zero-emission flying. All layers of government are increasingly emphasizing the transition toward a clean energy industry.11 Crises require adjustments to regulation. At the start of the COVID-19 crisis in March 2020, the European Union decided to override the strict rules for allocating landing rights at major airports in order to accommodate as many air cargo aircraft as possible, in order to ensure the supply of protective medical clothing.12 This turned out to be a good move. In the second half of 2020, the German aviation sector requested that the current exemptions for the EU slot rules be extended until the end of the next winter season (i.e., the end of March 2021).

5. Conclusions The aviation sector is a capital-intensive business that is subject to rapid technological and organizational changes. It is in constant need of knowledge about future market evolutions. Industrial economic insights indicate that there is no single homogeneous air transport market. Instead, the aviation business consists of a configuration of many different, yet interconnected sub-markets that are engaged in interaction with each other. Although the future of the aviation market is surrounded by considerable uncertainty, one thing seems to be clear: societies will behave differently in the wake of COVID-19. This raises a number of questions for the aviation market. To what extent will further specialization occur? Will new alliances be established (e.g., as a result of international capital movements)? Will the international aviation market, in all its dimensions, continue to be an example of easy entry and exit? Will government intervention increase or decrease, and at which levels? These questions require continuous analysis in order to substantiate meaningful decisions. This chapter provides a closer examination of the relationships between market actors that could potentially give rise to potential problems in the long term. Quite a few of these relationships exist at the level of all actors.

The European Union would like to contribute to this transition (e.g., by shortening flight routes, reducing fragmentation of the airspace above Europe and continuing to develop a Single European Sky). An additional alternative could involve a CO2 tax on kerosene. 12 The “20/80 rule” obliges airlines to use 80% of the requested landing rights in accordance with the application. If they do not, they will lose their historical landing rights at that airport. 11

24

The Air Transportation Industry

The crucial question in this regard concerns the extent to which strategic decisions are driven by the search for scale effects. Due to the derivative nature of the demand for air transport, the COVID-19 pandemic has had a major impact on all market players. The various actors are now being confronted with specific conflict situations that require their own strategic adjustments. Mutual relationships will change, and these changes will be driven to a significant extent by international capital flows. There remains a clear need for the further development and updating of a battery of models and instruments that allow the estimation and quantification of the consequences of all possible changes in strategies and mutual relationships.

References Balliauw, M., Meersman, H., Onghena, E., Van de Voorde, E., 2018. US all-cargo carriers’ cost structure and efficiency: a stochastic frontier analysis. Transport. Res. Pol. Pract. 112, 29e45. Belobaba, P.P., 2016. Overview of airline economics, markets and demand. In: Belobaba, P., Odoni, A., Barnhart, C. (Eds.), The Global Airline Industry, second ed. Wiley, Chichester, pp. 47e73. Bringmann, K., De Langhe, K., Kupfer, F., Sys, C., Van de Voorde, E., Vanelslander, T., 2018. Cooperation between airports: a focus on the financial intertwinement of European airport operators. J. Air Transport. Manag. 69, 59e71. Fageda, X., Flores-Fillol, R., Theilen, B., 2019. Joint ventures in the transatlantic airline market. In: Cullinane, K. (Ed.), Airline Economics in Europe. Emerald, Bingley, pp. 117e136. Flemish Airport Commission, 2012. Concurrentiepositie luchthaven Zaventem versus luchthavens Charleroi en Luik, Brussels. Flemish Airport Commission, Brussels. Gillen, D., Niemeier, H.-M., 2008. The European Union: Evolution of Privatization, Regulation and Slot Reform. In: Winston, C., de Rus, G. (Eds.), Aviation Infrastructure Performance. A Study in Comparative Political Economy. The Brookings Institution, Washington D.C., pp. 36e61. Graham, A., 2010. Airport strategies to gain competitive advantage. In: Forsyth, P., Gillen, D., Müller, J., Niemeier, H.-M. (Eds.), Airport Competition. The European Experience. Ashgate, Farnham, pp. 89e102. Greer, M.R., 2016. In: Bitzan, J.D., Peoples, J.H., Wilson, W.W. (Eds.), Airline Mergers in the United States since 2005. What Impact Have They Had on Airline Efficiency, Airline Efficiency. Emerald, Bingley, pp. 161e195. Gudmundsson, S.V., 2019. European air transport regulation: achievements and future challenges. In: Cullinane, K. (Ed.), Airline Economics in Europe. Emerald, Bingley, pp. 9e56. IATA, 2000-2020. World Air Transport Statistics (WATS). Montreal. ICAO, 2015. Air Transport in Figures. Economic Development at a Glance, Montreal. Kupfer, F., Kessels, R., Goos, P., Van de Voorde, E., Verhetsel, A., 2016. The origindestination airport choice for all-cargo aircraft operations in Europe. Transport. Res. E Logist. Transport. Rev. 87 (1), 53e74. Kupfer, F., Meersman, H., Onghena, E., Van de Voorde, E., 2017. The underlying drivers and future development of air cargo. J. Air Transport. Manag. 61, 6e14.

Economic structure of the air transport business

25

Mahoney, D., Wilson, W.W., 2012. The size and growth of airports. In: Peoples, J. (Ed.), Pricing Behavior and Non-Price Characteristics in the Airline Industry. Emerald, Bingley, pp. 233e273. Meersman, H., Van de Voorde, E., Vanelslander, T., 2011. The future air transport sector. A modified market and ownership structure. In: Macario, R., Van de Voorde, E. (Eds.), Critical Issues in Air Transport Economics and Business. Routledge, Abingdon, pp. 10e28. Merkert, R., Van de Voorde, E., de Wit, J., 2017. Making or breaking: key success factors in the air cargo market. J. Air Transport. Manag. 61, 1e5. Morrison, S.A., Winston, C., 2008. Delayed! U.S. Aviation infrastructure policy at a crossroads. In: Winston, C., de Rus, G. (Eds.), Aviation Infrastructure Performance. A Study in Comparative Political Economy. The Brookings Institution, Washington D.C., pp. 7e35. Odoni, A.R., 2016a. The international institutional and regulatory environment. In: Belobaba, P., Odoni, A., Barnhart, C. (Eds.), The Global Airline Industry, second ed. Wiley, Chichester, pp. 19e46. Odoni, A.R., 2016b. Airports. In: Belobaba, P., Odoni, A., Barnhart, C. (Eds.), The Global Airline Industry, second ed. Wiley, Chichester, pp. 361e393. Starkie, D., 2008. Aviation Markets. Studies in Competition and Regulatory Reform. Ashgate, Farnham. Templin, C., 2010. Competition for airport services e ground handling services in Europe: case studies on six major European hubs. In: Forsyth, P., Gillen, D., Müller, J., Niemeier, H.-M. (Eds.), Airport Competition. The European Experience. Ashgate, Farnham, pp. 393e412. Vennix, S., 2017. Economic Importance of Air Transport and Airport Activities in Belgium, Working Paper 324. National Bank of Belgium, Brussels.

This page intentionally left blank

CHAPTER 2

The burden of a ton CO2! Emission trading systems and the air transport business Chaouki Mustapha

Air Transport, ICAO, Montreal, QC, Canada

1. Introduction Climate change1 has become an issue of global importance. Rising temperatures, more severe weather, increases in flooding and desertification, sea level rise, and losses in biodiversity are among the impacts of this phenomenon, which is caused by emissions of greenhouse gases2 (GHG) due to human activities. While the contribution of air transport in those emissions is low (around 2%), the sector’s impressive growth, up until the COVID-19 outbreak in early 2020, and its high visibility have attracted the attention of the public, media, and policymakers in many regions of the world and have amplified the urgency to act and reduce emissions. Emissions trading3 is one of the measures available to address the problem in a cost effective way. This chapter aims to answer the following questions: (1) why a global emission trading system is critical to the achievement of the sector’s climate change goals? (2) to what extent would such a system affect the air transport business? (3) what would be the role of in-sector emission reduction measures? The chapter starts with an introduction to the global challenge of “A change of climate which is attributed directly or indirectly to human activity that alters the composition of the global atmosphere and which is in addition to natural climate variability observed over comparable time periods,” Article 1 of the United Nations Framework Convention on Climate Change (UNFCCC). 2 Greenhouse gases slow the process by which the earth releases the energy it receives from the sun, leading to the warming the earth’s surface and the lower atmosphere. These gases include water vapor (H2O), carbon dioxide (CO2), methane (CH4), nitrous oxide (N2O), ozone (O3), Chlorofluorocarbons (CFCs), Hydrofluorocarbons (including HCFCs and HFCs), Sulfur hexafluoride (SF6) and Perfluorocarbons (PFCs). 3 There are three types of emissions trading systems: (1) Cap-and-trade; (2) Baseline-andcredit; and (3) Offsetting. 1

The Air Transportation Industry ISBN 978-0-323-91522-9 https://doi.org/10.1016/B978-0-323-91522-9.00002-6

© 2022 Elsevier Inc. All rights reserved.

27

28

The Air Transportation Industry

climate change and CO2 emissions and the global policies designed to address it. This is followed by a presentation of the air transport industry traffic and CO2 emissions as well as the policies and measures adopted to limit or reduce the latter. Then, the Carbon Offsetting and Reduction Scheme for International Civil Aviation (CORSIA) is presented followed by an analysis of the supply and demand for carbon offsets and laying the grounds for a detailed assessment of the impact of CORSIA on air traffic and airline financial results, presented in the subsequent and penultimate section of this chapter.

2. The global challenge of climate change and CO2 emissions It is estimated that global warming due to human activities has reached approximately 1.2 C above pre-industrial levels in 20204 driven by past and current anthropogenic5 greenhouse gas (GHG) emissions, which have been increasing steadily since 1990. Carbon dioxide is the most significant greenhouse gas representing about 75% of total emissions while the energy sector (comprising mainly emissions due to fuel combustion) has the highest share of 71% amongst all sectors of activity. CO2 emissions have increased by more than 50% since 1990 to reach around 38 billion tons in 2019.6 At the current rate of warming, the world will reach 1.5 C increase sometime between 2030 and 2050 and 3 C by 2100, according to the 2018 IPCC Special Report on Global Warming of 1.5 C (SR15) (IPCC, 2018) Limiting global warming at 1.5 or 2 C requires cumulative net CO2 emissions (past, present, and future emissions) to stabilize. To achieve this, global emissions of CO2 need to reach a peak and then start declining to finally reach net zero. The IPCC SR15 identifies pathways limiting global warming to 1.5 C. Every fraction of global warming matters in terms of the scale and severity of climate related risks. In SR15, the IPCC highlights that in order to achieve the goal of limiting global warming to 1.5 C, global net anthropogenic CO2 emissions should decline by about 45% from 2010 levels by 2030 and reach net zero 4

World Meteorological Organization, The State of the Global Climate 2020 (https:// public.wmo.int/en/our-mandate/climate/wmo-statement-state-of-global-climate). 5 caused by human activities. Anthropogenic greenhouse gases include carbon dioxide (CO2), methane (CH4), nitrous oxide (N2O), hydrofluorocarbons (incl. HCFCs and HFCs), sulfur hexafluoride (SF6) and perfluorocarbons (PFCs). 6 World Resources Institute, Climate Watch; Global Carbon Project.

The burden of a ton CO2! Emission trading systems and the air transport business

29

around 2050. The IPCC also notes that limiting global warming to 2 C would require CO2 emissions to be reduced by about 25% by 2030 and reach net zero around 2070. Therefore, global CO2 emissions in 2030 should not exceed 20 billion tons for the 1.5 C scenario and 26 billion tons for the 2 C scenario. Reaching net zero emissions in 2050 would require removing CO2 from the air.

3. Overview of global policies to address climate change After gaining the attention of both the public and governments, concrete actions at the global level were sought to address climate change. In 1992, the United Nations Framework Convention on Climate Change (UNFCCC) was opened for signature at the United Nations Conference on Environment and Development.7 The objective of the convention is “to achieve stabilization of greenhouse gas concentrations in the atmosphere at a level that would prevent dangerous anthropogenic8 interference with the climate system.”9 The convention itself sets no mandatory limits on greenhouse gas emissions for individual countries but it provides for updates (called “protocols”) that would set mandatory emission limits. The first update was the Kyoto Protocol, which was adopted in Kyoto, Japan, on December 11, 1997 and entered into force on February 16, 2005. The protocol set binding emissions targets for 37 industrialized countries for reducing greenhouse gas (GHG) emissions, which amount to an average of 5% against 1990 levels over the 5-year period 2008e2012. A second commitment period from 2013 to 2020 aimed at reducing GHG emissions by at least 18% below 1990, was adopted in December 2012 in Doha, Qatar. Under the Kyoto Protocol, countries must meet their targets primarily through national measures. However, the Protocol offered them an

7

The United Nations Conference on Environment and Development, also known as the Rio de Janeiro Earth Summit, the Rio Summit, the Rio Conference, and the Earth Summit, was held in Rio de Janeiro from June 3e14 1992. 8 Caused by human activities. Anthropogenic greenhouse gases include carbon dioxide (CO2), methane (CH4), nitrous oxide (N2O), hydrofluorocarbons (incl. HCFCs and HFCs), sulfur hexafluoride (SF6) and perfluorocarbons (PFCs). 9 Article 2 of the United Nations Framework Convention on Climate Change (UNFCCC).

30

The Air Transportation Industry

additional means of meeting their targets, in a more cost-effective way, through the use of three market-based mechanisms: • Emissions trading • Clean development mechanism (CDM) • Joint implementation ( JI). The second update is the Paris Agreement covering the period beyond 2020. It was adopted during the 21st session of the Conference of the Parties to the UNFCCC (COP 21) held in Paris from 30 November to December 13, 2015 and entered into force on November 4, 2016. The main goal of this agreement is “holding the increase in the global average temperature to well below 2 C above pre-industrial levels and pursuing efforts to limit the temperature increase to 1.5 C above pre-industrial levels, recognizing that this would significantly reduce the risks and impacts of climate change”.10 The agreement aims to reach net zero emissions in the second half of the century. States are requested to prepare, communicate, and maintain their plans, referred to as Nationally Determined Contributions (NDCs), designed to contribute to reaching this goal. As was the case for the Kyoto Protocol, the Paris Agreement permits, in its Article 6, to States to seek voluntary cooperation in order to meet their goals in a cost-effective way. This would “involve the use of internationally transferred mitigation outcomes toward”11 the NDCs. Discussions on the implementation of Article six and in particular development of robust accounting rules to ensure the avoidance of double counting are still ongoing. Contrary to GHG emissions from domestic civil aviation, GHG emissions from international civil aviation are not included in national inventories and are dealt with through the International Civil Aviation Organization (ICAO). International air transport is by far the main component of international civil aviation.

4. Air transport industry traffic and CO2 emissions This section briefly describes the contribution of air transport to society and the economy, explains the past progress of its traffic and CO2 emissions including efficiency improvements and presents current projections, which show that substantial efforts are needed in order to overcome the challenge of meeting a 1.5 C or even a 2 C scenario. 10 11

Article 2 of the Paris Agreement. Article 6 of the Paris Agreement.

The burden of a ton CO2! Emission trading systems and the air transport business

31

4.1 Historical development The air transport industry plays a significant role in today’s world bringing people together, linking businesses, facilitating trade and e-commerce, supporting tourism and delivering much needed supplies such as medicines and vaccines. It is estimated that in 2019, the industry contributed 3.5 trillion US$ to the world economy and supported 87.7 million jobs worldwide.12 In that same year, the industry recorded 4.5 billion passengers13 and 57.6 million tons of freight representing an average annual growth of 4.8% and 4.0%, respectively, since 1990. Taking the distance flown into consideration,14 the corresponding growth rates amount to 5.4% and 4.7%, respectively over the same period. Combining both metrics, the average annual growth in terms of total revenue ton kilometers (RTKs) stands at 5.2%. During the same period, air transport CO2 emissions have increased at a much lower rate of 2.45% growing from around 453 million tons in 1990 to around 914 million tons in 2019 (ATAG, 2020a). The cumulative emissions during this period stands at 19 billion tons of CO2. The share of these emissions in total CO2 emissions (including energy, industrial processes, land use, and forestry) has also increased from 1.8% in 1990 to around 2.4% in 2019. Air transport CO2 emissions are proportional to the quantity of fuel burned by aircraft during flight15 and on the ground. The aircraft fuel productivity, measured in terms of RTKs per liter of fuel burned, has more than doubled between 1990 and 2019 implying that in the absence of technological and operational improvement introduced over the years, the industry CO2 emissions would have been more than double what they actually were in 2019. That represents a cumulative saving of around 12.3 billion tons of CO2 over the 29-year period (Fig. 2.1). Between 1990 and 2019, newer technology aircraft have entered service such as A330, A340, B777, Bombardier CRJs, Embraer ERJs, A380, B7478, B787, B737 max, A320 neo, A330 neo, A220, Embraer E-Jets, and A350. Each new generation of aircraft was between 15% and 20% more efficient than the previous one in the same category. Improvements were 12

Air Transport Action Group (ATAG) (https://aviationbenefits.org/economic-growth/ adding-value-to-the-economy/). 13 ICAO, Annual Report of the Council (2019). 14 Using the revenue-passenger-kilometer and revenue-ton-kilometer of freight. 15 This includes takeoff, cruise approach and landing.

32

The Air Transportation Industry

Figure 2.1 Historical trends of air traffic, fuel productivity and CO2 emissions. (Sources: ICAO and ATAG.)

mainly the result of newer and more efficient engine technologies such as the adoption of higher pressure ratios or the introduction of the geared turbo-fan technology. Other sources of technology related efficiency improvements include changes in aircraft aerodynamics and weight reduction. These technological improvements were infused into the fleet of aircraft in service through the purchase of new aircraft. Operational improvements were the result of sustained gains in load factors, enhanced flight planning, better fuel management, improved flight profiles (Continuous descent approaches and continuous climb operations), implementation of new air navigation initiatives (Reduced vertical separation minima (RVSM) and Required Navigation Performance (RNP), NextGen in the United States, SESAR in Europe, etc.), modernization of air traffic management systems and enhanced flight planning. 4.2 Projections The forecast prior to the COVID-19 pandemic shows that air traffic will continue to grow at rates higher than those of the overall economy (as measured by gross domestic product (GDP)). Table 2.1 illustrates the future average annual growth rates projected for passenger and cargo traffic, as well as for the combined RTKs up to the year 2050.

The burden of a ton CO2! Emission trading systems and the air transport business

Table 2.1 Total air traffic forecasts 2020e2050. 2020e2030 2030e2040

2040e2050

RPKs FTKs RTKs

4.2% 3.5% 4.1%

4.4% 3.7% 4.3%

4.3% 3.6% 4.2%

33

The above forecasts, which represent a central scenario, were prepared based on the assumption of an average annual growth rate of GDP of 3.2% in the first decade, progressively declining in subsequent periods.16 Prices (airfares and cargo rates) were also assumed to decline slightly at the rate of 0.2% per annum17 in real terms. Table 2.4 presents two additional scenarios High and Low, based on alternative GDP growth rate assumptions of 3.9 and 2.5% per annum for the first forecast period, progressively declining in subsequent periods with no change in the price assumptions. The pandemic has resulted in an unprecedented decline in air traffic in year 2020. While substantial uncertainties remain at the time of writing this chapter, the recovery is expected to start in 2021, but the 2019 traffic levels will not be regained before 2023 as illustrated in Fig. 2.2.

Figure 2.2 Impact of the COVID-19 pandemic on air traffic forecasts.

16

The GDP forecast assumption was based on long term forecasts by the Organization for Economic Co-operation and Development (OECD) (http://www.oecd.org/economy/ growth/scenarios-for-the-world-economy-to-2060.htm) and PricewaterhouseCoopers (https://www.pwc.com/gx/en/research-insights/economy/the-world-in-2050. html#data). 17 Based on an expected continuation of past trends.

34

The Air Transportation Industry

Growth in air traffic will inevitably result in growth in CO2 emissions, despite all efforts deployed by the industry to reduce them. Assuming a frozen technology and operations scenario, over the next 30 years, air transport CO2 emissions will amount to around 1400 and 3250 million tons in 2030 and 2050, respectively. Under a more optimistic scenario, assuming 1.5% average annual efficiency improvement over the same period, air transport CO2 emissions will amount to around 1225 and 2050 million tons in 2030 and 2050, respectively. In a 1.5 C scenario (a decline of 45% from 2010 levels by 2030 and reach net zero around 2050), air transport CO2 emissions should decline to below 365 million tons in 2030 and reach net zero emissions in 2050. In a 2 C scenario (a decline of 25% from 2010 levels by 2030 and reach net zero around 2070), these emissions should decline to 500 million tons in 2030 and reach net zero emissions in 2070. It is therefore, obvious, that substantial efforts need to be deployed in order to overcome the challenge of meeting a 1.5 C or even a 2 C scenario. The air transport sector would need to seek emissions reductions in the sector as well as in other sectors using available mechanisms.

5. Policies to address CO2 emissions from international air transport Discussions at the government and industry level on how to limit the impact of aviation on the climate and in particular, to reduce CO2 emissions have begun since the adoption of the Kyoto Protocol in 1997. In 2001, the 33rd ICAO Assembly endorsed the development of an open emissions trading system for international aviation. In 2010, the 37th Session of the ICAO Assembly adopted resolution A37-19 related to international aviation and climate change including the following provisions: (1) endorsement of a global goal of annual average fuel efficiency improvement of 2% until 2020 and of a global aspirational goal of 2% annual fuel efficiency improvement from 2021 to 2050; (2) a medium-term global aspirational goal of carbon neutral growth from 2020 onward; (3) a decision to explore the feasibility of a long-term global aspirational goal; (4) development of a framework for market-based measures (MBMs); (5) development of a global CO2 Standard for aircraft and (6) support for the development, deployment and use of sustainable aviation fuels.

The burden of a ton CO2! Emission trading systems and the air transport business

35

In 2016, the 39th Session of the ICAO Assembly adopted the Carbon Offsetting and Reduction Scheme for International Aviation (CORSIA), which aims to cap aviation emissions at 2020 levels by requiring airlines to offset the growth of their emissions beyond that year. In addition, industry has adopted the following goals: (1) achieve an average improvement in fuel efficiency of 1.5% per year from 2009 to 2020; (2) cap net aviation CO2 emissions at the 2020 level (carbon-neutral growth); and (3) reduce net aviation CO2 emissions by 50% by 2050, relative to 2005 levels. At a regional level, the European Union had adopted in 2008 a legislation to include CO2 emissions from aviation into the EU Emissions Trading System (EU ETS) starting 2012. The legislation was designed to apply to emissions from flights from, to and within the European Economic Area (EEA).18 The EU, subsequently, decided to limit the scope of the EU ETS to flights within the EEA until 2016 in support of the development of a global measure by ICAO. After the adoption of CORSIA by ICAO, the EU decided to maintain the limitation of the geographic scope of the EU ETS from 2017 onwards.

6. Measures available to reduce air transport CO2 emissions While the policies listed in the previous section outline the goals and broadly define what needs to be done to reduce CO2 emissions from air transport, a basket of measures is required to determine how those goals will be attained. Numerous measures are available to this end, which can be grouped into the following categories: (1) Technological developments; (2) New standards; (3) Airline operations; (4) Air navigation service provider’ operations; (5) Sustainable Aviation Fuels (SAFs) and (6) Market based measures. The first four categories are considered as in-sector and the last two as out-of-sector reduction measures. All measures will contribute to achieving the sector’s climate goals; however, the level, timing and cost of their contribution vary considerably. 6.1 Technological developments New technologies have long been a major driver in modernizing aircraft and engines. Their development goes typically through a nine-stage process 18

EU Member States, plus Iceland, Liechtenstein and Norway.

36

The Air Transportation Industry

before being introduced commercially,19 spanning from the observation of basic principles and the formulation of technology concept to the test and demonstration and successful operation stages. This involves a great deal of research and development by aircraft and engine manufacturers and also by governments. New technologies that are currently at the latest stages of development include advanced engine concepts and new engine architecture as well as improved aerodynamics and aircraft systems.20 These technologies would be introduced onto new aircraft types overtime between 2020 and 2035 and benefits will be achieved through aircraft fleet renewal and expansion. Existing technologies may also be introduced as retrofits and upgrades to aircraft in service and may apply to aerodynamics, material and structure, aircraft systems and advanced engine components. These technologies are expected to result in a further fuel efficiency improvement of 30% by 2030. More innovative technologies would be needed beyond 2035 to progress toward the achievement of the set goals. Technologies under development include open rotor engines, flying without landing gear, windowless fuselage, morphing wing, blended wing-body aircraft, hybrid-electric aircraft and fully electric aircraft. There remain substantial uncertainties on whether these technologies will be commercialized in the end. Once they do, their introduction into the aircraft fleet will take some time. The development and deployment of new aircraft and engine technologies is capital intensive and therefore costly. Their environmental benefits are significant but given that their commercial introduction occurs through aircraft fleet renewal and retrofit and that the useful life of an aircraft is in the range of 25e30 years, the resulting benefits will be slow to materialize. Accelerating the fleet renewal will lead to higher costs. 6.2 New standards New standards play an important role in advancing the integration of new technologies into new aircraft. In March 2017, the ICAO Council adopted a new aircraft CO2 emissions standard. This design certification standard will be reviewed periodically to take advantage of new technologies as they become available. New standards serve to harmonize the safety and

19 20

NASA’s Technological Readiness Levels. ATAG, Waypoint 2050, (September 2020b) (https://aviationbenefits.org/environmentalefficiency/climate-action/waypoint-2050/).

The burden of a ton CO2! Emission trading systems and the air transport business

37

efficiency requirements worldwide and certification authorities of ICAO member States are expected to adopt ICAO standards in their national regulations. Similar to new technologies, the environmental benefits of new standards are gained through fleet renewal and retrofits and are significant but will be slow to materialize. 6.3 Airline operations Improvements in airline operations, such as load factor increase, selection of aircraft best suited to the route, weight reduction, reduced speed, minimizing the use of thrust reversers, minimizing flaps and engine wash have also been a source of fuel efficiency gains and are expected to offer further savings in fuel burn and reduction in CO2 emissions. In general, these measures are not costly. Their environmental benefits can be immediate but are relatively small. In addition, as operations become more efficient, incremental benefits become more difficult. 6.4 Air navigation service provider’ operations Similarly, the modernization of air navigation services’ provision such as the continuous descent operations (CDO), continuous climb operations (CCO), airport collaborative decision making, improvements in aircraft departure and arrival management (DMAN and AMAN), measures to improve aircraft taxiing and guidance on apron and facilitation of aircraft high speed turnoff, will continue to deliver additional fuel savings and reduction in CO2 emissions. While in general, these measures are not costly; their environmental benefits can also be immediate but relatively small. Similar to airline operations, incremental benefits become more difficult to gain as air navigation service provider’ operations become more efficient. 6.5 Sustainable aviation fuels Given the limitations in the benefits from technological and operational measures, at least in the medium term, the development, deployment and use of Sustainable Aviation Fuels is key to the achievement of the goals set by the air transport industry. These fuels differ from conventional fossil fuels in that they are produced from a different feedstock such as cooking oil, plant oils, municipal waste, waste gases, and agricultural residues. They must meet certain sustainability criteria such as reducing GHG emissions on a life

38

The Air Transportation Industry

cycle basis,21 not competing with food, not impacting water and not causing deforestation. While both SAF and fossil jet fuels produce similar quantities of CO2 emissions during flight, SAF absorb CO2 during their production stage. When considering lifecycle CO2 emissions, SAF fuels produce typically 70 to 80% less than fossil jet fuel.22 SAF currently available are drop-in-fuels and are mixed with fossil fuel within a limit of up to 50% depending on the SAF fuel used. This limit is expected to be increased overtime. The development and deployment of SAF is at its early stages. The first test flight of a SAF took place in 2008. Since then, seven SAF production pathways have been approved and around 40 million liters have been produced. A solid ramp-up is underway and industry is projecting that by 2025, production will increase to 2986 million liters in a baseline scenario and to around 761 million liters in a low scenario. The baseline scenario includes an assumption of supportive policy interventions by governments around the world. The challenge with SAF is that their production cost is 2e3 times higher than conventional (fossil) fuels. In addition, they compete with other sectors for the same feedstock. It is hoped that the production cost will decline over time with increases in supply and improvements in technology; and that competition with other sectors be somehow regulated by governments in order to assign priority to the air transport sector given its dependence on fossil fuels and expensive in sector abatement solutions. Since the early days of air transport, the basic aircraft shape (tube and wing) and energy source of aircraft did not change much. A long term solution to the problem of emissions in general could reside in switching to a different source of energy such as electricity or hydrogen. While the electric aircraft technology has evolved substantially over the past decade and a few prototypes have been flying, the operating range remains limited. In the future, these aircraft could be operated on regional and short haul flights. Medium and long haul flights will continue to rely on liquid fuels including SAF.

21 22

Extraction, production, distribution and consumption. International Air Transport Association (IATA), IATA Technology Roadmap (https:// www.iata.org/en/programs/environment/technology-roadmap/).

The burden of a ton CO2! Emission trading systems and the air transport business

39

6.6 Market-based measures Market-based measures can take the form of a tax, a charge or an emissions trading system. A tax is normally designed and collected by a government authority and is used for general purposes. A charge (or a fee) is collected against the provision of a service and its proceeds are assigned to the continuation of that service. Both taxes and charges have historically been applied to address the impact of aviation on the environment, initially for noise and subsequently for emissions. Since 1981, the ICAO Council has adopted a specific ICAO policy on noise-related charges. In 1996, the ICAO Council adopted a Resolution on Environmental Charges and Taxes calling on States, if they want to introduce an environmental levy, to avoid imposing taxes and that the funds collected should be used, primarily, in the mitigation of the environmental impact of emissions. The charges should be cost-related and should not discriminate against air transport. Emissions trading is a flexible mechanism allowing polluters with higher abatement cost to meet their target by paying for emissions reductions by other polluters with lower abatement cost. This permits the attainment of the collective environmental goals with lower costs. Emissions trading can be used as part of a broad basket of measures to reduce emissions and is often an essential solution for hard-to-abate sectors such as air transport. In the long run, however, and as emissions reductions have to continue until reaching net zero emissions by all sectors, the cost of this solution will increase progressively. 6.7 Comparison of measures Table 2.2 provides a summary of the relative environmental benefit, cost, and timescale of the various categories of measures: In light of the above, improvement through operations and market based measures are more practical in the short term (2020e2035). Benefits from sustainable aviation fuels and new technologies and standards become more feasible in the medium term (2035e2050) and would become the main source of CO2 emissions reductions beyond 2050.

7. Carbon Offsetting and Reduction Scheme for International Civil Aviation (CORSIA) As concluded in the previous section, market based measures constitute an important cost effective solution to alleviate the climate change impact of

40

The Air Transportation Industry

Table 2.2 Relative benefit, cost and timing of the various categories of measures. Environmental benefits Cost Timescale

Technological developments New standards

Air navigation service Provider’s operations Sustainable aviation fuels

Relatively high Relatively high Relatively small Relatively small High

Market based measures

High

Airline operations

High High Relatively low Relatively low Relatively high Relatively low

Medium and long terms Medium and long terms Short term Short term Medium and long terms Short and medium terms

air transport in the short and medium terms. This feature of market based measures is not unique to air transport but is valid for other sectors as well. Several schemes have been adopted at the international, regional and national levels and others are under development. Existing schemes include the UNFCCC Clean Development Mechanism (CDM) and Joint Implementation ( JI) mechanism, as well as the European Emissions Trading Scheme (or EU ETS). National and sub-national schemes exist in Australia, Canada, China, Korea, Japan, Switzerland, and the United States. The International Maritime Organization (IMO) has also been working on a scheme, which is not yet finalized. In the air transport sector, CORSIA, adopted by ICAO in 2016, is an emission trading scheme aimed at capping the net emissions from international civil aviation (mainly international air transport) at the 2020 level.23 The duration of the scheme is currently set from 2021 to 2035 and is composed of three phases: • Pilot Phase: from 2021 to 2023 • First Phase: from 2024 to 2026 • Second Phase: from 2027 to 2035 State participation in the pilot phase and the first phase is voluntary. From 2027, all States whose individual share in total international RTKs in 23

The baseline for CORSIA has been changed to 2019 due to substantial decline in air traffic in 2020 due to COVID-19 and to avoid inappropriate economic burden on the aviation industry (https://www.icao.int/Newsroom/Pages/ICAO-Council-agrees-tothe-safeguard-adjustment-for-CORSIA-in-light-of-COVID19-pandemic.aspx).

The burden of a ton CO2! Emission trading systems and the air transport business

41

2018 was above 0.5% or whose ranking was within the top 90% are included. Least Developed Countries, Small Island Developing States and Landlocked Developing Countries are exempt unless they volunteer to participate. Other exemptions apply to aircraft operators with less than 10,000 metric tons of CO2 emissions per year from international operations and to aircraft with less than 5700 kg of maximum takeoff mass (MTOM). While all States, who meet the conditions above, are expected to participate in the scheme, those who do not wish to do so can notify ICAO by filing a ‘difference’. They may always decide to join at a later stage. From 2021, the growth in emissions above 2020 level is to be offset. The share of each operator in the quantity to be offset is determined using a growth factor that is based on the overall sector’s growth and the operator’s own growth using respective weights that change over time. From 2021 to 2029, the growth factor is equal to the sector’s growth. From 2030 to 2032, the growth factor is equal to a maximum of 80% of the sector’s growth and a minimum of 20% of the operator’s own growth. From 2033 to 2035, the growth factor is equal to a maximum of 30% of the sector’s growth and a minimum of 70% of the operator’s own growth. The reason for this dynamic formula is to limit the impact (at least between 2021 and 2029) on the fast growing operators that tend to be young and to belong to developing States while slow growing operators tend to be mature and to belong to developed States. In order to ensure a level playing field and avoid unfair competition, a route-based approach was adopted for CORSIA. If one State is not participating in CORSIA, all international traffic between this State and all other States is excluded from CORSIA coverage, regardless of the State of aircraft operator. While offsetting CO2 emissions is required only for international routes covered by CORSIA, monitoring, reporting and verification (MRV) of CO2 emissions is required for all international routes from January 1, 2019 until December 31, 2035. CORSIA allows operators to account for emissions reductions resulting from the use of sustainable aviation fuels. These fuels have to be approved by ICAO in order to ensure that they meet a set of sustainability criteria. So far, the organizations, International Sustainability and Carbon Certification and Roundtable on Sustainable Biomaterials, have been approved by ICAO to certify the sustainability and life-cycle emissions values of candidate fuels. The certified fuels are referred to as “CORSIA eligible fuels.”

42

The Air Transportation Industry

The carbon offsets used by aircraft operators to meet their offsetting requirements under CORSIA are subject to a set of eligibility criteria which are divided into two subcategories: • program design elements criteria (11); and • carbon offset credit integrity (8). An ICAO Technical Advisory Body (TAB) is responsible for the assessment of emissions unit programs against the emissions units’ criteria and develop recommendations on the list of eligible emissions unit programs (and potentially project types) whose emissions units would be eligible for use under CORSIA, for consideration by the ICAO Council. The approved offsets are referred to as “CORSIA eligible emissions units.” Projects generating units must have started their first crediting period from January 1, 2016 and emissions reductions must occur no later than December 31, 2020.24 The scheme includes several other provisions related to fuel use monitoring methods, CO2 emissions estimation and reporting methods and tools, emissions monitoring plans, treatment of new entrants, CORSIA Central Registry (CCR) and CORSIA review process. As of December 2020, 88 States have informed ICAO of their participation in the scheme starting from its pilot phase, including Australia, Canada, Japan, Qatar, Republic of Korea, Malaysia, Singapore, Thailand, Turkey, the United States, States of the European Union, the United Kingdom, and the United Arab Emirates.

8. Analysis of the supply and demand for carbon offsets for CORSIA A carbon offset is a reduction, avoidance or removal of greenhouse gas emissions in a sector that is not subject to an emissions cap.25 Quality standards are essential to ensure that emissions are actually being reduced in the other sector or location and are not “double-counted” against multiple targets. The unit of the carbon offset is 1 ton CO2eq. The supply of carbon offsets is provided by two main types of programs: regulatory compliance 24

ICAO, Understanding CORSIA Eligible Emissions Units (https://www.icao.int/ environmental-protection/CORSIA/Documents/TAB/TAB%202020/TAB_Webinar_ August_2020_CEUs.pdf). 25 International Emissions Trading Association (IETA) (https://www.ieta.org/resources/ Resources/101s/Offsets.pdf).

The burden of a ton CO2! Emission trading systems and the air transport business

43

and voluntary programs. Regulatory compliance programs or markets are created by government legislation at the national and sub-national levels or by an international agreement at the regional and international levels. Examples of compliance markets include the Kyoto Protocol’s Clean Development Mechanism (CDM) and Joint Implementation ( JI) mechanism, as well as the European Emissions Trading Scheme (or EU ETS). Voluntary markets are created by non-governmental entities (independent crediting mechanisms). They typically allow businesses and individuals to offset the emissions they create. Examples of independent crediting mechanisms include Verified Carbon Standard and Gold Standard. It is estimated that almost four billion ton CO2eq of carbon offset credits have been issued so far, more than half by the CDM.26 Future supply is in general difficult to estimate as it typically adjusts to demand and is subject to technical, economic and political uncertainties. An analysis issued in November 201927 projects the total supply of offsets up to 2035 at 18 billion ton CO2eq. These projections cover emissions reduction from four major programs (the Clean Development Mechanism, the Verified Carbon Standard, the Gold Standard and the Climate Action Reserve) over the period from 2013 to 2035. Another analysis issued in March 202028 finds that a supply of 386 million tons of CO2eq, that meets ICAO eligibility criteria, already exists and that another 183 million tons of CO2eq is in the pipeline for a total of 569 million tons of CO2eq. The study concludes that this supply is more than sufficient to cover CORSIA requirements for the pilot phase. For international air transport and in light of the continuing growth in CO2 emissions and the high abatement cost within the sector, it is expected that the air transport sector will be a net buyer of carbon offsets from other carbon markets. Based on this analysis, the carbon offset requirements for CORSIA over the period 2021e2035 are estimated to be in the range of 1000 to 2100 million tons of CO2, depending on the forecast scenario.

26

27

28

State and Trends of carbon Prices 2020 (https://openknowledge.worldbank.org/ bitstream/handle/10986/33809/9781464815867.pdf?sequence¼4&isAllowed¼y). New Climate Institute, Offset credit supply potential for CORSIA, 2019 (https:// newclimate.org/wp-content/uploads/2019/11/Offset-credit-supply-potential-for-COR SIA.pdf). Ecosystem marketplace, Carbon markets are well-positioned to meet CORSIA demand projections(https://app.hubspot.com/documents/3298623/view/69866075?accessId¼ fe65d3).

44

The Air Transportation Industry

As of December 2020, the CORSIA eligible emissions unit programs approved by ICAO Council include the following29: • American Carbon Registry (ACR) • Architecture for REDD þ Transactions • China GHG Voluntary Emission Reduction Program • Clean Development Mechanism (CDM) • Climate Action Reserve (CAR) • The Gold Standard (GS) • Verified Carbon Standard (VCS) Table 2.3 summarizes current carbon credit supply data for the programs above. It should be noted however that this data includes some elements that are not eligible for use under CORSIA. ICAO has launched a second call for applications and eight new applicants have applied for the 2020 assessment. The International Air Transport Association (IATA) has also launched the Aviation Carbon Exchange (ACE) to help airlines and other aviation stakeholders trade CORSIA eligible emission units. Table 2.3 Carbon credit supply data (Million-ton CO2 eq.). Credits issued (million-ton Credits retired CO2 eq.) or canceled

American Carbon Registry (ACR) Architecture for REDD þ Transactions China GHG Voluntary Emission Reduction Program Climate Action Reserve (CAR) Clean Development Mechanism (CDM) Gold Standard Verified Carbon Standard (VCS) Total

Balance

50

8.5

41.5

Not available

Not available

53

Not available

Not available 53

69

40

29

2002

1192

810

97 410

59 251

38 159

2681

1299.5

1130.5

Based on: The World Bank, 2020. State and Trends of Carbon Prices.

29

It should be noted however that not all units from these programs are eligible for use in CORSIA.

The burden of a ton CO2! Emission trading systems and the air transport business

45

The main purpose of carbon pricing policies is to send a signal to the investment community to invest in the carbon market. A low carbon price does not guarantee the financial viability of many carbon reduction projects and leads to low investment rates. In 2019, the average carbon price was relatively low in the range of 2e3 US$ per ton, due in particular to low demand. In order to cost-effectively reduce CO2 emissions in line with the temperature goals of the Paris agreement, carbon prices in the range of 40e80 US$ per ton in 2020 and 50e100 US$ per ton in 2030 are required.30 In addition to CORSIA, several other compliance programs are under development in the framework of the implementation of the Paris Agreement. This would likely help raise carbon prices. Under the UNFCCC, negotiations are continuing to reach an agreement on “robust accounting” rules for the implementation of the “internationally transferred mitigation outcomes” toward the NDCs. The conclusion of these negotiations may also lead to increasing the demand for carbon credits.

9. The impact of CORSIA on air traffic and airline financial results The aim of this section is to offer an analysis of the impact of CORSIA on airline results, starting with a brief description of the model used and related assumptions followed by a presentation of the model output and an analysis of the impact in terms of equivalent increase in ticket costs. 9.1 Assumptions and brief model description For now, CORSIA is set to extend from 2021 to 2035. Depending on the evolution toward the goals and the progress achieved in the implementation of other measures, the scheme may be extended beyond that period. Therefore, this analysis covers the period 2021e2050 and is based on three air traffic forecast scenarios, projections of aircraft fuel efficiency improvement and carbon price as well as assumptions on the availability and price of sustainable aviation fuels. Two alternative options are considered on whether airlines will pass on the cost to passengers and freight shippers or 30

High-Level Commission on Carbon Pricing and Competitiveness, Report of the High Level Commission on Carbon Prices (2017) (https://www.carbonpricingleadership.org/ s/CarbonPricing_FullReport.pdf).

46

The Air Transportation Industry

Table 2.4 Main input and assumptions. 2020e2030

2030e2040

2040e2050

5.4% 4.7% 5.3%

5.3% 4.6% 5.2%

5.2% 4.5% 5.1%

4.4% 3.7% 4.3%

4.3% 3.6% 4.2%

4.2% 3.5% 4.1%

3.4% 2.7% 3.3%

3.3% 2.6% 3.2%

3.2% 2.5% 3.1%

High scenario

RPKs FTKs RTKs Central scenario

RPKs FTKs RTKs Low scenario

RPKs FTKs RTKs

Average annual aircraft fuel efficiency improvement (%)

1.5%

1.0%

0.75%

30

40

50

3% 20%

5% 20%

10% 20%

Carbon price (US$)

Sustainable aviation fuels

Availability as % of total Price differential

no (100% cost pass-through and 0% cost pass-through). The analysis covers the total global international traffic. Table 2.4 summarizes the main assumptions. The model used evaluates the operational and financial impact of CORSIA by comparing a case where the scheme is implemented (MBM case) to a case where it is not (Base case). Under each case, a set of operational and economic relationships determine the traffic (both passenger and cargo), the operating revenues and the operating costs. Since operators have a choice between using sustainable aviation fuels and/or buying emission units, the model assesses both options and selects the cheapest. In the case of a cost pass-through, the additional costs are added to the fares

The burden of a ton CO2! Emission trading systems and the air transport business

47

(and cargo rates) leading to a reduction in traffic. Operators are assumed to match this reduction by reducing capacity. This, in turn, leads to a reduction in cost. 9.2 Results The results of the analysis of the impact of CORSIA on air traffic and airline financial performance for the central, high and low scenarios, are presented in Tables 2.5e2.7. 9.3 Analysis The analysis shows that under the Central scenario, the cost of CORSIA as a percent of airlines operating revenues for the period 2021e2035 is in the range of 0.8%. This represents around 1.6 US$ increase per ticket for a trip of an average distance of 3000 km and a 5.2 US$ increase for an average distance of 10,000 km. Considering the High and Low scenarios, those values would range from 0.5 to 1%, from 1 US$ to 2.1 US$ and from 3.5 US$ to 6.9 US$ respectively. The pandemic will lead to a reduction in future traffic in the long run and the impact of CORSIA will also be reduced in volume as the level of offsetting requirements will be lower. As a result, the cost of CORSIA as a percent of airlines operating revenues for the period 2021e2035 would be in the range of 0.6% while the corresponding cost per ticket would be in the range of 1.2 US$ for a trip of an average distance of 3000 km and a 3.9 US$ increase for an average distance of 10,000 km. Similarly, considering the High and Low scenarios, those values would range from 0.4 to 0.7%, from 0.8 US$ to 1.5 US$ and from 2.7 US$ to 5.1 US$ respectively. If the period is extended to 2021e2050 in the absence of the pandemic, the cost of CORSIA as a percent of airlines operating revenues would be in the range of 1.5% and the equivalent increase in cost per ticket would be in the range of 3.2 US$ for a trip of 3000 km and 10.6 US$ for a trip of 10,000 km. Under the pandemic, the cost of CORSIA would decline to 0.6 per of operating revenues while the increased cost per ticket would be in the range of 3.0 US$ for a trip of an average distance of 3000 km and a 9.9 US$ increase for an average distance of 10,000 km. In the case of a 100% cost pass-through, the airlines will not recover all costs. While the costs will be passed on to passengers and freight shippers through fare and rates increases, traffic will decline leaving a gap that airlines will still incur despite adjusting capacity to meet the demand.

48

Table 2.5 Results for the central scenario. No pandemic impact 2021e2050

2021e2035

2021e2050

No cost passthrough

100% cost passthrough

No cost passthrough

100% cost passthrough

No cost passthrough

100% cost passthrough

No cost passthrough

100% cost passthrough

2386

2355

10,996

10,827

1496

1476

8556

8421

86.6

85.5

497.2

489.4

46.9

46.2

336.7

330.5

0.0

0.0

0.0

0.0

59.5

59.5

374.4

374.4

0.0

0.0

0.0

0.0

7.5

7.5

55.0

55.0

0.0

0.0

0.0

0.0

40.4

40.4

269.5

269.5

0.0%

0.3%

0.0%

0.6%

0.002%

0.22%

0.002%

0.57%

0.6

0.2

1.2

0.4

0.5

0.2

1.3

0.5

Central scenario

Cumulative offset requirements (million tons) Cost of carbon (billion US$) Quantity of sustainable aviation fuels (million tons) Savings due to sustainable aviation fuels (billion US$) Cost of sustainable aviation fuels (billion US$) Change in air traffic (billion RTK) Change in operating margin (percentage points)

The Air Transportation Industry

2021e2035

With pandemic impact a

a

0.8%

0.4%

1.5%

0.8%

0.6%

0.3%

1.4%

0.8%

1.6

1.6

3.2

3.2

1.2

1.2

3.0

3.0

5.2

5.2

10.6

10.6

3.9

3.9

9.9

9.9

It should be recalled the CORSIA has been adopted for the period 2021e2035, only.

The burden of a ton CO2! Emission trading systems and the air transport business

Cost as proportion of revenue (%) Average cost per ticket (3000 km) (US$) Average cost per ticket (10,000 km) (US$)

49

Table 2.6 Results for the high scenario. With pandemic impact

2021e2050

2021e2035

2021e2050

No cost passthrough

100% cost passthrough

No cost passthrough

100% cost passthrough

No cost passthrough

100% cost passthrough

No cost passthrough

100% cost passthrough

3436

3390

16,607

16,351

2053

2026

12,490

12,291

124.9

123.2

753.9

742.1

66.7

65.7

507.7

498.5

0.0

0.0

0.0

0.0

65.5

65.5

451.0

451.0

0.0

0.0

0.0

0.0

8.3

8.3

66.6

66.6

0.0

0.0

0.0

0.0

44.5

44.5

325.2

325.2

0.0%

0.4%

0.0%

0.8%

0.002%

0.28%

0.002%

0.72%

0.7

0.2

1.5

0.5

0.6

0.2

1.6

0.6

1.0%

0.5%

1.9%

1.0%

0.7%

0.4%

1.7%

1.0%

2.1

2.1

4.0

4.0

1.5

1.5

3.7

3.7

6.9

6.9

13.3

13.3

5.1

5.1

12.5

12.5

High scenario

Cumulative offset requirements (million tons) Cost of carbon (billion US$) Quantity of sustainable aviation fuels (million tons) Savings due to sustainable aviation fuels (billion US$) Cost of sustainable aviation fuels (billion US$) Change in air traffic (billion RTK) Change in operating margin (percentage points) Cost as proportion of revenue (%) Average cost per ticket (3000 km) (US$) Average cost per ticket (10,000 km) (US$)

The Air Transportation Industry

2021e2035

50

No pandemic impact

Table 2.7 Results for the low scenario. No pandemic impact 2021e2035

With pandemic impact

2021e2050

2021e2035

2021e2050

100% cost passthrough

No cost passthrough

100% cost passthrough

No cost passthrough

100% cost passthrough

No cost passthrough

100% cost passthrough

1437

1418

6436

6337

977

964

5265

5182

52.1

51.4

290.2

285.7

28.6

28.1

194.5

190.7

0.0

0.0

0.0

0.0

54.0

54.0

311.2

311.2

0.0

0.0

0.0

0.0

6.8

6.8

45.5

45.5

0.0

0.0

0.0

0.0

36.7

36.7

223.6

223.6

0.0%

0.2%

0.0%

0.4%

0.002%

0.15%

0.002%

0.4%

0.4

0.1

0.9

0.3

0.3

0.1

0.9

0.3

0.5%

0.3%

1.1%

0.6%

0.4%

0.2%

1.0%

0.5%

1.0

1.0

2.2

2.2

0.8

0.8

2.1

2.1

3.5

3.5

7.5

7.5

2.7

2.7

6.9

6.9

Low scenario

51

Cumulative offset requirements (million tons) Cost of carbon (billion US$) Quantity of sustainable aviation fuels (million tons) Savings due to sustainable aviation fuels (billion US$) Cost of sustainable aviation fuels (billion US$) Change in air traffic (billion RTK) Change in operating margin (percentage points) Cost as proportion of revenue (%) Average cost per ticket (3000 km) (US$) Average cost per ticket (10,000 km) (US$)

The burden of a ton CO2! Emission trading systems and the air transport business

No cost passthrough

52

The Air Transportation Industry

The results above demonstrate that the impact of CORSIA would be manageable, under the assumptions made and in particular those related to carbon price.

10. Conclusion While its share in CO2 emissions is relatively small, the air transport industry has been very active in seeking ways to reduce it by setting ambitious goals and developing strategies. This chapter has presented the challenge posed by these emissions along with the policies and measures adopted by the sector to address it, including new technologies, new standards, enhanced airline operations, improved air navigation service provider’ operations, sustainable aviation fuels and market based measures. Since the cost of in-sector reductions remains high and most potential technical solutions either are many years away or would be restricted to certain markets, the industry is relying on two near-term solutions: sustainable aviation fuels and offsetting. Both come at a cost that is still manageable as demonstrated in this chapter. Although a number of previous studies have reached a similar conclusion, this chapter takes into consideration the most recent developments in the sector including the impact of the COVID-19 pandemic. While CORSIA is now on track, Government support remains key to achieving the goals set for the sector in particular through the promotion of research and innovation and through the setting of policies that favor the development and commercialization of sustainable aviation fuels. Cooperation between stakeholders within the air transport industry and with energy producers is also critical to meeting the goals.

References ATAG, 2020a. Tracking Aviation Efficiency, Fact Sheet # 3. ATAG, 2020b. Waypoint 2050. High-Level Commission on Carbon Pricing and Competitiveness, 2017. Report of the High Level Commission on Carbon Prices. ICAO, 2019. Annual Report of the Council 2019. IPCC, 2018. Summary for policymakers. In: Masson-Delmotte, V., Zhai, P., Pörtner, H.O., Roberts, D., Skea, J., Shukla, P.R., Pirani, A., Moufouma-Okia, W., Péan, C., Pidcock, R., Connors, S., Matthews, J.B.R., Chen, Y., Zhou, X., Gomis, M.I., Lonnoy, E., Maycock, T., Tignor, M., Waterfield, T. (Eds.), Global Warming of 1.5 C. An IPCC Special Report on the Impacts of Global Warming of 1.5 C above Preindustrial Levels and Related Global Greenhouse Gas Emission Pathways, in the

The burden of a ton CO2! Emission trading systems and the air transport business

53

Context of Strengthening the Global Response to the Threat of Climate Change, Sustainable Development, and Efforts to Eradicate Poverty. World Meteorological Organization. Geneva, Switzerland, 32 pp. available in:SR15_SPM_version_stand_alone_LR.pdf (ipcc.ch). The World Bank, 2020. State and Trends of Carbon Prices.

Further reading Air Transport Action Group, ATAG, 2020. Adding Value to the Economy. Ecosystem marketplace, 2020. Carbon Markets Are Well Positioned to Meet CORSIA Demand Projections. Future Earth, 2020. Global Carbon Project. IATA, 2020. IATA Technology Roadmap. ICAO, 2020a. ICAO Council Agrees to the Safeguard Adjustment for CORSIA in Light of COVID19 Pandemic. https://www.icao.int/Newsroom/Pages/ICAO-Councilagrees-to-the-safeguard-adjustment-for-CORSIA-in-light-of-COVID19-pandemic. aspx. ICAO, 2020b. Understanding CORSIA Eligible Emissions Units. International Emissions Trading Association (IETA), 2019. Offsets: The Basics. NASA, 2012. Technology Readiness Level. New Climate Institute, 2019. Offset Credit Supply Potential for CORSIA. OECD, 2018. The Long View: Scenarios for the World Economy to 2060. PricewaterhouseCoopers, 2017. The World in 2050. UNFCCC, 1992. The United Nations Framework Convention on Climate Change (UNFCCC). UNFCCC, 2015. The Paris Agreement. World Meteorological Organization, 2020. The State of the Global Climate 2020. World Resources Institute, 2020. Climate Watch Data, GHG Emission Module.

This page intentionally left blank

CHAPTER 3

Labor in the aviation industry: wages, disputes, and shocks Heather McLaughlin1 and Colm Fearon2 1 2

De Montfort University, Leicester, England; Business School, University of Birmingham, Birmingham, England

1. Introduction The aviation industry employs a diverse range of people with different characteristics and skill sets. At a macro level, the labor market is dependent on the market for the air transport services provided. But at a more granular level, the market is segregated according to job type, geographical location, and company, forming different markets with different characteristics which determine the wage that will be offered. It is a sector that has undergone substantial change over the decades, moving from strongly national structures to more deregulated competitive global operations. New business models have been built on leaner and agile cost structures, achieved partly through more creative employment practices. This has impacted the labor markets and the relative power base between employer and employee. Economic shocks have also played their part in this evolution, with aviation usually a major casualty of economic downturns. Over the last year, COVID-19 has seriously curtailed worldwide travel, causing a 43% reduction in direct aviation jobs (ATAG, 2020). Although we expect a recovery post pandemic, there are some indications of a more fundamental change in international travel patterns, particularly the business sector. This chapter examines labor in the aviation industry from an economics perspective and uses this lens to assess the impact of trade unions as well as economic shocks on the labor market. More specifically, it asks: What are the characteristics of the labor market? Is there a model which describes how wages are determined? How have wage conflicts affected the sector? What does the future hold for employment in the aviation industry in the post COVID-19 era? In answering these important questions, we highlight the significance of demand and supply issues and their sensitivity to wages (elasticity) in order to present a conceptual argument around the model of wage determination predominantly in the airline sector of the industry. The Air Transportation Industry ISBN 978-0-323-91522-9 https://doi.org/10.1016/B978-0-323-91522-9.00013-0

© 2022 Elsevier Inc. All rights reserved.

55

56

The Air Transportation Industry

The chapter is structured as follows: Section 2 examines the various types of labor in aviation and their characteristics. Wage determination is explored in Section 3, with the impact of monopsony and monopoly power presented in Sections 4 and 5. Section 6 discusses the bargaining power of employers and employees based on sensitivity to wage rates and the increasing levels of competition in the industry itself. Section 7 presents some recent evidence of industrial action to highlight the disruption that it can cause, and Section 8 discusses the serious impact of economic shocks and the COVID-19 pandemic in particular. Conclusions are drawn in Section 9, which also considers what the future holds for labor in this turbulent industry.

2. Employment in the aviation industry Air transport is a global industry, with some 1478 commercial airlines operating a total fleet of 33,299 commercial aircraft (ATAG, 2020). Although it represents a relatively small share of GDP, it supports a number of allied industries such as airports and aircraft manufacturing which are collectively known as the “aviation industry” (OECD.org, 2020). Table 3.1 shows that the global aviation industry (pre-COVID-19) supported some 87.7 million jobs distributed across all regions, with AsiaPacific accounting for the largest, number followed by Europe (ATAG, 2020). This total comprises direct, indirect, induced, and aviation-enabled tourism. Direct employment accounts for around 11.3 million; indirect which is generated from the purchase of goods and services of the companies in the industry is 18.1 million; induced from spending of industry Table 3.1 Estimated jobs in aviation by region. Region

Estimated jobs (millions)

Africa Asia-Pacific Europe Latin America and Caribbean Middle East North America TOTAL

7.7 46.7 13.5 7.6 3.3 8.8 87.7

Source: Air Transport Action Group, September 2020. Powering Global Economic Growth, Employment, Trade Links, Tourism, and Support for Sustainable Development through Air Transport, Despite Global Crisis. (Online) Available at: https://www.atag.org/our-publications/latest-publications.html.

Labor in the aviation industry: wages, disputes, and shocks

Table 3.2 Direct employment. Number of jobs (millions) Job type

0.648 5.5

Airport operators Other airport

3.6

Airlines

1.3 0.237

Civil aerospace Air navigation service providers

57

Description

Airport retail, car rental, customs, freight forwarders Pilots, cabin crew, executives, ground service, check in staff, training Engineers, designers

Source: Air Transport Action Group, September 2020. Powering Global Economic Growth, Employment, Trade Links, Tourism, and Support for Sustainable Development through Air Transport, Despite Global Crisis. (Online) Available at: https://www.atag.org/our-publications/latest-publications.html.

employees is estimated at 13.5 million; and aviation enabled tourism accounts for around 44.8 million jobs. The ATAG (2020) report further categorizes direct employment as shown in Table 3.2. Aviation is noted for its high rates of pay. The average salary for those in direct employment at European airports in 2016 was $45,310 (40.400 Euro) which is higher than the average in the overall economy. This was found to be the case in all countries. This “above average” salary is a reflection of the high level of skill required in many of these jobs. Data for airline salaries in the United States show an average of $89,000 per person in 2016, which again is significantly in excess of the average national private sector rate of $59,000 (ATAG, 2018, p. 29). Each type of job effectively represents a different market with potentially differing characteristics. It is these differences which segment the market and allow wage differentials to develop. More specifically, we can say that (Lipsey, 1992): a. Wages vary with the type of job. For example, flight attendants earn less than pilots. b. Wages vary with education and training. For example, pilots have to undergo rigorous training in order to obtain a license to fly. They will invest in that training on the basis that their average earnings will exceed those of workers which require a lower level of training. c. Wages vary with levels of experience. d. Wages vary according to the market in which labor is sold. For example, wages are often higher in heavily unionized markets. e. Wages sometimes may vary according to gender or race. As awareness of these gaps have increased, measures are being taken to address them.

58

The Air Transportation Industry

The following section examines how wages are determined and the part played by these characteristics and market imperfections.

3. Wage determination Labor markets bring together demanders (employers) and suppliers (employees) to determine the price of labor which is the wage rate (Lipsey, 1992). There are numerous models which describe the way in which markets operate, emanating from the levels of competition and the relative power of the individual actors (Lipsey, 1992). Booth (2014) observes that, even as recently as 25 years ago, many economists viewed labor markets as intrinsically competitive, consisting of many buyers and sellers who could not individually influence the wage rate. The assumptions of this perfectly competitive model also include no barriers to entry or exit, perfect factor mobility, perfect information and no transaction costs. While these assumptions can easily be challenged for labor markets, the perfect competition model provides a good starting point for a discussion on wage determination. In a perfectly competitive market, there are many buyers and sellers such that no individual has the power to influence the price. Both demanders and suppliers of labor are price (wage) takers and will adjust the level of employment based on this market determined wage. In Fig. 3.1, the supply curve is upward sloping because at higher wage rates, more workers are willing to supply their labor. The demand curve is downward sloping, indicating that firms will employ more labor as wage rates fall. The wage rate is established where supply and demand intersect at the equilibrium E, where the wage rate is we and the quality of labor is qe.

Figure 3.1 Wage determination in a competitive market.

Labor in the aviation industry: wages, disputes, and shocks

59

If disequilibrium occurs, there would be shifts in employment from lower income to higher income jobs until the differentials were reduced and the equilibrium restored. Wages in a competitive environment will thus tend toward the “opportunity wage”, i.e., what could be earned in similar jobs in alternative sectors because there is perfect mobility of labor between sectors (op cit. Lipsey). Mobility is far from perfect. Short term movement can be affected by the search for employment and, in the medium to longer term, more significant barriers are created by qualifications, skills, and experience. A pilot, for example, has to complete a rigorous program of training which takes around 2 years and costs around $80,0001 (prices in 2021) and thus time and money can prevent entry. Availability of training places can further exacerbate the situation as competition for places on these programs is fierce. Mobility may therefore only be possible within a particular area of specialization. In the aviation industry, more highly skilled workers such as pilots will earn more than flight attendants, and engineers will earn more than general ground staff but investment in training is required to achieve those higher earnings. Mobility in these more highly skilled professions is restricted in the short term because of the need to invest money and time in qualifications and experience, and wages will be determined by supply and demand within these non-competing groups. This investment can sometimes lead to scarcity of labor which shifts the supply curve to the left (Supply *) and sets the wage at a higher rate w* (see Fig. 3.1). Longer term, mobility is easier but many established airlines have used the “seniority rules” which afford pilots roster and pay privileges as a disincentive to move to other airlines (Capelli, 1985).

4. Monopsony (the power of the employers) In some markets, the supply of labor is competitive (i.e., there are lot of suppliers) but buyers are few in number. Where there is only one buyer, we have a situation of monopsonistic demand, where the buyer sets the rate and the suppliers decide whether they want to work at that rate or move into other areas of activity. Employment and wages will be lower than under perfectly competitive conditions. There is evidence that imperfections in the labor market are more prevalent where the market for product 1

According to https://atpflightschool.com/become-a-pilot/flight-training/pilot-trainingcost.html.

60

The Air Transportation Industry

or services is also imperfect (Stewart, 1990; Booth, 2014). This was certainly the case in the aviation industry pre-deregulation which was dominated by the so-called legacy carriers. In the United States, pre-1978, air travel was regulated by the Federal Civil Aeronautics Board (CAB). CAB controlled route allocation, pricing and areas served. It was a model which required considerable cross subsidy, where affordable services on unprofitable routes were financed by the more profitable high-traffic routes (Wolla, 2018). In Europe and Asia, the market was dominated by national carriers which had an effective monopoly in that country, and where the airline’s operations and employees were primarily based in that home country. In this scenario, there are a small number of buyers of labor who have the power to influence the wage rate and workers must choose whether or not to work at that wage or move to another market. In Fig. 3.2, the supply curve slopes upward indicating that more labor will be supplied at higher rates of pay and reflects the average cost to the buyer of each unit of labor. If an additional worker is employed, the marginal cost of employing that additional worker will exceed the wage currently paid in order to attract more labor. The new rate then has to be paid to all those workers. The marginal cost curve shown in Fig. 3.2 is above that of the supply. The monopsonist equates the marginal cost of employment with the marginal revenue product curve. Recruitment will continue until the last unit recruited increases total costs as much as it increases total revenue and so the equilibrium is established at Em where the wage is wm and the quantity of labor is qm. Therefore, monopsonistic demand for labor results in a lower wage rate and a lower level of employment than under perfectly competitive conditions.

Figure 3.2 Wage rates with monopsonistic demand.

Labor in the aviation industry: wages, disputes, and shocks

61

5. Monopoly (the power of the unions) The previous analysis assumes that there are many suppliers of labor. However, the presence of unions changes this model by forming a collective which can act as an effective monopoly of supply. But how significant are the unions? Table 3.3 shows that union density in nearly all OECD countries have fallen over the last 18 years. That said, union density in the airline industry is significantly higher than other sectors. In the United States in 2005 for example, more than 49% of the airline industry were unionized compared to just under 8% for the work force as a whole. These union contracts provided benefits which were higher than those seen in other sectors of the economy. They were particularly sizable for pilots but also generous for other unionized air transport employees (Hirsch, 2007).

Table 3.3 Union density for selected countries (where data available for 2000 and 2018). Country Source 2000 2018

Australia Belgium Canada Czech Republic Denmark Finland France Germany Hungary Iceland Italy Japan Mexico Netherlands Norway Spain Sweden Turkey United Kingdom United States

Survey data Survey data

Administrative data Administrative data Administrative data Administrative data Administrative data Survey data Administrative data Administrative data Survey data

24.9 56.2 28.2 27.2 74.5 74.3 9.5 24.6 23.8 89.4 34.4 21.5 16.9 22.6 52.4 17.4 79 16 29.7 12.9

13.7 50.3 25.9 11.5 66.5 60.3 8.8 16.5 7.9 91.8 34.4 17 12 16.4 49.2 13.6 64.9 9.2 23.4 10.1

Source: Oecd.org., 2020. COVID-19 and the Aviation Industry: Impact and Policy Responses. (Online) Available at: www.oecd.org/coronavirus/policy-responses/covid-19-and-the-aviation-industry-impactand-policy-responses-26d521c1/. (Accessed 14 December 2020).

62

The Air Transportation Industry

It is important to note that union density and union coverage are not the same thing. In many cases, firms adopt a collective contract which covers all employees, not just those who are members of the union. Coverage or reach of the union is perhaps the more important determinant (Fitzenberger et al., 2013). This also rings true when considering the union density in France which is only around 9% and yet unions are notoriously effective in mounting large scale industrial action, notably the air traffic controllers! In the aviation industry, unions operate differently from country to country. In some countries, there may be a single union for all workers of a company, and in others, they may be more profession-specific, or a combination of both. This issue is further discussed in Section 6.3, in relation to the global nature of the industry. Unions representing or covering the suppliers of labor can drive up the wage rate beyond its equilibrium level and establish a minimum wage below which no one will work. In Fig. 3.3, the union negotiates the wage rate at wu, higher than the equilibrium rate of we. The supply curve becomes elastic up to the quantity willing to work at the wage set, as the supply cannot vary with price. The amount of labor willing to work at wu is q*, but actually only qu will be employed based on the employer’s demand curve. Thus, lower employment and higher wages are established than would be the case under competitive conditions. The union therefore create an excess supply which puts pressure on wage reduction if more workers are going to be employed. Employment does not always fall with an increase in wages as can be seen in Section 6 which explores the concept of elasticity in more detail.

Figure 3.3 Wage rates with union influence (single union with many buyers).

Labor in the aviation industry: wages, disputes, and shocks

63

6. Bargaining power The previous two sections have outlined the models of monopsony and monopoly and their impact on wages and employment. These market imperfections have implication for the bargaining power of the employers and employees from an economic perspective. We have seen that elasticity of demand and supply is important to the analysis, but what determines theses elasticities? Another important economic determinant is the market for the service itself, namely air transport. There is evidence to suggest that there is a correlation between the imperfections in the product/service market and the labor markets (Stewart, 1990). These issues are explored in the context of the aviation industry, in order to help explain the impact of collective bargaining. 6.1 Elasticity of demand The elasticity of demand for labor depends on the proportion of total costs represented by labor, and, the extent to which labor may be substituted. The lower the proportion of total cost that labor represents, the more inelastic the demand. In other words, if the cost of labor is a relatively small proportion of total costs, then a change in the wage rate will not significantly affect the level of employment. Similarly, if there are no real substitutes for labor, then a change in the wage rate will not lead to a significant change in the labor demanded because there are no alternatives. If there is inelastic demand for labor, then a higher wage rate may not lead to a reduction in employment. The unions would be in a relatively strong position in trying to push up the wage rate in that the pressure for a work force reduction would be much less. In Fig. 3.4, demand for labor is perfectly inelastic, in other words, the amount qe is demanded regardless of the wage rate. In this case a wage rise to w* will not lead to a fall in the quantity of labor qe although more people will want to work at that rate as indicated by the supply curve. If demand is increased and the curve shifts to the right (Demand 1), the wage rate could remain at w* up to employment of q*. Conversely, with a very elastic demand for labor (i.e., a more horizontal demand curve), a small increase in the wage rate would cause a larger reduction in employment. In the airline industry, labor has traditionally constituted a high proportion of the total cost base which makes demand more elastic. A small change in the wage rate will cause a larger change in demand. However, with increasing competition and different methods of employment, labor

64

The Air Transportation Industry

Figure 3.4 Wage determination with inelastic demand.

costs have fallen proportionately over recent years in Asia, Europe, and the United States, with Low Cost Carriers (LCCs) leading the way. For example, in 2010, labor costs as a percentage of revenue was around 25% of revenue for British Airways (BA) and 12% for Ryanair. By 2020, the percentage for BA had fallen to 19%, with Ryanair at 11% (British Airways, IAG and Ryanair Annual Reports, 2010 and 2020). Availability of substitutes for the labor depends on the job itself and the scarcity of labor able to do that job. If there is a shortage of supply, then a large wage increase may only attract a few more workers. Airline pilots are not easily substituted because of the skills required including the number of flying hours needed to obtain a commercial pilot’s license. The global supply of newly qualified pilots has not increased in line with growth in demand in recent years. It has been estimated that the industry will need average annual growth of more than 4% per annum in the numbers of active commercial airline pilots in the decade up to 2027.2 US data (from the Federal Aviation Administration) and UK figures (from the Civil Aviation Authority) indicate average annual increases of only 1% in pilot licenses over the past 10 years. The scarcity of pilots in the Asia Pacific and the Middle East regions, has led to higher pilot pay in those regions. It is interesting to note that some experienced pilots from Europe have taken up better remunerated jobs in the Middle East and Asia Pacific. It will be interesting to see if this excess demand for pilots persists post-pandemic, as airlines rebuild their businesses.

2

According to CAPA analysis of forecasts by the flight training provider CAE.

Labor in the aviation industry: wages, disputes, and shocks

65

Over recent years, both the falling proportion of labor costs and the scarcity of certain types of labor, notably pilots, have created a tendency toward a more inelastic demand. However, it is likely that supply may outstrip demand post-pandemic (see Section 8). 6.2 Elasticity of supply The sensitivity of the supply of labor to the wage rate determines to the slope of the supply curve. Manning (2013) observes that: the most direct way to establish the existence of employer market power over its workers is to estimate the wage elasticity of the labor supply curve facing the firm (p.80). Studies of elasticity of supply are few, but the evidence suggests that, where there is monopsony or oligopsony power, elasticity is low. A decrease in the wage rate would not induce labor to move elsewhere. Increasing competition increases the elasticity of supply, such that a decrease in the wage rate would encourage movements of labor to alternative employment. Firms can offer lower wages without immediately losing workers. In a monopsonistic labor market, the existence of little alternative employment makes it difficult for the employees to go elsewhere, regardless of the wage rate, and gives power to the employer. In this case, the supply of labor is inelastic and a decrease in the wage rate would not induce any labor to move to alternative employment (see Fig. 3.5). Conversely, if the supply of labor is elastic, a small decrease will cause considerable movement of labor.

Figure 3.5 Inelastic supply of labor.

66

The Air Transportation Industry

Elasticity of supply is therefore an important consideration for the employer. Pre-deregulation, the supply of labor was relatively inelastic because of the organization of the large parts of the industry into segregated national carriers. Many countries had their own airline which tended to employ their own nationals. National regulations and requirements segregated these markets, making it difficult to move from one to another, whatever the wage. The resulting inelasticity of supply of labor meant that in the period before deregulation, airlines offered high wages when times were good, and lower wages when times were bad. In the latter case, demand shifted to Demand* (as in Fig. 3.5), but employment remained unchanged. The situation was very different during the era of deregulation and increasing competition (see Section 6.3). However, during the pandemic, there has been evidence of wage reduction to retain jobs in a number of airlines (see Table 3.5). 6.3 Competition in the aviation market itself We observe a number of changes over the last few decades which have dramatically affected the market model of the global industry. Deregulation in both the United States and Europe increased the level of competition. Airlines had to find new ways of competing either in terms of their service provision or in terms of their cost base. The latter became very significant in the future of aviation and its labor. The low-cost model was pioneered by Southwest airlines in the US domestic market and is often held out as the epitome of good practice in employee relations (Southwest.com, 2020). It remains the most highly unionized airline but takes seriously the importance of high-quality staff for customer service. Working in partnership with the union, the airline has only experienced one strike in its 50-year history. This model was adopted and substantially adapted by the LCCs in other parts of the world, which arguably did not share the same values as Southwest. LCCs were largely catering for the economy market of leisure travel where demand was very price sensitive and seasonal, with higher demand in summer and lower demand in winter. Airlines such as Easyjet and Ryanair started as niche short haul providers but became dominant players. The new business model changed employment practices to create a low labor cost and high labor productivity model (Gittell et al., 2015). Seasonality required flexibility which could be achieved by employing pilots and cabin crew through an

Labor in the aviation industry: wages, disputes, and shocks

67

“agency” (which pays workers only for the hours worked) or on short term contracts. This outsourcing of crew to third parties further converted labor into a more variable cost and disempowered the workers in terms of union representation, shifting power to the employer. That said, competition between Ryanair and others for pilots is not based solely on gross or net wages. The “5/4” rosters offered by Ryanair, that is, 5 days working followed by 4 days off meant that pilots could calculate months (even years) in advance exactly when they had to work, and proved to be very attractive. The unbundling of the air travel experience with additional charging for seats, priority boarding and luggage also introduced an “inflight performance” component to pay (Harvey and Turnbull, 2020). Internationalization and open skies agreements have further increased the competitive structure of the global marketplace and added a new layer of complexity with companies taking advantage of different jurisdictions. Despite the single market in Europe, labor laws remain fragmented relating to national governments of member states. It was therefore possible to employ staff in jurisdictions which offered more favorable employment law and potential for lower wages. Ryanair began using Irish employment law for employees across Europe, regardless of their nationality. Norwegian Air International, a subsidiary of Norwegian Air Shuttle also obtained an Irish Air Operator’s Certificate (AOC) which enabled them to employ workers on Irish contracts in order to take advantage of less onerous terms and conditions than in Norway. Such moves have been criticized as “flags of convenience,” having no genuine link to the state of registration (Harvey and Turnbull, 2015). So, what does all this mean for the relative bargaining power of the employer and employee in the aviation sector? The market for air transport is becoming increasingly competitive which would suggest an increase in the elasticity of supply of labor, as worker are able to explore outside options. We are also seeing a lowering of the proportion of wages in total costs particularly in the case of the LCCs which decreases the elasticity of demand for labor, as does the scarcity of some types of labor. The combination of a more elastic supply of labor and a more inelastic demand for labor should theoretically increase the power of the unions in a negotiating situation. However, in recent years unions have been relatively ineffective against the rise of the new LCCs and also the larger global alliances. This can be partly explained by the different employment practices, outsourcing to third parties and using different jurisdictions. However, it has also been argued that unions are still organized to tackle the national flag airlines of

68

The Air Transportation Industry

yesteryear rather than the new reality (op cit. Harvey and Turnbull). The International Transport Workers’ Federation (ITF) boasts “a global reach in over 135 countries with 250 affiliated aviation workers organisations” (Itfglobal.Org, 2020). There are more transport unions that are not affiliated to the ITF. It is a similar picture for pilots. For example, the European Cockpit Association (ECA) as an umbrella association of EU national pilot associations has national member associations in 33 countries representing more than 40,000 individual pilots (Eurocockpit.be, 2020). Each of these will negotiate at their own national level. The plethora of different unions operating at national level have limited impact or power in an industry which is increasingly organized across international lines. Harvey and Turnbull (2015) suggest that a supranational approach is required in order to properly address the changing business model of LCCs and the threat of “flags of convenience”. Only this approach will give the union that monopoly power that is required as part of collective bargaining.

7. Industrial action The previous section examined the theoretical aspects of bargaining based on the model of the labor market, the underlying market for the service and the elasticities of demand and supply. If the unions are able to form a strong collective voice, they can be very powerful in dealing with employers on issues such as wages and terms and conditions of employment. In reality, there will be a collective bargaining process between the unions and management which will determine the actual wage rate. Where there is a situation of monopoly power of the union and monopsony power of the employer, there is more than one economic outcome and the final agreement will depend on the collective bargaining process in which a whole range of factors may come into playdpsychological, political, and cultural (Kochan, 1980). In Fig. 3.6, the current level of employment is q. The supply curve indicates that that number would work for wage w but the demand curve suggests that the company would go up to w*. The wage will be set between w and w* depending on the collective bargaining. Both sides would prefer the wage to be somewhere in this range rather than have no agreement. Employers will consider how probable strike action may be in deciding on the final position. Although a higher wage rate will increase costs, a strike could have a considerably greater impact through loss of

Labor in the aviation industry: wages, disputes, and shocks

69

Figure 3.6 Collective bargaining (single union and single buyer).

revenue and customer loyalty. The union will consider how employers are likely to react in determining how far they will go. A strike will mean loss of wages but this may be a price worth paying if the future wages and terms and or terms and conditions are improved. Strikes lead to lost revenue for the airlines through canceled flights. Most airlines refund the tickets or help passengers with rearrangement of travel plans. There are also compensation claims for disruption, not to mention the reputational harm that a strike causes.3 For workers, a strike results in loss of pay for the duration of the strike action. Withdrawal of labor will only take place if it is likely to result in higher pay, or better terms and conditions, or both. Both sides will be keen to avoid this situation but there are numerous examples where negotiations have broken down and a strike ensues. Table 3.4 shows examples of strike actions over the last 2 years, with an indication of the cost to the airline. As can be seen, the strike action caused numerous flight cancellations, with disruption to many thousands of passengers, and significant costs to the airlines through lost revenue and compensation. It is interesting to note that the frequency of strikes since deregulation in the US is much lower than that prior to deregulation, despite some significant industry downturns (Gittell et al., 2015). Strike action also creates a cost for the airports through reductions in ground

3

Article 5(3) of the directive EU261/2004 states that an operating air carrier is not obliged to pay compensation only if it can prove that the cancellation was caused by extraordinary circumstances which could not have been avoided even if all reasonable measures had been taken. Strikes appear in the list of extraordinary circumstances.

70

Table 3.4 Examples of strikes and their impact 2018 and 2019.

Lufthansa TAP Air Portugal Vueling Ryanair

Air France

2018

Impact

Disruption 90,000 passengers 55,000 passengers 37,000 passengers 287,000 passengers

Cost

1.2 million passengers 15 days of strike

335 m Euro (FT 19/ 02/19)

Strike

2019

Impact

British Airway strike over pay

Disruption 2325 flights

Cost V137 m (FT 26/09/19)

Alitalia national transport strike Iberia Lufthansa 2- day strike over pay and conditions

Norwegian Air pilots in Norway, Sweden and Denmark strike over concerns about operations being moved to the UK Ryanair action from pilots and cabin crew in the UK, Belgium, Netherlands, Italy, Spain and Portugal. Scandinavian Air Services (SAS) had a 6-day pilot strike in April

Source: skycop.com.

Hundreds of flights 40% increase in canceled flights 1300 flights 180,000 passengers affected

Customers experienced increased delays but overall cancellations for 2019 were 70% fewer than in 2018 4015 flights were canceled and 360,000 passengers affected.

Lufthansa paying for train tickets to travel to other airports and travel on other planes

Ryanair employed strikebreaking crews from other countries, so the strike had little impact on customer service. Potential compensation could cost up to V90 million

The Air Transportation Industry

Strike

Labor in the aviation industry: wages, disputes, and shocks

71

handling, retail, and other services, as well as impacting on local businesses that supply airports, and revenue from tourism (De Langhe et al., 2013). These are not represented in the table. The table shows industrial action affecting LCCs as well as the legacy carriers, despite the attempts to reduce the power of the unions. The disruption affected millions of passengers and cost the airlines millions of dollars in funding alternative travel and in compensation. However, it is interesting to note that the Ryanair strike had little impact on customers because the airline hired strike breaking crews from other countries. The data on strike action appears to support the view that unions need to organize along international lines especially when dealing with the more progressive LCCs.

8. Economic shocks The industry has felt the full force of economic shocks. In recent history, the aftermath of 9/11 in 2001 and of the financial crisis in 2007/8 both had a detrimental impact on the sector with demand for air travel curtailed as economies went into recession (Wolla and Backus, 2018). However, the COVID-19 pandemic which hit us in 2020 has been catastrophic in its impact on world economies and the airline industry in particular. In the early days of the coronavirus, some companies furloughed employees hoping for a short-term resolution to the situation. One year on, and many countries have experienced a second and even a third wave, with continuing travel restrictions in order to contain the spread of the virus. The International Air Transport Association which represents 290 airlines is forecasting that companies could shed a million jobs as they counter the impact of $84billion of predicted losses this year (The Times 13/7/20). A recent McKinsey report stated that: By April 2020, US airline capacity declined about 70% from that in 2019, a decline nearly four times greater than seen after the September 11, 2001, attacks and six times greater than seen after the 2008e09 financial crisis (McKinsey, 2020a,b). On a global scale, it is estimated that there are now 4.8 million aviation jobs at risk, representing a 43% reduction compared to pre-COVID-19 levels (ATAG, 2020). This total is made up of 1.3 million at airlines

72

The Air Transportation Industry

(a reduction of 36%), 220,000 at airport operators ( 34%), 3.2 million other on-airport ( 55%), and 151,000 in civil aerospace ( 11%). Furthermore, 46 million jobs normally supported by aviation are also at risk.4 Table 3.5 highlights the impact on some or the airlines which have already announced redundancies. Allied industries have also been affects with Rolls Royce cutting 8000 jobs as demand for its engines has been decimated by the pandemic (The Times, 4/5/20), and Airbus shedding 10% of its workforce as airlines canceled their orders (The Times, 2/7/20). COVID-19 vaccines have provided hope of a recovery in 2021, but there is no doubt that the experience of pandemic in 2020 could herald structural changes in post-COVID-19 era, particularly in terms of business travel. Edgecliffe-Johnson et al. (2020) paint a bleak picture of the return of business trade, suggesting that companies have developed effective alternative practices during the crisis, and recognize the benefits of fewer business class flights in reducing carbon footprints. McKinsey (2020a,b) estimates that business flyers account for only one in ten airline passengers but they generate up to 75% of the airline profits. Brian Pearce, Chief Economist of the International Air Transport Association (IATA) recently estimated that labor costs would have to be cut by around 52% through pay and job losses in order to survive through to 2022. IATA forecasts that we will not see a full global recovery until 2024 (Iata.org., 2020). Out of necessity, we have discovered new ways of working and have realized that much can be achieved in a virtual environment. Added to this, the drive toward environmental sustainability and reduced carbon footprint is increasing pressure to limit international travel. Businesses more generally are moving to more agile and flexible structures and the airline industry will be no different. In Table 3.5, the British Airways dispute was about changing terms and conditions of employment, with BA threatening a “fire and rehire approach,” if terms could not be agreed. This is bound to have implications for the staff who work in the sector. The need for agility is likely to put yet more downward pressure on the cost of labor and drive more flexible of contracts as airlines fight for survival.

4

Analysis based on traffic forecasts, business activity, announcements of redundancies, and modeling. Could be more or less severe as the situation evolves.

Labor in the aviation industry: wages, disputes, and shocks

Table 3.5 Impact of COVID 19 on airline employment. Airline Response

Delta

United American

British Airways Ryanair Emirates Virgin Atlantic Air FrancedKLM Lufthansa American and United Cathy Pacific

18,000 staff taking early retirement and a further 40,000 to take voluntary leave 36,000 redundancies 25,000 redundancies and some 40,000 have taken early retirement, or reduced working hours or taken partially paid leave 1130/4300 pilot redundancies and 12,000/45,000 other employee redundancies 330 pilots have accepted a 20% cut in pay Predicted to cut 15% of their 60,000 employees Shedding 3150 jobs and not expecting to return to pre-COVID19 levels of employment Has already cut thousands of jobs, with another 6000 planned in coming years Already cut 30,000 jobs and expected to shed 10,000 more in Germany Have put a further 27,000 employees on notice that it might stop paying them on April 1, 2021 Cut nearly 25% of its staff (8500)

73

Source

Financial Times 13/10/20 The Times 13/7/20 The Times 1/5/20

Financial Times 27/8/20 The Times 2/7/20 The Times 13/7/20 The Times 6/5/20

Financial Times 18/2/21 Financial Times 4/3/21 Financial Times 3/3/21 Financial Times 10/3/21

9. Conclusions This chapter has used an economic lens to examine labor in the aviation industry. Economic theory describes how wages are determined in an industry which is subject to a number of market imperfections. In the regulated era, which was dominated by national carriers operating in segregated markets, companies had monopsony power, but workers were also heavily unionized creating an effective monopoly of labor supply. This powerful collective voice was very effective in driving up wages in a growing world economy.

74

The Air Transportation Industry

Following deregulation and the more recent Open Skies Agreements, the market for air transport has become increasingly competitive with more operators and different business models. While competition provides alternative employment options, it has also driven companies to find leaner operating models, lowering the proportion of wages to total costs, particularly in the case of the LCCs. The combination of a more elastic supply of labor and a more inelastic demand for labor should increase the power of the unions in a negotiating situation, but they have been less effective against the rise of the new LCCs and also the larger global alliances. The creative employment practices of such as outsourcing to third parties and exploiting different jurisdictions have, in some cases, weakened union power. Unions continue to organize at national level, and thus find it difficult to challenge companies with global operations. Without a more international collaborative approach, unions are only likely to be effective in areas where labor is scarce and markets are segmented. That said, where industrial action does take place, it does have serious consequences in terms of reduced revenue, increased costs and loss of customer base. There is also a “knock on” effect to airport operations, as well as the wider supply chain and aviation enabled tourism. More recently, the COVID-19 shock has devastated the industry which has seen planes grounded, wage cuts, and mass redundancies. This huge economic shock is likely to be a catalyst for structural change as the aviation industry fights for survival. For many airlines, labor still accounts for a substantial proportion of operating costs and is a cost that is likely to be squeezed some more as the industry strives for greater agility and flexibility. This chapter makes two contributions. Firstly, it conceptualizes and discusses the market for aviation labor using an economic lens, where other works use an industrial relations approach. Secondly, it explores the impact of COVID-19 on the sector’s labor market using secondary data sources. Future research could build on these contributions by conducting empirical studies based on this analysis.

References Air Transport Action Group, October, 2018. (Online) Available at: https://www.atag.org/ our-publications/latest-publications.html. Air Transport Action Group, September, 2020. Powering Global Economic Growth, Employment, Trade Links, Tourism, and Support for Sustainable Development through Air Transport, Despite Global Crisis (Online) Available at: https://www.atag. org/our-publications/latest-publications.html.

Labor in the aviation industry: wages, disputes, and shocks

75

Booth, A.L., 2014. Wage determination and imperfect competition. Lab. Econ. 30, 53e58. Cappelli, P., 1985. Competitive pressures and labour relations in the airline industry. Ind. Relat. 24 (3), 316e338. De Langhe, K., Sys, C., Van de Voorde, E., Vanelslander, T., 2013. Economic effects and costs of a temporary shutdown of an airport. In: Proceedings of the WCTR 2013 Conference, the 13th World Conference on Transport Research, July 15e18, Rio de Janeiro, Brazil e 2013, pp. 1e31. Edecliffee-Johnson, A., Hancock, A., Buckley, C., 25 September 2020. The Death of the Business Trip. Financial Times. Eurocockpit.be, 2020. About Us (Online) Available at: www.eurocockpit.be/about-us. (Accessed 14 December 2020). Fitzenberger, B., Kohn, K., Lembcke, A.C., 2013. Union Density and Varieties of Coverage: The Anatomy of Union Wage Effects in Germany. ILR Rev. 66, 169e197. Gittell, J.H., von Nordenflycht, A., Kochan, T.A., Greg, J., 2015. Labor relations and human resource management in the airline industry. In: Belobaba, P., Odoni, A., Barnhart, C. (Eds.), Global Airline Industry. Wiley, Chichester. Harvey, G., Turnbull, P., 2020. Human resource management and industrial relations. In: Budd, L., Ison, S. (Eds.), Air Transport Management: An International Perspective, second ed. Taylor & Francis. Harvey, G., Turnbull, P., 2015. Can labor arrest the “sky pirates”? Transnational trade unionism in the European civil aviation industry. Labor Hist. 56 (3), 308e326. Hirsch, B.T., 2007. Wage determination in the US airline industry: union power under product market constraints. In: Lee, D. (Ed.), Advances in Airline Economics, vol. 2. Elsevier B.V. Iata.org, 2020. COVID-19 Can Costs be Downsized to Make the Industry Cash Positive? (Online) Available at: https://www.iata.org/en/iata-repository/publications/economicreports/can-costs-be-downsized-to-make-the-industry-cash-positive/. (Accessed 14 December 2020). Itfglobal.org, 2020. Civil Aviation (Online) Available at: www.itfglobal.org/en/sector/civilaviation. (Accessed 14 December 2020). Kochan, T.A., 1980. Collective Bargaining and Industrial Relations. Richard D. Irwin, Homewood, IL. Lipsey, R.G., 1992. An Introduction to Positive Economics, seventh ed. Weidenfeld and Nicolson, London. Manning, A., 2013. Monopsony in Motion: Imperfect Competition in Labor Markets, Course Book edn. Princeton University Press, Princeton, NJ. McKinsey, 2020a. For Corporate Travel, A Long Recovery Ahead (Online) Available at: https://www.mckinsey.com/industries/travel-logistics-and-transport-infrastructure/ our-insights/for-corporate-travel-a-long-recovery-ahead#. (Accessed 13 August 2020). McKinsey, 2020b. Coronavirus: Airlines Brace for Severe Turbulence (Online) Available at: https://www.mckinsey.com/industries/travel-logistics-and-transport-infrastructure/ourinsights/coronavirus-airlines-brace-for-severe-turbulence. (Accessed 22 April 2020). OECD.org, 2020. COVID-19 and the Aviation Industry: Impact and Policy Responses (Online) Available at: www.oecd.org/coronavirus/policy-responses/covid-19-and-theaviation-industry-impact-and-policy-responses-26d521c1/. (Accessed 14 December 2020). Ryanair Annual Report, 2020 (Online) Available at: https://investor.ryanair.com/results/. (Accessed 11 March 2021). Ryanair Annual Report, 2010 (Online) Available at: https://investor.ryanair.com/results/. (Accessed 11 March 2021). Southwest.com., 2020. Southwest Citizenship (Online) Available at: www.southwest.com/ html/southwest-difference/southwest-citizenship/. (Accessed 14 December 2020).

76

The Air Transportation Industry

Stewart, M.B., 1990. Union wage differentials, product market influences and the division of rents. Econ. J. 100 (403), 1122e1137. Wolla, S.A., Backus, C., 2018. The Economics of Flying: How Competitive Are the Friendly Skies? (Online) Available at: https://research.stlouisfed.org/publications/ page1-econ/2018/11/01/the-economics-of-flying-how-competitive-are-the-friendlyskies/. (Accessed 14 December 2020).

CHAPTER 4

The air transportation vertical channel, the global value added, and the role played by private versus public control Gianmaria Martini

Università degli studi di Bergamo, Department of Economics, Bergamo, Italy

1. Introduction In recent years, the air transport sector has been characterized by strong growth both in terms of passengers and goods. Up until the end of 2019, predicted growth was positive: Airbus (2019) and Boeing (2019) forecasts, respectively, were þ4.3% and þ4.6% annual increases in air transportation demand for the 2019e2038 period. New aircraft demand reaching about 39,000 and 44,000 confirmed the resilience of the industry despite financial, economic, and geopolitical crises. (Boeing in its report (Boeing, 2019) had reported a þ6.7% annual increase in passenger demand since 2010). This positive sentiment was destroyed by the COVID-19 pandemic crisis, which started in China in January 2020, spread to Europe, and around the world. ICAO (2021) estimates an overall reduction of 50% of seats offered by the airlines, a reduction of 2.7 million passengers (60%) and about USD371 billion loss of gross operating revenue over the entirety of 2020. The crisis from the COVID-19 pandemic is not the first systemic shock to impact the aviation sector. The terrorist attack on the Twin Towers in New York in 2001, the SARS epidemic, and the global economic crisis all put a strain on the resilience of companies in this sector. However, this turbulence generated cases of bankruptcy on the part of companies belonging to the last stage of the supply chain, namely airlines, and partially, handlers. As a response, some national governments invested to rescue these companies. The result is a further increase in the mix of different forms of capital control in the vertical channeldi.e., the simultaneous presence of public and private shareholdersdthat often also gives rise to mixed ownership. The Air Transportation Industry ISBN 978-0-323-91522-9 https://doi.org/10.1016/B978-0-323-91522-9.00014-2

© 2022 Elsevier Inc. All rights reserved.

77

78

The Air Transportation Industry

The main objective of this work is to analyze effects and economic performance on the different forms of ownership existing in various vertical-channel stages and between companies in the same stage. Specifically, the vertical channel is composed of a mixture of public and private companies, and in some cases mixed ownership in which both national and local governments are involved in together with private agents, funds, or companies. These different ownership types may generate alignment in companies’ objectives at different phases of the vertical channel. For instance, a public ownership airport may cooperate with a public ownership flag carrier, as in the case with Gulf carriers. Emirates, Etihad, Qatar, and Turkish Airlines are controlled/subsidized by their national governments, and their hubs in Dubai, Abu-Dhabi, Doha, and Istanbul are public companies. They coordinate operations at the hubs, providing advantages to the flag carriers. However, the different ownership types may also generate a conflict of interestdfor example, a private airline may want to operate flights at night, while a public management airport aims to impose night curfews. Furthermore, analyses (e.g., Martini and Scotti, 2010) have shown that surplus created throughout the vertical channel is divided between the various stages and why operators of one or more phases periodically suffer from profitability issues. Most contributions to the literature have focused on examining the individual supply chain stages. For example, Mason (2007) studied the aircraft construction industry and the strategies of the two main large aircraft manufacturers, Airbus and Boeing. The ground handling segment has been studied by Soames (1997), Schmidberger et al. (2009), and Burghouwt et al. (2014). Alamdari and Mason (2006) explored the final distribution sectordi.e., travel agencies and global distribution systems (GDS). Airports and airlines are the two vertical channel sectors most studied. Regarding the airlines, research has focused on the existence of supply excess (Oum et al., 2005), the welfare effects of mergers between carriers (Brueckner and Pels, 2005), the variables that affect the choice of airlines by passengers (Pels, 2008), and airline efficiency (Scotti and Volta, 2017). Other research has included the low-cost airline segment (Piga and Polo, 2003) and the effects of liberalization of the skies in Europe (Arrigo and Giuricin, 2006) and in the North Atlantic markets (Brueckner et al., 2019). With regards to the airport sector, many contributions have been related to efficiency (Gillen and Lall, 1997; Pels et al., 2001, 2003, Martini et al., 2020) and the effects of privatization (Oum et al., 2008). The current chapter is structured as follows: Section 2 provides a brief

The air transportation vertical channel

79

description of the air transport chain, while Sections 3e9 investigate each stage of the vertical channel. Some final conclusions are reported in Section 10.

2. The air transportation vertical channel The air transport sector includes numerous production stages that operate between the upstream stage of aircraft construction and the downstream stage of air ticket sales to end consumers (Button, 2005). The air transport vertical channel is shown in Fig. 4.1. Players in the vertical channel phases have different weight in the total turnover generated. Martini and Scotti (2010) reported that almost two-thirds of the total turnover of the vertical channel goes to the airline phase, about 16% to the aircraft manufacturer sector, 11% to the airport sector, 6% to engine manufacturers, 1.6% to leasing companies and 1.2% to distribution companies. As such, there is a very high asymmetry in the various economic performances. The first stage is that of aircraft manufacturers. Strategic suppliers for this sector are the companies producing engines for air propulsion that, due to high R&D costs, are vertically separated from the previous stages. Continuing along the vertical channel, leasing companies take advantage of their financial leverage, buying aircraft and then leasing it to the airlines for a fee. Airlines are at the bottom of the vertical channel and provide the Aircraft manufacturers Engine manufacturers

Ownership: private with public subsidies

Leasing companies

Ownership: private with public subsidies

Ownership: private

Airports

Handlers Ownership: private

Ownership: mainly public

Airlines Ownership: mainly private with public bailouts

GDS Ownership: private

Passengers

Freights

Figure 4.1 The air transportation vertical channel with prevailing capital controls.

80

The Air Transportation Industry

service directly to final consumers, thus playing a central role. Airlines are also the only players in the vertical channel to have a direct relationship with the operators of all the other stages. Furthermore, we have airports and firms that deal with ground airport handling services, GDS, and all companies operating in the air ticket distribution sector.1 There are three fundamental elements that differentiate companies operating in the different supply chain stages: (1) the relevant geographic market in which they compete,2 (2) the ownership types, and (3) the economic performance. Regarding the relevant geographic market, the air transportation sector presents a high heterogeneity between the various phases. Aircraft manufacturing and engine manufacturing are global markets, which is also true for leasing companies and GDS. In regards to airlinesdthe central players in the sectordsome are global players while others are local. Airports and handlers, on the other hand, have a relevant local geographic market. The breadth of the geographic market is reflected in the type of ownership that operates in the various phases. In the global market vertical channel phases, private ownership prevails. Aircraft manufacturers and engine manufacturers are private companies. There may be public subsidies for companies that produce aircraftdoften the subject of lawsuits between the US and the European Union where the two main manufacturers residedbut the companies are private. Likewise, for leasing companies and GDS, the distribution of the ownership type among airlines is interesting. In general, airlines are private. However, in the past there was a strong public presence, in the so-called flag carriers, which were subsequently almost completely privatized. Some of these global players are still heavily publicly controlled, while most of the airlines operating on a continental horizon and all low-cost airlines are private. The situation is different for airports and handlers, as the referent context is local. In general, ownership at airports is public, often in a mixed formdi.e., local public capital (regions or provinces) and private. There is a trend toward airport privatization, especially in some European countries, but in most cases control is public. In the handler

1

Airlines can use vertical integration to operate in the distribution segment by selling tickets on their websites. 2 For some sectors (for example, aircraft and engine manufacturers, leasing companies, as well as GDS and online travel agencies) we speak of a real-world market, while for others the dimension is purely local (for example, handling companies, airports, and off-line travel agencies).

The air transportation vertical channel

81

phase, the situation is heterogeneousdin Europe, as a result of the liberalization process, ownership is mainly private (even if privatization has not always been completed). In other parts of the world, there is both public and private ownership. In regard to economic performance, Martini and Scotti (2010) emphasized the significant differences between the weight of the phases in terms of total turnover and profitability. They presented estimates of profitability given by the operating profit margin (OPM and defined as operating profit over sales), which include handlers, and provided an entirely different view from that obtained looking at total turnover. Airports are making the highest OPM (25%); ICAO (2014) reports 15.9% OPM worldwide. Leasing companies have a 22% OPM, which has risen to 25.6% in 2020 according to Macrotrends. The OPM for engine manufacturers is 16%, and Deloitte (2018) showed an increase to 24.3% in 2017 for GE Aviation. GDS have 10%3 of OPM, aircraft manufacturers about 8%, while airlines had 3% in the period 2000e2010 period. Recently, IATA (2017) reported an OPM between 4% and 9% in the 2011e2017 period. Handlers have strong variabilities, with even periods (e.g., 2005e2010), with negative OPM (8%); more recently, CAPA reports an average OPM of about 7% in the 2011e2012 period. The asymmetrical distribution of revenues is evident. It is necessary to emphasize the data relating to airlines. That is, airlines generate the largest volume of turnover within the vertical channel, but extract a minimum percentage of the surplus. Furthermore, airlines do not appear able to cover their cyclical long-term costs. Recent cases of bankruptcy or situations of great difficulty by many carriers confirm this trend. The limited resilience of airlines emerged even more clearly with the COVID-19 crisis. Many airlines have gone bankrupt, were forced to ask for government help, or have even been nationalized. Fig. 4.1 also shows the prevailing forms of capital control for the various stages of the vertical channel. By inspection, the proprietary forms are different, and the main forms are shown. In reality, mixed forms (public and private capital) and public and private properties coexist in every stage. In our analysis of the various phases of the vertical channel, we discuss the forms of capital control and highlight possible conflicts of interest that affect performance. 3

Financial statements point out 8.9% OPM for Travelport in 2018, 9% for Sabre in the same year, and 10% for Amadeus in 2019. All these data are available online.

82

The Air Transportation Industry

3. The aircraft manufacturers The aircraft manufacturing sector can be considered the upstream sector in the vertical channel of air transport. In particular, the sector of large aircraft manufacturers is a symmetrical duopoly (Airbus and Boeing). As the largest commodity market, which includes all manufacturers of aircraft of all sizes, the sector is a concentrated oligopoly. Table 4.1 shows the aircraft sales between Airbus and Boeing in the last decade. Generally speaking, the orders are received by Airbus, a private company owned by the European Aeronautic Defence and Space Company (EADS), and Boeing, another private company. The two companies’ sales are equivalent, with a recent prevailing trend by Boeing, but due to the grounding of the B737 MAX in 2019, there has been a drastic reduction. The leading company depends upon technological development. Airlines are looking for cost savings and environmentally friendly aircraft; thus, if either of the global players introduce a new aircraft model with such characteristics, it gains a short-run competitive advantage. The other operators present in this phase are mostly specialized in the production of regional jets or turbo engines with the relevant exception of the possible development of COMAC, a Chinese aircraft maker. Along with this market, there is also aircraft used in so-called general aviation, which differs both in the number of manufacturers and in the sizes of available aircraft. These are very light jets, and the main players are Cessna and Embraer. The high market concentration and the presence of a symmetrical duopoly do not necessarily imply strong bargaining power toward its customersdleasing companies. In particular, leasing companies are increasingly involved in R&D financing of new models due to their Table 4.1 Sales of new aircraft for Airbus and Boeing. Year

Airbus

Boeing

2011 2012 2013 2014 2015 2016 2017 2018 2019

534 588 626 629 635 688 718 800 863

477 601 648 723 762 748 763 806 380

Source: Airbus and Boeing financial statements.

The air transportation vertical channel

83

financial resources. As previously mentioned, the high costs and risks of these investments push aircraft manufacturing companies to diversify their financing sources so as not to limit innovative activity. By exploiting this situation to their advantage, leasing companies obtain better prices in their purchase of aircraft, and they can sometimes influence the decisions of the manufacturing companies. This is particularly salient because before manufacturing companies can start construction on a new aircraft, they need a “portfolio of potential orders” whose cancellation can have particularly serious effects on the companies’ viability. A controversial issue regarding the aircraft manufacturer sector is the difference in subsidy sources. In the state subsidies received by Airbus and Boeing, the former is often directly helped by the governments of France, Germany, Spain and Great Britain, while the latter is subsidized mainly “indirectly,” thanks to military investments of the American government. The costs involved in the development and construction of commercial aircraft have often made the involvement of governments, directly or indirectly, necessary to bear the levels of risk involved. Irwin and Pavcnik (2004) provided empirical evidence that the subsidy reductions to Airbus and Boeing following the 1992 bilateral agreements resulted in price increases of 3.7%. Martin and Valbonesi (2006) and Lane (2020) have made similar arguments. The companies operating in this sector, are formally privately owned but they receive a lot of public subsidies, and conflict exists in the relationships among manufacturers in regards to possible public subsidies. As an example, there has been a long dispute between the US and the European Union in the WTO regarding unfair public subsidies that might distort trading.4 The main relationships are with leasing companies, airlines, and engine manufacturers. The interests of engine manufacturers and aircraft manufacturers are aligned. Interests are often aligned with leasing companies and airlines, such as in the development of new aircraft models that allow cost savings and less environmental impact. Conflict is related to normal commercial relations, such as sales prices, order compliance, and supply conditions.

4

Information regarding the dispute may be found on the WTO website. It started in 2006 when the U.S. government filed a complaint arguing that Airbus received $22 billion in illegal subsidies. Over the years WTO has ruled that both sides unfairly subsidized their aircraft makers.

84

The Air Transportation Industry

4. The engine manufacturers The air propulsion sector is particularly interesting because, as we subsequently discuss, only three engine suppliers exist for widebody aircraft (this is therefore also a highly concentrated oligopoly). These companies have specialized expertise that guarantees them considerable bargaining power vis-à-vis aircraft manufacturers. In fact, such companies tend to extract a significant share of the vertical channel surplus, lower only than that achieved by leasing companies and airport companies, and much higher than that achieved by aircraft producers. Martini and Scotti (2010) showed the main players are CFM5 and International Aero Engines (IAE)6di.e., a consortium in which General Electric (CFM), Rolls Royce, and Pratt & Whitney (IAE) are relevant members and also receive individual orders. Between the consortium and individual orders, their market share is approximately 93%, a robust collective dominance position. The CFM consortium is the leading company with 39% of the market, followed by Pratt & Whitney (26%), Rolls Royce (18%), and General Electric (16%).7 The global players in this phase are private companies. Public subsidies are generally related to governmental funds for innovation. As previously mentioned, the interests are aligned with those of the aircraft manufacturers. Conflict can be linked to the specificity of investments. That is, if a type of engine can only be used for a particular aircraft model (net of possible and often practiced variations) then a lock-in phenomenon emerges, which benefits engine manufacturers and explains the higher profit margins compared to aircraft manufacturers. The absence of that particular type of engine for the specific aircraft model weakens the position of aircraft manufacturers, who need essential input to fulfill orders.

5. The leasing companies A strategically important stage in the air transport chain is that of leasing companies, which use financial leverage to assist airlines in the aircraft purchase phase. In fact, international regulations in the aviation sector 5

The CFM International acronym derives from the names of the commercial engines of the two parent companies: CF6 for GE and M56 for Snecma. 6 Rolls Royce, Pratt & Whitney, MTU, and JAEC cooperate in the International Aero Engines consortium. 7 Data obtained from T4 website, www.t4.ai.

The air transportation vertical channel

85

provide for the possibility for carriers to use, in addition to their own aircraft, aircraft that is available thanks to a leasing contract, which is normally stipulated by a financial company. In the air transport sector, there are two types of leasing contracts regulated on the basis of article 83 bis of the Chicago Convention (ICAO), as well as national regulations. The “dry lease” is normally used for a prolonged lease period, and the “wet lease” is used shorter leasing periodsdfor example in peak phases or for launch of new routes. A dry lease means a case in which the airline uses, as part of its operating license, an aircraft that they do not own. A wet lease is a rental of the flight by the airline. The rental includes not only the plane, but also the crew, which is made available by the leasing company that owns the aircraft. The decision to lease an aircraft rather than directly is one of the most complex strategic decisions for an airline, as shown by Gavazza (2011), and the decision is based on the optimal balance between advantages (financial and operational flexibility) and disadvantages (excessive cost, lack of capitalization, rigidity in the configuration of the aircraft). Gavazza (2011) showed that about one-third of aircraft operated by major carriers are under operating leases. Within the vertical channel, the leasing company sector (which has a relevant global geographic market) is the sector that manages to extract the largest surplus share, as shown by Martini and Scotti (2010). This information is important, because it indicates that actors operating in this stage are able to obtain better economic results even though they are not perceived by end consumers as protagonists in the vertical channel (many users are unaware of its existence), or have the technological know-how available (aircraft manufacturers and engine manufacturers) or management (companies that deal with the airline ticket distribution). Leasing companies are mostly represented by banks and finance companies, the so-called Commercial Aircraft Sales and Leasing (CASL). Among these, the two largest and best known are General Electric Commercial Aviation Services (GECAS) and AerCap.8 Table 4.2 shows the distribution of the fleet size and the corresponding market share by lessor. As leasing companies are private and the forms of public presence at this stage are modest, there is no conflict linked to different forms of capital 8

In 2013 AerCap acquired AIG the International Lease Finance Corporation, the previously second-most important lessor.

86

The Air Transportation Industry

Table 4.2 Fleet size and market shares by Lessor (2019). Leasing company Fleet size

Market share

GECAS AerCap Avolon BBAM Nordic Aviation Capital SMBC Aviation Capital ICBC Leasing Air Lease BOC Aviation DAE Capital Aviation Capital Group Aircastle Total

18.4% 16.4% 8.4% 8.2% 7.9% 6.7% 6.6% 6.2% 5.9% 5.7% 5.1% 4.4% 100.0%

1143 1016 524 511 488 417 412 383 363 354 319 274 6204

Source: Statista, see website.

control. However, there is conflict with companies active in the production phase of aircraft and engines, because the presence of some engine manufacturers in the capital of leasing companies can create dominant positions due to the control of essential inputs (engines), which exploits the companies’ financial leverage.

6. The handlers The term “airport handling” is used to indicate all those services that fall within airport assistance. The airport infrastructure is normally divided into two parts: the airside activity, including equipment and services for the movement of aircraft; and the landside activity, including equipment, structures, and services related to passengers and goods. Dussart-Lefret and Federlin (1994) and Burghouwt et al. (2014) reported that although these services do not constitute a homogeneous block, and it is possible to divide them into narrower markets, the sector is further divided into three relatively separate sub-markets, including self-handling: that is, services rendered to passengers, services rendered to aircraft, and services relating to the handling of goods. The stage of handling activities was the subject of an intense liberalization process within the European community, which as similar to that

The air transportation vertical channel

87

which occurred in the airline segment.9 In fact, European legislation considers ground handling services subject to the liberalization regime (see EC Directive 96/97), and such services belonging to the following categories includes: ground administrative assistance and supervision, passenger assistance (ticket control, travel documents, baggage registration, and transport up to the sorting systems), baggage sorting, cargo and mail sorting, assistance in runway operations, cleaning and airport services (for example, the de-icing of aircraft), fuel and oil assistance, aircraft maintenance assistance, air operations assistance and crew management, ground transport assistance, and catering and catering assistance.10 These activities require access to various centralized airport infrastructures such as the baggage sorting and delivery system, the piers for boarding and disembarking passengers, the de-icing systems of aircraft, centralized computer systems, static fuel distribution systems, catering, and so forth. Of course, some of these infrastructures, due to complexity, cost, or environmental impact, cannot be duplicated and or even used simultaneously by several operators. In this case, the legislation provides for a possible limit on the number of authorized providers,11 and the selection of operators takes place through a tender. For all other types, however, the presence of a plurality of operators is possible for the same service. The impact of increased competition obtained from the opening of the market is well illustrated by the sparse data on the margins achieved in the handling sector of the vertical channel, which even show negative values. The sector therefore has some of the lowest added value in the supply chain and cannot be classified among those in which annuity positions are nested. These considerations reinforce the belief, already partially reported in the 9

In fact, European legislation considers ground handling services subject to the liberalization regime (see EC Directive 96/97) and activities belonging to the following categories: ground administrative assistance and supervision, passenger assistance (ticket control, travel documents, baggage registration, and transport up to the sorting systems), baggage sorting, cargo and mail sorting, assistance in runway operations, cleaning and airport services (for example the de-icing of aircraft), fuel and oil assistance, aircraft maintenance assistance, air operations assistance and crew management, ground transport assistance, catering and catering assistance. For more details on the operations included in each category, see the EC Directive 96/97. 10 For more details on the operations included in each category, see EC Directive 96/97. 11 European legislation specifies that member states may limit the number of providers authorized to provide the following categories of services: baggage handling, runway operations assistance, fuel and oil assistance, goods and mail sorting. However, the number of lenders cannot be less than two for each of these categories.

88

The Air Transportation Industry

literature (Button, 2005; Button and McDougall, 2006), that the liberalization of the air transport sector is currently incomplete given that in some stagesdthose in which the liberalization process was more intensedmargins are reduced, while in others dnot affected by the market opening processesdmonopolistic rent positions with decidedly large profit margins are embedded. The handler phase is certainly the most fragmented of the whole vertical channel. It is also the one with a local reference context. Services are very often assigned to a specific company through the lowest price offer among several bidders. These offers lead to lower costs for both airlines and airports, but they also expose the winning company to the risk of bankruptcy due to tight operating margins. From an ownership viewpoint, which is generally private, it is interesting to note how often the handlers’ capital sees airport management companies among the shareholders. This obviously creates a conflict with the liberalization process and poses problems for free competition between handlers to acquire the management of a particular service in a certain period at a specific airport. A further cause of conflict concerning handlers within the vertical channel relates to airport operations and possible delays in flights. As shown by Malandri et al. (2019), possible strikes or malfunctions in handler operations can generate cascading delays for airlines given the complexity of the route networks, the need to have short aircraft turnaround periods at airports, as well as multiple use of the same aircraft on different routes on the same day. They also provided evidence that if the number of ground handler’s operators decreases, turnaround operations require more time to be performed, resulting in delayed departures and knock-on delays. In a case study of Lisbon “Humberto Delgado” Airport, turnaround time increases more than linearly with respect to the decrease of staff resources in service. When the number of operators decreases under a certain threshold, turnaround lengthening cannot be absorbed by buffer times, and departure delays propagate and cascade over the day. This implies that a higher number of ground handlers would benefit airlines, although it would also reduce individual handler profits due to intense competition.

7. The distribution: GDS and others The distribution phase in the air transport sector concerns all activities related to air connection information (prices, timetables, connections, and

The air transportation vertical channel

89

so forth), as well as the purchase and issuance of air or travel tickets. The distribution system therefore operates on the basis of two different types of networks: (1) the network of available connections, consisting of the offer formulated independently of each other by the various carriers,12 and (2) ticket booking and sales networks. The choices relating to the first type form part of the airline’s “product” decisions because they are linked to the choice of opening to a new market (i.e., a new point-to-point connection). Those relating to the booking and ticket sales networks are, on the other hand, considered part of the distribution policy, which concerns the choice between two options that are not mutually exclusive: vertical separation or vertical integration.13 In the first option, the distribution phase is carried out by companies independent of the airlines (for example, travel agencies); in the second option, the purchase and issuance of the ticket is handled directly by the airline. As such, it should be noted that the vertical integration option has recently become more attractive to carriers because of an important innovation of recent years: the introduction of the electronic ticket and thus an increase in online sales.14 Travel agencies, tour organizers, GDS, consolidators, tour operators and, in recent years, online travel agencies operate under vertical separation. Travel agencies are traditional travel brokerage organizations that sell products provided by third parties. Such agencies are mostly retailers who work to match airlines and customers. Tourist services have particular characteristics: they are intangible, they cannot be shown before being used, they cannot be stored, and the production and consumption phases coincide. The GDS are telematic booking systems created on the initiative of the large American airlines in order to connect is an extensive sales network with a central office that collects bookings, optimizes available aircraft seats,

12

13

14

A degree of collaboration between airlines in forming a link network exists between partner airlines of the same alliance system, namely Sky Team, One World, and Star. Airlines normally adopt “mixed” distribution strategies with vertically integrated sales channels (for example, online shopping) and vertically separated channels (i.e., purchasing through GDSs or travel agencies). Low-cost carriers only use vertical integration. It should be emphasized that the electronic ticket is also used by other companies operating in the distribution phase, such as GDSs and travel agencies. In fact, in June 2004, IATA decided to impose a goal of achieving 100% electronic tickets in 4 years. This goal was achieved on June 1, 2008 by the 230 airlines that adhere to IATA: airlines, travel agencies, airports, system providers and GDSs have moved an entire industry sector from paper to digital tickets.

90

The Air Transportation Industry

and also provides the customer with immediate answers on availability. The segment represents a worldwide oligopoly, with only three major companies active: Sabre, Amadeus, and Travelport.15 According to Statista,16 Amadeus has the largest market share of 44%, Sabre has 35%, and Travelport has the remaining 21%. The consolidator is an intermediary that deals with placing seats on the market that, a few days after take-off (generally one or 2 days before), are still unsold and therefore represent an effective channel for last-minute rates among customers. Tour operators are companies that organize trips to the public through retail agencies. Online travel agencies can use all the advantages of the internet to channel and market products online and bypass traditional distribution channels to reach end customers directly. The GDS receive a fee from airlinesdi.e., a distribution cost. The airlines try to reduce these costs with internet sales, through both the creation of alternative portals to GDS and direct sales from their own website. The GDS in turn reacts to these dynamics by expanding the online ticket segment. According to Martini and Scotti (2010), GDS still represent the main component of distribution costs; the FAA (2021) showed that distribution costs may account for 5% of airline operating costs. From an ownership viewpoint, companies operating in the distribution segment are generally private. In the past, GDS have seen some form of integration with airlines, which no longer exists.17 As we have seen, distributors often have a competitive advantage over airlinesdi.e., they are the essential points of sale for passengers, which allows for good margins that are again extracted to the airlines. However, this advantage is decreasing with the spread of computer knowledge among the population, which has led to a strong reduction in the use of distributors in favor of direct ticket sales. However, the GDS have the advantage, even online, of providing comparisons between alternatives and thus great added value to the consumer when purchasing a ticket. Hence, the prevailing form of capital control in both stages is private ownership, and we do not identify potential conflict from capital control in the relation between airlines and GDS.

15

Travelport merged the Galileo and Worldspan GDSs under its control in 2008. Online information available on Statista website. 17 Sabre started in 1964 to boost American Airlines sales through integrated computer systems in collaboration with IBM. Amadeus was founded in 1987 by Lufthansa, Iberia, SAS, and AirFrance. In the same year British Airways, KLM, and AerLingus started Galileo, later acquired by Travelport. 16

The air transportation vertical channel

91

8. Airports Airports are essentially local monopolies, as they are the only option for air travel in most areas of the world. In some areas, for example, London, Paris, Milan, New York, Seoul, Tokyo, and Istanbul, among others, there are several nearby airports and thus competition between airports is possible. However, any competition between airports can only occur if the management companies are different. As local monopolists, the companies that manage airports can exploit their market power vis-à-vis both airlines and handlers. For this reason, they achieve very high profit margins. Such high profitability is realized both in the presence of publicly controlled companies and with implementation of regulatory forms of airport tariffs. From the ownership point of view, the situation is very diversified. In the United States, for example, all airports are publicly controlled, in most cases by the port authorities. The same is true in most countries, with the exception of Europe where the situation is more varied. In Spain, for example, all airports are publicly controlled. In other countries, such as Italy, the United Kingdom, and Germany, the situation is mixed. Budd and Ison (2021) showed that in the United Kingdom, most airports have been privatized. Although this trend has increased since 2000, the number of airports with mixed ownership has also increased, while only a small number of airports have public ownership. In Italy, almost all the airports have mixed ownership with many regional governments and local municipalities as shareholders together with private agents. However, there is a growing number of privatized airports, such as in Rome and Venice. The largest airport in Germany, Frankfurt, is privatized. Regardless of their ownership form, airports are subject to regulation. This can apply airport tariffs, operations, or even environmental impacts. The transport regulatory authorities or the relevant ministries regulate the tariffs. Regulatory models are generally of an incentive type for efficient management, and they impose limits on tariff growth. Operations are restricted by transport ministriesdfor example, limits on night flights. Environmental impacts are subject to action by both regulatory authorities and ministries. As we have previously seen, despite these forms of regulation, airports are among the players that manage to generate high profit margins within the vertical channel. The business risk is in fact limited, given that airport attractiveness is generated by the vitality of the local economy. While

92

The Air Transportation Industry

airlines have to contend with load factors, airport management companies are less dependent on aircraft-fill rates. Their profitability is linked to overall volumes as well as freight transport, a rapidly expanding segment especially after the COVID-19 pandemic. However, some airports also suffer from negative performance. This occurs at airports with excess capacity due to stagnant demand. In such cases, management companies often have the benefit of mixed ownershipdi.e., a settlement of losses by public shareholders. The presence of local governments also allows the adoption of subsidies to both airport management companies and to airlines with the aim of stimulating traffic, especially in areas with excess capacity or that wish to increase tourist flows. In such situations, we are witnessing a reversal of market power between airports and airlines. Airlines exploit the advantage of offering air connections and therefore a greater demand for both tourist flows and business opportunities for the local economy to obtain subsidies or discounts on airport rates. The cases discussed by the European Commission against the main low-cost companies, namely Easyjet and Ryanair, were accused of exploiting this advantage to obtain subsidies from airport management companies, in turn financed by local governments, are well known. They have pointed out possible distortions arising from these subsidies, such as lower entrepreneurial risk and obstacles for potential and effective competition. Conflict and tension can be strong in the airport phase both with airlines and handlers, and some problems may be due to the specific type of capital control. Compared to airlines, the disputes are mainly related to fares and slots in the most congested airports. Fares are not only subject to regulation (often as a ceiling), but also to bilateral negotiation between the airport management company and the airline. This negotiation involves the parties dividing the surplus generated by the passenger or the goods shipped. Slots are also a way to extract surplus. That is, if slots are awarded through competitive bidding, they allow the management companies to obtain greater margins by exploiting the competition between companies to obtain high-demand slots. If, on the other hand, the slots are assigned through the grandfather rule, they allow airlines to have margins deriving from position income, that is, slots are always assigned to them in highdemand routes. The possible presence of public control in an airport and in the main airline operating in it may generate a strong conflict of interest, since airport management can introduce an unfair competitive advantage in favor of the airline with the same public capital control (e.g., more slots, or at better times).

The air transportation vertical channel

93

Relations with handlers are more often subject to conflict. On the one hand, the presence of handlers in airport operations removes surplus from the management companies. This is evidenced by the attempt to hinder the entry of handlers, which is often contested by the European Commission to the airport management companies. On the other hand, inefficient handler management also has repercussions on airport management, and it is often the subject of litigation. Airports and handlers have in general different forms of capital controls: the former is public, while the latter is private.18 Hence, the situation is similar to that of comparing capital controls between airports and airlines. Again, a possible conflict may arise if a publicly owned airport provides unfair competitive advantages (for instance, privilege in tenders) in favor of a handler with the same capital control.

9. Airlines Downstream of the vertical channel we have the key players in air transport, namely the airlines. It is the best-known segment of the air transportation sector; therefore, its main characteristics can be briefly summarized. In the airline realm, there are global players, offering long-haul connections, and local players. Companies have very large fleets and small companies with few aircraft. Low-cost companies and carriers offer network services with attention to service (they are in fact called full service). Airlines exist that provide both passengers and freight movements, as well as specialized airlines, such as integrated freighters (see Malighetti et al., 2019a,b). In terms of ownership, most airlines are now privatized. There are some exceptions, such as Gulf airlines, but the general trend is private ownership. Numerous discussions remain open, however, regarding state intervention in the rescue of troubled airlines, a circumstance sharpened by the crisis due to the COVID-19 pandemic. To save airlines from bankruptcy, especially those that in the past were classified as flag carriers, some national governments, especially in Europe, have intervened, creating strong tension between non-bailed airlines (especially low cost carriers) and disputes with the DG Competition of the European Commission.19 18

There are relevant exceptions such as dnata, operating in Dubai airport and controlled by Emirates Airlines. 19 For instance, in 2013, Ryanair sued Alitalia first at the DG Competition and then at the European Court of Justice for illegal state aid granted to Alitalia by the Italian government. Ryanair claimed that these subsidies altered free competition.

94

The Air Transportation Industry

As we have seen, although airlines are the main players in aviation, they are unable to extract most of the surplus generated by the sector. In part this is due to strong competition between airline companies. In addition to this, however, the various relationships with companies active in the other stages of the vertical channel show that airlines are in a weak position in the negotiation process. Financial difficulties expose them toward leasing companies, relations with airport management companies take their local monopoly position into consideration (which is strengthened especially for the most important airports), and handlers certainly have little market power in bilateral negotiations with airlines, but can generate significant delays with a global effect on operations and therefore accumulate delays in flights. In summary, airlines are perhaps the weakest link in the supply chain, with the only advantage of obtaining possible bailouts by national governments (or soft budget constraints in case of a government presence in the ownership) in the event of a financial crisis. Possible conflict of interest due to different forms of capital control may arise for unfair competitive advantages granted by the companies managing the airportde.g., airports and airlines under public control, as in the Gulf carriers.

10. Conclusions This contribution examines the vertical channel of air transport by highlighting the main characteristics, the level of competition, the economic performances and types of capital controldi.e., public, private, or mixed forms. In particular, the objective is to identify whether the different forms of control can generate conflicts of interest between companies both within the same phase and in different phases of the vertical channel. The analysis makes it possible to identify three critical issues. First, although airlines are the main players in the sector, they have, in comparison with companies operating in the other stages of the vertical channel, lower economic performances. This is due to erosion of the final surplus operated by firms working in the phases of the vertical channel that have the advantage of greater counterbalancing power. This applies, in particular, to airports (which undermine local monopoly power) and leasing companies (which play on financial resources and a lower level of entrepreneurial risk). The second critical issue is the incompleteness of the liberalization process within the various phases and their level of mutual competition.

The air transportation vertical channel

95

Handlers and airlines are the most competitive sector, as they are the phases in which the liberalization process, even if incomplete, has had the greatest impact. Aircraft manufacturers are practically a duopoly with a competitive fringe, engine manufacturers a triopoly, leasing companies a concentrated oligopoly, GDS a triopoly, and airports are instead local monopolists. In contrast, there is strong local competition in the handlers’ stage. This differential in competitive levels between the various phases also explains the greater concentration of surplus in the upstream phases. The third issue involves the prevailing forms of capital control that differ at various stages. In general, the prevailing form is private ownership, with the sole exception of the airport segment, in which the prevailing form is public, and second form is mixed (privately owned airports are growing, particularly in the United Kingdom and in some European countries, but they remain a marginal share). However, the links with public spending are much stronger in some phases. For example, in the aircraft manufacturer segment, there are numerous disputes regarding public subsidies guaranteed to the two main manufacturers. The same is true in some cases (think the Gulf companies) for airlines, which can be controlled by national governments. However, it is precisely the airport dimension in which the different forms of capital control generate the greatest conflict of interest and as well as unfair competitive advantage, especially toward airlines (favoring the publicly owned airline) or handlers. These forms of conflict due to public capital are particularly serious if adopted in more open markets, because they generate undue competitive advantage. In these cases, policies should be implemented that reduce the impact of such conflict. In countries with greater political control over economic decisions, any distortions can only be discussed at the level of transnational bodies.

References Airbus, 2019. Airbus Global Market Forecast 2019e2038. Airbus Market Forecast. Alamdari, F., Mason, K., 2006. The future of airline distribution. J. Air Transp. Manag. 12, 122e134. Arrigo, U., Giuricin, A., 2006. Gli effetti della liberalizzazione del trasporto aereo e il ruolo delle compagnie low cost un confronto USA e Europa. In: XVIII Riunione Scientifica Società Italiana di Economia Pubblica. Boeing, 2019. Commercial Market Outlook 2019e2038. Boeing Commercial. Brueckner, J.K., Pels, E., 2005. European airline mergers, alliance consolidation, and consumer welfare. J. Air Transp. Manag. 11, 27e41.

96

The Air Transportation Industry

Brueckner, J.K., Lee, D., Singer, E.S., 2019. Airline competition and domestic US airfares: a comprehensive reappraisal. Econ. Transp. 2, 1e17. Budd, L., Ison, S., 2021. Public utility or private asset? The evolution of UK airport ownership. Case Stud. Transp. Pol. 9 (1), 212e218. Burghouwt, G., Poort, J., Ritsema, H., 2014. Lessons learnt from the market for air freight ground handling at Amsterdam airport Schiphol. J. Air Transp. Manag. 41, 56e63. Button, K.J., 2005. How stable are scheduled air transport markets? Global competition in transportation markets: analysis and policy making. Res. Transp. Econ. 13, 27e48. Button, K.J., McDougall, G., 2006. Institutional and structural changes in air navigation service-providing organizations. J. Air Transp. Manag. 12, 236e252. Deloitte, 2018. Global Aerospace and Defence Industry Financial Performance Study (Available online). Dussart-Lefret, C., Federlin, C., 1994. Ground handling services and EC competition rules. Air Space Law 19 (2), 50e61. FAA, 2021. Economic Values for FAA Investment and Regulatory Decisions, A Guide: 2021 Update. FAA, Washington. Gavazza, A., 2011. Leasing and secondary markets: theory and evidence from commercial aircraft. J. Polit. Econ. 119, 325e377. Gillen, D., Lall, A., 1997. Developing measures of airport productivity and performance: an application of data envelopment analysis. Transp. Res. Part E 33, 261e275. IATA, 2017. State of the Airline Industry. www.iata.org/economics. ICAO, 2014. State of Airport Economics (Available online). ICAO, 2021. Effects of Novel Coronavirus (COVID-19) on Civil Aviation: Economic Impact Analysis, ICAO Uniting Aviation. ICAO, Montreal. Irwin, D.A., Pavcnik, N., 2004. Airbus versus Boeing revisited: international competition in the aircraft market. J. Int. Econ. 64, 223e245. Lane, N., 2020. The new empirics of industrial policy. J. Ind. Compet. Trade 20, 209e234. Macrotrends. Air Lease Operating Margins 2011e2020. (Available online). Malandri, C., Mantecchini, L., Reis, V., 2019. Aircraft turnaround and industrial actions: how ground handlers’ strikes affect airport airside operational efficiency. J. Air Transp. Manag. 78, 23e32. Malighetti, P., Martini, G., Redondi, R., Scotti, D., 2019a. Air transport networks of global integrators in the more liberalized Asian air cargo industry. Transp. Pol. 80, 12e23. Malighetti, P., Martini, G., Redondi, R., Scotti, D., 2019b. Integrators’ air transport networks in Europe. Netw. Spatial Econ. 19, 557e581. Martin, S., Valbonesi, P., 2006. State aid to business. Int. Handb. Ind. Pol. 134e152. Martini, G., Scotti, D., 2010. Potere di mercato e distribuzione dei profitti nella filiera del trasporto aereo. Mercato Concorrenza Regole 12, 127e161. Martini, G., Scotti, D., Viola, D., Vittadini, G., 2020. Persistent and temporary inefficiency in airport cost function: an application to Italy. Transp. Res. Pol. Pract. 132, 999e1019. Mason, K.J., 2007. Airframe manufacturers: which has the correct view of the future e a customer perspective. J. Air Transp. Manag. 13, 9e15. Oum, T.H., Fu, X., Yu, C., 2005. New evidences on airline efficiency and yields: a comparative analysis of major North American air carriers and its implications. Transp. Pol. 12, 153e164. Oum, T.H., Yan, J., Yu, C., 2008. Ownership forms matter for airport efficiency: a stochastic frontier investigation of worldwide airports. J. Urban Econ. 64, 422e435. Pels, E., Nijkamp, P., Rietveld, P., 2001. Relative efficiency of European airports. Transp. Pol. 8, 183e192. Pels, E., Nijkamp, P., Rietveld, P., 2003. Inefficiencies and scale economies of European airport operations. Transp. Res. Part E 39, 341e361.

The air transportation vertical channel

97

Pels, E., 2008. Airline network competition: full-service airlines, low-cost airlines and longhaul markets. Res. Transp. Econ. 24, 68e74. Piga, C., Polo, M., 2003. Il giro del mondo in 80 euro. Mercato concorrenza e regole 2, 281e296. Schmidberger, S., Bals, L., Hartmann, E., Kahns, C., 2009. Ground handling services at European hub airports: development of a performance measurement system for benchmarking. Int. J. Prod. Econ. 117, 104e116. Scotti, D., Volta, N., 2017. Profitability change in the global airline industry. Transport. Res. E Logist. Transp. Rev. 102, 1e12. Soames, T., 1997. Ground handling liberalization. J. Air Transp. Manag. 3, 83e94.

This page intentionally left blank

CHAPTER 5

Exogenous shocks on the air transport business: the effects of a global emergency Cristiana Piccioni, Andrea Stolfa and Antonio Musso

DICEA, Department of Civil, Building, and Environmental Engineering, “Sapienza” University of Rome, Rome, Italy

1. Introduction The COVID-19 public health crisis is commonly recognized as the biggest social and economic shock at the global level after the Second World War. Its impact on the ways we currently live, work, do business, and trade has already been enormous but still difficult to evaluate. The long-term scenarios of COVID-19 disruptions in transport systems and on the economy as a whole are not easy to predict, also considering that such systems affect the spatial distribution of population and economic activities and have both positive externalities (e.g., transport connectivity, social cohesion) and negative ones (e.g., CO2 emissions, global warming). In this context, the whole current mobility patterns and tools are rapidly changing, and we have to start re-thinking what will be the “new normal” and how do we get there. Aviation is undoubtedly one of the sectors most hit by the economic crisis induced by the COVID-19 health emergency, with many countries closing their borders. The domestic air travel supply has also been affected by social distancing, confinement measures, and shrinking economic activities. The year 2020 witnessed a decline of 65% in the world airline capacity. Particularly, the United States recorded a 72% decrease, followed by 71% in China and 48% in Japan. In Europe, Italy, Spain, France, and Germany lost 90% of their air traffic compared to the same period of 2019. What happened in 2020 has shown that new business models are required both for airlines and for airport management. Besides, restoring air connectivity is essential for economic recovery. The COVID-19 crisis is still causing a massive economic output loss, creating the biggest shock in a century for many economies. Such a loss would be translated into a contraction in world 2020 GDP ranging The Air Transportation Industry ISBN 978-0-323-91522-9 https://doi.org/10.1016/B978-0-323-91522-9.00007-5

© 2022 Elsevier Inc. All rights reserved.

99

100

The Air Transportation Industry

between 4.4% and 5.2%, far worse than during the 2008 global financial crisis (cf. IMF, 2020; World Bank, 2021). The government’s potential agreements not already carried out make scenarios for recovery particularly uncertain. Even if some of the initial harsh restrictions have been quite relaxed, the future economic output will be likely depressed because of business failures, canceled investments, and long-term unemployment. The hope is that the new normal in the post-COVID-19 will be significantly different and better than before. Some argue that it could be more car-dependent, but others suggest it could be a significant chance for more local living and virtual communications to replace longer trips. Anyway, the main outcomes will be the result, in substantial part, of the policy choices made over the coming months and years. The air market’s emergency crisis impacts will be outlined in the following sections. A focus on the supply revision policies will highlight the new mission of the nationalized airlines and the presence of the State, also in its new role of entrepreneur. Special attention will be paid to the core measures put into action by the main incumbents and low-cost air carriers in response to the sanitary and economic crisis, thus investigating the stability of the respective business models, according to the different airlines’ viewpoint.

2. Impact of COVID-19 on the airline businessdthe worst crisis since ever This section presents an overview of facts and figures of air traffic volumes at the global and European levels by investigating the dramatic drop in travel demand due to government restrictions. Even in the brightest forecasts and considering a greater domestic market’s resilience than the international one, such analysis highlights that traffic volumes of 2019 will be expected only in late 2024. Nevertheless, such a result is not unique; it rather frames different geographical contexts because today, regions such as the United States, Central and South America, and India are replying differently in terms of timing and types of containment measures the virus’ spreading. On the supply side, a critical discussion on whether and how it is possible to create a margin for air carrier survival is provided. Besides, the main value drivers affecting the airlines’ operation model’s efficiency are presented, thus linking the aircraft capacity with the resources’ productivity concept.

Exogenous shocks on the air transport business: the effects of a global emergency

101

Lastly, starting from the key figures of the worst month since ever, the main reactions addressing the global air market are described. 2.1 Air transport demand hugely impacted The reduction in passenger mobility demand, arising from governments’ restrictions and individuals’ fears of contracting the virus by using mass transit systems, is one of the major social and economic impacts to be traced back to the COVID-19. What is caused by such a pandemic is an impressive crisis negatively impactingdin all its dramadalso on the shrinkage of air transport demand, both short and long haul. Even if this sector has undergone many declines over time (e.g., due to the 2001 terrorist attack on the Twin Towers or the global economic crisis of 2008), such a sudden and profound emergency is putting a strain on air transport, thus announcing epochal changes in the industry itself. The decline process of global air transport services slowly started in early 2020, then turned into a free fall over the following 3 months. In February 2020, the international passenger capacity decreased by 10%, affected mainly by traffic from/to countries experiencing an early outbreak and those deeply linked to China. In March 2020, such a capacity was further reduced by 48% at the global level. The bottomdthe historical minimum ever experienced in modern aviation historydwas reached on April 2020 due to strict constraints imposed worldwide for national and international air travel (Table 5.1). As of May 18, 2020, almost all worldwide destinations have been subjected to travel restrictions. About 85% of destinations have entirely or partially closed their borders, and a further 5% have suspended international flights almost wholly. What happened in Europe presents almost similar dynamics. The following three months’ scenario has been beyond any plausible gloomy forecast, resulting in a more than 60% reduction of passenger traffic because of COVID-19 restrictions measures. The above, in turn, was translated into a significant loss of resources, in terms of money and workforce, along with a decrease of the Gross Added Value (GAV).1 The key figures describing the economic impact affecting air carriers, regardless of the airline 1

GAV is a measure of the contribution to GDP made by an individual producer/industry/sector. As it includes all primary incomes, it provides a better measure of the economic welfare of population. The GVA and GDP relationship is given, as follows: GVA ¼ GDP þ subsidies on products  taxes on products.

102

The Air Transportation Industry

Table 5.1 International passenger capacity loss: the snapshot of April 2020. Capacity loss from Country originally planned COVID-19 virus spread

United States United Kingdom Germany Spain China France Italy United Arab Emirates Japan Turkey Thailand

22,976,621 22,345,210 19,374,444 18,041,897 16,683,876 13,480,021 12,464,502 11,009,896 9,501,833 8,798,224 8,441,105

88% 90% 92% 94% 95% 91% 94% 89% 88% 94% 94%

Republic of Korea

7,960,525

86%

Hong Kong SAR of China (CN) Netherlands Singapore Canada India Switzerland Russian Federation Malaysia

7,122,206

93%

6,960,693 6,596,279 6,288,656 6,286,458 5,990,424 5,747,918 4,959,606

89% 93% 90% 89% 93% 87% 85%

Portugal Saudi Arabia Australia

4,913,803 4,193,572 4,115,805

95% 77% 92%

Mexico Austria Qatar Indonesia

4,104,882 3,812,866 3,760,492 3,723,583

78% 91% 80% 87%

Vietnam

3,681,731

89%

Ireland Poland Denmark

3,595,318 3,449,632 3,417,729

92% 79% 93%

Belgium

3,323,135

87%

Greece Philippines

3,078,774 2,993,741

94% 86%

50,000 confirmed cases

10,000  confirmed cases  49,999 50,000 confirmed cases 1000  confirmed cases  9999 10,000  confirmed cases  49,999 100  confirmed cases  999 10,000  confirmed cases  49,999 50,000 confirmed cases 10,000  confirmed cases  49,999 50,000 confirmed cases 1000  confirmed cases  9999 10,000  confirmed cases  49,999 1000  confirmed cases  9999 10,000  confirmed cases  49,999 1000  confirmed cases  9999 100  confirmed cases  999 10,000  confirmed cases  49,999 1000  confirmed cases  9999 10,000  confirmed cases  49,999 1000  confirmed cases  9999

Exogenous shocks on the air transport business: the effects of a global emergency

103

Table 5.1 International passenger capacity loss: the snapshot of April 2020.dcont'd Country

Capacity loss from originally planned

Sweden

2,941,579

89%

Norway

2,476,519

90%

Egypt

2,248,437

78%

Brazil Israel

2,214,850 2,196,238

92% 91%

COVID-19 virus spread

10,000  confirmed cases  49,999 1000  confirmed cases  9999 10,000  confirmed cases  49,999 50,000 confirmed cases 10,000  confirmed cases  49,999

Source: Authors’ elaboration from ICAO dataset, 2020.

registration region, for an international framework are summarized in Table 5.2. In this context, the UK and Western European markets result as the most impacted in terms of passenger demand and economic losses. To better understand the magnitude of COVID-19 impacts on the European skies, an actual comparison (Fig. 5.1) between the daily flights performed in 2019 (light blue; light gray in print) and 2020 (dark blue; black in print) provides evidence of what happened during the last three months of restrictions when the air services were almost entirely stopped compared to the pre-emergency conditions. This is also confirmed by the descending trend of the moving average of the daily variation over the last seven days (red color; light gray in print). A focus at the country level allows stressing how in the second half of June 2020, which is around two months away from the plunge in transport supply, most countries were still coping with a shortfall of air transport services: Macedonia, Malta, Morocco, Georgia, and Israeldfollowed by Ireland, Spain, UK, Greece, Portugal, Ukraine, and Latvia recorded a daily average flight number more than 80% below what performed in 2019, underlining that some of them will take several years to compensate the traffic lost in the first semester of 2020. In early August 2020, stricter air travel restrictions were lifted and, still missing the evidence of the second epidemic wave, the European market’s business confidence continued its slowly ascending trend toward a new equilibrium point. This is shown in Fig. 5.1, where the trend started in mid-June, compared to the Marche June past direction, confirms an increase, albeit relatively thin, of daily flights performed in the 2020 early second half (dark blue).

104

The Air Transportation Industry

Table 5.2 Impacts of COVID-19 on the air industry. Airline O-D Passenger passengers revenues Employment [$ billion] [units] [million] Country [%]

1 2 3

Austria Belgium Czech Republic 4 Finland 5 France 6 Germany 7 Greece 8 Hungary 9 Ireland 10 Israel 11 Italy 12 Netherlands 13 Norway 14 Poland 15 Portugal 16 Romania 17 Russia 18 Spain 19 Sweden 20 Switzerland 21 Turkey 22 UK 23 Ukraine Average value

GVA* total [$ billion]

58 61 61

16.5 18.6 10.6

2.6 2.6 1.3

52,600 66,700 33,800

4.6 6.4 1.2

61 61 63 58 59 61 55 59 60 76 57 60 59 58 59 67 60 55 61 55 L60%

9.6 88.7 113.4 28.8 9.8 21.3 13.1 92.0 32.6 26.0 22.8 28.6 12.3 67.5 124.5 22.6 29.6 58.7 154.6 10.8 L45.55

1.5 15.7 19.5 4.1 1.2 2.7 3.2 12.6 6.0 3.6 2.6 4.0 1.4 9.0 16.8 3.0 5.6 7.0 28.7 1.3 L7.03

40,600 434,700 534,000 260,100 42,300 86,800 95,300 345,300 177,600 105,900 64,600 187,800 54,300 428,800 983,100 112,200 120,000 559,600 732,500 81,100 L243,465

3.6 38.9 37.6 11.2 1.7 12.5 8.3 23.5 14.4 11.5 2.1 8.0 1.3 9.9 64.7 11.2 15.9 24.8 55.7 0.8 L13.90

GAV, Gross Added Value. Source: Authors’ elaboration from IATA dataset, 2020.

According to the broad set of forecasts (IATA, 2020; ICAO, 2021) describing possible global and local scenarios, it is reasonable to assume a rise in the air traffic demand could be achieved not before Summer 2021. The mid-year IATA forecasts affirmed that it would need to wait until 2024 to win back traffic levels of 2019. That was because air traffic recovery speed in May and June 2020 was lower than expected, mainly due to uncertainties about the restrictions to entry imposed for international flights. Such an outlook is, in turn, confirmed by the IATA forecasts based on end2020 data: airlines are not expected to turn liquidity positive until 2022 or see their traffic levels recovery until 2024. As of this stage of total uncertainty, coming months will be decisive to understand better whether such projections will be totally or partially denied or, at best, confirmed.

Exogenous shocks on the air transport business: the effects of a global emergency

Figure 5.1 EU air market: MarcheDecember trend and comparison with 2019. (Source: EUROCONTROLdAIU, January 1, 2021.)

105

106

The Air Transportation Industry

2.2 Deep impact on operating model: resources’ productivity and aircraft capacity The leading airlines’ financial statementsdgiven a relatively fragile business model as exposed to cyclical crises of drops in demand in an environment always highly competitivedinevitably suffered from the pandemic effect. Compared to the previous ones, such a crisis is still casting doubt also on the survival of the strongest and most consolidated carriers. Even in the past, air fleets were grounded, but there had never been a collapse of nearly 90% of passenger traffic lasting for at least three months. But that’s exactly what happened at the global level during the first phase of the COVID-19 health crisis. According to a literature review in the field (cf. Lan et al., 2006; Daft and Albers, 2014; Zhou et al., 2020), the robustness of the airline business model could mainly be traced to value drivers, as follows: (1) efficiency in terms of productivity, to be intended both for airplane fleets and crew; and (2) density, in terms of seats sold and flights operated, to measure how efficiently the single plane is used. They both have been entirely destroyed by the economic crisis induced by the pandemic, causing revenue to be zeroed. The reduction of offered capacity, that is, number and frequency of flights, due to destinations canceled because politically quoted and reduced flight frequencies due to low demand, negatively impact the productivity of the business model as a whole. That happens because the reduction in the total number of hours flown, one of the main parameters affecting the efficiency, has to be distributed over the entire airline schedule, thus entailing an increase in the single flight unit cost. In other words, the amount of “no-flown hours” determines a drop of main production factors such as crews (employment contracts) and airplanes (leasing contracts) whose fixed component remains constant. Consequently, the cost per single hour flown increases while the productivity ratios invoiced, both for personnel and aircraft, significantly fall. Besides, adopting spacing measures imposed for passenger safety purposes implies a reduction of on-board capacity that, in turn, causes an increase in the unit cost of the seat (CASKdCost of Available Seat Kilometer). All the above, considering the suspension of some ancillary items (e.g., luggage, food) as an essential source of additional revenues (up to 40%e50% for aggressive low-cost airlines), put a strain on the survival of airlines. Air carriers try to increase the ticket price to compensate for the

Exogenous shocks on the air transport business: the effects of a global emergency

107

very low-profit margins; however, the high demand elasticity and the low structural traffic volumes make such tactics quite tricky to apply. As a final consequence of such a vicious circle, the aircraft’s filling coefficients are significantly reduced and remain far from the economic break-even load factor.2 That generally ranges between 70% and 75% on the short-medium haul and 80% on the long haul in normal conditions. The IATA’s (2020) year-end forecasts estimate that about 40 million jobs in global aviation and its related value-chains, including the tourism sector, are at risk in the current crisis. Besides, passenger revenues fell to $191 billion, less than a third of $612 billion earned in 2019. Again, they predict industry losses of $118.5 billion in 2020 and $38.7 billion in 2021, thus stressing that both estimates are more profound than previously expected (Table 5.3). After a gradual lifting of domestic markets restrictions started in early June 2020, followed by a reopening of regional and intercontinental Table 5.3 Impact of COVID-19 at the global level: current trend and expected one. Demand Capacity 2020 2020 Demand Capacity e2021 2021 e2021 2019 2019 2020 (D 2019) (D 2019) profits Continent e2020 e2020 profits

North America Europe Asia Pacific Middle East Latin America Africa World

66.0%

51.6%

70.0%

62.4%

62.0%

55.1%

73.0%

64.5%

64.0%

60%

72.0%

62.8%

66.3%

57.6%

$45.8bn

þ60.5% (45%) $26.9bn þ47.5 (56%) $31.7bn þ50.0% (43%) $7.1bn þ43.0% (61%) $5.0bn þ39.0% (50%) $2.0bn þ35.0% (62%) $118.5bn þ50.4% (50%)

þ36.4% (34%) þ35.5% (49%) þ38.4% (38%) þ23.6% (56%) þ34.3% (46%0 þ21.5% (55%) þ35.5% (43%)

$11.0bn $11.9bn $7.5bn $3.3bn $3.3bn $1.7bn $38.7bn

Source: Based on IATA, 2020. Annual Review. https://www.iata.org/contentassets/c81222d96 c9a4e0bb4ff6ced0126f0bb/iata-annual-review-2020.pdf. (Accessed 4 January 2021).

2

It is given by the ratio between the CASK and the Yield (revenue per passenger km), thus making clear that a drop of revenues leads to an increase in the numerators that can rise indefinitely (up to over 100%).

108

The Air Transportation Industry

traveling, the last quarter of 2020 was affected by a tightening of restrictions globally due to the second pandemic wave affecting worldwide. At this point, we don’t know how long it will take to come back to pre-crisis conditions. Undoubtedly, COVID-19 related impacts will change the way the air sector industry operates shortly and beyond and, surely, it will never be the same again. 2.3 Immediate reactions from the vast majority of the operators: fleet and capacity reductions By the first week of April, governments in 75% of the markets monitored by the IATA3 banned entry, while an additional 19% imposed travel restrictions or mandatory quarantine for international arrivals. The above resulted in a 98.3% collapse of global passenger demand (measured in Revenue Passenger Kilometers) in April 2020, compared to April 2019, thus worsening the 58.1% drop recorded in March. Besides, available capacity (expressed by Available Seat Kilometers) decreased by 95.1% and the load factor4 (computed as a percentage of ASKs used) fell 46.6% over the 2019 same period up to 36.6%. Key figures of the worst month since ever, April 2020, spitted by the leading international markets, are summarized in Table 5.4. Table 5.4 International passenger market in April 2020: the worst month since ever. International markets

World share

RPK

ASK

PLF (%-PT)

PLF (LEVEL)

Africa Asia-Pacific Europe Latin America Middle East North America Total market

2.1% 34.7% 26.8% 5.1% 9.0% 22.2% 100.0%

98.3% 88.5% 98.1% 96.0% 97.3% 96.6% 94.3%

88.4% 82.5% 94.9% 94.0% 92.4% 80.5% 87.0%

62.8% 28.2% 53.2% 27.1% 52.1% 69.9% 46.6%

11%, 1% 53%, 8% 32%, 0% 55%, 0% 28%, 4% 15%, 0% 36%, 6%

ASK, Available Seat Kilometers; PLF, Actual load factor; RPK, Revenue Passenger Kilometers. PLF (%PT): Year-on-Year change in load factor. Source: Based on IATA, 2020. Annual Review. https://www.iata.org/contentassets/c81222d96 c9a4e0bb4ff6ced0126f0bb/iata-annual-review-2020.pdf. (Accessed 4 January 2021).

3

IATA (International Air Transport Association) represents 290 airlines comprising 82% of air traffic at global level. 4 It is given by the ratio between the RPK and the ASK.

Exogenous shocks on the air transport business: the effects of a global emergency

109

Table 5.5 Main domestic passenger markets in April 2020: the worst month ever. Domestic World PLF markets share RPK ASK PLF (%PT) (LEVEL)

Australia Brazil China Japan Russia USA Domestic market

0.8% 1.1% 9.8% 1.1% 1.5% 14.0% 36%, 2%

96.8% 93.1% 66.6% 88.7% 82.7% 95.7% 86%, 9%

92.5% 91.4% 57.2% 54.6% 62.4% 72.9% 72%, 1%

46.1% 15.9% 18.6% 51.8% 43.8% 72.3% L44%, 3%

34.6% 65.9% 66.4% 17.1% 37.1% 13.5% 39%, 5%

ASK, Available Seat Kilometers; PLF, Actual load factor; RPK, Revenue Passenger Kilometers. PLF (%PT): Year-on-Year change in load factor. Source: Based on IATA, 2020. Annual Review. https://www.iata.org/contentassets/c81222d96c9a4e 0bb4ff6ced0126f0bb/iata-annual-review-2020.pdf. (Accessed 4 January 2021).

As far as the domestic passenger markets are concerned (Table 5.5), last April, the traffic component decreased by 86.9%; the deepest drop respectively occurred in Australia (96.8%), Brazil (93.1%), and the United States (95.7%). This was a further deterioration from a 51% decrease already recorded in March. Moreover, the ASK declined by 72.1% and the related load factor fell from 44.3% up to 39.5%. It is worth mentioning how the initial flight increases, following the partial reopening of international skies in early June 2020, focused on national markets. Late May 2020 data showed that flight levels in the Republic of Korea, China and Vietnam increased by up to 22%e28% less than in 2019. Such latter figures suggest that the airline industry market, after reached the crisis peak, was giving the first sign of a new beginning to relaunch connectivity in the medium term, least for national and continental destinations. Throughout the full year 2020 (cf. ICAO, 2021), the COVID-19 impact on scheduled passenger traffic, compared to the baseline scenario (Business-as-usual) planned before this pandemic, would be likely translated into a consistent reduction of 67% of seats globally offered by airlines, an overall traffic volume reduction of 1470 million passengers and a potential loss of airlines’ gross operating revenues of $263 billion. To this end, it seems useful to look at the corresponding figures at the global level (Fig. 5.2). The above indicatorsdnamely capacity offered, mobility demand, and revenuesdconfirm that the European air market, followed by Asia and Pacific ones, are still suffering the impact of the crisis more than other countries.

110 The Air Transportation Industry

Figure 5.2 Impact on international passenger traffic and revenues by region during 2020. (Source: Based on ICAO, 2021. Effects of Novel Coronavirus on Civil Aviation: Economic Impact Analysis. Air Transport Bureau, Montréal, Canada. January 7th, 2021. https://www.icao.int/ sustainability/Documents/COVID-19/ICAO%20COVID%202021%2001%2007%20Economic%20Impact.pdf. (Accessed 24 January 2021).)

Exogenous shocks on the air transport business: the effects of a global emergency

111

From the passenger demand side, it is plausible to suppose that the public’s response to the COVID-19 emergency will probably be delayed by the current greater perceived risk of infection than other sanitary crises that occurred in the past. A confirmation of that arises from two surveys involving a large sample of travelers living in different countries worldwide; the former5 focuses on the public attitude to risk while the latter6 deals with the passenger behavioral approaches in China and Australia domestic aviation markets. Both underlined how people take more time, up to six months, before feeling safe enough to restart travel and, anyway, they will wait until their financial condition comes to stable again. Lessons learned from previous crises suggest that people often adopt new mobility practices during and immediately following an emergency, which can sometimes become permanent. Traveling for business purposes could be the case, at least for all Companies replacing business trips with videoconferencing, also in virtue of the wide spreading of more robust digital technologies. Such a consideration would not be applied to leisure purpose trips distinguished in the long and short-haul. The former requires a more complex planning phase so, they likely perform a slow recovery, while the latter could show a more dynamic trend once travelers have assured that flying and traveling is safe again. Despite the high number of flights canceled by April 2020, the willingness to travel showed slight recovery signs in the following months. Late June 2020, flight capacity was still around 60% of what it used to be in the pre-crisis condition. After some countries announced to remove travel limitations for the summer season, air traveling increased between July and September 2020. Then, a growing of infection cases at the global level imposed a new round of gradual lockdowns and travel restrictions in many regions affecting Eastern and Western Europe, the Middle East, the United States, Latin America, India, and Asia. The sanitary equilibrium and, so the economic and social ones, continue to be still very labile. The awareness the pandemic has not yet been defeated causes such instability and until the virus is completely eradicated, 5

The Ipsos survey was conducted April 16e19, 2020 on the Global Advisor online platform among 28,000 adults (approximately 2000 individuals in each country) respectively aged 18e74 in Canada and the United States and 16e74 in Australia, Brazil, China, France, Germany, Italy, India, Japan, Mexico, Russia, and the United Kingdom. 6 It was commissioned by IATA to define a set of confidence-boosting measures aimed at speeding up the recovery in air travel demand. The survey focused on the domestic markets of China and Australia.

112

The Air Transportation Industry

also considering the vaccination campaign in progress, it will call for stricter sanitary measures to be translated in a suspension of all flights with the risk regions. Such safety measures are commonly applied in all those countries still hostage to the first infection wave or those affected by a second wave because of the virus’s resurgence after the summer season or those who fear the third wave because of the virus’s alterations and variants. Whether and when passenger air transport demand returns to pre-crisis levels will depend on a broad set of parameters, all affecting people’s decision-making process. However, once countries lifted restrictions due to virus’ regression and/or large-scale vaccine administration, it is also reasonable thinking that confidence-boosting measures, along with support actions to the air market by States, will play a key role in restart travelingrelated activities, thus speeding up the recovery of the global economy.

3. A new era of nationalization? 3.1 The financial support to traditional flag carriers In common with the rest of the economy, government assistance has been essential in the aviation industry to preserve incomes and jobs through the COVID-19 crisis. Specific measures have included suspending requirements for “use it or lose it” landing rights at slot-coordinated airports, suspension of landing charges in Iceland and Norway and protecting services to remote communities. There was no single approach to supporting airlines during the crisis (Table 5.6). The US government earmarked $50 billion for a mix of grants, loans, and equity options for all airlines until July 2020 while Australia turned down support applications, partly because of the potential negative impact on competition. In Europe, the UK has considered support once all commercial finance options were exhausted. French government support to Air France-KLM required to bring forward the Company’s commitment to halve CO2 emissions per passenger from 2020 to 2024 for domestic flights. Different is the Italian airline case where the government will bring at least V3 billion capital into Alitalia, preparing itself to retake control of the airline after 11 years of difficult private management and three failed restructuring attempts. By summarizing, initially, governments were to provide temporary relief to airlines until travel demand recovery; however, such support has been unevenly distributed across all regions: the USA, European and parts of Asian airlines have generally received relevant governments support,

Exogenous shocks on the air transport business: the effects of a global emergency

113

Table 5.6 Support to airlines in selected countries in addition to payroll support until July 2020. The USA France Germany Italy Netherlands

Airlines eligible Grants Equity stakes Loan guarantees Government loans Total

All airlines (pax) $17.5 billion Optional

Air France

Lufthansa

Alitalia

KLM

0

0

0

0

0

V4 billion

$32.5 billion $50 billion

V3 billion

V300 millionc V3 billion V5.7 billion V9 billion

V1.3 billiona To be definedb n.a.

V7 billion

V3.5 billion V4.8 billion

0 V2.4 billiond V1 billiond V3.4 billion

a

Paid by the Italian State, in the pre-COVID era (when the Company was commissioned). The State will likely take a 100% stake, at least initially, and set aside V700 million to support Alitalia. c For 20% government stake in the Company. d Bank loans of up to V2.4 billion and a direct loan from the Dutch State of up to V1 billion. Source: Authors’ elaborations. b

while the assistance for airlines in Latin America, the Middle East and Africa was quite limited. As of the end of November 2020, government aid has globally amounted to $173 billion. The core of them made up of loans ($58 bn), wage subsidies to preserve jobs ($46 bn), loan guarantees ($24 bn) then followed by capital injections ($23 bn), deferring the taxes payment and reducing tax liabilities ($23 bn). 3.2 A new model of nationalized airlines The COVID-19 crisis shows that the decision-makers in governments have enormous power to exercise to the extent that they can decide which companies survive and which companies should go bankrupt or be liquidated. After the initial crisis is over, the policymakers are likely to develop the temptation to exercise their power over the private sector airlines in which the government owns significant portions of their shares and/or bonds. Such government power may lead to inefficiency and/or corruptive practices that should be avoided considering the following best practices: A. Governments should purchase “non-voting” shares rather than buying bonds. Non-voting shares make it difficult for the government to chair the Company or change the Top Management. The government could

114

B.

C.

D.

E.

The Air Transportation Industry

recover some returns to taxpayers’ money by selling those shares at a higher price later. State intervention in airline activities should be needed at the operational level, but more likely in tied bailouts or conditional restructuring loans. The governments around the world have to start specifying what they expect to extract for their interventions. These conditions may counter what an unchained business community would ordinarily want, but it’s necessary for this difficult situation. Noteworthy is the case of the German government’s support to Lufthansa, the second largest European airline getting strong financial support during the crisis of V9 billion (cf. Table 5.6). From the first lockdown end, the German carrier increased the number of flights handled after a second quarter with a considerable decrease of the revenues (96%) and a first semester 2020 with a reduction of the passengers handled (66%) compared to the 2019. Governments should also take a long-term vision of their air connectivity needs and airlines’ commercial standing when defining what financial aid to make available. During the COVID-19 crisis, flight cancellations have had mostly a serious impact on islands and peripheral regions, many losing all international connections. Many governments supported routes providing essential air connectivity before the crisis. Since the crisis, some have introduced new aids where financial support was provided for services on particular routes rather than to specific carriers. The benefit of such an approach is its focus on connectivity outcomes. Subsidies should be non-discriminatory toward carriers, including specific clauses to ensure that they will be periodically revised and revoked once the crisis is over. A return to hub-and-spoke operations could be foreseen in some cases. During the early 21st century, airlines globally used 300-seats planes such as the Boeing 787 Dreamliner and the Airbus A350 to connect new city pairs, linking secondary cities to each other, bypassing the hubs. After COVID-19, however, with a severe curtailment to international routes, only the main routes will survive. The Governments should re-think the ownership of privately-owned national carriers, reversing 40 years of liberalization. The national government’s ability to protect air routes into their territories, while loosened over decades, provided some cover after 2001 September 11, with insolvencies limited to Belgian national carrier Sabena and Swissair in Europe while US airlines were able to enter bankruptcy protection.

Exogenous shocks on the air transport business: the effects of a global emergency

115

3.3 How the new market environment impacteddreactions from low-cost carriers Low-cost carriers (LCCs) have presented a better reaction to the crisis than the incumbents in financial terms because of the more flexible network they own (the leading European airlines canceled almost totally intercontinental EuropeeUSA flights). LCCs’ business model seems to be more robust and has proved to be resilient in the face of many previous crises. It is likely that LCC’s will be structural winners again. Still, looking at both the USA and the European market, the responses to the crisis by the leaders in this segment were quite different as described next. 3.3.1 Southwest Airlines It is a US airline known as the major low-cost carrier worldwide. As of 2014, Southwest connects 93 cities in 41 states, Puerto Rico and overseas, although most routes are within the United States. In the pre-COVID-19 era, Southwest Airlines operated nearly 30,000 flights a week. By mid-April 2020, its weekly seat supply was 2,401,224 seats, putting it ahead of its national competitors and the leading Chinese airlines. Late April to mid-May brought the lowest flights in the United States over a decade, but as the restrictions began to ease, the major airlines, namely American Airlines, Delta, and United, improved their supply. Southwest increased its capacity to 15,000 flights/week in early June plus other 5000 flights at the end of June, while the three majors applied a more conservative strategy, approaching 10,000 flights/week. By early August 2020, Southwest operated around 21,000 flights/week, thus maintaining its leadership in the US market ahead of its top competitors, that is, China Southern, China Eastern, American and Air China, Delta, ANA, Wizz Air, Xiamen, and Qatar. Such success is due to a well-defined operating strategy: the airline has limited sections of each plane dedicated to business or first-class seating, allowing more capacity on each flight. Besides, it manages only a few international flying. It is most successful in US areas where lockdowns are limited or delayed, likely serving many smaller hubs in the Midwest and Western regions. 3.3.2 Ryanair Ryanair is the largest and most profitable European LCC, operating 450 Boeing 737e800 aircraft, generating V1 billion profits with 149 million passengers flown in its last financial year. As Europe moved into lockdown,

116

The Air Transportation Industry

Ryanair was entirely grounded until July 2020, when it began to reintroduce around 40% of its average capacity. It moved over 4 million people in July 2020 with a load factor of 72%, representing a fall of 70% versus 2019. A progressive ramp-up was planned to thank also to the government financial aids of V4.1 billion. Still, a tight quarantine regime restrained the Company in its home country, Ireland and the UK, imposed on visitors returning from holidays countries such as France, Spain, and Croatia. Thus, Ryanair cut the planned SeptembereOctober capacity by 20% by indicating a possible contraction in the airline capacity also over this year. 3.3.3 EasyJet It’s the second largest European LCC, operating over 300 Airbus A320 aircraft and flying 96 million passengers in the financial year ended September 2019. easyJet began its restart flying program in mid-June 2020, with activity limited to only 10 aircrafts, operating mainly in the French and UK domestic markets. Since then, it has stepped up activity and had planned to be at 40% of normal capacity by September 2020, already considerably more cautious than Ryanair. easyJet will have a fleet of around 30 fewer planes in the coming months, while the 24 aircraft delivery will be postponed beyond 2025. All of this means that the Company was unable to maintain its current network size and shape without reducing capacity in the face of the COVID-19 crisis. Given that a fundamental part of easyJet’s strategy is to preserve number one and two positions in key slot constrained airports, it has its work cut out simply defending these positions as other airlines see an opportunity to move in and claim territory. 3.3.4 Wizz Air The Hungarian LCC Wizz Air handled 40 million passengers delivering V281million profit for the past financial year. As the COVID-19 crisis has reduced its power, Wizz Air has moved more quickly than its rivals to higher activity levels. By July 2020, it operated nearly 75% of planned capacity and carried 47% of its regular traffic with 1.8 million passengers handled. Moreover, 200 new routes and four new bases were also announced. To enable these new opportunities 22 aircraft in the current fleet have been moved. It is also balancing its fleet toward larger aircraft, shifting from the A320 with 180 seats to the A321 version with up to 239 seats. This will further reinforce its capacity growth. The larger type accounts for around 50% of the fleet and rises to almost 90% by 2027. Ryanair will remain far ahead, but Wizz Air’s size will increase relative to easyJet,

Exogenous shocks on the air transport business: the effects of a global emergency

117

due to the latter’s growth constraints. Wizz Air is gradually moving beyond its core territory and is confident in coming up against its competitors in some of their more productive leisure markets in Western Europe. This is a calculated risk, but with the benefit of a very low-cost base overall and specifically the lower unit seat costs with its increasing number of Airbus A321 will drive, the airline appears well placed to handle a substantial improvement of its performances.

4. How the industry is going to face the crisis This section describes the core measures put into action by the main incumbents and low-cost air carriers to cope with the global crisis’s economic impacts. Although the flight restrictions were imposed at the country level, a common approach in implementing actions at the international levels can be recognized. To this end, the recurring measuresdin terms of fleet and capacity optimization, resizing of airline capacity because of the introduction of seat spacing measures, synergies between alliances, postponement of new aircraft orders and review of leasing contractsdare presented, thus going to investigate the stability of the respective business models, according to the different airlines’ viewpoints. 4.1 How major airlines have been considering to face the midterm issues One of the features of the COVID-19 sanitary emergency was the virus’s geographical spread, which appeared in Asia and then affected the rest of the world in the space of just over a month. Most airlines tried to keep their schedule unchanged until mid-March 2020 when drastic mobility restrictions were imposed. Among them, a common policy response across the world was the border closures that, in turn, resulted in a sudden drop in the flight’s number, first intercontinental and international ones. As a result, global markets suffered a stronger impact than domestic ones, and, to this end, it is likely to be expected that longhaul connections will be the last to be reestablished. Moreover, due to travel bans and restriction setting at the national level, such a process likely occurs unevenly worldwide. On the contrary, at least at the early of crisis, the national market experienced a less incisive and heterogeneous reaction, as it was lived as a sort of airlines’ buffer to preserve a certain level of operation, just before the stop of air traffic during the late March 2020 lockdown. According to the above, the major airlines were the main losers, and they will be so, both in the medium and long-run.

118

The Air Transportation Industry

Following what previously discussed, it is quite evident that most airlines worldwide searched the first aid through government grants, loans at preferred conditions/state guarantees, or subsidies. It is equally clear that without such a government-based strategy, all the leading airlines would not have been able to safeguard their financial status quo and, in some cases, avert the risk of bankruptcy. Besides, to cope with such a dramatic impact on air carriers’ assets and meet the governments’ safety requirements, all the airlines were obliged to implement further countermeasures by resizing both the fleets’ and aircrafts capacity. While a relevant transport demand drop has directly imposed the former, the latter are related to the social distancing due to seat spacing measures within the aircraft.7 Although the flight restrictions were imposed at the country level, common measures and actions implemented at the global level can be recognized. The main ones deal with a proper reorganization of fleets, fleet conversion from passenger to cargo, synergies between alliances, postponement of new aircraft orders and rearrangement of leasing contracts. As far as the reorganization of the fleet is concerned, it can be achieved: • By cutting fleetsdi.e., Austrian Airlines (25%) Brussels Airlines (30%) or planning a deep reduction into the near future, i.e., Lufthansa; • By replacing larger and older wide-body aircraft (i.e., the B747 and the A380) with narrow-body aircraft for long-haul routes. The above has been carried out along with working together with regional airlines to have a feeding fleet more appropriate for shorter links. This is the case of Air France bringing forward A-380 retirement, restructuring its domestic network with fewer flights and more low-cost connections (by Transavia) as well as Lufthansa itself grounding A380s likely permanently and replacing 20% with smaller aircrafts. As an immediate response to the economic crisis, the most dynamic airlines started converting (temporary or permanently) their fleet by reconfiguring their airplanes to cargo. Such a change in the airframe implies 7

According to the European Union Agency for Aviation Safety (EASA) guidelines, the obligation of 1-m interpersonal distance onboard aircraft was confirmed. It is permitted to derogate from that, if carriers, in addition to meeting a series of requirements, will define with airport managing bodies specific procedures that allow boarding of hand baggage of permitted dimension for placement in the overhead bins putting in place appropriate and selective embarking/disembarking measures about assigned seats on board (source: ENAC).

Exogenous shocks on the air transport business: the effects of a global emergency

119

removing passenger seats and fitting freight pallets on seat tracks. Austrian airline was among the first to opt for this solution, followed by Icelandair and Swiss Air. In September 2020, German maintenance specialist Lufthansa Technik started converting an Airbus A380 to offer temporary frompassenger-to-cargo modification services. In this context, the health emergency has been a catalyzer for new air market logistics, mostly related to the supply of essential medical equipment and medicines. In a middle and long-run perspective, a further measure deals with the exploitation of synergies between alliances. Such a strategy allows air carriers to consolidate or extend their network, thus providing customers with more compelling and competitive travel options and, at the same time, achieving cost savings and better economies of scale. Among the leading incumbents choosing such an approach, Air France-KLM signed a joint venture agreement with the US company Delta Air Lines and the British airline Virgin Atlantic in early February 2020, just before the sanitary emergency. Besides, British Airways engaged in a UKeAustralia partnership with Qatar Airways that came into effect on May 29, 2020 and will run for five years. Such new partnerships shift the balance by creating new market equilibria, apparently disjointed each other, but affecting the air industry dynamics internationally. Proof of that is the announcement made in midMay 2020 by Etihad Airways, Qatar Airways’ Middle Eastern rival, about its intention to restore service from Melbourne to London Heathrow, via its Abu Dhabi hub, after its suspension due to the COVID-19 crisis. At the same time, Emirates Airlines started in May 2020 to open new intercontinental connections from Dubai, also offering a seamless flight experience to customers traveling between the United Kingdom and Australia. A further key point deals with postponing new aircraft’s orders along with reviewing leasing contracts. According to data retrieved by IBA’s intelligence platform (cf. IBA.iQ, 2020), it is expected that older aircraft will be considered surpluses to airlines’ requirements over the next 18 months. This may be likely translated into more than 1800 12-year-old widebody aircrafts8 to be alienated by the global airline system shortly.

8

IBA has identified 4-engine type are the vulnerable ones, such as the Boeing 747 and Airbus A380, in addition to older twins including Boeing 777-200/ER/LR models, Airbus A340s and mature A330s as those likely to suffer most.

120

The Air Transportation Industry

However, the considerable fall in airlines’ revenues under their almost complete impossibility to operate along with a drop in travelers’ confidence around future bookingsdthus avoiding capturing the benefits of the advance ticket sales practicedrequired carriers to scale back their growth. This implies that leading airlines are rushing to postpone aircraft deliveries in 2020. They are still reevaluating long-term orders while trying to reorganize their strategies considering the industry’s austerity prospects extended well beyond 2021. 4.2 The reactions of low-cost carriers In the low-cost market, a dichotomous scenario emerges. There is the drama of closing several air carriers not able to contain the economic impact derived from the safety restriction measures on passenger traffic. This is contrasted by a different reality in which airlines have enough resources to manage such a complex situation. In some cases, they can take opportunities from this crisis, for example, by reconsidering their fleet configuration or extending their network by using new hubs no longer congested by the presence of incumbent carriers. On the one hand, Europe’s ultra-cheap flights could disappear due to the most fragile airlines’ closure in the “low-cost” market. In this regard, the privately owned Italian airline Air Italy ceased operations in February 2020; in the same month, the Turkish carrier AtlasGlobal Airlines was filed for bankruptcy while the Norwegian-owned Swedish airline Braathens was filed for court administration in April 2020 and Lufthansa closed its subsidiary Germanwings. Again, the British carrier Flybe went into administration after the COVID-19 outbreak. On the other hand, the historical players of such a “no-frills” market live their match according to different strategic plans, all thought for surviving and, in some cases, to enlarge their market shares or conquer new ones. In this context, easyJet falls into the former group, deciding to cancel new aircraft orders and, at the same time, rearranging a leaner fleet. As well, Sun Express, born as a joint venture of Lufthansa and Turkish Airlines, has revised its strategy in the running by switching a part of its aircraft fleet to cargo operations. Conversely, Helvetic Airways falls in the latter group. To tackle the current situation due to COVID-19 induced crisis but positioning itself as ideally as possible for the future, the Swiss regional airline is diversifying its fleet capacity by acquiring four larger E195-E2

Exogenous shocks on the air transport business: the effects of a global emergency

121

Embraer aircraft.9 Besides, the Spanish low-cost airline Volotea has planned 40 new seasonal routes while Wizz Air expects to enter new markets in Europe and increase ancillary service-related business. Finally, as one of the leading companies at the global level for traffic volume and the undisputed queen of European low-cost carriers, Ryanair still has the power of driving the air market dynamics both in terms of connectivity and price competition. Numbers in hand, after just under two months from the reopening of the skies, Ryanair has flown 99% of its airplanes, followed by Wizz Air with 97%, at the same time, easyJet has made 71% of its fleet available. Ryanair’s strategic plan is evident: to put their entire fleet into circulation to grab the European market, the most flexible one, considering that, from a global perspective, intercontinental flights will require a lot of time before reaching the numbers of 2019. Despite the still uncertain prospects affecting the LCCs industry, Ryanair seems to be once again decreed leader of the European market. As previously mentioned, it is the carrier with the biggest fleet (around 450 aircraft against WizzAir’s 130), which may be translated into 1900 flights/day: a value more than two and a half times higher than what was achieved by easyJet (700 flights/day). It is still challenging to get an idea of how long the aftermath of the crisis will be and what further measures could be put in place by the airlines to contain the economic damage and plan their recovery. However, what is clear to date is the widespread awareness that the most efficient narrow-body aircraft used to bypass hubs will be useful also to low-cost airlines for longhaul traffic revitalization. Besides, the significantly lower traffic levels than in the past, the potential entry of new competitors in hub airports (i.e., Wizz Air), the late recovery of long-haul demand as well as the reduction in feeding traffic indicate that the strategic challenge could be played mainly in primary hub airports: i.e., Dubai is one of them for many LCCs.

5. Conclusions The worst global crisis has hit the aviation sector ever since: an air traffic drop of more than 90% along with entire aircraft fleets grounded all around the world featured half of 2020. 9

The Embraer E195-E2 strikes a good balance between seating capacity (between 120 and 150 seats), range, fuel consumption, and environmentally friendly operation. It seems to have virtually no competitor in the regional aircraft segment.

122

The Air Transportation Industry

Recovery will be slower than originally expected, and even if traffic increase signals mainly in domestic markets were arising, it is commonly recognized that 2019 air traffic levels will be likely achieved not before late 2024 at best! Added to that, the scenario of business traffic is featured by drastic and permanent drops in international demand because of radical changes in customer behaviors shifting to digital videoconferencing. At the global level, airlines, regardless of financial background and fleet size, all immediately reacted with capacity reductions. Nevertheless, the operating model has been deeply hurt in its key fundamentals, namely, aircraft and capacity productivity. The above has also given rise to undeniable and unpredictable consequences involving the medium and longrun. Governmental support at the global level has played a crucial role in avoiding most airlines falling bankrupt by providing several social measures and fiscal incentives. However, in some countries, governmental initiatives moved-up to acquisitions of shares and new nationalization models, thus determining a potential conflict between the State’s dual role of “regulator” and “shareholder” Besides, most support was devoted to domestic flag carriers, thus triggering strong reactions from leading low-cost carriers claiming against asymmetric incentives criteria. Despite the context described in this chapter, since it has been said “in the worst scenarios, a great opportunity arises” the post-pandemic setting will surely provide the leading airlines with a unique chance to reset their business models. It will be possible to translate it into reality by rethinking the airline paradigm according to several emerging value drivers, mainly traced to both the aircraft fleet purchasing and ground and flew personnel recruiting processes. The former would be due to the lowest fleet ownership cost ever. Indeed, the abnormal availability of aircrafts grounded will determine unique opportunities to buy a new and second-hand fleet at the cheapest conditions ever. The above implies that the smartest carriers with financial solidity could place an order at up to 70% discount versus previously listed prices for both new and second-hand aircraft. The latter would be linked to a highly efficient labor market. As a matter of fact, there will be an unprecedented excess of pilots and crews on the market, allowing air carriers to renegotiate contractual agreements much more in line with the current traffic demand conditions. That would also be an excellent opportunity to recover competitiveness on the cost per available seat kilometer.

Exogenous shocks on the air transport business: the effects of a global emergency

123

Finally, on the top-line potential revenue, the adoption of new management tools implementing recent developments in data analytics technologies and AI would allow establishing new criteria of value-based segmentation, dynamic pricing and customer management, thereby boosting revenue growth. Combining such drivers will likely determine a new competitive landscape where boundaries between full-service carriers and low-cost ones will disappear, thus paving the way for establishing a new post-pandemic airline model.

References Daft, J., Albers, S., 2014. An Empirical Analysis of Airline Business Model Convergence, Working Paper, No. 112. University of Cologne, Department of Business Policy and Logistics, Cologne. IATA, 2020. Annual Review. https://www.iata.org/contentassets/c81222d96c9a4e0b b4ff6ced0126f0bb/iata-annual-review-2020.pdf. (Accessed 4 January 2021). ICAO, 2021. Effects of Novel Coronavirus on Civil Aviation: Economic Impact Analysis. Air Transport Bureau, Montréal, Canada. January 7 2021. https://www.icao.int/ sustainability/Documents/COVID-19/ICAO%20COVID%202021%2001%2007%20 Economic%20Impact.pdf. (Accessed 24 January 2021). IMF, 2020. EURO AREA POLICIES, Staff Report for the 2020 Article IV Consultation with Member Countries, Country Report No. 20/324. December 2020. Lan, S., Clarke, J.P., Barnhart, C., 2006. Planning for robust airline operations: optimizing aircraft routings and flight departure times to minimize passenger disruptions. Transp. Sci. 40, 15e28. World Bank, 2021. Global Economic Prospects. International Bank for Reconstruction and Development/The World Bank, Washington DC, ISBN 978-1-4648-1613-0. Zhou, L., Liang, Z., Chou, C.A., et al., 2020. Airline planning and scheduling: models and solution methodologies. Front. Eng. Manag. 7, 1e26.

Further reading Albers, S., Rundshagen, V., 2020. European airlines strategic responses to the COVID-19 pandemic (JanuaryeMay, 2020). J. Air Transp. Manag. 87 (C). Bailey, J., 2020. How Southwest Became the World’s Largest Carrier (For A Week), August 28th. https://simpleflying.com/southwest-largest-carrier/. (Accessed 12 October 2020). Curley, A., Dichter, A., Krishnan, V., Riedel, R., Saxon, S., 2020. Coronavirus: Airlines Brace for Severe Turbulence. McKinsey. April 22nd. https://www.mckinsey.com/ industries/travel-logistics-and-transport-infrastructure/our-insights/coronavirus-airlinesbrace-for-severe-turbulence?cid¼soc-app. Dunn, G., 2020a. The Story of the Coronavirus Impact on Airlines in Numbers. https:// www.flightglobal.com/strategy/how-the-airline-industry-has-been-hit-by-the-crisis/ 138554.article?adredir¼1. (Accessed 3 September 2020). Dunn, G., 2020b. Air France Terminates the A380 Fleet with Immediate Effect. https:// www.flightglobal.com/fleets/air-france-terminates-a380-fleet-with-immediate-effect/ 138463.article. (Accessed 3 September 2020).

124

The Air Transportation Industry

ENAC. COVID-19 e Informazioni ENAC per passeggeri, operatori e industria. https:// www.enac.gov.it/news/covid-19-informazioni-enac-per-passeggeri-operatori-industria. (Accessed 14 September 2020). EUROCONTROL e Aviation Intelligence Unit, Dashboard. https://ansperformance.eu/. (Accessed 8 January 2021). Fonte, G., 2020. UPDATE 1-Italy to Inject 3 bln Euros in New Alitalia. https://www. reuters.com/article/italy-alitalia-minister/update-1-italy-to-inject-3-bln-euros-in-newalitalia-minister-idUSL8N2CP4B6. (Accessed 12 July 2020). IATA. COVID-19 Puts over Half of 2020 Passenger Revenues at Risk. https://www.iata. org/en/pressroom/pr/2020-04-14-01/. (Accessed 4 July 2020). IATA. Traveler Survey Reveals COVID-19 Concerns. https://www.iata.org/en/ pressroom/pr/2020-07-07-01/. (Accessed 6 September 2020). ICAO. COVID-19 Response and Recovery Platform. https://www.icao.int/covid/Pages/ default.aspx. ITF, 2020. Restoring Air Connectivity under Policies to Mitigate Climate Change, Transport and Covid-19: Responses and Resources. https://www.itf-oecd.org/covid19. (Accessed 4 July 2020).  Mihajlovic, Z., Draskovic, B., Cokorilo, O., 2020. The influence of passenger behavior and economy measures on air traffic recovery after COVID-19 crisis. In: Zanne, M., Bajec, P., Twrdy, E. (Eds.), ICTS 2020 Maritime, Transport and Logistics Science Conference Proceedings. Faculty of Maritime Studies and Transport, Portoroz, ISBN 978-961-7041-08-8, pp. 236e245. Reuters, 2020. UPDATE 1-Wizz Air Expands in Western Europe as Crisis Brings Opportunities. https://www.reuters.com/article/wizz-air-expansion/update-1-wizzair-expands-in-western-europe-as-crisis-brings-opportunities-idUSL8N2DB5SW. (Accessed 4 July 2020). Welle, D., 2020. Lufthansa Grounds Germanwings and Cuts Fleet Size. https://www.dw. com/en/lufthansa-grounds-germanwings-and-cuts-fleet-size/a-53053398. (Accessed 3 September 2020).

CHAPTER 6

The impact of regulation on the airport industry: the Italian case Carlo Cambini and Raffaele Congiu

Department of Management, Politecnico di Torino, Turin, Italy

1. Introduction An airport is a complex system, both in terms of operations and in the relationship between the variety of its users, from airlines to consumers. This multi-faceted nature of airports is reflected in the complexity of economic regulation of the airport industry. If, at first, airport charges were a sort of taxation for the use of state-owned infrastructure, over time they have changed to better reflect costs and revenues incurred by airport managing companies. An increasing emphasis was put into cost efficiency as a requirement to be met alongside allowing interconnections in a secure way, leading to a more diffusely adoption of incentive-based regulatory mechanisms. As commercial activities started providing an ever-increasing share of overall revenue, single- versus dual-till has become a subject of discussion. All these factors have led to the increasing complexity of regulatory frameworks. As regulation evolves to adapt to an evolving market, it becomes fundamental to assess the impact of regulatory policies in order to guide this process. In Italy, in 2014, a newly appointed regulation authority passed a new regulation which introduced a negotiation procedure for airport charges, in application of the EU Directive 2009/12/EC, together with a price cap mechanism with dual-till. Our analysis aims to assess the impact that this regulatory change has had on the average cost of affected airports; i.e., we compare the performance of airports adopting the new regulatory scheme with that of airports which did not apply the new mechanism yet. We exploit this event as a quasi-natural experiment, which allows us to adopt a difference-in-differences methodology. We build on the approach taken by Conti et al. (2019) to provide an analytical framework that can be useful to assess the impact of regulatory measures in the airport industry. Although we analyze the impact of this shift in regulation on the Italian airport system, this method can be used to assess the effect of a change in regulatory The Air Transportation Industry ISBN 978-0-323-91522-9 https://doi.org/10.1016/B978-0-323-91522-9.00016-6

© 2022 Elsevier Inc. All rights reserved.

125

126

The Air Transportation Industry

approach in any country. Indeed, the approach we propose might be applied to study the effectiveness of any regulatory change (e.g., a shift from cost plus to incentive-based regulation, from single- to dual-till, from the absence to the presence of a negotiation procedure) in any country where airports are subject to ex-ante scrutiny. Our analysis finds evidence of a positive effect of the regulation in term of overall efficiency: after an initial increase in airport average cost following the adoption of the new scheme, average costs significantly decline in the following years. The remainder of the work proceeds as follows. Section 2 presents a literature review on the assessment of airport managers’ performance and the regulation of airport companies. Section 3 describes the evolution of the Italian regulatory framework in the airport industry. Section 4 explains the adopted empirical strategy. Section 5 describes the data used. Section 6 discusses the results of the analysis. Section 7 provides conclusions.

2. Airport regulation: a literature review This work is related to the rich body of literature on regulation of airport companies and on the assessment of airport managers’ performance, in particular concerning cost efficiency. This section provides an overview of relevant literaturedboth empirical and theoreticaldfrom which we draw in structuring the following analysis. In airport management, several studies have been focused on benchmarking and on the comparison of airport efficiency. Based on a sample of 34 European airports observed between 1995 and 1997, Pels et al. (2003) make use of both stochastic frontier analysis and data envelopment analysis to estimate production frontiers for aircraft traffic and passenger traffic. The study shows that European airports are generally inefficient. It also finds increasing returns to scale in terms of passenger traffic for airports with at least 12.5 million passengers/year, whereas the returns to scale are constant for freight movements. Oum et al. (2008) consider the effects of ownership type on airport cost efficiency. They use a sample of major airports in the world, showing that airports controlled and operated by privately owned companies reach greater efficiency levels than those operated by mixed companies whose majority is owned by the government. Barros (2008a,b) carries out frontier analyses on the operating costs of British and Portuguese airports, respectively, ranking airport by efficiency level and finding in both cases substantial levels of cost inefficiency. Martín et al. (2009) analyze the

The impact of regulation on the airport industry: the Italian case

127

efficiency of Spanish airports operated by a single operator (AENA). The authors estimate a long-term total cost frontier. Their sample includes 37 airports over the period from 1991 to 1997. The results show that cost inefficiency range from 15% to 26% on average, and that larger airports are typically more efficient. Regarding the analysis on the Italian airport industry, Curi et al. (2010, 2011) investigate efficiency determinants looking at ownership type and concession agreements. Their first study employs a dataset of 36 airports ranging from 2001 to 2003. The second one looks at 18 airports over the period from 2000 to 2004. They find higher levels of efficiency in publicowned airports than mixed-owned or private ones and a generally low efficiency level. Furthermore, they determine that the type of concession agreement can be a driver of operational efficiency, especially for smaller airports. The study by Scotti et al. (2012) analyses a production frontier with a sample of 38 Italian airports observed between 2005 and 2008. In particular, the research examines the efficiency level with respect to the degree of airport competition, which is measured by an index based on the ratio between the number of seats available for each route at an airport and the total number of seats available at the same airport, that is, an indicator of potential demand in the airport under consideration. Lo Storto (2018) estimates a meta-frontier through data envelopment analysis to study the effect of ownership types on technical, cost and revenue efficiencies. His data covers 45 Italian airports in the year 2012. His results underline that private owned airports have higher levels of technical efficiency, but perform as good as and worse than public ones in term of cost and revenue efficiency, respectively. Finally, and more recently, Martini et al. (2020) estimate a cost frontier for the Italian airport system to disentangle longterm and short-term inefficiencies. They employ a panel of 15 airports for years 2010e15. They find that most of the inefficiency is a short-term one and that private-owned airports appear to be more efficient than public-owned ones. Our analysis also draws from the literature on regulation of airport companies. This body of literature analyses various aspects of airport economics, such as the determinants of airport charges, the effect of the ownership structure, the impact of single- versus dual-till regulation. Bel and Fageda (2010) study the determinants of airport charges using data on 100 large European airports. They find a positive correlation between charges and passenger volume and a negative one with the presence of competing airports or other transport modes. They also find evidence of

128

The Air Transportation Industry

higher charges in non-regulated airports, although the regulation mechanism does not appear to matter. Bilotkach et al. (2012) also look at airport charges determinants through a panel of 61 European airports observed over 18 years. Their focus is on the effect of regulation policies and privatization. Their findings suggest that single-till regulation and privatization lead to lower charges. They also find tentative evidence of ex-post regulation having a lowering effect on charges. Furthermore, their results suggest that price cap regulation and the presence of rival airports do not lead to lower charges. Adler et al. (2015) analyze the impact of privatization on the governance of airport companies, assess how well regulatory authorities reflect “good regulation” principles, and review benchmarking studies to assess the allocative efficiency performance of new regulatory practices. They also perform a data envelopment analysis to study the impact of incentive regulation on productive efficiency, finding that it performs better than cost plus. Czerny (2006) studies the impact of single-versus dual-till regulation. He shows thatdfor non-congested airportsdsingletill overperforms dual-till in terms of welfare maximization. Gillen and Mantin (2014), Ivaldi et al. (2015), and Malavolti (2016) study the airport as a two-sided platform. Gillen and Mantin (2014) analyze the trade-off between aeronautical and commercial revenues to determine when privatization is to be preferred. They find that, if passengers show a sufficiently large valuation for concession goods, lowering charges attracts more passengers and leads to an increase in commercial revenues. In such cases, welfare loss in the aeronautical segment due to privatization is minimized. Otherwise, privatization is not recommended. Ivaldi et al. (2015) highlight the presence of externalities between airlines and passengers. By performing an empirical test on major US airports, they show that airport companies do not internalize these externalities. Malavolti (2016) studies the airport as a platform that matches passengers and commercial activities. She identifies a trade-off between the waiting time and the number of passengers: single-till charges can be respectively higher or lower depending on which of the two factors has a greater impact on shop demand. She points out how dual-till masks the externalities existing between the two sides of the market. Particularly relevant to our analysis is the work by Conti et al. (2019). In this work, the authors assess the impact that the Eu Airport Charges Directive had on the level of aeronautical charges. They perform a difference-in-differences analysis using a panel of 113 EU airports over the period 2008e17. They find that the transposition of the Directive into

The impact of regulation on the airport industry: the Italian case

129

national law by member states has led to a decrease in the level of airport charges after some years.

3. Airport regulation in Italy In Italy, until the year 2000, airport charges were periodically updated by decree of the Ministry of Transport. The charges were fixed without considering neither the airport size nor the costs incurred, as they had the nature of a tax for the use of state-owned infrastructure. With the CIPE Resolution n. 86/2000, a dual-till price cap mechanism was adopted for the first time. The tariff model was determined considering a series of parameters such as costs, volumes, productivity, investments and environmental protection. The determination of these parameters was under the responsibility of ENACda public organization established with the Decree n. 250/1997. The role of ENAC was to stipulate contracts with airport management companies, with a validity of at most four years, aimed to define aspects of technical-operational and economic-financial nature. However, the CIPE Resolution was never actually applied, as operators judged it to be too complex and penalizing compared to European tariff levels, and a regulation not related to costs continued to be followed (Sciandra, 2009). The CIPE Resolution n. 38/2007 transposed the new regulatory criteria defined by Law 248/05, summarized in the following points: • The recognition of costs directly and indirectly attributable to regulated services; • The return on net invested capital and on new investments; • The recognition of depreciation costs and operating expenses deriving from new planned investments; • The deduction of at least 50% of the commercial margin to be made to the costs of regulated activities (hybrid-till). The implementation of the 2007 CIPE resolution has led to the definition of numerous contractsdcalled program contracts (“Contratti di programma”)dwhich has resulted in the diversification of the tariff system within the sector, i.e., differentiated tariffs depending on whether the airports have signed such contracts. More precisely, the airports that signed a program contract were able to adjust charges to the costs incurred, while the others could adjust charges only to the inflation rate. In this scenario, ENAC defined three tariff schemes according to airports’ passenger volume. These tariffs adopted a sort of cost-based mechanism.

130

The Air Transportation Industry

The regulatory scenario completely changed in 2013, with the establishment on September 17, 2013 of the Italian transport regulation authority (in what follows, “the Authority” or “ART”). The role and operation of the Authority are codified in decree-law December 6, 2011 n. 201, converted, with amendments, into Law December 22, 2011, n. 214. The establishment of the Authority rounded out the framework for the independent ex-ante economic regulation of public services that, in Italy, had been introduced by Law July 14, 1995 n. 481. The latter set off the creation of regulators in the energy and telecommunication sectors, the competencies of which have later been extended to, respectively, the water and postal services sectors. Reflecting the distinctive nature of the industries and markets concerned, at present, the three authoritiesdthe energy, telecoms and transport regulatorsdmaintain specificities in their mandate. In all cases, they operate alongside the Italian antitrust authority, which is in charge of regulation on individual cases. In exercising its role, the ART is fully independent of government, regulated companies and infrastructure operators. The requisite of independence extends to the institution as a whole, the president and the other two members of the council, and the staff, who are all bound by a strict regime of incompatibility with any other activity. In particular, the president and the members of the council have a seven-year non-renewable mandate, they may not be nominated members of the board of other independent authorities for at least five years, and they are bound to a cooling-off period of two years. The measuresdregulatory or individualdthat the Authority adopts are subject to administrative law and controls, including judicial oversight. The first decision of the Authority for the airport industry was to transpose EU Directive 2009/12/EC, which reformed the European regulatory framework for the definition of airport charges. Under this regulation, the revision of the charges results from a process of negotiation between the airport manager and the users. Should the parties not reach an agreement on the revision of the charges, either one may appeal to the independent surveillance authority to arbiter the dispute: a role that in Italy is conferred upon the Authority. In this regard, the Authority submitted to consultation and finally approved in September 2014 the Deliberation n. 64/2014. This Deliberation defines three airport categories based on the yearly volumes of passenger traffic: above 5 million, between 5 and 3 million, and below 3 million. Each category is subject to a different regulatory model, which

The impact of regulation on the airport industry: the Italian case

131

defines the procedures to be followed by airport managers for calculating access charges, for consulting the users, for solving disputes and for ensuring transparency in the flow of information to and from airport users. Regulatory constraints and procedural onuses on the airport managing company decrease with traffic volumes. Since the EU Directive does not contain any prescriptions as to which charging method to applydwhether single-till, dual-till or hybrid-tilldthe Authority considered it appropriate, in the early phases of concrete application of the models, to allow the use of the dual-till system. Accordingly, in the course of the negotiation process with the users, airport managing companies are free to assess whether and to what extent to take into account revenues resulting from commercial activities. However, as part of its supervisory activities, the Authority reserves the right to apply the corrective measures that it deems necessary (i) to promote competition with due attention to costs and profitability of the sector; (ii) to ensure adequate access to infrastructure; (iii) to encourage the productive efficiency of airport managing companies and the containment of costs for airport users where the industrial policies adopted by the managers are deemed inadequate. The need to enforce corrections on the charging system proposed by airport managers is assessed through a detailed analysis of the documentation submitted to the Authority concerning cost accounting on commercial activities, investment plans and trade policy. The thrust of the measures adopted is summarized next. • Airports above 5 million passengers/year. A price cap mechanism is adopted with productivity return on opex. The tariff and its dynamic are defined for each aviation service (e.g., passengers’ rights, handling services, takeoff and landing fees). The regulatory control period lasts 4 years. The tariff depends on admissible opex and capex (with reference to the base year), the retail price index, a rebalancing effect (so that the flow of revenues is equal to the flow of costs), a pass-through term for new investments and unexpected changes in legal or regulatory frameworks, and quality and environmental targets. The efficiency level is evaluated through a simplified approach. Opex evolves in function of the expected traffic growth, of the estimated cost elasticity per cost category and taking into account a parameter for incentives, negotiated between users and the airport manager and representing the estimated productivity growth. Capex is the regulatory asset base multiplied by WACC. The approach allows profit sharing for productivity return in

132

The Air Transportation Industry

excess of the negotiated values (according to a threshold defined in the consultation process). A procedure to mitigate the traffic risk is included (based on a threshold which is negotiated during the consultation). • Airports with traffic between 5 and 3 million passengers/year. The tariff dynamic is simplified, and the opex dynamic applies an incentive parameter equal to the 70% of the inflation rate for the first regulatory period. The approach allows profit sharing for productivity return in excess of the negotiated values. The value of WACC is obtained as for airports above 5 million passengers/year, with benchmarking for notional leverage (i.e., Debt/Equity) based on national companies belonging to the same traffic range. For the first regulatory period, the values of debt and equity are set at a level of 30% and 70%, defined through international benchmarking. • Airports below 3 million passengers/year. For airports in this range, a simplified dynamic for opex is adopted, which evolves with the inflation rate. The WACC is also determined through a simplified formula by using notional values. The costs for quality and environmental targets pass through the tariff and are negotiated with users. The regulatory approach has been slightly revised later on with Deliberation n. 92/2017, where minor interventions have been introduced on the negotiation procedure between airport managers and users as well as in setting the WACC. Lastly, with the new Deliberation n. 136/2020, the Authority further modifies the previous regulatory framework by revising the negotiation process and slightly adjusting the tariff dynamics. Moreover, it introduces a new procedure for the definition of the efficiency parameter (i.e., the X factor) within the price cap mechanism: although the Authority evaluates this parameter using a stochastic frontier analysis based on accounting data provided by airports, this parameter is not imposed by the Authority but left to the negotiation procedure prescribed by the EU Directive. The final value is thus set during the negotiation process between the airport manager and the airline companies. This implies that the final value of the X factor may vary if the parties negotiate a higher service quality or more investments. In other words, the new process leaves to a market mechanism (the negotiation) how to balance static efficiency targets and investment targets, according to the specific demand and supply characteristics of the airport. Due to the pandemic crisis that completely unsettles the industry in 2020, the new Deliberation will start being applied from July 2021.

The impact of regulation on the airport industry: the Italian case

133

The Italian regulatory framework is not dissimilar from the ones of most EU countries. The study by Steer Davies Gleave (2017) for the European Commission provides a complete comparison of the implementation of the EU Directive 2009/12/EC in 89 European airports across all countries. The majority (51%) of EU airports are now subject to the negotiation procedure previously described. However, in some cases (35%) airport charges are strictly defined by the National Regulator, and only in few cases (12%) access charges are determined after a market power assessment analysis. In terms of commercial margin, in the EU the system that is most frequently adopted is dual-till (37%), followed by single-till (21%), then hybrid-till (18%) while, to a lesser but significant extent, the commercial margin is not subject to any ex-ante intervention (15%). For example, in Spain and Germany only dual-till is adopted, in Portugal only single-till, while in Sweden there is a hybrid-till mechanism. However, different airports within the same country can be subject to different regulations. For example, in France the prevailing system is single-till, but CDG (Paris) and LYS (Lyon) have a hybrid-till instead. In the UK, single-till is adopted for Heathrow and Gatwick airports, while in the remaining airports the prevailing option is no regulation. Regarding the tariff dynamics, as reported in Table 6.1, the most prevalent regulatory mechanism used in EU is the rate of return (36% of the sample), followed by the price cap mechanism (24%) and a mix of the two systems (hybrid incentive schemes, 7%). The other airports adopt no regulation or a light-handed approach to airport charges.

4. The empirical analysis The goal of our analysis is to assess whether the ART regulation has had an impact on airports economic performances. In particular, we look at its Table 6.1 Regulatory mechanisms of EU airports. Regulatory mechanism

Number of airports

%

Rate of return Price cap No regulation Hybrid incentive schemes Light-handed regulation nd Total

32 21 19 6 4 7 89

36 24 21 7 4 8 100

134

The Air Transportation Industry

effect on airports average cost, defined as operating costs over work-load units (WLU).1 The introduction of the regulation provides a quasinatural experiment that allows us to quantify this impact within a difference-in-differences (DID) framework. This approach enables us to estimate the effect of the regulation by comparing the average change in average costs of airports that adopted the new ART models (i.e., the treatment group) with that of airports that are not yet subject to the new regulation (i.e., the control group). The advantage of the DID design is that it allows us to ascribe this difference to a causal impact of the regulationdunder the assumption that the control group provides an adequate representation of the trend that the treated group would have shown if the regulation had not been adopted. 4.1 The average cost function This empirical strategy requires the estimation of an average cost function, to which we add some variables that capture the effect of the implementation of the regulation. Cost functions in the airport industry have been the subject of a host of empirical research, such as Oum et al. (2008), Martín et al. (2009), Martini et al. (2020). Average cost functions have been estimated in several studies (Hubbard and Dawson, 1987; Filippini and Wild, 1998; Stratopoulos et al., 2000). A multi-output and multi-input average cost function can be expressed as:   ACit ¼ f y0it ; p0it ; s0it where the index i, i ¼ 1,2, ., n represents the i-th airport in the sample; the index t, t ¼ 1,2, ., T denotes the t-th period of the analysis; ACit is the dependent variable representing the average cost; y0it is the vector of the K outputs of the airport; p0it is the vector of the J inputs, s0it is the vector of the W time-varying controls. A CobbeDouglas functional form was preferred to more flexible ones (such as the translog) because of the limited number of observations that constitute our dataset (i.e., 242), as it requires to estimate fewer parameters. Expressed in logarithmic form, the average cost function is as follows: ln ACit ¼ b0 þ

K X k¼1

1

bk ln ykit þ

J X

gj ln pjit þ

j¼1

A WLU is defined as one passenger or 100 kg of cargo.

W X w¼1

dw swit þ εit

The impact of regulation on the airport industry: the Italian case

135

We defined the average cost as operating costs over WLU, thus decreasing in passengers or cargo. We include two output measures: the number of aircraft movementsdto characterize the aviation-side of the businessdand commercial revenuesdto characterize the non-aeronautical one. Other output measures were considered but were discarded due to very high multicollinearity with the other variables. We include the price of three inputs: labor, capital and other inputs (i.e., non-labor and noncapital). The price of labor was calculated as the cost of labor over the full-time equivalent (FTE). The price of capital was proxied by the sum of depreciation and financial costs over the total terminal area. The price of other inputs was defined as the sum of expenditures for materials, services and the use of third parties’ assets over the airport grounds area. We include three time-varying controls: the share of international traffic on overall traffic, the percentage of low-cost carrier (LCC) flights, and a dummy taking the value of 1 if the airport has public ownership. Table 6.2 presents the variables included in the cost function. We impose linear homogeneity in factor prices by using the price of capital as a numeraire, thus expressing the model as: ln

ACit ¼ b0 þ b1 lnðmovementsit Þ þ b2 lnðcom revit Þ price capitalit     price laborit price otherit þ g2 ln þ g1 ln price capitalit price capitalit þ d1 share intit þ d2 share lctit þ d3 publicit þ εit

Table 6.2 Definition of variables. Variable

Description

average_cost movements com_rev price_capital price_labor price_other share_int share_lct public

Average operating costs Aircraft movements Commercial revenues Price of capital Price of labor Price of services and materials Share of international traffic Share of low-cost traffic Dummy taking value 1 if the airport has public ownership

Unit of measure

V Count V V V V % % Boolean

136

The Air Transportation Industry

In the CobbeDouglas functional form, the coefficients bk can be interpreted as the cost-output elasticities, while the coefficients gj give us the cost-input price elasticities. For simplicity’s sake, from now on, we will refer to the normalized variables omitting the numeraire. We’ll further simplify the notation by referring to the deterministic part of the equation M P simply as b0 þ bm lnðdetmit Þ, with M ¼ K þ J þ W and detmit referring m¼1

to outputs, inputs and controls. 4.2 The identification approach To assess the impact of the regulation, we perform a difference-in-differences analysis (DID). We include the DID-coefficient in the average cost function to capture differential in average costs due to the implementation of the new regulation. As mentioned in Section 3, the new regulatory framework introduces a new negotiation procedure as well as a price cap mechanism with dual-till. As the previous agreements had different expiration dates, the adoption of the ART models happened at different points in time for the treated airports. We therefore adopt a generalized DID, which requires the estimation of the function: ln ACit ¼ b0 þ

M X

bm lnðdetmit Þ þ a1 artit þ mi þ mt þ εit

(6.1)

m¼1

where the determinants of the cost function detmit have the same meaning as before; artit is a binary variable that identifies the treated airports in the years of the treatment, taking the value of one from year t onwards for the airport i that in year t has adopted the new ART model; mi is an airport-fixed effect that controls for any unobserved determinant of airport costs that do not change in time within each airport; mt is a year-fixed effect that controls for any unobserved determinant of airport costs that do not change across airport within each year. As pointed out by Conti et al. (2019), including airport-level fixed effects helps to increase the comparability of control and treatment groups. The main coefficient of interest is that of the interaction term, a1 , which captures the average differential level of average costs for the airports that have adopted the regulation compared to the control ones. We do not expect that the adoption of the new ART models can have an impact on airport costs right away: we believe that its effect may require some years to manifest, as the airport adapts and responds to the new

The impact of regulation on the airport industry: the Italian case

137

regulation. This is even more so as the adoption can happen in between a year. Following the analyses of Autor (2003) and Conti et al. (2019), we take this into consideration by estimating a second model where the effect of the regulation can change over time: ln ACit ¼ b0 þ

M X

bm lnðdetmit Þ þ a1 artit þ a2 art timeit þ mi þ mt þ εit

m¼1

(6.2) This model differs from the previous one as it includes an additional term, art timeit . This variable takes the value of one in the year of adoption of the regulation by the airport i, the value of 2 after one year from the adoption, and so on. It takes the value of 0 in all other cases. This allows to study the impact of the regulation as a linear function of the time passed since its adoption by the i-th airport. This impact can therefore be calculated as the combination of a1 þ a2 art timeit , which varies with the value of art timeit . In both models, the error terms are clustered at the airport level to allow for heteroskedasticity and correlation within clusters. Due to the low number of clusters, we also estimated the models with unclustered standard errors. The results do not substantially differ.

5. Data The dataset comprises data on 22 airports from 2008 to 2018, forming a balanced panel of 11 years. The dataset was built by collecting financial data from airport companies’ financial statements through AIDA, a private database containing information on Italian companies (BVD, 2020). Data on airport operations was extracted from ENAC’s yearly reports on airport traffic and from the website of Assaeroporti, the Italian association of airport management companies (ENAC, 2020; Assaeroporti, 2020). Airport companies do not always follow the same logic in compiling their financial statements; neither do they provide the same level of detail. This is especially true for companies managing smaller regional airports. Although great care was put into making the data as comparable as possible, we acknowledge that as a possible source of bias. A further problem was given by the impossibility of attributing costs to a specific airport where a company controlled more than one. For the cases of Rome Fiumicino and Rome Ciampino; Milan Linate and Milan Malpensa; Florence and Pisa, this

138

The Air Transportation Industry

forced us to consider them as a single airport.2 As pointed out by Curi et al. (2008), this is not problematic since they constitute an airport system with the same management and catchment area. In cases where the differences among the airports belonging to a single group were too large (e.g., Aeroporti di Puglia, which controls four airports in the Apulia region), we were forced to exclude them from the analysis. All the monetary variables have been converted to 2006 prices by using the annual consumer price indices published by the Italian National Institute of Statistics (ISTAT). Table 6.3 presents the descriptive statistics for the variables included in the model. Although Italy has 42 airports open to commercial traffic, our sample of 22 represents 93% of aircraft movements and passenger traffic and 98% of cargo volume, considering 2018 data. Fig. 6.1 shows traffic trends for the years of the analysis. Among the airports in our sample, 12 of them have adjusted their aeronautical charges in compliance with the regulatory models introduced by ART with Resolution 64/2014 during the time of our analysis. Data on the time of their adoption has been gathered from ART’s website, which publishes all deliberations announcing adjustments to airport charges (ART, 2020). Table 6.4 reports the year in which each treated airport has adopted the new incentive regulation scheme.

6. Empirical results This section presents and discusses the results of the analysis. Section 6.1 estimates the two models represented by Eqs. (6.1) and (6.2), which quantify the impact that the regulation has had on the affected airports’ average costs. Section 6.2 tests the solidity of the results obtained by introducing two robustness checks. 6.1 Baseline We first estimate the static model as per Eq. (6.1). Table 6.5 reports the estimated coefficients: fist without including the controlsdcolumn (1)d then with the time-varying controlsdcolumn (2). We can see that the average cost function is well specified as it is decreasing in outputs and shows significant input prices. The results are 2

The analysis was repeated excluding the airport systems of Rome and Milan, leading to very similar resultsdas reported in Appendix.

Average cost Movements Commercial revenues Price of labor Price of capital Price of other inputs Share of international traffic Share of low-cost traffic Share of public ownership Observations

mean

sd

min

max

p25

p50

p75

11.04 57,250 26,985,899 45,641 159 70,775 46% 49% 59% 242

3.86 84,759 54,660,131 8485 128 51,425 25% 25% 36%

5.03 2837 229,415 21,012 16 1912 0% 0% 0%

28.02 392,246 270,101,351 79,088 680 321,555 87% 100% 100%

8.27 12,262 2,218,490 41,521 73 31,335 25% 31% 25%

9.85 26,591 7,466,633 46,984 136 65,069 45% 46% 58%

13.11 63,400 22,918,956 49,913 185 99,769 70% 65% 98%

The impact of regulation on the airport industry: the Italian case

Table 6.3 Descriptive statistics.

139

140

The Air Transportation Industry

Figure 6.1 Traffic trends. Table 6.4 Timeline of regulation adoption. Airport

Olbia, Toscana Bologna, Cagliari, Catania, Genova, Lamezia, Napoli, Palermo, Torino, Trieste Bergamo

Year of adoption

2015 2016 2017

stable to the addition of the controls. The number of aircraft movements has a statistically significant coefficient as expected. Its negative sign implies that average cost tends to decrease the higher is the number of movements, i.e., the larger is the scale of the airport. The share of low-cost carriers (share_lct in the table) is significant and with a negative sign: an effect that is in line with previous literature (e.g., Martini et al., 2020). The share of international passengers (share_int in the table)don the other handdis not significant. We find evidence of lower average costs per movement on airport companies having prevalently public ownership (public in the table)din line with Curi et al. (2010) and Scotti et al. (2012), but in contrast with Martini et al. (2020). The coefficient of the variable art, although negative, is not statistically significant: we cannot infer with a

The impact of regulation on the airport industry: the Italian case

Table 6.5 Baseline estimates. Average costs (Log) (1)

movements com_rev price_labor price_other art

(2)

(3)

(4)

0.631** (0.115) 0.106 (0.0809) 0.400** (0.0603) 0.429** (0.0543) 0.0262 (0.0333)

0.607** (0.104) 0.108 (0.0637) 0.372** (0.0546) 0.456** (0.0537) 0.0143 (0.0359)

0.885 (2.271) 242 No Yes Yes

0.0705 (0.284) 0.514* (0.221) 0.0672* (0.0315) 0.949 (1.764) 242 Yes Yes Yes

0.609** (0.110) 0.105 (0.0785) 0.409** (0.0567) 0.424** (0.0512) 0.0802 (0.0486) 0.0561* (0.0214)

0.585** (0.104) 0.107* (0.0615) 0.382** (0.0515) 0.450** (0.0506) 0.117* (0.0504) 0.0551* (0.0226) 0.0392 (0.280) 0.522* (0.222) 0.0588* (0.0285) 0.657 (1.687) 242 Yes Yes Yes

art_time share_int share_lct public constant Observations Controls Fixed effects - airport Fixed effects - year

141

0.612 (2.166) 242 No Yes Yes

Notes: This table presents the estimates of Eqs. (6.1) and (6.2). The main coefficients of interest are those of art and art_time. art is a dummy variable taking the value of one from the year of adoption of the regulation by airport i onwards, zero otherwise. art_time takes the value of one on the year of adoption of the regulation by airport i, 2 after one year, and so on. It takes the value of zero in all other cases. Standard errors are shown in parenthesis, clustered at the airport level. ***, **, and * indicate statistical significance at the 1%, 5% and 10% levels, respectively.

reasonable level of certainty that this coefficient is different from zero. However, as discussed in Section 4.2, we can reasonably expect that any impact that the regulation could have an effect in years after its first adoption. We explore whether the time since the implementation of the regulation is an important factor for airports average costs by estimating Eq. (6.2). This model allows for a differential effect of the new incentive regulation that varies according to the number of years elapsed since its adoption by airport i. This effect is expressed in the model by the coefficient of the art timeit term. The results of this estimation are reported in

142

The Air Transportation Industry

columns (3) and (4) of Table 6.5, which estimate the model respectively without and with controls. Column (3) shows that the art_time coefficient is negative and significant at a 10% level, while art is still not significant. This implies a progressive cost reduction starting from the adoption of the new regulation. In column (4), which includes the controls, this finding is confirmed while art also becomes significant. This shows that there is an increase in costs sustained by the airport at the moment of the adoption of the regulation of about, but also a yearly decrease as time goes by. Both effects are statistically significant at the 10% level. These findings can be attributed to the effect of the adoption of the new models, which raises the airport’s costs as it has to adapt to the new regulation, and then to an efficiency increase due to the effect of the incentive mechanism. The second output measure, commercial revenues, also becomes significant in this specification. We see that the regulation has had a dual effect, with its two components leading to opposite directions. We further investigate this dynamic by disentangling the aggregated effect of the regulation into yearly components. We use the estimated coefficients of columns (2) and (4) of Table 6.5 to calculate the annual combined effect for any year after the adoption of the regulation, which equates to computing the linear combination a1 artit þ a2 art timeit . We report these results in Table 6.6. Table 6.6 Effect of the regulation depending on the time elapsed since its adoption. Years of implementation Without controls With controls

Year 0 Year 1 Year 2 Year 3

(1)

(2)

0.0241 (0.0357) 0.0319 (0.0333) 0.0880* (0.0431) 0.1440** (0.0594)

0.0623^ (0.0380) 0.0072 (0.0368) 0.0479 (0.0478) 0.1030^^ (0.0650)

Notes: This table shows the estimated impact of the regulation on airport average cost depending on the years passed since its adoption. Estimates in column (1) are based on the specification reported in column (2) of Table 6.5. Estimates in column (2) are based on the specification reported in column (4) of Table 6.5. Standard errors are shown in parenthesis and are clustered at the airport level. ***, **, and * indicate statistical significance at the 1%, 5% and 10% levels, respectively. Coefficients marked with^and^^ have a P-value of 0.12 and 0.13, respectively.

The impact of regulation on the airport industry: the Italian case

143

Column (1) shows an effect of cost reduction that is statistically significant at the 5% level from the second year since the adoption of the ART models. When adding the controls, the coefficients for year 0 and three are almost significant, even though below the standard values (the P-value is around 0.12). These results seem to suggest that the regulation indeed has had a positive impact on airports average costs, with an initial slight increase and a progressive decrease over the years. However, the statistical significance of the estimates requires us to be cautious in assessing these results. 6.2 Robustness checks We test the solidity of our analysis by performing some robustness checks. Eq. (6.2) assumes that the effect of the incentive regulation on airport costs is linear: the effect of a2 does not change within years. Following Autor (2003) and Conti et al. (2019), we relax this assumption by estimating a model where these effects are free to change on a per-year basis. This more general specification takes the following form: ln ACit ¼ b0 þ

M X m¼1

bm lnðdetmit Þ þ

3 X

a1 artitþj þ mi þ mt þ εit

(6.3)

j¼0

where artitþj is a variable that takes the value of 1 j years after airport i has adopted the regulation, the value of 0 otherwise. Note that the variable is equal to one only in the relevant year. We set J ¼ 3 as it covers the whole four years (including the first one, J ¼ 0) of the new regulatory scheme. We estimate the coefficients of Eq. (6.3) in columns (1) and (2) of Table 6.7. In both the specification without and the one with controls, we find a statistically significant effect at the 5% level after three years from the adoption of the ART regulation. We believe that this specification is already stretching the limit of what can be done with such a database, as the number of clusters and of observations limits the number of parameters that we can reasonably estimate. Moreover, the quality of the datadwhich comes mainly from financial statements, as discussed in Section 5dadvises us to look skeptically at the magnitude of the effect estimated. Nevertheless, the results from columns (1) and (2) of Table 6.6 are in line with the expected result of a delayed effect of cost reduction. We then perform another test to assess whether one of the main assumptions of any DID analysis holds: that of parallel trends for the treated

144

The Air Transportation Industry

Table 6.7 Robustness tests. Average costs (Log) (1)

movements com_rev price_labor price_other

(2)

(3)

(4)

0.599** (0.109) 0.107 (0.0784) 0.422** (0.0557) 0.412** (0.0493)

0.573** (0.101) 0.109* (0.0600) 0.397** (0.0497) 0.437** (0.0483)

0.00680 (0.0364) 0.0266 (0.0430) 0.0463 (0.0404) 0.300** (0.0867)

0.0420 (0.0366) 0.0158 (0.0471) 0.0000493 (0.0368) 0.286* (0.104) 0.0244 (0.276) 0.542* (0.227) 0.0576* (0.0282) 0.231 (1.676) 242 Yes Yes Yes

0.599** (0.108) 0.0967 (0.0785) 0.427** (0.0553) 0.414** (0.0477) 0.0400 (0.0348) 0.0705 (0.0420) 0.0191 (0.0469) 0.0490 (0.0484) 0.0711 (0.0469) 0.328** (0.0902)

0.575** (0.0983) 0.102 (0.0612) 0.400** (0.0499) 0.437** (0.0479) 0.0251 (0.0315) 0.0416 (0.0371) 0.0244 (0.0460) 0.000122 (0.0527) 0.0175 (0.0448) 0.304* (0.108) 0.00363 (0.285) 0.530* (0.224) 0.0589* (0.0264) 0.113 (1.726) 242 Yes Yes Yes

artit2 artit1 artitþ0 artitþ1 artitþ2 artitþ3 share_int share_lct public constant Observations Controls Fixed effects - airport Fixed effects - year

0.203 (2.156) 242 No Yes Yes

0.00194 (2.170) 242 No Yes Yes

Notes: This table presents the estimates of Eqs. (6.3) and (6.4). The coefficients of artit2 to artitþ3 are dummy variables taking the value of one in year t þ j, and zero otherwise. Standard errors are shown in parenthesis and are clustered at the airport level. ***, **, and * indicate statistical significance at the 1%, 5% and 10% levels, respectively.

and control groups prior to the exogenous shock (i.e., the adoption of the incentive regulation). To do this, we follow the method adopted by Autor (2003) and Conti et al. (2019) of introducing a lead of our treatment variable. This results in a model very similar to Eq. (6.3), which is as follows:

The impact of regulation on the airport industry: the Italian case

ln ACit ¼ b0 þ

M X m¼1

bm lnðdetmit Þ þ

3 X

a1 artitþj þ mi þ mt þ εit

145

(6.4)

j¼2

where the lead term is captured by artitþj when j is a negative number. A statistically significant coefficient for artit1 or artit2 would imply a violation of the parallel trends assumption, as we would have an anticipatory effect. Columns (3) and (4) of Table 6.6 report the results of the estimation of this model. We can see how these coefficients are not statistically significant in both specifications, which suggests that the parallel trends assumption is not violated. This finding strengthens the solidity of the results from Section 6.1.

7. Conclusions Analyzing the impact of a regulatory change on a regulated company, such as an airport, is generally made complex by the presence of multiple effects that have to be taken into account. In this work, we try to provide a method to study the impact that a change in regulation has had on the average costs of Italian airport managing companies. While we apply it to this specific case, this method can be used to assess the effect of a change in regulation in any country. Such an approach, however, requires a change in regulation that only affects a subset of airportsdor that affects them at different points in timedwhich could reduce its applicability. We consider the Italian airport system not only because of our better knowledge about this industry, but also because in 2013 we observe a structural change in the regulatory intervention due to the establishment of a newly appointed independent regulatory Authority. The new regulation adopted in 2014, following EU 2009 Directive, introduced a negotiation procedure for airport charges alongside a price cap mechanism with dual-till. This event provides a quasi-natural experiment that allows us to quantify its impact within a difference-in-differences framework. We find some evidence of an effect of cost reduction articulated in two components: an initial increase in cost, followed by a decrease over the years. We ascribe this effect to the change in regulation, which has had a positive effect on the overall efficiency of the Italian airport system. While a larger and higher quality dataset could overcome some of the limitations of

146

The Air Transportation Industry

the studydand it could allow for a more precise quantification of the effectdthe analysis provides a framework that can be useful to assess the impact of the implementation of policies in the airport industry in any country.

Appendix See Tables I 1eI 3.

Table I 1 Baseline estimates excluding Rome and Milan. Average costs (Log) (1) (2) (3)

movements com_rev price_labor price_other art

0.613** (0.110) 0.125 (0.0836) 0.404** (0.0620) 0.432** (0.0574) 0.0231 (0.0386)

0.589** (0.102) 0.128* (0.0630) 0.373** (0.0564) 0.465** (0.0564) 0.0169 (0.0388)

0.806 (2.209) 220 No Yes Yes

0.0537 (0.299) 0.529* (0.230) 0.0641* (0.0348) 0.889 (1.700) 220 Yes Yes Yes

art_time share_int share_lct public constant Observations Controls Fixed effects - airport Fixed effects - year

0.584** (0.101) 0.124 (0.0806) 0.417** (0.0566) 0.425** (0.0527) 0.0960* (0.0496) 0.0639** (0.0223)

0.474 (2.070) 220 No Yes Yes

(4)

0.560** (0.102) 0.127* (0.0602) 0.387** (0.0517) 0.456** (0.0520) 0.131* (0.0520) 0.0628* (0.0234) 0.00888 (0.293) 0.539* (0.231) 0.0532* (0.0302) 0.530 (1.601) 220 Yes Yes Yes

Notes: This table presents the estimates of Eqs. (6.1) and (6.2) having excluded the observations for Rome and Milan. The main coefficients of interest are those of art and art_time. art is a dummy variable taking the value of one from the year of adoption of the regulation by airport i onwards, zero otherwise. art_time takes the value of one on the year of adoption of the regulation by airport i, 2 after one year, and so on. It takes the value of zero in all other cases. Standard errors are shown in parenthesis, clustered at the airport level. ***, **, and * indicate statistical significance at the 1%, 5% and 10% levels, respectively.

The impact of regulation on the airport industry: the Italian case

147

Table I 2 Effect of the regulation depending on the time elapsed since its adoption, excluding Rome and Milan. Years of implementation Without controls With controls

Year 0 Year 1 Year 2 Year 3

(1)

(2)

0.0321 (0.0390) 0.0319 (0.0396) 0.0958* (0.0511) 0.1597** (0.0682)

0.0681^ (0.0405) 0.0053 (0.0410) 0.0575 (0.0530) 0.1203^ (0.0710)

Notes: This table shows the estimated impact of the regulation on airport average cost depending on the years passed since its adoptiondhaving excluded the observations for Rome and Milan. Estimates in column (1) are based on the specification reported in column (2) of Table 6.5. Estimates in column (2) are based on the specification reported in column (4) of Table 6.5. Standard errors are shown in parenthesis and are clustered at the airport level. ***, **, and * indicate statistical significance at the 1%, 5% and 10% levels, respectively. Coefficients marked with^have a P-value of 0.11.

Table I 3 Robustness tests excluding Rome and Milan. Average costs (Log) (1) (2)

movements com_rev price_labor price_other

0.575** (0.101) 0.126 (0.0808) 0.430** (0.0553) 0.412** (0.0506)

0.549** (0.0993) 0.129* (0.0586) 0.402** (0.0498) 0.442** (0.0496)

0.0155 (0.0389) 0.0268 (0.0505) 0.0531 (0.0504) 0.312** (0.0923)

0.0481 (0.0391) 0.0145 (0.0522) 0.00904 (0.0428) 0.300* (0.110)

artit2 artit1 artitþ0 artitþ1 artitþ2 artitþ3

(3)

(4)

0.573** (0.0988) 0.116 (0.0795) 0.432** (0.0549) 0.420** (0.0479) 0.0783 (0.0462) 0.0509 (0.0378) 0.0174 (0.0526) 0.0561 (0.0573) 0.0856 (0.0587) 0.346** (0.0964)

0.550** (0.0958) 0.120* (0.0588) 0.404** (0.0494) 0.446** (0.0480) 0.0539 (0.0428) 0.0391 (0.0349) 0.0215 (0.0518) 0.0105 (0.0590) 0.0367 (0.0532) 0.329** (0.114) Continued

148

The Air Transportation Industry

Table I 3 Robustness tests excluding Rome and Milan.dcont'd Average costs (Log)

(1)

(2)

0.404 (2.102) 220 No Yes Yes

0.00475 (0.289) 0.558* (0.236) 0.0523* (0.0300) 0.437 (1.617) 220 Yes Yes Yes

share_int share_lct public constant Observations Controls Fixed effects - airport Fixed effects - year

(3)

(4)

0.165 (2.103) 220 No Yes Yes

0.0556 (0.303) 0.547* (0.232) 0.0533* (0.0282) 0.255 (1.661) 220 Yes Yes Yes

Notes: This table presents the estimates of Eqs. (6.3) and (6.4) having excluded the observations for Rome and Milan. The coefficients of artit2 to artitþ3 are dummy variables taking the value of one in year t þ j, and zero otherwise. Standard errors are shown in parenthesis and are clustered at the airport level. ***, **, and * indicate statistical significance at the 1%, 5% and 10% levels, respectively.

References Adler, N., Forsyth, P., Mueller, J., Niemeier, H.-M., 2015. An economic assessment of airport incentive regulation. Transp. Pol. 41, 5e15. https://doi.org/10.1016/ j.tranpol.2015.03.008. ART, 2020. Delibere. Autorità di Regolazione dei Trasporti. URL. https://www.autoritatrasporti.it/indice-delibere/ (Accessed 11.5.20). Assaeroporti, 2020. Dati Annuali. URL. https://assaeroporti.com/dati-annuali/ (Accessed 11.13.20). Autor, D.H., 2003. Outsourcing at will: the contribution of unjust dismissal doctrine to the growth of employment outsourcing. J. Labor Econ. 21, 1e42. https://doi.org/ 10.1086/344122. Barros, C.P., 2008a. Technical efficiency of UK airports. J. Air Transp. Manag. 14, 175e178. https://doi.org/10.1016/j.jairtraman.2008.04.002. Barros, C.P., 2008b. Technical change and productivity growth in airports: a case study. Transp. Res. Pol. Pract. 42, 818e832. https://doi.org/10.1016/j.tra.2008.01.029. Bel, G., Fageda, X., 2010. Privatization, regulation and airport pricing: an empirical analysis for Europe. J. Regul. Econ. 37, 142e161. https://doi.org/10.1007/s11149-009-9110-7. Bilotkach, V., Clougherty, J.A., Mueller, J., Zhang, A., 2012. Regulation, privatization, and airport charges: panel data evidence from European airports. J. Regul. Econ. 42, 73e94. https://doi.org/10.1007/s11149-011-9172-1. BVD, 2020. Aida e Italian Company Data [WWW Document]. URL. https://www. bvdinfo.com/en-gb/our-products/data/national/aida (Accessed 11.13.20). Conti, M., Ferrara, A.R., Ferraresi, M., 2019. Did the EU Airport Charges Directive lead to lower aeronautical charges? Empirical evidence from a diff-in-diff research design. Econ. Transp. 17, 24e39. https://doi.org/10.1016/j.ecotra.2018.12.001.

The impact of regulation on the airport industry: the Italian case

149

Curi, C., Gitto, S., Mancuso, P., 2011. New evidence on the efficiency of Italian airports: a bootstrapped DEA analysis. Soc. Econ. Plann. Sci. 45, 84e93. https://doi.org/10.1016/ j.seps.2010.11.002. Curi, C., Gitto, S., Mancuso, P., 2010. The Italian airport industry in transition: a performance analysis. J. Air Transp. Manag. 16, 218e221. https://doi.org/10.1016/ j.jairtraman.2009.11.001. Curi, C., Gitto, S., Mancuso, P., 2008. Un’applicazione della Data Envelopment Analysis (DEA) per la misurazione dell’efficienza degli aeroporti italiani dopo la privatizzazione del settore. L’industria 29, 689e712. Czerny, A.I., 2006. Price-cap regulation of airports: single-till versus dual-till. J. Regul. Econ. 30, 85e97. https://doi.org/10.1007/s11149-006-0010-9. ENAC, 2020. Dati di traffico [WWW Document]. URL. http://www.enac.gov.it/ trasporto-aereo/compagnie-aeree/dati-di-traffico (Accessed 11.13.20). Filippini, M., Wild, J., 1998. The Estimation of an Average Cost Frontier to Calculate Benchmark Tariffs for Electricity Distribution (Working Paper No. 9803). Working paper series/Socioeconomic Institute. University of Zurich. https://doi.org/10.5167/ uzh-51848. Gillen, D., Mantin, B., 2014. The importance of concession revenues in the privatization of airports. Transp. Res. E Logist. Transp. Rev. 68, 164e177. https://doi.org/10.1016/ j.tre.2014.05.005. Hubbard, L.J., Dawson, P.J., 1987. Ex ante and ex post long-run average cost functions. Appl. Econ. 19, 1411e1419. https://doi.org/10.1080/00036848700000128. Ivaldi, M., Sokullu, S., Toru, T., 2015. Airport Prices in a Two-Sided Market Setting: Major US Airports (SSRN Scholarly Paper No. ID 2619233). Social Science Research Network, Rochester, NY. Lo Storto, C., 2018. Ownership structure and the technical, cost, and revenue efficiency of Italian airports. Util. Pol. 50, 175e193. https://doi.org/10.1016/j.jup.2018.01.003. Malavolti, E., 2016. Single till or dual till at airports: a two-sided market analysis. Transportation research procedia. Transp. Res. Arena 14, 3696e3703. https://doi.org/ 10.1016/j.trpro.2016.05.489. Martín, J.C., Román, C., Voltes-Dorta, A., 2009. A stochastic frontier analysis to estimate the relative efficiency of Spanish airports. J. Prod. Anal. 31, 163e176. https://doi.org/ 10.1007/s11123-008-0126-2. Martini, G., Scotti, D., Viola, D., Vittadini, G., 2020. Persistent and temporary inefficiency in airport cost function: an application to Italy. Transp. Res. Pol. Pract. 132, 999e1019. https://doi.org/10.1016/j.tra.2019.12.016. Oum, T.H., Yan, J., Yu, C., 2008. Ownership forms matter for airport efficiency: a stochastic frontier investigation of worldwide airports. J. Urban Econ. 64, 422e435. https://doi.org/10.1016/j.jue.2008.03.001. Pels, E., Nijkamp, P., Rietveld, P., 2003. Inefficiencies and scale economies of European airport operations. Transp. Res. E Logist. Transp. Rev. 39, 341e361. https://doi.org/ 10.1016/S1366-5545(03)00016-4. Sciandra, L., 2009. Regulatory Assessment of Italian Airports: Some Critical Issues. MC. https://doi.org/10.1434/29270. Scotti, D., Malighetti, P., Martini, G., Volta, N., 2012. The impact of airport competition on technical efficiency: a stochastic frontier analysis applied to Italian airport. J. Air Trans. Manag. 22, 9e15. https://doi.org/10.1016/j.jairtraman.2012.01.003, 14th Air Transport Research Society Conference.

150

The Air Transportation Industry

Steer Davies, G., 2017. Support Study to the Ex-Post Evaluation of Directive 2009/12/EC on Airport Charges: Final Report. (No. Study Contract No. MOVE/E1/SER/2016592/SI2.748561). European Commission Directorate-General for Mobility and Transport, Luxembourg. https://doi.org/10.2832/926068. Stratopoulos, T., Charos, E., Chaston, K., 2000. A translog estimation of the average cost function of the steel industry with financial accounting data. Int. Adv. Econ. Res. 6, 271e286. https://doi.org/10.1007/BF02296108.

CHAPTER 7

Airline pricing, incumbents, and new entrants Rosário Macário1, 2 and Eddy Van de Voorde2 1

CERIS, Instituto Superior Técnico, Universidade de Lisboa, Lisbon, Portugal; 2Department of Transport and Regional Economics, University of Antwerp, Antwerp, Belgium

1. Introduction Pricing remains one of the most important and most difficult tasks in business management. For transport companies, and particularly for airlines, pricing is highly complex, due to major differences in terms of transport performance and the impossibility of stock production. Airlines generally charge many different prices at the same time for transport performance that is the same in terms of distance. Such wide variations in pricing might indeed be part of an optimal strategy. The variations can be based on origin and destination, taking into account the circumstances under which incumbent the transport takes place, even with large fluctuations over time. Prices can thus vary drastically from one daydor even hourdto the next. Such complexity in pricing is due primarily to the inability to deliver from stock, combined with the stochastic nature of fluctuations in demand over time. Another factor that contributes to pricing complexity is the lack of cost orientation. Exponential increases in the complexity of pricing have been observed ever since pricing began to be based on differences between customer segments with regard to their willingness to pay. One practical implication has been the need to take probability processes into account. Under normal economic circumstances, the primary criterion continues to be the objective of profit maximization (Blauwens et al., 2020). As with all transport companies, the availability of useful data is also a problem for airlines. In practice, the only exact data available have to do with cost calculation. Moreover, cost calculations include the addition of a margin, the size of which often remains an intuitive decision. Such

The Air Transportation Industry ISBN 978-0-323-91522-9 https://doi.org/10.1016/B978-0-323-91522-9.00006-3

© 2022 Elsevier Inc. All rights reserved.

151

152

The Air Transportation Industry

management intuition often makes the difference between success and failure, including with regard to the fine-tuning of the market and the accuracy of assessments of the traveler’s willingness to pay. The body of scientific economic literature on profit maximization in airline pricing is not extensive. The most challenging element in such price calculations is waiting time. In the current case, waiting time can also emerge from aircraft moves, which are non-storable, meaning that delivery cannot be made from stock. In other words, deliveries are impossible in the absence of sufficient production capacity. It is not always possible to coordinate production capacity to the wishes and expectations of individual travelers. For airlines, pricing thus constitutes a crucial tool for competition. Although prices must be at least high enough to recover costs, they should ideally contribute to achieving the objective of profit maximization. If prices are too high, there is a risk that the competition will gain market share. If they are too low prices, there is a risk of jeopardizing the relationship of trust with the final customer. Pricing is thus a crucial element in any competitive strategy. This chapter examines the importance of pricing for airlines in greater detail. It begins with a brief literature review and an overview of the theory underlying aviation pricing. This is followed by a detailed discussion focusing on one example of deviant behavior, in which an incumbent makes effective use of pricing as a barrier to entry. The discussion explores the extent to which this type of behavior can be generalized and which alternative strategies might be possible.

2. Pricing principles: theory and literature This section is divided into three blocks. The first block provides a brief description of the historical situation, followed by a presentation of the theoretical foundation of pricing in all its aspects, particularly as applied to the aviation industry. The third block provides a more detailed exploration of the most recent trends in pricing in the aviation sector, based on recent scientific literature, amongst other sources. 2.1 A fault line in aviation history Prior to 1978, the aviation industry in the United States was highly regulated, with schedules and rates strictly monitored by the Civil Aeronautics Board. The rates were high enough to guarantee an acceptable

Airline pricing, incumbents, and new entrants

153

return on investment. The situation was similar throughout the rest of the world (Phillips, 2005, p. 121). The Airline Deregulation Act of 1978 called for the deregulation of the airline industry by 1982. The act aimed to achieve the complete elimination of restrictions on domestic routes and new services (with a target date of December 31, 1981) and the elimination of all fare regulations beginning on January 1, 1983. One of the drivers underlying this call for deregulation was the objective of encouraging creative new opportunities for market entry, with low rates. This was expected to force incumbents to lower their rates only if they also lowered their cost base. The result would be beneficial to the ultimate customer, while create new market segments that could expand the overall dimensions of the “revenue pie.” One convincing argument was that deregulation would generate a win-win situation. One incumbent in these developments was American Airlines, which was facing stiff competition from the newcomer PeopleExpress at that time. American Airlines managed to compete with the newcomer by offering low prices as well, while selling some of their seats at higher rates. This plan was based on an ingenious pricing system, in which the airline divided the market between leisure and business travelers. In addition, tiered prices were used to create new artificial markets (thus expanding the pie) and to attack the competitor. This strategy can be regarded as the beginning of what would later be referred to as “revenue management” (Phillips, 2005, p. 122). Before deregulation, the industry was protected by national governments throughout the world, and there was a lack of full competition. Within these regulated markets, most airlines used relatively static pricing mechanisms, resulting in simple pricing structures based primarily on distance. The aforementioned deregulation and liberalization of the market ushered in greater competition, thereby forcing airlines to pay more attention to cost efficiency, cost recovery and profitability. Following the example of American Airlines, many airlines introduced much more finegrained and dynamic price-discrimination systems (Van de Voorde, 1992, p. 521e522; Starkie, 2008, p. 21e24). This development generated changes in the pricing strategies that, in many markets, led to decreasing operating costs and average ticket prices. The ultimate result was significant growth in air traffic due to price stimulation, as reflected in declining actual yields (Holloway, 2003, p. 130).

154

The Air Transportation Industry

2.2 Theoretical foundations of pricing in aviation This sub-section provides a closer examination of the economic principles, ratio and competitive forces that can help to explain the evolution of airline fare structures. These developments can be summarized as a transition from more traditional restricted approaches to simplified rates. In practice, although different airlines can have different fare structures, the basic principles and range of options remain the same. In the pricing process, airline pricing and revenue management are often mentioned in the same breath. There is nevertheless a clear distinction between the two concepts. Pricing involves the process of determining fare levels, along with various service features and restrictions, resulting in a range of fare products within an origin-destination (O-D) market. Revenue management is the associated process of determining how many seats should be made available at each rate level (Belobaba, 2016, p. 75e125), thereby optimizing the final revenue. One important basic principle is that airline fares should be defined for specific O-D markets, and not for specific airline flight legs. Travelers can choose from several options, ranging from non-stop to various numbers of connecting flights. At the same time, a single flight can serve several O-D markets, each with its own set of prices. The ultimate prices will be a function of supply and demand, as well as of competition within clearly separate markets and sub-markets, in which price differences can be explained in part by such factors as distance and the services offered. When setting prices, airlines can apply either cost-based or service-based economic principles. In practice, most airline pricing strategies reflect a mix of these three theoretical principles (i.e., O-D market, supply/demand and cost/service). The competitive environment also bears a strong influence on prices. Most airline fare structures reflect the “differential pricing” features of “price discrimination” and “product differentiation” (Belobaba, 2016, p. 79). Price discrimination essentially consists of responding to different levels of “willingness to pay” on the part of consumers, by charging different prices for the same (or very similar) product with the same production cost (Lipczynski et al., 2009, p. 321e331). Product differentiation consists of charging different prices for products with different quality-ofservice characteristics and the different production costs derived from them (Lipczynski et al., 2009, p. 383e394). Travelers are offered a set of

Airline pricing, incumbents, and new entrants

155

fare-product options, balancing the inconvenience of fare restrictions and lower fares against the higher prices of tickets without restrictions. One major challenge faced by airlines involves designing mechanisms that can prevent consumers who would be willing to pay high prices from being diverted to less-expensive fare classes. This involves avoiding imperfect segmentation, cannibalization and arbitrage (Phillips, 2005, p. 77). The ultimate goal for management agents is obvious: to offer potential customers a sufficiently wide range of fare-product options at differing price levels, thereby effectively capturing greatest possible share of the surface area below the demand curve. Each fare product is aimed at a specific demand segment, which is characterized by a specific willingness-to-pay level. One major challenge in this regard involves establishing a series of fare-product restrictions that can avoiddor at least counteractdthe potential flow of customers from more-expensive to less-expensive fare classes. In practice, the fundamental challenge of airline pricing also consists of knowing and systematically making use of the differing price elasticities for each sub-class of potential travelers. Knowledge of customer preferences is critical to the development of pricing strategies, as it makes it possible to stimulate the demand of leisure travelers with low rates, without distracting business travelers from paying higher rates. Pricing strategies also make the greatest possible use of information on developments in demand over time (e.g., by day, hour, week, or season), such that pricing also responds to factors including the cyclical variability of demand. The principles of demand segmentation and differential pricing described here thus constitute the foundation for the evolution of the airline fare structures, as currently occurring throughout the world. All of these principles are applied within a framework aimed at maximizing the total revenue generated, as linked to specific flight schedules and fixed flight capacity. To this end, the next step in the pricing process is revenue management, which affects the prices that travelers pay for particular flight departures. Revenue management refers to the process of determining the number of seats to be made available for each fare class on a specific flight, given a fare structure that offers a variety of different prices with different travel characteristics within the same O-D market. These “artificial markets” are created in order to defeat the competition.

156

The Air Transportation Industry

The objective of price and revenue management is to set and update the price of any combination of product, customer segment, and channel. The approach used must be geared to the specifications of the underlying market structure. Based on the philosophy of profit maximization, the ultimate goal of revenue management is always to keep a certain number of seats available for business travelers who will book later, at higher rates. This is done with systems aimed at filling each available seat on each future flight departure with the traveler who will contribute to the highest possible income. Revenue management applies under the following conditions: (1) the airline sells a fixed stock of non-storable capacity; (2) all passengers book their seats before departure; (3) the seller manages a series of rate classes, each of which has a fixed price (at least in the short term); (4) the seller may change the availability of rate tiers over time (Phillips, 2005, p. 120). Revenue management can be regarded as a special case of limited-supply pricing (Phillips, 2005, p. 120). The method and approach became necessary in response to the development of differential fare practices in which the same seat on a flight could be sold at different prices (Belobaba, 2016, p. 99). In practice, several generations of computerized revenue-management systems have been created and used effectively. In fact, the thirdgeneration approach to revenue-management systems was designed to maximize the revenue (or, more specifically, the yield) of each flight leg, but not necessarily the total revenue on the airline’s network. The next (i.e., fourth) generation will thus direct more attention toward revenue management at the level of the airline network (Belobaba, 2016, p. 112), where code-sharing strategies are used to increase revenues without the corresponding increase in costs. 2.3 Recent trends in airline pricing The entry and rapid growth of low-fare airlines in the late 1990s led to relatively structural changes in airline pricing strategies, especially involving a shift toward more simplified fare structures with fewer restrictions. This pushed airline pricing to return to the basics of fare-product differentiation as a way for all types of airlines to increase their revenues in the face of rising costs (Belobaba, 2016, p. 89). The results of these developments have included new pricing strategies aimed at cultivating various “branded-fare families” that offer travelers

Airline pricing, incumbents, and new entrants

157

explicit differences in services and travel conditions, accompanied by the return of old restrictions (e.g., such as minimum stays and round-trip purchase requirements). In an effort to increase revenues, new rates are being introduced, and airlines are shifting toward “à la carte” pricing (also known as the “unbundling” of rates). Additional efforts are being made to generate ancillary revenues, including the sale of food and beverages on board as a means of increasing profits. For many airlines, ancillary fees have become a central component of revenue maximization (Belobaba, 2016, p. 93). Within the context outlined earlier, it could be extremely instructive to gain insight into future trends in airline pricing and revenue management. Such insight can be generated in several ways. For example, stakeholders are constantly working on projects, and the results could potentially be used to derive trends. Another way to develop the needed insight could involve the synthesis of current scientific research in the field of pricing and revenue management. As technology becomes more sophisticated, airlines will be able to elaborate their pricing strategies with much greater accuracy. Five trends that are having (and will continue to have) an enormous impact on pricing and revenue management in the future are presented in Table 7.1. The next steps that can be expected in the pricing process in the short to medium term are unclear. It is not easy to gain insight in this regard, given the confidential nature of such strategic information. Current research nevertheless offers some insight into possible future strategies. Scientific research is translated into scientific publications. We therefore analyzed a number of important publications emerging in the period 2016e20. Our analysis focusses more on information relating to pricing and revenue management than it does on the relevance of the methodology used and/or the empirical results generated. One important line of research involves the calculation of costs in prices. In the past decade, airline costs have been highly volatile, due primarily to wide fluctuations in jet fuel prices. This has forced airlines and regulators to consider the extent to which airlines can pass these cost changes along to consumers through pricing. Economic theory is quite clear in this regard: the ability to pass costs along to consumers is strongly dependent on the type of cost increase (e.g., sector-wide or company-specific) and the

158

The Air Transportation Industry

Table 7.1 Trends in pricing and revenue management. Trends Explanation

Passenger profiling

Demand forecasting

Artificial intelligence (AI) and machine learning

Dynamic pricing and fare optimization

Total-offer optimization

Airlines use advanced methods to profile their potential customers in order to adjust prices. One example involves classifying passengers as either business or leisure travelers (with further subdivisions) and introducing a different pricing structure for each group. Although airlines usually set fares based on historical data, such information is limited. Currently available tools are able to predict demand for particular routes based on a range of external factors (e.g., planned sports and cultural events). Airlines are experimenting with AI algorithms that can predict the impact of pricing on demand, and they are inserting these algorithms into their ongoing forecasting processes. They are also using AI to predict demand for related products and services (e.g., extra baggage, priority boarding, on-board sales). AI and various algorithms are used to adjust rates based on a range of real-time data. For example, a potential customer who has previously flown in business class or is about to achieve loyalty status may be offered a higher price than a customer who typically flies economy class. The next phase of dynamic pricing involves the application of this technology to ancillaries and optimizing the entire product bundle, including rates and extras. Airlines design and price packages for each individual customer (i.e., “travel-experience” packages).

Source: Based on fairfly.com (https://www.fairfly.com/insights/top-five-pricing-trends-in-airlinerevenue-management/).

prevailing market conditions (e.g., monopoly, oligopoly, perfect competition). According to Koopmans and Lieshout (2016), airlines can first choose their quantities (e.g., flight schedules) and adjust their prices to existing demand (yield management), thereby making it possible to apply the Cournot model. In such markets, company-specific cost changes will be

Airline pricing, incumbents, and new entrants

159

passed along at a rate of less than half, with sector-wide cost changes being implemented at a rate of more than half. Throughput speed may differ in specific situations (e.g., with limited airport capacity). Another group of studies focusses on generating additional insight into the variables that can explain “yield.” Such factors include the interplay of operational and spatial factors on average passenger revenues from non-hub airports. The average passenger yield of busier airports with longer routes tends to be lower than that of quieter airports with shorter routes (Gao and Sobieralski, 2021; Gillen and Hazledine, 2015). Competition is another important factor. For example, in China, the rapid development of highspeed trains appears to have implications for the intertemporal pricing strategies of airlines. The increase in the demand for leisure activities has been accompanied by increases in the average prices of airlines (especially along routes where high-speed trains also run), while the average prices of airlines have remained relatively low on routes with high business demand (Su et al., 2019). Another striking study concerns price differences as a function of the number of seats purchased by a single consumer. Quantum discounts appear to be greater for flights with a higher proportion of available seats at the time of booking, as well as for seats booked longer in advance (Cattaneo et al., 2016). A considerable body of recent research concerns the pricing of extras, linked to the estimation of the willingness to pay for those extras. A typical example involves the choice of a preferred seat (Rouncivell et al., 2018). Another example concerns a strategy of passenger order when boarding, which is disrupted by airline priority fares, as well as by priority passengers who board in any order, thus destroying the desired boarding order (Kisiel, 2020). Another study examines how much additional air travelers are willing to pay in order to upgrade to premium economy class. It is believed that, in the future, many more airlines will introduce a premium economy class aimed at both price-sensitive business customers and comfort-seeking leisure travelers (Kuo and Jou, 2017; Jeon and Lee, 2020). Another “extra” that has been the subject of research concerns luggage. On a number of routes, the deployment of wide-fuselage aircraft leads to the under-utilization of belly-space capacity, particularly for airlines that make little use of freight transport. Airlines can respond to this situation by proposing extra baggage service, with prices corresponding to the traveler’s

160

The Air Transportation Industry

willingness to pay, thereby generating additional income (Shaban et al., 2019). Such strategies are in line with research demonstrating that “free” luggage service is not free at all. This line of research has demonstrated that airlines that do not charge baggage allowances tend to increase the prices of their airline tickets in response to the baggage charges imposed by their rivals. The order of magnitude depends on the market power prevailing within particular societies, as well as on the competition (Zou et al., 2017). Another striking research result that has been reported is that the airfreight market is also increasingly shifting toward new pricing techniques and revenue management (Lin et al., 2017). In the future, prices will also be weighed against other variables. Although price remains important, and although airlines will continue to make efforts to keep ticket prices reasonable, it is no longer the only influencing factor. Other variables (e.g., the quality of the service offered, the proven safety record and the reputation of the airlines) will also continue to play an important role in the selection process (Truong et al., 2020). One aspect that is often neglected is the relationship between price and quality, which represents a particular cognitive language used between customer and company. The impact of flight delays has been similarly neglected (Yimga, 2020). The price mechanism clearly continues to be a crucial weapon in the competition between airlines. Recent studies have analyzed the competitive market situation in the air transport industry, taking into account fullservice carriers (FSC), low-cost subsidiaries (LCC) and rival LCCs at the flight-leg level. Subsidiary LCCs are established by FSCs as a means of competing against other LCCs, in order to maintain market share and, above all, make more profit (Dae Ko, 2019). As indicated by similar research within the field of freight transport, a unified pricing approach does not consider changes in demand in response to the price of freight space during different sales periods. Nevertheless, two competing air freight companies can generate more revenue with an optimized differentialpricing approach than they could with a single-pricing approach (Yu et al., 2019). Other interesting research concerns the effect of new entry on pricing, including the entry of rivals. For example, one study examines the effects of the entry of Southwest Airlines on the prices of its rivals, with a comparison of non-stop and connecting flights. The results of the study indicate that Southwest’s non-stop access depressed the connecting rates of its rivals, but

Airline pricing, incumbents, and new entrants

161

not their non-stop rates. Such research results have important policy implications for anti-trust analysis (Ren, 2020). We address this point in greater detail in the following section.

3. Deviant behavior: pricing as a barrier to entry Many studies within the aviation literature refer to the research of Morrison and Winston (2000, p. 18 ff.), who estimate the effect of the presence or absence of Southwest Airlines based on a model of rate changes, within the context of deregulation (see also Belobaba, 2016, p. 95). According to the results, the presence of Southwest Airlines ensures a decrease in average of fares, regardless of whether the airline actually serves a route, whether it may eventually serve a route or whether it serves an adjacent route. The presence of Southwest on a particular route causes competitors to leave some airports and increases fares on routes not served by Southwest. This suggests that fares increase along with the dominance of a particular airline. An airline can thus benefit from ousting a competitor, even if it does not result in a subsequent increase in rates. Pricing is an art unto itself. It can nevertheless also be used as a strategic weapon to combat competitors and/or to create barriers to entry for potential competitors. The literature refers to predatory pricing, which occurs when a large incumbent uses a very low price to drive smaller rivals out of the market. Predatory pricing serves a dual purpose: to drive out existing rivals and to make future rivals reconsider entering a given market. The strategy is clearly aimed at discouraging market entry. One risk associated with predatory pricing is that, in some cases, the incumbent may stand to lose more than the newcomer does. The use of predatory pricing as a weapon should therefore be predicated on the availability of large amounts of cash and a firm belief that there are many potential entrants. Recent research has also addressed a case in which a relatively small incumbent carrier (Brussels Airlines, a subsidiary of Lufthansa) pursued a pricing strategy against two other carriers, including a much larger LCC (Ryanair and Vueling). The study investigates the hypothesis that, under heavy pressure from the incumbent (in this case, Brussels Airlines), both LCCs (Vueling and Ryanair) had reduced their expansion plans. In this case the expansion plans were reduced in response to Brussels Airlines’ aggressive strategy of combining a substantial capacity increase with lower fares on all routes targeted by the newcomers. This strategy initially resulted in a significant drop in the yield of Brussels Airlines, and thus in its profitability.

162

The Air Transportation Industry

The central question addressed in the aforementioned study concerned how airline pricing strategies can be used to support strategic aims and what the consequences of such an approach are likely to be (Macário et al., 2019). The authors attempted to create a detailed reconstruction of the strategic decisions taken by the three airlines involved: who decided what, and at what time? No published data were available on the applied fares and settled revenues, nor on the revenue per route. Only a few indicators could be calculated and analyzed in order to obtain a “proxy yield” for Brussels Airlines. The current study is based on gross yield, as reflected in the operating income of each RPM. Before presenting the case study, we provide a brief overview of the historical framework. In November 2013, the Spanish low-cost carrier Vueling (a subsidiary of IAG) announced that it would increase the number of destinations from 4 to 10 beginning with the 2014 summer timetable at Brussels Airport, with announced prices starting at V29 (one-way), with some fares being as low as V19. A week later, Ryanair announced that it would start operations at Brussels Airport. Ryanair also planned to fly to 10 destinations, nine of which had previously been announced by Vueling. The most important difference between the two airlines was that Ryanair stationed four planes in Brussels, and Vueling only one. Ryanair launched its operations in late February 2014, slightly earlier than Vueling. The crucial question concerned whether Brussels Airlines would be able to survive under such pressure, given that the airline was operating on seven of the common destinations announced by both Vueling and Ryanair. The previously stated ambition of Brussels Airlines was to become profitable beginning in 2014. The entry of the two LCCs clearly forced the incumbent to rethink and adjust its own commercial strategy. In an initial response, Brussels Airlines asked the government to guarantee a “level playing field”, given that, as a registered Belgian company, it was required to comply with Belgian labor legislation and social security regulations, resulting in wage costs that were much higher than those paid by some competitors (e.g., by Ryanair, which was subject to the much lower social security contributions of Ireland). In 2014, this wage handicap was estimated at V30 million per year. For a chronological overview of all steps taken by the various actors, we refer to the publication by Macário et al. (2019, p. 83). The initial commercial steps taken by Ryanair were directed against Vueling. Surprisingly, however, the most prominent reaction came from the incumbent, Brussels Airlines, with a significant increase in the number of European destinations,

Airline pricing, incumbents, and new entrants

163

frequency and, therefore, capacity. The result was enormous over-capacity, with four competing airlines going to and from Brussels on some routes (e.g., Rome, Berlin, Lisbon, Madrid). Fares dropped, and none of the airlines was able to operate these routes profitably. In addition, Brussels Airport Company (BAC) refused to change its pricing strategy, despite Ryanair’s request for lower fares and/or compensation. As a result of these developments, Vueling (and easyJet, for that matter) decided to reduce the routes it would offer at Brussels Airport. Ryanair decided not to realize its original expansion plan and to remain with its original four aircraft based in Brussels. As in the rest of the economy, competitive pressure can force airlines to evolve toward a new business model. Such was the case for Brussels Airlines, which developed into a “hybrid” business model, in which it offered inexpensive tickets for comfortable European flights, in combination with an extensive network in Africa. In practice, this model posed strong competition for Vueling and Ryanair, as Brussels Airlines offered a different and sufficiently differentiated product. In the end, it seemed that Vueling (and also easyJet) had lost the competition at Brussels Airport, and not Brussels Airlines, as had been expected. On the contrary, however, the reduction in the routes offered by all low-cost airlines, including Ryanair, caused airfares to rise to profitable levels again in 2017. One important question concerns the extent to which the consequences of the competitive struggle described above can be quantified. Airlines are unwilling to disclose their commercial data and/or strategic steps, even after the fact, and Brussels Airlines does not publish data on yield. An approximation of the yield has nevertheless been estimated by Macário et al. (2019): the operating result per passenger-kilometer flown, assuming that the network did not change much throughout the period 2012e17. Their calculation might be subject to a slight bias, in the sense that the operating result also includes operating income from freight. Given that the share of freight in total operating income is quite small, however, this potential bias can also be regarded as quite small. As shown in Table 7.2, the returns of Brussels Airlines fell dramatically (23.3%) in the period 2012e16, when it was forced to compete with the operational expansion of Vueling and the new entry of Ryanair. Returns once again increased dramatically following the strategic policy changes enacted by both Vueling (partial withdrawal from Brussels and concentration

164 The Air Transportation Industry

Table 7.2 Operating income per passenger-kilometer flown. Indicator

2012

2013

2014

2015

2016

2017

Operating income (V million) Passenger-kilometers flown (thousands) Operating income per passenger-kilometer (V)

776.9 9,215,250 0.0843

748.8 9,753,577 0.0768

762.1 10,755,768 0.0709

830.1 11,842,966 0.0701

820.6 12,681,136 0.0647

1354.7 15,257,103 0.0888

Source: Macário, R., Meersman, H., Van de Voorde, E., 2019. Competition among European airlines: pricing strategies, yields and profits. In: Cullinane, K. (Ed.), Airline Economics in Europe. Emerald Publishing, Bingley, 77e89.

Airline pricing, incumbents, and new entrants

165

on Spanish destinations) and Ryanair (ending the original expansion plan). Total revenues for 2017 increased by 37.2% relative to 2016 (Macário et al., 2019, p. 86e87). In the calculation discussed earlier, the return is based on all operations of Brussels Airlines, both European and intercontinental. The exercise demonstrates that the company as a whole was affected by the aggressive newcomers. In this case, the incumbent was able to stop the newcomers, albeit at a hefty price. Combined with fare reductions, the strategy to expand the range resulted in a continuous decline in revenue. The aggressive counteroffensive launched by Brussels Airlines had a negative impact on both the operating result and the net result. In this respect, the case described here differs from the example of Southwest Airlines, as presented at the start of this section.

4. A possible generalization and alternative strategies One important question raised by the case developed in this chapter concerns its relative uniqueness. Although it may be one of the few price wars to have been documented in detail, it is reasonable to assume that other similar cases exist as well. In recent decades, despite constant accusations of predatory practices on the part of airlines, few if any airlines have been convicted of predation. As argued by Morrison and Wilson (2000, p. 7), predatory behavior (and the extent to which it can be measured) has long posed an unresolved problem for courts, regulatory agencies and academics. The situation has not changed. Ryanair’s decision to include Brussels Airport in its network and to station aircraft there represented a departure from the airline’s original strategy of flying exclusively at secondary airports. In addition to Brussels, the airports of Rome Fiumicino (Italy) and Lisbon (Portugal) were also added to the network. Ryanair’s choice was clearly focused on the “captive” routes, which are characterized by highly inelastic demand, due to the location of the European Commission (LisboneBrussels, Romee Brussels), and which are therefore quite resistant to higher prices, and Ryanair benefited from this. One important question concerns the response of the respective home carriers (i.e., Alitalia and TAP). Such a response need not take the form of a price war. It can also involve indirect influence (e.g., in the allocation of requested slots).

166

The Air Transportation Industry

Two developments within the Ryanair strategy merit specific attention. Although Ryanair may have already discontinued the price war with Brussels Airlines, it continues to exert pressure on all variables that could potentially influence its own cost structure. In early 2021, Ryanair called upon the Belgian government to reduce airport charges, in order to provide the much-needed impetus to restore passenger traffic following the COVID-19 crisis. The proposed incentives should be available to all airlines on an equal basis. Ryanair claims to have worked with airports and governments all over Europe to reduce airport charges. In Belgium, however, these efforts have apparently not been successful to date. At the same time, Ryanair has denounced as illegal the state aid that various European airlines (e.g., Lufthansa, Brussels Airlines, Alitalia, KLM and Air France) have received in order to survive the crisis. European Commission proceedings against such assistance have nevertheless been rejected. A new case emerging in Italy exhibits a number of similarities to the case of Brussels Airlines, Ryanair and Vueling, as analyzed in this chapter. Ryanair is also now responding to a strategic move by a competitive lowcost airline. On September 8, 2020, Wizz Air announced the opening of a new base in Catania, Italy. The intention is to station two aircraft there and to offer flights to 20 destinations, including five newly introduced routes. This new move in Wizz Air’s continuing strategy to serve the Western European market elicited a reaction from Ryanair. On September 11, 2020, Ryanair announced that it would significantly increase its number of flights in response to high demand. The announcements made by the two carriers are identical with regard to the increased capacity within the domestic Italian network. Ryanair’s decision to effect a drastic increase in capacity was undoubtedly in response to the decisions taken by Wizz Air. In the short term, travelers will benefit from lower ticket prices. This situation is clearly not sustainable, however, and its ultimate goal is to cause a cash drain for the other airline (in this case, Wizz Air), after which it will withdraw from these routes. The two low-cost carriers involved may have embarked upon a war of attrition on Italian territory. Can this situation be avoided? In theory, each of the parties involved could employ a different strategy. There are many forms of cooperation on many routes offered worldwide, both international and national. Codeshare agreements are visible examples of this. An airline could also opt for much more far-reaching strategies, including mergers or acquisitions, which would involve an explicit search for economies of scale, density or

Airline pricing, incumbents, and new entrants

167

scope. Mergers make it possible to increase prices in the short term without attracting new market entry. Another option could be to adopt a strategy of tacit collusion, which would not require a formal agreement and no direct communication between companies. Such approaches are quite common with regard to operations in “captive routes”. With an aggressive market player like Ryanair, this does not seem to be much of a risk within the current context. Market conditions can change over time, however, as is the case for all of the market parties involved. In the future, it will be important for academics and the relevant authorities to continue following the various types of potential strategies in detail through cases like the one discussed in this chapter, and the results of this follow-up should be evaluated. For example, one relevant question concerns the extent to which Ryanair is currently a dominant force in Italy. Is it possible to establish any generalization of the strategies used in the cases described above? Given that prices are the synthesis of movements in supply and demand, linked to strategic decisions, the “laboratory” of the aviation sector is in need of a price observatory. The observation of prices and the market would enhance transparency within all markets (and sub-markets).

5. Conclusions Liberalization has become a global phenomenon in recent decades. The resulting increase in competition has forced almost all airlines to investigate ways of operating more efficiently. They have succeeded, and carriers have indeed become significantly more efficient, and passengers (as consumers) have emerged as the big winner. At the same time, however, airlines have usually failed to translate the many cost and productivity improvements that they have developed, implemented and translated into financial benefits. Many of these improvements have been passed along to customers through lower prices, with benefits ultimately returning to the airlines through the expansion of their base markets with clients that had previously been unable to afford air transport. As a result, the shareholders of airline companies failed to achieve the anticipated profits, as increased competition eliminated the potential benefits of the expanded market. As demonstrated throughout this chapter, pricing appears to be a crucial instrument within the highly competitive development of the aviation business. This is also evident from the successive generations of revenuemanagement systems that have been developed, in which sophisticated

168

The Air Transportation Industry

discrimination of the potential market is used to help maximize yield. It is further illustrated by the recent evolution toward a network-transcending approach. In this chapter, we have presented a detailed case analysis to demonstrate that aggressive pricing can also be used as a barrier to entry. In the case at hand, an incumbent faced with competition from new entry fought the newcomers with a combination of capacity increases and price cuts. Entry was halted (at least in part), albeit at a high cost. Combined with strong rate reductions, the incumbent’s strategy of strongly expanding its own offerings over the course of several years resulted in a continuous decline in the company’s results. Calm would not return until the end of hostilities, which immediately translated into a substantial increase in yield. No clear generalization can be derived from the conclusions arising from a single case. Although lessons can undoubtedly be drawn from each case, it is necessary to take into account specific, often local, characteristics. Given that the aviation market is in constant development, a pricing laboratory is needed, in which all possible and applied strategies in the field of pricing can be monitored, analyzed and interpreted. Such a laboratory calls for re-thinking regulation and vesting it in more sophisticated agencies that have knowledge of the market. Measurement is knowledge, and it can thus enhance the transparency of the aviation market and its pricing.

References Belobaba, P.P., 2016. Airline pricing theory and practice. In: Belobaba, P., Odoni, A., Barnhart, C. (Eds.), The Global Airline Industry, second ed. Wiley, Chichester, pp. 75e98. Blauwens, G., De Baere, P., Van de Voorde, E., 2020. Transport Economics, seventh ed. Van In, Wommelgem. Cattaneo, M., Malighetti, P., Morlotti, C., Redondi, R., 2016. Quantity price discrimination in the air transport industry: the easyJet case. J. Air Transp. Manag. 54, 1e8. Dae Ko, Y., 2019. The airfare pricing and slot allocation problem in full-service carriers and subsidiary low-cost carriers. J. Air Transp. Manag. 75, 92e102. Gao, Y., Sobieralski, J.B., 2021. Spatial and operational factors behind passenger yield of U.S. nonhub primary airports. J. Air Transp. Manag. 90. Article 101967. Gillen, D., Hazledine, T., 2015. The economics and geography of regional airline services in six countries. J. Transp. Geogr. 46, 129e136. Holloway, S., 2003. Straight and Level: Practical Airline Economics, second ed. Ashgate, Aldershot. Jeon, M.-S., Lee, J.-H., 2020. Estimation of willingness-to-pay for premium economy class by type of service. J. Air Transp. Manag. 84. Article 101788. Kisiel, T., 2020. Resilience of passenger boarding strategies to priority fares offered by airlines. J. Air Transp. Manag. 87. Article 101853.

Airline pricing, incumbents, and new entrants

169

Koopmans, C., Lieshout, R., 2016. Airline cost changes: to what extent are they passed through to the passenger? J. Air Transp. Manag. 53, 1e11. Kuo, C.-W., Jou, R.-C., 2017. Willingness to pay for airlines’ premium economy class: the perspective of passengers. J. Air Transp. Manag. 59, 134e142. Lin, D., Ka Man Lee, C., Yang, J., 2017. Air cargo revenue management under buy-back policy. J. Air Transp. Manag. 61, 53e63. Lipczynski, J., Wilson, J.O.S., Goddart, J., 2009. Industrial Organisation. Competition, Strategy, Policy, third ed. Pearson, Harlow. Macário, R., Meersman, H., Van de Voorde, E., 2019. Competition among European Airlines: pricing strategies, yields and profits. In: Cullinane, K. (Ed.), Airline Economics in Europe. Emerald Publishing, Bingley, pp. 77e89. Morrison, S.A., Winston, C., 2000. The remaining role for government policy in the deregulated airline industry. In: Peltzman, S., Winston, C. (Eds.), Deregulation of Network Industries. What’s Next? Brookings, Washington, pp. 1e40. Phillips, R.L., 2005. Pricing and Revenue Optimisation. Stanford, California. Ren, J., 2020. Fare impacts of Southwest airlines: a comparison of nonstop and connecting flights. J. Air Transp. Manag. 84. Article 101771. Rouncivell, A., Timmis, A.J., Ison, S.G., 2018. Willingness to pay for preferred seat selection on UK domestic flights. J. Air Transp. Manag. 70, 57e61. Shaban, I.A., Wang, Z.X., Chan, F.T.S., Chung, S.H., Eltoukhy, A.E.E., Qu, T., 2019. Price setting for extra-baggage service for a combination carrier using the newsvendor setup. J. Air Transp. Manag. 78, 1e14. Starkie, D., 2008. Aviation Markets. Studies in Competition and Regulatory Reform. Ashgate, Farnham. Su, M., Luan, W., Sun, T., 2019. Effect of high-speed rail competition on airliners’ intertemporal price strategies. J. Air Transp. Manag. 80. Article 101694. Truong, D., Pan, J.Y., Buaphiban, T., 2020. Low cost carriers in Southwest Asia: how does ticket price change the way passengers make their airline selection? J. Air Transp. Manag. 86. Article 101836. Van de Voorde, E., 1992. European air transport after 1992: deregulation or re-regulation. Antitrust Bull. 507e528. The journal of American and foreign antitrust and trade regulation. Yimga, J., 2020. Price and marginal cost effects of on-time performance: evidence from the US airline industry. J. Air Transp. Manag. 84. Article 101769. Yu, S., Yang, Z., Zhang, W., 2019. Differential pricing strategies of air freight transport carriers in the spot market. J. Air Transp. Manag. 75, 9e15. Zou, I., Yu, C., Rhoades, D., Waguespack, B., 2017. The pricing responses of non-bag fee airlines to the use of bag fees in the US air travel market. J. Air Transp. Manag. 65, 209e219.

Further reading Asahi, R., Murakami, H., 2017. Effects of Southwest airlines’ entry and airport dominance. J. Air Transp. Manag. 64A, 86e90.

This page intentionally left blank

CHAPTER 8

The fight for airport slots: the case of Amsterdam Airport Schiphol1 Lisanne van Houten and Guillaume Burghouwt Royal Schiphol Group, Schiphol, Netherlands

1. Introduction Global air transport demand is expected to continue to grow in the decades ahead, despite the COVID-19 pandemic in 2020e21. Yet, Europe’s airports will not be able to accommodate all flight demand, as the currently envisioned capacity expansion plansdif they can be realized at alldare not sufficient. Increasing capacity saturation levels at airports are likely to be observed. Growing excess demand will have substantial implications for European airports. These implications relate not only to increasing average aircraft size and load factors, but also to the crowding out of certain traffic segments such as full freighter flights, traffic spill to other airports, higher ticket prices and gaming of the slot system. Furthermore, the question rises how governments and airports can ensure that scarce capacity is used most efficiently from a productive and allocative perspective. In this chapter, we explore how scarce capacity can be put to its optimal use. We first assess the implications and symptoms of growing capacity constraints at European airports, with a focus on Amsterdam Airport Schiphol until March 2020 when COVID-19 started to spread globally. In 2017, the airport hit its capacity limit of 500,000 aircraft movements per annum. Second, this chapter discusses the different options that airports and governments realistically have to ensure an efficient use of scarce capacity, taking into account the upcoming revision of the EU Slot Regulation.

1

The authors have written this chapter on personal title. Conclusions and viewpoints do not necessarily represent the position of Royal Schiphol Group.

The Air Transportation Industry ISBN 978-0-323-91522-9 https://doi.org/10.1016/B978-0-323-91522-9.00012-9

© 2022 Elsevier Inc. All rights reserved.

171

172

The Air Transportation Industry

2. The EU Slot Regulation At airports where demand for air transport services exceeds supply, slot allocation is used to define a set of rules for the allocation of airport capacity. In the EU, the allocation of slots is governed by Council Regulation 95/93, also known as the “Slot Regulation.” The Slot Regulation applies to all slot coordinated airports in the European Economic Area, and is largely based on the Worldwide Airport Slot Guidelines (WASG). Although the WASG, since 2019 jointly published by airports, airlines and coordinators, are not legally binding for EU Member States, they provide the global air transport community with a single set of standards for the management of slots and are to be taken into account by EU coordinators provided they comply with the EU Slot Regulation (ACI et al., 2019). Ahead of the allocation process, the airport issues its capacity declaration expressed in coordination parameters to the coordinator. The coordination parameters entail the total capacity available for slot allocation in a particular season, reflecting all technical, operational and environmental constraints. The coordinator allocates the slots to airlines within the limits of the declared capacity. The current slot allocation rules first and foremost recognize historic precedence, which refers to the entitlement to use a series of slots, provided the airline has already made use of that series for at least 80% of the time during the preceding, equivalent season (the “80/20 rule”). If the 80% threshold has not been met, the slots are reallocated to other airlines (ACI et al., 2019; European Parliament and Council of the EU, 2004). After the historic slots have been allocated, a maximum of 50% of the slot pool is set aside for new entrants. Any remaining slots can be used to accommodate slot requests of incumbent airlines. For competing slot requests belonging to the same priority class, coordinators may apply secondary criteria for tie-breaking purposes. The WASG provides a set of secondary criteria, which coordinators may apply to their own discretion (for example, route development or operational factors). Throughout all stages of the allocation process, coordinators must fulfill their tasks in a transparent, nondiscriminatory and independent manner (European Parliament and Council of the EU, 2004; ACI et al., 2019).

3. The changing context for slot allocation: COVID-19 and the airport capacity crunch Both the WASG and the Slot Regulation still resemble largely the system that has been developed by IATA since 1974, before deregulation and

The fight for airport slots: the case of Amsterdam Airport Schiphol

173

liberalization saw the light of day. When the key concepts of the current WASG were drafted, international air traffic was still heavily regulated and dominated by the so-called “flag carriers.” Capacity constraints were much less of an issue than they are today. Since then, the market circumstances have changed substantially (Odoni, 2020). In particular the COVID-19 crisis, as well as growing excess demand at large airports demonstrate the limitations of the current slot system. 3.1 COVID-19 Aviation has seen its largest demand shock in history due to the COVID-19 crisis from early 2020 onwards. Worldwide, passenger demand was 90% lower in Q2 2020 than in the year before.2 In September 2019, Amsterdam Airport Schiphol accommodated flight volumes equal to those in the early 1990s. On the one hand, the dramatic drop in passenger demand would make it financially impossible for airlines to meet the “80/20 rule,” which is crucial to retain slots. On the other hand, slots returned to the pool following the drop in demand (due to network rationalization, bankruptcies, etc.) were expected to dramatically impact the network of Europe’s airports and substantially change the European airline landscape in the longer term. After all, slots returned to the pool would have to be allocated according to the standard procedure, without taking into account preCOVID-19 utilization. Hence, in order to ensure schedule stability at Europe’s airports and allow airlines flexibility in light of the continued government restrictions disabling travel, as well as to mitigate the risk of airlines operating low load factor flights to protect historic slots, the European Parliament (2021) agreed to slot alleviation measures for 2020, 2021 and, depending on market developments, potentially also 2022, entailing that the “80/20 rule” has been temporarily and partially suspended at all coordinated airports. The Commission’s decision to provide alleviation was preceded by extensive consultations and discussions among industry stakeholders, since the Slot Regulation does not provide solutions for exceptional circumstances such as a pandemic (European Commission, 2020). Interestingly, COVID-19 has not changed the fact that around 13% of Schiphol’s declared capacity in 2021 has been placed on the no-slot waitlist,

2

ACI World traffic statistics.

174

The Air Transportation Industry

an indication of the still scarce capacity at Amsterdam Airport Schiphol.3 The percentage decreased only slightly compared to 2019 (15%). The demand shock resulting from COVID-19 hasdif anythingd shown that the current Slot Regulation is not suitable for dealing with these exceptional circumstances and that structural updates are needed to make the Slot Regulation more future-proof. 3.2 Demand growth and the airport capacity crunch Despite the COVID-19 crisis and the potentially longer-term impacts on the industry (Czerny et al., 2021; Suau-Sanchez et al., 2020), many expect the demand for air transport to grow in the decades ahead. ICAO still expects global passenger demand to grow by 4.2% per annum toward 2038 with slightly lower growth rates in the maturing European market. Rising disposable incomes, urbanization, liberalization, competition, globalization, and more efficient aircraft drive long-term growth. However, airport capacity in Europe may not be able to keep pace with demand growth. Eurocontrol (2018) expects that 1.5 million flights cannot be accommodated at Europe’s airports by 2040 and that 16 airports will face Heathrow-like congestion levels. It is not so much that the capacity in the entire aviation system will be falling short. However, half of global air traffic is concentrated at just 3% of the largest airports (Gelhausen et al., 2020). These are the airports that are or will be confronted with (severe) capacity problems. The growing pressure on airport capacity does not only follow from industry forecasts like those of Eurocontrol, but also from the number of airports that is slot coordinated. The number of coordinated or Level 3 airports continues to increase globally: 136 in 2000, 155 in 2010 and 198 in 2021.4 In 2019, Level 3 airports accounted for 46% of global seat capacity offered and 38% of the number of scheduled passenger flights (Fig. 8.1). Capacity constraints are most visible at the so-called super-congested airports, often providing their countries with the majority of long-haul destinations (Egeland and Smale, 2017). These airports have little to no slots available to accommodate new slot requests, since all slots are covered by incumbents’ historic rights. At such super-congested airports, route entry is often difficult, making the markets served from and to the hub less contestable (Odoni, 2020). 3 4

Slot Allocation List W20 and S21, Airport Coordination Netherlands. List of slot coordinated and facilitated airports (WASG Annex 12.7).

The fight for airport slots: the case of Amsterdam Airport Schiphol

175

Figure 8.1 Share of Level 3 airports by region in 2019 in global seat capacity and scheduled passenger flights.5 (Source: Diio; IATA; author’s analysis.)

The capacity problem is especially prevalent in Europe, which hosts about half of all Level 3 airports. European Level 3 airports account for approximately 70% of seat capacity and flights offered (Fig. 8.1). This number reflects the chronic difficulty that many European countries face regarding the expansion of airport capacity (Odoni, 2020). Growing capacity shortages are reinforced by growing public concerns with regard to noise exposure, carbon emissions and land use planning. Environmental concerns are the primary reason for the observation that airport capacity in large urban areas is lagging behind demand growth. For example, Amsterdam Airport Schiphol has a long history of strong environmental opposition to airport expansion and flight volume growth. The cap of 500,000 aircraft movement at the airport reflects not technical but environmental constraints. Environmental impacts were also the main reason for delaying new runways at Vienna and Munich airport. Environmental concerns go hand in hand with a societal debate regarding how the airport capacity can be used optimally to the benefit of the economy and welfare. One reasoning in the public debate is that if 5

Level 3 airports for summer 2020. Scheduled traffic statistics for 2019 (JaneDec). Passenger services only.

176

The Air Transportation Industry

airport capacity is used for those flights that deliver most socioeconomic value, there is no or less need for airport expansion. Yet, the current administrative slot regime does not provide any guidance to slot coordinators as to how to allocate slots according to their socioeconomic value. In sum, airport capacity will become increasingly scarce in large urban areas in the long run, globally but particularly in Europe. Slot alleviation measures in fact “froze” the 2019 capacity shortages. Although the current airport slot system is intended to allocate capacity efficiently, it was designed in a time with relatively modest capacity constraints. In our view, it is questionable to what extent the Slot Regulation is fit for purpose in ensuring the efficient use of increasingly scarce capacity. We will discuss this at the end of this chapter.

4. The implications for growing excess demand for slots: theory and research findings When an airport reaches its capacity limitdenvironmental or technicald not all cargo and passenger demand can be accommodated. This will have various demand and supply implications (SEO and Cranfield University, 2017). One of the main impacts may be higher ticket prices. When demand exceeds supply, producer prices normally rise to clear the market and balance supply and demand. If the airport would price efficiently through its airport charges, higher airport charges would reflect scarcity. However, most large European airports are regulated and cannot increase their charges in response to scarcity. In other words, constrained, regulated airports are likely to charge inefficient low prices in case of severe capacity constraints (SEO and Cranfield University, 2017). Instead, various scholars expect airlines to raise ticket prices to clear the market instead, according to what the market can bear. The premium on the “normal” ticket prices is called the scarcity rent (Starkie, 1998; Forsyth, 2004; Gillen and Starkie, 2016). When additional slots are made available, new entry may compete (part of the) scarcity rents away. Hence, some studies argue that scarcity rents can be a disincentive for airlines to actively support airport capacity expansion (Forsyth, 2004; Gillen and Starkie, 2016). The issue of scarcity rents has been subject to fierce debates, fueled by controversies between airports and airlines (Batley et al., 2019; FTI Consulting, 2018). A number of studies have attempted to empirically demonstrate the existence of scarcity rents (Abramowitz and Brown, 1993;

The fight for airport slots: the case of Amsterdam Airport Schiphol

177

Borenstein, 1989; Dresner et al., 2002; Evans and Kessides, 1993; Frontier Economics, 2014; Fukui, 2019; Morrison, 2001; PWC, 2013; SEO and Cranfield University, 2017; Van Dender, 2007). For example, SEO and Cranfield University (2017) find an average ticket price premium of approximately 6 euros per return trip paid at congested airports in Europe. Furthermore, a 10% increase in airport congestion is associated with a 1.4e2.2% increase in air fares. However, these are average values for Europe. Impacts by individual airports may differ: Frontier Economics (2014) estimates that air fares at Heathrow and Gatwick are respectively 18% and 7% higher than at other London airports. Others cast doubt on the existence of scarcity rents (SLG, 2013; British Airways, 2015; IAG, 2018), where and to what extent these rents will accrue in the value chain (Batley et al., 2019) and the underlying drivers for higher ticket prices (Batley et al., 2019; Starkie, 2020). When ticket prices rise, airport users with the lowest willingness to pay will be “crowded out.” Generally, these are transfer passengers, price sensitive, lower yielding leisure passengers and cargo. Part of these passengers may substitute to other airports (spill-over), other modes of transport or may decide not to travel all together (Gudmundsson et al., 2014; SEO and Cranfield University, 2017). Capacity constraints will also impact network development. First, constraints may result in a stabilization of direct connectivity or flight volumes. Secondly, airlines will adjust network development, with a growing focus on denser, higher yielding routes at the expense of lower yielding routes. For example, owing to the introduction of the 480,000 movement capacity limit at Heathrow in 2008, many regional routes were crowded out at Heathrow while spill of long-haul traffic to other airports inside and outside London has been substantial (Gudmundsson et al., 2014). Another source for “crowding out” is the slot allocation system itself: operations which cannot meet the “80/20 rule” because of their irregular nature and dependency on nonhistorical slots (such as freighter flights) will be pushed out of congested airports. Finally, in response to airport capacity constraints, airlines are expected to optimize load factors and increase average seat capacity above the average market trend. Average aircraft capacity at Heathrow is amid the highest in the world, reflecting among other things the severe constraints and the fact that the secondary market has become the primary source of slots at Heathrow.

178

The Air Transportation Industry

5. The implications for growing excess demand for slots: the case of Amsterdam Airport Schiphol In 2017, Amsterdam Airport Schiphol reached its capacity limit of 500,000 aircraft movements per annum. The capacity constraint was an environmental limitation with the aim of balancing the development of the airport with noise reduction and local environmental quality, and not a reflection of operational capacity limits. It was the outcome of a 2-year-long negotiation process that resulted in the so-called Alders Agreement of 2008 between various industry parties, regional stakeholders and local residents. The Agreement was subsequently adopted by the government.6 The 500,000 movement limit has substantially affected connectivity and traffic at the airport. A number of descriptive statistics already allow us to draw in this section the contours of how the capacity limit impacted Amsterdam Airport Schiphol, until the spread of COVID-19 from early 2020 onwards. With respect to traffic and connectivity, aircraft size and traffic, but also regarding airline strategies to gain access to constrained infrastructure. 5.1 Stabilizing connectivity The capacity limit resulted in the number of flights leveling off after 2017 (Fig. 8.2). Surprisingly, the number of passenger destinations increased slightly over the same period, presumably because of the use of former freighter slots for passenger operations as well as the tendency of airlines to serve some destinations at a lower frequency. Yet, the net increase in total number of passenger and freighter airport-reported destinations since 2017 has been modest. 5.2 Seat capacity and load factor The impact of the capacity constraints can also be observed in the development of the average seat capacity per flight (Fig. 8.3). Before 2017, annual average aircraft size growth ranged between 0.5% and 2%, in line with other large European airports. Since 2017, airlines have been up-gauging their fleet, leading to high growth rates in average number of seats per plane. At other European hubs with less constraints, such a spike in aircraft size growth cannot be observed. 6

For the Alders Agreement of 2008, see https://www.omgevingsraadschiphol.nl/wpcontent/uploads/2020/01/advies-alders-middellange_termijn.pdf (in Dutch).

The fight for airport slots: the case of Amsterdam Airport Schiphol

179

Figure 8.2 Development of number of flights and number of passenger destinations 2009e19.7 (Source: SEO, 2020; Amsterdam Airport Schiphol traffic statistics.)

In 2019, the growth rate in average aircraft size decreased rapidly. One reason for the decrease may be found in the so-called operational peak-parameters in the capacity declaration. In 2018/2019, the airport implemented parameters in the capacity declaration to limit the number of wide-body aircraft in the morning peak hours, since gate and terminal capacity was becoming scarce. In other words, further growth of aircraft size was for one part suppressed by the capacity declaration. Another explanation is the possibility that airlines have already picked the low-hanging fruit in terms of up-gauging. Finally, the period since 2017 has also witnessed wide-body aircraft replacement in the KLM fleet,8 resulting a downward pressure on seat capacity. The growth in aircraft capacity and load factors absorbed part of the demand growth that could not be accommodated by movement growth, but most likely not all potential demand growth at Amsterdam Airport Schiphol could be accommodated during the 2017e20 period. 7

Notes: (1) destinations representative week: number of passenger destinations measured in third week of September of each year based on airline schedules data (SEO, 2020); (2) airport-reported destinations: destinations available during the year, both for passenger and full-freighter traffic, as reported by Amsterdam Airport Schiphol. 8 Boeing 747-400 aircraft were taken out of service, whereas Boeing 787 aircraft with a lower seat capacity were added to the fleet.

180

The Air Transportation Industry

Figure 8.3 Y-o-y growth of average seat capacity per flight at Amsterdam Airport Schiphol versus six other large European hubs (FRA, MUC, CDG, MAD, ZRH, FCO). (Source: Diio; authors’ analysis.)

Regarding load factors, Amsterdam Airport Schiphol ranked among the highest of seven largest European hubs and showed a steady increase in average load factor over the past decade, reaching a record high level in 2019 (Fig. 8.4). However, the load factors leveled off between 2017 and 2019 and other airports such as Munich and Zurich had similar load factors as Schiphol. We note that load factor growth is the consequence of various forces: the sophistication of revenue management systems among the home carriers, traffic mix, capacity constraints and the share of low-cost carriers (LCCs), which tend to have higher load factors on average. At airport level, it may also be difficult to grow load factors beyond a ceiling of approximately 88%, due to temporary reductions in demand and seasonality of demand, the need for airlines to avoid denied access for highyield bookers and heterogeneity in the sophistication of yield management systems between airlines. In that respect, it could well be that since 2017 load factors have not substantially increased at Amsterdam Airport Schiphol because it was already close to its natural “load factor ceiling.”

The fight for airport slots: the case of Amsterdam Airport Schiphol

181

Figure 8.4 Development of estimated average airport load factors 2009e199 at Amsterdam Airport Schiphol versus other airports. (Source: Diio; authors’ analysis.)

5.3 Transfer traffic In line with the theory regarding the impact of capacity constraints (see Section 4), a sharp decrease in the growth of transfer traffic can be observed since 2017 as the hub carrier may prioritize origin-destination traffic over (part of its) transfer traffic10 due to thedon averagedhigher yields (Fig. 8.5). The share of transfer traffic showed a declining trend over the 2014e2019 period, eventually reaching a level comparable to the early 1990s when KLM started to build its extensive wave-system at Amsterdam Airport Schiphol. The declining transfer share can be partly explained by capacity scarcity, but also by the expansion of (OD-focused) LCCs at the airport between 2014 and 2017.

9

Note: load factors calculated based on available scheduled seat capacity per year and annualized adjusted passenger booking data for each airport. Estimated load factors may deviate slightly from official airport reported load factors. 10 However, a small part of the transfer segment is (extremely) high yield, in particular uniquely served transfer markets where few or no travel alternatives are available.

182

The Air Transportation Industry

Figure 8.5 Year-on-year growth of transfer passenger traffic (excl. transit) and transfer shares at Amsterdam Airport Schiphol. (Source: Amsterdam Airport Schiphol annual traffic statistics.)

Figure 8.6 Development of scheduled full-freighter flights at AMS and six other large hubs (FRA, MUC, CDG, MAD, ZRH, FCO) (May 2015 ¼ 100). (Source: Diio; authors’ analysis.)

5.4 Full freighter traffic Before the COVID-19-crisis started in 2020, the number of full-freighter flights at Amsterdam Airport Schiphol was already declining for a

The fight for airport slots: the case of Amsterdam Airport Schiphol

183

number of years, while other European hub airports still faced significant growth (Fig. 8.6). The impact also became visible as several full freighter carriers started moving freighter traffic to other airports, such as Brussels Airport and Liège. In part, the reason lies in the business model of full freighter operations which makes it more difficult to fly according to a fixed on-time schedule to comply with the “80/20 rule.” Unlike passenger services, the air cargo industry often does not offer scheduled air services with balanced, bidirectional demand with fixed time schedules (Commission of the European Communities, 2009). At Amsterdam Airport Schiphol, the lack of any available slots from 2017 onwards exacerbated the decrease, as full freighter airlines were no longer able to pick-up alternative slots to continue pre2017 operations. 5.5 The role of slots in retaliatory actions in the aeropolitical arena The lack of available slots at capacity-constrained airports may affect the ability of airlines to exercise traffic rights granted under air services agreements. In particular at constrained airports where no substitute airports serving the same market is available, it may be practically impossible for an airline to access that market. Despite the fact that the grant of traffic rights does not imply free access to coordinated airports in the EU because of the existence of a so-called ’operational link’ between traffic rights and slots, this lack of effective airport access despite the availability of traffic rights, may result in an alleged breach of the terms negotiated between two States. Consequently, the opposite State may adopt retaliatory measures (García-Arboleda, 2013). If the slots are refused because there are not any left, a State may argue that its designated carriers do not have a fair and equal opportunity to compete with the national carriers of the other State (Mendes de Leon, 2013). This issue became clearly visible at Amsterdam Airport Schiphol in 2017 as several full freighter carriers started moving freighter traffic to other airports. Following the increased slot shortage, Russian-registered AirBridgeCargo was not allocated all of the slots it previously held. As part of the commitments in the air services agreement between The Netherlands and Russia, Russia invoked the principle of a fair and equal opportunity to compete. It threatened to close Russian airspace to Dutch aircraft, including KLM, if AirBridgeCargo was refused more slots at Amsterdam Airport Schiphol. After several rounds of negotiations,

184

The Air Transportation Industry

AirBridgeCargo and KLM came to an agreement, which allowed AirBridgeCargo to re-establish its operations at the airport instead of moving to Liege airport in Belgium. The agreement was described as a codeshareagreement, through which KLM could lease slots to AirBridgeCargo. The then Managing Director of Airport Coordination Netherlands (ACNL), responsible for slot allocation at Dutch coordinated airports, stated in Dutch newspaper NRC that the government of The Netherlands “could set a dangerous precedent by giving in to the threats over Russia’s airspace, as other carriers may try to gain space at European airports in the same way” (NRC Handelsblad, 2017). A similar dispute occurred in 2019, when US freighter carrier Kalitta Air filed a complaint with the US Secretary of Transportation, arguing that its slots at Amsterdam Airport Schiphol were “wrongfully withheld” by ACNL, Amsterdam Airport Schiphol and the Dutch government. 5.6 Use of remedies at Amsterdam Airport Schiphol Another way to access a severely congested airport is through the use of “remedy slots,” if available. In its assessment of airline mergers and alliances, and more recently also in State aid cases, the European Commission may make its approval conditional upon the offering of slot commitments in order to reduce barriers to entry and facilitate new entry, in particular at constrained airports. Following the 2004 Air France-KLM merger, the Commission had doubts over the potentially dominant position of Air France-KLM on identified long-haul city pairs, including the AmsterdameNew York (JFK and Newark) route. To remedy the concerns, Air France-KLM committed to make slots available in order to enable new entry, not less than six times a week. Besides brief operations by British Airways subsidiary OpenSkies for 3 months in 2009, there had been no applicants for the remedy slots between Amsterdam and New York until 2017, presumably because there were still slots available at the airport through the normal slot procedure (European Commission, 2019). When the capacity ceiling was reached in 2017, Norwegian applied for the remedy slots on the route. According to the Commission, the remedy slots between Amsterdam and New York would restore the competitive situation that existed prior to Air France’s acquisition of control over KLM (European Commission, 2019). This situation shows that when capacity constraints really start to bite, airlines are exploring alternatives to access the market, particularly when attractive slot times are involved.

The fight for airport slots: the case of Amsterdam Airport Schiphol

185

5.7 Deficiencies of grandfather rights: slot hoarding and babysitting The slot rules appear to encourage high levels of slot utilization by means of the “80/20 rule,” since airlines will lose them to the pool if they do not operate them according to the threshold. However, grandfather rights have also frequently been critcizied for preventing an optimal use of scarce airport capacity, especially at airports with severe excess demand. The “80/20 rule” may provide a strong incentive for airlines to hold on to slots and deter entry. Research by, among others, the European Commission also found that, instead of returning slots, slots are hoarded to safeguard them for future operations, even if their use is not financially viable at the time (European Commission, 2011). Although there may be other justifiable reasons for airlines to hold onto slots, safeguarding slots prevents slots from ending up with competitors, which could potentially make more efficient use of them (Finger et al., 2019). For example, in order to retain slot portfolios at super-congested Heathrow, airlines resort to flying smaller planes than necessary in order to spread seat capacity across their slots. Heathrow is also familiar with the so-called “ghost flights”: airlines have operated (nearly) empty flights to ensure that the airport is busy at the appointed time (Competition and Markets Authority, 2019). In 2018, roughly 7000 slots were handed back too late at Amsterdam Airport Schiphol for other airlines to use them. This resulted in a waste of scarce and highly sought-after capacity that could have been recycled to other airlines (PA Consulting, 2019). Thus, there is no guarantee that incumbent slot holders are also the most efficient users of airport capacity from a productivity perspective (Finger et al., 2019). This is not solely the result of the incumbent lacking the incentive to use a slot efficiently, but also because it might have an active, strategic interest in not freeing up capacity for competitors (Competition and Markets Authority, 2019) and/ or because of uncertainties as regards planned operations. In addition to slot hoarding, actual slot utilization patterns show that already scarce airport capacity may be used inefficiently through, among others, overbidding, flights operated without slots and mismatches between allocated slot times and actual operated times, causing congestion externalities to competitors and sometimes even foregoing competitive market entry (Rupp, 2009; Madas and Zografos, 2010; Steer Davies Gleave, 2011; Haylen and Butcher, 2017).

186

The Air Transportation Industry

Slot noncompliance is not uncommon at constrained airports throughout the world, but the prevalence of these events at Amsterdam Airport Schiphol appears to be above industry norms according to PA Consulting (2019). Steer Davies Gleave (2011) already concluded about a decade ago that the scope of sanctions for slot misuse in The Netherlands are less than required by the Slot Regulation. With slot monitoring being more rigorously enforced at other airports in other States, there is a risk that airlines sanctioned for noncompliance at other airports will effectively “export” their delays to Amsterdam Airport Schiphol in order to avoid fines and/or withdrawal of slots down route, because there are hardly any penalties. The WASG includes a new chapter on slot monitoring since 2019, which provides useful guidelines for slot compliance that could be adopted in The Netherlands in order to ensure that Schiphol’s scarce slots are used efficiently and in line with their initial allocation (PA Consulting, 2019). 5.8 Gaming of the new entrant rule In Europe, an airline only qualifies as a new entrant if it holds fewer than five slots at a particular airport on a particular day, and if its total slot holdings do not exceed 5% of the total number of available slots at that airport, or 4% at the airport system level (European Parliament and Council of the EU, 2004). Per these requirements, airlines with only a limited presence at an airport would not satisfy the new entrant criteria, including the ones with as much as 2% of the total slot holdings, as the allowed maximum is easily exceeded. At the London airport system level, for example, LCCs such as easyJet or Ryanair would not qualify for new entrant slots at Heathrow because their slot holdings within the system exceed 4% of total slot holdings, even though they have no operations at Heathrow (Competition and Markets Authority, 2019). These airlines would have to acquire slots at Heathrowdof which there currently are nonedthrough the regular procedure, buy slots on the secondary market or engage in a slot lease. In addition, airlines with multiple Air Operator Certificates (AOCs) and airline groups may use the loopholes of the new entrant rule. For example, UK-based slot coordinator ACL has found that the same airline may claim new entrant priority with different AOCs, or even though it already has a presence at the airport, either individually or as part of an airline group. Moreover, airlines conducting shared operations may freely swap slots between them, sometimes also including slots allocated with new entrant priority, thus foregoing the requirement that new entrant slots may not be

The fight for airport slots: the case of Amsterdam Airport Schiphol

187

transferred for a period of two equivalent seasons. Although not prohibited by the Slot Regulation, this behavior has been labeled as a “potential form of gaming” of the system by ACL and needs legislative clarification (Airport Coordination Limited, 2019). The specific characteristics of the new entrant rule have induced excessively fragmented outcomes, with a large number of small airlines with only very limited slot holdings and often less favorable times operating at large hub airports (Competition and Markets Authority, 2019). At major European hub airports, slots are concentrated among a few larger players, but the “tail” of slot holdings is highly fragmented among many airlines. Fragmentation makes it less likely that new entrants with modest slot holdings are able to compete effectively with incumbent carriers at constrained airports (DotEcon, 2006; Steer Davies Gleave, 2011; Competition and Markets Authority, 2019). These characteristics go against the raison d’être of the current system, that is to, among others, promote effective competition and remedy schedule fragmentation. Subsequently, the Commission proposed in 2011 to structurally broaden the definition of a new entrant airline (European Commission, 2011). However, this proposal has not been acted upon in the European Council at the time of writing. In 2021, the European Parliament adopted a temporary amendment to the new entrant rule by increasing it to 10% at airline group level as part of its temporary COVID-19 relief measures from the slot utilisation rules, as embodied in EU Regulation 2021/250.

6. Conclusions Our assessment has provided an overview of the symptoms and implications of airport capacity constraints by focusing on the Amsterdam Airport Schiphol case. Consequences do not only relate to traffic changes, crowding out and (inefficient) market outcomes, but also to the strategies that allow airlines to gain access to a scarce resource. It goes without saying that the growing capacity scarcity calls for an efficient use of that capacity. In turn, a more efficient capacity utilization may reduce the need for considerable airport expansion. We argue that the current administrative Slot Regulation is not optimally suited for dealing with severe excess demand for slots in terms of delivering an optimal allocation and use of scarce airport capacity, nor is the system sufficiently flexible in dealing with large demand shocks exemplified by COVID-19.

188

The Air Transportation Industry

We highlight the most important shortcomings in the current slot regime in Europe and discuss potential remedies to enhance the efficient use of capacity through the revision of the current Slot Regulation. We argue that a future-proof slot system needs changes is mainly three areas: (1) fixing the bugs in the current slot system, (2) creating a category of supercongested airports in the Slot Regulation and (3) introducing a flexible approach toward market-based mechanisms such as secondary trading, however at the discretion of individual EU Member States depending on local circumstances. 6.1 Fixing the current system With more and more airports declared to be congested, existing airport capacity should be allocated and used in the most efficient way to benefit consumers and the broader economy/society. The current slot rules, with grandfather rights at the heart of the system, come with certain strengths in terms of schedule continuity for airlines. However, the system also causes concerns of anti-competitive side-effects and slot gridlock, in turn negating the system’s objective of improving the efficient use of scarce airport capacity (NERA, 2004; Steer Davies Gleave, 2011; Odoni, 2020). Hence, a revised Slot Regulation should address that airlines are expected to make full and proper use of the slots allocated to them, and that any forms of non-justified misuse are sanctioned through enhanced slot enforcement (ACI Europe, 2020). Furthermore, we suggest that the definition of a slot is amended. This means that a slot is not just allocated as a permission to use congested infrastructure. Its allocation should encompass an obligation to use the congested infrastructure with the exception of justified non-use (ACI Europe, 2020). This ensures that scarce slots are not wasted and allows for the introduction of sanctions for, among others, overbidding and late handbacks. Airlines that engaged in such practices in the previous season would then be sanctioned for misuse, for example through receiving lower priority during initial allocation in the following season. Synergies could be sought with the recently added Chapter 9 of the WASG on slot monitoring. In line with the above, and in order match the pro-competitive spirit of the new entrant rule, qualification for new entrant priority should be measured at airline group level as airline groups should be treated as single entities for allocation purposes. Looser forms of cooperation, such as codeshare agreements, should no longer fall within the terminology of “shared operations.” Besides strengthening the new entrant rule to

The fight for airport slots: the case of Amsterdam Airport Schiphol

189

circumvent abuse, it could be strengthened to deliver pro-competitive impacts, instead of promoting highly fragmented market outcomes, which are of little interest to the airport’s connectivity, consumer or promising new entrants. The new entrant rule as we currently know it does not provide sufficient room for airlines with limited slot holdings to establish a competitive foothold at congested airports. There have already been proposals to raise the 5% threshold or the limit of five slots per day (Steer Davies Gleave, 2011; ACI Europe, 2020). Alternatively, consideration could be given to applying the new entrant rule at route level. After all, the level of competition is often measured at route or market level, as we have learned from the imposition of slot remedies in mergers and alliance cases. Further analysis should be carried out to explore whether this could be a feasible and effective solution. The practice and formulation of the 80% usage threshold could also be improved. It is questionable whether an isolated increase of the threshold would be feasible and effective, as that is likely to put greater pressure on the interpretation of force majeure (20% are currently for commercial cancellations only, cancellations for which force majeure may be granted fall outside the scope of the 80/20 rule), by coordinators, for which a clear definition is currently lacking. Any increase should be paired with a clear definition of what constitutes force majeure. Finally, as the demand shock resulting from the COVID-19 crisis followed by extensive discussions on slot alleviation have shown, the Slot Regulation should be equipped with a provision that allows for applying modulated usage thresholds to grant relief in exceptional circumstances. 6.2 The concept of the super-congested airport As we have seen at Amsterdam Airport Schiphol and Heathrow, severe slot scarcity may crowd-out certain traffic segments, while existing slots at highly constrained airports may be increasingly used to increase frequencies at already served destinations instead of adding new ones. Over time, this could lead to less network diversity, which may not necessarily be aligned with efficient use of airport capacity from a societal point of view. At super-congested airports, airlines should not only be obliged to use the slots allocated to them, but they should also continue to operate the slots in line with the conditions imposed upon initial allocation, be it new entrant priority, year-round criteria or other airport-specific strategic conditions. Governments should be able to develop criteria for coordinators to apply at these airports located within their territories, entailing certain strategic

190

The Air Transportation Industry

objectives for the coordinator to apply in the allocation of slots at their own discretion. Although we acknowledge that airlines need flexibility to respond to changing consumer demand, introducing such a requirement helps governments to ensuredvia the independent coordinatordthat slots are put to a more socio-economically beneficial use. In its simplest form, such (secondary) criteria would apply to competing slot requests, after the application of historic rights, new entrant rule and year-round criteria. To effectuate such a requirement to work in practice, a provision enabling the coordinator to earmark the slots for two consecutive, equivalent seasons after initial allocation could be introduced. This would compel airlines to operate slots in line with imposed conditions for a minimum of two years. After the two years have lapsed, any changes are subject to the approval of the coordinator who has to take into account any applicable local rules and strategic objectives of the airport concerned. Within the 2-year timeframe, slots may not be transferred for use by other airlines or alternative operations. As such, airlines are encouraged not to hold and use their slots in a way contrary to the objectives and the spirit of the slot system. This could enable airports to maintain or develop connectivity that is under pressure (for example, full-freighter flights) or develop connectivity that is considered to be of strategic importance for the local community (e.g., certain regional connections, unserved business destinations, new long-haul routes or hub-related traffic in the peaks), whilst mitigating further allocation fragmentation, which may add little value at already constrained airports.11 6.3 Market-based measures Many studies and scholars have advocated secondary slot trading as a market-based solution to induce slot mobility and increase efficiency, as airlines will take into account the opportunity costs of not trading the slots. In its revision of the Slot Regulation, the Commission will likely consider secondary slot trading as a way to enhance slot mobility. Nonetheless, “grey trading” or “trading in disguise” (for example, through mergers, take-overs and slot leases) may already be going on a number of highly constrained airports across Europe and even globally, although any hard evidence on 11

An alternative market-based measure would be charges differentiation. However, the regulatory framework of many airports limits significantly the possibilities for differentiating charges on a destination-basis. Yet, incentives schemes used by many airports can play a role to incentivize a certain connectivity development.

The fight for airport slots: the case of Amsterdam Airport Schiphol

191

trading is lacking outside the UK (SEO, 2018). Formalized secondary trading could help to establish a more liquid and transparent slot market with clear rules and regulations that enhances slot mobility and aligns slot utilization better with the willingness to pay of airlines. Furthermore, trading may allow effective entry and add competition at severely constrained airports. It will also inform governments, regulators and the general public about the economic value of slots (Odoni, 2020). However, some caution is needed with respect to formalizing secondary trading. This should depend on local circumstances. Secondary slot trading bears the risk of inducing further market concentration around already dominant airlines at already concentrated and constrained hub airports, as the London experience has demonstrated.12 This may allow the dominant carrier to start new routes and frequencies that could strengthen the position of the airport as a hub, but it may also limit downstream competition as the dominant airline may intensify its gaming behavior and pre-empt competition, for example by discriminatory practices among potential slot buyers. Also, the concern of entry of airlines with “deep pockets” has frequently been mentioned as a concern related to secondary trading. The trade-off between the consequences of better network quality and a stronger hub position on the one hand and competition concerns on the other hand, is not clear-cut. Although many studies relate higher fares at congested airports to market concentration and scarcity rent issues, it is also recognized that higher fares at constrained airports may correspond “with a differentiated and more costly network quality. [.] it is the hub-andspoke-system that provides the origin-destination market at the hub with an excellent accessibility product of direct connections to a disproportionate set of worldwide destinations [.]” (De Wit and Burghouwt, 2008). Furthermore, secondary trading will ensure a more efficient allocation of slots from an airline point of view (otherwise, the trades would not have taken place), but it is beforehand not a given that secondary slot trading will also enhance efficient use of capacity from an airport’s or societal point of view (SEO, 2018). In sum, the pros and cons of secondary trading call for a careful and tailormade approach toward secondary slot trading. Preferably, it should be possible but explicitly left to the discretion of individual EU Member States to implement secondary trading or not, depending on the local market circumstances. Where trading is permitted, it should be delineated with clear 12

Although it may also open up the market for new competition on certain markets.

192

The Air Transportation Industry

rules and regulations. For instance, with close monitoring and evaluation, as well as the supervision/regulation of any adverse competitive or societal impacts of slot trading. With lack of empirical evidence on the impacts of secondary slot trading on the wider European market, it could be worthwhile exploring the possibility to allow a maximum share of slots to be traded annually on the secondary market and increase this share if trading results to be beneficial from a societal point of view. Further research is clearly needed regarding the pros and cons of secondary slot trading in the European context, while taking into account the lessons learned from the UK. 6.4 Concluding remarks The fight for slots at major European hubs such as Amsterdam Airport Schiphol and Heathrow provides already a first outlook on the consequences of extreme scarcity. This outlook calls for a revision of the Slot Regulation to ensure an optimal use of airport capacity toward the future. Such a revision includes not only the fixing of a number of shortcomings in the current system, but also the designation of a category of “super-congested airports” where provisions are applicable to prevent slot stagnation, preserve airline competition and to stipulate airport-specific strategic criteria for slot allocation in order to put available capacity to its most socio-economically optimal use. These provisions could potentially include market-based mechanisms such a secondary slot trading, at the discretion of individual EU Member States and depending on local market circumstances. In conclusion, growing excess demand for airport capacity will be among the set of developments that will affect the global aviation industry most in the decades ahead. Growing public awareness about aviation’s impact on climate change and the policy measures to mitigate these impacts are likely to put the allocation of scarce airports capacity at the forefront of the debate.

References Abramowitz, A.D., Brown, S.M., 1993. Market share and price determination in the contemporary airline industry. Rev. Ind. Organ. 8 (4), 419e433. ACI EUROPE, January 2020. ACI EUROPE Position Paper on Airport Slot Allocation. ACI, IATA and WWACG, 2019. Worldwide Airport Slot Guidelines (WASG) Effective 1 June 2020. Airport Coordination Limited, 2019. ACL Response to Sections 3.46 to 3.65 of the Consultation Document of Aviation 2050: The Future of UK Aviation.

The fight for airport slots: the case of Amsterdam Airport Schiphol

193

Batley, R., Wheat, P., Mackey, P., 2019. Independent Peer Review of Recent Research on the Existence of Scarcity Rents at Heathrow. Study Funded by the Civil Aviation Authority (CAA). Borenstein, S., 1989. Hubs and high fares: dominance and market power in the US airline industry. RAND J. Econ. 344e365. British Airways, 2015. British Airways Response to the Airports Commission Public Consultation on New Runway Capacity in the South East. 3 February 2015. Commission of the European Communities, 2009. CASE M.5403 e Lufthansa/BMI. Article 6(1)(b) NON-OPPOSITION, 14/05/2009. Competition and Markets Authority, 2019. Aviation 2050: Response from the Competition and Markets Authority. Czerny, A.I., Fu, X., Lei, Z., Oum, T.H., 2021. Post pandemic aviation market recovery: experience and lessons from China. J. Air Transp. Manag. 90. De Wit, J., Burghouwt, G., 2008. Slot allocation and use at hub airports, perspectives for secondary trading. Eur. J. Transp. Infrastruct. Res. 8 (2), 147e164. DotEcon Ltd., 2006. Alternative Allocation Mechanisms for Slots Created by New Airport Capacity. Final report by DotEcon Ltd. Dresner, M., Windle, R., Yao, Y., 2002. Airport barriers to entry in the US. J. Transp. Econ. Pol. 36 (3), 389e405. Egeland, J., Smale, P., 2017. Capacity building through efficient use of existing airport infrastructure (2017). In: International Transport Forum Discussion Paper 2017. Eurocontrol, 2018. European Aviation in 2040. Challenges of Growth. European Commission, 2011. Explanatory Memorandum to the Proposal for a Regulation of the European Parliament and of the Council on Common Rules for the Allocation of Slots at European Union Airports, COM(2011) 827 Final. European Commission, 2019. CASE M.3280 e Air France/KLM. Decision on the Implementation of the Commitments e Waiver of the Commitments, 06/02/2019. European Commission, 2020. Slot Relief Measures in Light of the COVID-19 Pandemic, SWD(2020) 341 Final. European Parliament and Council of the European Union, 2004. Regulation (EC) No 793/ 2004 of the European Parliament and of the Council of 21 April 2004 Amending Council Regulation (EEC) No 95/93 on Common Rules for the Allocation of Slots at Community Airports. OJ L 138. European Parliament and Council of the European Union, 2021. Regulation (EU) 2021/ 250 of the European Parliament and of the Council of 16 February 2021 Amending Council Regulation (EEC) No 95/93 as Regards Temporary Relief from the Slot Utilization Rules at Union Airports Due to the COVID-19 Crisis. OJ L 58. Evans, W.N., Kessides, I.N., 1993. Localized market power in the U.S. Airline industry. Rev. Econ. Stat. 66e75. Finger, M., Montero-Pascual, J.J., Serafimova, T., 2019. Navigating Towards a More Efficient Airport Slots Allocation Regime in Europe. Florence School of Regulation. Forsyth, P., 2004. Locational rents at airports: creating them and shifting them. J. Air Transp. Manag. 10, 51e60. Frontier Economics, 2014. Impact of Airport Expansion Options on Competition and Choice. A Report Prepared for Heathrow Airport. FTI Consulting, 2018. A Critique of Published Reports Regarding Scarcity Rents at Heathrow. A Report for the Civil Aviation Authority. Fukui, H., 2019. How do Slot Restrictions Affect Airfares? New Evidence from the US Airline Industry. Economics of Transportation 17, pp. 51e71. García-Arboleda, J.I., 2013. Airport slot regulation in Latin America: between building the fortress and protecting the newcomers. Aviation Law Pol. 12 (3), 573e614.

194

The Air Transportation Industry

Gelhausen, M.C., Berster, P., Wilken, D., 2020. Airport Capacity Constraints and Strategies for Mitigation. A Global Perspective. Academic Press, London. Gillen, D., Starkie, D., 2016. EU slot policy at congested hubs, and incentives to add capacity. J. Transp. Econ. Pol. 50 (2), 151e163. Gudmundsson, S., Paleari, S., Redondi, R., 2014. Spillover effects of the development constraints in London Heathrow Airport. J. Transp. Geogr. 35, 64e74. Haylen, A., Butcher, L., 2017. House of Commons Library Briefing Paper on Airport Slots. IAG, 2018. Letter to Stephen Gifford, Consumer & Markets Group, Civil Aviation Authority. https://www.caa.co.uk/uploadedFiles/CAA/Content/Accordion/Standard_ Content/Commercial/Airports/Files/IAG%20CAP1658%20response%20FINAL.pdf/ (Accessed 15.03.21). Madas, M.A., Zografos, K., 2010. Airport slot allocation: a time for change? Transp. Pol. 17 (4), 274e285. Mendes de Leon, P.M.J., 2013. A multifunctional approach towards slot allocation. German J. Air Space Law 62 (4), 553e578. Morrison, S.A., 2001. Actual, adjacent, and potential competition estimating the full effect of southwest airlines. J. Transp. Econ. Pol. 35 (2), 239e256. NERA, 2004. Study to Assess the Effects of Different Slot Allocation Schemes. A Report for the European Commission. NRC Handelsblad, 2017. Boodschap is dat chantage loont. https://www.nrc.nl/nieuws/2017/ 11/02/boodschap-is-dat-chantage-loont-13815483-a1579681/ (Accessed 15.03.21). Odoni, A.R., 2020. A Review of Certain Aspects of the Slot Allocation Process at Level 3 Airports Under Regulation 95/93. PA Consulting, 2019. Improving Slot Compliance: Addressing Slot Scarcity at Schiphol Airport. Report Commissioned by the Ministry of Infrastructure and Water Management. PWC, 2013. Fare Differentials. Analysis for the Airports Commission on the Impact of Capacity Constraints on Air Fares. Rupp, N.G., 2009. Do carriers internalize congestion costs? Empirical evidence on the internalization question. J. Urban Econ. 1, 24e37. SEO, 2018. Secundaire Slothandel Op Schiphol. Report Commissioned by the Dutch Ministry of Infrastructure and Water Management (In Dutch). SEO reportnr, pp. 2018e2029. SEO, 2020. Amsterdam Airport Schiphol traffic statistics. SEO and Cranfield University, 2017. The Impact of Airport Capacity Constraints on Air Fares. Report Commissioned by ACI EUROPE. SEO Reportnr. 2017-04. SLG, 2013. Q6 Review of the Distribution of Economic Rent between Airport, Airlines and Passengers and Cargo Users at Heathrow and Gatwick. Report prepared for the CAA by SLG Economics Ltd. Starkie, D., 1998. Allocating airport slots: a role for the market? J. Air Transp. Manag. 4 (2), 111e116. Starkie, D., 2020. Economic Rents at Heathrow Airport. Case Associates. Steer, D.G., 2011. Impact Assessment of Revisions to Regulation 95/93. Report prepared for the European Commission. Suau-Sanchez, P., Voltes-Dorta, A., Cugueró-Escofet, N., 2020. An early assessment of the impact of COVID-19 on air transport: just another crisis or the end of aviation as we know it? J. Transp. Geogr. 86, 102749. Van Dender, K., 2007. Determinants of fares and operating revenues at US airports. J. Urban Econ. 62 (2), 317e336.

CHAPTER 9

Different approaches to airport slots. Same results, same winners? S. Sera Cavusoglu

CERIS, Instituto Superior Técnico, DECivil, Transportation Systems, Lisboa, Portugal

1. Introduction The rapid transformation of the air transport system and the rapidly growing amount of traffic have caused considerable congestion and delay issues at airports and has become a major transport strategy concern (Cavusoglu and Macário, 2021, p. 1). Airports are facing capacity problems more and more everyday, often the runway capacities are constrained, and airports are unable to accept additional aircrafts (Cavusoglu and Macário, 2021, p. 1). Technological advances will reduce the need for policy limitations. However, it can also be considered that the production and full introduction of these developments would most likely take them years to complete. As the airport capacities have been becoming limited more and more everyday, the discussion of how the airport capacity is allocated and shared between the airlines has taken airlines’ extra attention in the past decades. Limited capacity leads to less competition for the airlines; however, the capacity that is available needs to be allocated as efficiently as possible but also in a fair and competitive way between incumbent and new-comer airlines. Airports in the Europe are regulated by an administrative mechanism that restricts aircraft movements according to the declared runway capacity of the airport (number of slots available at airport per hour). On the contrary, market-based mechanisms (a first comeefirst served basis rule) are applied in the United States with the exemptions of five busiest airports: Newark Liberty International Airport (EWR), John F. Kennedy International Airport (JFK), Los Angeles International Airport (LAX), Chicago O’Hare International Airport (ORD), and San Francisco International Airport (SFO) which are restricted according to high density rules and subject to International Air Transport Association’s Worldwide Slot The Air Transportation Industry ISBN 978-0-323-91522-9 https://doi.org/10.1016/B978-0-323-91522-9.00017-8

© 2022 Elsevier Inc. All rights reserved.

195

196

The Air Transportation Industry

Guidelines while a secondary trading between carriers is also allowed (Cavusoglu and Macário, 2021, p. 2). In our study, we present airport slot allocation approaches around the world and uncover the problems and inefficiencies created by them. The solution alternatives to both approaches are discussed briefly and a new auction mechanism and a model for airport slots allocation (ASAM) is proposed. This model is presented as a useful and generic model to be used by airport authorities around the world. Most of the studies in literature related to auctions focus on bidding strategies and equilibrium but not on the components of decision making that define the bidding behavior of a firm. Therefore, with this study it is aimed to contribute to these gaps in the field of transportation systems as well. Components of bid price, bidding strategies and airlines’ learning mechanisms in the decision-making process is determined and related algorithms are developed in this study. A mini auction market of Heathrow Airport with experimental results of application of ASAM at this airport is presented as a case-study. Finally, the proposed model and results are discussed, and conclusions of the research are summarized.

2. Airport slot allocation approaches in the world and the problems emerging Airport capacity can be allocated in various ways; slot allocation and traffic distribution rules for administrative capacity management; market-based capacity management; combination of both mechanisms; or no regulatory framework as implemented in the United States, with the exception of the busiest airports, EWR, JFK, LAX, ORD, and SFO (Fig. 9.1) (Cavusoglu and Macário, 2021, p. 2).

Figure 9.1 Approaches to airport slot allocation (Cavusoglu and Macário, 2021).

Different approaches to airport slots. Same results, same winners?

197

All over the world airports are categorized at levels according to their degree of congestion as Level 1, 2, and 3 (International Air Transport Association (IATA), 2019, p. 13). At Level 3 airports, demand may significantly exceed the airport’s capacity and slot coordination and its approval is required to prevent inherent delays (Cavusoglu and Macário, 2021, p. 2; International Air Transport Association (IATA), 2019, p. 24). An airport slot is defined as a “permission given by a coordinator for a planned operation to use the full range of airport infrastructure necessary to arrive or depart at a Level 3 airport on a specific date and time” (International Air Transport Association (IATA), 2019, p. 14). Slot allocation mechanism is the standard way to control capacity in slot-coordinated airports around the world (International Air Transport Association (IATA), 2019). The allocation of airport slots is a strategic tool used 5e6 months before the start of planning operations for the allocation of capacity over the long term (Fig. 9.2). European airports enforce regulations and administrative structures that regulate the number of allowed aircraft movements according to the declared runway capacity of the airport (number of slots available at each airport per hour) while in the United States a first comeefirst served basis rule applies (except for five airports that are restricted according to High Density Rules) (Cavusoglu and Macário, 2021, p. 2). Delays are significantly lower in Europe than in the United States (The European Parliament, 2016, p. 13). However, the US slot allocation system is more able to take advantage of existing airport infrastructure while the overweighted European slot allocation schemes do not allow the efficient use of current infrastructure (Cavusoglu and Macário, 2021, p. 2) (Fig. 9.3).

Figure 9.2 Slot allocation at the strategic and operational levels in slot coordinated airports (Cavusoglu and Macário, 2021).

198 The Air Transportation Industry

Figure 9.3 EU versus USA-Delays by cause-2017. Based on EUROCONTROL on behalf of the European Union and FAA Air Traffic Organization System Operations Services, 2019. 2017 Comparison of Air Traffic Management-Related Operational Performance: U.S./Europe, p. 58. (Based on EUROCONTROL on behalf of the European Union and FAA Air Traffic Organization System Operations Services, 2019. 2017 Comparison of Air Traffic Management-Related Operational Performance: U.S./Europe, 58p(EUROCONTROL on behalf of the European Union and FAA Air Traffic Organization System Operations Services, 2019).)

Different approaches to airport slots. Same results, same winners?

199

In the EU, airport slots are defined as take-off and landing rights. However, even though the trading of slots is permitted, the absolute selling or buying of slots is not permitted, because the slots have no economic value under the law (Cavusoglu and Macário, 2021). In the EU, airport slot coordinators either allocate slots to operators by placing or rescheduling (alternative slot times) slot requests or reject some requests if no possible slot is available in order to meet capacity constraints (Fig. 9.4). “Grandfather Rights” and “Historical Slots-Time Adjustment” rules are applied by the coordinator that the carrier reserves the right of continuing its operation next season under the condition that it operates its slots at least 80% of the time during that current season (use-it-or-lose-it rule). Both in the United States and in Europe, difficulties of administrative slot control mechanisms have been published within numerous researches (DotEcon Ltd., 2001; Haylen and Butcher, 2017; Mott MacDonald and European Commission, 2006; NERA Economic Consulting, 2004). These studies showed the inefficiencies created by the overruling that administrative slot allocation mechanisms caused and emphasized the obstacles that exist for the potential market entrants and that very little room is left for competition within the system (Cavusoglu and Macário, 2021, p. 5). A recent analysis regarding airport seat capacity distributions at London Heathrow and Paris CDG indicates that nearly 50% of the seats at Heathrow and 60% of the seats at CDG are allocated to the incumbent flag carriers and the rest of the slots are mainly shared by international fullservice airlines (CAPAdCentre for Aviation and OAG, 2020).

Figure 9.4 Slot allocation scheme in the EU (Cavusoglu and Macário, 2021).

200

The Air Transportation Industry

Although slots have no economic value under the regulation in the EU, some slots are very valuable assets for some of the airlines and, although no ownership is defined under the regulation, airlines prefer selling them to new entrants despite the regulation. Under the table negotiations have been reported for quite some time and recently, the information regarding sale of one AF-KLM slot to a Gulf operator for 75 million USD was dispatched. “Fake” or “artificial” slot exchanges have been practiced as the key method of buying, selling, and leasing of slots since many years to comply with the regulatory requirements (Steer Davies Gleave, 2011, p. 84e85). An improved proper mechanism that ensures fair competition between incumbent and newcomer airlines is apparently necessary. Thus, it is also necessary to re-define slots in terms of ownership rights in this context (Cavusoglu and Macário, 2021, p. 4; Gillen and Starkie, 2015). There is ample anecdotal proof that airlines tend to hang on to slots not just for their own use, but also to retain their competitive advantage in the market and usually apply for a larger number of slots than they currently require. Eventually, a large level of slots is returned too late to be reallocated to another carrier. This impacts the efficiency of the slot distribution system and increases the residual workload for both the coordinators and the airline planning departments (Cavusoglu and Macário, 2021, p. 4). In the administrative regulatory system, incumbents and dominant carriers are favored. They maintain the right to take-off and land on the same slots they previously used. Therefore, few slots are left for potential market entrants and they are off-peak hour slots most of the time.

3. Discussion of the solution alternatives to the problems emerged from allocation approaches Political, economic, and social workability are requirements to apply the transport regulations which are primarily concerned with safety, quality, and competition. Transport regulations have to be cost-effective, create negligible transaction costs and account for its own sectoral impact (Cavusoglu and Macário, 2021, p. 2). When the airport authority simply cannot answer additional required slots and congestion problems occur due to this limited capacity available, this is called scarce capacity. Thus, scarce capacity and congestion are associated. The congestion-management approaches are classified into two major classifications: price-based and quantity-based. Either the airport authority declares a charge per flight that airlines must pay to use a congested airport (price-based) which is congestion pricing or the airport

Different approaches to airport slots. Same results, same winners?

201

authority claims a total flight volume that is allocated as fixed number of airport slots with no charge (quantity-based) (Brueckner, 2009, p. 681). This regime with an allowed secondary trading between carriers is existent at slot-controlled airports of the United States (Cavusoglu and Macário, 2021, p. 3). Another quantity-based method is slot auctioning where the airport authority would allocate slots through an auctioning mechanism. Slot auctioning has been proposed by Federal Aviation Administration in 2008 but the proposal was canceled shortly before getting into force due to the economic crisis the nation was facing. Two extensively researched approaches by the policy and research network to the slot allocation are market-based mechanisms using economic instruments by economic theory justifications and, IATA-based administrative slot allocation mechanism improvements (De Neufville et al., 2013, pp. 409e445). Fig. 9.5 addresses alternatives to IATA based administrative solutions and market-based mechanisms that have been proposed so far in order to improve deficiencies of the current system (Cavusoglu and Macário, 2021, p. 4). Alternate market mechanisms allow the allocation of airport slots to the ones who value them most and promote competition between all types of carriers (incumbent airlines and new entrants) (Cavusoglu and Macário, 2021, p. 4). Market mechanisms mainly uncover the economic value of the slots, allow the development of a productive secondary market between airlines, provide use of larger aircraft resulting in carrying more passengers and advance airlines to use slots more efficiently (Harsha, 2009).

Figure 9.5 Slot allocation solution alternatives (Cavusoglu and Macário, 2021).

202

The Air Transportation Industry

The main problem with respect to development in the welfare economy is whether a commercial approach would restore the socially optimal form of goods (prices, quality, market entry) (Cavusoglu and Macário, 2021, p. 4). Three wide factors will lead to problems: distributional equity, externalities and the business structure (Dixit and Stiglitz, 1977, p. 297). Airlines do not want to start paying for the slots and do not want these prices to be released in such a scenario. In theory market-mechanisms increase equity but also may raise distributional issues that’s why they may need to include appropriate compensation mechanisms, too. Slot auctions is an influential mechanism that represents the prevailing market demand trends, the wishes of all bidders and an instrument that produces an equal and fair atmosphere in which all carriers are interested. Prices are decided by competition between prospective buyers and with proper modeling, transparency, and price discovery could be made possible. The auction mechanism should be so designed that the airlines with different market and purchasing powers could have the possibility of getting slots of peak-hours at congested airports. While slot auctions have the potential to raise regulatory concerns, it is especially important to focus on the economics behind this initiative. There are many questions that has to be answered such as how many companies will attend the auction, how will the bids be assessed and awarded? An auction’s performance strongly depends on the details of its mechanism design. The design of the auction mechanism has to take into consideration the characteristics of the market and provide transparency. Especially in government applications, the goal of the auction mechanism is social surplus maximization. However, forming an incentive mechanism in the auction design in which bidders face no issue of transaction specific assets is also necessary. It should not be assumed that any market-based mechanism will produce allocation efficiency where some actors have market power. In the next chapter, the features of the proposal that may affect its performance is discussed as well as some of the important characteristics of the market that should be taken into account when designing the auction and airline bidding behavior.

4. Proposal of a new and untraditional auction mechanism for airport slot allocation A new approach to airport slots allocation auction mechanism inspired by Sponsored Search Auctions (internet keyword search auctions) is proposed

Different approaches to airport slots. Same results, same winners?

203

in this study and it is intended to provide a comprehensive approach to slot management at congested airports. The proposed model “Airport Slots Auction Model (ASAM)” aims to develop a market and induce competition by seasonally auctioning a limited number of slots (peak-hour slots). A web search results page includes a list of query results and, a list of sponsored links which have been paid by advertisers. Each time a keyword or a combination of keywords are searched in a search engine, an auction is run in real time between advertisers who had submitted bids on these keywords and trying to maximize their utilities and the auctioneer (the search engine) who is to decide which advertisers’ links are displayed and in which order (ranking) in results pages (Roughgarden, 2016, p. 16e17). Each ad is assumed to have a click-through-rate (CTR) depending on its slot (position, rank). Slots higher on the search page are more valuable than lower ones (Roughgarden, 2016, p. 17). Each slot can only be assigned to one advertiser. In this auction mechanism, the advertiser is charged when a user actually clicks on the Ad. In other words, the search engine matches the subset of selected Ads to slots based on a matching criterion called ranking function. Sponsored search auction mechanism has many arguments matching with airport slot allocation. Assuming that advertisers are airlines, and the auctioneer is the airport slot allocation authority whose objective is allocating this public resource in a fundamental aspect that maximizes social surplus, one can employ a similar incentive compatible mechanism where truthful bidding is encouraged. Each slot position in the results page can be assumed as the airport slot listed in a particular time-window which is going to be auctioned and each winner airline is matched with only one slot in that particular time-window similar to the mechanism in sponsored search auctions. 4.1 ASAM model architecture The proposed model for slot allocation is intended to be used at a congested Level 3 airport during strategic level of allocation (Fig. 9.6). This model architecture is the baseline framework for further analysis in this study. Institution, Environment, and Behavior are the three distinctive dimensions in the design of any market (Smith, 1989, p. 153). The description of the model hereinafter follows the natural ingredients of this designed market. The environment consists of collection of all agents and

204 The Air Transportation Industry

Figure 9.6 Global framework of determination of slot capacity and its allocation between airlines.

Different approaches to airport slots. Same results, same winners?

205

their characteristics. Agents’ utility functions, preferences, resources, cost functions, individual demand, and supply curves, they all define the environment. The Institution consists of the messages and actions that were defined within the language of communication. The institution specifies the order of agent movements and messages and their related rules, conditions and terms of contracts and allocations. Behavior is concerned with each agents’ choices of actions subject to defined environment and institution. In this section, the end-to-end auction process that is proposed with this model is presented by this means. The Institution An airport take-off and landing slots auction based on auction based on keyword search auctions ranking, awarding and payment mechanism is defined as the Institution. New entrants are treated fairly by the rules of this mechanism. All bids will be conducted as sealed bidding. Every bidder will bid one value per slot set in the defined auction set. Landing and take-off slots are auctioned separately. Different than keyword search auctions, there are four simultaneous bidding cycles (rounds) per auction set per time window. At the end of each round ranking is announced. After the fourth ranking, awarding is announced. Announcement windows is the only communication channel between bidder agents and the airport agent. Messaging is unrestricted among bidders at any time during auctions. The Environment Airport Landing and Take-off Slots within time-windows are sold by the Slot Coordinator in a digital platform for real-time bidding cycles and communication. The Slot Coordinator is the seller, and the Airlines are the buyers therefore the bidders. The number of slots is evaluated by the airport and this information is shared with the Slot Coordinator. The number of airlines competing, and the number of slots offered are limited in experiments. The number and type of airlines competing are varied in experiments (homogenous and heterogenous distribution). Number of airlines bidding for slots may vary per auction set based on slot demand profiles. The Agents’ Behavior According to worldwide slot guidelines, at Level 3 airports, the role of the slot coordinator is to allocate slots to airlines and other aircraft operators in a neutral, transparent, and non-discriminatory way and also perform slot monitoring and address problems (International Air Transport Association

206

The Air Transportation Industry

(IATA), 2019, p. 25). In the model the slots are allocated by the slot coordinator and thus he is the auctioneer. The Slot Coordinator and the Airline agents are described as follows: The Slot Coordinator The Slot Coordinator is the seller of the airport slots and runs the auction process as the auctioneer. The Slot Coordinator is entitled to be fair and transparent. His responsibilities are the following: • 40 days prior to auctions the auctioneer (slot coordinator) states fully the terms and conditions upon which the sale will be made in a public announcement through his channels. • The Slot Coordinator uses airport declared capacity information received from the airport and evaluates the list of all slots with timewindows and provides this information to all bidders. • The Slot Coordinator requests pre-bids and preferences from all interested parties to be sent in 20 days in “Bidder Information and Pre-bid Form Package.” The Slot Coordinator requests “aircraft type information” within these pre-bids. • The Slot Coordinator determines a quality score for each operator and also demand profiles of each slot before the auctions. • At the end of each auction, the Slot Coordinator evaluates the payment prices and awards the slots to their new owners. • Under certain conditions, in which there is no articulation of interest for a specific slot other than what is stated by the winning bidder, the Slot Coordinator sets a reserved price for the payment. The Airlines Airlines plan their flight schedules, the aircraft types on these routes, flight patterns and arrival/departure times depending on their revenue and cost models and by doing this they create their slot valuations and decide which slots they will propose and bid for. The airlines try to get the arrival/ departure slots suiting their schedules and they evaluate substitute and complementary slots. Airlines’ true valuations depend on their business models and strategies. All airlines try to maximize their profit while some of them are also trying to enter the market. Airlines do not want to deviate much from their previous season’s schedules and want to get the slots in similar time-windows to conduct both their regular businesses and external businesses related to their scheduled flights. However, some Airlines can have more resilience to deviations, too. Some Airlines operate connecting

Different approaches to airport slots. Same results, same winners?

207

flights like regional carriers while others operate point-to-point flights. Airlines target certain periods of the day/week according to their business plans. Therefore, it is critical for the airlines to prefer proper slots according to these business plans. That is why, the airline agents are categorized according to their strategic positioning such as the following: • Low-Cost Carriers, • Legacy Carriers, • Network Carriers, • Point-to-Point Carriers, • Regional Carriers, • Cargo Carriers. Airline agents have different bidding strategies during the auction process. They learn to re-evaluate the revenue and cost margins over the auction process using the information about the previous bidding and auction cycles. Agents choose from bidding strategies provided and decide. Agents evaluate their strategies and switch to another strategy if they consider that they can achieve better outcomes by doing so. 4.2 Adaption of ASAM auction mechanism to internet keyword search auctions An airport slot-based auction system starts with dividing a day into a set of time-windows and declaration of the available number of slots in each time-window. Each slot has a different operational value for each airline since each airline has a particular passenger demand for that particular slot. Some slots might have similar values for airlines while some slots do not. In the model it is assumed that advertisers are airlines, and the auctioneer (search engine) is the Slot Coordinator whose objective is allocating slots in a fundamental aspect that maximizes social surplus. Each slot position in the results page is assumed as the airport slot listed in that particular timewindow which is going to be auctioned and each winner airline is matched with only one slot in that particular time-window similar to sponsored search auctions. Time windows can be arranged as consisting of similar-value slots (similar demand-rates). In ASAM, Click-through-rates of a slot is based on the probability of a passenger flying with the flight at that particular slot. ASAM aims to discourage airlines from competing for similar slots with time-window flexibility. Airlines would not schedule same origindestination flights very closely and also; they would not try to hold unused slots in order to maintain their grandfather rights. The proposed

208

The Air Transportation Industry

model is targeted to be used only twice a year during schedule conferences in the strategic level of slot allocation before the coming season of operations. It is not a dynamic, continuous auction mechanism like internet keyword search auctions where advertisers have standing bids on the keywords that were searched on. That is why, different than sponsored search auctions, four simultaneous bidding cycles(rounds) per auction set per time window are introduced. At the end of each round ranking is announced and after the fourth ranking, awarding is announced. Payments are not received right after auctions are completed. An airline pays a price depending on the number of tickets it sold during the season for which auctions were held. Therefore, payments are done after the season ends and “sold-ticket” numbers are evaluated. In current practice, airline service and operation quality (e.g., on-time performance, load factors, passenger satisfaction) is not an argument that is taken into account when allocating slots at the strategic level. In sponsored search auctions, the bidders can be assigned quality scores (Roughgarden, 2016, p. 17). Similarly, in our mechanism, each airline bidder has simply a quality score and the click through rates of a bidder airline in a specific slot is the product the Airline Quality Score and the Click Through Rate of that slot. With this option, the auctioneer has the flexibility to define and identify quality scores for airlines for the coming season’s auction and formulate fair options for new entrants who have no previous flight histories in order to encourage and support their participation in the auction. In ASAM, ex ante passenger-slot numbers play role in determination of click-through rates and so the prices, but payments are done according to ex post passenger numbers carried per slot after the seasons ends. If a carrier pays a slot relatively a lower price but carries higher numbers of passengers than expected, this high passenger per slot numbers will be reflected on the following seasons’ click-through rates therefore on following season’s prices. 4.3 The conceptual structure of the model and the auction mechanism The model mechanism allows segmentation of peak and off-peak hours and time windows of a day. Time windows can be arranged depending on the airport’s specific congestion profile and on the on the pre-bids of airlines. The number of actual landing and take-off slots at a congested airport is large enough to decide beginning auctioning from peak-hour slots. In this

Different approaches to airport slots. Same results, same winners?

209

Figure 9.7 Conceptual structure of ASAM: four bidding and ranking cycles (rounds) in each auction.

conceptual model, bids are made for time windows that consist of peak hour slots (Fig. 9.7). The number of slots generated by the Slot Coordinator in a time window can be any distribution. For the applicability of the model, each auction consists of four simultaneous bidding cycles (rounds), at the end of each round ranking is announced. Each simultaneous bidding cycle occur every 15 min in each auction set. All bids are conducted as sealed bidding. Every bidder bid one value for the slot set auctioned in each cycle of the auction. The information about each airline’s bids is not known to bidders during the auction. The Slot Coordinator shares incomplete information about the bid prices. At the end of fourth ranking, the Slot Coordinator evaluates the payment prices and awards the slots to their new owners. In the scope of the model, in the cases in which the model works is that for an airline bidding, all the slots auctioned in an auction set is same with having just one of the slots auctioned. For example, let us assume that three slots are auctioned in an auction set, and four airlines are participating in the auction. If these three slots are like having one “keyword” in search auctions meaning that these four airlines want one of three slots, then the model can be used. Of course, some of the airlines may have a preference for slot 1, slot 2 or slot 3 but if they are awarded a different slot than their preference, they will accept it anyway because they want to fly to that airport although within the pool of slots there is one better than the other. Consequently, for an airline participating in the slot auction, all three slots

210

The Air Transportation Industry

will have more or less the same value for her. As a result of the mechanism used, one of the airlines will not be awarded any slots. In the real world, the slot policy regulation prevents every airline from getting a slot and in our model, the looser airline will still have the opportunity to participate in another auction set.

5. Analysis of airline agents’ bidding behavior in ASAM There are three non-trivial questions a player/bidder face and has to make a decision in our Airport Slots Auction Model (ASAM): 1. How much should he/she bid in the first round? 2. How much bid-increment should he/she apply in the following round? 3. What would be his/her walk-away price (Threshold) in the final round? In practice, there are multiple criteria that affect how bids are evaluated (Cagno et al., 2001, p. 313). However, in real-life, in a bidding environment, firms may use subjective evaluation methods. Most of the time the level of competition and customer expectations force the bidders re-evaluate their competing bids. Bidders face the uncertainties over estimated costs, other bidders’ profiles, estimated revenues and evaluation of probability of winning desired slots. These uncertainties start to fade when more information is known to the bidder during the bidding process. 5.1 Components of bid price determination The competitive value of the bid depends on factors that define the profile of a firm. The attributes of a firm given in Fig. 9.8 are found to have direct impact on its bidding behavior. In our model the financial status and market-share strategy of the airline determine the airline’s risk attitude and thus bid price determination and bidding behavior. Relatively new

Figure 9.8 Components of bid price determination.

Different approaches to airport slots. Same results, same winners?

211

companies prefer “Building Strategies” in order to increase market share and whenever the market position of a company is satisfactory, or growth seems costly it follows the “Holding Strategy” and aims to hold its statusquo (Buzzell et al., 1975). Maintaining market share has strong dependency on rate of return on investment and both company’s competitive position and other competitors’ strategies. “Harvesting Strategy” permits the market share to decline while short-term income and cash flow is maintained. The connection between market-share and operating profit margin has been now widely recognized and it has been also clear that the market share and the return on investment are related. Companies want both high operating profit margins and big market shares. Operating profit margin is a ratio that describes how much profit a company makes for every sale and is expressed as a percentage. On the other hand, market share describes how much that company sold compared to all sales by all companies in the market. It takes time for a company to gain market share and it is an indicator for growth. Profitability ratios help to determine the efficiency and success of the business and in fact, operating profit margins exceeding 12% is rare for airline industry (Vasigh et al., 2015, p. 792). Although legacy or flag carriers make more revenue than low-cost carriers, low-cost carriers are also significantly profitable due to the average fares paid per km. In general, an operating profit margin greater than 10 is assumed high, 7e10 normal, 5e7 low, and 0e5 very low. Operating profit margins can fluctuate during fiscal years while the market share stays steadier and is a more reliable indicator. Companies having larger market shares have strong competitive positions and are highly profitable as a result of their greater market power (Buzzell et al., 1975). Higher prices and quality present a unique competitive position for market leader. Nevertheless, a company with a large market share can find itself in a difficult position when a company with a smaller market-share is selling cheaper products with comparable service quality. All airlines are not alike, and they have different financial risk preferences. Their business strategies differ, and these strategies determine the extent of risk and uncertainty that they are prone to take in. These decisions define their risk attitudes. The bidding behavior is not only a result of the bidding firm’s single attribute; risk preference; but also, a combination of all previously defined attributes of a firm as decision factors and their interaction among eachother contributing to bidder’s bid decision. Therefore, a flowchart of decision making is developed as follows (Fig. 9.9).

212

The Air Transportation Industry

Figure 9.9 Determination of business and bidding strategies in ASAM.

The risk for airlines associated to bidding is the risk of not getting the slot they bid for, therefore the probability of losing the slot. Losing a slot has a direct impact on an airline’s network performance. The risk for an airline bidding is the impact of losing the desired slot on the network performance. The relationship between not getting the slots desired and its impact on the network performance of the airline are translated into the risk attitudes of the firms. In the model bidders consider a risk allowance in the bid value that is usually added to the valuation as a percentage of valuation. Risk attitudes are classified into four groups in the model: • More risk-tolerant, • Moderately risk-tolerant, • Moderately risk-averse, and • More risk averse. Usually risk-averse companies have high budget thresholds and want to avoid big losses. A company’s decision to how much bid depends on its evaluation of risk amount versus uncertainty. Risk-tolerant companies take big risks to gain potential high profits and usually they have low budget thresholds (Kim and Reinschmidt, 2013, p. 280). 5.2 Determination of bidding strategies and introducing learning mechanisms to airline agents in ASAM Our system is a simulation of auction games with different types of agents having different business models and risk attitudes. An airline does not know exactly how much another airline’s valuation of a particular slot is.

Different approaches to airport slots. Same results, same winners?

213

In reality airlines will try to learn about their rivals’ past bidding prices when the system of auctions is in place for several seasons. Our agents in simulation use learning mechanisms in parallel to this expected future reality. A minimum increase price (MIP) adjusting strategy to determine it in each round was also designed (e.g., if the bidder bids a new price, the increase must be higher than the MIP). MIP can be assumed as a threshold of offering a new bid price and protects the auction from underbidding of agents. Before the start of the auctions and during auctions, airlines are expected to set their goals regarding the total number of slots they target to get at the end of the auctions, and they are expected to behave in parallel to their goals while deciding their bidding strategies for each auction. Most of the time bidding starts with all airlines bidding below their maximum bids. This process continues till last cycle and until the highest bid is reached by one of the airlines. At the end of each bidding cycle during an auction, the Slot Coordinator shares information to each bidder; a ratio of how close the bidder’s bid to the highest bid in the last round was. In our model the bid thresholds (initial bid and the walk away price) of airlines are determined by setting upper and lower limits. There are three main greedy bidding strategies for players during sponsored search auctions; Balanced Bidding; Competitor Busting and Altruistic Bidding (Cary et al., 2007). An airline chooses a bid for the next round to maximize its utility assuming that all other bids of other bidders for the next round will remain fixed to the bids of previous round while using a balanced bidding strategy. An airline might choose to bid as high as possible in order to make other players pay as much as possible or bid as low as possible to try to hurt others as much as possible. An airline can decide to apply one of these strategies during auctions depending on its goal. Bid increment decision in each following cycle is the most important and critical decision that an airline has to make during the game played and these decisions translate into airline’s behavior. The bid increment of an airline will try to serve the optimum for her to reach her goals during the auction. Learning mechanisms are formulated in the model as decision functions for airline agents to help them decide their optimal bids during each bidding cycle of an auction including the decision about entering the auction before auctions; thus, there are five learning mechanism algorithms developed for this model (Fig. 9.10).

214

The Air Transportation Industry

Figure 9.10 Introducing learning mechanisms for decision-making during the bidding process of an airline.

Learning mechanisms are based on collecting the historical bidding data from previous seasons and making assessments based on these data and also based on risk attitudes, goal ratios and true slot valuations and thresholds. It is very likely expected from a learner airline agent to assess its chances of winning an auction for a specific time-window when past history is available. The agent checks past history of all bids given so far for the timewindow regardless of assessing who its competitors are since this information is unknown before the auctions. For example, in the first cycle of auctions, an airline agent knows its competitors. It can identify the maximum bids that had been given by its competitors in the previous seasons for this particular time-window and therefore can make an assessment. It is expected from the agent to bid as low as possible if its chances are too low and it is likely for an agent to apply balanced bidding if its chances are close to high or high. At the end of each cycle, an airline with a learning capability can make an evaluation of its position in the auction, can assess the bid increment it can, or it should apply and decide what to bid in the next cycle. In the next section, ASAM is furthermore applied to a mini auction market of a historically busy airport that experiences heavy traffic flow and serves as one of the main hubs of international travel. Among the most congested airports in the EU, it is also one of the most desired airports that airlines from all over the world want to enter its market and plan operations to.

Different approaches to airport slots. Same results, same winners?

215

6. Case study: application of ASAM to a synthetic auction market of Heathrow Airport In this section, allocation of an airport’s peak-hour arrival slots to the airlines by the application of ASAM is conceptualized and will be tested in a mini auction market and competition between all types of carriers (incumbent airlines and new entrants) will be endorsed. New entrants are expected to obtain slots easily if they have financial resources. The economic value of the slots which will be set by the market will be uncovered with this application. A synthetic mini auction market of Heathrow Airport arrival slots is implemented in this section. The reason for choosing a specific market is both decreasing the level of complexity of the model and focusing on a congested level three airport to capture the market dynamics with capability of predicting and interpreting the results. Considering the expansion plans of Heathrow Airport and the number of new available slots of the future runway, the allocation of slots will come under close examination and the battle for slots will take a new shape between incumbents and new entrants. The airlines’ data used in this section were taken from annual financial statements of actual, real airlines that reflect airline types used in the experiments. Legacy1 is a dominant incumbent carrier that uses Heathrow Airport as base and Low-cost1 is a dominant incumbent carrier in the United Kingdom. Other legacy carriers are also actual airlines that operate to Heathrow regularly. Regional1 and Low-cost2 airlines are operating to London City and London Gatwick Airports in real-life but in the experiments, it is assumed that all airlines are operating to Heathrow Airport. All other data are collected from internationally recognized world-wide organizations’ reports and statistical bulletins. Passenger per slot numbers in the experiments reflect real statistical numbers of passengers arriving at this airport daily, during these specified slots and CTR values are derived from these numbers. Airline schedules, interests of new entrants and the airline quality scores are hypothetical. 6.1 Specifications of the auction market The nature of operations at Heathrow is such that demand has historically been very high in early morning hours. There are nine airlines that are competing for a total of 112 peak-hour arrival slots at Heathrow Airport (Table 9.1). Two of these airlines are new entrants each having a different profile (low-cost and legacy). Since Heathrow airport is an international

216

The Air Transportation Industry

Table 9.1 Airline agents business strategies, risk attitudes, and quality scores.

market for carriers, each carrier plans international routes on the slots they require. The Slot Coordinator assigns “Airline Quality Scores” to each carrier out of 100 and lets the new entrants hold a score of 100 for their first season of operation. This score is assumed to be based on “On-time Performances” of the airlines at this airport and re-evaluated at the end of each season. Table 9.2 shows the existing schedules of the airlines at this airport for the time windows that are going to be auctioned for the following season. New entrants who have already shown interest on bidding for the slots they desire are shown on the first column. Passenger per slot numbers and the click-through rates of slots can be found in Table 9.3. These numbers are derived from statistical arrival passenger data of Heathrow and calibrated Table 9.2 Existing schedule of airlines for the auction time-windows.

Different approaches to airport slots. Same results, same winners?

217

Table 9.3 Passenger per slot numbers and click-through rates of slots and the auction groups.

accordingly for the model. The Slot Coordinator arranges 10 auction groups by considering existing schedules of the airlines and new entrant’s desires (Table 9.3). 6.2 Experiments and results Two experiments and their results are presented in this chapter. Experiment 1 formulates an auction market scenario where there exists a past history of 10 seasons of bidding. The Slot Coordinator Agent does not consider any Airline Quality Scores. Experiment 1 can be assumed like a base-scenario experiment. In Experiment 2, the Slot Coordinator Agent applies the Airline Quality Scores in the final round of each auction which means the quality score has the power of creating an impact on the airlines’ ranking after the bids are given in the final round, therefore it is significantly important for each airline. This impact is shown in the results of Experiment 2. In each experiment a total number of 10 Auctions each having four simultaneous bidding rounds (cycles) are run and 112 peak-hour slots are auctioned.

218

The Air Transportation Industry

Figure 9.11 Experiment 1dhow rankings and prices change in each cycle in an auction.

Fig. 9.11 shows results of only one of the auctions (Auction 1) that were run in Experiment 1. It shows how rankings change in each cycle and bid prices vary depending on the learning mechanisms of airline agents. Valuation of a slot and the bidding price is highly affected by the route-km. The loosing airline’s bid price has a direct impact on the payment prices while the highest winning bid does not have a direct impact on the payment prices but effect the bid increments of all players during the auctions thus the mechanism encourages the bidders bid truthfully since otherwise, they’d make a negative utility. Table 9.4 gives the new schedules of airlines with the payments and profits and Fig. 9.12 shows the new slot distribution in the airport. At the end of experiment 1, the regional airline couldn’t get any slots and was removed from market in the new season and the newcomers (legacy and low-cost) entered the market instead. One incumbent low-cost (dominant) lost one of its weekly slots (7 slots per week) while one legacy (not a dominant) added weekly slots to its operations (7 slots per week).

Different approaches to airport slots. Same results, same winners?

219

Table 9.4 Experiment 1dairlines’ schedules, payments, and profits.

Figure 9.12 Experiment 1dthe new airport slots distribution without airline quality score application.

220

The Air Transportation Industry

Figure 9.13 Experiment 2dimpact of airline quality scores on the results.

In Experiment 2, the Slot Coordinator applied the Airline Quality Scores at the final round of each auction (Fig. 9.13). With the application of the quality scores, one of the incumbent low-costs is removed from the market while New-entrant1 (low-cost) replaced its slots instead.

7. Discussion of the proposed model ASAM and results In the model, the slots were awarded to the airlines who value them most and it was possible for the new-entrant airlines with different profiles (lowcost, legacy, etc.) to enter this market. With this model, an incumbent network carrier and an incumbent low-cost have the possibility of getting similar slots for the same route. Since the flight route together with the airline’s unit measures of revenue and cost per available seat-km has strong impact on valuation and bidding power of the airline in its auction group, the participant selection for the auction set is found to be particularly important. That is why the pre-bids collected prior to auctions (40 days prior to auction date in our proposed model) is extremely important for the Slot Coordinator to picture the demand and prices for each slot evaluated by airlines. If the Slot Coordinator applies the “Airline Quality Factor” at the final round of auctions, it can be a game changer. All airlines would be encouraged to have better “On-time Performances” by this way regarding how efficiently they use the slots they were awarded within the system that is proposed.

Different approaches to airport slots. Same results, same winners?

221

In the auction model proposed the airlines were encouraged to bid truthfully since otherwise, they’d make a negative utility. Airlines will always have positive utilities in this mechanism because the awarding prices are always significantly lower than their true valuations. If a bidder bids unnecessarily high, this would have an impact on upcoming seasons’ bidding behavior of both his competitors and his, because every bid is recorded in history and past bidding is one of the input variables of learning mechanisms (bid value decision criteria). Therefore, no bidder would like to choose overbidding or underbidding that would harm himself in the future auctions. However, the looser airline’s bid has substantial importance on the payment prices of each winner in the auction by mechanism design. Network effects Although the auction mechanism proposed provides an airline high chances of getting a slot from the time-window auctioned by taking into consideration the airline’s previous schedules, the impact of auctions on the network performance of an airline is inevitable. An airline’s slot valuation is inseparable from its other coordinated and connecting flights. This includes origin-destination scheduling and coordination of flights and airplanes. The complexity emerges for the airlines from here that they have to consider also these network effects before valuing the slots for the auctions. The relationship between not getting the desired slots and its impact on the network performance of the airline were translated into the risk attitudes of the firms in the proposed model. From airlines’ perspective, how much return on investment is convenient for them will be determined in their risk allowances they defined. In the model, bidders consider a risk allowance in the bid value that is usually added to the valuation as a percentage of valuation. Auctioning slots periodically has the potential of creating practical problems for airlines in terms of network planning and this may need strategically restructuring the cost balancing in the short and long-term and network route optimization. Airline industry shows changes in earnings depending on the season and for each airline the earnings per quarter may differ depending on the airline’s business strategy, the regions operations are planned to and the markets they serve, as a result of travel demand (Vasigh et al., 2015). Therefore, the impact of auctions on each airline’s aggregated allocation efficiency would be different depending on above mentioned factors but again this is not a barrier for the slots to be awarded to the

222

The Air Transportation Industry

airlines who value them most. These factors can be managed by the authorities by defining a suitable period for auctioning slots and by defining slot lease periods. However, losing auctions and not being able to operate specific flights at a congested airport will further promote competition and interestingly has also the potential of causing the airlines to turn their faces to other airports. In order to restructure costs, they would optimize flight plans and use alternative airports (Vasigh et al., 2015, p. 247). It is therefore necessary to also look at the positive impacts of auctions on other airports that have the unused capacity to serve as a hub or that is an alternative airport that could help shifting the excessive traffic from congested hubs to them which in turn would stimulate growth in alternative airports.

8. Conclusions Due to the lack of available capacity at congested airports and high demand, a new and untraditional auction mechanism has been proposed to deal with the dilemma: How can the delays be managed while endorsing entry to airlines wanting to operate at the airport and endorsing competition? Development of a fair, transparent, and competitive primary allocation auction market that also stimulates an efficient, transparent and formalized secondary market between airlines is reasonable and possible. Apparently, the slots have great economic value to all airlines and these values were uncovered by the experiments presented. A net total benefit and cost estimation that would result under the proposed scheme should be evaluated before implementation of such a scheme. Over the course of the last several years, many airlines raised concerns that incumbent airlines were holding slots and acting to limit competitors’ entry into market. With the proposed mechanism, incumbent airlines would no longer hold slots with low load factors in order to maintain their competitive positions. On the contrary, they would try to have better performances and higher “Airline Quality Scores” to guarantee stronger bidding power during auctions. One of the expected outcomes of this mechanism is that airlines would find themselves encouraged to use larger aircraft that would result in carrying more passengers and thus more efficient use of slots. However, with regards to the growing traffic, it is unquestionably appearing socially attractive to allow more passengers to be able to benefit from existing

Different approaches to airport slots. Same results, same winners?

223

airport capacities through the operation of larger aircraft. Increasing numbers of passengers carried will produce increasing social surplus and net economic benefits. On the other hand, the size of the aircraft is not a complete indicator of slot usage efficiency since airlines have several operations from largest airports serving differing markets by using various aircraft types depending on the traffic and their market strategies. Some challenges regarding development, implementation and maintenance of a web-based slot auctioning platform with secure information sharing or challenges regarding the difficulties for airlines in terms of schedule inflexibility to secure matching slots at origin-destination airports, are foreseen. The proposed scheme has the possibility of producing higher operating costs for some of the airlines. But from another standpoint, the mechanism has the capability of limiting the market dominance by its mechanism design. Airline resistance is expected during a possible transition to a scheme similar to the one proposed in this study. Nevertheless, the use of market-based instruments is the basic piece of answer for justification of the demand and airport airside capacity management. This model can be assumed as the first progressive step to a more extensive path.

Acknowledgments The chapter presented is based on my on-going Ph.D. work and I acknowledge Prof. Dr. Rosário Macário and Dr. Vasco Reis for their significant contribution, supervision and support during our discussions in the development of my thesis and research.

References Brueckner, J.K., 2009. Price vs. quantity-based approaches to airport congestion management. J. Publ. Econ. 93, 681e690. https://doi.org/10.1016/j.jpubeco.2009.02.009. Buzzell, R.D., Gale, B.T., Sultan, R.G.M., 1975. Market share- a key to profitability. Harv. Bus. Rev. JanuaryeFebruary. Cagno, E., Caron, F., Perego, A., 2001. Multi-criteria assessment of the probability of winning in the competitive bidding process. Int. J. Proj. Manag. 19, 313e324. https:// doi.org/10.1016/S0263-7863(00)00020-X. CAPA e Centre for Aviation and OAG, 2020. Paris CDG Airport Now Busiest in Europe, Ahead of London Heathrow [WWW Document]. URL. https://centreforaviation.com/ analysis/reports/paris-cdg-airport-now-busiest-in-europe-ahead-of-london-heathrow542752 (Accessed 3.13.21). Cary, M., Das, A., Edelman, B., Giotis, I., Heimerl, K., Karlin, A.R., Mathieu, C., Schwarz, M., 2007. Greedy bidding strategies for keyword auctions. EC’07 e Proc. Eighth Annu. Conf. Electron. Commer. 262e271. https://doi.org/10.1145/ 1250910.1250949.

224

The Air Transportation Industry

Cavusoglu, S.S., Macário, R., 2021. Minimum delay or maximum efficiency? Rising productivity of available capacity at airports: review of current practice and future needs. J. Air Transp. Manag. 90. https://doi.org/10.1016/j.jairtraman.2020.101947. De Neufville, R., Odoni, A.R., Belobaba, P., Reynolds, T., 2013. Airport Systems: Planning, Design, and Management, Second. Mc Graw Hill Education, New York, Chicago, San Francisco, Lisbon, London, Madrid, Mexico City, Milan, New Delhi, San Juan, Seoul, Singapore, Sydney, Toronto. Dixit, A.K., Stiglitz, J.E., 1977. Monopolistic competition and optimum product diversity. Am. Econ. Rev. 67, 297e308. DotEcon Ltd., 2001. Auctioning Airport Slots. London. EUROCONTROL on behalf of the European Union and FAA Air Traffic Organization System Operations Services, 2019. 2017 Comparison of Air Traffic ManagementRelated Operational Performance: U.S./Europe. Gillen, D., Starkie, D.N., 2015. Congested hubs, the EU slot regulation and incentives to invest. Soc. Sci. Res. Netw. 1e19. Harsha, P., 2009. Mitigating Airport Congestion: Market Mechanisms and Airline Response Models. Massachusetts Institute of Technology. Haylen, A., Butcher, L., 2017. Briefing Paper: Number CBP 488 Airport Slots [WWW Document]. House Commons Libr., UK. URL. http://researchbriefings.files. parliament.uk/documents/SN00 488/SN00488.pdf (Accessed 10.4.20). International Air Transport Association (IATA), 2019. Worldwide Slot Guidelines, 9th Edition. Kim, H.-J., Reinschmidt, K.F., 2013. Effects of contractors’ risk attitude on competition in construction. J. Construct. Eng. Manag. 1e13. https://doi.org/10.1061/(ASCE) CO.1943-7862. Mott MacDonald, European Commission, 2006. Study on the Impact of the Introduction of Secondary Trading at Community Airports Volume 1- Report. Surrey, UK. NERA Economic Consulting, 2004. Study to Assess the Effects of Different Slot Allocation Schemes. A Report for the European Commission, DG TREN. Roughgarden, T., 2016. Mechanism design basics. In: Twenty Lectures on Algorithmic Game Theory. Cambridge University Press, 2016, Stanford University, California. Smith, V.L., 1989. Theory, experiment and economics. J. Econ. Perspect. 3, 151e169. https://doi.org/10.1257/jep.3.1.151. Steer Davies Gleave, 2011. EUROPEAN COMMISSION Impact Assessment of Revisions to Regulation 95/93 Final Report (Sections 1e12). London. The European Parliament, 2016. Research for TRAN Committee: Airport Slots and Aircraft Size at European Airports - IN DEPTH ANALYSIS. Vasigh, B., Fleming, K., Barry, H., 2015. Foundations of Airline Economics, Second ed. Routledge Taylor & Francis Group.

CHAPTER 10

Black swans or gray rhinos on the runway? The role of uncertainty in airport strategic planning Jaap de Wit

Emeritus Professor, University of Amsterdam, Amsterdam, The Netherlands

1. Introduction and research questions Annual air traffic volumes worldwide reflect political, economic, and natural conditions ranging from slowly moving changes to disruptive highimpact shock events. COVID-19 is an example of such a disruptive event in the global economy of 2020. The impact on the air transport system has been unprecedented since World War II. Fig. 10.1 shows a 60% setback in passenger volumes for almost two decades. Recovery may take several years: IATA (Pearce, 2020) and Eurocontrol (2020) expect air traffic only to return to pre-COVID-19 levels in 2024. Gudmundsson et al. (2020) estimate recovery times on average of 2.4 years starting from 2020. This global average regionally differs. 2.2 years recovery time is estimated for Asia Pacific, 2.5 years for North America and 2.7 years for Europe. In the most pessimistic scenario, recovery of air transport demand will take six years on average. The impact of COVID-19 not only differs among regions but also among individual airports with regard to traffic volume and recovery time. For example, the traffic mix of international and national traffic accommodated by each individual airport is an important factor. International traffic is subject to border closures and is more dependent on travel bans, quarantine restrictions, and testing regulations elsewhere than domestic traffic is. Recovery times for individual airports will also depend on the mix of carrier types accommodated. Low-cost carriers (LCCs) are more flexible in adapting labor costs and the scale of their point-to-point operations than

The Air Transportation Industry ISBN 978-0-323-91522-9 https://doi.org/10.1016/B978-0-323-91522-9.00005-1

© 2022 Elsevier Inc. All rights reserved.

225

226

The Air Transportation Industry

Figure 10.1 COVID-19 impact on air transport. (Based on source: Economic Impact Analysis of COVID-19 on Civil Aviation. Note: (a) Historical figures are subject to revision and update; (b) For latest update please refer to ICAO Economic Impact Analysis of COVID-19 on Civil Aviation at: https://www.icao.int/sustainability/Pages/EconomicImpacts-of-COVID-19.aspx.)

network carriers. Network carriers have to operate with new state aid obligations after the COVID-19 outbreak and will restrict these carriers’ decision autonomy in routes to be served (Albers and Rundshagen, 2020; CAPA, 2020). All in all, traffic volatility for individual airports is not a new phenomenon. During the last few decades more and more airports have learned to cope with increasing air traffic volatility. The disruptive COVID-19 event is however unique in the size and duration of its impact on individual airports. The uncertainty that accompanies the causes of this increased volatility renders two research questions: how to characterize a high-impact event such as the COVID-19 pandemic in the context of deep uncertainty and how to cope with such an event in airport strategic planning?

2. Increasing year-to-year traffic volatility at airports Burghouwt (2005) distinguishes three sources of year-to-year volatility in airport traffic, namely a cyclical component, random events and factors related to a free market regime. De Neufville and Barber (1991) observe

Black swans or gray rhinos on the runway? The role of uncertainty

227

significant differences between the traffic volatility1 of individual US airports in the periods before and after the introduction of the 1978 Airline Deregulation Act for the US domestic market. Although each of the three volatility sources might have been relevant for the increased volatility, the obvious conclusion is that deregulation related factors played the dominant role. Burghouwt (2005) applied a comparable analysis to EU airports, be it that the breaking year between the pre- and post-liberalization period is less obvious than the breaking year of 1978 in the United States. The liberalization of the EU internal market was a stepwise process during the period 1988e93. On top of that, the United States introduced a less restrictive international open skies policy in the nineties that strongly stimulated transatlantic traffic at some major EU airports. Furthermore, the resulting volatility ratios for EU airports not only reflect the impact of liberalization, but some major random events also contributed to this volatility in the eighties and nineties (see Fig. 10.1). Examples of both sources of volatility are considered in more detail next. 2.1 Increased traffic volatility at airports in more competitive air transport markets In the US domestic market, the deregulation and emerging new price competition urged major carriers to transform their linear coast-to-coast networks into competing star-shaped hub-and-spoke networks (Borenstein, 1992). Consequently, airports with limited local demand such as Raleigh Durham and Charlotte got the opportunity to be upgraded to new artificial hubs by major carriers. Thereafter, they also ran the risk to be de-hubbed, when majors started to restructure their networks in times of economic downturns and airline bankruptcies or take-overs. In the new competition within the liberalized European market, national carriers transformed their star-shaped networks at their national home bases into hub-and-spoke systems by improving the connectivity between continental and intercontinental flights. These new competitive conditions affected the viability of the smaller national carriers in Europe (Burghouwt et al., 2015). Secondary hubs such as Brussels, Zürich, Budapest, and Milan Malpensa, permanently or temporarily lost their home carrier. This dehubbing is a high-impact event that creates substantial airport traffic 1

Traffic volatility of an airport is the average of annual differences between the actual annual traffic levels and the annual traffic trend. The quotient of the volatility in period t1 and period t0 renders the volatility ratio.

228

The Air Transportation Industry

Figure 10.2 Changing passenger traffic mix of Zürich airport after the collapse of Swissair. (Based on source: Statistical Yearbooks Zurich airport.)

volatility. Fig. 10.2, for example, shows that Zürich airport was only in 2010 able to return to the pre-collapse passenger numbers of 2000. In the meantime, the traffic mix noticeably changed due to the growing number of origin-destination passengers and a stagnating number of transfer passengers. The collapse of Swissair also shows the relationship between transfer passenger volumes (Fig. 10.2) and the number of direct intercontinental connections provided at its hub (Fig. 10.3). If the hub carrier goes out of

Figure 10.3 Direct Euro and ICA connections of Zürich Airport after the Swissair collapse. (Based on source: Statistical Yearbooks Zürich Airport.)

Black swans or gray rhinos on the runway? The role of uncertainty

229

business the transfer volume diminishes and the direct ICA connectivity of the hub shrinks. The local demand from and to Zürich’s catchment area is not sufficient to sustain the total set of existing ICA services. Although Swissair was quickly succeeded by a new national carrier Swiss, the ICA-connectivity of Zürich airport did not return to the pre-collapse level (Fig. 10.3). It is possible that the later take-over of Swiss by Lufthansa also contributed to this development when Zürich airport turned into a secondary hub of the new multi-hub network dominated by Frankfurt and Munich. In more generic terms, Redondi et al. (2012) conclude that the 37 de-hubbed airports in their study took more than five years on average to return to pre-de-hubbing traffic levels. In the meantime, the traffic mix of long haul and short haul flights and the overall connectivity of the de-hubbed airports substantially changed after the influx of the new LCC entrants with their continental point-to-point operations. Bilotkach et al. (2014) demonstrate this impact for Budapest airport and Kincaid et al. (2012) for Baltimore/Washington International airport. The new business model of LCCs is another outcome of deregulation and liberalization that resulted in additional traffic volatility. Traffic volumes at LCC airports and bases clearly demonstrate that LCCs are rather footloose due to limited switching costs between alternative airports and bases. Reasons for switching to other airports are diverse, ranging from disputed airport charges (Belfast City and Shannon) to night-curfews and limited availability of slots (Eindhoven) or better market opportunities at neighboring airports (Barcelona vs. Girona). These traffic downsizings or even complete abandonments of bases strongly contribute to the traffic volatility of individual airports involved. Malighetti et al. (2016) found 109 airport cases worldwide with LCC traffic reductions in offered seats by at least 50%. They also found that the presence of neighboring middle-size airports increases the likelihood of such downsizings. The Spanish airport Girona clearly shows in Fig. 10.4 the volatility of this traffic up- and down-sizing after becoming a Ryanair base. An upsizing surprise in air traffic emerged at Amsterdam airport Schiphol when the United States concluded the first Open-Skies agreement with the Netherlands in 1992 in combination with an antitrust immunity to

230

The Air Transportation Industry

Figure 10.4 Fluctuating passenger volumes at Girona airport. (Based on source: Gironaairport.net.)

allow for a deeper alliance between North West and KLM.2 Amsterdam Schiphol became the European hub in a new transatlantic double-hub system to connect behind- and beyond-traffic in Europe and the U.S. on transatlantic hub-to-hub routes. To that end, connecting times between intercontinental and continental traffic at Schiphol were improved by three daily connection waves in 1992 and this system was extended to six daily waves in 1997e1998. The resulting unexpected traffic growth went hand in hand with a classic mistake in static airport planning: the middle traffic scenario from a set of three alternative scenarios was selected to establish noise contours and insulation programs for Amsterdam airport. Although the three traffic scenarios were based on a broad range of contrasting assumptions and each of them could be equally viable, political, decisionmaking under time pressure tried to speed up the noise contour calculations by choosing only one traffic scenario as input. Due to the new role of Amsterdam airport as an alliance hub, the actual traffic levels surpassed the selected scenario by 50% in 1997. This created substantial 2

This bilateral air service agreement clearly reflected a try-out of the United States new international air transport policy: the Dutch designated carrier was allowed to provide third and fourth freedom services to any point in the United States, while vice versa US designated carriers were allowed to serve any point in the Netherlands. This de-facto boiled down to Amsterdam.

Black swans or gray rhinos on the runway? The role of uncertainty

231

political turmoil (Dutch Parliament, 1998) because an environmental runway capacity crunch threatened. An early master planning revision became inevitable. All in all, due to the increasingly competitive conditions in the air transport market, individual airports in the United States and Europe have been confronted with substantially more traffic volatility in the last three decades. Another cause of increasing airport traffic volatility concerns the earlier mentioned random events. 2.2 High-impact shock events from outside the air transport system Some structural changes in the world have increased the probability of external high-impact shocks in the international air transport system. Climate change and global warming go hand in hand with a growing number of natural disasters. Population growth, global interconnectedness, urbanization, expansion into wild frontiers and global warming contribute to the steeply rising threat of new infectious diseases such as COVID-19. The impact of earlier pandemics of SARS, MERS, Ebola, Zika, Swine flu, and HIV/AIDS mainly differs from COVID-19 with respect to contagion and number of deaths. COVID-19 has caused a shock in air travel that will affect airport traffic worldwide for several years. This makes COVID-19 essentially different from other external high-impact shocks on airports. The Islandic Eyjafjallajökull volcano eruption in 2010 only fully disrupted air traffic at airports in Northern and Central Europe and only for six and a half days. A longer lasting but airport-specific external shock event was hurricane Katrina that hit New Orleans in 2005, causing widespread flooding, billions of dollars of property damage, and more than 1300 deaths. Although New Orleans International airport itself was not significantly damaged, passenger volumes decreased by 39% in 2006 and passenger traffic was still 21% below pre-Katrina levels in 2009 (Kincaid et al., 2012). A further reaching high-impact shock event in duration and geographical impact were the 9/11 terrorist attacks shortly after the burst of the Dotcom Bubble in 2000. The impact of 11/9 was far more pronounced for US airports than elsewhere. It resulted for these airports in two to three years of decreasing traffic from peak to through. Table 10.1 shows the impact for a selection of US airports. The global financial crisis starting in 2008, followed by an economic downturn, was an unforeseen high-impact event for air traffic levels of airports worldwide. The peak-to-through decline in RPKs was not

232

The Air Transportation Industry

Table 10.1 U.S. passenger enplanements after 9/11. Lowest traffic Pre-9/11 level after Percentage of traffic level of 9/11 total decrease 2000 Airport

Atlanta San Francisco Chicago Dallas Las Vegas Minneapolis Newark Miami US total (BTS)

Year of lowest traffic level

80.162 41.041

75.859 28.786

5.4% 29.9%

2001 2003

72.144 60.687 36.866 36.772 34.189 33.621 669.282

66.566 52.830 35.001 32.630 29.221 29.596 616.238

7.7% 12.9% 5.1% 11.3% 14.5% 12.0% 7.9%

2002 2002 2002 2002 2002 2003 2002

Based on source: USA DOT Bureau of Transportation Statistics, and individual airport statistics.

especially deep, compared with the foregoing SARS pandemic and Iraq war impact, 6.3% versus 15.8% according to IATA (2018). The SARS pandemic mostly hit East-Asian airports and the peak-to-through took only 5 months followed by a 1-year recovery. The global financial crisis reached its through in 12 months and the resulting economic downturn took several years of recovery time for airports worldwide. As an example, traffic at Amsterdam airport took four years to surpass the former peak year of 2007 and the decrease from peak-to-through took 9.1%. All in all, the unforeseen high-impact shock events from outside and within the aviation system show diverse but relatively limited impacts on the traffic level and traffic mix of the airports involved. COVID-19 is the only event with an exceptional impact on aviation in the combination of decrease from peak-to-through and expected duration. 2.3 Shock events and economic characteristics of airports The impact of disrupting shock events in the airport industry can only be completely understood in the context of the industry’s economic characteristics. At the heart of the matter, the core business of airports remains it existence as a node in multimodal traffic networks, irrespective of the broader functions an airport may fulfill in broader concepts such as the Airport City-concept (Peneda et al., 2011) and the Aerotropolis-model (Kasarda, 2019). Since airports are primarily a part of the traffic infrastructure, they are characterized by economic characteristics such as longevity, lumpiness and scale economies as well as substantial sunk costs

Black swans or gray rhinos on the runway? The role of uncertainty

233

related to the limited options of its alternative use. But airports as well as ports differ from other traffic infrastructure as corporate entities charging their users to fully cover the infrastructure costs. These characteristics make the airport pre-eminently vulnerable to uncertainty in future traffic volumes. Periodic under- and overcapacity in a stepwise expansion to accommodate growing demand is not a novelty for airports due to their infrastructure indivisibilities, but the COVID-19 impact is very different in creating an unprecedented overcapacity. The resulting financial losses strongly affect the financial robustness of airports as a fixed costs industry. The perception of the COVID-19 pandemic as an unforeseen phenomenon is further explored in the next section.

3. High-impact shock events and deep uncertainty The high impact of the COVID-19 pandemic has been a surprise to many people. It apparently incentivized the idea that this pandemic is a Black Swan. Taleb (2001) developed the black swan theory with regard to financial events and later extended this metaphor to events outside the financial markets (Taleb, 2007). A Black Swan event can be defined as an event with three attributes: rarity, extreme impact and only retrospective plausibility and predictability. The concept of the falsifiable white swan goes back to John Stuart Mill (1835) and Popper’s falsifiability slogan “No number of sightings of white swans can prove the theory that all swans are white. The sighting of just one black one may disapprove it” (Popper, 1934). Taleb’s Black Swan concept is different due to the addition of the high-impact attribute. 3.1 Black Swans or fat tails Since Black Swan events can be characterized as “unknown unknowns,” a term Donald Rumsfeld popularized during a Pentagon news briefing in 2002, no probabilities can be attached to such events, because their existence is unknown. This concept contrasts with known unknowns applied to events we know they exist and to which we attach a probability. Rare but possible events we know, can’t be captured in Gaussian bell-curve distributions. Only “fat-tailed” probability distributions will do. The more extreme case of a fat tail distribution emerges if the tail decays like a power law. A particular class of these probability distributions is described by Mandelbrotian randomness based on a fractal power law distribution and baptized by Taleb (2007) as Gray Swans. It should be noted that all other

234

The Air Transportation Industry

known unknowns of which we know the stochastic distributiondeither bell-shaped or fat-taileddare White Swans (if the tail is not too fat). This distinction between Black and White Swans corresponds with the watershed between Knight’s computable economic risks and incomputable uncertainty (Knight, 1921). One can however dispute whether a Black Swan is a deep uncertainty phenomenon as stated by Marchau et al. (2019). The epistemic issue of an unknown makes a Black Swan an extreme phenomenon that belongs to the category of total ignorance in the model of various uncertainty levels developed by Walker (2003). The Black Swan-concept has increasingly become part of the common language in describing a major disaster. A growing number of scientific publications apply the “Black Swan”-label to the COVID-19 pandemic and its predecessors. Antipova (2021), for example, characterizes the 21st century so far as an age of Black Swans culminating in the COVID-19 pandemic. Sehl (2020) observes that every few years or so the airline industry gets rattled by a “Black swan” event. Mishra (2020) declares COVID-19 to be a Black Swan event due to the “unthinkable” consequences of the nonlinearity in contagions. Apparently, a reference to the Black-Swan label is easily made, if an individual is surprised by a disruptive event. 3.2 Color blindness regarding swans Although the COVID-19 pandemic is frequently called a Black Swan by politicians, academics, and professional consultants, Cirillo and Taleb (2020) and Flyvbjerg (2020) argue that pandemics such as COVID-19 follow a classic fat-tailed distribution and are predictable as White Swans. This corresponds with the evidence of two reports the Center for Strategic and International Studies (November 2019) and the Global Preparedness Monitoring Board (September 2019), issued just before the outbreak of COVID-19. Both reports underline the probability of “a rapidly spreading, lethal respiratory pathogen pandemic.” Lempert et al. (2002) also emphasize that not earlier predicted “surprises” hardly exist. As an example, they refer to the story of General Ridgeway, who as a young officer wrote a war game scenario about a surprise attack on the US fleet at Pearl Harbor. His fellow officers refused to play it because they regarded it as “possibly too improbable.” Paté-Cornell (2012) also argues that true Black Swan-events are extreme ones. In most cases important warning signs can be detected, monitored an adequately responded in proactive strategic planning. The 9/11 attacks for example, were imaginable. The FBI knew that questionable people were taking flying lessons on large aircraft and the attacks were a replica of the 1994 hijack of an Air France aircraft in Algiers bound to

Black swans or gray rhinos on the runway? The role of uncertainty

235

Paris. During a stopover in Marseille the hijackers required far more fuel than needed for the second part of the trip. The intention was to make the aircraft a flying bomb to be dropped on the Eiffel tower. Although 9/11 was called unpredictable, the Director of the National Commission on Terrorist Attacks upon the United States (2004) called the misreading of the precursors to these events “a failure of imagination.” Hurricanes and tsunamis are also often incorrectly labeled as Black Swans. According to Wucker (2016), hurricane Katrina that hit New Orleans mid 2005 was neither a Black Swan, because many predictions on hurricane risk were already available for New Orleans. A year before Katrina hit Louisiana, the Federal Emergency Management Agency sponsored a disaster planning exercise called “Hurricane Pam” based on a scenario of a devastating hurricane hitting New Orleans. However, the lessons learned from the Pam exercise arrived too late. The global financial crisis has also been perceived as a Black Swan by people such as Alan Greenspan (2013) and Végh et al. (2018). Anyway, some people did expect this burst of the housing bubble and the implosion of the related collateralized debt obligation market and anticipated it accordingly (Lewis, 2010). All in all, the Black Swan-label is too often applied in the hindsight to rare, high-impact events, because the signs are missed due to a lack of proactive risk management. This lack may even concern high-impact events which are easily imaginable and/or highly probable. The reason is that they are conveniently ignored as the elephant in the room and primarily reflect an organizational complacency. Wucker (2016) extends the bestiary a little by labeling these high-impact, high-probability events, we do not like to know, as Gray Rhinos. Actually, COVID-19 has been a Gray Rhino for many well-informed public authorities. This is the category of the unknown known, that does not concern epistemic uncertainty about its existence but aleatory uncertainty about its frequency as a fat tail phenomenon (Fox and Ulkümen, 2011; Walker et al., 2003). All in all, most rare, high-impact shock events appear to have another color than the Black Swan and might even represent another animal metaphor. As a consequence, the possibility of these shock events can be detected in advance and should also play a role in airport strategic planning. In response to the COVID-19 pandemic, Linden (2021) proposes a range of generic strategic management theories for the aviation industry “to embrace higher levels of uncertainty of their environment.” However, airport managers need more elaborated applications than this summary of generic theories to cope with the impact of COVID-19.

236

The Air Transportation Industry

4. Absorbing rare, high-impact shock events in airport strategic planning Airport strategic planning is a dynamic and continuous process. Iterative scans of the continually changing external environment to detect future external threats and opportunities is a key component of strategic planning. It should enable an airport organization to anticipate and adapt to these threats and opportunities by proactively thinking about the future and by mapping it. As such, airport strategic planning has a much broader scope than the airport master planning. The airport strategic plan guides all other plans, ranging from the risk management plan to the marketing plan, business plan, master plan and land use plan. The master planning process only focuses on the need for future airport facilities, based on a choice among various planning alternatives for terminals, runways etc. (Ricondo & associates, 2009). Consequently, the risks and uncertainty resulting from the threats and opportunities detected in the strategic planning process should also resonate in the airport master plan. Judgment-based approaches such as scenario building and simulations by serious gaming and red teaming can be helpful to cope with high-impact, rare shock events. Here we only focus on scenario building. 4.1 Scenarios in aviation The first step in exploring the external environment usually focuses on the identification of key driving forces that will shape future opportunities and threats including their levels of uncertainty (Kincaid et al., 2012). SOIF and IATA (2018) provide an extensive risk register for the airline industry containing 50 drivers categorized according to the “STEEP” framework (Society, Technology, Environment, Economy, and Politics). This risk register results from interviews with global trend specialists and experts inside and outside the airline industry. The scenario development team prioritized a set of 13 important drivers in an impact-uncertainty diagram. The 13 drivers concern alternative fuels and energy sources, cybersecurity, environmental activism, extreme weather events, level of integration along the air-industry supply chain, new modes of consumption, price of oil, strength, and volatility of global economy, geopolitical (in)stability, infectious diseases and pandemics, tensions between data privacy and surveillance, terrorism and international regulation of emissions, and noise pollution. Pandemics are among the highest scoring drivers in this set. Two of the most relevant driversdgeopolitics and datadare used by the scenario

Black swans or gray rhinos on the runway? The role of uncertainty

237

development team as axes in a scenario matrix, providing four permutations to consistently develop scenarios by including the other mentioned drivers (Maack, 2001). Although the perspective of the airline industry differs from the airport business, there is substantial overlap in common threats and opportunities. An illustrative example of a scenario drivers risk map in the airport industry in Fig. 10.5 shows some similarities with the airline industry, discussed earlier. Kincaid et al. (2012) not only propose the application of judgment-based approaches, such as scenario building, but also data-driven procedures such as Monte Carlo simulations based on assigned probability distributions to relevant variables. This however ignores some complications demonstrated by COVID-19. This pandemic also triggers knock-on effects of other variables such as the intensified use of internet for meetings, increased airport security requirements, new modes of consumption, an economic recession and the rise of populist movements, to name just a few of the 50 drivers

Figure 10.5 Illustrative example of a risk map for airports: the case of Bellingham International Airport. (Based on source: Kincaid, I., M. Tretheway, S.Gros, D. Lewis, 2012. ACRP report 76: Addressing Uncertainty about Future Airport Activity Levels in Airport Decision Making. Transportation Research Board of the National Academies, Washington, D.C., Figure 33, p. 85. Copyright, National Academy of Sciences. Reproduced with permission of the Transportation Research Board.)

238

The Air Transportation Industry

identified in the airline risk register developed by SOIF and IATA (2018). This interdependence cannot be reflected in the probabilities of individual drivers. Furthermore, some drivers concern rare shock events with fat tail probability distributions instead of the more conventional Gaussian distributions used in Monte Carlo simulations. Therefore, it can be disputed whether the application of data-driven procedures can adequately cope with uncertainty of such disruptive events. 4.2 Scenarios to be integrated in a responsive strategic plan Scanning opportunities and threats enables the development of more robust, risk responsive strategies. Marchau et al. (2019) summarize various approaches such as dynamic adaptive planning, info-gap decision theory and dynamic strategic planning with real options analysis (De Neufville, 2003).3 The dynamic adaptive planning approach in five steps according to the scheme of Fig. 10.6 has been developed by Walker et al. (2019). The robustness of the initial strategic plan can be improved in step III by various actions with regard to the observed threats and opportunities depending on their level of uncertainty. Walker et al. (2019) apply this planning approach to Amsterdam airport for an eclectic set of stand-alone threats and opportunities without the context of coherent scenarios. For example, the EU slot allocation procedures make the indicated threat of de-hubbing conditional to the second indicated threat of a substantial LCC-influx and the third indicated threat of long-haul, low-cost (LHLC) operations. Although the relevance of the third threat as such may be disputed due to the low viability of this business concept, it also needs a rehubbing process (Zuidberg and De Wit, 2020). The absence of a scenario context results in responses to the individual threats and opportunities that can evoke the opposite of planning flexibility due to irreversible solutions, such as too early redesigned terminals (Walker et al., 2019). The same problem arises in the next stage IV of contingency planning. In order to continuously track the risk environment, trigger levels of signposts for individual threats and opportunities are established as well as the required actions when these trigger levels are reached. Maack (2001) also emphasizes the monitoring of progress indicators in the context of determining which scenario is coming to pass. These progress indicators are strongly related to 3

Real option theory can be illustrated by the option to further expand an airport on the existing location in a densely populated area versus the real option of making a spatial reservation for a new offshore airport location.

Black swans or gray rhinos on the runway? The role of uncertainty

239

Figure 10.6 A dynamic adaptive planning approach in five steps. (Based on source: Walker, W.E., Marchau, V.A.W.J., Kwakkel, J.H., 2019. Dynamic adaptive planning (DAP). In Marchau, V.A.W.J., Walker, W.E., Bloemen, P.J.T.M., Popper, S.W. (Eds.), Decision Making under Deep Uncertainty. From Theory to Practice. Springer, Cham. https://doi.org/10.1007/978-3030-05252-2. Unaltered version under Creative Commons license http://creativecommons. org/licenses/by/4.0/.)

the major driving forces included in the scenarios. However, problems arise if these drivers concern uncertain, high-impact shock events. They require a thorough analysis of their causes to detect adequate turning points and trends that enable a continuous scanning of the business environment. The combination of this information for various drivers in a coherent setting may provide a better insight in the timing of contingency actions, but it is not obvious in advance whether such signposts can be determined in case of low-frequency disruptive events. 4.3 Robust airport strategic planning and flexible master planning In order to mitigate the risks and uncertainties of the various scenarios, coherent risk reducing actions have to be developed as contingency plans in the airport strategy. For example, the threat of de-hubbing in a broader

240

The Air Transportation Industry

scenario setting can be mitigated by the hub airport’s airline marketing. A hedging strategy can increase the number of visiting airlines. This reduces the risk of exposure to a single dominant hub carrier. Such an approach sharply contrasts with the current selectivity strategy of Amsterdam airport Schiphol within its current capacity constraints. In cooperation with the Dutch government, “hub related” routes are prioritized based on a minimum amount of transfer passengers. This strategy results in a set of destinationsdespecially intercontinental destinationsdthat is assumed to optimally serve the needs of the Dutch economy. Such an approach ignores, however, the risks and uncertainties for the airport due to the increased dependence on its home carrier. A further reaching diversification strategy for airports can be realized through a stronger focus on non-aeronautical activities (Broekema et al., 2020). This strategy intends to mitigate the financial risks of shrinking traffic volumes. Only passenger independent diversification options have a risk mitigating effect such as stand-alone real estate developments on- and off-airport,4 consultancy services to other airports and stakes in airports elsewhere. Other non-aeronautical activities are directly grafted on the air passenger flows themselves, such as on-airport shopping malls, restaurants, hotels and car parking facilities. They all contribute to a further risk accumulation for the airport. The implementation of such a diversification strategy, however, also comes with new risks and uncertainties depending on the question whether the current core competences are transferrable to the targeted industry (Markides, 1997; Andersen et al., 2011). The various contingency plans based on the triggers for the major drivers in coherent sets of scenarios also affect the airport master planning. The airport master plan must be clearly distinguished from the strategic planning stage since this stage only details the necessary airport expansion on the long run. In its traditional version the most appropriate expansion alternative is chosen from a few long-term expansion alternatives by confronting the existing capacities with one or more traffic forecasts (ICAO, 1987). The underlying traffic forecasts are usually based on a set of assumptions or a demand model that reflects changes air fares, GDP, world trade, and oil prices. Some uncertainty in these assumptions may be reflected in high and low traffic forecasts, but usually the traditional master plan is inflexible and tends to be based on one single future. 4

The original policy of Schiphol airport to primarily attract air traffic related business in on-airport real estate facilities seems to dilute more recently.

Black swans or gray rhinos on the runway? The role of uncertainty

241

4.4 Increasing master plan flexibility In order to integrate strategic and master planning, the existing traffic forecasts have to be confronted with the uncertainties in the threats and opportunities of the various scenarios. The then resulting preferred masterplan alternative is more flexible and responsive to the futures described by the scenarios. Various actions can be taken to improve the master plan flexibility (Kincaid et al., 2012). Common use of aircraft stands through a multi-aircraft ramp system reduces the risk of an unexpected changing traffic mix. This also holds in Europe for the shared usage of gates and terminals through swing gates and piers for Schengen and Non-Schengen passenger flows and for commonuse terminal equipment and common-use self-service. Such measures do not only improve flexibility, but they can also improve efficiency. Land banking through spatial reservations in the wider airport region can be incorporated in the public physical planning process and provides real options for additional runways and terminal expansion. The spatial reservation for a sixth runway at Amsterdam airport Schiphol is an example of such an approach. Spare capacity in terminal space creates more flexibility in handling a changing traffic mix and in implementing new security regulations, such as social distancing requirements during a pandemic. Linear and modular terminal design enables optimal flexibility in terminal extension in different directions in case of a changing traffic mix or an unexpected traffic growth. The design of a dedicated low-cost carrier pier at a hub airport, such as the one at Schiphol airport, creates the opposite in more terminal inflexibility. Self-propelled people movers can better mitigate uncertainty in traffic growth than fixed transit systems of piers and gates. Such a choice also affects the quality of the airport service. The same holds for the flexibility of a tug-and-trolley baggage systems with simple conveyor belts compared with a fixed automated baggage handling system. For example, the then new Denver International Airport was opened in 1995 after 16 months of delay, because one had to return to a less complicated baggage handling system after 10 years of failures in planning and construction of an automated baggage handling system (Lukaitis, 2004). Larger hub airports usually do not have this freedom of choice, because an automated baggage handling system is a key success factor in hubbing. Transfer times between connecting flights are determined by the quality of the baggage handling system and as such it strongly defines the competitive strength of a hub.

242

The Air Transportation Industry

Most of these options will contribute to more flexibility in airport master planning if unexpected changes emerge in the traffic mix and traffic volume. However, if air traffic implodes due to an unforeseen high-impact event and its knock-on effects, such as COVID-19, most of these design principles only have a soothing effect at best.

5. Final observations and conclusions The global aviation industry has experienced an increasing traffic volatility in the last 40 years. Among its causes was a growing number of virus outbreaks and pandemics. The first research question concerned the characterization of such disruptive shock events in terms of deep uncertainty. Due to its unexpected high impact, COVID-19 has been frequently but incorrectly labeled as Taleb’s Black Swan. This pandemic is no more and no less than a White Swan with a fat tail for the aviation industry like any other industry. Only for national public health services COVID-19 has been a Gray Rhino due to a lack of pandemic preparedness and a neglect of early warnings to timely contain this outbreak. The second research question concerned the role of airport strategic planning in coping with such a disruptive event. The focus of the analysis was on subjective judgment approaches and more specifically on scenarios. The scenario drivers represent the main opportunities and threats derived from an extensive scan of the business environment. Pandemics play an important role in scenarios of the aviation industry as a high-impact, highuncertainty driver. It can be disputed however, whether signposts and trigger values can be determined for this type of drivers to improve resilience and robustness of strategic airport planning. To a certain extent, proactive contingency actions can be incorporated in an airport’s strategic plan. Examples of such contingency actions may concern the development of more flexible airport master plans and the implementation of dedicated diversification strategies to improve the airport’s financial robustness. The fact remains however, that the massive impact of COVID-19 on the air travel industry can hardly be mitigated by the industry itself. A public awareness and preparedness for new pandemics by national health organizations remain conditional to a timely containment.

References Albers, S., Rundshagen, V., 2020. European airlines’ strategic responses to COVID-19 pandemic (January-May 2020). J. Air Transp. Manag. 87, 101863. Andersen, R.I., Stowe, J.D., Xing, X., 2011. Does Corporate Diversification Reduce Firm Risk? Evidence from Diversifying Acquisitions. https://doi.org/10.2139/ssrn.1755654. Available at SSRN.

Black swans or gray rhinos on the runway? The role of uncertainty

243

Antipova, T., 2021. Coronavirus pandemic as black swan event. In: Antipova, T. (Ed.), Integrated Science in Digital Age 2020. ICIS, Lecture Notes in Networks and Systems, 136. Springer, Cham. https://doi.org/10.1007/978-3-030-49264-9_32. Bilotkach, V., Mueller, J., Németh, A., 2014. Estimating the consumer welfare effects of de-hubbing: the case of Malév Hungarian Airlines. Transport. Res. E Logist. Transp. Rev. 66, 51e65. Borenstein, S., 1992. The evolution of the U.S. Airline competition. J. Econ. Perspect. 6 (2), 45e73. Broekema, H., Crombach, l., Horst, P. van der, Marey, R., Pauwels, K., Vijlbrief, M., 2020. How Real Estate Can Help Airports Build Resilient Business. NACO White Paper. Burghouwt, G., 2005. Air Line Network Development in Europe and its Implications for Airport Planning. PhD thesis. Utrecht University, ISBN 90-393-3924-4. Burghouwt, G., Mendes de Leon, P., Wit, J.G. de, 2015. EU Air Transport Liberalisation, Process, Impacts and Future Considerations. Discussion Paper No. 2015-04, ITFOECD. www.internationaltransportforum.org/jtrc/DiscussionPapers/jtrcpapers.html. CAPA Centre for Aviation, 2020. Europe’s LCCs Navigate the Crisis Better than Legacy Airline Rivals. Center for Strategic and International Studies, 2019. Ending the Cycle of Crisis and Complacency in U.S. Global Health Security. https://csis-website-prod.s3.amazonaws.com/ s3fs-public/publication/191122_EndingTheCycle_GHSC_WEB_FULL_11.22.pdf. Cirillo, P., Taleb, N.N., 2020. Tail Risk of Contagious Diseases arXiv, 18 April. https:// arxiv.org/abs/2004.08658. De Neufville, R., Barber, J., 1991. Deregulation induced volatility of airport traffic. Transp. Plan. Technol. https://doi.org/10.1080/03081069108717476. De Neufville, R., 2003. Real options: dealing with uncertainty in systems planning and design. Integrated Assess. 4 (1), 26e34. Dutch Parliament, 1998. Growth figures Schiphol, report nr. 2, doc. 26265, 27-10-1998. (Kamerstuk 26265, Groeicijfers Schiphol, nr. 2 rapport, Tweede Kamer der StatenGeneraal, 27-10-1998.). Eurocontrol Statfor, 2020. Five-Year Forecast 2020-2024, European Flight Movements and Service Units, Three Scenarios for Recovery from COVID-19. Flyvbjerg, B., 2020. The law of regression to the tail: how to survive COVID-19, the climate crisis and other disasters. Environ. Sci. Pol. 114, 614e618. Fox, C.R., Ulkümen, G., 2011. Distinguishing two dimensions of uncertainty. In: Brun, W., et al. (Eds.), Perspectives on Thinking, Judging, and Decision Making. Universitetsforlaget, Oslo. Global Preparedness Monitoring Board, 2019. A World at Risk: Annual Report on Global Preparedness for Health Emergencies. World Health Organization, Geneva. License: CC BY-NC-SA 3.0 IGO. Greenspan, A., 2013. Never Saw it Coming. Why the Financial Crisis Took Economists by Surprise, Foreign Affairs. Gudmundsson, S.V., Cattaneo, M., Redondi, R., 2020. Forecasting Temporal World Recovery in Air Transport Markets in the Presence of Large Economic Shocks: The Case of Covid-19. https://www.researchgate.net/publication/342051795. IATA, 2018. Global Financial Crisis Cost the Industry $280 Bn in Lost Revenues. IATA Economics’ Chart of the Week. ICAO, 1987. Airport Planning Manual, Part 1, Master Planning, 2d ed. ICAO, Montreal. Doc. 9184-AN/902. Kasarda, J.D., 2019. Aerotropolis. Published online by John Wiley & Sons, Ltd. https:// doi.org/10.1002/9781118568446.eurs0436. Kincaid, I., Tretheway, M., Gros, S., Lewis, D., 2012. Addressing Uncertainty about Future Airport Activity Levels in Airport Decision Making. ACRP Report 76. http://www. trb.org/Publications/Blurbs/167934.aspx. Knight, F.H., 1921. Risk, Uncertainty and Profit. Boston MA: Hart, Schaffner and Marx. Houghton Mifflin Co.

244

The Air Transportation Industry

Linden, E., 2021. Pandemics and environmental shocks: what aviation managers should learn from COVID-19 for long-term planning. J. Air Transp. Manag. 90, 1001944. https://doi.org/10.1016/j.jairtraman.2020.101944. Lukaitis, S., Cybulski, J., 2004. The Denver International Airport Baggage Handling System. Papers from the Information Systems Foundations: Constructing and Criticizing Workshop at the Australian National University 16e17 July 2004. Lempert, R., Popper, S., Bankes, S., 2002. Confronting surprise. Soc. Sci. Comput. Rev. 20, 420e440. http://ssc.sagepub.com/cgi/content/abstract/20/4/420. Lewis, F., 2010. The Big Short. Inside the Doomsday Machine. W.W. Norton & Co, New York. ISBN 0-393-07223-1. Maack, J.N., 2001. Scenario analysis, a tool for task managers. In: Krueger, R.A., et al. (Eds.), (2001) Social Analysis, Selected Tools and Techniques, Social Development Papers, Paper Nr. 36. The World Bank, Washington DC. Malighetti, P., Paleari, S., Redondi, R., 2016. Base abandonments by low-cost carriers. J. Air Transp. Manag. 55, 234e244. Decision making under deep uncertainty. In: Marchau, V.A.W.J., Walker, W.E., Bloemen, P.J.T.M., Popper, S.W. (Eds.), 2019. From Theory to Practice. https:// doi.org/10.1007/978-3-030-05252-2. Markides, C.C., 1997. To diversify or not to diversify. Harv. Bus. Rev. 75 (6). Mill, J.S., 1835. A System of Logic. University Press of the Pacific, Honolulu, 2002, ISBN 1-4102-0252-6. Mishra, P.K., 2020. COVID-19, Black Swan event and the future of disaster risk management in India. Prog. Disaster Sci. 8, 100137. https://doi.org/10.1016/j.pdisas.2020.100137. National Commission on Terrorist Attacks Upon the United States, 2004. The 9/11 Commission Report. Government Printing Office, Washington DC. Paté-Cornell, E., 2012. On “black swans” and “perfect storms”: risk analysis and management. When statistics are not enough. Risk Anal. 32 (11), 1823e1833. https://doi.org/ 10.1111/j.1539-6924.2011.01787.x. Pearce, B., 24 November 2020. Outlook for Air Transport and the Airline Industry. IATA Annual General Meeting. Peneda, M.J.A., Reis, V.D., Macário, M.R., 2011. Critical factors for development of airport cities. Transportation research record. J. TRB. https://doi.org/10.3141/2214-01. Popper, K., 1934. Logik der Forschung. Abingdon: Routledge revised edition 2005. ISBN. 1-1344-7002-9. Redondi, R., Malighetti, P., Paleari, S., 2012. De-hubbing of airports and their recovery patterns. J. Air Transp. Manag. 18, 1e4. Ricondo & associates, Booz Allen Hamilton, George Mason University, National Service research, 2009. Strategic Planning in the Airport Industry. ACRP report 20, Washington DC: TRB. Sehl, K., 2020. How the Airline Industry Survived SARS, 9/11, the Global Recession and More. https://apex.aero/articles/aftershocks-coronavirus-impact/. SOIF (School of International Futures), IATA, 2018. Future of the Airline Industry 2035. www.iata.org/iata-future-airline-industry-pdf. Taleb, N.N., 2001. Fooled by Randomness, the Hidden Role of Chance in Life and in the Markets. Penguin Books Ltd, London, ISBN 978-0-141-03148-4. Taleb, N.N., 2007. The Black Swan: The Impact of the Highly Improbable. Penguin Books Ltd, London, ISBN 978-0-1410-3459-1. Végh, C.A., Vuletin, G., Riera-Crichton, D., Pablo Medina, J., Friedheim, D., Morano, L., Venturi, L., 2018. “From Known Unknowns to Black Swans: How to Manage Risk in Latin America and the Caribbean,” LAC Semiannual Report (October). The World Bank, Washington, DC. https://doi.org/10.1596/978-1-4648-1373-3. License: Creative Commons Attribution CC BY 3.0 IGO.

Black swans or gray rhinos on the runway? The role of uncertainty

245

Walker, W.E., et al., 2003. Defining uncertainty: a conceptual basis for uncertainty management in model- based decision support. Integrated Assess. 4 (1), 5e17. Walker, W.E., Marchau, V.A.W.J., Kwakkel, J.H., 2019. Dynamic adaptive planning (DAP). In: Marchau, V.A.W.J., Walker, W.E., Bloemen, P.J.T.M., Popper, S.W. (Eds.), Decision Making under Deep Uncertainty. From Theory to Practice. Springer, Cham. https://doi.org/10.1007/978-3-030-05252-2. Wucker, M., 2016. The Gray Rhino: How to Recognize and Act on the Obvious Dangers we Ignore. St. Martin’s Press e-book, New York. Zuidberg, J., Wit, J.G. de, 2020. The development of long-haul low-cost networks in the North Atlantic airline market: an exploratory data approach. Transp. Pol. 95, 103e113.

This page intentionally left blank

CHAPTER 11

Making sense of airport security in small and medium-sized airports Duarte Cunha

CERIS, Instituto Superior Técnico - Universidade de Lisboa, Lisbon, Portugal

1. Introduction Regular air transport services are a critical for remote areas, connecting them to each other and to larger population centers, and, consequently, to certain needed services (such as healthcare and education, for example), providing socioeconomic benefit (Halpern and Bråthen, 2011). An effective air transportation system is linked to the accessibility and quality of life of remote communities and to their economy (Baker and Donnet, 2012). Regional airports are vital for economic growth of Europe’s regional communities. The worldwide connection and speed of air travel gives remote regions more accessibility than other means of transport being an enabler for social development and economic growth. In Norway, for example, a study concluded that residents in remote regions have a higher frequency of travel by air on domestic services than the national average (Halpern and Bråthen, 2011). Regardless of catchment area size, the airports efficiency plays a main role on the airport’s success (Merkert and Mangia, 2014). In Norway, smaller airports are losing market share due to better land accessibility to larger airports, as passengers (especially in the leisure segment) choose to spend longer times to the airport, to benefit from lower fares and more convenient airline services provided in larger airports, even though smaller airports are covered by Public Service Obligations that set maximum fares, but, nevertheless, are still high and provide indirect services (Lian and Rønnevik, 2011). Regulation has imposed additional challenges by weighing heavily on the operation of small and medium-sized airports (SMSA). Regarding safety and security, all airports are required to have a minimum level of infrastructure and operational capability, and maintaining these requirements The Air Transportation Industry ISBN 978-0-323-91522-9 https://doi.org/10.1016/B978-0-323-91522-9.00011-7

© 2022 Elsevier Inc. All rights reserved.

247

248

The Air Transportation Industry

comes at a high cost, which is not directly equivalent to the volume of traffic of each airport. SMSA have high fixed costs imposed by regulations, which may be justified at larger airports, and cannot be reduced as function of the lower traffic presented at these airports. Regarding safety, airports must maintain fire cover for commercial operations, and the fire trucks used must be the same specifications and standards in all airports, small or major, representing a very high capital expenditure for the smaller airport. As for security, the same security processes and regulations are applied in all airports, increasing the fixed costs in terms of equipment and manpower. Additionally, SMSA are subject to rigorous audits, both on safety and security, that are arguably not appropriate in their demands given the small volumes of traffic. Besides verifying the airport operational capability, they also require significant manpower to work with the auditors and to complete paperwork and documentation of the operation, consuming precious management time that could be applied in more fruitful actions (European Parliament, 2015). A risk-based security is deemed much more cost-effective than the current paradigm. An efficient allocation of resources is needed to improve cost-efficiency. A risk-based assessment can help as it will identify threats and vulnerabilities and, with this, where the resources need to be spent. Hence, our research intends to develop a new approach to aid in the implementation of risk-based security in SMSA.

2. A brief history of air transportation security Over the years, air transportation security has evolved. This evolution is linked with the evolution of aviation terrorism, for which three distinct phases have been identified (European Commission, Irish Aviation Authority, and Avia Solutions, 2004). Phase 1 (1948/1968) was characterized by air piracy or hijacking of aircraft by individuals who, to avoid persecution or prosecution, escaped from a State and considered this strategy as a fast and convenient way to achieve their aim. Since this period, hijacking of aircraft for these purposes has reduced, except for 2003, when hijackings were reported where individuals were attempting to reach Taiwan from China (Unitn et al., n.d.). Phase 2 (1968/1994): In 1968, the world saw the beginning of “modern terrorism” and the link between politics and terrorist acts against civil aviation. The hijacking was still among the most popular tactics along with bombings. Terrorist organizations used these tactics as a way of calling

Making sense of airport security in small and medium-sized airports

249

attention to their cause and to disconcert or humiliate governments or other terrorist organizations who opposed them, damage the economy of a targeted state, demand the release of imprisoned fellows and/or to demand money, or to provoke fear in the population. The current phase of Civil Aviation Terrorism (Phase 3) started in 1994 with the Algerian terrorists’ hijack of an Air France flight. The French Government, having received intelligence that the terrorists intended to blow the aircraft over Paris, diverted the flight to Marseille, where a successful rescue operation was executed. This was the first event where the aircraft was intended to be used as a weapon. With the events of September 11, 2001, an immediate and drastic heightening of air transportation security occurred across the world (Unitn et al., n.d.). The international community has undertaken a variety of security measures to prevent and mitigate unlawful acts against civil aviation. The most notable success was emerging of an international consensus that aviation security issues are of international concern, as air transport itself, and security insufficiencies in any part of the world can affect the entire global air transport system. Consequently, no country in the world dares to provide opportunities for the perpetrators under the current international political environment. Aviation security has been a concern of the International Community since the inception of air travel in the early 1900s. As air transport developed technically and international travel started in the period following World War I, governments realized that air transport needed to be regulated on an international level. In 1919, at the Paris Peace Conference, the first International Air Convention was signed, and the International Commission for Air Navigation was established. The Convention on International Civil Aviation in 1944, known as the Chicago Convention, even though not only security-related, generally regulates civil aviation from passenger safety to technical aspects of flight. It was then that the International Civil Aviation Organization was established, being formally instituted in 1947. Later, in 1963, in the Convention on Offences and Certain Other Acts Committed on Board Aircraft, known as the Tokyo Convention, the concern on establishing a jurisdiction over offences committed on board an aircraft appeared. With the large increase in aircraft hijackings in 1969 and 1970, the international community became aware of a need to have legislation to prevent hijacking. The Hague Convention in 1970 (Convention for the Suppression of Unlawful Seizure of Aircraft), had this objective, by establishing hijacking as a crime

250

The Air Transportation Industry

and enforcing parties to legally set it as punishable and with severe penalties. The 1971 Montreal Convention (Convention for the Suppression of Unlawful Acts against the Safety of Civil Aviation), following a model similar to the Hague Convention, dealt with the sabotage of aircraft and was supplemented by the Montreal Protocol for the Suppression of Unlawful Acts of Violence at Airports Serving Civil Aviation in 1988, which tuned illegal or offensive to carry out armed attacks, to cause damage or disruption at civil aviation airports, also instating penalties for such offences. It was during the previously described as phase 2 that three significant attacks on civil aviation occurred that later defined some security regulations as a reaction. In 1985, Lebanese Terrorists hijacked a TWA flight diverting it to Beirut. The event lasted for two weeks with 1 of the 156 hostages being killed. This originated several US actions, including the International Security and Development Cooperation Act of 1985. It was also in 1985 that ground staff allowed a bag with no passenger on the airplane to be checked through in an Air India Flight, in which an Explosive Device was concealed, resulting in one of the worst aviation bombings with the loss of 329 people. It was also in 1988, after the Lockerbie Incident in which the same tactics as the Air India flight were used, that improvements in aviation security were implemented all over the world, with the aviation security plan adopted by ICAO, which are considered as the base for the current aviation security structure. Measures such as screening of hold baggage, cargo and mail, baggage reconciliation, control access in sensitive areas of airports and the reinforcement of powers and organization of ICAO, to enable it to implement more actively the security and safety standards. In the aftermath of the 9/11 terrorist attacks, in the ICAO Assembly in 2001, the ICAO Council was directed to develop an action plan addressing the threat to civil aviation and to perform a review of the international conventions and Annex 17 of the Chicago Convention. In 2002, the ICAO implemented a comprehensive strategy reinforcing aviation security throughout the world. The key in this strategy was the ICAO Aviation Security Plan of Action that included regular, mandatory, systematic and harmonized audits to evaluate implemented aviation security in all Member States. It also included the identification, analysis and development of an effective global response to emerging threats. The improvements to aviation security have been historically reactive, responding to a crisis as it occurs. The adoption of recommendations from

Making sense of airport security in small and medium-sized airports

251

the various international bodies by States has resulted in an approach that is disjointed and incremental, instead of creating a coherent global standardized system that is essential to fight the growing terrorism threat. Despite the extensive focus on aviation security in past years, Elias et al., (2016) identifies some challenges that remain: To increase the efficiency of detection of explosive threats, development and implementation of risk based methods for screening passengers and others who access sensitive areas. Identify potential threats thru use of existing watch lists and available intelligence; Development and implementation of strategies and defenses for aircraft against shoulder-fired missiles or other standoff weapons; Address the potential security implications of the use of unmanned aerial vehicles or remote piloted aircraft (UAV).

3. Regulatory framework The world has changed and will change at an accelerated pace, with the society, economics, international relationships and even weather patterns becoming increasingly complex over the years. The complexities of globalization, public expectation, regulatory requirements, transnational issues, multijurisdictional risks, crime, terrorism, advances in information technology, cyber-attacks, and pandemics have created a security risk environment that has never been more challenging. Security has been continuously developed as a discipline, yet no single framework pulls together all the excellent but disparate work that practitioners and researchers are continually working on. Although there is little dispute that risk is a factor that must be considered by decision-makers when deciding what, if anything, should be done about a risk that falls within their responsibility, in Security there has been less than total agreement as to what this means in practical terms (Shim et al., 2014). Before entering the Aviation Security domain, one needs to understand what security is. Kölle et al. (2011) start by examining the dictionary definition of it and define three different concepts for security. The authors point out that security is expressed as a state, is described as a process, and is a process. Security is the condition of being protected against danger or loss (Talbot and Woodward, 2009). Such protection is done by the mitigation of negative consequences from the intentional and unwarranted action of others. The authors explain that the unwarranted actions are those that are illegal and unacceptable, at least from the defender’s perspective, while the

252

The Air Transportation Industry

intentional adjective clearly separates security and safety. While security is the protection from deliberate acts, safety involves the protection from unintended events such as accidents or falls, for example. Other definition of security states that security provides “a form of protection where a separation is created between the assets and the threats” (Institute for Security and Open Technologies). Most of the existing literature on risk was developed for issues that do not have the same complexity, uncertainty, or ambiguity as modern security has. For example, managing financial or operational risks can be quantified more easily than most of the concept security managers handle. Nevertheless, there are insights in the tools and techniques in areas such as safety, project methodologies, engineering, operations research, and information technologies that may offer security new ways to handle risk. The United Kingdom defines three objectives for national security: protect the people in all territories, the territories, economy, infrastructure, and way of life; protect the nation’s global influence, reducing the likelihood of threats to the nation’s interests and those of its allies; and promote the nation’s prosperity, seizing opportunities, innovating and supporting the industry (British Government, 2018). The ICAO Annex 17 defines aviation security as “Safeguarding civil aviation against acts of unlawful interference. This objective is achieved by a combination of measures and human and material resources.” It also considers unlawful interference to be “acts or attempted acts such as to jeopardize the safety of civil aviation (.).” Safety and security both intend to protect aviation from losses. While security protects aviation from a foe, safety protects it from errors. What both have in common, is that, when the defenses are insufficient or fail, losses can be sustained. While the foe in security is an attacker, frequently a terrorist, in Safety the enemy seeking for breaches in the defenses is mischance, casually known as Murphy’s law. In a short description, ICAO sets the minimum standards that every State Member needs to fulfill to have flights in its territory, which means that a State Member needs to draw up and implement a civil aviation structure that complies with the minimum standards. State Members can also create a different organization to implement security standards. For example, the FAA in the USA or the EASA in the EU. These organizations can even go beyond the minimums set by ICAO. Fig. 11.1 describes the level of detail considering the amount of information given at each level and who they are addressed to, that

Making sense of airport security in small and medium-sized airports

253

Figure 11.1 Regulative scheme.

increases from the ICAO, as the legislator, to the airport stakeholders, for example, the airport operators themselves, airlines, ground handlers, etc. On the other way, the “lower ranks” must comply with the ranks above them. 3.1 ICAO In Annex 17 of the Chicago Convention, ICAO identifies the security objectives that each Member State shall: • Have as its primary objective the safety of passengers, crew, ground personnel and the public in all matters related to safeguarding civil aviation against unlawful interference. • Establish an organization to develop and implement regulations, practices, and procedures to safeguard civil aviation against unlawful interference, while considering the safety, regularity and efficiency of flights. • Ensure that the above-mentioned organization, regulations, practices, and procedures protect the safety of passengers, crew, ground personnel and the general public in all matters related to safeguarding against acts of unlawful interference in civil aviation and that they are capable of rapidly respond to any increased security threat. • Ensure that the appropriate authority arranges for the supporting resources and facilities required by the civil aviation security services to be available at each airport serving civil aviation. • Ensure that persons other than passengers, and carried items, who have access to security restricted areas are screened. If the principle of 100%

254

The Air Transportation Industry

screening cannot be accomplished, other security controls shall be applied following a risk assessment carried out by the respective national authorities. Until the 9/11 events, the ICAO model was regarded as an adequate and sufficient structure to ensure the security of civil aviation. However, the changes implemented after the 9/11 drastically imposed stricter requirements and are the standing currently, such as: • Aircraft Security CheckdRequirement to inspect the interior of an aircraft to which passengers may have had access, and an inspection of the hold for suspicious objects, weapons, and other dangerous devices. • Background checkdAny individual requiring unescorted access to a security restricted area shall be subject to an identity and previous experience check, including criminal history. • ScreeningdThe addition of the term “identify and/or” to the requirement to “detect weapons, explosives, or other dangerous devices which may be used to commit an act of unlawful interference” places an additional responsibility on screening operators on, not only to find, but also identify dangerous objects. It also added the requirement for additional equipment for effective detection and identification. • SecuritydBy changing “international civil aviation” to just “civil aviation,” it implied that security procedures should be extended to domestic air transportation, also changing the objectives to ensure that the principles governing measures designed to prevent acts of unlawful interference are also applied to domestic operations as far as is feasible. This means that is up to the States responsible authorities to analyze if the application of additional security measures in domestic air transport is practicable. • Security Restricted AreasdExtension of the security restricted areas to areas outside the actual aircraft and apron areas, covering a much wider physical area. This turned out to be one of the changes responsible for the increase in needs for security services and manpower required. • International cooperationdEach Contracting State must share with other Contracting States any threat information that applies to the security interests of those States. • National organization and appropriate authoritydAs stated earlier, each State is required to create and empower the appropriate authority to manage a national civil aviation security program by defining and allocating tasks and coordinating activities, not only between the departments, agencies and other State organizations, but also with airport

Making sense of airport security in small and medium-sized airports

255

and aircraft operators, and other stakeholders in Aviation Security. The coordinating function is to be assumed by a national aviation security committee selected for that purpose. • Security ProgramdThe changes also introduced the requirement for airports and airlines to establish and implement a written security program conforming with the requirements of the national aviation security program. Although these changes may seem simply changes in terminology, they introduce a significant new prominence. In an overall view, these changes represent a distinct hardening of airport controls and security measures, with all the repercussions associated (Minnaar, 2003). Annex 17 and Doc 8973 of ICAO are constantly being reviewed and amended, not only considering new threats, but also under new technological developments that may have influence on the effectiveness of measures designed to prevent acts of unlawful interference in civil aviation. 3.2 European Framework The aviation security framework/legislation in Europe has changed over the years with the EC issuing several laws. At present time, the European top framework on Aviation Security is reflected on Regulation (EC) No 1998/2015, which establishes the common rules in the field of civil aviation security and is complemented by Commission Regulation (EC) No 185/2010, which goes deep into details, regarding the implementation of the rules. The main objectives of the CE regulations are to establish and implement appropriate Community measures, in order to prevent acts of unlawful interference against civil aviation, and to provide a basis for common interpretation of the related provisions of the Chicago Convention, and in particular, of the Annex17. To achieve these objectives, the CE set up a common basic standard for civil aviation security measures and a fitting compliance monitoring system. These rules cover several topics: • Airport security • Demarcated areas of airports • Aircraft security • Passengers and cabin baggage • Hold baggage • Cargo and mail

256

• • • • • •

The Air Transportation Industry

Air carrier mail and air carrier materials In-flight and airport supplies In-flight security measures Staff recruitment and training Security equipment Classification of weapons and prohibited items Therefore, complementing the ICAO regulation, the European Regulation mandates in rule 13 of Regulation (EC) No 300/2008 that every state should implement a National Security Program (NSP), in which are defined the general rules for airport operators, airlines, etc. to abide in terms of airport and on-board security, screening of passengers, luggage, cargo and mail, and airport and on-board supplies, and also in terms of employee and personnel qualification and recruitment. Due to the need to monitor the implementation of rules defined in civil aviation security scope by every stakeholder, rule 14 of the same Regulation also states that every State Member must implement an Aviation Security Quality Control Program defining how the compliance of the methods and procedures in the NSP are to be monitored by the responsible authority. Such program shall notably fix the specifications as to security audits and inspections, including their frequency. It is also an obligation of the Member States to impose penalties in case of infringements, through the aviation security authority, and to cooperate with and provide assistance to the Commission while it conducts inspections to monitor the compliance with EU rules, as well as ensure that the notification of an inspection is kept confidential and to have qualified auditors available to participate in such inspections. The testing of new technologies in security measures which are not provided for by EU law is also permitted. These tests have to be reported to the Commission so that it is guaranteed that it will not affect the overall level of security and may not be implemented for more than 30 months. This was the basis for some Member States to evaluate the use of “body scanners” before they were introduced in the framework in late 2011. It is also the Member States option, having as basis a security risk assessment, to apply more rigorous rules in accordance with EU law. For example, there is no imposition for the deployment of inflight security officers, it is the sole decision of the Member State. The EU framework also imposes that all airports, air carriers, and other operators draw up and implement their own security programs, complying with the national civil aviation security program of the Member State in which they are located and with EU law.

Making sense of airport security in small and medium-sized airports

257

The airport security programs must be submitted for approval by the national authority responsible for civil aviation security, while the air carriers or other operators’ security programs have to be submitted upon request. Furthermore, when an airline’s security program has been approved and validated by a Member State which also approved the operating license, the latter is considered compliant with the requirements of national security programs throughout the EU.

4. Air transport security costs The current EU legislative framework lets Member States decide how air transport security costs are covered. A report on financing aviation security required by the Regulation (CE) No 300/2008 (COM 30/2009) proposed a directive on air transport security charges to ensure the application of key principles such as nondiscrimination between airlines or passengers and cost-relatedness. However, this proposal did not decide between public financing and the user-pays principle, turning to subsidiarity the responsibility to arrange who covers the security expenses. Security taxes are charged to passengers and cargo differently. The amount passengers pay is divided into two components fixed by the regulator and may be different according to the destination of the passenger. One is the counterpart of the general costs and specific security costs of the regulator, security forces, and border control services. The second component is counterpart for the airport authority/manager for the specific security procedures costs and for installation. Although security costs are partially on either to the taxpayer or to the passenger, they have a significant effect on the demand and supply of air transportation services. Some defend that it should not only be the passengers to be charged with security taxes. There are other airport users that require security checks, such as the airport employees themselves, and airline ground staff and even crews, ground handler staff and other people involved in airport activities. Airport security taxes also impact the demand, with airlines recognizing that the rising security taxes have a negative impact on demand (Prentice, 2015). Another argument is that security is a public good, and, nowadays, the cost of air security is not being taken by society in general, as it happens in other transport modes, but the benefit of the whole society is being supported by the passengers (Prentice, 2015).

258

The Air Transportation Industry

5. Proportionality of security in airports Looking at the aviation security framework, there is a set of standards and rules that do not differentiate different situations. Although it has a wide range covering all the operations and off operations of air transportation, it does not consider different scenarios for the same sector. Several critics to the current paradigm of aviation security can be found in literature, as well as possibilities to overcome the identified problems. There is an ongoing discussion on risk-based security and how it can enhance the current aviation security paradigm. As Safety Management Systems (SMS), Security Management Systems (SeMS) is a growing trend in aviation security due to the concern on enhancing efficiency. In the following paragraphs we will explore this. 5.1 Criticisms to aviation security framework Aviation is a critical part of local and national economies (Szyliowicz, 2004). Stricter security requirements and screening procedures and equipment require larger areas in an already constrained infrastructure, which, in its turn, demands for increased throughput strategies (CSES UK, 2011). Until now, the evolution of Aviation Security approach has been a target hardening approach, but it is not conceivable to fully harden all targets, as the resources are not unlimited (Poole, 2008). Hence, the challenge in dealing with security threats is deciding where to invest scarce resources to maximize the benefit (Poole, 2009). The same author argues that a cost-effective security must be risk based, and that current policy is only risk-based in name. Another negative aspect of target hardening is that Intelligent adversaries can monitor defenses and learn how to counter-react them (Tambe, 2011). Over the years, there has been cases where perpetrators tried to infiltrate prohibited items into Restricted Areas using dissimulation techniques, which decreases the efficiency of security procedures and equipment. The aviation security framework has been developed based on strategies of anticipation, focusing on preventing the entry of prohibited items and protecting targets based on known threats and problems detected from previous security incidents. This approach is somewhat identical to what happens in Safety. Nevertheless, due to the security framework, aviation organizations do not have a degree of functional authority and flexibility that is present in Safety. The Aviation Security Regulations impose standardized and detailed routine ways of achieving security, that do not allow

Making sense of airport security in small and medium-sized airports

259

adaptations (Pettersen and Bjørnskau, 2015). The same authors consider this a paradox, due to the uncertainty that security must handle in terms of terrorism threats against civil aviation, some flexibility would be desirable. Stewart and Mueller (2014) assessed the risks and cost-effectiveness of measures designed to further protect airports and associated facilities, considering the threat likelihood, cost of security measures, hazard likelihood, risk reduction and expected losses to realize a cost-benefit analysis comparing the optimal security measures in airports. The authors concluded that attack probabilities are low and do not justify additional security measures, and even imply that the current security impositions may be in excess considering an airport specific risk management. But this may contradict the idea of a one stop security screening for passengers, cargo and mail in Europe, since the general idea is that a passenger or cargo can go from airport A to airport C, passing thru airport B without losing time with another security check, which would be mandatory, if the security check in C was not the required for airport B. Nevertheless, this also brings up another question: who is responsible for the security costs at airport A, if this airport has a higher security restriction that is necessary, considering its own risk exposure. The EU regulation allows for security derogation in airports but with some very restrictive criteria for commercial airports. Limiting the MTOW to 15,000 kg strikes out most of low movement regional airports from being able to derogate security measures or apply alternative ones which may turn out to be as much efficient but with much lower costs. For example, Flores Airport in Flores Island, also in the Azores Archipelago, with up to 44,000 passenger movements (enplaned and deplaned) in 2014, is subject to the full extent of security standards in the EU framework. Since the EU leaves the decision between screening methods open to those performing the security controls or Member States, a lack of harmonization between screening procedures is present in the EU. Many countries detail the practice at national level, but not in the same way. As a result, some screening methods are not admissible in every country, raising the possibility of undermining certain aspects of internal movement. Regarding in detail the cargo and mail air transport, EU air cargo security regulations give security operators the choice of which of the above presented pathways to apply. SCS implies a more structured effort in terms of investment in designing facilities, training, and certification, but its procedures are less arduous, and it is more cost-efficient for organizations that handle large amounts of cargo and mail. As a result, Screening is the

260

The Air Transportation Industry

most common method to ensure the Security in Air Cargo and Mail, especially in airports with lower volumes of traffic (Amorim da Cunha et al., 2017). Nevertheless, in some circumstances, there is no option regarding what security procedures are applied. For example, in flights to the United States, all cargo and mail must be screened, regardless of the application of SCS controls due to USA own Air Cargo and Mail Security Framework. Also, some aircraft operators impose the screening of all cargo and mail, even for intra-European flights. Aviation Security is not different from police, military or border control activities. It is part of the Security System of a nation or group of nations in this matter, and everyone benefits from it, and no one should be excluded from it. Aviation Security is then a public good. Economic theory states that public goods cannot be efficiently delivered by free market forces due to the free-rider problem and, therefore, public goods are either financed from the general treasury, or under produced (Prentice, 2015). 5.2 Risk based security Poole (2008) argues that cost-effective security must be risk-based and that current policy is only risk-based in name. An equal-risk assumption is seen in the current European aviation security framework. All airports, disregarding their characteristics, must implement a certain set of Security measures and procedures. Risk-Based Security would allocate security resources in proportion to the posed threat/risk. This is the policy for Critical Infrastructure protection. For example, air, road, rail, and sea cargo do not have 100% physical inspection, either road, train or sea passenger transportation. As an alternative, various procedures are in place to secure these transportation modes (Poole, 2015). Nevertheless, taking a risk-based Security approach is often seen as a slackening in Security (Wong and Brooks, 2015). One example is the exception of small pocketknives allowed onto restricted areas and ultimately to the aircraft which may cause a negative impact in Public Perception, stating that the threat of using them as a weapon would increase. However, taking into consideration other implemented measures, such as the reinforced cockpit doors, the risk posed by these objects is still reduced. Another stated reason against implementing a risk-based model is discrimination, as different procedures may be seen as against the principle that a government must treat all citizens equally (Poole and Passantino, 2003).

Making sense of airport security in small and medium-sized airports

261

The idea that underlies Risk-Based Decision making is to rank all the risks and proportionally apply measures to mitigate each risk level. This means that high-level risks will have extensive scrutiny, medium-level will also be scrutinized, but at a lower intensity, but most importantly, the low-level risks will intentionally receive the least attention (Poole, 2003), which may free resources to deal better with high-level and medium-level risks. Applying this paradigm to preventing acts of unlawful interference on air transportation will require reducing the focus on finding bad objects and increasing the focus on identifying people most worthy of additional scrutiny. Meaning that there is a need to develop a system that identifies the risk category in which people (passengers and others) and items (including baggage) for differential processing (Domingues et al., 2014). This is the fundamental idea behind the TSA Pre-Check/Fast Track Security implemented in the USA. Passengers who sign up for this program must comply with specified pre-requisites (for example, US citizens or Green Card Holders with clean criminal and immigration records), and provide information to the TSA, and are subject to some form of scrutiny by the authorities before being accepted. With this, the TSA classifies passengers accepted in the Pre-Check as low-risk and, therefore, these may be subject to lower scrutiny before boarding an aircraft, increasing the processing speed, comfort and, therefore, reducing costs at the airport security. Compared to other countries, the TSA screening workforce is more expensive. Considering worst assumptions, the program can slightly reduce risk (0.3%), while in a common case can reduce up to 1%. By allowing a faster screening process, it increases the efficiency of the screening checkpoint, while providing a better passenger experience, which can result in a benefit of over billions of dollars per year (Stewart and Mueller, 2018). Other similar programs include the CAPPS (Computer-Aided Passenger Pre-screening System) that was already in place before the 9/11 attacks that divide passengers into selectees and non-selectees thru a risk assessment. This is a pre-screening system that assigns an assessed threat to a passenger considering certain characteristics and requires additional screening for selectees (the higher risk passenger). In a way, doe limited due to the perception of threat at the time, it was close to a risk-based model. This program was even upgraded after the 9/11 attacks to the CAPPS II, but was later dismantled due to privacy concerns.

262

The Air Transportation Industry

The most rigorous airport security is in Israel (Poole, 2007), and it is a good example of a risk-based security model. Comparing to the European and American model, in which the primary target are dangerous items, it aims at potentially dangerous people instead, hence, changing the focus as stated earlier by sorting passengers into groups, according to a risk assessment, who will receive different airport processing based on the calculated risk level. Another example is the Supply Chain Security (SCS) for air cargo transportation in the European Framework, in which regulated entities, such as Known Consignors (KC), Account Consignor (AC) and Regulated Agents (RA) must have a Security Programme implemented and guarantee that the cargo is protected from unauthorized interference until it is loaded onto an aircraft. In a way, instead of examining thoroughly the agents, these elements must undertake specific actions in the selection and training of personnel and protection of cargo and mail that limits the possibilities of acts of unlawful interference to the consignment and are subject to a Quality Control Audits and Inspections to guarantee the compliance and, therefore, the risk reduction. 5.3 SeMSdsecurity management system To ensure the effective application of preventive security measures, the level of threat should be continually reviewed, considering the international, national and regional situation. Security measures and procedures should be flexible and proportionate with the threat assessment which may differ given various changing factors. A security management system (SeMS) provides an airport with a structured approach to managing security as an integral part of its overall business. SeMS serves as a tool for systematically integrating security risk management into the day-to-day operations in close alignment with other risk management systems. SeMS also provides organizations with a disciplined, consistent and risk-driven method for identifying and closing critical security gaps and provides the means to implement the best practices in security and demonstrate an airports’ commitment to security through an emphasis on accountability and due diligence. This approach assists airports to integrate and manage security risks holistically and systematically, with better results at less cost for government and the private sector over time. SeMS serves to enhance security in areas of higher risk and priority, and therefore its complexity and costs should be determined by the size of the

Making sense of airport security in small and medium-sized airports

263

operation and the results of threat, risk and vulnerability assessments. It is more than just a quality control framework as it assesses the efficiency of the implemented procedures but at the same time can detect new Security related needs.

6. A new approach for security in a network of airports The security framework decisions are done by the regulator at a strategical level. The implementation of the security measures in airports is done at a tactical and operational levels by the airport managers, but it is strongly limited by the framework. Risk-based security can improve the efficacy of security in a network of airports in which SMSA are included. It is possible to reach the security levels established in the regulatory framework, but with greater efficiency in the related costs management, by using a different approach to security. Instead of the top-down approach currently in practice that imposes the same standards to all airports, a risk-based approach that considers each airport and its specificities integrated into an airport network allows a reorganization of processes and procedures that improves the efficiency of the security system, both at a local level and system-wide. It is then possible to identify semi-isolated sub-networks or clusters of SMSA that are connected to a small number of larger airports in which the existing threat, and, consequently, the current security risk is lower than in larger airports. Using risk assessment techniques, it is possible to determine alternative or compensatory security measures that can have little effect on the security risk for SMSA but largely improve efficiency. This means that the resulting security system will have different security levels throughout the airport network, as depicted in Fig. 11.2.

Figure 11.2 Multi-level security airport network.

264

The Air Transportation Industry

At some point of the risk reduction, to further decrease it, we need to invest much more resources (Fig. 11.3). Either this is resolved by acquiring new processes and technologies that are more efficient, or we consider that a certain level of risk is acceptable considering the available resources. To say that additional security is “worth it” is not accurate. If all considerations were put aside, security would be enhanced to the extent that every passenger would have to answer myriad questions, strip naked and be body searched, and then pass-through multiple scanners (Gerstenfeld and Berger, 2011). This is the point where marginal cost of security is extremely high, meaning that the risk reduction is much lower than the investment one must make to achieve it (blue [dark gray in printed version] arrow Fig. 11.3). It is also possible to optimize security by considering that a reduction of costs can produce a much smaller increase in risk (inverse of blue [dark gray in printed version] arrow Fig. 11.3). Having this into consideration, a hierarchical security structure in an airport network, with different airport security levels, resulting from a riskbased approach can reduce costs, and increase efficiency. An agent-based simulation model is under development to demonstrate the advantages of the new approach to security in a network of airports. The model works as a proxy of the real world to determine the security risk and the cost of the implemented security. A series of tests under different scenarios can be run to collect evidence in favor or against the hypothesis. Risk-based security is characterized by three fundamental variables: risk, cost, efficiency. The model’s assumption and architecture are projected taking into consideration the hypothesis and the case study. The case study to be used is a semi-closed air transport network consisting of airports and flight routes connecting these airports, and some of these airports to the exterior of the considered network.

Figure 11.3 Risk variation versus cost variation.

Making sense of airport security in small and medium-sized airports

265

An agent-base structure was chosen for the model as it enables a set of properties to recreate real-work conditions, namely, the model can • represent an air transport network, consisting of airports and flight routes; • simulate the movements from airport to airport in the network and to the outside of the network; • represent each airport in the network and its main areas; • simulate the security procedures and equipment in each airport; • simulate the decisions and actions of an attacker and the defender; • provide the performance indicators, namely risk, cost and efficiency of the simulated scenario which will be used in the analysis tool pack. The model also includes some elements of Discrete Events simulation and an Analysis tool pack that processes the outputs of the simulation (Fig. 11.4). The agent-based simulation consists of the environment, adversaries, defenders and others, and the relations between them. The environment is the air transportation network in which all other agents are inserted. The structure of the model is built on three modules, presented as follows: The physical module of the model is comprised of the air transport network, the airports’ physical structure, including terminal area, check-in area, security screening area, departure area, baggage and cargo handling areas and curbside, and the flight paths that allow the connection between these airports. It is in these areas that the agents will interact with each other, making this the environment. The processes module represents the sequential events that the considered agents may perform while in the system, according to each type of agent, and is based on a Discrete Events modeling approach.

Figure 11.4 Projected simulation model.

266

The Air Transportation Industry

Finally, the Agents’ module feeds the processes module and the physical module. The adversaries will have a set of instructions to cause harm to the system, either by attacks, or by monitoring the defenders, or by influencing other agents or interacting with each other. The Defenders will interact with each other and with the Attackers to find unwanted behavior or forbidden items from adversaries or others. There are different types of defenders. A defender can be a security officer patrolling an area, a CCTV system, an X-ray machine or ETD machine, or a security officer in the screening checkpoint performing hand searches. Each type of defender can detect a certain threat with a pre-determined efficiency. To better simulate the intelligent behavior for both attackers and defenders, Game Theory elements will be included in the agent’s behavior, as, in this case, agents will seek maximization of their output. Adversarial risk analysis concept will be used to determine the agents’ actions. As they are treated as utility maximizers, one will need to depict the agents’ decision process and the costs and benefits of the available action. By others, we refer to passengers, crews and airport employees that may be “used” by the attackers. As for passengers, passengers can pass thru the airport in two channels: the departing passengers and the arriving passengers. Some passengers go thru the airport as transit passengers, meaning that they use the two channels (arriving plus departing). The departing passengers that board a flight in an airport are arriving passengers in the destination airport. Crews have the same behavior as passengers, and airport employees may be stationed in any area of the airport, meaning that they may have to go thru security controls to reach their designated area or not. The result of the implementation of the described layers then creates the simulation depicted in Fig. 11.5. An attacker or a threat enters the air transport network at a determined airport and goes thru flights and airports to reach its target. Defender agents are stationed in all airports and, according to the predetermined efficiency levels, they will successfully detect or not the attacker, not allowing it to reach its target.

Figure 11.5 Depiction of the simulation.

Making sense of airport security in small and medium-sized airports

267

The first phase of the attacker behavior (Fig. 11.6), planning of the attack, is modeled using Utility Theory, in which its objective will be to maximize the impact. The utility function considers, as variables, the target value, the attack type, the probability of detection and the penalty of detection. The target value and the attack type combined is the impact the attacker can make. For simulation purposes, the target value used will be the passenger movement value of the airport, as a proxy for the overall value of the airport and its catchment area which can be targeted by the attacker. The attack type will be a multiplier to the impact, as different types of attack can have different outcomes. But there is also a negative value on this Utility Function which is the Cost of Being Detected. This will be a function of the Probability of Detection perceived by the attacker and the resulting penalty. Since the attacker may not have perfect information on the security system efficiency, although it may have some information on the systems and procedures from the Defender point of view. From the Defender point of view, the simulation will have implemented different security equipment and procedures in the different areas of the airport. Although the largest security concentration of security is visible in the passenger screening points, the model will simulate security both in the landside and airside of the airport. This will be of great importance as the alternative measures will mostly be applied in the airside, especially for transit passengers, as we will see ahead.

Figure 11.6 Attacker behavior.

268

The Air Transportation Industry

The Analysis tool pack comprises Risk Assessment and Cost-Benefit Analysis. The Risk Assessment will rely on multiple runs of the simulation with different threat scenarios to determine the resulting Risk of the System. Together with the Risk Assessment, Cost-Benefit Analysis is used to measure the cost-effectiveness of security countermeasures at airports. It will be of extreme importance to use this to compare the current security framework with the hierarchical airport network security framework in terms of implementation and operational costs and the reduction in risk they produce (benefit). These will be the outputs of the simulation: risk and security costs. With this, it will be possible to compare the resulting security network with the base case, which will validate, or not, our hypothesis (Fig. 11.7). To validate our theory, we will need to assess the risks and costs associated with the current air transportation security framework and the ones associated with our proposed risk-based approach. A CBA will then be executed to determine if there is an advantage for the proposed network security against the current security framework (base). The implementation of security measures and procedures in the network airports will be determined by the simulation model, which will consider alternative security scenarios with different levels of efficacy and different costs than the ones demanded by the current framework. The base scenario, with the current standards of security, considers the same level of efficiency in all airports. This means that the same cost for each security procedure and equipment is present in all airports, but not necessarily the same total cost for each of them since different passenger or cargo movements may require a different amount of equipment. For example, considering passenger screening, the cost per passenger smaller

Figure 11.7 Scenario comparison.

Making sense of airport security in small and medium-sized airports

269

airports undertake is higher than larger airports, although larger airports have more X-ray machines. We shall consider an extreme example: Airport A has only one X-ray equipment to scan passenger hand luggage and has 35,000 boarding passengers per year while Airport B has four X-ray machines and 1,000,000 boarding passengers per year. Considering the passenger movements and if an X-ray machine has an initial cost of V240,000, it means Airport A had a cost of V6.52 per passenger and Airport B under V1 per passenger in one year, just considering the initial cost. Different other scenarios will consider different levels of efficiency of security procedures and equipment in all airports. As an example, considering the detection of IED in hand baggage, smaller airports will be set with lower efficiency security equipment and procedures at the passenger security and, therefore, lower costs, and the resulting risk will be determined to be compared with the base scenario. Since only reducing the efficiency may end up increasing the risk, alternatives will have to be tested. Using the same example, the efficiency of detection of IED in transit areas in larger airports will be increased. This way, if an attacker may choose a path that starts in a lower efficiency airport, trying to take advantage of the one-stop security concept, and reach a larger airport (his target) via an intermediate airport, it will be subject to additional security in the latter.

7. Conclusion Airports are part of the Critical Infrastructure of countries for the on-going health of the economy and for people’s lives and livelihoods and, therefore, are potential targets for serious crime and terrorism, not only due to the stated reasons, but also because airports are areas of great concentration of people. Since the 9/11 attacks, in 2001, Air Transport Security has evolved drastically in terms of procedures and equipment applied in response to the also evolving threat. This evolution has had its benefits but also its drawbacks, such as the imposition of higher costs and inefficiencies on SMSA. An optimized airport security network and the ability to adapt to new threats while guaranteeing the necessary security risk level, not only for passenger and airport operators but to society in general, can greatly increase the efficiency of the air transportation security system. The approach under development intends to aid in the implementation of risk-based security in SMSA. Security in these airports has not been studied deeply, so it can bring new insights into this matter. It is expected to

270

The Air Transportation Industry

show that SMSA can be subject to “lighter” security rules, without prejudice to their security level, but downsizing the associated costs. It will allow to identify and assess compensatory measures, such as procedures and equipment, more suitable to the specificities of SMSA while exploring the current regulatory framework. On a larger scale, the simulation developed may later be implemented at a larger scale, extrapolating to the whole air transportation network, where all airports can have different characteristics from each other. Terrorist acts are intentional events. The probability of such events is very hard to quantify, even more, when considering extreme attacks, with very low probabilities but with major impacts. Another reason that can turn the probabilities hard to estimate is the possibility of strategic behavior, as terrorists adapt their strategies as the security environment adapts to the terrorists’ behavior in its turn. Hence, the reduction of efficiency in smaller airports may increase the threat and, therefore, the risk in the network, with new gaps to be explored by the attackers, by simulating the attacker decision it will be possible to determine where to increase the efficiency or simulate how the defender reacts to this change, to maintain the acceptable risk level in the network. On the other hand, we are considering that the Attacker has a certain behavior that is described in different literature, but there is no direct way to do an accurate model calibration. Considering a network of airports, instead of looking for an airport isolated can create a hierarchical security with increased efficacy. The proposed simulation model works as a proxy of the real world to determine the security risk and the cost of the implemented security.

Acknowledgments The author thankfully acknowledges all the interviewees for all the insights and perceptive suggestions, constructive comments, and remarks, and for the financial support of Fundação para a Ciência e Tecnologia (MIT Portugal Program scholarship PD/BD/113720/2015). The author would also like to thank the guidance from Rosário Macário and Vasco Reis in this research.

References Amorim da Cunha, D., Macário, R., Reis, V., 2017. Keeping cargo security costs down: a risk-based approach to air cargo airport security in small and medium airports. J. Air Transp. Manag. 61. https://doi.org/10.1016/j.jairtraman.2017.01.003. Baker, D., Donnet, T., 2012. Regional and remote airports under stress in Australia. Res. Transp. Bus. Manag. 4, 37e43. https://doi.org/10.1016/j.rtbm.2012.06.011.

Making sense of airport security in small and medium-sized airports

271

British Government, 2018. National Security Capability Review. CSES, U.K., 2011. Framework Service Contract for the Procurement of Studies and Other Supporting Services on Commission Impact Assessments and Evaluations Ex-Post Evaluation of the Preparatory Action on Security Research (PASR) Interim Evaluation of FP7 Security Research Final Report Ex-Post Evaluation of PASR and Interim Evaluation of FP7 Security Research. www.cses.co.uk. Domingues, S., Macário, R., Pauwels, T., Van de Voorde, E., Vanelslander, T., Vieira, J., 2014. An assessment of the regulation of air cargo security in Europe: a Belgian case study. J. Air Transp. Manag. 34, 131e139. https://doi.org/10.1016/ j.jairtraman.2013.10.001. Elias, B., Peterman, D.R., Frittelli, J., 2016. Transportation Security : Issues for the 114 th Congress. European Commission, Irish Aviation Authority, and Avia Solutions, 2004. Study on Civil Aviation Security Financing. Summary of Final Report (TREN/F3/51-2002). Website Der Europäischen Kommission, No. September 2004. European Parliament, 2015. Current Challenges and Future Prospects for EU Secondary Airports. DG for Internal Policies, Policy Department B: Structural and Cohesion Policies, Transport and Tourism. Gerstenfeld, A., Berger, P.D., 2011. A decision-analysis approach for optimal airport security. Int. J. Crit. Infrastruct. Prot. 4 (1), 14e21. https://doi.org/10.1016/ j.ijcip.2011.01.002. Halpern, N., Bråthen, S., 2011. Impact of airports on regional accessibility and social development. J. Transp. Geogr. 19 (6), 1145e1154. https://doi.org/10.1016/ j.jtrangeo.2010.11.006. Kölle, R., Markarian, G., Tarter, A., 2011. Aviation security engineering: a holistic approach. Artech House. Lian, J.I., Rønnevik, J., 2011. Airport competition e regional airports losing ground to main airports. J. Transp. Geogr. 19 (1), 85e92. https://doi.org/10.1016/ j.jtrangeo.2009.12.004. Merkert, R., Mangia, L., 2014. Efficiency of Italian and Norwegian airports: a matter of management or of the level of competition in remote regions? Transp. Res. A Pol. Pract. 62, 30e38. https://doi.org/10.1016/j.tra.2014.02.007. Minnaar, A., 2003. Policing the ports. Reducing illicit trafficking in South Africa. Institute for Security Studies Monographs 2003 (84), 101. Pettersen, K.A., Torkel, B., 2015. Organizational contradictions between safety and security e perceived challenges and ways of integrating critical infrastructure protection in civil aviation. Saf. Sci. 71 (PB), 167e177. https://doi.org/10.1016/j.ssci.2014.04.018. Poole, P.R.W., 2003. A Risk e Based AirPort Security Policy. No. May). Poole, R.W., 2008. Toward risk-based aviation security policy. Int. Transp. Forum. Poole, R., George, P., 2003. Risk-based airport security policy. Reason Public Policy Institute, Los Angeles, California (No. Policy Study 308). Poole, R.W., 2009. The case for risk-based aviation security policy. World Cust. J 3 (2), 3e16. Poole, R.W., 2015. Fresh thinking on aviation security. J. Air Transp. Manag. https:// doi.org/10.1016/j.jairtraman.2015.06.014. Elsevier Ltd. Poole Jr., R.W., 2007. 4. Airport security: time for a new. The Economic Costs and Consequences of Terrorism 67. Prentice, B.E., 2015. Canadian airport security: the privatization of a public good. J. Air Transp. Manag. 48, 52e59. https://doi.org/10.1016/j.jairtraman.2015.06.012. Shim, W., Massacci, F., Tedeschi, A., Pollini, A., 2014. A relative cost-benefit approach for evaluating alternative airport security policies. In: Proceedings e 9th International Conference on Availability, Reliability and Security, ARES 2014, 514e22. https:// doi.org/10.1109/ARES.2014.76.

272

The Air Transportation Industry

Stewart, M.G., Mueller, J., March 2014. Cost-benefit analysis of airport security: are airports too safe? J. Air Transp. Manag. 35, 19e28. https://doi.org/10.1016/ j.jairtraman.2013.11.003. Stewart, M.G., Mueller, J., 2018. Improving checkpoint efficiency: evaluating PreCheck. In: Are We Safe Enough?, 135e53. Elsevier. https://doi.org/10.1016/b978-0-12811475-9.00005-3. Szyliowicz, J.S., 2004. Aviation security: promise or reality? Stud. Conflict Terrorism 27 (1), 47e63. https://doi.org/10.1080/10576100490262160. Talbot, S., Woodward, A., 2009. Improving an organisations existing information technology policy to increase security. Tambe, M., 2011. Security and Game Theory, vol. 9781107096. Cambridge University Press, Cambridge. https://doi.org/10.1017/CBO9780511973031. Unitn, W.S., Cano, J., Rios Insua Urjc, D., Williams, J., Collinson, M., Houmb, S.H., Nguyen Snok, T.-son., Rey, U., Carlos, J., Calle, T.N., n.d. Airport Requirements. Wong, S., Brooks, N., 2015. Evolving risk-based security: a review of current issues and emerging trends impacting security screening in the aviation Industry. J. Air Transp. Manag. 48, 60e64. https://doi.org/10.1016/j.jairtraman.2015.06.013.

CHAPTER 12

How can airports influence airline behavior to reduce carbon footprints? Vasco Reis and Laura Khammash

CERIS, Instituto Superior Técnico, Lisbon, Portugal

1. Introduction With the rise in travel opportunities for passengers and the prospective rise in air traffic offer and demand, working toward reaching environmental impact reduction goals at all levels of the aviation sector has become an increasing impelling need. Not neglecting all other spheres of corporate sustainability, professionals from the industry generally recognize that more needs to be done to offer greener services and products in aviation. At the airport level, an environmental sustainability practice and effective impacts and emission reduction measures often constitute a license to grow, if not to operate (Bylinsky, 2019). Demand for air traffic has been object of great and optimist forecasts in the last decades, and prospect traffic levels were expected to grow at rates that would counteract, in absolute terms, the benefits resulting from the mix of impact reduction measures implemented so far.1 In numbers, the International Civil Aviation Organization predicted a 3.9% average annual growth for passenger traffic and a 4.2% growth for freight traffic at a global scale until 2035, using as base year 2015, with peaks in developing regions such as North Asia with 7.2% annual growth for passenger traffic (ICAO, 2018). A full return to actual operations before the pandemic started to spread is not yet predictable, even though projections estimate that preCOVID-19 traffic levels will be reached again in 3, 5, or even more years. Focusing on the emissions and impacts provoked within the airport boundaries, the heterogeneity of actors that have a direct control over the 1

It is to note that this chapter has been written during the COVID-19 pandemic crisis, during which the world witnessed huge drops in air transport demand and almost total traffic reductions (Le Quéré et al., 2020).

The Air Transportation Industry ISBN 978-0-323-91522-9 https://doi.org/10.1016/B978-0-323-91522-9.00008-7

© 2022 Elsevier Inc. All rights reserved.

273

274

The Air Transportation Industry

polluting sources represents an issue for the airport management decision making process. Airports, however, representing the “face of aviation to the general public” (Bylinsky, 2019) have a strong motivation to act toward environmental sustainability and impact reduction also as a result of local political pressure, legal requirements, and voluntary commitment. The heterogeneity of playing actors and the often underlying discrepancy among their goals and objectives is to be framed within a wider and more complex context where the environmental goal, on a general basis, is felt and dealt with diversely in different regions of the world, depending on the level of political maturity and economic development of each country and region, mainly. Providing an overview of the actual and current practice concerning environmental sustainability options from an airport operator’s perspective, the objective of this chapter is to propose and discuss options to engage with tenant and third-party operators toward the effective implementation of measures, to mitigate the adverse impacts of aircraft movements and other external sources of pollution operating at airports. In particular, different negotiation strategies are explored with the objective to understand the potential benefits of each, based on specific airport characteristics. Drawing from non-aviation fields of negotiation theory, the question whether negotiating could lead to positive agreements toward greener operations on the airside is assessed, in a way to illustrate that win-win situations in the airport-airline interactions are possible when including environmental impact reduction goals. The chapter is structured as follows: Section 2 is an overview of airport operations environmental impacts; Section 3 depicts airport management options to contain impacts from airport controlled sources, while Section 4 focuses on sources which are not controlled by airport operators, and the related issues; Section 5 explores the fundamentals of negotiations strategies and techniques with a potential benefit for airports; finally, Section 6 concludes on the negotiation opportunities and recommendations to be considered in airport sustainability practice.

2. Evolution in air transport traffic and impacts worldwide Global air traffic has been growing above 4% per year over the last 30 years (ICAO, 2018). Available forecasts predicted a steady growth in the coming decades. However, the current pandemic situation changed the global

How can airports influence airline behavior to reduce carbon footprints?

275

picture, with heavy traffic reductions taking place in the last three trimesters of 2020 and an overall drop of 60% in passengers for the same year (ICAO, 2021). The expectations are for reductions to stay, although forecasts are cautious and not updated for the long term (ICAO, 2021). Considering the actual environmental impacts of the transport sector in the past years, and narrowing down to the evolution of carbon emissions within the European Union starting from base year 1990, of all activity sectors, the evolution of transport alone has come with an overall increasing tendency regarding the emission of greenhouse gases (EC, 2020). Breaking down to the different transport modes, air transport grew the most, in relative terms, since 1990 (EC, 2020): This growth is not compatible with the environmental strategies and sustainability development practices envisaged by the European Union. It is generally not straightforward to find aggregated or complete information regarding environmental impacts of aviation in other parts of the world, this observation acting as an indicator of the level of attention that is attributed to environmental sustainability regionally. A consolidated environmental practice is the result of political action and regulation implementation, among regional and local voluntary initiatives. Economic development and political maturity contribute to define the priority of the environmental sustainability agenda. In other words, the lack of information reporting itself is to be seen as a signal that the environmental impacts are disregarded or regarded at in a way that there can be found room for more attention and concrete action. Breaking down the accounting of emissions within the air transport value chain, 80% of total international aviation emissions occur at high altitudes (ATAG, 2020), so that 20% of emissions not only contribute to global climate change but also produce a local impact at and around airports with a consequent concern for air quality and health issues. Narrowing the scope at the individual airport level, all the activities carried out by sources that fall under the direct control of airport operators, contribute to a smaller percentage of total aviation climate impacts, if compared to aircraft operators’ activities. Estimates of the ACI Airport Council International (ACI Europe, 2018) point out that up to 20 million annual tons of CO2, corresponding to 3% of the total aviation industryrelated impacts, are produced directly by airports. Among these impacts, opportunities for further reductions can be found, not counting the emissions that are produced at airports, at a local level and by other actors and operators, the reduction of which would still lay among the interests of

276

The Air Transportation Industry

airport operators. Airport operators as key stakeholders in the aviation industry, moreover, represent a connecting node between relevant stakeholders, such as airlines, ground handlers, air traffic controllers, among others. Airports are required to deliver GHG inventory and corporate social responsibility reports on a regular basis by national governments as a way to promote environmentally and socially responsible behavior of firms. Guidelines for the development of the former in the context of airport operations have been made available by international organizations (NASEM, 2009) and suggest that emissions are classified and then monitored. The air pollutants to be monitored are those reported to have a climate change effect on the atmosphere (Kyoto Protocol) while the classification is based on the responsibility of emissions. Three classes of emissions are identified (NASEM, 2009), depending on the level of control that the airport operator can exert on its sources, with the last class or Scope 3 emissions being those that take place within the boundaries of the airport infrastructure over which the airport operator has no direct control or influence, such as Landing and Take-Off (LTO) cycle emissions. These are often referred to as “optional” (NASEM, 2009), and the airport cannot be considered as responsible for their production. As of the Airport Cooperative Research Program (NASEM, 2009), Scope 3 emissions “include tenant emissions, public ground travel on- and off-airport, and airport employee commute emissions,” to mention the major ones. Tenant activities are necessary for the undertaking of airport activities but are not directly controlled by the airport operator, thus including also aircraft operations, third-party ground support equipment, and vehicles for the access of passengers and staff to the airport. This makes such that, even in the academic literature, studies that consider the airport operator point of view discard the optional or Scope 3 emissions during the assessment and evaluation phase on environmental impacts. The reason for this is the limited availability of data or the complete lack of it (Kılkı¸s and Kılkı¸s, 2016; Monsalud et al., 2015), so that a sort of vicious circle is initiated: The lack of data can be implicitly felt as a way to address the matter as non-relevant, or might lead to letting the issue of Scope 3 emissions fall behind in the priorities to be addressed. Most likely, Scope 3 emissions represent the most significant contribution in terms of annual amount released in the air within the airport boundaries (NASEM, 2009). Nonetheless, it is difficult to picture the

How can airports influence airline behavior to reduce carbon footprints?

277

problem in absolute terms: As an example, an airport with a compact and organized airside layoutdin terms of relative disposition of aprons, taxiways, and runwaysdwill show a better LTO carbon footprint than airports where aircraft take longer distances from apron to runway and vice versa. This adds up to the issue of data availability (Kılkı¸s and Kılkı¸s, 2016; Monsalud et al., 2015), making the estimation of total Scope 3 emissions even harder. As discussed earlier, environmental related data is scarce and of difficult reporting. Scope 3 emissions are no exception due to various reasons: legal requirements to do so, they change from region to region, variability in airport conditions for operations, or privacy of data as companies do now want to share how they carry out operations in close detail. Among scope 3 emissions, the agents responsible for the majority of impacts are ground handlers and airlines. Both are tenants that use the airport infrastructure, make profit through operations that come with significant local air pollution effects, while being inserted in a complex and nested chain of operating emissions sources as schematized in Fig. 12.1. The implementation of mitigation measures affecting Scope 3 emissions sources is understandably complex and difficult. The European Aviation Safety Agency surveyed the hundred major European airports regarding the application of air and noise pollution charging schemes to find out that 35 out of these did not implement any environmental charging scheme so far, whereas only 16 currently charge users for the emission of air polluting substances (EASA, 2019). Liberalizing the air transport market has turned operating an airport into an extremely competitive business in what concerns the stipulation of agreements with airlines. This applies whenever the national regulations and geographical distribution of airports clear the way for airlines’ route planning, and it is especially the case for intra-European routes operated by low-cost airlines. Any negative alteration in prices, such as implementing additional charging schemes, has a direct and immediate impact on the interests of airlines and hence on the airport business itself. Given the vast range of aspects affecting the airport environmental practice, and the relatively low influence of airport operators over polluting sources, the negotiation and mediation promises to be one way for engaging the actors in reducing their environmental impacts in the near and prospect future.

278

The Air Transportation Industry

Emissions at airports Scope 1 Airport-controlled sources

Scope 2 Purchased electricity

Scope 3 Other sources from airport activity Aircraft landing

GSE belonging to the airport Heating

Aircraft taking off

On-site waste management Cooling On-site waste water management

Lighting

Aircraft ground movements APU

On-site power generation GSE belonging to third parties Firefighting exercises Boilers, Furnaces

Passenger travel to the airport Staff commute Off-site waste management Off-site water management Staff business travel

Figure 12.1 Sources of emissions at airport by Scope. (Adapted from Airport Council International (ACI) Europe, 2018. Airport Carbon Accreditation. Annual Report 2017e2018. Retrieved from: https://www.airportcarbonaccreditation.org/component/attachments/? task¼download&id¼132.)

3. Airports environmental practice and carbon reduction initiatives When tackling Scope 1 and 2 emissions (on the sources of which airports have either direct or indirect control, respectively), airport operators can implement measures in autonomy, in compliance with internally set goals, regulations, and environmental pressure, with opportunities to be found among all processes (Fig. 12.2). On the landside, the emissions resulting from the use of motor vehicles for within-airport traffic and operations could be reduced by replacing old vehicles with non-polluting ones. Similarly, other dated support equipment could be replaced or improved in terms of energy use. Among the sources of pollution operating on the landside, it is to mention private and public transport to and from the airport. Even though an increasing number of airports are now linked to their catchment areas and served cities through the rail or light rail modes, road access to the airport still represents a wide share in the way passengers reach the airport, be it by public bus, taxi or

How can airports influence airline behavior to reduce carbon footprints?

279

Mitigation

Landside

Aeronautical

Airport traffic

Airside

Nonaeronautical Electrical appliances

Vehicle traffic

GSE

HVAC

other

Airport maintenance

Aircraft

engines

Oeprational

taxiing

Airport

Runways

APUs

Services

other

Figure 12.2 Major sources of pollution with a potential for mitigation in the airport setting. (Adapted from Ribeiro, J.A.M., Macário, R., Reis, V., 2011. Proposals for mitigation of pollutant emissions from aircraft: application to Lisbon airport. TRANSPORTES 19(2), 34e41.)

similar, or private car, not to mention the frequent case of passengers being taken to the airport by an accompanying driver. Further landside emissions result from the operation and maintenance of terminal buildings, through HVAC, waste and water treatment, the functioning of appliance to handle equipment and luggage. Accounting should indeed include the emissions produced off-site for the provision of electricity: Improving the management and operations of the sources of emissions and implementing measures, be them operational or technological ones, to improve the carbon footprint of these sources, lies in the full responsibility and control of the airport operator. The management and control over sources of air and noise pollution gets complex especially on the airside, with aircraft representing the major sources of pollution at airports: These are not designed to be efficient during taxiing and maneuvering on airport grounds, causing great amounts of fuel to be consumed, with no technological substitute existing for aviation jet engines and for the time being. Technological alternatives are easier to implement when it comes to road vehicles, compared with aircraft engines. Different solutions to turn operations greener on the airside exist

280

The Air Transportation Industry

and include installing electric equipment, replacing the use of APUs with no internal combustion engine alternatives such as GPUs, replacing motor vehicles for ground handling with more efficient or sustainable ones, among others. Among direct measures for airport operators are the modernization of HVAC, lighting, and other appliances in use in terminal buildings, installation of LED lighting systems on apron, taxiways and runways, as well as the installation of renewable energy plants (e.g., solar plant), and finally and for carbon offsetting, planting is also an option. Although there are some emerging collateral issues related to the installation of power plants in the vicinity of runways, mostly due to flare to pilots during approach and landing, and to fauna and bird nest creation occurring in solar panel fields, it can represent a valid source of air impact reduction for airport operations. Regarding carbon offsetting in general and planting in particular, the general idea is that plants will help capture emitted CO2 thus contributing as an offsetting measure to airport practice. This is a responsibility of the airport alone and is more and more considered as a green measure to achieve carbon neutrality goals. It is to be considered an indirect measure for compensation and not for effective reduction of the amount of CO2 and other polluting substances emitted into the air. Tenants, or third-party agents that pollute within the airport boundaries, can also act proactively toward a better carbon footprint of operations. For these actors, the effort to pollute less in the vicinity of airports is subject to a looser commitment with local authorities and the general public, since the use of airport resources and the basis for profit generation is stipulated with the airport operators at fixed stages. In this scenario and with a margin for improvement, measures include the switch to more efficient or renewable energy vehicles for ground handling operations, alternative taxi for ground movements (either electrified or by use of pushback tractors), single engine taxi, the use of sustainable aviation fuels (SAF), as well as LTO efficiency improvement through ATC management. The first mentioned measure is technologically safe and far-sighted, although its main constraint for application is the requirement of strategic initial investment. Moving to aircraft-related emissions, these represent a source of local air quality impacts throughout the whole duration of the landing and take-off (LTO) cycle, with taxi in and taxi out operations holding the biggest share due to a combination of 1. The fact that aircraft engines are not designed to be the most efficient when taxiing on the ground but instead at maximum thrust, and 2. The fact that congestion, and the formation of queues with

How can airports influence airline behavior to reduce carbon footprints?

281

consequent inefficiencies, takes place at major airports, which frequently operate close to capacity limits on the airside (Adacher et al., 2018; Dissanayaka et al., 2019; Simaiakis et al., 2014). The use of Alternative Ground Propulsion Systems is a technological measure to reduce the time aircraft engines are on and polluting, but operational measures are also an option and aim at reducing impacts not by replacing the sources but by through making a more rational and efficient way of the existing ones. Options to minimize impacts at an operational level include performing single engine taxi, using biofuels or sustainable aviation fuels (SAF), continuous descent and climb operations (CDO/CCO), all of which ultimately tackle the emissions of Scope 3 sources but do depend on the decision-making of additional intervening agents, spanning from the aircraft pilot for the case of single engine taxi, to air traffic control (ATC) management for the coordination of CDO and CCO. Airport operators have a wide span of opportunities to work on toward the reduction of their carbon footprint. Opportunities include the improvement of technologies and procedures affecting ground handling and aircraft ground movements. However, as opposed to Scope 1 and 2 emissions, airports do not have the power to tackle the emissions provoked by these sources directly, as the related decision-making involves the intervention of actors which are not strictly committed to airport carbon reduction plans, opening the way for airport operators to engage in communication and sensibilization to nudge the environmental impact reduction goals when negotiation and setting new agreements with these actors.

4. Challenges in environmental sustainability practice at airports and ways forward Several issues arise when talking about the application of environmentoriented policies and impositions at airports. First, a free competition market could represent a limit for the design of incentives to Scope 3 agents for climate intervention, since such incentives could fall under the risk of favoring one actor above other competitors; second, there is a risk, especially for developing countries situations, that the endeavor for business models to improve the companies’ competitiveness does not prioritize or even discard the need for standard environmental action and impact reduction measures to be addressed (Salem et al., 2020); finally, for historical, political, and nonetheless economic reasons, the environmental

282

The Air Transportation Industry

concern is felt and dealt with differently around the world, with developing economies lagging behind on the environmental agenda in result of a different legal and regulatory maturity on the issue. For these reasons, the actual practice that airports and other agents adopt is heterogeneous in maturity and effectiveness. A common practice initiated in Europe and that spread successfully to the rest of the world in the latest years, is to join voluntary-based initiatives. The voluntary connotation adds to the complexity of framing environmental practice and regulation worldwide: If voluntary and not based on legal requirements, actions are only taken if they come with real economic benefit, leaving little to no space for intervention toward options that are not disadvantageous but at least neutral, from an economic efficiency perspective. As put by the authors of a European collaboration project toward carbon reduction measures implementation at airports (dAIR Project, 2014, p. 14), “economic efficiency is, in fact, the only tool that can force or even enable stakeholders to implement CO2 reduction measures.” This is valid not only among the private sector activities, also public sectors in some countries have restraints to apply measures that are not economically advantageous. It is the case for example of Prague Airport, under Czech national laws, which cannot implement any costly measures unless strictly required by law (dAIR Project, 2014). Despite the success of airport carbon accreditation schemes, dealing with environmental issues on a voluntary basis can represent an invitation not to act in situations where the economic benefit is marginal or nonexistent. The voluntary connotation is not the only door for a potential non-attractiveness of engaging in environmental impact reduction. Especially when the focus shifts from a global perspective to the one of developing countries, contemporary literature debates on whether it pays to be green, that is, there is ongoing debate, at all corporate levels, whether being green can lead to significant benefits, notwithstanding the need to go green in a way that is economically efficient (Salem et al., 2020). In terms of competitiveness, adopting an environmental sustainability and impact reduction practice within a company, leads to no significant benefits if the results of environmental practice are not visible in the final products delivered by those companies. For the airline and ground handling industry, it could only translate into engaging with the passengers and logistic clients in a way to inform and raise awareness concerning the way the operations are conducted, in a sustainable and cleaner way. The companies’ and airlines’ focus is to be competitive and attract more customers and revenues,

How can airports influence airline behavior to reduce carbon footprints?

283

and a carbon reduction strategy will only be interesting if it is capable to address and positively affect potential customers. There is argument supporting the idea that, whatever the type or nature of the environmental technologies to be implemented, engaging in a proactive environmental governance toward cleaner practice improves the competitive position of a company (Salem et al., 2020), therefore providing a basis for discussion toward the inclusion of carbon and impact reduction as top list objectives for both airports on one side, and tenant agents operating at the airport on the other, starting from airlines and ground handlers. Therefore, a simple incentive and imposition mechanism might not be the most efficient way to tackle Scope 3 emissions. A relevant challenge is constituted by the fact that imposing limitations over airlines could be counterproductive for the airport in that, in a dense and competitive network of airport, airlines can choose to opt out from negotiated agreements and choose instead another airport to operate. The airport manager walks hence over a thin edge in which it must nudge those agents toward a more sustainable way but without losing them. With due differences, airport managers are faced with a dilemma similar to that of governments when they want to mobilize society and the economy on the path of sustainable development. This is because, although governments may impose certain behaviors, there may be negative or unpredictable impacts, the result of resistance or deviant behaviors of economic agents and populations. As such, in a way to nudge airport agents toward adoption of sustainable practices and behaviors, airport managers can use similar options as those deployed by governments. A key purpose of a public policy is to correct market failures or other competitive limitations (Weimer and Vining, 2011) by inducing agents to change and adopt an expected behavior. Four factors shaping the design of public policies can be identified: desired action to be taken, the target (agent), the nature of barriers (internal or external) preventing the desire action to take place and the type of the public policy. From the airport manager’s perspective, the desired action to be taken refers to the implementation of new energy efficiency and sustainable solutions. The target of the incentives refers to the society. The barriers are endogenous, when the barriers are established by the airport agents, or exogenous, when the barriers are created by factors outside the scope of the airport agents, factors or practices that prevent the adoption of the technologies. These three factors determine the type of public policy (e.g., regulation or directive) to implement.

284

The Air Transportation Industry

Four types of public policy measures may be identified (BemelmansVidec et al., 2011; Vieira et al., 2007): regulation, incentive, transport supply, and information. Regulation can be defined as a “measure taken by governmental units to influence people (and organisations) by means of formulated rules and directives which mandate receivers to act in accordance with what is ordered in these rules and directives” (Bemelmans-Videc et al., 2011, p. 10). The regulation entails an authoritative relation between the ruler and the controlled people or organizations. Two types of regulation may be identified: technical regulation and economical regulation. Technical regulation defines the properties and characteristics of products and services. Economical regulation defines the conditions of access and operation within the market. Regulation tends to be quite an effective incentive because it is compulsory. Setting a parallel between regulatory bodies and the airport management context, the compulsory components of airport practice can be identified in the clauses set by the airport operator to any service level agreements or other contracts to be stipulated with tenant agents to operate at the airport. The regulation or set of clauses can be either technical or economical; however, the freedom for the definition of clauses is limited in that the contracts can be negotiated or rejected by tenants in the case of economically non-advantageous prospect scenarios. The demand and ambitiousness of clauses will therefore result from the actual market power of the airport operator over the airline market in the region. Economic incentives can be defined as measures of “handing out or taking way material resources while the addressees are not obliged to take the measures involved” (Bemelmans-Videc et al., 2011, p. 11). This type of incentive uses the market properties (namely: willingness to pay) to lead people and organizations, on a voluntary basis, to behave in a certain way. This type of incentive is recurrently used by the European Commission. A typical example of economic incentives are the subsidies (e.g., the promotion of the short sea shipping). Airport managers resort to incentives on a recurring basis and in many and varied ways, namely, offering discounts, privileged locations, or personalized services. Transport supply are key incentives that the governmental bodies have to drive the market behavior. In this incentive, the governmental bodies change the properties of the transport network, in terms of capacity, technological properties and even access (construction or removal of tracks). These incentives target the transport network and not the stakeholders, so that they can be considered an indirect type of incentive. They are

How can airports influence airline behavior to reduce carbon footprints?

285

nonetheless effective since the stakeholders will change their behavior accordingly to the properties of the transport network. The final type of incentives, information, can be defined as “attempts at influencing people (and organisations) through the transfer of knowledge, the communication of reasoned argument and persuasion” (BemelmansVidec et al., 2011, p. 11). This type of incentive aims to convince people or organizations to behave in a certain way by demonstrating the benefits or disadvantages of certain actions. Aviation in general, and the airport sector in particular, is known for group work and uninterrupted information sharing between agents. In fact, the A-CDM initiative2 is a paradigmatic case. The selection of the public policy measures should be done according to two conditions. Firstly, the “good” properties that an incentive should exhibit. So far, no general agreement has been reached on what defines a good incentive (Vieira et al., 2007); yet, consensus has been reached on two properties: coherence and effectiveness. Coherence refers to alignment between the public policy and the ultimate purpose to be achieved. In the case of airport context, coherence refers to that the public policy actually nudges airport agents toward a sustainable path including energy consumption and greenhouse gas emissions reduction. The effectiveness of an incentive refers to how close the market’s behavior is to what was initially foreseen. Again, in the case of the airport sector, the effectiveness refers to how close the airport agent’s behavior is of the desired targetdi.e., energy consumption and green gas house emissions reduction. The choice of the mix of options depends on a wide range of factors, such as airport manager negotiation capabilities, airport market power, current situation and envisaged objectives, overall economic cycle, available technology, timelines and deadlines, and so on. Among these options, negotiation is undoubtedly one of the most relevant since the choice of options and the exact terms of their definition and implementation will depend on the airport manager’s ability to negotiatedi.e., present, influence, convinceda win-win situation accepted by all parties.

2

Airport CDM (A-CDM) aims to improve the efficiency and resilience of airport operations by optimising the use of resources and improving the predictability of air traffic. It achieves this by encouraging the airport partners (airport operators, aircraft operators, ground handlers and ATC) and the Network Manager to work more transparently and collaboratively, exchanging relevant accurate and timely information. It focuses especially on aircraft turn-round and pre-departure processes.

286

The Air Transportation Industry

To conclude, the design of a suitable, clear and effective mechanism to tackle environmental emissions at airports is subject to a set of issues and sources of complexity resulting from the heterogeneity of actors’ nature and goals and the complexity of the aviation value chain. Therefore, traditional practices such as incentives or prohibitions may not be the best option to tackle environmental emissions at airports in a competitive, liberalized, and yet regulated framework. Other than legally not always possible, these practices might not be economically feasible. On the contrary, negotiation as a way to overcome obstacles and barriers can offer opportunities to engage with all actors in meaningful dialogue toward win-win agreements.

5. Negotiation Negotiation can be defined as the decision making process that takes place whenever two parts with divergent or opposed objectives discuss in order to achieve a common agreement as opposed to a no-agreement situation, its final objective being the advantageous satisfaction of the needs of all parties involved (Carvalho, 2020). Different negotiation strategies and approaches can be undertaken depending on each involved negotiators’ characteristics and goals. Academic literature extensively addresses negotiation in public policy, and more recently has been reaching the commercial sector (Zachariassen, 2008). Commercial negotiations, in brief, involve the reduction of set prices and the maximization of profit for both sides, with the airporteairline interaction showing its peculiarity in the strive for airport operators to meet high demands and standards for carbon footprint and environmental impact reduction. Also, the type of airline, be it a lowcost or full carrier, and the shared strategic view for future traffic in the specific airport, can influence the type of approach the airport operator might want to adopt. In practical examples, two extreme situations might occur, based on a combination of relationship type and negotiation strategy (Zachariassen, 2008). In the case of an airline-airport solid relationship with common long-term growth commitment, it is fair to assume a partnership kind of interaction, with a desired negotiation situation to be an alignment of goals and action plans. Partnerships involve a share and coordination of information for the prosecution of shared goals in an optimal corporate and working environment, suggesting a win-win solution can be achieved through an integrative negotiation strategy. Partnerships in airline-airport interaction are common at hub airports.

How can airports influence airline behavior to reduce carbon footprints?

287

The opposite extreme occurs in situations where the type of relationship is a one-off and there is no mutual interest in sharing information within an integrative approach, leading to a negotiation that is perceived like a repeated, pre-determined ritual (Zachariassen, 2008). Low-cost carriers at regional airports frequently enter the market with such a strategy, mainly with a predatory behavior over the reduction of prices. The predatory behavior manifests as a consequence of the small traffic volumes at regional airports, which put airport operators in a reduced market power situation, a condition that arises from the need to catch more transport offer, and hence more passengers and passenger revenues. A negotiation toward the implementation of carbon reduction goals will be a non-collaborative one, in that the implementation of measures will increase the distance between the overall goals of the airports and those of the airline. The literature and industry agree to describe the negotiation process as divided into four main phases or stages, as depicted in Fig. 12.3 (Broekens et al., 2010), and as a macro-simplification of more iterative and detailed models (Baber, 2018): preparation, joint exploration, bidding, and closure.

Figure 12.3 Four stages of the negotiation process. (Adapted from Broekens, J., Jonker, C.M., Meyer, J.J.C., 2010. Affective negotiation support systems. J. Ambient Intell. Smart Environ. 2(2), 121e144.)

288

The Air Transportation Industry

The preparation phase starts with a clear statement of the goals and objectives of the negotiator and is generally aimed at developing a plan of action for the upcoming phases, taking into account BATNAs (Best Alternative to a Negotiated Agreement) (Baber, 2018). Interaction between involved parties officially starts at what is called a joint exploration or exchange phase and can be either formal, with each party providing a formal presentation of their companies and goals, or take place smoothly from the previous phase through personal communication of responsible stuff. The core of the negotiation takes place during the bargaining phase. Offers and counterproposals are successfully put into the table keeping in mind the no-deal scenario or BATNA and the initial stated goals and objectives. Finally, the agreement is acknowledged and recognized by the parties and often takes the form of a signed contract, although this is might not always be the case, with some airline-airport vertical interaction components and negotiation outcomes to result from unstructured communication and hence take place in implicit, non-written form (D’Alfonso and Nastasi, 2014). At all stages of negotiation, each partner and actor are moved by their own objectives. Having set different objectives, the strategy, techniques, and tactics followed and implemented by the negotiator will be unique both in its contents and the responses it will generate on the other side. The objectives of each negotiator are not the only aspects to take into account when assessing the optimal negotiation strategy. Indeed, a set of factors comes into place playing a key role in the way agents can deploy their opportunities for achieving their objectives in a context-sensitive framework which considers the specificities and peculiarity of all involved parties. For the situation at hand, a key role is played by the market power and positioning of the airport. Indeed, the mix and type of airlines operating at an airport, as well as the business model of the airport itself in relation with airline agreement, can determine the effective negotiation power the airport operator can exert over an individual airline. Airport-airline interaction will be also influenced by the type of services as well as the volume of demand that each can offer to the respective negotiator, among other factors. A discourse on negotiation opportunities will make the most sense when talking about Scope 3 emissions. The aviation regulatory framework strives to ensure perfect competitiveness for new entrants and established

How can airports influence airline behavior to reduce carbon footprints?

289

players, especially in the airline sector. This translates in limited room for action and inventiveness for negotiators. That is, from the airport operator’s perspective, negotiation should be carried out at all phases bearing in mind what are the limits set by the need not to benefit one airline over the others. Therefore, addressing Scope 3 emissions, and considering the specific nature of the airline market, the airport will need to guarantee perfect competition between different airlines and potential entry airlines, which translates into limitations for certain practices such as incentives or “prizes” to be given by the airport to the better-performing airlines in terms of environmental impacts and reduction goals. For example, assigning faster and closer/contact gates (privileged spots) to “better” airlines might represent an issue on the point of view of competition, as airlines would potentially be disadvantaged in terms of service levels and operational efficiency since such a practice could affect directly their access to the same resources, the privileged gates or aprons. In the case of successful negotiations, the gain manifests not only in the quality of operations but also in creating a positive collaboration and working environment between parties (NASEM, 2010). Regardless of the outcomes, the negotiation process requires a prior assessment of other parties’ needs and wants and for this reason can provide the side benefit of allowing the airport operator to better acknowledge airlines’ and other tenants’ decision-making processes, thus representing a future basic tool for engaging in better informed business and management choices and decisions. Policy and political issues, that can be either local, national, or regional, are also relevant aspects to be considered during negotiation from both parts, as they may, either positively or not, influence the outcome of the negotiation itself and on the final agreements to be reached (NASEM, 2010). Ultimately, the preparation and consideration of all interaction components and possible outcomes, should be taken as an opportunity for the airport operator to nudge tenant agents toward the adoption of a more sustainable, with reduced environmental impacts, practice. Ideally, in a negotiation a successful negotiation setting and relationship, the final objective of all actors involved is reaching an agreement that represents a win-win solution, where all parties are better off and partly meet their original goals, with a no-agreement to represent a commercial loss for both sides in terms of revenues and expansion opportunities.

290

The Air Transportation Industry

6. Good practice recommendations and opportunities The major source of air pollution within the air transport value chain is represented by aircraft. International aviation, that is, flights with a 1500 km or more distance for which no other modal alternative exists, provokes the emissions of substances that is dispersed both globally and locally: 20% of these are released at and around airports causing an additional concern for local impacts (ATAG, 2020). Considering the expected growth for overall traffic worldwide, even assuming a temporary halt with an uncertain deadline for recovery due to the pandemic currently taking place, designed and implementable measures to reduce impacts will not be able to counteract the increase in overall emissions resulting from increased traffic (ICAO, 2019). A set of sources operate at airports to guarantee full operations. Some of these sources are under the direct responsibility of the airport operator, some depend on third-party agents, conditional to airport management approval through contracts, while others involve the responsibility of important clients and partners over which the airport operator might not have sufficient market power to wedge in demanding conditions for the reduction of environmental impacts, a market power indeed threatened by the risk of losing a strategic partner or client. Depending on which the sources to be tackled are, different implementation options exist. As a major distinction, Scope 1 emissions will fall under a bigger strategic view of the airport operator itself, Scope 2 emissions will result from new contracts between suppliers and the airports, with a relative freedom for airports to opt for environmentally sustainable options, and finally, Scope 3 emissions will be successfully addressed as a result of a productive negotiation between airports and tenants, namely ground handlers, and most of all airlines. More generally, several options to engage in negotiations with airlines are feasible and will depend on the actual characteristics of the involved actors. Table 12.1 is intended as a recap of the possible measures at all stages of airport activities, together as a tool to depict mitigation options for the airport manager, depending on the desired measure to implement. In a liberalized market, it is the market power of an airport to determine its space and capacity for negotiation with tenants. It is especially the case when talking about service level agreements and contracts with airlines, being the airline market a free though regulated one concerning competition and competitiveness. Plus, the airport is strongly relying on airlines for airport revenues, and if the passenger share is distributed unevenly between

How can airports influence airline behavior to reduce carbon footprints?

291

Table 12.1 Summary of impact reduction measures and airport mitigation options by Scope. Emission Source Involved Mitigation type description Mitigation measure actors options

Scope 2

Electric appliances (landside)

Modernization HVAC, lighting, and other appliances in use in terminal buildings LED lighting systems on apron, taxiways, and runways

Electricity suppliers, airports

Switch to more efficient or renewable energy vehicles for ground handling operations Alternative taxi for ground movements, either electrified or by use of pushback tractors Single engine taxi

Ground handling providers

Aircraft

Use of sustainable aviation fuels

Aircraft LTO

LTO efficiency measures through ATC management Installation of renewable energy plants, such as solar power plants, on the green fields within the airport boundaries Planting as an option for carbon offsetting

Airlines, aircraft refuelers Airlines, ANSPs

Electric appliances (airside)

Scope 3

Airside operating fleet Aircraft

Aircraft

All scopes

Electricity provision

Overall carbon footprint

Electricity suppliers, airports

Airlines, Ground handling providers, ANSPs Airlines, ANSPs

Environmental sustainability strategic planning, Investment Environmental sustainability strategic planning, Investment Contracts, Incentives

Contracts, Incentives

Incentives, Information/ cooperation Contracts, Incentives Information/ cooperation

Airports

Environmental sustainability strategic planning, Investment

Airports

Environmental sustainability strategic planning

292

The Air Transportation Industry

different airlines, one airline can implicitly impose its own conditions on the premises that their economic contribution to the airport operations is crucial. Let’s take a look at the airside passenger market alone: we are talking about passenger airlines of any kind (charter/seasonal, low-cost, and full legacy), as well as all other service providers to ensure airline operations on the ground, starting from ground handling. This market has been privatized in the last decades in various countries and the overall tendency is toward a progressive liberalization. In such a setting, national governments can impose further limitations or demand minimum services in the name of social obligations and overall sustainability. Airports and tenants are then forced to operate and renegotiate contracts within the limits imposed on market competition and competitiveness laws on one side, and minimum service requirements on the other. Although many countries allow a liberalized market regarding the provision of air passenger transport, the system is not completely unregulated. Which turns the discourse over the opportunities for negotiation and incentivizing an interesting one.

7. Conclusions Regulation imposes that free competition is ensured and hence no special treatment is applied to any of the competing actors (e.g., different airlines). Plus, the mix of airlines operating at a certain airport, and the characteristics of airports themselves, open opportunities for discussing over the effective market power for negotiating environmental objectives with tenants. In the case of reduced market power on the side of the airport management, it will be harder to impose any new clause compared to old contracts, or clauses that induce no economic benefits to tenants. Market power exerts a significant role on the negotiation of Scope 3 emission reduction measures, as airports in general, and in particular small and regional ones, have less power if compared to airlines when designing new contracts. In light of the reduced market power that airport operators face as the major constraint for the achievement of sustainable development goals, negotiation can represent a useful practice to overcome impasse situations over the prioritization of environmental impact reduction goals by airports. Finally, negotiation represents a key tool to support the strategic decisionmaking of the airport operator who intends to effectively take the road toward carbon footprint reduction. Negotiations in this setting will need to be able to bridge and link the necessities and objectives of each agent involved in

How can airports influence airline behavior to reduce carbon footprints?

293

order to reach a win-win solution that satisfies all parties. Notwithstanding, strategies applied by airports will have to be molded out of each tenant’s specific assets and power, in addition to the relative market power and consequent space for imposing clauses that is created between actors.

References Adacher, L., Flamini, M., Romano, E., 2018. Airport ground movement problem: minimization of delay and pollution emission. IEEE Trans. Intell. Transp. Syst. 19 (12), 3830e3839. Air Transport Action Group (ATAG), 2020. Factsheet 2: Aviation and Climate Change. Retrieved from. https://aviationbenefits.org/media/167159/fact-sheet_2_aviationand-climate-change.pdf. Airport Council International (ACI) Europe, 2018. Airport Carbon Accreditation. Annual Report 2017e2018. Retrieved from. https://www.airportcarbonaccreditation.org/ component/attachments/?task¼download&id¼132. Baber, W.W., 2018. Identifying macro phases across the negotiation lifecycle. Group Decis. Negot. 27 (6), 885e903. Bemelmans-Videc, M.-L., Rist, R.C., Vedung, E.O., 2011. Carrots, Sticks, and Sermons: Policy Instruments and Their Evaluation. Transaction Publishers, New Jersey. Broekens, J., Jonker, C.M., Meyer, J.J.C., 2010. Affective negotiation support systems. J. Ambient Intell. Smart Environ. 2 (2), 121e144. Bylinsky, M., 2019. Airport Carbon Accreditation: The Global Carbon Management Standard for Airports [Presentation]. https://www.icao.int/Meetings/ GREENAIRPORTS2019/Green%20Airports%20Presentations/ Marina%20Bylinsky%20session%203.pdf. Carvalho, J.C., 2020. Negociação, sixth ed. Edições Sílabo, Lisbon. D’alfonso, T., Nastasi, A., 2014. AirporteAirline interaction: some food for thought. Transp. Rev. 34 (6), 730e748. Dissanayaka, D.M.M.S., Adikariwattage, V., Pasindu, H.R., October 2019. Evaluation of emissions from delayed departure flights at Bandaranaike International Airport (BIA). In: 11th Asia Pacific Transportation and the Environment Conference (APTE 2018). Atlantis Press, pp. 183e186. dAIR Project, 2014. dAIR: Decarbonising Airport Regions [final Report]. European Union Aviation Safety Agency (EASA), 2019. European Aviation Environmental Report 2019. https://doi.org/10.2822/309946. Available at: European Commission (EC), 2020. EU Transport in Figures e Statistical Pocketbook 2020. Available at: https://ec.europa.eu/transport/facts-fundings/statistics/pocketbook-2020_en. International Civil Aviation Organization (ICAO), April 2018. ICAO Long-Term Traffic Forecasts: Passenger and Cargo. Retrieved from. https://www.icao.int/sustainability/ Documents/LTF_Charts-Results_2018edition.pdf. International Civil Aviation Organization (ICAO), 2019. Aviation Environmental Report. Available at: https://www.icao.int/environmental-protection/Documents/ICAOENV-Report2019-F1-WEB%20(1).pdf. International Civil Aviation Organization (ICAO), 2021. Effects of Novel Coronavirus (COVID-19) on Civil Aviation: Economic Impact Analysis [PowerPoint Slides]. https://www.icao.int/sustainability/Documents/Covid-19/ICAO_coronavirus_Econ_ Impact.pdf.

294

The Air Transportation Industry

Kılkı¸s, S., Kılkı¸s, S., 2016. Benchmarking airports based on a sustainability ranking index. J. Clean. Prod. 130, 248e259. Le Quéré, C., Jackson, R.B., Jones, M.W., et al., 2020. Temporary reduction in daily global CO2 emissions during the COVID-19 forced confinement. Nat. Clim. Change 10, 647e653. https://doi.org/10.1038/s41558-020-0797-x. Monsalud, A., Ho, D., Rakas, J., 2015. Greenhouse gas emissions mitigation strategies within the airport sustainability evaluation process. Sustain. Cities Soc. 14, 414e424. National Academies of Sciences, Engineering, and Medicine (NASEM), 2009. Guidebook on Preparing Airport Greenhouse Gas Emissions Inventories. The National Academies Press, Washington, DC. https://doi.org/10.17226/14225. National Academies of Sciences, Engineering, and Medicine (NASEM), 2010. Airport/ Airline Agreements Practices and Characteristics. The National Academies Press, Washington, DC. https://doi.org/10.17226/22912. Salem, M.A., Shawtari, F., Hussain, H.B.I., et al., 2020. Environmental technology and a multiple approach of competitiveness. Futur. Bus. J. 6, 17. https://doi.org/10.1186/ s43093-020-00012-1. Simaiakis, I., Khadilkar, H., Balakrishnan, H., Reynolds, T.G., Hansman, R.J., 2014. Demonstration of reduced airport congestion through pushback rate control. Transp. Res. Pol. Pract. 66, 251e267. Vieira, J., Moura, F., Manuel Viegas, J., 2007. Transport policy and environmental impacts: the importance of multi-instrumentality in policy integration. Transp. Pol. 14 (5), 421e432. https://doi.org/10.1016/j.tranpol.2007.04.007. Weimer, D.L., Vining, A.R., 2011. Policy Analysis. Retrieved from. https://books.google. pt./books/about/Policy_Analysis.html?id¼mZEnAQAAMAAJ&pgis¼1. Zachariassen, F., 2008. Negotiation strategies in supply chain management. Int. J. Phys. Distrib. Logistics Manag. 38 (10), 764e781. https://doi.org/10.1108/096000 30810926484.

CHAPTER 13

The measurement of accessibility and connectivity in air transport networks Augusto Voltes-Dorta1 and Juan Carlos Martín2 1

The University of Edinburgh Business School, Edinburgh, United Kingdom; 2Institute of Tourism and Sustainable Economic Development, University of Las Palmas de Gran Canaria, Las Palmas de Gran Canaria, Spain

1. Introduction Accessibility was defined in the seminal paper by Hansen (1959) as “the potential of opportunities for interaction.” Twenty years later, Morris et al. (1979) defined three proxies for the concept of “interaction”: (1) activities that can be reached from (2) a certain place and with a specific (3) transport mode. Gutiérrez (2001) contended that Morris et al. accessibility concept initiated a number of operational indicators which were mainly based on travel costs from a certain place to a selection of potential destinations within a demarcation area, and the attractiveness of all the locations selected in the study. Geurs and van Wee (2004) proposed the four main components model (Where-How-When-Who) in the analysis of accessibility: (1) land use (where to go); (2) the transportation system (how to go); (3) the time dimension (when to go); and (4) the individual or segmentation analysis (who is going). More recently, Levine (2020) pointed out that the accessibility concept has evolved over a number of dimensions such as quantity, levels of aggregation, and place versus person-based measures, and Handy (2020) argued that accessibility reflects the distribution of destinations around a given place, the ease of getting to those destinations provided by transport systems, and the individual benefits that can be achieved on the destinations. The land use component usually determines the type of activities that can be developed by individuals or firms at destinations. The second component (the transportation system) measures the travel cost between the areas as affected by the available services. The third component relates to the time dimension and it has been empirically analyzed in relation to peak versus off-peak periods, different days of the week and seasons such as The Air Transportation Industry ISBN 978-0-323-91522-9 https://doi.org/10.1016/B978-0-323-91522-9.00010-5

© 2022 Elsevier Inc. All rights reserved.

295

296

The Air Transportation Industry

summer versus winter. The fourth component reflects the characteristics of the individual travelers, firms or population segments determined by age, income, gender or race. The literature on accessibility is abundant, but air transport is clearly underrepresented. Most of the studies are only focused on one mode, and cities are the most common demarcation areas of study. Focusing on air transport, land use is clearly determined by the airports’ proximity and the different role that airports play in different cities which are closely related to the topological property of airport centrality (Rodriguez-Deniz et al., 2013; Suau-Sanchez et al., 2015; Zhu et al., 2019). The transport system has been mainly analyzed from the dynamics of airline networks (Dai et al., 2018; Jia et al., 2014; Wang et al., 2016; Zhang et al., 2017). The time dimension is especially relevant in some airports in which airline networks are highly constrained by the slots available, especially at peak periods, and there exists an important gap in the literature analyzing the loss of accessibility and connectivity by this issue. And finally, regarding the individual component, the literature is also scant with the exception of Bombelli et al. (2020) who analyzed the global air cargo network including passenger airlines, fullcargo airlines, and integrators. They used complex network analysis (CNA) theory and found that the cargo network exhibits more concentration than the air passenger transport network and that cargo airports possess larger catchment areas than passenger airports. Air transport accessibility has mainly been analyzed from the airlines’ perspective studying the main topological properties of the networks (Borodako and Rudnicki, 2014; Neil, 2013). Nevertheless, Van Wee (2016) reviewed the literature on accessibility concluding that air transport accessibility should be further studied beyond this as other components have also an important effect on regions, firms, and citizens. With this idea, the contribution of this chapter is threefold. First, it will present the main methodologies that have been applied in the literature of air transport accessibility. Second, it will provide a consistent review of air transport accessibility jointly with other related concepts such as connectivity, resilience and vulnerability. And third, it will provide a future research agenda analyzing the main gaps existing in the literature.

2. An overview of air transport accessibility A number of partial indicators have been developed in the last decades to measure air transport accessibility. The current chapter focuses on air

The measurement of accessibility and connectivity in air transport networks

297

transport but it is evident that a door-to-door trip that involves air transport is clearly multi-modal as other transport modes are necessary to transfer to/ from the airports of origin and destination. Bowen (2000) studied accessibility of Southeast Asian hubs for 1979 and 1997 using only graph theory topological indices (Shimbel and centrality indices). The Shimbel index measures the average minimum number of segments necessary to connect each hub airport with the rest of the network nodes (i.e., airports) included in the study. Thus, the nodes can be ranked by order of importance according to a decreasing order of the index so that the minimum value of the index (i.e., one) would correspond to an airport that is fully connected with all other airports. The centrality index is measured using an adjacency matrix in which each element indicates whether a nonstop connection exists or not between a pair of airports. The sum of the column and row sums for each airport yields the number of connections that exist for each airport in the set of airports. The figure is normalized so the airports can be ranked according to the increasing order of the centrality index. Yamaguchi (2007) analyzes the effects of air transport accessibility on percapita GDP growth using a dataset of 47 prefectures in Japan between 1995 and 2000. The author defines air transport accessibility between two prefectures in relation to the generalized cost of air transport between the origin and destination prefectures, including the air ticket price and the value of time, the share of air transport trips between the prefectures, and the relative economic importance of the origin and destination prefectures as a way to measure the relative interaction. This index is more similar to the partial indicators that have been developed in the accessibility literature, as it includes the generalized costs and the attractiveness (economic level and relative share of air transport) components as defined in Gutiérrez (2001). Hesse et al. (2013) analyzed air transport accessibility for 82 major European cities using four partial accessibility indicators (potential, daily, location, and relative network efficiency). These four indicators have been used more extensively in other transport modes such as public transport, HSR, and road transport (Martín and Reggiani, 2007). 2.1 The potential indicator The potential indicator is a gravity-model-based index that takes into account travel costs and attractiveness. The index ranks the accessibility of the airports or catchment areas as high values mean that a great economic

298

The Air Transportation Industry

potential is achievable from the origin. Thus, the potential Pi for an airport i can be defined as: Pi ¼

n X GDPj j¼1

tij

(13.1)

GDPj is equal to the gross domestic product of the NUTS3 area in which the airport is located, and tij denotes the travel time between the airports, which includes a scheduled delay time penalty which depends inversely on flight frequency during an average week. It is not clear how the time was calculated for airports that are connected by two or more trip segments, or how competition between direct versus connecting flights was taken into account. 2.2 The daily accessibility indicator Daily accessibility indicators use a travel time threshold to measure the possibilities for interaction and they are also known as a “cumulative opportunities” model. These indicators have more relevance when normative aspects of accessibility are highlighted (Páez et al., 2012). For example, when policy planners expect that citizens must not travel to some facilities such as hospitals, medical doctors, pharmacies, or schools more than 30 min. Air transport is not an exemption to these normative aspects, as, for example, 90% of travelers within the EU are expected to be able to complete their door-to-door trips within 4 h (European Commission, 2011). Hesse et al. (2013) chose a threshold figure of 3 h to measure the daily accessibility provided by air transport. The daily accessibility indicator DAi is obtained as follows: DAi ¼

n X

Popj *dij

(13.2)

j¼1

where dij ¼ 1 if

tij  3;

and 0 otherwise

POPj considers the population of the NUTS2 area in which the airport is located, and the delta parameter denotes the threshold figure. In our view, 3 h is a very stringent threshold unless the authors want only to consider the accessibility for business passengers who do not usually check

The measurement of accessibility and connectivity in air transport networks

299

in luggage at airports. This figure should also vary according to the number of daily frequencies that exist between each city pair. For example, in 2018, according to OAG, Seoul-Jeju was the busiest air transport route in the world with 79,640 flights, an average daily frequency of around 218 flights per day, which is equivalent to approximately one flight per each 4 min (considering an average operating schedule day of 14 h). These frequencies can vary drastically across countries and markets, thus affecting potential daily accessibility. 2.3 The location indicator The location indicator measures a weighted average of travel times, travel costs or equivalent distances from a given origin to all destinations included in the analysis. The higher the index, the lower is the accessibility of the node. It is highly correlated with the centrality index measured by the network analysis and reflects the locational disadvantage that is experienced by peripheral cities. Hesse et al. (2013) calculated the location indicator according to: Li ¼

n X

tij sj

(13.3)

j¼1

where GDPj sj ¼ P n GDPk k¼1

The variables included in the index are the same ones as in the potential indicator. It does not define a decay function for distant nodes, so the calculation considers each city pair as equally representative, which is a very strong assumption made by this indicator. As in the rest of the indices, we miss the role that flight frequency can play in the analysis. 2.4 The relative network efficiency indicator The relative network efficiency indicator NEi is an interesting extension of the location indicator. In this case, the travel time between the two airports is compared to a hypothetical “optimal” travel time. This optimal travel

300

The Air Transportation Industry

time can be either estimated or obtained from empirical observations included in the analysis. This index has been used by Hesse et al. (2013) and Martín et al. (2004), and can be obtained as follows: NEi ¼

n X tij j¼1

et ij

sj

(13.4)

where GDPj sj ¼ P n GDPk k¼1

Hesse et al. (2013) calculated the optimal travel time for each pair ( et ij ) as the straight line distance between i and j converted to time with a ratio of 600 km per hour. Again, this could be normalized according to the densest route in the dataset in terms of flight frequency. Another interesting issue is the sensitivity of the index to non-stop, one-stop, and two-stop connections to analyze in which cities the accessibility index is more or less affected by the structure of airline networks. Ultimately, Hesse et al. (2013) performed a principal component analysis to calculate a synthetic air transport accessibility index for the 82 cities, finding that the results were quite robust to including or not the airline flight frequency information, although some partial indicators differed substantially. The results could be partly explained by the size of the cities used in the analysis, but caution should be exercised as air transport networks are dynamic entities that evolve rapidly in time and space. A better approach to capture these effects is to use complex network analysis, which is based on the representation of the network as static or dynamic graphs (Zweig, 2016). For that reason, the next section will present the relationship between the intertwined concepts of accessibility, connectivity, resilience, criticality and vulnerability which are salient phenomena from the perspective of complex network theory (Reggiani et al., 2015).

3. Air transport accessibility and related concepts Air transport accessibility is an important driver for economic growth. In fact, both developed and developing countries usually foster this attribute in their national economic development strategies for the apparent link that exists between air transport accessibility and economic growth. At the same

The measurement of accessibility and connectivity in air transport networks

301

time, economic growth is one of the key drivers of air transport accessibility, among other drivers such as environmental regulations, aircraft design, airport investments, market liberalization, airline business models, and competing transport modes. All these drivers cannot be studied in isolation, and the role of the EU in their development is crucial. Thus, in December 2015, the European Commission (2015) presented an “Aviation Strategy for Europe,” setting out an ambitious vision for the future of European aviation that aims to boost Europe’s economy, strengthening its industrial base and reinforcing its global leadership position in the aviation sector while addressing both environmental and climate challenges. The EU is also setting out the conditions to establish aviation agreements with third countries in order to create legal stability and certainty for investors. The strategy will foster a complete liberalization of the skies with as many non-EU Member States as possible. Thus, it is highly plausible that airline networks will evolve as a product of the liberalization strategy affecting air transport accessibility, connectivity, resilience, criticality and vulnerability. For this reason, the next subsections will provide a good understanding of the relationship that air transport accessibility has with the commented intertwined concepts. 3.1 Accessibility and connectivity Arvis and Shepherd (2011) proposed an air connectivity index (ACIN) based on a gravity model that captures the essence of market potential. For the first time, the authors focused on the country as the level of analysis and measured connectivity having in mind the concept of accessibility. The authors contended that a good definition of connectivity should have the following properties: (1) it should be realistic and linked to transportation models such as the gravity-type model; (2) it should control for the problem of self-potential in the sense that connectivity should be the same for two nodes with the same number of connections independently of the size of the node; (3) it should be dimensionless and normalized; and (4) it should be global and not sensitive to the units of the demarcation area. The authors found that past air connectivity measures were mainly based on simple metrics, concentration indices, clustering methods and centrality measures. Their proposed ACIN takes into account both push and pull factors. The authors found that ACIN is a function of both geographical (i.e., remoteness) and connectivity aspects, such as the diversity of a node’s connections that is highly affected by policy measures like, for example, open skies agreements.

302

The Air Transportation Industry

Burghouwt and Redondi (2013) contended that the level of air accessibility of airlines and airports cannot be exclusively measured in terms of supply measures such as enplanements or aircraft movements, and that air accessibility depends on both direct and indirect connectivity. They classified the measures according to the following six dimensions: (1) accessibility versus centrality; (2) temporal coordination at the hubs; (3) routing factor; (4) quality of a connection; (5) maximum number of legs for the trip; and (6) demarcation area of the study (local vs. global markets). The authors compared eight different connectivity models employed in air transport literature and applied them to a set of European airports. They conclude that traditional size-based measures tend to underestimate the accessibility of small airports when these are well connected to important hub airports. At the same time, these measures also tend to overestimate the centrality of large airports that are important economic centers that attract passengers from different regions. Allroggen et al. (2015) considered for the first time all scheduled flight connections to measure a “quality-weighted” connectivity index that considers the “value” of each connection from the observed passenger behavior. The authors contended that the indices are sensitive to deregulation and liberalization processes seen in different world regions, as well as the emergence of global airline alliances and new airline business models. The quality of the connection was assumed to depend on both frequency and directness factors that affect passenger utility. The quality of the destination is also included in the model in order to analyze the potential for the interaction in the accessibility literature. Unfortunately, due to problems on data availability, the role of fares and competition was not included in the model. The analysis of the key drivers of air transport connectivity is another interesting strand of the literature. There are two recent studies Zhang et al. (2017) and Antunes et al. (2020) that analyzed the drivers on the air connectivity at Chinese airports during the period 2005e2016 and in Europe for the period 2009e16. Zhang and his colleagues found that airline competition is a positive driver on airport connectivity, as well as city population, GDP per capita, being a hub airport or a highly-dominated one by the presence of a single low cost carrier (LCC). Strangely, more competition from high-speed rail (HSR) acts as a barrier to airport connectivity. Meanwhile, Antunes and his colleagues found that a good level of regional air connectivity is built on a sort of “crowding-out” effect, that is, decreasing the connectivity of the neighboring regions. They also found

The measurement of accessibility and connectivity in air transport networks

303

that GDP per capita, population density, tourist arrivals are important key drivers of air regional connectivity in Europe. Interestingly, they also showed that LCCs do only improve the connectivity in remote regions. 3.2 Accessibility, resilience, criticality, and vulnerability Reggiani et al. (2015) analyzed the transportation field and concluded that resilience, vulnerability, connectivity, accessibility, and security are all intertwined concepts (Caschili et al., 2015; Mattsson and Jenelius, 2015; Östh et al., 2015). Jenelius and Mattsson (2015) contended that vulnerability is usually measured analyzing the effects of the potential degradation of some elements of a transport network, the second component of the accessibility concept. Indeed, air transportation systems can be affected by volcanos eruptions, snowstorms, electrical power breakdowns, terrorist attacks, or computer reservation system collapses. Jafino et al. (2020) contended that previous literature on transport resilience and vulnerability has used different terminologies such as criticality and importance analysis. The broad spectrum of the field regarding transport modes and the respective measures has produced inconsistent findings creating confusion among practitioners and policy makers. After a systematic literature review, the authors found 17 measures that analyzed resilience, criticality, and vulnerability in transportation systems. They also concurred with Jenelius and Mattsson (2015) in that the selection of measures should consider all functionality, ethical and aggregation dimensions. A five-step guideline to select the measure was finally proposed. According to Martin (2012), the four main reasons to study resilience in transportation systems are: (a) the impact of natural and man-made disasters that have afflicted airports and airlines; (b) the influence of other disciplines, such as ecology, where the main interest is in how ecosystems respond to shocks; (c) recognition that major disruptions can affect the air transport system; and (d) the effects at both local and regional levels of the airports’ catchment areas to understand the interactions between airports, airlines and governmental policies pursued at different levels. Furthermore, Manca et al. (2017) argued that the duration of the disruption and its effects determine different types of resilience namely absorptive, adaptive and transformative (Giovanni et al., 2020). Wong et al. (2020) compared the data-driven methods based on the Mahalanobis distance with more traditional CNA measures such inverse average path length, cluster size, and algebraic connectivity. Their results

304

The Air Transportation Industry

showed that all the approaches can determine airline resilience but they admitted not to have a clear suggestion for coming up with the most appropriate method. Thus, they conclude suggesting that different approaches can provide a complementary vision of airline resilience. Sun et al. (2020b) contended that the analysis of the efficiency and resilience of air transportation will be one of the major challenges in this century. For this reason, the authors proposed a method to analyze the resilience of the 5000 largest cities in the presence of airport disruptions. In this sense, the analysis is based on the loss of air accessibility of the cities when a single airport is disrupted from the network. The analysis concluded that about 70% of the sample cities have at least one airport which can be reached within 2 h of driving time. A significant proportion of those cities have two or more reachable airports. The two cities with the largest number of reachable airports are Antwerpen and Oxford with 10 airports, followed by 14 cities that can reach nine airports, e.g., New York, Brussels, Cologne, Wuppertal, Liege, Gent, Eindhoven, Newark, Reading, Luton, Paterson, Arnhem, ‘s-Hertogenbosch, and Maastricht. Voltes-Dorta et al. (2017) analyzed the vulnerability of the European air transport network measuring the delays imposed to disrupted airline passengers when one of the 25 busiest European airports is disrupted from the network during a whole day. The authors assumed that all the affected passengers can be relocated through minimum-delay itineraries within their own airlines’ alliance, and the aggregation of delays is used to rank each airport’s criticality. The airport criticality is also compared with three well known CNA indicators such as degree centrality, closeness, and between centrality. The authors found that the resilience of European airports rests mainly on the ability of airlines to find local surrogate airports, so intermodal aspects can play a key role. It is still soon to anticipate how the air transport industry is going to recover after the severe troubles caused by the COVID-19 pandemic in the form of travel bans or barriers such as quarantines or mandatory testing, and the recovery sharp V-shaped patterns observed in the past after other disruptions seem to be wishful thinking (Suau et al., 2020). Suau and his colleagues conducted a qualitative research through semi-structured interviews to a convenience sample of 16 senior aviation executives. The respondents considered that regional airports were going to suffer more than other airports due to airline consolidation trends. From the demand side, the analysis concluded that business travel will be reduced as companies need to reduce costs and videoconferencing might stay in the future. Nevertheless, for leisure passengers, the respondents highlighted that the

The measurement of accessibility and connectivity in air transport networks

305

impact of the pandemic will be less intense, and they expect that tourism authorities will actively support the air traffic recovery through incentive programs. Finally, regarding the regulatory aspects, the respondents expressed concern on how health-screening processes at the airports and other social distance measures will affect airline load factors and their longterm profitability. All in all, it seems clear that the literature on air transport accessibility will continue to mature as an important research field with new methods, concepts and applications. To this end, the next section will provide a tentative research agenda that could address in part some of the gaps that exist. Obviously, the list of topics is not exhaustive, and some topics can stretch across different academic disciplines.

4. A tentative future research agenda The analysis of the 19 cluster events organized by the accessibility cluster of Nectar (Network on European Communication and Transport Activities Research), as well as the special issues and books published on the topic (e.g., Geurs and Macharis, 2019) show that air transport is clearly underrepresented in the accessibility literature. Following some of the ideas and topics raised by Macharis and Geurs (2019) and van Wee (2016), this section adapts them to the aviation industry in order to provide a tentative research agenda for the coming years. 4.1 Short distances and greener modes Air transport accessibility can be constrained by social norms or regulations regarding the environmental damage that is produced by the mode. In this respect, Gössling (2019) argued that the existing debate between climate change and air transport is relevant because the social discussion of decarbonization has changed for the first time from the producer (supply) side to the consumer (demand) side. Another important aspect to highlight is the role that social norms play in determining what society as a whole decides what can be consumed (Bicchieri et al., 2018). Once an individual has learned to behave in a manner consistent with the interests of the group, they will tend to persist in the learned behavior unless it is clear that, on average, the cost of maintaining the norm significantly outweighs the benefits. Small groups are generally able to monitor their members’ behavior and successfully punish whenever an individual is found to be breaking the rules.

306

The Air Transportation Industry

In light of the challenge to decarbonize transport, and the need to reduce greenhouse gas emissions, Gösling et al. (2019) discussed the utility value of air transport from individual and societal perspectives, while considering moral and economic viewpoints. It is evident that the values are subjective and cannot be easily generalized. For example, the authors found that the subjective values of 29 international students at Lund University (Sweden) were affected by motives, experience, life stage, or other situational factors. As expected, leisure flights are considered less important than other flights, and, at the same time, it is the motive most frequently cited. The authors concluded that moral concerns and social norms could coevolve so social practices, and policy support for them can be highly affected. This discussion has not entered in the field of air transport accessibility and it should be incorporated in the future agenda as it will gain more relevance. 4.2 Intermodality Most air transport accessibility studies are based on only the air transport network but it would be necessary to extend the scope of analysis to include road, railways and high speed railways (HSRs) as a way to complement part of the trip by one of these modes. In this sense, we highlight the study of Sun et al. (2020a) in which the authors analyzed the airports’ road access at a global scale using an open-source database. Accessibility profiles and catchment areas for the sample airports were determined. Sun and her colleagues found that, in Europe and North America, more than 90% of the population can reach at least one airport within 90 min, and in Southeast Asia, the value is much lowerdonly 44%. The authors also found that there is a high variability in the ratio of potential over actual demand that can be explained by either the different hierarchy of hub airports or by the functioning role of airports that serve tourism hotspots. The authors mentioned that future studies that consider other surface transport modes such as public transport, railways, buses, or subways are needed. For example, Gallotti and Barthelemy (2015) integrated the timetables for air, rail, metro, coach, bus, and ferry as a multi-layered network in the United Kingdom for a week in October 2010. In the future, better multi-layered networks could include all the transport modes that can be used to access/egress to/from the airports, especially the HSRs. Sun et al. (2017) contended that the new developments in HSRs have gradually blurred the line between competition and cooperation between air and HSR. The integration of high-speed

The measurement of accessibility and connectivity in air transport networks

307

trains and air transport gained academic attention because of the political support given in the EU to relieve capacity constraints of some of the main EU airports, and to reduce environmental costs (Givoni, 2007; Givoni and Banister, 2006). Zhang et al. (2019) reviewed studies on the impacts of HSRs on airports and found that some hub airports with good air connectivity could gain traffic from air-HSR integration. Nevertheless, Givoni and Chen (2017) analyzed whether the potential for airerail integration at Shanghai Hongqiao Integrated Transport Hub is achieved and found that the actual level of integration is low despite the high-quality infrastructure. The authors concluded that although the potential benefits from air and HST integration could be huge, there are important barriers that hamper real integration. In China, the main barrier is both institutional and cultural caused by the division of air transport and HST as well as the promotion of modal competition at any cost. The problem could be resolved with the creation of a truly integrated transport system in China. A number of attributes that could be attractive to passengers of the integrated air-HSR product such as travel time, price, baggage handling integration, fare ticket integration, timetable coordination, and creation of bimodal stations, among others, have been analyzed using stated-preference experiments (Brida et al., 2017; Román and Martín, 2014; Song et al., 2018). Thus, intermodal passengers should be included in the accessibility studies as a new transport mode. 4.3 ICT Van Wee (2016) contended that the use of Information and Communication Technologies (ICT) has changed dramatically the four accessibility components. Users can find restaurants and/or attractions to visit using their own smart devices. Similarly, it is easier than ever in several respects to ework, e-shop, e-learn, or e-medical visit, so the land use components will be highly affected in the future. Users can also employ ICT to find the best transport alternative, and most available apps are based on the integrated transport networks. The disutility associated with the components of waiting time and travel time has also been affected by the use of ICT because travelers can still work or be entertained during their trips. Additionally, satnav systems do not only reduce travel times but also travel time uncertainties associated to congestion that affects the reliability of transport systems. Thus, the static nature of accessibility analysis can be improved with a more dynamic approach.

308

The Air Transportation Industry

This dynamic approach has already been used in urban contexts (GarcíaAlbertos et al., 2019; Moya-Gómez et al., 2018). Both studies analyze the accessibility performance of the transport network as well as the attractiveness of the destinations. The main differences of the studies can be seen in the datasets used for the analysis as in the first paper the authors use mobile digital phones to construct origin-and-destinations travel matrices in which the travel times are obtained by the use of Google Maps API. Meanwhile, the second study uses the information provided by TomTom and Twitter, respectively. In both studies, urban accessibility profiles vary highly at different times of the day as each transport zone has its own peculiarities that needs time-specific solutions. Similarly to the studies commented above, dynamic accessibility in the case of air transport can be focused in measuring the attractiveness of different airports taking into account different seasons, days or even times of the day. Clearly not all airports have the same levels of activity during different seasons, days, or times of the day. It is also true that access/egress times to/from the airports also vary during the day in respond to the patterns of congestion and public transport service frequencies. Traditional measures have used static information that measures the relative attractiveness by population or GDP, but real market data based on mobile phone logs or social networks could be used to measure the relative attractiveness of some destinations. Thus, future studies need to include the temporal dimension to analyze a type of dynamic accessibility that could assess the evolution of door-to-door accessibility in relation to complementary transport modes used to access to the airport of origin and egress from the airport of destination. 4.4 Equity The concept of accessibility has been linked to issues such as equity, distributional effects, and social exclusion (van Wee, 2016). This aspect is related to the fourth component and normative use of the analysis. In the case of air transport, this is a relevant issue as an average citizen might not represent accurately the traveler population (Gössling et al., 2019). In developing countries, financial constraints and inadequate air transport development prevent large shares of the population from flying. Even in wealthy countries, there are many citizens who do not fly regularly. Besides the popularization of low-cost carriers, Banister (2018) found that the representativeness of any given sample of air travelers is still very limited. In fact, the author finds that 8% of those in the highest personal income group (>£65,000) fly a disproportionate amount.

The measurement of accessibility and connectivity in air transport networks

309

The legitimacy of aviation as a mode of transport has been questioned by the emissions caused, although the majority of air travelers are not highly concerned by the environmental externality. Alcock et al. (2017) studied whether people with concerns about the risks of climate change and pro-environmental concerns would also be more likely to abstain from air travel or at least fly less distances. The authors found no significant evidence in support of that hypothesis. Thus, they concluded that policies that aim to reduce discretionary flights cannot be based on increasing proenvironmental engagement. In this regard, accessibility indicators should include the environmental costs that are not internalized by the individuals. Thus, it would be possible to compare the gains in accessibility by income groups, and, if possible, by different types of trips (i.e., discretionary vs. mandatory). Since individuals and societies need to balance air transport accessibility with issues such as climate change, emissions reductions, and subsidies given to airlines, airports, or aircraft manufacturers (Gössling et al., 2017), the accessibility indicators should be sensitive to the relevant policies in order to evaluate whether the latter are progressive or regressive. 4.5 Big data and open sources Witlox (2015) discussed the recent ICT apps as valuable open sources that generate valuable data for accessibility studies. The apps provided by social media and service firms can generate data that can be geotagged and constitute the basis for what is known as “big data.” According to GarcíaAlbertos et al. (2019), examples of big data include geolocated mobile phone data, GPS data, smart card data from public transport, and social network data. Many accessibility studies have benefitted from the use of big data. One of the challenges of these applications is to analyze to what extent the dataset is representative of the population as young generations use smart devices more intensively than the average citizens. Moya-Gómez et al. (2018) made use of big data for the dynamic analysis of accessibility in Madrid. The authors used Twitter data as a proxy to estimate the degree of opportunities that exist in each transport zone of Madrid, and TomTom Speed Profiles data to calculate the travel time between each pair of zones. In both cases, a temporal dimension was considered for each day of the week and time bands of 15 min to calculate the dynamic accessibility indicators. Meanwhile, García-Albertos et al. (2019) proxied the destination attractiveness for each hour of the day using mobile phone data. The authors contended that mobile phone data present clear advantages, mainly in representativeness, over other sources such as Twitter.

310

The Air Transportation Industry

On the other hand, Sun et al. (2020a) made use of open-source databases to estimate the airport road access at a global scale. For example, the authors used Openstreetmap (a community project with the goal to create an editable map of the entire world); Open Source Routing Machine (a scalable method that calculates the travel time between two points at continental scale based on openstreetmap) and Gridded Population of the World (an open-source initiative that provides human population data worldwide). Air transport accessibility indicators studies do not abound by the lack of available data. For ground transport, the studies are mainly based on the two first components, namely land use proxied by population, GDP, jobs or more recently with Twitter or mobile phone data, and transportation costs proxied by travel times, generalized costs or a threshold figure that measures the interaction between zones. The temporal dimension has recently been included in studies of dynamic accessibility, and there exists an important gap with the individual characteristics of certain population groups. Sun et al. (2017) saw a need for the construction of an open-source dataset that includes a large-scale multimodal transport network. In our view, it is time to build a global dataset that can be complemented with different regional, national and supranational datasets from different organizations such as Official Airline Guide (OAG), International Civil Aviation Organization (ICAO), Airport Council International (ACI), Marketing Information Data Transfer (MIDT), and the national Civil Aviation Authorities among others. The open-source datasets should also be accompanied by opensource tools that calculate air transport accessibility indicators at different scales using the four components approach.

5. Conclusions This chapter analyses a number of related concepts in air transport such as accessibility, connectivity, resilience, criticality and vulnerability. We first analyze the four main components model (Where-How-When-Who) in the analysis of accessibility from general and classical studies to more specific air transport studies. We contend that air transport accessibility studies are scarce, and we provide an agenda for future studies in air transport extracted from van Wee (2016). The chapter has three main objectives: (1) to present the methodologies that have been applied in the main empirical studies that have analyzed air transport accessibility; (2) to provide a consistent review of air transport

The measurement of accessibility and connectivity in air transport networks

311

accessibility jointly with related concepts such as connectivity, resilience and vulnerability; and to provide a future research agenda analyzing the main gaps obtained in the literature. To our knowledge, there is only one paper (Hesse et al., 2013) that analyzed air transport accessibility for 82 major European cities. The analysis is based on four well known classical accessibility indicators: the potential, daily accessibility, location and relative network efficiency indicators. The literature related to other concepts such as connectivity, resilience, criticality and vulnerability is more abundant, but unfortunately, the methods are more based on complex network analysis, and the associated topological features of airline networks. Thus, there is still an ample room to analyze these concepts with accessibility indicators. We end the chapter with a tentative future research agenda that is based on four main pillars: (1) The role of short-distance air travel and greener modes; (2) Intermodality with ground transport modes, especially highspeed rail; (3) The role of information, communications, and technology; (4) Equity considerations as air transport in mainly used by the high-income segment; and (5) The development of new sources of information known as big data and new open sources that could allow for scalable analyses of the global multimodal transport network. It is clear that the above pillars are interrelated, so future research could also be developed based on a combination of them.

References Alcock, I., White, M.P., Taylor, T., Coldwell, D.F., Gribble, M.O., Evans, K.L., Corner, A., Vardoulakis, S., Fleming, L.E., 2017. ‘Green’on the ground but not in the air: pro-environmental attitudes are related to household behaviours but not discretionary air travel. Glob. Environ. Change 42, 136e147. Allroggen, F., Wittman, M.D., Malina, R., 2015. How air transport connects the world - a new metric of air connectivity and its evolution between 1990 and 2012. Transp. Res. E Logist. Transp. Rev. 80, 184e201. Antunes, A., Martini, G., Porta, F., Scotti, D., 2020. Air connectivity and spatial effects: regional differences in Europe. Reg. Stud. 54 (12), 1748e1760. Arvis, J.-F., Shepherd, B., 2011. The air connectivity index. Measuring integration in the global air transport network. In: The World Bank Policy Research Working Paper 5722. Banister, 2018. Inequality in Transport. Oxfordshire. Alexandrine Press. Bicchieri, C., Muldoon, R., Sontuoso, A., 2018. Social norms. In: Zalta, E.N. (Ed.), The Stanford Encyclopedia of Philosophy (Winter 2018 Edition). https://plato.stanford. edu/archives/win2018/entries/social-norms. (Accessed 18 October 2020). Bombelli, A., Santos, B.F., Tavasszy, L., 2020. Analysis of the air cargo transport network using a complex network theory perspective. Transp. Res. E Logist. Transp. Rev. 138, 101959.

312

The Air Transportation Industry

Borodako, K., Rudnicki, M., 2014. Transport accessibility in business travelea case study of Central and East European cities. Int. J. Tourism Res. 16 (2), 137e145. Brida, J.G., Martín, J.C., Román, C., Scuderi, R., 2017. Air and HST multimodal products. A segmentation analysis for policy makers. Network. Spatial Econ. 17 (3), 911e934. Bowen, J., 2000. Airline hubs in Southeast Asia: national economic development and nodal accessibility. J. Transp. Geogr. 8 (1), 25e41. Burghouwt, G., Redondi, R., 2013. Connectivity in air transport networks: an assessment of models and applications. J. Transp. Econ. Pol. 47 (1), 35e53. Caschili, S., Medda, F.R., Reggiani, A., 2015. Guest editorial: resilience of networks. Transp. Res. A Pol. Pract. 81, 1e3. Dai, L., Derudder, B., Liu, X., 2018. The evolving structure of the Southeast Asian air transport network through the lens of complex networks, 1979e2012. J. Transp. Geogr. 68, 67e77. European Commission, 2011. Flightpath 2050. Europe’s Vision for Aviation. Report of the High Level Group on Aviation Research. European Commission, Luxembourg. European Commission, 2015. Communication from the Commission: ‘An Aviation Strategy for Europe’. Brussels: COM/2015/0598 Final. Gallotti, R., Barthelemy, M., 2015. The multilayer temporal network of public transport in Great Britain. Sci. Data 2 (1), 1e8. García-Albertos, P., Picornell, M., Salas-Olmedo, M.H., Gutiérrez, J., 2019. Exploring the potential of mobile phone records and online route planners for dynamic accessibility analysis. Transp. Res. A 125, 294e307. Geurs, K.T., Van Wee, B., 2004. Accessibility evaluation of land-use and transport strategies: review and research directions. J. Transp. Geogr. 12 (2), 127e140. Geurs, K., Macharis, C., 2019. The future of European communication and transportation research: a research agenda. Region 6 (3), D1eD21. Giovannini, E., Benczur, P., Campolongo, F., Cariboni, J., Manca, A., 2020. Time for Transformative Resilience: The COVID-19 Emergency. EUR 30179 EN. Publications Office of the European Union, Luxembourg. Givoni, M., 2007. Environmental benefits from mode substitution e comparison of the environmental impact from aircraft and high-speed train operation. Int. J. Sustain. Transp. 1 (4), 209e230. Givoni, M., Chen, X., 2017. Airline and railway disintegration in China: the case of Shanghai Hongqiao Integrated Transport Hub. Transp. Lett. 9 (4), 202e214. Givoni, M., Banister, D., 2006. Airline and railway integration. Transp. Pol. 13, 386e397. Gössling, S., 2019. Celebrities, air travel, and social norms. Ann. Tourism Res. 79, 102775. Gössling, S., Fichert, F., Forsyth, P., 2017. Subsidies in aviation. Sustainability 9 (8), 1295. Gössling, S., Hanna, P., Higham, J., Cohen, S., Hopkins, D., 2019. Can we fly less? Evaluating the ‘necessity’ of air travel. J. Air Transp. Manag. 81, 101722. Gutiérrez, J., 2001. Location, economic potential and daily accessibility: an analysis of the accessibility impact of the high-speed line MadrideBarcelonaeFrench border. J. Transp. Geogr. 9 (4), 229e242. Handy, S., 2020. Is accessibility an idea whose time has finally come? Transp. Res. D Transp. Environ. 83, 102319. Hansen, W.G., 1959. How accessibility shapes land use. J. Am. Inst. Plan. 25 (2), 73e76. Hesse, C., Evangelinos, C., Bohne, S., 2013. Accessibility measures and flight schedules: an application to the European air transport. Eur. Transp. 55, 1e23. Jafino, B.A., Kwakkel, J., Verbraeck, A., 2020. Transport network criticality metrics: a comparative analysis and a guideline for selection. Transp. Rev. 40 (2), 241e264. Jenelius, E., Mattsson, L.-G., 2015. Road network vulnerability analysis: conceptualization, implementation and application. Comput. Environ. Urban Syst. 49, 136e147.

The measurement of accessibility and connectivity in air transport networks

313

Jia, T., Qin, K., Shan, J., 2014. An exploratory analysis on the evolution of the US airport network. Phys. Stat. Mech. Appl. 413, 266e279. Levine, J., 2020. A century of evolution of the accessibility concept. Transp. Res. D Transp. Environ. 83, 102309. Manca, A.R., Benczur, P., Giovannini, E., 2017. Building a Scientific Narrative towards a More Resilient EU Society. Part I: A Conceptual Framework. Luxemburg: Publications Office of the European Union. Martín, J.C., Reggiani, A., 2007. Recent methodological developments to measure spatial interaction: synthetic accessibility indices applied to high-speed train investments. Transp. Rev. 27 (5), 551e571. Martin, J.C., Gutiérrez, J., Román, C., 2004. Data envelopment analysis (DEA) index to measure the accessibility impacts of new infrastructure investments: the case of the highspeed train corridor Madrid-Barcelona-French border. Reg. Stud. 38 (6), 697e712. Martin, R., 2012. Regional economic resilience, hysteresis and recessionary shocks. J. Econ. Geogr. 12 (1), 1e32. Mattsson, L.-G., Jenelius, E., 2015. Vulnerability and resilience of transport systemsda discussion of recent research. Transp. Res. A Pol. Pract. 81, 16e34. Morris, J.M., Dumble, P.L., Wigan, M.R., 1979. Accessibility indicators for transport planning. Transp. Res. A Gen. 13 (2), 91e109. Moya-Gómez, B., Salas-Olmedo, M.H., García-Palomares, J.C., 2018. Dynamic accessibility using Big Data: the role of the changing conditions of network congestion and destination attractiveness. Network. Spatial Econ. 18, 273e290. Neal, Z., 2013. Evolution of the business air travel network in the US from 1993 to 2011: a descriptive analysis using AIRNET. Res. Transp. Bus. Manag. 9, 5e11. Östh, J., Reggiani, A., Galiazzo, G., 2015. Spatial economic resilience and accessibility: a joint perspective. Comput. Environ. Urban Syst. 49, 148e159. Páez, A., Scott, D.M., Morency, C., 2012. Measuring accessibility: positive and normative implementations of various accessibility indicators. J. Transp. Geogr. 25, 141e153. Reggiani, A., Nijkamp, P., Lanzi, D., 2015. Transport resilience and vulnerability: the role of connectivity. Transp. Res. A Pol. Pract. 81, 4e15. Rodríguez-Déniz, H., Suau-Sanchez, P., Voltes-Dorta, A., 2013. Classifying airports according to their hub dimensions: an application to the US domestic network. J. Transp. Geogr. 33, 188e195. Román, C., Martín, J.C., 2014. Integration of HSR and air transport: understanding passengers’ preferences. Transport. Res. E Logist. Transp. Rev. 71, 129e141. Song, F., Hess, S., Dekker, T., 2018. Accounting for the impact of variety-seeking: theory and application to HSR-air intermodality in China. J. Air Transp. Manag. 69, 99e111. Suau-Sanchez, P., Voltes-Dorta, A., Cugueró-Escofet, N., 2020. An early assessment of the impact of COVID-19 on air transport: just another crisis or the end of aviation as we know it? J. Transp. Geogr. 86, 102749. Suau-Sanchez, P., Voltes-Dorta, A., Rodríguez-Déniz, H., 2015. Regulatory airport classification in the US: the role of international markets. Transp. Pol. 37, 157e166. Sun, X., Wandelt, S., Hansen, M., 2020a. Airport road access at planet scale using population grid and openstreetmap. Network. Spatial Econ. 20 (1), 273e299. Sun, X., Wandelt, S., Zhang, A., 2020b. Resilience of cities towards airport disruptions at global scale. Res. Transp. Bus. Manag. 100452. Sun, X., Zhang, Y., Wandelt, S., 2017. Air transport versus high-speed rail: an overview and research agenda. J. Adv. Transp. 2017, 8426926. Van Wee, B., 2016. Accessible accessibility research challenges. J. Transp. Geogr. 51, 9e16.

314

The Air Transportation Industry

Voltes-Dorta, A., Rodríguez-Déniz, H., Suau-Sánchez, P., 2017. Vulnerability of the European air transport network to major airport closures from the perspective of passenger delays: ranking the most critical airports. Transport. Res. A 96, 119e145. Wang, J., Bonilla, D., Banister, D., 2016. Air deregulation in China and its impact on airline competition 1994e2012. J. Transp. Geogr. 50, 12e23. Witlox, F., 2015. Beyond the data smog? Transp. Rev. 35 (3), 245e249. Wong, A., Tan, S., Chandramouleeswaran, K.R., Tran, H.T., 2020. Data-driven analysis of resilience in airline networks. Transp. Res. E 143, 102068. Yamaguchi, K., 2007. Inter-regional air transport accessibility and macro-economic performance in Japan. Transp. Res. E Logist. Transp. Rev. 43 (3), 247e258. Zhang, A., Wan, Y., Yang, H., 2019. Impacts of high-speed rail on airlines, airports and regional economies: a survey of recent research. Transp. Pol. 81, A1eA19. Zhang, Y., Zhang, A., Zhu, Z., Wang, K., 2017. Connectivity at Chinese airports: the evolution and drivers. Transp. Res. A Pol. Pract. 103, 490e508. Zhu, Z., Zhang, A., Zhang, Y., Huang, Z., Xu, S., 2019. Measuring air connectivity between China and Australia. J. Transp. Geogr. 74, 359e370. Zweig, K.A., 2016. Network Analysis Literacy. Springer-Verlag GmbH Austria, Vienna.

CHAPTER 14

Fighting for market power: the case of Norwegian Airlines Siri P. Strandenes

Department of Economics, Norwegian School of Economics, Bergen, Norway

1. Introduction In Norway, air transport is more heavily used than in most European (and other) countries and has been so for a long time. The country is characterized by long distances, deep fjords and high mountains resulting in long travel times by car. In addition, the quality of the highways varies, and, in most areas, they are quite narrow with moderate speed limits. To illustrate, the flight OsloeTromsø takes two hours (1116 km) which is like going from Oslo to Frankfurt (1142 km). Even for shorter distances such as OsloeBergen (325 km) a 50-min flight, the alternatives’ travel times are seven hours by train and by car almost the same even though the ground distance is 462 km. Consequently, aviation is more important in domestic passenger transport in Norway than in most countries. The result is a high frequency of flights also in the domestic market. Following this, Norway has a very high number of commercial airports for a country with 5.37 million inhabitants. In 2021, Norway has 45 airports of which 44 are state owned. Local authorities own and operate the Sandefjord Airport Torp, while a private company leases the state-owned Haugesund airport. In 2019, the number of passengers in domestic flights reflects that every person living in Northern Norway travels six times a year, persons living in Mid and Western Norway two to four trips a year. These are the regions with the longest distances and few alternatives for traveling. People in the Oslo area travel twice per year on average. Overall, the 13 million domestic trips in 2019 imply 2.42 trips per inhabitant (Avinor, 2020). As a result, the main routes in the Norwegian domestic network are among the densest domestic routes in Europe. In the years up to deregulation in 1994, the structure of the Norwegian airline industry remained stable. Three airlines shared the market. The partly state-owned Scandinavian Airlines (SAS) (established in 1946) and The Air Transportation Industry ISBN 978-0-323-91522-9 https://doi.org/10.1016/B978-0-323-91522-9.00018-X

© 2022 Elsevier Inc. All rights reserved.

315

316

The Air Transportation Industry

the private Braathens (also established in 1946) served the primary airports in the domestic route network. The two airlines were of similar size in the Norwegian domestic market and operated separate routes. Widerøe (established in 1934) was partly owned by the two other and later fully owned by SAS until sold in 2013 and operates the short runway airport network with small aircraft feeding passengers into the primary network (NOU, 2019). Before deregulation, airlines needed approval by the authorities before changing prices or flight schedules and each route was a legal monopoly operated by one of the carriers. Deregulation at first did not bring about new entry, but the two incumbents serving the primary network entered each other’s densest routes and thus started competing as duopolists. The first new entry occurred four years later in 1998. In this chapter, we discuss why deregulation at first had limited effects on the airline structure in the Norwegian domestic market by investigating the industrial-economic consequences of the interplay of airport policies and airline strategies. The low-cost carrier Norwegian successfully entered the Norwegian domestic market in the autumn of 2002. Adding to its domestic growth, this airline after few years initiated a strong international expansion establishing subsidiaries abroad. In so doing, Norwegian chose a different strategy from the traditional full-service carriers, which has remained an industry of mainly national air carriers. In the next section, we analyze why deregulation took time to affect the Norwegian domestic airline market. The destabilization and fight for market power that eventually followed when deregulation took effect, is the topic of Section 3. Section 4 discusses competition between airports in the wider Oslo area. Section 5 analyzes the innovative strategy for international expansion following the successful entry into the domestic market by the low-cost carrier Norwegian. Section 6 offers concluding remarks.

2. Why did EU deregulation initially not affect the Norwegian domestic airline market? Norway deregulated domestic aviation in April 1994. Deregulation at first did not significantly affect competition there. The three incumbents, SAS and Braathens plus the regional airline Widerøe, kept their market shares. Why did Norway experience such stability in its domestic airline market at a time of large changes in aviation in Europe following EU airline deregulation?

Fighting for market power: the case of Norwegian Airlines

317

The main reason was the airport structure and capacity. Private air travel started in the interwar period mainly using seaplanes. Norway’s airports were mostly built during and after World War II, several of them initially for military use. After the war, the country established airports close to Oslo and to the larger cities along the coast, thereby replacing services by seaplanes and much of the passenger traffic by coastal steamers. The air services supplemented a small number of regional railway lines and the sparsely developed regional road network. The construction of inter-urban motorways between Oslo and the three larger cities in southern Norway; Bergen, Trondheim, and Stavanger proposed in 1962 was not completed. Instead, a close knit pattern of short take-off and landing airports (STOL) were built in western and northern Norway between 1968 and 1975 (NOU, 2019). These STOL airports enabled opening short haul routes served by small aircraft carrying 15e20 passengers. In 2021, 27 of the 45 Norwegian commercial airports are STOL airports (Avinor, 2020). The high number of airports, the long distances, and the small but geographically dispersed population plus the limited availability of efficient surface transport alternatives is the main reason for the high frequency of flying among Norwegians. The Oslo airport is the hub airport in the primary domestic routes network and the routes from the STOL airports on the west coast and in northern Norway feed passengers into this domestic hub and spoke primary routes network. Furthermore, most international routes go to and from Oslo. Most Norwegian passengers traveling abroad thus transit via Oslo and or Copenhagen, the latter being the hub airport in the SAS international route network. SAS before regulation was the only Norwegian airline operating international scheduled flights (NOU, 2019). When assigning airport slots the Norwegian aviation authority follows the traditional policy of grandfather rights. As Oslo Airport Fornebu became crowded following the rise in frequencies of flights by the incumbent carriers after deregulation, potential entrants could not acquire slots. Deregulation induced the two incumbent airlines SAS and Braathens to compete for market shares. They formed duopolies in the denser routes in the primary routes network and competed in capacities. Salvanes et al. (2005) also found that deregulation led to a clustering of flights in the duopoly routes compared to flights in the thinner routes that remained a monopoly for either of the two incumbents. The opening of the new and larger Oslo Airport Gardermoen in the autumn of 1998 lifted the capacity problems at the domestic hub and the

318

The Air Transportation Industry

first aircraft to depart from the new airport was a flight from the new upstart airline Color Air. From the start Color Air operated three jet aircraft and introduced routes from the new Oslo Airport Gardermoen to Bergen, Trondheim, and the smaller city Ålesund, which was the hometown of the owner (Lian et al., 2002). The main reason for the delayed entry of new airlines into the large and growing Norwegian domestic market was the capacity problems at the existing hub Oslo Airport Fornebu. No new slots were available for any start-up until after the airport moved from the old, cramped airport west of Oslo to the new larger Gardermoen airport north of Oslo. The capacity problems at the old Oslo Airport Fornebu affected airline competition more severely than capacity problems at other domestic airports since Oslo is the national hub airport. The delayed effects of airline deregulation described above, illustrate the importance of adequate airport capacities for well-functioning airline markets and furthermore point at the high importance of sufficient capacity at the hub airport. Lack of free capacity thus is an efficient barrier to new entry and thereby functions as a protection for the incumbent airline or airlines (Strandenes, 2004a).

3. Phases in airline strategic behavior following the deregulation In many ways, developments in the Norwegian domestic airline industry upon deregulation offer a laboratory on industrial economics. In the sixyear period after deregulation the market structure in the primary routes network switched from a stable duopoly formed by the two incumbents, to a brief year of intense competitive among three airlines one of them a new entrant, followed by the two incumbents continued competition which ended in a monopoly after the one incumbent acquired the other. This monopoly period with reduced supply and sharply rising fares induced another new entry into the market. In this section we analyze the effects of deregulation focusing on the shifts in the airline’s strategies over time. 3.1 The initial period 1994e1998 Before 1987, regulatory authorities divided the primary routes between the two airlines each having a monopoly in its specific routes (NOU, 2019). The two incumbent airlines SAS and Braathens were of similar size. After 1987, but before domestic deregulation, these two airlines operated as a

Fighting for market power: the case of Norwegian Airlines

319

regulated duopoly in most routes since the regulators allowed each airline a few flights in the other’s former monopoly routes. This increased the airlines knowledge of their competitor’s operations. An interlining system, that is, a system for coordination of prices for full fare flexible tickets allowed passengers a late change of flights from one airline to the other. To coordinate prices on these tickets, the airlines could consult on future prices. As a result, the two incumbents knew each other well when domestic market deregulation came about in 1994. In addition to the postponed new entry caused by the lack of slots discussed above, both the interlining that allowed consultation on prices and the similar size of the two incumbents contributed to reducing the effects of the initial deregulation in the Norwegian domestic market. Any price reductions in this period mainly resulted from allocating more seats in each flight to the inflexible mini-price tickets (Steen and Sørgard, 2002). Following the EU deregulation packages foreign EU/EEA airlines might enter the Norwegian domestic markets from 1997. Thus, free cabotage was introduced (NOU, 2019). Even so, no foreign airline entered the Norwegian domestic routes which were and still are some of the densest domestic routes in Europe. Nonetheless deregulation had effects. Competition for market shares between the two Norwegian incumbents started in the primary routes network. The competitive setting resembles a prisoner’s dilemma. Each airline will gain by increasing their frequencies if the other airline does not answer by increasing theirs. When both increase frequencies, both will loose from lower load factors. This is so unless they adjust fares downward to increase total demand for air travel. Hence, both will be better off by not increasing the frequencies and thereby keeping a higher load factor. Both airlines increased the capacity offered, and both faced reduced the load factors. For both airlines, the load factor stood between 50% and 55% in 1999 and lower than their load factors around 60 in 1992 (Lian et al., 2002). As Salvanes et al. (2005) argued, deregulation also led to a clustering of flights on the new duopoly routes compared to the spacing of flights in those domestic routes that remained monopoly routes for either of the two incumbents. By clustering the departures, the airlines tried to increase their market share at the cost of the other. Allowing the airlines to set the fare structure, frequencies of flights and departure times resulted in a clustering of flights in the duopoly routes in line with Hotelling competition (Hotelling, 1929). Consequently, the first period after deregulation 1994e98 saw capacity competition, but little effect on prices and no new entry. The market shares

320

The Air Transportation Industry

of the two incumbents remained stable (Steen and Sørgard, 2002). This stability was to change, however, after 1998 when the capacity in the hub airport increased, thereby allowing for entry. 3.2 Second phase 1998e2002 When the new Oslo Airport Gardermoen opened in October 1998, the newly established airline Color Air started serving three domestic routes: OsloeBergen, OsloeTrondheim and OsloeÅlesund, operating three aircraft. Both Braathens and SAS operated the routes to Bergen and Trondheim and Braathens also served the OsloeÅlesund route. For one year, until Color Air withdrew from the market after heavy losses, the frequencies increased markedly in these routes and SAS entered the OsloeÅlesund routeea rather thin route reflecting the size of Ålesund at close to 38,000 inhabitants in 1998. The number direct of flights (one way) per day in this route increased from 6 to 17 daily departures from 1996 to 1999 but fell back again to 11 daily departures in 2001, after Color Air withdrew. In the denser routes OsloeBergen and OsloeStavanger departures increased from 27 (25) to 32 (32) respectively to fall back to 29 (26) in 2001. The strongest rise in the number of departures came in the OsloeTrondheim route, where departures rose from 28 in 1996 to 39 daily departures in 1999. In 2001, this route had 31 daily departures (Lian et al., 2002). Even though there was some reduction in frequencies as Color Air left the market, Lian et al. (2002) points out that the contraction in frequencies was slow as neither SAS nor Braathens accepted a reduction market share. During the months the new entrant stayed in the market, that is, until late 1999, the index of airfares remained in line with the general consumer price index for Norway both being around 103 from a basis of 100 in 1998. The index reflects developments in airfares for the main fare groups; flexible and inflexible tickets in the Norwegian market, based on the denser domestic routes and two intra-Nordic flights out of Oslo ( Johansen, 2007). Even though this airfare index is not only influenced by changing competition in the domestic market, we shall see that changes in the competitive setting clearly affect the changes in the index. One reason being the high weight of domestic to international air travel in Norway following the high average number of flights per person per year. The overall fare stability changed when Color Air went bankrupt and withdrew from the market. The incumbents initially reduced their capacities

Fighting for market power: the case of Norwegian Airlines

321

(Steen and Sørgard, 2002) and the market saw a steep rise in the price index for airfares up to 118e120 relative to the overall consumer price index rising from 105 to 110 the following year. After the initial competition against the newcomer, the two incumbents returned to competing as duopolists for market shares in the primary domestic routes, but the competition intensified compared to the situation before the opening of the new Oslo airport and Color Air’s entry. The developments led to a destabilization between the two incumbents. The situation also escalated by SAS and Braathens’ intensified competition for large customer contracts. The fare reducing effect from such contracts is not visible in the airfares index since the index reflects the listed fares. The number of contracts and the discounts offered increased markedly just before the opening of the new larger airport. The contracts had strict secrecy clauses, but it has been assumed that the large customer contracts, which were negotiated with both private firms and state institutions for example universities, might offer discounts up to 50% of the listed fares for flexible tickets (Lian et al., 2002). The contracts pressed the airlines’ profitability, even though they pressed up listed prices for business class tickets to smaller customers. The large customer contracts together with the frequent flyer programs that both airlines offered also hampered competition by locking in customers (Steen and Sørgard, 2002). The pressure on profitability was further deepened by the dot.com crises from 2000 and the resulting 2e3-year dip in passenger growth. On top of this the air passenger tax levied since 1994 was adjusted upwards in 2000 and extended to more domestic routes in 2001. The competition for market shares that induced the high frequency of flights in the primary domestic routes drained the competing airlines’ resources. Whereas SAS partly compensated the loss by income earned in the intra Scandinavian market and from their international operations, Braathens depended more strongly on the Norwegian domestic market. Braathens had also challenged SAS on the OsloeStockholm route and experienced big losses from this and from acquisitions (Steen and Sørgard, 2002). Competition law do not allow two duopoly firms to merge and form a monopoly, except when one of the firms is found to be a so-called failing company. By the end of 2001, the Norwegian competition authority accepted that the losses faced by Braathens implied that the airline had become a failing firm during the autumn of 2001 (Konkurransetilsynet, 2001). Thus, SAS could buy its competitor in the Norwegian domestic market. The immediate effect was another hike in the airfares index to a

322

The Air Transportation Industry

maximum of 158 in the first quarter of 2002 at a time when the consumer price index reached 110. This steep rise in airfares followed the strong reduction in the number of seats and flights offered by the merged SAS Braathens. 3.3 Third phase 2002 onward Early 2002, some months after SAS acquired Braathens, Norwegian Air Shuttle (NAS), which since 1993 had operated smaller aircraft as a supplier into Braathens’ routes network, announced that they planned to enter the primary domestic market under the brand name Norwegian the same autumn. From the start, Norwegian served four of the denser domestic routes. The airfare index fell by 25% upon Norwegian’s entry. This impact on fares resembles the fare effect estimated in a theoretical assessment of potential fare reductions in the OsloeStockholm route if one airline were to enter this traditional SAS monopoly route. This intra-Scandinavian route resembles the denser routes in southern Norway. Norman and Strandenes (1994) simulated the effect of entry into this SAS monopoly route and found, based on fares, prices, and costs at the 1990 level, that the simulation model indicated fare reductions of just below 30%. After entering Norwegian grew fast in the domestic market in Norway. In the first year after entering the domestic market the airline expanded into the European market (Norwegian, 2004), and within 11 years the airline entered intercontinental routes (Norwegian, 2013). Why did Norwegian’s start up succeed so much better than Color Air, which entered just four years earlier? When Norwegian entered in 2002, airline markets were depressed both from the dot.com bubble and the effects of 9/11. Worldwide several aircraft were grounded. As a result, the leasing market for aircraft was favorable to a new entrant. This differed from the leasing cost Color Air met in 1998, when the airline markets were strong. Furthermore, SAS0 acquisition of Braathens and the consolidation process starting thereafter meant that experienced pilots and crews laid off by the merged SAS Braathens were available to Norwegian, which hired several. Norwegian’s start up strategy was more effective than the one used by Color Air four years earlier. For the first months, Norwegian operated four aircraft in four domestic routes: Oslo-Stavanger, OsloeBergen, OsloeTrondheim and OsloeTromsø. To be able to offer business passengers acceptable frequencies of flights during morning and afternoon

Fighting for market power: the case of Norwegian Airlines

323

hours, Norwegian parked their aircraft at the gates for several hours midday irrespective of the cost from parking at a gate and the less favorable working hours for crews. This contrasted the strategy chosen by Color Air which instead spread their flights evenly over the day and thereby became less attractive for business travelers since flexible full fare tickets did not offer much flexibility with such limited frequency of flights during morning and afternoon. Norwegian chose to follow a low-cost carrier strategy. They offered no service on board and cut the cost of ticketing and boarding by introducing barcodes, which allowed passengers to book online and print the combined ticket/boarding card at home. They had only one class in the cabin and as other low-cost carriers introduced one-way tickets also for low fare inflexible tickets. The cost per available seat-km for Norwegian was approximately 80% of its domestic competitor SAS cost in 2003e2004 (Norwegian, 2005) (SAS, 2004). The airline also introduced charges for extra services as fast track, extra luggage and offered no free inflight service. Norwegian’s market shares in the routes where they competed head on with SAS, increased strongly after the start up in 2002, but the company early on chose to expand in international routes and the market share in the domestic market has stayed somewhat below SAS’s market share in the Norwegian domestic market (NOU, 2019). Thus, the newcomer did not challenge SAS as the largest operator at home. The policy of one-way economy tickets was followed up by SAS for their summer program in 2004. Thereby SAS broke with the traditional pricing policy for network airlines by only discounting round-trip tickets. This pressed the airfares down another 15% to approximately the level as the general consumer price index at the time. The strong dip in the air fares in the summer 2004, bringing the airfares index 10e15 points below the general consumer price index for the summer program, followed from SAS summer campaign offering very low campaign prices. Norwegian’s expansion was helped by the government’s prohibition of frequent flyer programs in the Norwegian domestic market from 2002. This ban lasted until 2013 and reduced passengers’ gain from customer loyalty. In addition, the new entrant won the large customer contract with the Norwegian state in 2002 (NOU, 2019). This increased the demand Norwegian faced in the business segment, even though at discounted prices. In the years after entering Norwegian’s routes network spread fast from the first four domestic routes in Norway. Norwegian very soon opened

324

The Air Transportation Industry

more domestic flights mainly out of Oslo, plus direct international flights from large regional airports such as Bergen, Stavanger, and Trondheim. The airline typically established routes that did not compete head on with SAS international flights but introduced direct flights as alternatives to flights via Oslo or Copenhagen to the final foreign destination, in this way repeating the initial strategy for growth by increasing demand from new passenger groups instead of competing for market share in the incumbent SAS main markets.

4. Airport competition The new hub airport’s location 45 km north of Oslo imposed longer travel time to the airport for most people and businesses mainly located in or to the west and south of Oslo, while the north was more sparsely populated. This location was disputed for many years, but the policy objective to use the airport as a trigger for economic growth north of Oslo was decisive for its location. To compensate for the longer travel distance for almost everyone, a fast speed airport express train started operating parallel to the opening of the new airport. Oslo Airport Fornebu locked down as the new airport opened and therefore did not influence the airport choice by airlines or passengers. But the location of the new hub airport improved the relative position of the smaller Torp Sandefjord Airport located south-west of Oslo. Its local market increased markedly. There was a suggestion to expropriate this airport and put it under the national airport authority’s control to hinder competition with the new Oslo Airport Gardermoen. Arguing for expropriation proved politically difficult though, since local and regional authorities in the area owned and operated Trop Sandefjord Airport after having bought out the private airport owners some years earlier. One result being that Oslo Airport Gardermoen from the start faced competition from Torp Sandefjord Airport. As we shall see next, Torp Sandefjord Airport came to influence the competition between airlines for international flights. Ryanair established themselves at Torp Sandefjord airport in 1997, the year before the new hub airport opened at Gardermoen. The airport is called Oslo Torp in Ryanair’s booking site even though the distance to Oslo is 116 km. The state-owned company Avinor operates Norwegian airports with a few exceptions. One effect is a unity in the airport charge structure. As Ryanair’s strategy is to operate from low-cost airports,

Fighting for market power: the case of Norwegian Airlines

325

Avinor’s tariff policy delayed Ryanair’s entry into the market for international flights to and from Norway. Ryanair expanded at Torp and before the COVID-19 disruption operated several routes to and from Torp, competing with routes operated by SAS and other airlines from Gardermoen. In the following years, the competition from Torp Sandefjord Airport intensified as other airlines; KLM, Norwegian and Wizz Air established routes there. Several other airlines have opened seasonal routes to and from Torp. Widerøe, the Norwegian airline mainly operating the STOL airports in western and northern Norway operated at Torp Sandefjord Airport since several years when Ryanair entered the airport and feeds passengers into the international routes. One year ahead of the move to Gardermoen, commercial interest southeast of Oslo suggested opening the military airport Rygge Moss Airport located 64 km southeast of Oslo for commercial traffic. In 2004, Rygge obtained concession to operate non-military aviation and commercial operations started in 2007. As a result Oslo Airport Gardermoen faced yet another competitor nine years after opening. The new competitor was located closer to the populous areas south east of Oslo. Norwegian established a base in 2008 and opened international routes out of Rygge plus domestic routes to feed passengers from west and north Norway into these international flights. To limit competition with Oslo Airport Gardermoen the government had fixed, however, a max number of 774,000 passengers and 15,000 flights allowed at Rygge. The passenger limit soon became a binding restriction. A political attempt to remove the passenger limit in 2008 at first was rejected by the parliament, but the following year the government removed the passenger limit but kept the limit of 15,000 aircraft movements per year. Ryanair in 2009 entered Rygge with several international flights and established its base there in 2010 at the same time reducing its presence at Torp. In the following years, both Torp and Rygge competed with Gardermoen, especially for the holiday market travelers. The airports at each side of the Oslo fjord Torp and Rygge also competed. Norwegian reduced the number of flights to and from Rygge after Ryanair moved in and left the airport in January 2012. Did the policy to limit the competition faced by Gardermoen Oslo Airport and thus Avinor, only aim at securing the new hub airport a high market share? One additional policy aim was to secure Avinor’s ability to cross subsidize between non-profitable and profitable airports. The hub airport together with the other larger airports, e.g., Bergen, Trondheim,

326

The Air Transportation Industry

and Stavanger create a surplus that cover losses from the 40 other airports operated by Avinor, many of them small but essential to communication in many remote localities in western and northern Norway. Ryanair in two instances played against Avinor. First, when entering the Norwegian holiday market and initiating negotiations with Avinor for lower airport charges. This proved unsuccessful at Gardermoen, but Ryanair accepted conditions obtained at Torp, which could set airport charges independent of Avinor policies since it is owned by local authorities and not by Avinor. Second, when seeking access to western Norway. Ryanair opened routes to and from western Norway in 2003 operating at Haugesund airport. Ryanair had negotiated with Avinor for flights from Bergen, Trondheim, and Stavanger, but failed to obtain airport charges they were willing to accept. At the same time, there was a political pressure in west Norway for access to low-cost flights like those offered passengers in east Norway. Lowering charges at these three profitable airports, would weaken Avinor’s financing of the whole airport network in Norway. Reduced charges for Ryanair would have a strong negative effect since airports are not allowed to charge different airlines different fees. This rule originally was introduced to secure that countries do not favor their flag carrier. If Avinor were to reduce airport charges on the relatively few Ryanair flights, they consequently also had to reduce airport charges for all domestic and international flights by other airlines and this would disrupt the cross-subsidization scheme in Norwegian airport financing. Following the political pressure from west Norway, the government decided to reduce airport charges by 90% for all airlines operating at the smaller airport Haugesund. Ryanair started operating international flights to and from this airport in 2003. Airport express busses were set up from the larger cities Bergen north of and Stavanger south of Haugesund (Strandenes, 2004b). When the airport discount was removed at Haugesund airport in 2006, local businesses in the Haugesund area established a support scheme to compensate Ryanair for the extra costs. When Ryanair made Rygge Moss Airport their main base in Norway in 2010, a reason for shifting from Torp to Rygge probably was the closer proximity to Oslo. When the Norwegian government announced the plan to introduce a passenger tax from April 2016 Ryanair announced that they would exit Rygge. The passenger tax was decided by the Parliament and Ryanair closed its base and terminated all flights to and from Rygge, concentrating again on Torp for the east Norwegian market opening only a few routes from Oslo Airport Gardermoen. Following Ryanair’s exit

Fighting for market power: the case of Norwegian Airlines

327

Rygge Moss Airport closed all commercial traffic from November 1, 2016. Before closing Ryanair operated 15 international routes out of Rygge on an annual basis plus extra 20 seasonal routes. Norwegian operated three international routes to and from Rygge. These and domestic feeder flights into the airport terminated when the airport closed to commercial air transport. The Norwegian case offers an example of airport competition and the links between airport capacities and policies and competition among airlines. This case illustrates that a smaller airport, Torp Sandefjord Airport may expand by attracting low-cost carriers and focusing on the holiday market even when serving the same market area as a new larger national hub airport. One important fact is that the two competing airports does not belong to the same owner (Oum and Fu, 2008). Torp was harder hit by competition from a second private airport targeting the same passenger groups and with a closer location to large passenger groups as illustrated by the opening of Rygge Moss Airport. When this airport was closed, not because it lost out in the competition, but following a policy decision on a passenger tax Torp Sandefjord Airport regained its position in the holiday market in eastern Norway.

5. The low-cost carrier Norwegian’s continued growth strategy Norwegian’s entry in the autumn of 2002 converted the Norwegian domestic market back to a duopoly now between the large incumbent SAS Braathens and the entrant Norwegian. In 2003 after some months in operation, Norwegian expanded domestically and in addition added some international routes. They opened international routes from Norway to holiday destinations in Spain and Portugal and by the end of the year operated 18 routes of which 6 were routes to foreign destinations in Europe (Norwegian, 2004). At first, SAS Braathens did not challenge Norwegian. Both airlines expanded in the domestic market as a main effect of the lowcost carrier’s entry was to expand the passenger base in the domestic market by low fares that opened aviation to new passenger groups. Thus, overall SAS Braathens experienced growth in the domestic market even as Norwegian’s market share increased (Avinor, 2020). Similarly, Norwegian’s strategy was to open direct scheduled flights to tourist destinations in Europe. As a result, more people traveled by air for private and family visits and new groups started going abroad for holidays.

328

The Air Transportation Industry

Early on, Norwegian stated a growth strategy necessary for profitable operations and that such growth was not attainable in the domestic market that the airline shared with SAS Braathens. Implementing this strategy, the airline established its first international base in 2006 in Warsaw (Norwegian, 2007). Why Warsaw? EU in 2004 opened to the east accepting Poland and other East European countries as members. This expansion widened differences in wage levels within EU and the common airline market. This meant that airlines based in the lower labor costs East European countries might operate all over EU/EEA. The airline industry furthermore expected East Europeans to experience a rising welfare level and thereby represent a growing market for private, family and holiday travel. Norwegian established five routes out of Warsaw to holiday destinations in the Mediterranean. In addition, the airline opened seven domestic Polish routes to and from Warsaw. This initial foreign base did not end up a commercial success though and was closed in 2010 after five years operation. An important cause being the economic downturn following the financial crises in 2008. For their Polish flights to and from Warsaw, Norwegian operated based on their ordinary Aircraft Operating Certificate (AOC) issued by Norwegian authorities to the mother company in Oslo. Following the international agreements on airline operations established by the Chicagoconvention on international Civil Aviation in 1944, every airline needs a national certificate to operate commercial flights. At first, one monopoly flag carrier served most country’s aviation market. An exception to this rule was SAS, which since its establishment in 1946, operates based on a Scandinavian AOC. SAS was owned partly by the governments in Denmark, Norway, and Sweden and partly by private owners in these countries. It operated as the Scandinavian flag carrier. The Chicago convention furthermore regulates international flights via bilateral agreements creating duopolies between the flag carriers from each nation. The airlines shared bilateral route revenue 50/50. Deregulation of the airline industry started in the United States in 1978 and was carried out in sequential steps toward a fully deregulated domestic market (Bailey et al., 1985). EU/EEA established an internal common airline market as part of the deregulation. This implied that the former international EU/EEA markets among the member countries converted into one large domestic market for carriers registered in an EU/EEA country. International markets outside the EU/EEA continue as duopoly markets regulated by the Chicago rules, except for those international routes opened to the signatories of the so-called Open Skies agreements.

Fighting for market power: the case of Norwegian Airlines

329

The first Open Skies agreement was negotiated by the United States and the Netherlands in 1992. Open Skies agreements replace the traditional bilateral agreements based on the Chicago convention, open routes between the two countries to several airlines from each country and stop regulating the market shares of the individual airlines operating between the countries. The Open Skies agreements, however, do not open the domestic markets in each of the signatory countries to airlines from the other. Consequently, when EU/EEA negotiated an Open Skies agreement with the United States, this resulted in an imbalance in market access between US and EU/EEA airlines. European airlines when flying to the US, cannot continue to another US destination since they then will enter the US domestic market. An US airline flying to, e.g., London, on the other hand, may continue to an airport in another EU/EEA country as this still is regarded as an international flight since the EU/EEA aviation markets remain separate for airlines registered outside EU/EEA. Norwegian continued its European expansion and in 2007 bought the airline FlyNordic that operated on a Swedish AOC. This airline in 2008 was renamed Norwegian Air Sweden but kept the Swedish AOC. In 2013, they also established a subsidiary in Ireland Norwegian Air International, operating on an Irish AOC. They also added an operating company in Great Britain named Norwegian Air UK and bases for pilots and crews in Great Britain and in other European countries plus in Thailand and the United States (Norwegian, 2019). These expansions show that Norwegian went from focusing on the domestic Norwegian market and international flights to and from Norway to an airline holding AOC’s in four European countries. The company thereby converted into a multinational European airline challenging the traditional structure in the airline industry established by the Chicago convention in 1944, which had remained the default for the industry for several decades. In 2013, Norwegian opened intercontinental flights as one of the first low-cost carriers to offer long-distance intercontinental flights (NOU, 2019), which primarily had been left to the network airlines and their alliances. The company opened routes to Bangkok and New York, and later expanded the network to other destinations in the United States and South America. As Norwegian expanded beyond the European market, they challenged the typical profile chosen by low-cost carriers. By the end of 2017, Norwegian operated 64 intercontinental routes (Norwegian, 2018). They established crew bases in Bangkok and the United States with crews on local wages and working conditions. This caused problems with

330

The Air Transportation Industry

labor unions both in Europe and in the United States. When starting flights to the United States, Norwegian first had to operate the routes based on its Norwegian AOC with Norwegian crews. Following a court decision, the airline could start operating flights to the United States based on a foreign AOC. The British subsidiary acquired a so-called foreign air carrier permit and thereby the right to serve different destination in the United States. The Norwegian mother company and the Swedish daughter also holds such permits. By 2018, Norwegian had established crew bases in New York, Fort Lauderdale, Bangkok, and London for its long-haul flights. “Recruitment to new bases takes place locally, at competitive local wages and benefits” (Norwegian, 2018). National crews still are the tradition for full-service network airlines on intercontinental flights. They may include a few foreigners among the cabin crew to offer passengers from the foreign country service by persons knowing their language and culture, however. The last step in the global expansion by Norwegian was the 2017 launch of Norwegian Air Argentina with an Argentinian AOC giving them the right to operate domestic flights in Argentina and international flights out of Argentina to other countries in South America and to Europe. Operations started in 2018. The Argentinian subsidiary was, however, sold to JetSMART airlines in 2019 (Norwegian, 2020). Norwegian was atypical when introducing intercontinental routes in their network. By this policy they challenged the traditional network carriers. The large low-cost carriers Ryanair and easyJet did not follow Norwegian into adding intercontinental routes to their operations, but easyJet started feeding passengers into Norwegian’s intercontinental routes. By establishing foreign subsidiaries holding their own AOC Norwegian started on the road to becoming a multinational company, not merely a national carrier facilitating economic globalization by operating intercontinental routes.

6. Concluding remarks By studying developments in Norwegian aviation after deregulation in 1994 and onwards the importance of airport capacity for airline competition becomes clear. We have seen that expected effects from deregulating Norwegian aviation were delayed almost four years by the congested hub airport Oslo Airport Fornebu. In this period the two incumbents SAS and Braathens

Fighting for market power: the case of Norwegian Airlines

331

who knew each other’s policies well after having operated the primary domestic network together for many years, competed for market shares in a stable duopoly with two airlines of similar size. Their capacity competition soon was limited by the capacity restrictions at the hub airport. When the new and significantly larger Oslo Airport Gardermoen opened in the October 1998, a new airline simultaneously entered. The two incumbents increased their frequencies in the four routes Color Line entered. Failing, the new airline closed after one year, but the entry has had lasting effects. It destabilized the competition between the former duopolists and made them continue the fight for markets shares. Fares rose as price competition concentrated on large customer contracts which also increase customer loyalty. The intensified competition between the two incumbents resulted in Braathens becoming a failing company and the competition authority allowing SAS to acquire Braathens in 2001. This established a domestic monopoly which as should be expected, lead to a sharp reduction in frequencies and a hike in airfares. Combined with the available capacity at the hub airport this situation was an invitation to new entrants. Government policies strengthened the case for entry by prohibiting frequent flyer programs in the domestic market, thereby reducing passenger loyalty for the incumbents and by awarding the newcomer a large customer agreement with the state. The Norwegian case also illustrates the potential for airport competition from different ownership and management for airports serving overlapping passenger locations. Since the new hub airport was located further away from most passengers compared to the one it replaced, the attractiveness of the existing smaller airport in the region increased, especially so for holiday travelers. Torp Airport Sandefjord was owned by the local authorities and free to negotiate lower airport charges than those set by the state owned airport operator Avinor, and thus illustrates the link between airport policies and airline competition. Finally, the growth strategy followed by the new low-cost carrier Norwegian attracted demand from new groups of passengers, that is, private and holiday travelers and thus enabled a passenger growth both for the new and the incumbent airline. The same has been observed in other markets where low-cost carriers entered. Focusing on continued growth, Norwegian in 2013 entered the intercontinental market to offer long distance flights and pursuing the same growth strategy of attracting new passenger groups also to long-distance markets. Even more interesting, Norwegian broke the tradition of one national AOC only by establishing subsidiaries

332

The Air Transportation Industry

abroad operating on foreign AOCs, thereby taking a first step, currently halted by COVID-19, toward establishing multinational airlines in this industry that has been so important for globalizing the economy.

References Avinor, 2020. Reisevaner på fly 2019 (Travel by Air 2019). Avinor, Rapport, Oslo, 20pp., (In Norwegian). Bailey, E.E., Graham, D.R., Kaplan, D.P., 1985. Deregulating the Airlines. The MIT Press, Cambridge, Massachusetts. Hotelling, H., 1929. Stability in competition. Econ. J. 39 (153), 41e57. Johansen, I., 2007. Ny metode for flyreiser i KPI (New Method for Assessing Airfares in the Consumer Price Index), Statistics Norway, Mimeo 2007/40 (In Norwegian). Konkurransetilsynet, 2001. SAS’ erverv av 68,8 prosent av aksjene i Braathens ASA (In Norwegian) (Last accessed 20210227). https://konkurransetilsynet.no/wp-content/ uploads/2018/08/a2001-21.pdf. Lian, J.I., Eriksen, K.S., Lauridsen, H., Rideng, A., 2002. Norsk Innenlandsk luftfarts e konkurranse og monopol (Norwegian Domestic Aviation e Competition and Monopoly), Oslo, Transportøkonomisk Institutt. TØI-Report 586/2002 (In Norwegian). Norman, V.D., Strandenes, S.P., 1994. Deregulation of Scandinavian airlines: a case Study of the Oslo-Stockholm route. In: Krugman, P., Smith, A. (Eds.), Empirical Studies of Strategic Trade Policy. The University of Chicago Press, Chicago, pp. 85e100. Norwegian, 2004. Annual Report 2003 (Last accessed 20210220). https://www.norwegian. no/globalassets/ip/documents/about-us/company/investor-relations/reports-and-presentations/annual-reports/norwegian-annual-report-2003.pdf. Norwegian, 2005. 4Q Presentation 2004 (Last accessed 20201220). https://www. norwegian.no/globalassets/ip/documents/about-us/company/investor-relations/ reports-and-presentations/quarterly-results/norwegian-2004-q4-presentation.pdf. Norwegian, 2007. Annual Report 2006 (Last accessed 20201227). https://www.norwegian. no/globalassets/ip/documents/about-us/company/investor-relations/reports-and-presentations/annual-reports/norwegian-annual-report-2006.pdf. Norwegian, 2013. Annual Report 2012 (Last accessed 20210220). https://www.norwegian. no/globalassets/ip/documents/about-us/company/investor-relations/reports-and-presentations/annual-reports/norwegian-as-asa-annual-report-2012.pdf. Norwegian, 2018. Prospect 2018 (Last accessed 20201121). https://www.norwegian.no/ globalassets/ip/documents/investor-relations/prospekt-norwegian-air-shuttle-asa.pdf. Norwegian, 2019. Annual Report 2018 (Last accessed 20201121). https://www.norwegian. no/globalassets/ip/documents/about-us/company/investor-relations/reports-and-presentations/annual-reports/annual-report-norwegian-2018.pdf. Norwegian, 2020. Annual Report 2019 (Last accessed 20201121). https://www.norwegian. no/globalassets/ip/documents/about-us/company/investor-relations/reports-and-presentations/annual-reports/annual-report-norwegian-2019.pdf. NOU, 2019. Fra statussymbol til allemannseie e norsk luftfart i forandring (From Status Symbol to Commonly Usede Changes in the Norwegian Aviation Industry), Norwegian Official Report 2019, # 22 (In Norwegian), (Last accessed 20201227). https:// www.regjeringen.no/no/dokumenter/nou-2019-22/id2680751/?ch¼5. Oum, T., Fu, X., 2008. Impacts of Airports on Airline Competition. Focus on Airport Performance and Airport -Airline Vertical Relations, Paris, OECD/ITF. Discussion Paper 2008e17.

Fighting for market power: the case of Norwegian Airlines

333

Salvanes, K.G., Steen, F., Sørgard, L., 2005. Hotelling in the air? Flight departures in Norway. Reg. Sci. Urban Econ. 35, 193e213. SAS, 2004. Q3 2004 Report (Last accessed 20210103). https://www.sasgroup.net/investorrelations/financial-reports/interim-reports/sas-interim-report-january-september-2004/. Steen, F., Sørgard, L., 2002. From a regulated duopoly to a private monopoly; the deregulation of the Norwegian airline industry. Swed. Econ. Policy Rev. 9, 191e221. Strandenes, S.P., 2004a. Airport capacities and airline competition. In: Meersman, H., Roosens, P., van de Voorde, E., Witlox, F. (Eds.), Optimizing Strategies in the Air Transport Business, Survival of the Fittest, Antwerp, Garant, pp. 17e28. Strandenes, S.P., 2004b. Konkurranse og konkurransehindringer i norsk luftfart (Competition and hindrances to competition in the Norwegian aviation industry). In: Hagen, K.P., Sandmo, A., Sørgard, L. (Eds.), Konkurranse i samfunnets interesse, Festskrift til Einar Hope, (Competition in the Interest of Society. Festschrift in Honor of Einar Hoper), Bergen, Fagbokforlaget, pp. 176e192 (In Norwegian).

This page intentionally left blank

CHAPTER 15

Is privatization of ATC an economic game-changer? Who gains and who loses? Sven Buyle

Department of Transport and Regional Economics, University of Antwerp, Antwerp, Belgium

1. Introduction Before the 1970s, air transport economics research focused solely on airlines and airports. However, this changed after the oil crisis in the 70s, when researchers started to look for ways to make aviation less fuel dependent. Policymakers and the research, and broader aviation community soon realized that how states organize Air Traffic Control (ATC) has a significant impact on airline and airport performance (Hislop, 1980). Following the 1944 Chicago Convention on Civil Aviation, air navigation services provision is the state’s responsibility. Traditionally, each country had a government department in charge of delivering ATC services in its territory. The Convention, however, does not oblige countries to offer the services themselves. States have the option to delegate the provision of ATC to another country or (commercial) company while remaining responsible for the supervision. Entirely in the spirit of the time, New Zealand used this option and commercialized its Air Navigation Service Provider (ANSP) in 1987. Many countries followed this example in the years after that. Since then, many researchers have investigated the consequences of ATC commercialization and privatization on the market. This chapter assesses whether the privatization of ATC has been an economic game-changer. It covers the effects of commercialization and privatization on economic efficiency, ATC prices, safety and quality of service, and ANSP profits. After discussing definitions, the first part of the chapter summarizes past research findings on the effects of ATC commercialization and privatization. In the second part, multivariate data analysis is conducted on the current business model characteristics of European ANSPs. European ANSPs seem to differ mainly in their corporatization level, which can be linked to commercialization and The Air Transportation Industry ISBN 978-0-323-91522-9 https://doi.org/10.1016/B978-0-323-91522-9.00019-1

© 2022 Elsevier Inc. All rights reserved.

335

336

The Air Transportation Industry

privatization. The third part analyzes the relationship between the identified factors and profits from Air Navigation Services (ANS), which leads to further conclusions regarding the effects of ATC privatization. While the literature review covers ANSP commercialization cases in different geographical regions, this chapter focuses on the effects in the European ANS market.

2. Definitions Before providing an overview of the current state of the literature, it is interesting to look at how various ANS economics authors view the terms “commercialization” and “privatization”. Commercialization is often linked with financial autonomy (McDougall and Roberts, 2007), bringing in private sector style incentives to the sector (Button and McDougall, 2006), managing the service “business-like” with for-profit objectives and free access to labor and capital markets (Andrew, 2012). In contrast, privatization is connected with private ownership, wholly or in the majority (Arblaster, 2018). A commercialized ANSP can still be publicly owned. In a privatized ANSP, however, at least part of the shares will be in the hands of the private sector. As different authors use different definitions, Materna and Galieriková (2019) tried to conceptualize commercialization. They emphasize that commercialization is a process that takes several years and that it is about providing commercial services in new (liberalized) markets as opposed to the traditional approach of service provision. The commercialized ANSP often follows new innovative strategies and management approaches, and additional revenue creation forms the main driver for engaging in commercial (non-regulated) activities. It becomes clear that in reality, there exists a continuum of governance forms with different gradations regarding the level of commercialization and privatization. On one end of this spectrum resides the ANSP as a government department offering a public service. On the other end, the fully privatized and commercialized ANSP (wherein the government is no longer a shareholder). In Europe, most ANSPs commercialized in the 1990s or early 2000s (see Fig. 15.1). Only a few, however, were also privatized, while the government remained the main shareholder. As a consequence, European ANSPs remained geographical monopolists after the commercialization wave.

Is privatization of ATC an economic game-changer? Who gains and who loses?

337

Figure 15.1 ANSP commercialization timeline. (Source: Devised by the author.)

3. Literature review One of the first academic ANS economics research papers that can be found dates back to 1980. Hislop (1980) describes how air traffic control economics’ interests emerged after the oil crisis and economic recession of the 1970s. As traffic and ANS charges grew after the recession, the first ideas for sector reforms emerged. In the 1990s, Majumdar (1995) identifies the motivations for a state to commercialize its national ANSP and possible issues in this change by analyzing the commercialization of the ANSP in New Zealand. Early concerns raised are the potential decreases in ANS safety levels due to new commercial and profit-making objectives. The main motivations for ANSP commercialization following Majumdar (1995) are e.g., budget constraints, procurement problems; civil service constraints; provision of a customer-driven business; etc. Research of 2003 suggests that the financing motive and having control over ANSP revenues are the prime motivation for privatization instead of performance benefits (Sclar, 2003). An evaluation of the public-private partnership (PPP) active in the UK since 2001 learned that the PPP has not realized its advocates’ financial expectations (Steuer, 2010). Another early research paper on ANSPs was written by Prins and Lombard (1995), although it does not solely focus on ANSPs. Similar to the research of Majumdar (1995), it concerns a case study of the commercialization of state-owned enterprises. The main conclusion is that if the commercialized airport or ANSP operates in a monopolistic setting, there is a need for a policy and legislative framework to stimulate efficient operations, as there are no competitive forces that do so (Prins and Lombard, 1995).

338

The Air Transportation Industry

3.1 Commercialization and competition Commercialization often raises the question of market liberalization and competition. Sclar (2003) argues that ANS could never be subject to competition because it fails what he calls the yellow pages test. There is no list of private organizations available to contact if an airspace user wants to obtain air navigation services in a given country, such as it would do for getting, e.g., handling services. In all cases, there is only one provider because, as Sclar (2003) argues, the high capital cost of the needed infrastructure. These high capital costs form a sunk cost for a possible new market entrant and could give rise to economies of scale advantaging the incumbent. On the other hand, airspace safety could be considered a public good that is non-excludable and non-rivalrous. Having two providers for the same airspace sector would possibly lead to safety issues. However, more recent research by Delhaye and Blondiau (2016) shows that, due to advances in technology, it becomes possible to introduce trajectory-based ANS concepts. Such concepts, in fact, do allow for competition within the same geographical area. The yellow page test argument would be no longer valid. The sunk cost argument partly fades away when air traffic data providers are entering the market. With the introduction of satellite-based surveillance systems and digital tower technology, there is no longer a need for the ANSP to invest in its ground-based communication, navigation and surveillance (CNS) equipment. The needed data can be supplied from the (private) satellite network operator, lowering barriers to entry for ANSPs. While the concept of such ANS Data Service Providers is now considered crucial to solving the European ANS system’s inefficiencies, it is up to the regulator to create the right regulatory and commercial incentives for those entrants to emerge (Finger and Serafimova, 2020). Golaszewski (2002) suggests that economies of scale are likely to exist in providing ANS. According to Sclar (2003), ANSPs are generally seen as natural monopolists, despite the current discussion in the academic literature. Some authors claim that the average ANSP faces large economies of scale (e.g., Holder et al., 2006 and Dempsey-Brench and Volta, 2018), while others find low economies of scale (e.g., COMPAIR, 2017 and Veronese et al., 2011) or only economies of density in the sample means (e.g., Buyle et al., 2020). While there is no consensus on the exact degree to which economies of scale are present in the market, it is plausible that there

Is privatization of ATC an economic game-changer? Who gains and who loses?

339

are limits to natural monopolies and consolidation opportunities within Europe. What is also certain is that due to the Chicago Convention, ANSPs are currently geographical monopolies. However, this does not mean that competition is nonexistent. There has always been some competition for en-route services since airspace users can choose the route they file in their flight plan. Delgado (2015) looked into flight data from the EUROCONTROL data demand repository to determine how airlines make their route decisions for intra-European flights based on a fuel and route charge trade-off. His findings are that airlines have incentives to prefer longer routes in exchange for lower route charges by using neighboring airspace. However, the competition for this en-route traffic is somewhat limited, as, on the busiest traffic day in 2014, only 6.4% of the flights preferred a longer route to save charges. On top of that, en-route competition seems to be limited to specific regions in the European airspace. Delgado (2015) identifies three primary traffic flows where airlines can take advantage of the route charge/ fuel cost trade-off: flights between Northern Europe and the Canary Islands, flights to/from Central and Northern Europe to/from the GreeceCyprus-Turkey area and north-south routes within Eastern Europe. The commercialization of ANS also brought new forms of competition in the means of competition for the market instead of competition in the market. In competition for the market, separate independent ANSPs can bid to obtain a (temporary) license to provide services at a given airport or en-route sector. The competition is in the bidding process. Once the license is granted, the winning ANSP has a monopoly position for the geographical area in which the license is valid. An example of competition for the market can be found at National Air Traffic Services (NATS), the former national UK ANSP and one of the few European privatized ANSPs. NATS no longer possesses an institutional monopoly position, especially for terminal services. Many UK airports have their own terminal ANSP or have the ability to buy services from other providers. Even for en-route services, NATS is no longer sure to be the sole provider, as other providers are free to bid for a license (Goodliffe, 2002). The common understanding among ANS academics is that the opening of terminal markets in the UK has positively affected ANS cost-efficiency, even at airports where the market was not liberalized. As Adler et al. (2018) demonstrate, the introduction of competition for the market of en-route ANS could significantly reduce charges. It may lead to defragmentation when providers win multiple auctions. However, Arblaster and Zhang

340

The Air Transportation Industry

(2020) argue that, given the currently available information on the liberalization of terminal ANS in Spain, it is disputable whether competition for the market leads to economic welfare changes. The introduction of competitive tendering procedures in ANS markets must be accompanied by the proper institutional framework and regulation lowering entering barriers, e.g., guaranteeing access to specialized ATCO labor, which can be costly (Arblaster and Zhang, 2020). If adopted, the SES 2þ regulatory package aims to follow the example of the UK, Spain, Germany, Sweden and Norway and stimulate member states to introduce competition for terminal and en-route ANS in their territories. It is important to note that the commercialization of ANSPs is a prerequisite for the liberalization of terminal and en-route markets (e.g., in Spain ENAIRE was separated from AENA, the national airport operator). However, it is the regulatory framework that leads to potential efficiency benefits rather than the commercialization or privatization by itself. 3.2 Effects of commercialization When more countries started to commercialize their ANSP, research interests moved toward analyzing the effects of commercialization and privatization efforts. Button and McDougall (2006) wisely state here that “conclusions must be tentative and limited” which means that it cannot be said with certainty that the commercialization or privatization of ANSPs by itself has improved performance. Other authors come to comparable conclusions. Some take it even further and state that, because it is difficult to know what would have happened if the commercialization did not occur, it is impossible to say that stakeholders are now better off with the commercialized ANSP (Steuer, 2010). There are just too many factors that can influence ANSP performance for a fair comparison. Today’s shared belief is that commercialization by itself is not enough to achieve the desired economic benefits. Governments also need to set up the right regulatory or competitive stimuli to counter ANSP monopoly power and maximize economic welfare. Without such incentives, commercialized or privatized ANSPs have the liberty to charge above their average cost level and will continue to produce inefficiently. The combined impact of competition, regulation and ownership on airport performance has been demonstrated by Adler and Liebert (2014). Similar effects can be expected for ANSPs.

Is privatization of ATC an economic game-changer? Who gains and who loses?

341

Although conclusions are limited, most research reviewed in this chapter provides arguments favoring commercialization. Sclar (2003) is the only author included in the review whose arguments are against the commercialization case. It should be noted, though, that his research into the pitfalls of ANSP privatization was commissioned by the American National Air Traffic Controllers Association. Sclar (2003) does, however, have some legit points of criticism toward the privatization advocates. He writes that privatization proponents see privatization as a solution but fail to specify the problem. For example, Abeyratne (2001) sees privatization as a tool for solving the aviation industry’s challenges but does not mention these challenges. Table 15.1 presents an overview of case studies that analyze the effects of commercialization and privatization, the analyzed time-frame, and the countries studied. Most of these studies focus primarily on European cases, but also other geographical regions (US, Canada, Australia, New Zealand, and South Africa) are represented. 3.2.1 Effects on safety and security An ANSP has to support the safe conduct of flights. A common fear in commercializing ANS has been that it would jeopardize safety. Most studies Table 15.1 Case-studies of ANSP commercialization. Author Years Countries studied

Majumdar (1995) Prins and Lombard (1995) Golaszewski (2002) Goodliffe (2002) Sclar (2003) Button and McDougall (2006)

1987e1994 1993e1995 2000 2000 1997e2001 1997e2004

Poole (2007)

2005

McDougall and Roberts (2007)

1997e2004

Steuer (2010)

1995e2009

Germany, New Zealand South Africa USA UK Australia, Canada, UK Australia, Canada, France, Germany, Ireland, Netherlands, New Zealand, South-Africa, Switzerland, UK, USA Australia, Canada, Germany, New Zealand, UK Australia, Canada, France, Germany, Ireland, Netherlands, New Zealand, South-Africa, Switzerland, UK, USA UK

342

The Air Transportation Industry

agree that there is no reason to state that commercialization would have harmed safety. In some cases, safety has even been improved. Furthermore, there is still the regulator whose role is to safeguard safety, supplemented by the safety incentives stemming from improved customer focus (Button and McDougall, 2006; McDougall and Roberts, 2007; Poole, 2007; Steuer, 2010). Apart from the operational safety question is the question of national security. Privatization of ANSPs implies giving private organizations access to vital information (not only the ANSPs but also its contractors and subcontractors), imposing a new risk of national security and, in particular, a potential threat for terrorist activities (Sclar, 2003). 3.2.2 Effects on costs Sclar (2003) estimated that privatization of the national ANSP in the US would increase ANS provision costs by 30%.1 Although the same study states that several private ANSPs have reduced costs, this cost reduction came with a price in safety issues and lower employee satisfaction, creating additional costs. This finding is, however, contradicted by other research. It sounds reasonable to accept that commercialization mainly affects the distribution of the cost burden created by ANS provision (Button and McDougall, 2006). Whereas before the commercialization wave, the gap between costs and revenues was mainly covered by the taxpayer, this is today no longer acceptable.2 Other studies also report a reduction in ANSP costs per IFR movement between 5% and 15% between 1997 and 2004 (McDougall and Roberts, 2007; Poole, 2007). In 2018, the unit ANS provision cost in Europe reached its lowest point since 2001 at 389 euros. When considering the total user costs, the evolution becomes less clear as periods with significant ATFM delay reductions are followed by periods with large increases in total ATFM delay costs (EUROCONTROL, 2020). Sclar (2003) also states that privatization fails to address efficiency concerns. On the contrary, McDougall and Roberts (2007) conclude that the increased focus on customer needs and the ability to make faster decisions after the commercialization led to a continuous efficiency improvement. 1

2

This estimation is based on a proprietary cost model. Sclar (2003), however, do not provide any details on the method and data used. Covering ANSP losses with public funds might, however, be an alternative to airline government bailouts in times of crises (e.g., during the COVID-19 pandemic). It could be a way to support the air transport sector without distorting the level playing field in the airline market.

Is privatization of ATC an economic game-changer? Who gains and who loses?

343

Although the latter might be focusing more on flight efficiency rather than the economic efficiency of the ANSP. Both authors, however, do not rely on empirical efficiency measures. When reading the more recent econometric research papers, one quickly learns that there is no clear consensus on whether ownership influences ANSP efficiency. While Dempsey-Brench and Volta (2018) suggest that ownership has no direct impact on ANSP cost structures or costefficiency, Button and Neiva (2014) find that ANSPs closer linked to the government are more efficient than the commercialized or privatized ANSPs. COMPAIR (2017) makes a distinction between the efficiency of terminal and en-route services. For the en-route services, COMPAIR (2017) finds the public-private ANSPs to be more efficient, while the government agencies and corporations outperform these in terms of efficiency of terminal services. COMPAIR (2017) also suggests that ANSP efficiency is higher when vital stakeholders are involved in the board or shareholder structure. 3.2.3 Effects on ANS prices Following Sclar (2003), private ANSPs have limited incentive to keep user fees low, which contradicts most national and EU aviation policies’ goals. Governments require low ANS user charges to have an internationally competitive transportation system. This statement is somehow inconsistent with the findings of Button and McDougall (2006), who state that commercialization, at least in the long term, leads to reductions of user charges. However, that user charges are lower does not mean that total user costs decreased after commercialization (e.g., cost of delay). Sclar (2003) states that case studies indicate that private ANSPs tend to impose higher costs on users. One could again argue that the effect of privatization on charges depends mainly on the economic regulation put in place when the privatized ANSP remains a monopolist. In Europe, most ANSPs are under a price-cap regulation. ANS charges are fixed beforehand for several years by dividing the forecasted determined costs by the forecasted traffic numbers, corrected by an efficiency improvement target. Such a regulation creates an incentive for the monopoly ANSP to outperform the efficiency target and produce at an even lower cost, which will generate additional revenues (and bonuses for the private management). However, generating additional revenues entails the risk that the regulator will set stricter price caps in the following reference period. It will become more difficult for the ANSP to continue to

344

The Air Transportation Industry

outperform the cap and achieve additional revenues in the long run. The regulation hence creates incentives for the monopolist to safeguard information asymmetry, make only incremental efficiency improvements over time (which leads to more extended periods of smaller additional revenues and more bonuses for the management), reduce the quality of service or to look for other revenue sources in unregulated parts of the market. If price caps are set too loose, information remains asymmetric, and ANSPs have too much influence on the efficiency target and the price cap, prices are likely to remain high. 3.2.4 Effects on service quality Following McDougall and Roberts (2007) and Poole (2007), commercialization led to improved service quality. However, neither of them explain what should be understood under quality. If improved quality means fewer delays, it can be said with caution that service quality at least remained equal (McDougall and Roberts, 2007; Steuer, 2010). One conclusion of Steuer (2010) after its evaluation of the UK’s PPP is that there is no considerable change in delays that can be attributed to the ANSP. Nevertheless, he emphasizes that due to the small sample analyzed, not too much importance should be paid to this result. When exploring the evolution in total ATFM delays in Europe, periods of improvements are alternated with deterioration periods (EUROCONTROL, 2020). Once again, a robust regulatory framework is vital to stimulate ANSPs to improve service quality in the absence of competition. Quality based targets are currently included in the performance regulation. However, there are no clear sanctions when ANSPs don’t meet those targets. 3.2.5 Effects on customer relationships Button and McDougall (2006) found evidence that the ANSP product portfolio improved after commercialization. ANS users also got more involved in decision-making processes leading up to a more transparent ANS organization (Button and McDougall, 2006). User consultation processes are an essential part of the performance regulation. The enhanced customer relationship is also to the government’s benefit, which finds itself now supported as a shareholder by airlines as user experts (Andrew, 2012). Before commercialization, the government had to decide on policy but had little experience with ANS. Bringing in the input from the airlines as customers solves this issue. More research into the relationships between the airspace users and ANSPs will benefit all stakeholders.

Is privatization of ATC an economic game-changer? Who gains and who loses?

345

Arblaster (2012) investigated the relationship between ANSPs and airspace users concerning technological developments. She finds that, since ANSPs are natural monopolists, airlines have little influence on ANS technology investments, although they are being consulted more often in consultation processes since commercialization. However, since most technologies benefit from network effects, it is crucial to get more airspace users involved. More research into this issue is needed to come to faster technology adoption. 3.2.6 Effects on government relationship Private ANSPs still have to rely on government backing during disruptive market events (Sclar, 2003). Different studies take notice of the financing problems many commercialized ANSPs faced after the market collapse in 2001. After the 9/11 attacks, the demand for air transport dropped, so did the need for ANS. Since ANSPs have high fixed costs stemming from highly specialized staff and infrastructure, they cannot match supply with decreasing demand as easy as a commercial organization in any other industry would be able to do. Without government intervention, many ANSPs would possibly not have survived, which is, of course, unfavorable given the importance of ANS for air transport. ANSPs are facing similar difficulties during the COVID-19 crisis (Button and McDougall, 2006; McDougall and Roberts, 2007). 3.2.7 Effects on labor and capital Private ANSPs tend to be more prone to technological failure and labor disputes, while its monopolistic nature and labor intensity might undermine adequate service provision (Sclar, 2003). However, McDougall and Roberts (2007) did not see any typical commercialization risksdthe erosion of government accountability, deterioration of labor relations, or worsened relationships between civil and military ANSdemerging.

4. The emergence of the ANSP business model and its impact on ATM/CNS profits A consequence of the commercialization and privatization wave is that ANSPs started to develop their business model. While the Single European Sky initiative initiates a further reform of the sector, ANSP management has to continue to look for ways to cope with the expected future increase in competition, innovation and pressure on cost-effectiveness. Further

346

The Air Transportation Industry

liberalization of the industry will require ANSPs to build a competitive advantage by further improving their business model. This part of the chapter identifies the current business model characteristics used to assess the commercialization wave’s impact on ANS profits. Section 4.1 discusses the definition of a business model and the quantitative multivariate method used to determine the primary business model components, Section 4.2 the business model variables used in this analysis, Section 4.3 the resulting business model constructs and Section 4.4 the impact on profits. 4.1 Methodology To get such insights, one needs a working definition of the business model. A good starting point is the work of Amit and Zott (2001). In their view, a business model consists of the content, structure and governance components of the transactions that create value for the firm. Later work by Casadesus-Masanell and Ricart (2010) builds further on the Amit and Zott (2001) definition. They see a business model as a collection of choices made by the firm and their consequences. This chapter relies on this framework of choices as it lends itself easily to be adapted to the context of the sector of interest. After having identified the essential business model choices, these can easily be quantified for further analysis. Table 15.2 provides an overview of all variables influencing the ANSP business model considered in this study. These variables are ordered according to the Casadesus-Masanell and Ricart (2010) framework, in strategic (or policy), asset, and governance choices and the strategy outcomes (or consequences).3 They are quantified using cross-sectional data collected between 2016 and 2019 from the 2016 ACE benchmarking report published by EUROCONTROL (2018), ANSP websites and available annual reports. Information on Horizon (2020) project participation is taken from the EU Open Data Portal (2018). The resulting dataset is a mixture of quantitative and qualitative information. A factor analysis method for mixed data described by Pagès (2004) is used to identify and measure the primary business model constructs of ANSPs. This method is a combination of principal component analysis (PCA) and multiple correspondence analysis (MCA), in which the quantitative and qualitative variables are compared at an equal level. The method is preferred above traditional factor analysis, which can only be used with

3

The variables are described in detail in Section 4.2.

Is privatization of ATC an economic game-changer? Who gains and who loses?

347

Table 15.2 Variables influencing the ANSP business model. Strategic choices Choice

Variables

Operational scope

Marketable service offer (yes/no) Military ANS integration (yes/no) International ANS services (yes/no) Number of ANSP-only alliances Number of mixed alliances Number of ANSP-only joint ventures Number of ANSP-supplier joint ventures Number of mixed joint ventures Horizon (2020) projects per 100 FTEs

Collaboration forms

Innovation strategy

Asset choices Choice

Variables

Factor inputs Make-or-buy choices

Labor to capital ratio Outsourcing of MET services (yes/no) Governance choices

Choice

Variables

Ownership structure

% of government shares % of private shares Government department (yes/no) Common airport-ANS entity (yes/no) Independent company (yes/no)

Corporate structure

Strategy outcomes Outcome

Variable

Cost structure

Cost-share of staff costs Cost-share of non-staff operational costs Cost-share of depreciation costs Cost-share of capital costs Unit cost of terminal services Unit cost of en-route services Revenue share of terminal services Revenue share of en-route services Revenue share of marketable services Unit revenue of terminal services Unit revenue of en-route services

Unit costs Revenue structure

Unit revenue

348

The Air Transportation Industry

quantitative variables. Factor analysis is used, as opposed to, e.g., cluster analysis, since it not only provides a ground to classify ANSPs based on their business model variables but also provides insight into the primary structures inherent in the dataset. The obtained factor scores are afterward regressed on the 2016 gate-togate ATM/CNS profits to identify significant links between the business model constructs and overall ANSP performance. Profits from commercial activities and subsidiaries are not included in the dataset. Furthermore, it should be noted that profitability is just one of the criteria that can be used to assess ANSP business model performance. ANSPs might also have other objectives than only profit maximization, e.g., enhancing national interests or consumer surplus (see Section 4.2.4). 4.2 Variables influencing the ANSP business model This section discusses the variables taken into account in the factor analysis and is structured based on the different blocks of the Casadesus-Masanell and Ricart (2010) framework. This section furthermore provides a general overview of the current state of the industry. 4.2.1 Strategic choices Strategic choices refer to those management decisions that affect all aspects of the firm operations. In this research, they are considered to be partially overlapping with the content and structure transactional components of Amit and Zott (2001). The specific choices are identified as the operational scope of the core ANS, innovation strategy, and whether the ANSP decides to cooperate with third parties. 4.2.1.1 Operational scope The operational scope of the ANSP comprises the services offered. Initially, most national ANSPs only provided ANS to civil flights within their national airspace under the federal government’s direct control. Over time some ANSPs started to extend their operations to more market-oriented commercial services. These ANSPs market their expertise in consulting services, open their air traffic controller (ATCO) training school to provide training to ATCOs of other ANSPs or actively bid for ANS contracts at airports in partly-deregulated markets. To stress that these services are offered in a market-like setting, they will be referred to as marketable services. The majority of these marketable services are non-core services, while only the commercially offered terminal ANS could be seen as a core service.

Is privatization of ATC an economic game-changer? Who gains and who loses?

349

As noted by e.g., Tomova (2016), these marketable services and the commercial revenues that they generate are gaining importance for European ANSPs. They are a consequence of the commercialization wave. As ANSPs commercialize and get economically regulated, they start to explore unregulated market segments to maximize their revenues. While civil and military ANS were strictly separated in the past, more European countries realize that having two providers with each their infrastructure is inefficient. Therefore, in 2016, already 14 out of 37 national ANSPs integrated military services into their product offer (EUROCONTROL, 2018). Both the kinds of services as well as where they are offered changed over time. Because of political reasons, a small minority of ANSPs started to internationalize to some extent. In such cases, a country’s government delegates ANS provision in a part of its airspace to a neighboring country (EUROCONTROL, 2018). An exception to this case is HungaroControl (Hungary (HU)), which is remotely controlling the airspace of Kosovo (under a NATO mandate) (HungaroControl, 2019). NATS (United Kingdom (UK)), DFS (Germany (DE)) and Austro Control (Austria (AT)) to the contrary are actively making use of the opportunities created by the opening up of the terminal ANS market by actively acquiring service licenses at foreign airports within Europe. DFS and NATS do so via a subsidiary set up for this purpose (Air Navigation Solutions, 2018; DFS, 2018; FerroNATS, 2019; NATS, 2019). For this study, an ANSP is considered to be international if it offers enroute services in a considerable part of the territory of a foreign country (i.e., Croatia Control and SMATSA in Bosnia and Herzegovina,4 HungaroControl in Kosovo and Skeyes in Luxembourg) or offers terminal services at foreign airports, whether these are provided via subsidiaries (i.e., NATS and DFS) or the mother company (i.e., Austro Control). 4.2.1.2 Collaboration strategy Many ANSPs are collaborating with other ANSPs or technology providers in joint ventures or alliances. Alliances between ANSPs and a technology provider (i.e., iTEC and COOPANS) are initiated to share technology development costs and reduce the fragmentation of technology used in the

4

Bosnian airspace is currently no longer controlled by SMATSA and Croatia Control.

350

The Air Transportation Industry

European ANS landscape (COOPANS, 2019; iTEC, 2019). The joint ventures often have a more commercial focus. Examples include ⁃ GroupEAD, a joint venture between ENAIRE (Spain (ES)), DFS (DE) and technology provider Frequentis, which is an aeronautical data management company operating on behalf of EUROCONTROL; ⁃ European Satellite Services Provider (ESSP), a joint venture owned by seven European ANSPs (DFS, DSNA, ENAIRE, ENAV (Italy (IT)), NATS (UK), NAV Portugal and Skyguide (Switzerland (CH))) operating a satellite-based system for navigation. ESSP has a pan-European certificate to act as an actual ANSP; ⁃ Aireon, a joint venture between NAV Canada, ENAV, IAA (Ireland (IE)), NAVIAIR (Denmark (DK)) and Iridium, that is setting up a system for space-based surveillance; ⁃ Entry Point North (EPN), a joint venture between LFV, Naviair and IAA that provides ATCO training to third parties (other ANSPs have set up joint ventures with EPN for local training schools) (Entry Point North, 2019); ⁃ Flight Calibration Services, a joint venture between Austro Control (AT), DFS and Skyguide that provides R&D, engineering, consultancy and inspection services for communication, navigation and surveillance systems (CNS) equipment (Flight Callibration Services, 2019); ⁃ Frequentis DFS Aerosense, a joint venture between Frequentis and DFS subsidiary DFS Aviation Services for the development and commercialization of a remote tower system (Anonymous, 2018a); or ⁃ Saab Digital Air Traffic Solutions (SDATS), a joint venture between LFV (Sweden (SE)) and Saab. The last one sells the remote tower system, but they also intend to operate it from their center in Sweden, making SDATS the first fully digital ANSP (Anonymous, 2018b). In the analysis, the number of joint ventures and alliances in which the focal ANSP participates is split up by type of partners (ANSP-only, ANSP, and supplier, mixed form). 4.2.1.3 Innovation strategy The extent to which a firm contributes to innovation in the sector is often measured via its research and development investments. Unfortunately, such data is not available for the European ANSPs. This study hence relies on the number of Horizon (2020) projects (linked to Single European Sky ATM Research (SESAR)) per 100 full-time equivalents (FTEs) in which

Is privatization of ATC an economic game-changer? Who gains and who loses?

351

the ANSP is participating or has participated. All projects to which the ANSP contributed between 2014 and 2018 are included regardless of whether they were finalized during the reference period or are still ongoing. The correction for the number of FTEs is made to overcome possible bias since it is assumable that larger organizations have more means to participate in a higher number of projects than smaller ones. However, this normalization does not distinguish between two ANSPs, each having the same number of FTEs and participating in the same project but contributing a different number of person-months. Such an approach would require reliable data on the person-month project contribution, which is currently not available. Only half of the European ANSPs participate in Horizon (2020) projects. While some of those ANSPs only participate in one or two projects, most are active in a wide range of projects. However, when analyzed in terms of projects per 100 FTEs, there is much more variation between the ANSPs. With six projects per 100 FTEs Oro Navigacija (Lithuania (LT)) scores the highest, followed by IAA (IE), LPS (Slovakia (SK)), and Naviair (DK) with between three and four projects per 100 FTEs. Only four ANSPs are taking up a leading role in these projects: ENAV (IT), DFS (DE), DSNA (France (FR)), and NATS (UK)5 (EU Open Data Portal, 2018). While this study uses Horizon (2020) project participation, it should be noted that other programs might have existed before, or simultaneously with, the Horizon (2020) program and that ANSP innovation projects are not necessarily always sponsored by such large (European or national) government initiatives. Therefore, the potential adoption of innovative solutions, such as remote tower technology, is considered in this study. It should be noted that both variables only measure the innovation effort and not to which extent they contribute to cost reductions or revenue improvements. 4.2.2 Asset choices There are two choices related to the assets used by the ANSP considered in this study. One is the labor to capital input ratio, and the other concerns the make or buy decision of support services such as meteorological services for air navigation (MET). The labor to capital input ratio is calculated by 5

Using the number of projects per 100 FTEs in which the ANSP has a leading role does not considerably change the factor interpretations.

352

The Air Transportation Industry

dividing the number of FTEs employed by the fixed asset’s net book value and corrected by the national capital goods price index as measured by Eurostat (2018). Six missing values for the price index are imputed via predictive mean matching. Note that only outsourcing of MET services is considered, as there is no publicly available and reliable data on outsourcing other kinds of services. 4.2.3 Governance choices Most European ANSPs today are independent but government-owned enterprises. Exceptions are NATS (UK), ENAV (IT) and Skyguide (CH). A consortium of airlines owns 49% of the shares in NATS, while 47% of the shares of ENAV are traded on the stock exchange. The other part is in the hands of the government. For Skyguide, only a tiny amount of the shares (less than 1%) is in private hands (EUROCONTROL, 2018). While most ANSPs are government-owned, national legislation frequently includes the option of private sector involvement, often with a golden share that remains with the government, which stresses the strategic importance of the ANS industry. The ANSPs which are not an independent company are either a government body (DSNA, PANSA, DCAC Cyprus, HCAA) or part of a common airport-ANS entity (Avinor, DHMI). Interesting here is that at the beginning of 2018, ANS Finland was split off from Finavia, the national airport operator in Finland. However, the Finnish government has the ambition to integrate ANS Finland with traffic managers of other transport modes under the same holding. They believe this would stimulate digitalization and create possibilities for commercializing new services, e.g., increased data sharing (ANS Finland, 2018; EUROCONTROL, 2018). 4.2.4 Strategy outcomes The strategic, asset and governance choices affect the ANSP cost and revenue structure, which can be considered a strategic outcome. The operational scope, service offer decisions, asset choices, collaboration and innovation strategies of the ANSP determine its revenue streams and cost structure. An ANSP with a high labor to capital ratio and which outsources noncore services is likely to have a different cost structure than an ANSP which is part of a common airport-ANS entity and has a low labor to capital ratio. In the end, costs and revenues will determine profits and part of the business model’s success. Profit maximization is, however, not the only objective of an ANSP. As postulated by Adler et al. (2020), the ANSP objective function is a mixture of the maximization of profits, consumer

Is privatization of ATC an economic game-changer? Who gains and who loses?

353

surplus and national interests. Consumer surplus and national interests are, although more challenging to quantify and thus excluded here. ANSPs are subject to the SES Performance Scheme with binding performance targets in four key performance areas (safety, environment, capacity, and costefficiency). This regulation has undoubtedly an influence on the success of the business model. 4.2.4.1 Cost structure The cost structure is measured by calculating the cost shares of the gate-togate staff costs, non-staff operational costs, depreciation costs and capital costs.6 The cost structure explains how the money is spent, but not how much is spent. Therefore, it is useful to take into account the unit costs of the terminal and en-route services. The unit cost of terminal services is calculated by taking the ATM/CNS cost attributable to terminal services and dividing it by the number of IFR airport movements (as reported in the ACE benchmarking reports). The unit cost of en-route services is calculated similarly by dividing the ATM/ CNS costs attributable to en-route services by the en-route IFR flight kilometers handled. 4.2.4.2 Revenue structure The revenue structure is measured based on the contribution of each product to the operational revenue. ANSP operating revenues are composed of revenues generated by terminal charges, en-route charges and the marketable service offer. According to the annual reports, 14 out of 37 ANSPs have revenues from marketable services ranging from 0.1% to 13% of the operational revenue. However, these revenues contribute no more than 1% to the operating revenues for four of them and are hence neglectable. Only for three ANSPs (ANS CR, NATS and LFV), marketable services have an operational revenue contribution of more than 5%. Next to revenue shares, the unit revenues are also taken into account. These are measured by taking the total revenues generated by the terminal and en-route services, dividing them by the number of IFR airport movements and IFR flight kilometers handled. The unit revenues are positively correlated with the unit costs, likely due to the price-cap regulation. 6

The total user cost of capital is calculated in the ACE benchmarking reports as the sum of the cost of equity and interest costs through the weighted average cost of capital.

354

The Air Transportation Industry

4.3 Business model constructs Based on the factor analysis of the collected data, five primary ANSP business model constructs are identified and quantified: corporatization, collaboration and terminal efficiency, capital intensity, the level of outsourcing and mixed alliance participation versus commercial focus. Table 15.3 provides an overview of the variance in the data explained by each factor. The most variance between European ANSPs can be explained by their corporatization level, which is a concept strongly linked to commercialization and privatization. Highly corporatized ANSPs tend to act more like large independent commercial businesses. The factor contrasts the ANSPs which are partly privatized, have a marketable service offer, an international scope and engage in joint ventures, with government departments only offering ANS nationally. This concept of corporatization hence more or less coincides with Materna and Galieriková’s (2019) understanding of the idea of commercialization. Another important ANSP business model construct is the level of collaboration and the extent to which the ANSP innovates to improve terminal services’ efficiency. ANSPs scoring high on this factor participate in many alliances, participate in more Horizon (2020) projects, invest in remote tower technology and have lower terminal unit costs. The three other business model constructs are the capital intensity, the level of outsourcing (military ANS, meteorological services, R&D via mixed joint-ventures) and a factor contrasting the ANSPs focusing mainly on mixed alliance participation with those having a commercial focus.

Table 15.3 Variance explained by the ANSP business model constructs. Mixed alliance versus commercial Capital Corporatization Collaboration intensity Outsourcing focus

Eigenvalue Percentage of variance Cumulative percentage of variance

6.77 25.09

3.71 13.76

2.73 10.12

2.11 7.80

1.60 5.91

25.09

38.85

48.97

56.77

62.68

Is privatization of ATC an economic game-changer? Who gains and who loses?

355

4.4 The impact on ATM/CNS profits Table 15.4 presents the results of three OLS estimations, regressing the identified business model constructs on ATM/CNS profits. All three estimated models apply with the heteroscedasticity and normality assumptions of the error term required in OLS estimation, as the null hypothesizes of the Goldfeld-Quandt test, Breusch-Pagan test, and KolmogoroveSmirnov test cannot be rejected at the 95% significance level. The models suggest a significant positive relation between the ATM/CNS profit and corporatization and outsourcing levels. The other business model components do not show substantial impacts. Table 15.4 Dependence of ATM/CNS profits on business model characteristics (OLS estimations). Model 1 Model 2 Model 3

(Intercept) Corporatization Collaboration Capital intensity Outsourcing Mixed alliance versus comm. focus

23,472.78*** (6039.62) 7656.25** (2481.34) 2510.20 (3347.85) 1875.16 (4172.17) 18,893.28*** (4583.39) 1211.10

59 998.90 (56,426.95) 6746.22* (2846.13) 1664.13 (3354.52) 1281.11 (4325.01) 11,201.25* (4842.89) 3987.36

(6078.42)

(5262.68) 0.05** (0.01) 1667.57 (2587.41) 439,990.99 (40,802.42) 0.69 0.60 36 30,179.82 0.0196 0.6190 0.0793 0.4291

Airspace Complexity Variability R2 Adj. R2 Num. obs. RMSE RESET test p.val. Goldfeld-Quandt test p.val. Breusch-Pagan test p.val. KolmogoroveSmirnov test p.val.

0.53 0.46 36 35,322.62 0.8433 0.8611 0.1150 0.0893

7023.32 (6631.16) 7306.67* (1946.90)

12,107.18** (3939.48)

0.05** (0.01)

0.66 0.63 36 29,300.43 0.1687 0.9226 0.0736 0.5485

356

The Air Transportation Industry

The findings from the regression analysis are logical. ANSPs with a higher level of corporatization tend to have a higher probability of being partly privately owned, to have an international scope, to have a commercial service offer and to participate more in joint-ventures (which often also have a commercial focus). It is reasonable to assume that those ANSPs put more attention to profit maximization. Also, the impact of outsourcing on profit seems to make logical sense. Commercialization, privatization or any form of corporatization is hence beneficial for the ANSP and its shareholders. It stimulates ANSPs to restructure their business model by developing a competitive advantage for certain well-chosen noncore services such that they can be offered commercially on the market. Other cost centers that cannot be turned into revenue centers (e.g., it is not easy to achieve competitive advantage) are better outsourced if they reduce overall costs. The findings furthermore suggest that ANSPs operating larger airspace generate significant larger ATM/CNS profits. Each additional square kilometer of airspace contributes an extra 50 euros to profits. From the current literature, it is unclear whether this profit contribution stems from economies of scale or increased monopolist market power. Either way, it explains the motive for ANSPs to extend their current market by, e.g., participating in tendering procedures in third countries. An important question to consider when ANS markets are further liberalized is whether European society will be better off with less, but larger monopolists (as Adler et al., 2018 suggest would happen after liberalization), especially when economic welfare impacts of these competitive tendering procedures are unclear (Arblaster and Zhang, 2020). While it is clear from the analysis that ANSP profits benefit from commercialization and privatization, the question remains (and this was the question this chapter started with) whether commercialization and privatization are also beneficial for airspace users and society as a whole. The main conclusion from the literature review is that it often depends on the effectiveness of market regulation. Even when the national ANSP is commercialized or privatized, it remains a monopolist that might abuse its market power if not appropriately regulated. Commercialization or privatization by itself cannot be considered as the holy grail if not complemented by a proper and effective regulatory and institutional framework. Simultaneously, regulating a market is costly due to, e.g., information asymmetry and regulatory capture risks. In the current European framework, the regulator has to invest in user consultation procedures that are

Is privatization of ATC an economic game-changer? Who gains and who loses?

357

both timely and costly, and the performance targets have to be monitored and benchmarked. The benefits of the regulation have to outweigh these costs. The fragmented ATC provision in Europe, with little coordination between sectors operated by different ANSPs, continues to lead to more delays and a higher user cost. Horizontal consolidation, to the contrary, could lead to higher monopoly power, diseconomies of scale (Buyle et al., 2020), higher charges and hence a need for even stricter economic regulation. Privatization and further liberalization of ATC are likely to lead to consolidation in the market. A structural market reform that allows for more competition in the market will have to go hand in hand with a fierce economic regulation that translates efficiency improvements in lower user costs to be a real economic game-changer.

5. Conclusions The chapter started with the question if ATC privatization, commercialization, or corporatization can be considered an economic game-changer. Who gains from it, and who loses? The question is answered in this chapter via an in-depth review of existing academic literature and by considering future ANSP business model developments in Europe, which is its main contribution. ANSP business model constructs are identified via a multivariate mixed data factor analysis and regressed on ATM/CNS profits to assess the impact of evolutions in business models due to commercialization and potential liberalization waves. The literature review shows that providing factual evidence on the effects of privatization is difficult due to the low number of privatized ANSPs and the high number of external factors that influence ANSP performance. General conclusions, however, are that privatization or commercialization of ATC, just as with airlines and airports, does not lead to safety infringement as governments seem to have established a wellfunctioning safety authority and regulator. The effects on service quality, the total user cost of ATC and ANSP cost efficiency, to the contrary, are less clear from the existing literature. Overall the privatization and commercialization of ATC have not been the economic game-changer that governments hoped for, despite that commercialization might have led to efficiency improvements. The winners are the shareholders, who achieve better returns and generate enough cashflows to make the necessary investments in new technologies and

358

The Air Transportation Industry

infrastructure. Airspace users who had hoped for lower air navigation charges often find themselves disappointed. The total user cost did not significantly decrease, as reductions in charges (if they exist) go hand in hand with higher delay costs. Efficiency improvements are not necessarily passed on to airspace users in terms of lower charges. The general conclusion from the analysis and discussion above is that commercialization and privatization are insufficient to achieve the desired economic benefits but can be necessary to achieve such goals. Competitive tendering procedures can create the right stimuli for ANSPs to improve their efficiency; however, its overall effect on economic welfare remains unclear. However, competitive tendering, or competition for the market, does not replace perfect competition in the market. Even if competitive tendering procedures are in place, there is a need for the proper institutional framework to keep entry barriers low, countering the incumbent’s market advantages.

References Abeyratne, R.I.R., 2001. Revenue and investment management of privatized airports and air navigation servicesda regulatory perspective. J. Air Transp. Manag. 7 (4), 217e230. https://doi.org/10.1016/S0969-6997(01)00008-4. Adler, N., Delhaye, E., Kivel, A., Proost, S., 2020. Motivating air navigation service provider performance. Transp. Res. Pol. Pract. 132, 1053e1069. https://doi.org/10.1016/ j.tra.2019.12.014. Adler, N., Hanany, E., Proost, S., 2018. Introducing Competition through Auctions in the Air Traffic Control Market Eighth SESAR Innovation Days. Salzburg, Austria. https:// nicoleadler.huji.ac.il/sites/default/files/nicoleadler/files/adler_hanany_proost_atc_ auction.pdf. Adler, N., Liebert, V., 2014. Joint impact of competition, ownership form and economic regulation on airport performance and pricing. Transp. Res. Pol. Pract. 64, 92e109. https://doi.org/10.1016/j.tra.2014.03.008. Air Navigation Solutions, 2018. Welcome to ANS. https://airnavigationsolutions.co.uk. Amit, R., Zott, C., 2001, JuneJul. Value creation in e-business. Strat. Manag. J. 22 (6e7), 493e520. https://doi.org/10.1002/smj.187. Andrew, D., 2012. Institutional policy innovation in aviation. J. Air Transp. Manag. 21, 36e39. https://doi.org/10.1016/j.jairtraman.2011.12.015. Anonymous, 2018a. Germany’s DFS launches remote tower operations at Saarbrücken from Leipzig. Air Traffic Manag. 2018 (4), 7. Anonymous, 2018b. SDATS takes over remotely operated air traffic control in Sweden, eyes UK market. Air Transp. Manag. 2018 (4), 7. ANS Finland, 2018. Annual Report 2017. https://www.ansfinland.fi/application/files/ 2115/2767/2183/FINAL_ANS_AnnualReport.pdf. Arblaster, M., 2012. Comparing consultation on investment and technology decisions in air traffic management in Australia and the UK. J. Air Transp. Manag. 22, 36e44. https:// doi.org/10.1016/j.jairtraman.2012.01.007.

Is privatization of ATC an economic game-changer? Who gains and who loses?

359

Arblaster, M., 2018. Air Traffic Management. Economics, Regulation and Governance [Book]. Elsevier, Amsterdam. http://search.ebscohost.com/login.aspx? direct¼true&AuthType¼ip,url,uid&db¼cat01187a&AN¼ABC.c.lvd. 14554561&lang¼nl&site¼eds-live. Arblaster, M., Zhang, C., 2020. Liberalisation of airport air traffic control: a case study of Spain. Transp. Pol. https://doi.org/10.1016/j.tranpol.2020.03.003. Button, K., McDougall, G., 2006. Institutional and structure changes in air navigation service-providing organisations. J. Air Transp. Manag. 12 (5), 236e252. https:// doi.org/10.1016/j.jairtraman.2006.07.001. Button, K., Neiva, R., 2014. Economic efficiency of European air traffic control systems. J. Transp. Econ. Pol. 48 (1), 65e80. http://dev.anet.ua.ac.be/eds./eds.phtml? url¼http%3a%2f%2fsearch.ebscohost.com%2flogin.aspx%3fdirect%3dtrue%26AuthType %3dip%2curl%2cuid%26db%3dbuh%26AN%3d93723167%26lang%3dnl%26site%3dedslive. Buyle, S., Dewulf, W., Kupfer, F., Onghena, E., Meersman, H., Van de Voorde, E., 2020. Does ANSP size and scope matter in the European ANS market? A multi-product stochastic frontier approach. J. Air Transp. Manag. 83, 101754. https://doi.org/ 10.1016/j.jairtraman.2019.101754. Casadesus-Masanell, R., Ricart, J.E., 2010. From strategy to business models and onto tactics. Long. Range Plan. 43 (2e3), 195e215. https://doi.org/10.1016/ j.lrp.2010.01.004. COMPAIR, 2017. Report on Economic Analysis. http://www.compair-project.eu/files/ 2017_05_27_d3_2_report_on_economic_analysis_compair_00_02_00.pdf. COOPANS, 2019. About COOPANS. http://www.coopans.com/. Delgado, L., 2015. European route choice determinants. In: 11th USA/Europe ATM R&D Seminar. Delhaye, E., Blondiau, T., 2016. Introducing More Competition into ATM: Possible Institutional Designs 6th SESAR Innovation Days. Technical University of Delft, the Netherlands. http://sesarinnovationdays.eu/files/2016/Papers/SIDs_2016_paper_19. pdf. Dempsey-Brench, Z., Volta, N., 2018. A cost-efficiency analysis of European air navigation service providers. Transp. Res. Pol. Pract. 111, 11e23. https://doi.org/10.1016/ j.tra.2018.02.019. DFS, 2018. Annual Report 2017. https://www.dfs.de/dfs_homepage/en/Press/ Publications/GB%202017%20EN.pdf. Entry Point North, 2019. Our Heritage: Decades of Training Experience. https://www. entrypointnorth.com/about/our-history-and-heritage. EU Open Data Portal, 2018. CORDIS e EU Research Projects under Horizon 2020 (2014e2020). https://data.europa.eu/euodp/en/data/dataset/cordisH2020projects. EUROCONTROL, 2018. ATM Cost-Effectiveness (ACE) 2016 Benchmarking Report with 2017-2021 Outlook. https://www.eurocontrol.int/sites/default/files/2019-08/ ace-2016-benchmarking-report-upd.pdf. EUROCONTROL, 2020. ATM Cost-Effectiveness (ACE) 2018 Benchmarking Report. https://www.eurocontrol.int/ACE/ACE-Reports/ACE2018.pdf. Eurostat, 2018. Producer Prices in Industry, Total e Annual Data. MIG e Capital Goods. https://appsso.eurostat.ec.europa.eu/nui/show.do?dataset¼sts_inpp_a&lang¼en. FerroNATS, 2019. About FerroNATS. http://ferronats.com/en/about-ferronats/ourcompany. Finger, M., Serafimova, T., 2020. Enabling Air Traffic Management (ATM) Data Services. https://cadmus.eui.eu/bitstream/handle/1814/66906/PB_2020_18_FSR.pdf? sequence¼1.

360

The Air Transportation Industry

Flight Callibration Services, 2019. About FSC. https://www.fcs.aero/english/content/ about-fcs#. Golaszewski, R., 2002. Reforming air traffic control: an assessment from the American perspective. J. Air Transp. Manag. 8 (1), 3e11. https://doi.org/10.1016/S09696997(01)00018-7. Goodliffe, M., 2002. The new UK model for air traffic servicesda public private partnership under economic regulation. J. Air Transp. Manag. 8 (1), 13e18. https://doi.org/ 10.1016/S0969-6997(01)00029-1. Hislop, A., 1980. The economics of air traffic control. J. Navig. 33 (1), 23e29. https:// doi.org/10.1017/S0373463300028216P. Holder, S., Veronese, B., Metcalfe, P., Mini, F., Carter, S., Basalisco, B., 2006. Cost Benchmarking of Air Navigation Service Providers: A Stochastic Frontier Analysis. http://www.eurocontrol.int/sites/default/files/publication/files/cost-benchmarkingstochastic-analysis-.pdf. HungaroControl, 2019. More than an ANSP: Annual Report 2017. https://en. hungarocontrol.hu/download/a941cf0edc766669051abfc40e9bf67f.2017-eves-jelentes. pdf. iTEC, 2019. Delivering a Single European Sky. http://www.itec.aero/. Majumdar, A., 1995. Commercializing and restructuring air traffic control: a review of the experience and issues involved. J. Air Transp. Manag. 2 (2), 111e122. https://doi.org/ 10.1016/0969-6997(96)00004-X. Materna, M., Galieriková, A., 2019. A new approach to classification of air navigation service providers in the context of commercialization. Transp. Res. Proc. 43, 139e146. https://doi.org/10.1016/j.trpro.2019.12.028. McDougall, G., Roberts, A., 2007. Commercializing Air Traffic Control: Have the Reforms Worked? http://papers.ssrn.com/sol3/papers.cfm?abstract_id¼1317450. NATS, 2019. Annual Report and Accounts 2018. https://www.nats.aero/wp-content/ uploads/2018/07/NATS6766_AnnualReport2018_FULL.pdf. Pagès, J., 2004. Analyse factorielle de données mixtes 52 (4), 93e111. Poole Jr., R.W., 2007. The Urgent Need to Reform the FAA’s Air Traffic Control System, pp. 1e38. http://research.policyarchive.org/5798.pdf. Prins, V., Lombard, P., 1995. Regulation of commercialized state-owned enterprises: case study of South African airports and air traffic and navigation services. J. Air Transp. Manag. 2 (3e4), 163e171. https://doi.org/10.1016/0969-6997(96)00002-6. Sclar, E., 2003. Pitfalls of Air Traffic Control Privatization. Report commissioned by National Air Traffic Controllers Association. Steuer, M., 2010. The partially private UK system for air traffic control. J. Air Transp. Manag. 16 (1), 26e35. https://doi.org/10.1016/j.jairtraman.2009.07.011. Tomova, A., 2016. Are commercial revenues important to today’s European air navigation service providers? J. Air Transp. Manag. 54, 80e87. https://doi.org/10.1016/ j.jairtraman.2016.03.023. Veronese, B., Conti, M., Pesendorfer, M., 2011. Econometric Cost-Efficiency Benchmarking of Air Navigation Service Providers. http://www.eurocontrol.int/sites/ default/files/content/documents/single-sky/pru/publications/other/anspseconometric-cost-efficiency-benchmarking.pdf.

CHAPTER 16

The forwarders’ power play effect on competition in the air cargo industry Thomas Van Asch

Department of Transport and Regional Economics, University of Antwerp, Antwerp, Belgium

1. Introduction Contrary to the maritime sector in which shippers more often deal directly with the shipping lines, air cargo shippers are almost always supported by an intermediary, often called the freight forwarder. In the past, some shippers and airlines have tried to bypass the freight forwarder. Despite these efforts, such attempts were rather unsuccessful, and the freight forwarders still control a significant share of the air cargo market. In the past, freight forwarders were merely the architect of the whole transport chain. However, the product offering is broader nowadays, and freight forwarders also offer value-added activities such as labeling, repacking, and quality control. Despite the significant role of freight forwarders in the air cargo industry, its position has not been sufficiently examined as yet and little is known about the economics of these companies. The lack of reliable datadespecially on the air cargo segmentdis one of the main reasons for this. The chapter will try to address the identified research gap and add value to the existing literature by examining the relationships between the actors in the air cargo transport chain and more specifically the role that freight forwarders play in the air cargo market. Quantifying and analyzing the air freight forwarding industry dynamics will further contribute to the understanding of the position of the freight forwarder in the air cargo industry. This is important for a variety of reasons. First, with the deregulation of air transport, not only has air traffic been growing, but competition between airlines, forwarders, and airports respectively has also increased significantly. On the other hand, consolidation in both the airline and forwarding business seems to be apparent. These developments raise different challenges for both companies and governments. As air transport contributes significantly to a nation’s economy, a healthy and active air The Air Transportation Industry ISBN 978-0-323-91522-9 https://doi.org/10.1016/B978-0-323-91522-9.00003-8

© 2022 Elsevier Inc. All rights reserved.

361

362

The Air Transportation Industry

cargo industry is considered highly valuable by many governments, at the local as well as at the national level. Furthermore, knowledge about the relationships between the different actors is useful for policymakers when drafting their transport policies. Last but not least, a good understanding of the role of the freight forwarder can lead to a more efficient organization of the air cargo transport chain. A better view of the decision-making process of freight forwarders, airlines, as well as airports can better meet the expectations of these companies to ensure an efficient, reliable and sustainable air transport market. The remainder of the chapter is structured as follows. In the next section, a literature review is completed to present relevant research dealing with freight forwarders. While freight forwarders are mentioned in a broad range of research dedicated to air cargo, there is minimal research focusing exclusively on the air freight forwarding industry itself. The role and position of the freight forwarder and its relationships with other actors in the air cargo market will be discussed in Section 3. Section 4 will further elaborate on the freight forwarding industry. Although it is believed that consolidation is critical in the industry, the concentration ratios calculated for the industry indicate the opposite. As forwarders are one of the major actors in the air cargo market, the relationship between the forwarder and airport will be discussed in Section 5. Finally, some conclusions and suggestions for further research are given in Section 6.

2. Freight forwarders in a literature review Freight forwarders are a crucial intermediate supplier in the air cargo industry. As Ankersmit et al. (2014) mentioned in their study about ground-handlers, the number of freight forwarders at an airport increased significantly over the last decades. A study by Chu (2014) about the air cargo industry in Taiwan confirmed this view and reported an increase of 105% in international freight forwarding services between 2000 and 2011. MergeGlobal (2008) estimated the revenue share of freight forwarders of total general air cargo to be around 85%, a figure that has been confirmed by Doganis (2019), Hellermann (2006), and Air Cargo World (2001). Small packagesdtraditionally carried by the integratorsdwere excluded from these estimates as it is generally considered to be a separate market segment. Despite their significant role and relevance in the industry, minimal academic literature is dedicated to the understanding of the market dynamics of the freight forwarder industry itself. Nevertheless, the vast

The forwarders’ power play effect on competition in the air cargo industry

363

majority of papers dealing with different topics within air cargo have already confirmed the relevance of these forwarding companies within aviation (e.g., Alkaabi and Debbage, 2011; Chu, 2014; Kupfer et al., 2011; Wan et al., 1998; Wong et al., 2009). An excellent location for their warehouses is a vital choice for freight forwarders to assure efficiency and customer convenience (Wan et al., 1998). Alkaabi and Debbage (2011) added on their part that large airports could more easily attract shippers and freight forwarders through their frequent flight schedules and more sophisticated cargo services. Rodrigue (2012) then observed that forwarders cluster together in the area surrounding major cargo airports. Kupfer et al. (2016) found that the presence of freight forwarders at an airport is a decisive factor in the airport choice of all-cargo aircraft operations as the presence of a wide variety of freight forwarders is perceived to be an indicator of the market size in the (local) area. Wan et al. (1998) defined the role of the freight forwarder in the business by indicating that freight forwarders serve as the middleman for the information flow between airlines and customers, the coordination of the movement of physical goods and the offeringdif asked for by its customersdof some additional services (e.g., labeling, repacking, quality control). This definition wasdamongst othersdsupported by Chu (2014). Whereas the client base in the past was somewhat limited to some large shippers, freight forwarders currently deal with thousands of shippers that prefer a so-called “one-stop total logistics service,” meaning that shippers ask forwarders to ship goods from A to B without taking care about any logistics themselves (Wong et al., 2009). The major decision process in shipping goods by air from A to B is the assignment of the shipments to a particular flight leg (Li et al., 2012). This decision is taken by the freight forwarder, which illustrates its crucial role in the air cargo market. By combining the goods of different shippers, freight forwarders can consolidate shipments, and by doing so, benefit from economies of scale (Wan et al., 1998). Airlines heavily rely on these freight forwarding companies to sell their cargo capacity to the market (Amaruchkul and Lorchirachoonkul, 2011; Gupta, 2008; Kupfer et al., 2016; Li et al., 2012; Wong et al., 2009). Lufthansa, for example, mentioned that 95% of its total air freight is delivered by different freight forwarding companies (Reuters, 2016). Many forwarders have contracts with multiple airlines which assure the forwarder of cargo space on a specific flight of a particular airline (Amaruchkul and Lorchirachoonkul, 2011; Gupta, 2008; Li et al., 2012). Airlines on their part mainly sell their cargo capacity in two stages.

364

The Air Transportation Industry

Typically, a particular amount of cargo capacity is pre-allocated to one or multiple freight forwarders. In the second stage, the remaining capacity is sold on the spot market (Amaruchkul and Lorchirachoonkul, 2011; Gupta, 2008). The choice of the preferred carrier by freight forwarders was investigated by Chu (2014), among others. The study by Chu (2014) indicated that freight forwarders choose an airline based on reliable and ontime air cargo services, offering express cargo shipments, services that are consistent with the contract, achieving a high level of customer satisfaction, and whether the carrier is known for its good reputation. Freight forwarders are mentioned in a vast majority of papers dealing with air cargo, which indicates their crucial role in the industry. Furthermore, it seems that these companies are strong decision-makers in the air cargo market. However, very few research is dedicated to the air freight forwarding industry itself. The dynamics within the air freight forwarding industry on the one hand and with the other actors along the supply chain, on the other hand, is not yet investigated sufficiently. Before looking into these dynamics, section three will elaborate on the business model of the air freight forwarder.

3. The business model of the air freight forwarder To better understand the role of the freight forwarder in the air cargo market and its interaction with the other market players (airlines, airports, etc.), it is necessary to identify the different business models in the industry. An essential characteristic of the air cargo market is that shippers are nearly always dealing with an intermediary to ship goods from A to B. As indicated earlier, shippers rarely contact a carrier directly to fly goods from A to B. Based on this very typical characteristic, two major business models can be distinguished in the air cargo industry. The first business model is built upon the integration of the transport chain by a single company. These fully integrated companies (e.g., DHL, FedEx, and UPS) will pick-up the goods at the shipper, transport them to the airport, fly them to the destination airport, and finally deliver the goods to the consignee; additionally, the integrator will take care of all documentation and customs clearances. The main advantages of this business model are its speed (e.g., next-day delivery) and reliability because all of the processes are done in-house. But this goes hand-in-hand with a premium price, which is often too expensive for shippers that mainly deal with general freight. Historically, the business model of integrators was based on

The forwarders’ power play effect on competition in the air cargo industry

365

the delivery of parcels and documents that need to be delivered as soon as possible. However, these companies also currently offer services for general freight. The second business model is built upon cooperation between the different actors in the transport chain. Instead of integrating all activities within one company, each subsequent step in the chain is executed by another partner who specializes in that specific aspect: road haulage companies transport the cargo from the shipper to the airport, ground-handlers put the cargo on the aircraft, airlines fly the goods to the destination airport, and last-mile delivery companies deliver the goods to the final consignee. The architecture of the entire transport chain is the responsibility of the freight forwarder who acts as a virtual integrator. The freight forwarder takes over the entire logistics process from the shipper. In addition, the forwarder not only organizes the physical transport of the goods but is also responsible for all of the documentation. If asked by the shipper, some value-added activities could also be done, such as labeling, repacking, and quality control. By definition, many different market players are involved in this business model. Therefore, it is generally considered to be slower and less reliable compared to the services offered by the integrators. However, shipping goods via a freight forwarder is supposed to be less expensive. Whereas previously the integrators mainly focused on parcels and documents, freight forwarders were more specialized in general air cargo (Chao et al., 2013; Feng et al., 2015; Neiberger, 2008; Popescu et al., 2011; Schramm, 2012). Fig. 16.1 depicts the differences between the two business models mentioned above. Integrators traditionally operate their own pick-up and delivery vehicles as well as their own aircraft. These companies collect the cargo themselves, sort all the freight at their own warehouse facilities and fly it most of the time on their own aircraft1 toward the destination airport after which they will deliver it to the final consignee. As shown in Fig. 16.1, only one company is responsible for the whole logistics chain, for both the physical goods flow as well as the organization and corresponding information flow. Contrary to the integrators, freight forwarders generally do not operate any vans or aircraft. Therefore, this business model demands the 1

Integrators also book cargo on aircraft (full freighters or in the belly of passenger aircraft) of other carriers. A lack of capacity of the integrator on a particular route (due to too much or too little demand) is one of the reasons that could explain this phenomenon.

366

The Air Transportation Industry

Figure 16.1 Integrator versus freight forwarding business model. (Source: Own composition.)

involvement of many more actors. Once the shipper has selected a particular freight forwarder, the forwarder will instruct a transport company to pick up the goods and deliver them to the warehouses of the freight forwarder (which are located near an airport most of the time). The freight forwarder will consolidate all shipments into some larger consignments. Then, the goods are transferred to the ground-handler who will put the freight on the aircraft of the airline selected by the forwarder. At the destination airport, the ground-handler will present the goods at the warehouses after which the goods can be collected by the transport company to deliver to the final consignee. As air cargo transport is generally transported cross-border and even cross-continental, customs are involved as well. Some goods/ consignments may have to be transferred to the customs warehouses where they can be checked before final delivery to the consignee. The freight forwarder is continuously managing the entire transport process and makes sure that every single actor knows what to do and when. In addition, freight forwarders are responsible for the documentation flow between the different companies involved in this whole process. Fig. 16.1 shows no link between the integrator and freight forwarder. This could suggest that both business models are separated, which is in

The forwarders’ power play effect on competition in the air cargo industry

367

reality not the case, as both are intertwined to some extent; freight forwarders sometimes book freight on the aircraft of integrators. The other way around is also occurring; integrators often book capacity on other airlines in case of capacity shortage to a particular destination.2 Customs are placed centrally in the air transport flow in Fig. 16.1 to show the importance of customs in the airport choice. The better (i.e., fast and competent) customs are at a particular airport, the more likely an airline and freight forwarder will choose this airport for its operations. Also, the integrators have to deal with customs affairs. Still, these companies are organized in such a way that customs clearance will happen in most cases at their own warehouses so that customs is not explicitly shown in the upper part of Fig. 16.1. Considering the global small packages market, together the large integrators (DHL, FedEx and UPS) controlled approximately 75% of the market in 2017 (Doganis, 2019). The forwarders on their part control around 85% of the market for heavier air cargo shipments (Doganis, 2019; MergeGlobal, 2008). Although the position of the freight forwarder in the air cargo industry has already been clearly explained above, the role of these companies in the industry will now be further elaborated on. In the beginning, the role of a freight forwarder was somewhat limited. Most governments were levying customs duties which resulted in the arrival of customs brokers and agents. Such companies acted with government agencies on behalf of shippers. Later on, these brokers and agents evolved into freight forwarding companies which coordinated the whole transport chain. Especially in air transport, shippers often lack the knowledge to send goods efficiently and adequately. Today, these companies have become real logistics service providers which are not just coordinating the whole transport process, but also offer value-added services, such as labeling, repacking, and quality control (Transport Logistics, 2006). One of the essential abilities of freight forwarders is their capacity to consolidate freight. Airlines typically charge by weight and volume, with the chargeable weight being the greater of the gross weight and the volume weight. Therefore, it is advantageous for freight forwarders to consolidate shipments of low-density items and high-density items in such a way that 2

It is difficult to obtain detailed insights into these operations as integrators as well as freight forwarders do not want to reveal information on this due to the commercially sensitive character of such data.

368

The Air Transportation Industry

the overall weight of a consolidated shipment is less than the sum of the weights of the individual pieces. By combining many different smaller shipments into a limited number of bigger shipments, freight forwarders can realize price advantages in the market. Freight forwarders are, therefore, continually trying to optimize their shipments. On top of that, airlines often offer a discount based on volume to freight forwarding companies (Huang and Chi, 2007). Therefore, the more cargo throughput a forwarder has, the lower its (theoretical) average cost per shipment will be. Historically, freight forwarders situated their consolidation centers near the large airports within Europe (e.g., FRA, CDG, AMS, and LHR). In that way, consolidated cargo could quickly be transferred to the groundhandler and time-losses were limited. The benefits of such a location close by major airports were multiple, as was already explained in the literature review. This has also benefitted the airports as the proximity of these freight forwarders were a crucial advantage in attracting the air cargo market. But these hub airports in Europe are becoming capacity constrained. LHR is probably the best example (98% of the maximum allowed flight movements are being used), as the airport has already been capacity constrained for many years (Heathrow Expansion, 2020). More recently, AMS has also had to deal with capacity limitations as the airport reached the agreed limit of 500,000 movements per year (Schiphol Group, 2020). Traditionally, when cargo operations are confronted with these capacity constraints, the relocation of full freighter activities is common. This trend could potentially be disadvantageous for the whole air cargo market at an airport. Although around 50% of total air cargo is flown via the bellyhold of passenger aircraft, full freighter capacity is still considered to be an essential asset for the cargo market. The absence of these flights at an airport are disadvantageous for the cargo market and could be a signal for freight forwarders and shippers to start looking for other airports to accommodate their future expansion plans. In case large freight forwarders begin to expand to other airports and build new warehouses in the surroundings of these airports or even re-locate, the whole cargo segment at the former airport will be under pressure. Therefore, airports that are becoming capacity-constrained have to be aware of the potential harm the decision to downscale full freighter activities may cause. Whereas integrators connect all the activities along the supply chain inhouse, the freight forwarder took a crucial position in the non-integrated air cargo market. Its role of being the coordinator is of utmost importance to this market segment. Although historically freight forwarders were

The forwarders’ power play effect on competition in the air cargo industry

369

only dealing with customs, today these companies are evolved toward real logistics service providers offering essential added-value services as well. By having a clear knowledge of the business model, position and role of the freight forwarder in the air cargo industry, the next section will look into the dynamics of the industry itself.

4. Concentration in the air freight forwarding industry The previous section clearly explained the role and position of the freight forwarder in the air cargo industry. This section will further elaborate on the competition in the air freight forwarding industry itself. Basically, a freight forwarding agency needs a computer and a telephone. Barriers to entry are low, which is supported by the fact that many small freight forwarding companies operate with only a limited number of staff. However, large freight forwarding companies do exist. In the last decade, it is said that a trend toward consolidation has started: DHL Supply Chain & Global Freight Forwarding was created out of Exel and Danzas in 2005, CEVA acquired EGL in 2007, Kuehne þ Nagel acquired ReTrans in 2015, Geodis Wilson acquired OHL in 2015 and very recently, DSV acquired UTi (2016) and Panalpina (2019). A potential explanation for this consolidation trend is the fact that these forwarding agencies were looking for additional sources of revenues. By taking over competitors in the market, growth can be realized, more cargo throughput handled, and economies of scale may occur. Second, economies of scope could be achieved by offering more value-added services, or by broadening the service toward different transport modes (air, maritime, railways). Finally, these mergers and acquisitions may lead to economies of density due to broader networks. At the same time, larger manufacturers were looking for freight forwarders that could arrange the full logistics process for them. The elements that play a role in choosing a particular freight forwarder are among others: expertise across different transport modes (air, sea, and land transport), presence in as many relevant locations as possible, and the ability to identify outsourcing opportunities in a client’s supply chain. It is clear that only the larger forwarders can offer these services (Bowen and Leinbach, 2004; Transport Intelligence, 2015). Therefore, broadening the product portfolio as a freight forwarder and trying to realize economies of scale, scope, and density scope could be a way to survive in the highly competitive market.

370

The Air Transportation Industry

In order to measure the degree of concentration in an industry, several ratios could be calculated. These ratios are traditionally split into absolute and relative ones. To determine the absolute concentration of the air freight forwarding industry in this study, the n-firm concentration ratio and the Herfindahl-Hirschman Index (HHI) are calculated (Laroche et al., 2019; Lipczynski et al., 2005; Sys, 2011). Other indicators (e.g., Hannah and Kay index or the entropy index) require data for the whole industry, and are not available for the air freight forwarding industry whereas the n-firm concentration ratio and the HHI only require data from the largest firms (Laroche et al., 2019; Lipczynski et al., 2005). The relative concentration is often calculated by the Gini-coefficient and the corresponding Lorenzcurve (Laroche et al., 2019; Lipczynski et al., 2005; Sys, 2011). To the best knowledge of the author, no studies have calculated these ratios for the air freight forwarding industry as yet. However, considering other (transport) markets, relevant studies have been conducted by Sys (2011) for the container liner shipping industry, Laroche et al. (2019) for the European rail freight market, and Naldi and Flamini (2014) for various sectors, including aviation. Very recently, a study by Zhang et al. (2020) examined the impact of high-speed rail on the HHI of China’s airline markets. Within the aviation literature, the use of HHI to measure industry competition is common (e.g., Belobaba and Van Acker, 1994; Luo and Lee, 2010; Wang et al., 2018). Each year a list of the 25 largest air freight forwarders based on air freight metric tons and gross revenue is published by Armstrong & Associates Inc. (2020). For the purpose of this study, the concentration on the industry level is measured, and the supply of the air freight forwarding companies is believed to be homogenous.3 Quite surprisingly, Fig. 16.2 shows that the total market share of the top-25 air freight forwarders dropped from 31.49% in 2009 to 25.82% in 2017, thus resulting in less concentration in the global market. This is the opposite of what many industry specialists strongly believe in. However, for these calculations, only 25% of the market is considered, meaning that 75% of the market remains outside the analysis. Many niche markets (e.g., live animals, African market, pharmaceuticals) exist in the freight forwarding industry, and small freight forwarding companies often specialize in a specific niche market. Due to their 3

If value added services are taken into account as well as service, delivery time, etc. the product becomes heterogeneous. However, due to lack of information, the product is believed to be homogeneous (transport of goods via air).

The forwarders’ power play effect on competition in the air cargo industry

35%

371

31.49%

30%

25.82%

31.05%

25% 20%

23.77%

23.60%

17.45%

15% 15.12%

10% 9.88%

5%

Top 4

Top 10

Top 20

2017

2016

2015

2014

2013

2012

2011

2009

2010

0%

Top 25

Figure 16.2 Evolution market share top-4, top-10, top-20, and top-25 freight forwarders in total air cargo market for the period 2009e17. (Source: Own composition.)

specialization, these forwarders can offer a premium service to their customers and can thus survive in the highly competitive market. By interpreting the results of the analysis in the next paragraphs, the fact that 75% of the market is not taken into account should be carefully kept in mind because the analysis cannot be applied to this remaining 75% of the market. Next, the Concentration Ratio Four (CR4) is presented in Fig. 16.3. CR4 is the sum of the market shares of the four largest forwarders in the considered market, here calculated for respectively the top-10 market, top-20 market, top-25 market, and the total freight forwarding market (Lipczynski et al., 2005). The higher the CR4, the higher the concentration in the market. The literature suggests that a CR4 above 25% indicates a

70%

64.06% 56.61%

60% 48.69%

50% 40%

41.57% 48.01% 38.27%

30% 20%

15.12%

9.88%

10%

Top 10

Top 20

Top 25

2017

2016

2015

2014

2013

2012

2011

2010

2009

0%

Total market

Figure 16.3 Evolution of CR4 in top-10, top-20, top-25, and total air freight forwarding market for the period 2009e17. (Source: Own composition.)

372

The Air Transportation Industry

loose oligopoly; in case the index is above 60%, a tight oligopoly will occur with a high risk of market power and collusion between the most significant firms (Martin, 2002; Shepherd, 1999). Whereas the CR4 only looks to the largest firms in the industry, the HHI takes into account the number of firms and the inequality in the market shares (Lipczynski et al., 2005). The higher the HHI, the more concentration in the industry. In a market with perfect competition, the HHI approximates zero, whereas in a monopoly market the HHI would be equal to 10,000. A study by Shepherd (1999) suggests the following benchmark: a market with an HHI below 1000 is unconcentrated and thus fully competitive, while a value between 1000 and 1800 generally indicates moderate concentration. Any value over 1800 indicates a highly concentrated market with companies that could potentially have some market power. Fig. 16.4 depicts the HHI for the air freight forwarding industry (top-25) for the period 2009 to 2017. The largest freight forwarders in terms of air cargo tonnage in 2017 were DHL Supply Chain & Global Forwarding, Kuehne þ Nagel, DB Schenker and Panalpina. Together, these four companies had a market share of approximately 40% of the top-25 in 2017. If the total air cargo market is considered, their market share was only 9.88% in 2017. Compared to 2009, the market share of the top-4 in the top-25 decreased by almost 10%. Moreover, the market share of these top-4 in the total air freight forwarding market dropped by approximately 5%; this clearly indicates less concentration in the industry. The HHI confirms this decreasing concentration trend and the existence of a market with (almost) perfect competition as the index for the top-25 market is far below 1000 (see Fig. 16.4). A decreasing trend over the 4,500

3,986

4,000 3,500 3,000

2,729

2,500 2,000 1,500

1,856 1,199

1,143

1,000 500

720

1,112

623

HHI top 4

HHI top 10

HHI top 20

2017

2016

2015

2014

2013

2012

2011

2010

2009

0

HHI top 25

Figure 16.4 Evolution of HHI for the period 2009e17. (Source: Own composition.)

The forwarders’ power play effect on competition in the air cargo industry

373

period 2009e17 confirms, again, a trend of more competition in the industry. If the data are considered in more detail, DHL Global Forwarding (the market leader) is especially losing its market share. Compared to 2010, the total air freight tonnage handled by DHL Global Forwarding in 2017 dropped by 10%. The most likely explanation for this declining trend is the leading position of DHL in the integrator market, which was, and still is, the primary market segment of the company. Especially in Europe, DHL grew significantly in the integrator market during the last couple of years. As available capacity was probably primarily assigned to integrator activities within the company and less to the freight forwarding activities, the decline in market share in this latter segment is reasonable. Finally, the Lorenz-curve for the top-25 freight forwarders is shown in Fig. 16.5 and shows the evolution in market concentration over time. On the x-axis, the cumulative percentage of the total number of freight forwarders is plotted; the y-axis shows the cumulative air cargo tonnage percentage. A market with perfect competitiondor with equal-sized firmsdwill be depicted by a 45 line. The more inequality in an industry, the more the Lorenz-curve will shift away from this 45 line. Fig. 16.5 suggests that the air freight forwarding industry has evolved toward more competition. Compared to 2009, the Lorenz curve of 2017 moved closer to the 45 line, which means less concentration and more competition in the industrydor at least within the top-25 air freight

100 90

% of air freight tonnes

80 70 60 50 40 30 20 10 0 0

20

40 60 % of top-25 freight forwarders Line of equal size companies

2017

80

100

2009

Figure 16.5 Lorenz-curve of top-25 freight forwarders for 2009 and 2017. (Source: Own composition.)

374

The Air Transportation Industry

forwarders. These results are confirmed by the Gini coefficient, which measures the area between the Lorenz curve and the 45 line. In 2009, the Gini coefficient equaled 0.2608, whereas in 2017, the coefficient was 0.1766. This decrease corresponds with the trend toward less market concentration and more competition. Freight forwarders benefit from the consolidation of different smaller shipments into a limited number of larger consignments. The more cargo a forwarder has, the more possibilities to consolidate. This could potentially explain the consolidation trend in the industry, characterized by many takeovers, mergers and acquisitions over the last few years. However, this study did not found any evidence for more concentration in the business. Rather to the contrary, based on some concentration indices, it must be concluded that the industry has become more competitive. The top-25 freight forwarders are losing market share in favor of companies outside of this top-25. Section 5 will look to the presence of freight forwarders at major European cargo airports.

5. Freight forwarders at major European cargo airports As was illustrated in the previous section, competition in the air freight forwarding industry has increased. This section will further elaborate on the presence of freight forwarders at some major airports in Europe. It is expected that when more freight forwarders are present at an airport, the more freight will pass through the airport, as some studies have suggested that the presence of freight forwarders is a proxy of the size of the local market (e.g., Kupfer et al., 2016). Other studies have indicated that freight forwarders cluster together around major cargo airports (e.g., Rodrigue, 2012). Fig. 16.6 shows the total freight in tons at the largest European cargo airports and the number of freight forwarders4 present at these airports. The larger the dots, the more cargo throughput (in tons) at the airport. The darkness of the dot reflects the number of top-25 freight forwarders are present at the airport: the darker the dot, the more top-25 freight forwarders are present at the airport while an almost white dot corresponds to very few or no freight forwarders from the top-25 list. 4

Only the top-25 freight forwarders in 2018 were considered (based on Armstrong & Associates Inc., 2020); however, it is believed that this is a good proxy for the total number of freight forwarders present at an airport.

The forwarders’ power play effect on competition in the air cargo industry

375

Figure 16.6 Freight in tons versus the number of top-25 freight forwarders at the top-40 cargo airports in Europe in 2016. (Source: Own composition.)

Regarding the total number of freight throughput in tons, the major European intercontinental hubs (FRA, CDG, AMS, and LHR) are performing the best. At the same time, these airports have the highest presence of top-25 air freight forwarders either situated at or near the airport. This is illustrated in Fig. 16.6 by the large red (dark gray in printed version) dots. A great deal of the European air cargo is arriving or departing from one of these airports as these airports accommodate both a significant share of full freighter flights (except for LHR) as well as a large number of intercontinental wide-body passenger flights. The latter have a high frequency of flights toward many different destinations, which offer freight forwarders the essential connectivity needed. The presence of many forwarding companies at these airports may, therefore, be not surprising. The more focused on freight and the fewer passenger activities developed at an airport, the fewer freight forwarders seem to be present. Airports such as LEJ in Germany (DHL hub), LGG in Belgium (FedEx/TNT hub) and CGN in Germany (UPS hub) are often called “integrators hubs” and primarily focus on air cargo. Also, LUX, EMA, and BGY are important European air cargo airports. All of these airports are characterized by high numbers of cargo throughput and at the same time by a limited number of freight forwarders, shown in Fig. 16.6 by a moderate white dot. First, it

376

The Air Transportation Industry

may be the case that freight forwarders do not want to locate themselves at integrator hubs as the competition may be too fierce. Second, freight forwarders rely heavily on the belly space capacity of passenger aircraft to ship their cargo around the world. Given the fact that these airports are mainly focusing on cargo, the number of intercontinental passenger flights is limited, and there is almost no belly space capacity available at these airports. In addition, many forwarders book last-minute cargo space via the spot market. Due to the volatile character of belly space capacity on passenger flights,5 such bookings are typically made on such passenger flights and less frequently on full freighter flights which are more pre-determined. For these reasons, forwarders may prefer a location at an airport which has significant passenger activities and concomitant belly space capacity than at a dedicated cargo airport. The vast majority of airports shown in Fig. 16.6 traditionally focus on the passenger segment but are significantly smaller than the major European intercontinental hubs. Most of the airports are characterized by a rather small or even tiny dot; however, the presence of top-25 freight forwarders at the airports varies. The rationale behind the forwarders’ choice to locate themselves at these airports may be diverse. Geographical coverage may be a good motive as was demonstrated by Crane Worldwide Logistics that opened a new branch office at BRU in 2019 to serve the European market from a strategically positioned airport (Air Cargo News, 2019a). Other forwarders may want to prepare themselves for capacity shortages at the major European intercontinental hubs and look for alternative locations, such as BUD (The Loadstar, 2019). Forwarders may choose a location in order to be close to one of its major clients, as DSV Panalpina did in locating at HEL (Air Cargo News, 2019b). The cargo throughput at these airports is mainly determined by the number of freight forwarders present at or nearby the airport. If these airports want to boost their cargo activities, airport management should actively try to attract forwarders to the airport. Within this last group, three airports are interesting to have a closer look at: BRU, MUC, and MAD. For all three airports, the dot is rather large (within the group) and dark red (dark gray in printed version), representing a significant cargo segment at the airport and a high number of top-25 freight forwarders present. Due to its central location in Western Europe 5

The belly space capacity available for cargo on passenger flights is dependentdamongst othersdon the total number of passengers and associated luggage that has to be taken on the flight.

The forwarders’ power play effect on competition in the air cargo industry

377

and its location between the four major European intercontinental hubs in Europe, BRU has already benefited from capacity constraints at these major hubs. As capacity constraints will undoubtedly worsen at the big four airports, BRU has the potential to boost its cargo segment even further. The high number of forwarders present at the airport also indicates that these companies believe in the airport’s potential in the cargo segment. The situation of MUC is comparable to the one in Brussels. MUC, similar to BRU, is a traditional passenger hub. Due to Lufthansa’s decision to begin operating a dual hub system with its second hub in Munich, not only have passenger numbers increased at the airport, but the network and connectivity of the airport has also improved, especially on long-haul routes (67 intercontinental passenger routes in 2013, 80 in 2017 and 77 in 2018). Typically, these flights are operated by wide-body aircraft which are traditionally well-designed to transport cargo in the bellyhold. Freight forwarders were aware of this increasing role of MUC in the cargo segment, which explains the dark red (dark gray in printed version) color of the dot. MAD is the only European airport with an extensive network (number of destinations as well as frequency) to Latin America. This probably explains why many forwarders have a base at the airport to do business in the Latin American market. For most airports, freight forwarders are crucial. Looking at some selected European airports, it seems that when more freight forwarders are present at an airport, the more cargo throughput is realized. However, this does not apply to one group of airports: the dedicated cargo airports (e.g., LEJ, LGG, CGN). These airports have a minimal number of forwarders present at their airport but are still very busy cargo airports. This can be explained by the role of the integrators at these airports. Most dedicated cargo airports are a major or at least a regional hub for an integrator. Probably, freight forwarders do not want to compete with these integrators at their hub airports because of the dominance of the integrators at these hubs.

6. Conclusions Despite their major role in the business, there has been as yet limited research focused on the role and position of the freight forwarders in the air cargo market. Freight forwarders are involved in almost all air cargo shipments that are not shipped via an integrator, which makes them of major importance in the business. Within aviation, there are two transport chains

378

The Air Transportation Industry

that can be distinguished. The first one is based on integration and is dominated by companies like DHL, FedEx, and UPS. In this instance, the integrator is responsible for the entire transport chain, from pick-up of the goods by the shipper to the transport toward the final consignee, with every single task integrated within one single company. On the other hand, the second transport chain is based on cooperation. Here, the primary role is played by the freight forwarder. The forwarder coordinates the whole chain without doing the physical transport itself which is outsourced to different specialized companies. Therefore, the freight forwarder is often called the “architect” of the transport chain. Because of their coordinating role in the business, freight forwarders are in contact with many different parties: road haulage companies, groundhandlers, customs, airlines, last-mile delivery companies, and so on. The forwarder makes sure every single actor knows what its responsibilities are to deliver the goods at its final destination successfully. Furthermore, freight forwarders can consolidate freight from many different shippers. Economies of scale occur, and a lower average cost per unit (or per parcel) can be reached compared to a situation in which a single shipper arranges the whole transport itself. The art of consolidation is the cornerstone in the success of these forwarding companies. Freight forwarders have not only looked to develop economies of scale, but due to continued competitive pressures, freight forwarders have also started to offer additional addedvalue services (e.g., labeling, repacking, quality control) to broaden their product portfolio and to increase their competitiveness in the business, which ultimately leads to economies of scope. Economies of scale and density are found to be important in the air freight forwarding business. The industry has appeared to consolidate: many mergers and take-overs have occurred in the past, which may indeed indicate a consolidation trend. Furthermore, the air freight forwarding business consists of thousands of companies. The vast majority of these companies are rather small focusing on specific niche markets such as live animals or a particular geographical market. Contrary to what is believed by many industry specialists, different concentration ratios indicate that the freight forwarding industry is not becoming more concentrated but rather the opposite. Compared to 2009, the market is getting more competitive, at least within the top-25. The HHI is far below 1,000, indicating a competitive market and has dropped even further over the period 2009e17, which means more competition in the industry. Looking at the Lorenz-curve, the same can be observed: the

The forwarders’ power play effect on competition in the air cargo industry

379

industry became more competitive over the period 2009e17. DHL Global Freight Forwardingdthe most significant air freight forwarderdhas mainly lost market share over the period 2009e17; however, this is probably because of its significant position in the integrator market. For most European airports, freight forwarders play a significant role. FRA and CDG, for example, benefit from their broad range of intercontinental passenger flights and the considerable number of full freighter flights to facilitate the cargo market. Furthermore, many forwarders started activities at these airports, which once again supported the air cargo market. At the dedicated cargo airports in Europe, the integrators play a major role. Because of the strong position of the integrators at many of these dedicated cargo airports, forwarders are not really interested in these airports. While the vast majority of European airports is primarily focusing on passengers, they also accommodate significant cargo activities. For these airports, forwarders were, and still are, essential to develop the cargo segment. The more forwarders, the more cargo is attracted to the airports. In case the major European intercontinental hubs become more capacity constrained shortly (e.g., AMS), these airports will be viable alternatives. Freight forwarders do not want to reveal too much information: neither on their relationships with other actors, nor on their actual air cargo throughput. Better and more reliable data would help to position the freight forwarder more precisely in the industry. The industry concentration ratios for the air freight forwarding industry, for example, are calculated based on data from Armstrong & Associates Inc. (2020). Although the data are widely used, the data collection method remains unclear. Moreover, the data are only available for the top-25 companies. More data would be helpful to make more accurate calculations. Further research could focus on quantifying the air freight forwarding industry to provide insights into its overall importance in the air cargo market. Similar analyses for the American and Asian market could be suggested as well. Also, the power of large freight forwarding companies in the industry would be interesting to assess. Last but not least, the many take-overs, mergers and acquisitions could be analyzed to gain a better understanding of the dynamics in the industry.

References Air Cargo News, 2019a. Crane Worldwide Logistics Opens Brussels Office. Air Cargo News, 2019b. DSV Panalpina to Provide E-Commerce Logistics for Marimekko.

380

The Air Transportation Industry

Air Cargo World, 2001. The 2001 MergeGlobal Air Cargo World Forecast. Alkaabi, K.A., Debbage, K.G., 2011. The geography of air freight: connections to U.S. metropolitan economies. J. Transp. Geogr. 19, 1517e1529. https://doi.org/10.1016/ j.jtrangeo.2011.04.004. Amaruchkul, K., Lorchirachoonkul, V., 2011. Air-cargo capacity allocation for multiple freight forwarders. Transp. Res. Part E Logist. Transp. Rev. 47, 30e40. https:// doi.org/10.1016/j.tre.2010.07.008. Ankersmit, S., Rezaei, J., Tavasszy, L., 2014. The potential of horizontal collaboration in airport ground freight services. J. Air Transp. Manag. 40, 169e181. https://doi.org/ 10.1016/j.jairtraman.2014.07.005. Armstrong & Associates Inc., 2020. A&A’s Top 25 Global Freight Forwarders List [WWW Document]. URL. https://www.3plogistics.com/3pl-market-info-resources/3pl-marketinformation/aas-top-25-global-freight-forwarders-list/. (Accessed 10.14.19). Belobaba, P.P., Van Acker, J., 1994. Airline market concentration. An analysis of US origindestination markets. J. Air Transp. Manag. 1, 5e14. https://doi.org/10.1016/09696997(94)90026-4. Bowen, J., Leinbach, T., 2004. Market concentration in the airfreight forwarding industry. Tijdschr. voor Econ. en Soc. Geogr. 95, 174e188. Chao, C.C., Lirn, T.C., Shang, K.C., 2013. Market segmentation of airline cargo transport. Serv. Ind. J. 33, 1672e1685. https://doi.org/10.1080/02642069.2011.638920. Chu, H.C., 2014. Exploring preference heterogeneity of air freight forwarders in the choices of carriers and routes. J. Air Transp. Manag. 37, 45e52. https://doi.org/10.1016/ j.jairtraman.2014.02.002. Doganis, R., 2019. Flying Off Course, Fifth. ed. Routledge, New York. Feng, B., Li, Y., Shen, Z.J.M., 2015. Air cargo operations: literature review and comparison with practices. Transp. Res. C Emerg. Technol. 56, 263e280. https://doi.org/ 10.1016/j.trc.2015.03.028. Gupta, D., 2008. Flexible carriereforwarder contracts for air cargo business. J. Revenue Pricing Manag. 7, 341e356. https://doi.org/10.1057/rpm.2008.29. Heathrow Expansion, 2020. Heathrow Today [WWW Document]. URL. https://www. heathrowexpansion.com/uk-growth-opportunities/facts-and-figures/ (Accessed 10.22.19). Hellermann, R., 2006. Capacity Options for Revenue Management - Theory and Applications in the Air Cargo Industry. Springer Berlin Heidelberg, Berlin. Huang, K., Chi, W., 2007. A Lagrangian relaxation based heuristic for the consolidation problem of airfreight forwarders. Transp. Res. C Emerg. Technol. 15, 235e245. https://doi.org/10.1016/j.trc.2006.08.006. Kupfer, F., Goos, P., Kessels, R., Van de Voorde, E., Verhetsel, A., 2011. The Airport Choices in the Air Cargo Sector e A Discrete Choice Analysis of Freighter Operations. Kupfer, F., Kessels, R., Goos, P., Van de Voorde, E., Verhetsel, A., 2016. The origindestination airport choice for all-cargo aircraft operations in Europe. Transp. Res. Part E Logist. Transp. Rev. 87, 53e74. https://doi.org/10.1016/j.tre.2015.11.013. Laroche, F., Sys, C., Vanelslander, T., Van de Voorde, E., 2019. Assessing competition in the European rail freight market: is there an oligopoly? Int. J. Transp. Econ. XLVI. Li, Z., Bookbinder, J.H., Elhedhli, S., 2012. Optimal shipment decisions for an airfreight forwarder: formulation and solution methods. Transp. Res. C Emerg. Technol. 21, 17e30. https://doi.org/10.1016/j.trc.2011.08.001. Lipczynski, J., Wilson, J., Goddard, J., 2005. Industrial Organization: Competition, Strategy, Policy, second ed. Prentice Hall, Harlow. Luo, J., Lee, L.W., 2010. Supramonopoly: theory and evidence from the US air passenger service markets. Int. J. Econ. Bus. 17, 405e426. https://doi.org/10.1080/ 13571516.2010.513817.

The forwarders’ power play effect on competition in the air cargo industry

381

Martin, S., 2002. Industrial Economics, Economic Analysis and Public Policy. Blackwell, Oxford. MergeGlobal, 2008. Forwarder Momentum e Opportunities for Value Creation in Freight Forwarding. Naldi, M., Flamini, M., 2014. The CR4 Index and the Interval Estimation of the Herfindahl-Hirschman Index: An Empirical Comparison. HAL Arch. https://doi.org/ 10.2139/ssrn.2448656. Neiberger, C., 2008. The effects of deregulation, changed customer requirements and new technology on the organisation and spatial patterns of the air freight sector in Europe. J. Transp. Geogr. 16, 247e256. https://doi.org/10.1016/j.jtrangeo.2007.09.003. Popescu, A., Keskinocak, P., al Mutawaly, I., 2011. The air cargo industry. In: Intermodal Transportation: Moving Freight in a Global Economy, pp. 208e237. Reuters, 2016. Air Cargo Firms Feel the Pressure from Plunge in Freight Prices. Rodrigue, J.-P., 2012. The geography of global supply chains: evidence from third-party logistics. J. Supply Chain Manag. 48, 15e23. https://doi.org/10.1111/j.1745493X.2012.03268.x. Schiphol Group, 2020. Network of Destinations [WWW Document]. URL. https://www. annualreportschiphol.com/our-results/network-capacity-and-security/network-ofdestinations. (Accessed 10.22.19). Schramm, H.-J., 2012. Freight Forwarder’s Intermediary Role in Multimodal Transport Chains e A Social Network Approach. Physica-Verlag. https://doi.org/10.1007/9783-7908-2151-2. Shepherd, W., 1999. The Economics of Industrial Organization. Waveland Press, Illinois. Sys, C., 2011. Inside the Box: Assessing the Competitive Conditions, the Concentration and the Market Structure of the Container Liner Shipping Industry. University of Antwerp. The Loadstar, 2019. Cargo City Opens at Budapest Airport: A “Real Game-Changer” for Air Freight. Transport Intelligence, 2015. The Future of Logistics e what Does the Future Hold for Freight Forwarders? Bath, UK. Transport Logistics, 2006. The Changing Role of the Freight Forwarder 1e8. Wan, Y.W., Cheung, R.K., Liu, J., Tong, J.H., 1998. Warehouse location problems for air freight forwarders: a challenge created by the airport relocation. J. Air Transp. Manag. 4, 201e207. https://doi.org/10.1016/j.apergo.2003.11.013. Wang, K., Zhang, A., Zhang, Y., 2018. Key determinants of airline pricing and air travel demand in China and India: policy, ownership, and LCC competition. Transp. Pol. 63, 80e89. https://doi.org/10.1016/j.tranpol.2017.12.018. Wong, W.H., Leung, L.C., Hui, Y. Van, 2009. Airfreight forwarder shipment planning: a mixed 0-1 model and managerial issues in the integration and consolidation of shipments. Eur. J. Oper. Res. 193, 86e97. https://doi.org/10.1016/j.ejor.2007.10.032. Zhang, Q., Yang, H., Wang, Q., Zhang, A., Zhang, Y., 2020. Impact of high-speed rail on market concentration and Lerner index in China’s airline market. J. Air Transp. Manag. 83. https://doi.org/10.1016/j.jairtraman.2019.101755.

This page intentionally left blank

CHAPTER 17

Fuel hedging: how many games can we play? Carlos Filipe Marques

Faculty of Business and Economics, Antwerp, Belgium

1. Fuel costs’ relevance in aviation Despite continued technological evolution, significant limitations render aviation essentially hostage to using fossil fuels, making it one of the most challenging transportation sectors to decarbonize. Until 2000, the range of USD22e28 remained the OPEC reference for pricing a barrel of oil. Oil prices would exceed USD30/bbl only under Middle East’s circumstantial conflicts. Following several decades of relative stability, oil prices surged following the September 11 terrorist attacks in the United States. The swings in oil price from 2000 until 2020 were remarkable, as captured in Fig. 17.1. This context has established the backdrop for multiple aviation mitigation strategies to reduce the sector’s exposure and vulnerability. A defining period for aviation came when oil price surged to around USD150/bbl in 2008. Jet fuel’s share of total operating costs grew from 13% in 2001 to 33% in 2008 (Martino et al., 2009), ranging from 10% to 45% for airlines in the United States (Pyke and Sibdari, 2018). Increasing recognition of the limits associated with the economics of exhaustible resources emerged then as a plausible explanation for oil price surge and fluctuations. An imminent shortage of energy supplies to sustain growing demand got increasingly accepted as fact, with oil’s scarcity rent finally coming into play. This conviction gained ground among major business stakeholders and governments, with the annual IEA 2008 report stating that “. it is becoming increasingly apparent that the era of cheap oil is over” (IEA/OECD, 2008). Jetfuel cost predictability turned then into a major challenge for airlines, unable to raise airfares due to intense competition, especially on the passenger side (Morrell and Swan, 2006). This situation often prevents airlines from reflecting swiftly higher jet fuel costs in their fares. Even if fuel hedging had been adopted by aviation since the late 1980s, it gained more momentum from the early 2000s, upon such increased oil price volatility. By negotiating futures’ contracts to lock in The Air Transportation Industry ISBN 978-0-323-91522-9 https://doi.org/10.1016/B978-0-323-91522-9.00020-8

© 2022 Elsevier Inc. All rights reserved.

383

384 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0 -10 -20 -30 -40 -50

The Air Transportation Industry

WTI

Brent

Figure 17.1 West Texas Intermediate (WTI) and Brent oil price (USD) per bbl (2000e20). (Based on data set obtained from https://www.eia.gov.)

what they believed were more favorable prices for a portion of forecasted needs, airlines wanted to avoid the impacts of sudden oil price swings on their bottom line. The remaining of this chapter will look at whether this strategy did work as expected.

2. Fuel hedging fundamentals Market fundamentals associated with fuel hedging is that the expected value of a fuel hedge is zero, with bets on oil evenly balanced between sellers and buyers. The oil market is characterized by high liquidity attracting multiple industries and stakeholders, rendering airlines’ role a relatively minor one, with the market price representing a well-examined consensus (Morrell and Swan, 2006). Despite reported cases where hedging may have returned circumstantial profits for airlines, they do not deliberately take an active role in the oil speculation market for financial gains. They primarily wish to increase earnings predictability through tight cost risk management. The second fundamental of hedging is that investors look for paying only for reductions in market risk, not to reward with stock valuation. Fuel risk hedging involves taking a position in a derivative instrument that gives an equal and opposite financial exposure to the underlying physical position to protect against fuel price changes (Dontwi et al., 2010). Fuel hedging uses financial instruments strategically to offset the risk of adverse oil price movements, betting on expectations that can bring some

Fuel hedging: how many games can we play?

385

degree of predictability to the business cost baseline. This means locking in the cost of future fuel purchases within a specific timeframe, protecting against sudden cost increases from rising fuel prices or fluctuations. This entails buying or selling the expected future oil price through a range of investment products. Like other nonfinancial firms, airlines use derivatives as a risk mitigation tool (Lim and Hong, 2014). Airlines expect to eliminate price exposure for a specified period by hedging their jet fuel needs. Such predictability of expected cash flows is supposed to give companies a competitive edge with better operational planning during volatile periods by locking in prices for future input requirements. However, fuel hedge positions can also significantly impact operating cash flows, given the corresponding cash collateral requirements associated. 2.1 Instruments A derivative transaction is a bilateral contract whose value is derived from some underlying asset’s value (Dontwi et al., 2010), such as a commodity (e.g., oil). The market can be divided into exchange-traded derivatives and over-the-counter (OTC) derivatives, that is, traded directly without an exchange’s supervision. Energy options are traded on the regulated futures exchanges and the unregulated OTC market. While airlines are interested in protecting themselves against jet fuel price fluctuations, the fact is that Jet fuel itself is scarcely traded in organized exchange markets. Apart from a limited Japanese market, there are no exchange-traded futures available in aviation fuel, which must be otherwise negotiated OTC, involving counterparty risk for both sides. Hence, depending on an airline’s financial situation, it might be difficult to find counterparts willing to take risks entailed in OTC contracts. The most closely related product in a liquid market is crude oil, Brent, and US West Texas Intermediate (WTI) crude. Airlines, therefore, tend to hedge oil prices as a proxy for its refined products, through futures markets or other financial derivatives such as options, collars, and swaps, or a combination of some of these instruments (Morrell and Swan, 2006). However, crude might not always be an ideal hedge against jet aviation price increases. This happens in circumstances of greater volatility in crude prices when the root cause is a war or a threat of a conflict due to increased military use of refined fuels with a switch of production fractions at refineries (Morrell and Swan, 2006). Regarding the choice of instruments or derivatives to hedge fuel price, decisions will

386

The Air Transportation Industry

depend on the airline’s strategy, having a significant impact on the financial outcome in case of unfavorable swings in oil price. Especially with forward contracts, compared to options, which losses are limited to the premium paid by the hedger for the respective option (EuroFinance, 2020a). 2.1.1 Futures Most fuel hedges imply purchases of an oil future as a cash bet on oil price on a particular date. In simple terms, and disregarding fees and other contractual costs associated, an airline may, for instance, acquire a future oil contract for a certain quantity at, for example, USD50/bbl. If the oil price soars to USD60/bbl, that contract will allow the company to offset a jet fuel price increase by USD10/bbl. Futures are agreements between parties to buy or sell oil at a particular time in the future and at a specific price. Futures are a popular risk management standardized instrument executed on an exchange, with predetermined oil quantities (e.g., 1000 bbl) and are characterized by high liquidity. In practice, futures’ oil contracts don’t necessarily result in delivery. One of the parties agrees to deliver a given quantity of oil at an agreed strike price on a specific future date. These are often reversed on the due date, so no physical delivery occurs. Less than 1% of trades result in delivering the underlying commodity (Morrell and Swan, 2006). Forward contracts are similar to Futures but traded on OTC markets, with customized terms negotiated (quantity, settlement process, etc.), bringing up the risk that the counterparty may not be able to honor the contract. This situation can put airlines under pressure for setting aside collateral that may be challenging to provide. 2.1.2 Derivatives Derivatives consist of an option is a right to buy or sell at a specific date at a stated “strike” price, both above and below current futures prices. Options are different from other trading instruments by allowing the flexibility to buy or sell fuel under specific conditions during an agreed period, but not the obligation. Hence, an option contract will only be exercised if the market moves favoring the option holder. If exercised, it will result in a corresponding futures position. If the company decides not to exercise the option, there is a specific premium cost to pay; hence options are more flexible than forwards or futures, although that flexibility comes at a cost.

Fuel hedging: how many games can we play?

387

The primary option positions are: (i) buy a call, that is, an airline can purchase an option to buy the underlying commodity or asset, by granting the holder the possibility to protect against a price rise, simultaneously giving the opportunity not to exercise and take advantage of a market price decline; (ii) buy a put, that is, purchase an option to sell the underlying commodity or asset, (iii) sell a call, that is, settle a contract that gives the other party the option to buy the underlying commodity or asset, and (iv) sell a put, that is, settle a contract that gives the other party the chance to sell the underlying commodity. When a “call” and a “put” option are combined, it creates a “collar.” Collars became popular with airlines for locking in the price for fuel between two known predictable values, limiting the speculative risk to a predetermined range of prices (Morrell and Swan, 2006). While the “call” protects the airlines from oil prices soaring beyond the strike price above the current future, the “put” determines the possibility of selling at another strike price, below the current future. The total cost of a collar is calculated as the call option premium minus the put option premium received. Swaps are privately negotiated transactions between two parties. They are customized to the two parties’ specific need, such as price, quantity, commodity, location, and settlement type. This way, an airline can buy a swap for 1 year at a specific strike price, assuming predetermined jet fuel quantities per month, similarly to a forward contract. Each month’s spot prices will then be compared with the swap price settled. 2.1.3 Long versus short hedging A long hedger is when an airline holds the short cash position (buys physical commodity) and the long future/forward position in the underlying commodity. The counterpart is designated as a short hedger. This will be an intermediary of the commodity transaction, holding a long cash position (sells the physical commodity) and the corresponding short future/forward position in the underlying commodity. Consequently, if an airline wants to hedge the risk of rising oil prices, it will take a “long position” and buy future/forward contracts to hedge rising commodity prices (Table 17.1). The counterpart will sell the commodity (short position) future/forward contracts, hedging the opposite risk of falling oil prices (Tokic, 2012).

388

The Air Transportation Industry

Table 17.1 Hedging strategies. Long hedge

• Buy commodity futures • Needs the commodity for future consumption • Hedging rising commodity prices

Short hedge

• Sell commodity futures • Has to sell the commodity from the inventory in the future • Hedging falling commodity prices

Reproduced from Tokic, D., 2012. When hedging fails: what every CEO should know about speculation. J. Manag. Dev. 31, 801e807. https://doi.org/10.1108/02621711211253259.

3. Hedging in reality 3.1 Introduction Although fuel hedging as a risk management strategy started to be adopted by aviation in the late 1980s (Morrell and Swan, 2006), it gained a more prominent role from the early 2000s to protect the business from the increasing oil price trends and volatility. But it also prevented savings when fuel prices drop against the expectations, putting airlines in a vulnerable situation, with unhedged competitors reaping the benefits of cheaper oil (Merkert and Swidan, 2019; Morrell and Swan, 2006). This dilemma begs the question of whether financial hedging is a sensible strategy. The extent of hedging will depend on airline expectations of future prices. Contracts are typically short-term, no longer than 1 year, with few hedges at more than 1 year ahead and almost none beyond 2 years (Lim and Hong, 2014; Morrell and Swan, 2006). Fuel hedging usually comes as part of a broader integrated risk management design to improve its flexibility to react to changing fuel costs. This often covers multiple vectors such as operational hedging, that is, adjusting to flight plans and operations, promoting fleet diversity, fleet fuel-efficiency, and optimizing capacity. Also, the hedging of foreign exchange rates as airlines operate in an international context subject to fluctuations in foreign currencies or USD dominated debts (e.g., related to aircraft acquisition). Fig. 17.2 depicts the options for airlines around realworld strategic fuel risk hedging, across financial, operational, and currency exchange hedging. 3.2 Fuel hedging drivers This section covers multiple relevant studies looking into the key factors driving airlines to hedge, discussing its actual benefits and effectiveness.

Fuel hedging: how many games can we play?

389

Jet Fuel Hedging

Natural FX Hedge

Passively

Enforced

Operational Hedging

Financial Hedging

Instruments

Products

Horizon

Quan ty

Short Term

Fares/Surcharges

Mid Term

Long Term

Capacity

Alterna ve Fuels

Network

Technology

Fleet

Figure 17.2 The spectrum of strategic jet fuel hedging. (Reproduced from Merkert, R., Swidan, H., 2019. Flying with(out) a safety net: financial hedging in the airline industry. Transp. Res. Part E Logist. Transp. Rev. 127, 206e219. https://doi.org/10.1016/j.tre.2019.05. 012.)

3.2.1 Company valuation An attempt to understand why nonfinancial firms hedge and whether it allows increasing a firm’s value was conducted by (Carter et al., 2006). Hedging of jet fuel price exposure by US airlines was evaluated from 1992 to 2003 to look for empirical evidence on the source of value from hedging operations. Results suggested that the firm value is positively associated with the amount of fuel hedging and that hedging provides an additional cash source for making acquisitions during periods where higher fuel costs would otherwise limit the ability to invest. These same authors (Carter, Rogers, and Simkins) have more recently coauthored the evaluation of new evidence from the US airline industry based on date expanding until 2008 (Treanor et al., 2014). While still finding evidence of a positive hedging

390

The Air Transportation Industry

premium, no evidence could be found to support that a firm’s value would change with the interaction of fuel hedging and risk exposure. An IATA report published in 2018 highlighted that seven decades of data analysis since the 1950s could actually not prove a relationship between oil prices and industry profitability (IATA, 2018a). The number and nature of factors influencing stock returns remain one of the most challenging research areas in financial economics, including whether oil price risk affects stock returns. While the direct impact of oil price risk on equity level returns is not clearly known, it is acknowledged that oil price fluctuations possibly affect stock returns through multiple indirect channels. These can be lowered earnings expectations due to increasing production costs or the increased discount factor due to rising inflation and interest rates (Azimli, 2020). Hence, and despite the range of theories on the impact of hedging, there is still today little empirical evidence of the effects on firm value (Nelson et al., 2005). Findings, therefore, suggest that the economic fundamentals of hedging hold valid in the real world, whereas permanent hedging of fuel costs should leave expected long-run airline’s profits unchanged (Morrell and Swan, 2006). The exception that may justify an airline hedging fuel might be when nearing bankruptcy by protecting the business from reaching a tipping point when at risk of a sudden loss of value, despite challenges in accessing the required collateral’s liquidity at such a vulnerable moment. 3.2.2 Mitigate volatility Financial hedging’s effectiveness seems to diminish during the global financial crisis, rendering the benefits of financial hedging questionable during such periods of heightened volatility (Laing et al., 2020). Other authors (Lim and Hong, 2014) have empirically examined the role of fuel hedging in reducing airlines’ operating costs used US airlines data from 2000 through 2012. Findings suggest that hedging had a negative but statistically negligible impact on operating costs after controlling for cost inefficiencies. The authors concluded that the prices paid by fuel hedging firms were lower yet not statistically different from fuel prices paid by nonhedging companies after accounting for cost inefficiencies. While airlines with fuel hedging programs could reduce fuel prices marginally, they did remain susceptible to fuel cost volatility, just like those that don’t hedge. Likewise, several other authors have analyzed risk exposure among US airlines, hedging strategies and the actual value derived out of those

Fuel hedging: how many games can we play?

391

strategies (Turner and Lim, 2015; Korkeamäki et al., 2016;Treanor et al., 2014). Daily data between 1994 and 2013 was analyzed by Turner and Lim (2015) to determine the minimum variance hedge ratio for airlines wishing to hedge jet fuel price risk with futures, looking into hedge effectiveness to establish the best cross-hedging asset. But results beyond a 3-month maturity were unclear and none of the cross-hedge proxies could be proven to be highly effective. At a global scale, and looking back into quantitative data around financial and operational hedging strategies Berghöfer and Lucey (2014) examined data from 64 airlines between 2002 and 2012: 20 Asian, 20 European, and 24 North American companies. USA airlines were the most exposed, followed by Asian carriers, with European airlines being the less exposed ones. It was noted that highly exposed airlines tend to have higher leverage from larger sizing, with a higher degree of operational hedging through fleet diversification and a higher load factor than low-exposed airlines. Yet, they did not differ significantly in their portion of fuel hedged or periods of hedging. Therefore, the hypothesis that financial hedging reduces exposure was rejected, even when the model was applied within regions. Such conclusions suggest that fuel hedging is significantly less effective than operational hedging in mitigating oil price volatility exposure. Findings could not confirm that fuel hedging could dampen out airlines’ exposure to volatility, with some data actually suggesting otherwise. 3.2.3 Management competence Fuel hedging was also pointed out as a way to show the executive team’s savviness to the shareholders, a signal of management competence, making it probably one of the most compelling arguments for airline’s hedging (Morrell and Swan, 2006). When oil spiked above USD150/bbl in 2008, and aviation investors realized that Southwest’s aggressive and successful hedging program was not replicated by other airlines, there was a strong critic. Consequently, many have partially blamed the poor financial performance on rising fuel costs not being sufficiently hedged (MarketWatch, 2008). Corporate risk management theories also suggest that managerial risk aversion might be an incentive for corporate hedging, with management bonus often tied to profits and stock prices (Stulz, 1996). An explanation on why firms hedge was described by (Breeden and Viswanathan, 2016), suggesting that hedging occurs when “higher ability managers are substantially

392

The Air Transportation Industry

different from lower ability managers, or the costs of hedging are low”. This, in turn, drives “the casual belief that hedging locks up higher profit opportunities.” Last but not least, people’s “intuitive expectations are governed by a consistent misperception of the world” (Tversky and Kahneman, 1971). Known as the Confirmation Bias (CB), defined as “an agent’s tendency to seek, interpret and use evidence in a manner biased toward confirming her existing beliefs or hypotheses” (Charness and Dave, 2017). Confirmation Bias has been extensively studied by behavioral economics and is acknowledged as a trap in trading for the risk of creating the reinforcing conviction, built on past successes that there is a way to beat the market, overlooking future risk for following an emotional rather than a logical approach. 3.3 Challenges 3.3.1 Contractual costs Fuel hedging comes at the cost of entailed guarantees. Acquiring oil price forward contracts, bonds or lines of credit may come at prohibitively costs for the business to cover for any possible losses incurred if the bet goes against the airline (Morrell and Swan, 2006). Indeed, many airlines may not even have the necessary cash collateral and credit conditions to engage in hedging strategies (Merkert and Swidan, 2019). Alternatively, airlines can buy a “call” option that pays off above some upper bound on oil prices. But even that cost might not always be feasible. When oil price risk hedging becomes more costly to the airline than to the oil market, airlines may find themselves in a situation of not being to hedge fuel risks at all. Recalling here that, ultimately, the reduction in risk for the airline is worth only the contract’s costs entailed (Morrell and Swan, 2006), which might be a challenge in itself. Some of these authors (Swidan et al., 2019) have more recently tried to quantify the benefits and the costs associated with fuel hedging in aviation based on a historical simulation based value at risk (VaR) model. By estimating the cash collateral required to support jet fuel hedging using exchange-traded futures, the authors have concluded that although hedging reduces risk exposure, the collateral costs required may offset the potential financial benefits. Cash collateral required was found to be “a key limiting factor in hedging decisions of airlines, even for financially stable airlines” (Swidan et al., 2019). Moreover, this effect grows steeper when companies hedging ratio surpasses 60%e70%, despite a marginal reduction in market risk exposure.

Fuel hedging: how many games can we play?

393

3.3.2 Accounting standards Following the period immediately after the increase in oil price volatility observed from the early 2000s, several airlines increased their fuel hedging practices globally. Soon after, most major passenger airlines in the USA, Europe, and Asia were hedging at least part of their future fuel needs (Morrell and Swan, 2006). While not many airlines were then reporting gains and losses from fuel hedging activity, they were, however, required to start informing the market value of unexpired contracts on their balance sheets. Although airlines supposedly looked merely to flatten cost volatility, something else could have been at play, tied not to economics but rather to the possibility of moving profits forward or backwards in their balances by timing the sale of oil futures. Hence, by conveniently moving costs across quarters, companies could overstate profits by understating the degree of otherwise immediate losses. But the stringent US GAAP (Generally Accepted Accounting Principles) set hedge effectiveness rules that require derivatives to cover a minimum and maximum proportion of underlying risk exposure. Companies whose hedges fail are obliged to report the mark-to-market1 swings on the derivatives in their income statement (EuroFinance, 2020a). The risk of fuel hedging creating additional earnings volatility is actually worsened if fuel derivatives do not qualify for hedge accounting under GAAP standards, for example, if significant fluctuation in energy market prices is observed (Southwest, 2020). Southwest Airlines, an exception among large US companies for its continued engagement in fuel hedging, acknowledges such complexity in its 2019 Annual Report, stating that “accounting pronouncements pertaining to derivative instruments and hedging are complex with stringent requirements, including the documentation of a Company hedging strategy, statistical analysis to qualify a commodity for hedge accounting both on a historical and a prospective basis, and strict contemporaneous documentation that is required at the time each hedge is designated” (Southwest, 2020). In Europe, the adoption of IFRS (International Financial Reporting Standards) allows mark-tomarket changes in derivative values to be kept out of income statements, especially regarding option contracts, which may drive airlines under IFRS to consider hedging. Although there is an ongoing convergence of Global Accounting Standards, the USA is the only country yet to switch to IFRS (Dotzlaw et al., 2020). 1

Accounts in a futures contract are marked to market on a daily basis, with profit and losses calculated between the long and short positions.

394

The Air Transportation Industry

3.3.3 Predictive modeling The attempt to understand the behavior of oil prices (Hamilton, 2009) has described three approaches to explain oil price changes: through statistical investigation of historical data, predictions of economic theory and fundamental determinants and prospects for demand and supply. Predictive econometric modeling underpinning fuel price evolution and hedging decisions come under two different types: structural models, relying on fundamental data such as demand and supply, implemented through the use of linear regression, factoring in explanatory variables beyond just the past data of oil prices. The other type of model relies on time series to determine future price changes. By capturing time-varying volatility, econometric models have improved oil price forecasting accuracy. Yet, there is a fundamental flaw in that they assume data to be stationary, regular, and linear. However, the reality concerning oil price prediction is complex, irregular, and nonlinear (Wang et al., 2018). There are multiple underlying variables at play, including nonlinear effects. For instance, the geopolitical events threatening the security of supply, demand elasticities, ambiguous, or imperfect information about the accurate stock of global oil resources and the complexity of the commodity and market mechanisms entailed (Kjärstad and Johnsson, 2009). Predictions for oil price are therefore predominantly focused on time-boxed periods. Artificial intelligence (AI) models are today suggested having the potential to tackle such constraints (Abdollahi, 2020). Oil price predictive modeling does remain an active topic of research, stirred by artificial intelligence developments (Abdollahi, 2020; Abdollahi and Ebrahimi, 2020; Bristone et al., 2020; Kristjanpoller and Minutolo, 2016; Li et al., 2020; Mostafa and El-Masry, 2016; Tontiwachwuthikul et al., 2020; Wang et al, 2018, 2020; Wu et al., 2021; Yu et al., 2014). Artificial intelligence has the potential to address the drawbacks of econometric methods that assume data to be stationary, regular and linear, thus failing to capture nonlinear characteristics. Leapfrogging advancements are not far-fetched futuristic scenarios (Abdollahi, 2020). The association of AI with Quantum Computing, a concept known as quantum artificial intelligence (QAI), while still some 5e10 years ahead, promises however disruptive wideranging implications across new, powerful applications (Ghose, 2020; Taylor, 2020), with near term value expected from finance and energy sectors’ applications (Ménard et al., 2020). Big banks are now investing in Quantum expertise, also. IBM is deep into quantum computing and has plans for a 1000-qubit machine ready by 2023. Google might have a

Fuel hedging: how many games can we play?

395

million qubits machine ready by the end of the decade. But for specific applications, using simpler algorithms, such revolution may come much earlier. And “whoever makes a breakthrough first may choose to reap the rewards in obscurity, rather than broadcast the fact to the world . that is how high-frequency trading got started” (The Economist, 2020a). 3.4 Regional perspective The impact of elevated jet fuel prices in 2018 was reportedly one of the industry’s key challenges (IATA, 2018b). The following chart represents the percentage of fuel consumption hedged by several airlines globally in 2019. Several airlines hedged their fuel consumption needs heavily in the first semester of 2019, while others had decided to not hedge at all. Indeed, heavy losses created in times of dropping oil prices led several airlines to ditch their fuel hedge programs altogether, starting with major US carriers. Notably, American Airlines terminating its hedging program entirely after oil prices plunged in 2014, with its president stating at the time that “hedging is a rigged game that enriches Wall Street” (W.S.J., 2016). And later in 2016, also Delta Airlines upon accumulating around USD8 billion in fuel hedging losses. Hence, fuel hedging has remained more popular in the last years among major European and Latin American carriers, hedging a significant portion of their planned needs (Fig. 17.3). Notably, Ryanair and Lufthansa, in

Figure 17.3 Major Global Airlinesdfuel consumption % hedged for 2019H1. (Reproduced from IATA, I.A.T.A., 2018b. FUEL HEDGING: LARGE DIFFERENCES BY REGION, FEWER BY MARKET.)

396

The Air Transportation Industry

2018, locking in 90% and 77% respectively of their expected 2019H1 fuel needs. Except for Southwest Airlines, still aggressively hedging some 60% of their fuel needs, most other major US airlines have been keeping their “nohedging” approach very consistent (Delta Air Lines, United Airlines, American Airlines). One reason for the decline observed in hedging among the big US players, other than fuel hedging negative experiences following the 2008 oil price peak, might be the US GAAP (Generally Accepted Accounting Principles). In Europe and for the last years, airlines tended to hedge way more than in the US, which might also be related to the adoption of IFRS (International Financial Reporting Standards). Finally, we find the largest Middle East (Emirates), Russian and all Chinese airlines, notably not hedging their fuel needs, with India or China having regulations not allowing carriers to hedge their future fuel consumption (IATA, 2018a). The review conducted showed that while some carriers hedge fuel substantially at a given moment in time, others do so only moderately or not at all, changing their strategy along the time. This owes to various factors, including business models, regional market dynamics, accounting practices or even the influence of positive versus negative experiences in the past. Also, the access itself to hedging instruments, either because contract costs are too expensive or simply because the political context plays a role. Regional differences are therefore noticeable, but, in general, airlines competing in the same markets tend to have similar hedging strategies, as they also tend to benchmark their prices against one another. 3.5 Cases We have selected two cases depicting different experiences, policies, and competitive landscapes concerning fuel hedging and derivatives’ accounting standards, bolstering the conclusions presented further in this chapter. Southwest Airlines is best known for its major fuel hedging success case in the period until the oil price peaked in 2008, earning more than USD4 billion in fuel hedging settlements from 2003 to 2008. However, Southwest has reported substantial financial hedging losses ever since. Still, the company remained faithful to its fuel hedging approach as an instrument to dampen oil price volatility. Operating out of Hong Kong, Cathay Pacific has seen significant market growth in Asia in a competitive regional environment where the three major Chinese carriers are actually unhedged. After a streak of

Fuel hedging: how many games can we play?

397

successful hedging bets, negative results in the last decade have dwarfed past gains. Still, Cathay kept an aggressive fuel hedging strategy until recently, even daring to take long hedging positions for up to 4 years above market price, which caught the company off guard. Cathay’s case sheds light on the risks entailed in fuel hedging in the attempt to gain a competitive edge against rivals which, contrary to the USA situation, are predictably unable to hedge and exposed to oil price volatility. 3.5.1 Southwest Airlines Southwest Airlines is a leading US low-cost airline that started its operations in 1971. A pioneer in low-cost air travel, the company adopted a short route point-to-point business model, a single flight strategy with “no-frills” service. This approach contrasted with the hub and spoke model adopted by multiple competitors, having also pioneered operational hedging with its commitment to operate a single type of plane, the Boeing 737. Southwest was an adopter of jet fuel price-hedging since the early 2000s, accompanying oil price evolution closely from around USD20e30/bbl in 2001 to reach around USD150/bbl in 2008 (Fig. 17.4). By 2008, when the oil bubble hit the market, Southwest had in place one of the most aggressive fuel-hedging programs out of all airlines (Tokic, 2012), covering around 80% of its planned needs, with about 70% of its fuel consumption in 2008 hedged at the advantageous cost of USD51/bbl (Pyke and Sibdari, 2018). The gains from fuel derivative contracts’ settlements in 2008 reached USD1.3 billion (Lim and Hong, 2014), with a net profit of

Figure 17.4 Monthly average oil price in the period 2001e09. (Based on data set obtained from https://www.eia.gov.)

398

The Air Transportation Industry

USD1.1 billion (Southwest, 2020). Southwest had been following a long hedging strategy for several years, given the relentless increase in jet fuel price, earning more than USD4 billion in fuel hedging settlements from 2003 to 2008 (Peterson, 2008). Hedging a sizable portion of its fuel use, Southwest enjoyed considerably lower real jet fuel prices than its main competitors. Other US airlines were impacted with financial losses during the first half of 2008, namely United, Delta and Jet Blue, which were hedging only up to 40% of their needs. By mid-2008, Southwest Airlines had emerged as a fuel hedging winner (MarketWatch, 2008). However, when fuel prices dropped sharply from around USD150/bbl to USD30e35/bbl by year-end, Southwest was in a tough spot, overexposed with more than 80% of its jet fuel needs long hedged for 2008. Moreover, Southwest Airlines was also exposed for long hedging 75% and 50% of its planned fuel needs in 2009 and 2010, respectively. This led Southwest to revisit its fuel hedging position in place for 2009 through 2013, acknowledging it was locked in hedging-related losses for 2009 through 2013. Southwest recognized in 2009 a total of $467 million in fuel hedging related losses and $426 million in 2010, already factoring in the net premium costs paid to enter into fuel derivative instruments. Thus, while before 2009, Southwest’s fuel hedging program seemed successful, the hedging program caused the company to pay a higher price than most in the industry in 2009 and 2010 (Lim and Hong, 2014). Southwest continued to adjust its fuel hedge portfolio more carefully to “layer back in some protection in the event of a significant surge in market price,” adding however in its annual report to shareholders in 2010 that there could be “no assurance that the Company will be able to continue to cost-effectively hedge against increases in fuel prices” (Southwest, 2020). But as of January 2011, Southwest was still heavily engaged in multi-year fuel hedging. Approximately 60% of Southwest Airlines estimated fuel consumption for 2012 was then already covered; 50% in 2013 and about 45% in 2014. Hence, and despite the losses incurred in hedging following the 2008 recession, Southwest still relies today on financial derivative instruments both on a short-term and a long-term basis as insurance against oil prices volatility. 3.5.2 Cathay Pacific Airways Established in 1946, Cathay Pacific Airways is Hong Kong’s flag carrier. In 2006, Cathay Pacific acquired Dragonair (now “Cathay Dragon”), a Hong Kong-based carrier within the AsiaePacific region and signed a strategic

Fuel hedging: how many games can we play?

399

partnership with national flag carrier Air China. According to IATA’s 2019 World Air Transport Statistics, the company is the world’s eighth-largest carrier of international passengers and the third-largest carrier of international air cargo. Founding member of the Oneworld alliance, operated 235 aircraft immediately before COVID-19, connecting Hong Kong to 119 destinations in 35 countries worldwide, including 26 destinations in the Chinese mainland, employing some 23,000 people as of early 2020. The company’s fuel hedging policy defines that “derivative financial instruments are used solely to manage exposures to fluctuations in foreign exchange rates, interest rates and jet fuel prices in accordance with the Group’s risk management policies. The Group does not hold or issue derivative financial instruments for proprietary trading purposes” (Cathay Pacific, 2016). As per accounting rules in Hong Kong (HKFRS), all derivative financial instruments are recognized at fair value in Cathay’s annual financial statements. But even if the purpose of fuel hedging is not to profit, Cathay, just like other companies (such as Southwest), was fortunate in obtaining significant gains from fuel hedging for a particular time interval (Fig. 17.5). From 2005 through 2014, those gains totaled around HK$4 bn (USD0.5 bn), even considering the significant losses in the second half of 2014 and the first half of 2015 (CAPA, 2016). Indeed, Cathay was on a streak of fuel hedge betting successes until oil prices started to fall by mid-2014 and then again in 2015. By early 2015, the company found itself over hedged and not just for the short term. In 2013, its hedging positions extended into 2016 and by

Figure 17.5 Monthly average oil price in the period 2013e20. (Based on data set obtained from https://www.eia.gov.)

400

The Air Transportation Industry

early 2015 and was 60% hedged until the end of the year with additional exposure extending to future years. The company disclosed unrealized hedging losses in 2014 at HK$14 billion, with actual losses valued at HK$0.9 billion. Some HK$6 billionn were planned to hit 2015, leaving still around HK$8 billion of unrealized losses for subsequent years. That was somehow betting on fuel prices to return to the levels of 2013. But fuel prices remained way lower and for longer than what Cathay had hedged for. The year 2015 marked the beginning of a fuel hedging nightmare for Cathay, with oil prices dipping from USD60/bbl to around USD30/bbl, which caught the company in a situation of enduring heavy losses. Its “into wing” fuel price, advantageous to Cathay until 2014, saw a significant negative gap in 2015 as per Fig. 17.6. Consequently, by the end of 2015, Cathay reported HK$8.4 billion in realized fuel hedging losses in the year, starting a spiraling downwards cycle after almost a decade of fuel hedging profits. This put regional competitors in advantage, including the mainland Chinese carriers and Emirates, who had not hedged at all.

Figure 17.6 Fuel hedging impact on Cathay Airlines fuel price (2011e 2015). (Reproduced from Annual Report Cathay Pacific, 2015. Cathay Pacific Airways Limited 2015 Annual Results.)

Fuel hedging: how many games can we play?

401

But it would still take the company a long time to come off its hedges and reap the benefit of low fuel prices. Unlike most airlines, which managed most of their exposure for up to 1 year, Cathay had long-hedging positions for up to 4 years. Eventually, a wrong bet because the oil price never recovered to the levels predicted (Fig. 17.7). The group’s fuel hedging exposure as of December 31, 2015, was described as per the chart below, illustrating the degree of the company’s realized losses up until 2019. The impact on the “into wing” fuel price in this period is clear from the chart below, putting Cathay in a vulnerable situation against its nonhedging competitors in the region for several years (Fig. 17.8), (Cathay Pacific, 2015; Cathay Pacific, 2019). Cathay Pacific could finally see its expensive fuel hedges rolling over in 2019. Fuel hedging losses for Cathay had added up to HK$25bn (nearly USD3.3 bn) (Table 17.2). This loss due to over-hedging was topped by around USD200 million in the first half of 2020 when oil prices plunged.

Figure 17.7 Cathay’s percentage consumption subject to hedging contracts by the end of 2015. (Reproduced from Annual Report Cathay Pacific, 2015. Cathay Pacific Airways Limited 2015 Annual Results.)

402

The Air Transportation Industry

Figure 17.8 Fuel hedging impact on Cathay Airlines fuel price (2015e19). (Reproduced from Annual Report Cathay Pacific, 2019. Annual Report 2019. Hong Kong.) Table 17.2 Cathay Airlines annual fuel hending losses (2014e19). Fiscal year 2014 2015 2016 2017

2018

2019

Fuel Hedging Losses (HK$ bn)

1.45

0.1

0.9

8.47

8.45

6.38

From Cathay Annual Reports 2014, 2015, 2016, 2017, 2018, 2019.

4. Recent developments Whichever lessons the aviation business might have learned from past fuel hedging initiatives, the oil price disruption that emerged in 2020 (Fig. 17.9) has added more controversial results to the debate, questioning such practices. Multiple airlines were locked in 2019 into what turned out to be disadvantageous fuel hedging contracts for 2020. Most European airlines had hedged above 60% of their fuel consumption until Covid-19 changed their operational plans dramatically. Companies that had hedged more than 50% of their demand in 2020 include Ryanair, Lufthansa, EasyJet, Singapore Airlines, Air France-KLM, SAS and Southwest, to name a few. The choice of hedging instrument has a significant impact on the financial outcome using forward contracts versus options where the loss is limited to the airline’s premium for the respective option. Ryanair and EasyJet favored

403

Fuel hedging: how many games can we play?

80 70 60 50 40 30 20 10 0 -10 Jan -20 -30 -40 -50

Feb

Mar

Apr

May

Jun

WTI

Jul

Aug

Sep

Oct

Nov

Dec

Brent

Figure 17.9 Brent and WTI. Oil prices (USD) in 2020. (Based on data set obtained from https://www.eia.gov.)

forward contracts and may have suffered significant impacts. EasyJet hedged its fuel consumption for 2020 at USD89/bbl, the highest among major airlines, likely losing over USD1 billion. Ryanair had hedged 90% of its fuel requirement at about USD77/bbl for the year beginning April 1st, with financial impacts also estimated at more than USD1billion. Air France KLM adopted a combination of options and swaps for fuel hedging, locked at about USD78.5/bbl for more than 60% of its 2020 needs. The company may have lost an estimated USD800 million. British Airways and Iberia (IAG) had also hedged more than 90% of its needs planned for the first three quarters of 2020, potentially facing multibillion hedging (EuroFinance, 2020a). “British Airways is literally selling off the family china to shore up its finances: half a dozen first-class teacups can be yours for just over $30” (The Economist, 2020b). Together, Ryanair, Air France, EasyJet and IAG were expected to sustain around USD4.5 billion in realized losses in 2020 owing to wrong fuel hedging bets. IAG and Lufthansa have been particularly hit and have stopped fuel hedging altogether, with some analysts suggesting they may never resume fuel hedging programs again. Steve Gunning, C.E.O. of IAG, noted that “we are actually doing a sort of fundamental review of that policy to see whether we need to learn from 2020 and take a different tack” (EuroFinance, 2020b). Adding further “no, we’re not taking out further hedges at this point”, which might also read like a “never say never” statement. Other companies such as Lufthansa and Southwest Airlines had optionbased strategies, hence remaining less vulnerable. Lufthansa airlines use options exclusively in its fuel hedging, with the CEO reinforcing in March’s earnings conference meeting that there was “no commitment to buy

404

The Air Transportation Industry

any fuel” (Reuters, 2020). Nevertheless, Lufthansa had hedged at comparatively the lowest level at USD63.32/bbl and for approximately 75% of its planned annual fuel demand. Southwest Airlines had hedged around 59% of its estimated fuel consumption, down from 73% in 2019. In Asia, Cathay Pacific reported a record net loss of nearly USD1.3 billion in 2020H1, under the coronavirus pandemic’s impact, having grounded most of its passenger flights. This negative result was aggravated by hedging contracts for fuel set before the oil futures dropped off. While this fall in fuel prices, as well as decreased activity, generated a cost reduction of approximately USD1.1 billion to Cathay, having fuel hedged at more than USD62/bbl for 40% of the Group’s typical consumption caused losses of around USD200 million (NikkeiAsia, 2020) in the first semester. An order of magnitude similar to the company’s net profit in 2019. In Australia, Qantas Group’s fuel needs were 100% hedged for most of FY20. Qantas closed out in April its over-hedged position until September 2020, significantly lowering the exposure to further hedging losses in the year. As of May 2020, all foreign exchange and fuel hedging’s cash impact was estimated at AUD 145 million, with hedging beyond September tied to options and collars (Qantas, 2020). These recent developments came to add new reasons for a seriousminded reflection about fuel hedging. Perhaps to the point of companies ditching programs altogether as major USA airlines did, years ago. Still, it will take time for the industry and analysts to process all that has happened in 2020 before the full consequences and sway on fuel hedging strategies are entirely understood.

5. Conclusions and future outlook We have looked for answers on whether there were real benefits for airlines to engage in fuel risk hedging systematically. A literature review revealed no clear evidence that the practice has contributed to adding value to the airlines in the long run while leaving airlines profits’ volatility fundamentally unchanged, at best. Despite anecdotal cases where airlines have obtained impressive results, fuel hedging fundamentals seem to hold solid. Some research findings have suggested that fuel hedging might even increase airline profits volatility, owing to the correlation between oil prices, air travel demand cycles and GDP growth fluctuations. Consequently, fuel hedging during periods of heightened economic volatility may actually be

Fuel hedging: how many games can we play?

405

counterproductive. Ultimately, the airline’s risk reduction is worth only the contract’s costs involved. Notwithstanding, fuel hedging has persisted as a risk management practice amongst many airlines. The backdrop of high oil price volatility showing no easing signals for the past two decades has undoubtedly contributed to that. Yet, oil market is characterised by high liquidity attracting multiple industries and stakeholders, rendering airlines’ role a relatively minor one. Capturing the complex factors at play in incidental oil price volatility has proven challenging, at best. Even if the fundamentals of economics of exhaustible resources’ may justify a mid-long term oil price upwards trend underpinning airlines’ betting strategies, volatility happens at unpredictable discrete times, often with dramatic swings, turning fuel hedging into a mere guessing game. When these swings occur at the timescale of financial reporting periods, companies often got caught in the wrong hedges. Still, fuel hedging practices are often mentioned in association with executive management competence and savviness. In general, airlines competing in the same markets tend to have similar hedging strategies, as they also tend to benchmark their prices against one another, turning fuel hedging into a high stakes gamble between competing management teams. Accounting standards followed by each company in the respective countries and regions may also allow convenient representations of a company’s results. This aspect can be seen either as a driver for EU companies and as an obstacle for USA based ones. Another element driving fuel hedging seems to lie in the reinforcing conviction of managers that “the market can be beaten”. Managers overlook future risk for following an emotional rather than logical approach, a risky way of thinking known to financial markets. Yet, the last two decades have bewildered all those players who, at some point, had wrongly assumed they had nailed the marked trends just right. These factors have emerged among some of the most compelling drivers for airlines to hedge their fuel needs. Playing against fuel hedging are the collateral costs required, identified as a critical limiting factor in airlines’ hedging decisions, every so often offsetting any potential financial benefits. Also challenging is the actual ability to forecast oil price volatility. Even the most advanced econometric predictive models have fundamental flaws in assuming relevant data to be stationary, regular, and linear when it is, in fact, complex, irregular, and non-linear in nature.

406

The Air Transportation Industry

Heavy losses have in many cases dwarfed past gains, and fuel hedging motivation seems to be hitting rock-bottom in 2020, upon detrimental fuel hedging contract settlements for most airlines. Consequently, and after spending billions to hedge against rising fuel costs, more and more airlines have decided to back off on fuel hedging altogether or significantly reduce the hedged percentage of their needs. Notwithstanding, those players who will emerge from this crisis “will also find the skies less crowded” (The Economist, 2020b). Governmental support to airlines and business consolidation have the power to change the competitive landscape in ways yet to be exposed. Nonetheless, jet fuel market stakes are high, constituting a market worth nearly USD0.2 trillion in 2019. Oil price predictive modeling remains a very active research topic, stirred by Artificial Intelligence developments in association with Quantum Computing. These may address the drawbacks of econometric methods, and leapfrogging advancements in this area of study are not far-fetched futuristic scenarios. If breakthroughs are achieved in interpreting and factoring in the interplay among crucial data and if the access of stakeholders to such foretelling information is somewhat asymmetric, then there is a chance of unbalancing the odds. Such a revolution may come much earlier for specific applications using simpler algorithms. And “whoever makes a breakthrough first may choose to reap the rewards in obscurity, rather than broadcast the fact to the world” (The Economist, 2020a). Therefore, the opportunity to adopt increasingly sophisticated predictive models seems far too tempting to ignore. Assuming that competition is not rendered to a moot-point in the post-COVID-19 aviation landscape, fuel hedging, though in a more sophisticated fashion, will likely remain enticing aviation stakeholders for the foreseeable future.

References Abdollahi, H., 2020. A novel hybrid model for forecasting crude oil price based on time series decomposition. Appl. Energy 267, 115035. https://doi.org/10.1016/ j.apenergy.2020.115035. Abdollahi, H., Ebrahimi, S.B., 2020. A new hybrid model for forecasting Brent crude oil price. Energy 200, 117520. https://doi.org/10.1016/j.energy.2020.117520. Azimli, A., 2020. The oil price risk and global stock returns. Energy 198, 117320. https:// doi.org/10.1016/j.energy.2020.117320. Berghöfer, B., Lucey, B., 2014. Fuel hedging, operational hedging and risk exposure evidence from the global airline industry. Int. Rev. Financ. Anal. 34, 124e139. https:// doi.org/10.1016/j.irfa.2014.02.007. Breeden, D.T., Viswanathan, S., 2016. Why do firms hedge? An asymmetric information model. J. Fixed Income 25, 7e25. https://doi.org/10.3905/jfi.2016.25.3.007.

Fuel hedging: how many games can we play?

407

Bristone, M., Prasad, R., Abubakar, A.A., 2020. CPPCNDL: crude oil price prediction using complex network and deep learning algorithms. Petroleum 1e9. https://doi.org/ 10.1016/j.petlm.2019.11.009. CAPA, 2016. Cathay Pacific’s Difficult 2016 [WWW Document]. Cent. Aviat. URL: https://centreforaviation.com/analysis/reports/cathay-pacifics-difficult-2016-hedgingloss-growth-cuts-a350-delay-unions-10-abreast-response-263367 (Accessed 11.30.20). Carter, D.A., Rogers, D.A., Simkins, B.J., 2006. Does hedging affect firm value? Evidence from the US airline industry. Financ. Manag. 35, 53e86. https://doi.org/10.1111/ j.1755-053X.2006.tb00131.x. Cathay Pacific, 2019. Annual Report 2019. Hong Kong. Cathay Pacific, 2016. Annual Report 2016. Cathay Pacific, 2015. Cathay Pacific Airways Limited 2015 Annual Results. Charness, G., Dave, C., 2017. Confirmation bias with motivated beliefs. Game. Econ. Behav. 104, 1e23. https://doi.org/10.1016/j.geb.2017.02.015. Dontwi, I., Dedu, V., Davis, R., 2010. Application of options in hedging of crude oil price risk. Am. J. Soc. Manag. Sci. 1, 67e74. https://doi.org/10.5251/ajsms.2010.1.1.67.74. Dotzlaw, R., Ipatova, I., Bascom, K., Santoro, J., 2020. IFRS® Compared to US GAAP. Retrieved from: https://assets.kpmg/content/dam/kpmg/xx/pdf/2020/03/ifrs-usgaap-2020.pdf2020. EuroFinance, 2020a. Covid-19 Puts Airline Hedge Strategies under New Focus [WWW Document]. Eurofinance News. URL: https://www.eurofinance.com/news/covid19-puts-airline-hedge-strategies-under-new-focus/ (Accessed 11.27.20). EuroFinance, 2020b. European Airlines May Quit Fuel Hedging after $ 4.66 Billion in Losses [WWW Document]. December. URL: https://www.eurofinance.com/news/ european-airlines-may-quit-fuel-hedging-after-4-66-billion-in-losses/ (Accessed 12.7.20). Ghose, S., 2020. Are you ready for the quantum computing revolution? Harvard Business Review. https://hbr.org/2020/09/are-you-ready-for-the-quantum-computing-revolution (Accessed 12.15.20). Hamilton, J.D., 2009. Understanding crude oil prices. Energy J. 30, 179e206. https:// doi.org/10.5547/ISSN0195-6574-EJ-Vol30-No2-9. IATA, I.A.T.A., 2018a. No simple link between oil prices and industry profitability. IATA, I.A.T.A., 2018b. Fuel hedging: large differences by region, fewer by market. IEA/OECD, 2008. World Energy Outlook 2008. https://doi.org/10.1787/weo-2008-en. Kjärstad, J., Johnsson, F., 2009. Resources and future supply of oil. Energy Pol. 37, 441e464. https://doi.org/10.1016/j.enpol.2008.09.056. Korkeamäki, T., Liljeblom, E., Pfister, M., 2016. Airline fuel hedging and management ownership. J. Risk Finance 17, 492e509. https://doi.org/10.1108/JRF-06-2016-0077. Kristjanpoller, W., Minutolo, M.C., 2016. Forecasting volatility of oil price using an artificial neural network-GARCH model. Expert Syst. Appl. 65, 233e241. https:// doi.org/10.1016/j.eswa.2016.08.045. Laing, E., Lucey, B., Lutkemeyer, T., Lucey, B.M., 2020. Which form of hedging matters d operational or financial? Evidence from the US oil and gas sector. Res. Int. Bus. Finance 51, 101088. https://doi.org/10.1016/j.ribaf.2019.101088. Li, J., Tang, L., Wang, S., 2020. Forecasting crude oil price with multilingual search engine data. Phys. A Stat. Mech. Appl. 551, 124178. https://doi.org/10.1016/ j.physa.2020.124178. Lim, S.H., Hong, Y., 2014. Fuel hedging and airline operating costs. J. Air Transp. Manag. https://doi.org/10.1016/j.jairtraman.2013.12.009. MarketWatch, 2008. Southwest’s Treacherous Oil Hedges. MarketWatch First Take. https://www.marketwatch.com/story/southwests-infatuation-with-fuel-hedgesbackfires (Accessed 11.03.2020).

408

The Air Transportation Industry

Martino, A., Casamassima, G., Fiorello, D., 2009. The Impact of Oil Prices Fluctuations on Transport and Its Related Sectors 96. Ménard, A., Ostojic, I., Patel, M., Volz, D., 2020. A game plan for quantum computing. McKinsey Q. 1e8. Merkert, R., Swidan, H., 2019. Flying with(out)a safety net: financial hedging in the airline industry. Transp. Res. Part E Logist. Transp. Rev. 127, 206e219. https://doi.org/ 10.1016/j.tre.2019.05.012. Morrell, P., Swan, W., 2006. Airline jet fuel hedging: theory and practice. Transp. Rev. 26, 713e730. https://doi.org/10.1080/01441640600679524. Mostafa, M.M., El-Masry, A.A., 2016. Oil price forecasting using gene expression programming and artificial neural networks. Econ. Modell. 54, 40e53. https://doi.org/ 10.1016/j.econmod.2015.12.014. Nelson, J.M., Moffitt, J.S., Affleck-Graves, J., 2005. The impact of hedging on the market value of equity. J. Corp. Finance 11, 851e881. https://doi.org/10.1016/ j.jcorpfin.2005.02.003. NikkeiAsia, 2020. Cathay Pacific Airways Ltd [WWW Document]. URL: https://asia. nikkei.com/Companies/Cathay-Pacific-Airways-Ltd (Accessed 11.27.20). Peterson, K., 2008. Oil Decline Erodes Airline Fuel Hedge Value [WWW Document]. Reuters. URL: https://www.reuters.com/article/us-airlines-hedging-sb/oil-declineerodes-airline-fuel-hedge-value-idINTRE4BM4TW20081223 (Accessed 10.30.20). Pyke, D., Sibdari, S., 2018. Risk management in the airline industry. In: Finance and Risk Management for International Logistics and the Supply Chain. https://doi.org/ 10.1016/b978-0-12-813830-4.00012-5. Qantas, 2020. ASX and Media Release. QANTAS group market update, Melbourne. Reuters, 2020. Edited Transcript of LHA.DE Earnings Conference Call or Presentation March 2020 [WWW Document]. URL: https://finance.yahoo.com/news/editedtranscript-lha-earnings-conference-225637786.html?guccounter¼1&guce_referrer¼aHR0c HM6Ly93d3cuZ29vZ2xlLmNvbS8&guce_referrer_sig¼AQAAANzMtwkgdcC9a0OViIx TcBb9h_vTULcpK57TjFBAP4-AmUl0Ts37rkQ8omBi2dNwL5mYhwcLw3SkxcX4Wh hN_O (Accessed 10.10.20). Southwest, 2020. Southwest Airlines - 2019 Annual Report to Shareholders. Stulz, R., 1996. Rethinking risk management. J. Appl. Corp. Financ. 9, 8e25. Swidan, H., Merkert, R., Kwon, O.K., 2019. Designing optimal jet fuel hedging strategies for airlines e why hedging will not always reduce risk exposure. Transp. Res. Part A Policy Pract. 130, 20e36. https://doi.org/10.1016/j.tra.2019.09.014. Taylor, R.D., 2020. Quantum artificial intelligence: a “precautionary” U.S. approach? Telecommun. Pol. 44. https://doi.org/10.1016/j.telpol.2020.101909. The Economist, 2020a. Wall Street’s Latest Shiny New Thing: Quantum Computing. December, 19th 2020. https://www.economist.com/finance-and-economics/2020/ 12/19/wall-streets-latest-shiny-new-thing-quantum-computing. The Economist, 2020b. Will Big Firms Benefit from the Covid Crunch? Nov 28th 2020, first ed. https://www.economist.com/leaders/2020/11/26/will-big-firms-benefit-fromthe-covid-crunch. Tokic, D., 2012. When hedging fails: what every CEO should know about speculation. J. Manag. Dev. 31, 801e807. https://doi.org/10.1108/02621711211253259. Tontiwachwuthikul, P. (PT), Chan, C.W., Zeng, F. (Bill), Liang, Z. (Henry), Sema, T., Chao, M., 2020. Recent Progress and New Developments of Applications of Artificial Intelligence (AI), Knowledge-Based Systems (KBS), and Machine Learning (ML) in the Petroleum Industry. Petroleum 0e1. https://doi.org/10.1016/j.petlm.2020.08.001. Treanor, S.D., Rogers, D.A., Carter, D.A., Simkins, B.J., 2014. Exposure, hedging, and value: new evidence from the U.S. airline industry. Int. Rev. Financ. Anal. 34, 200e211. https://doi.org/10.1016/j.irfa.2014.04.002.

Fuel hedging: how many games can we play?

409

Turner, P.A., Lim, S.H., 2015. Hedging jet fuel price risk: the case of U.S. passenger airlines. J. Air Transp. Manag. 44 (45), 54e64. https://doi.org/10.1016/ j.jairtraman.2015.02.007. Tversky, A., Kahneman, D., 1971. Belief in the law of small numbers. Psychol. Bull. https://doi.org/10.1037/h0031322. Wang, J., Niu, T., Du, P., Yang, W., 2020. Ensemble probabilistic prediction approach for modeling uncertainty in crude oil price. Appl. Soft Comput. J. 95, 106509. https:// doi.org/10.1016/j.asoc.2020.106509. Wang, M., Zhao, L., Du, R., Wang, C., Chen, L., Tian, L., Eugene Stanley, H., 2018. A novel hybrid method of forecasting crude oil prices using complex network science and artificial intelligence algorithms. Appl. Energy 220, 480e495. https://doi.org/ 10.1016/j.apenergy.2018.03.148. WSJ, 2016. Airlines Pull Back on Hedging Fuel Costs Reappraisal of Costly Strategy Comes after Some Carriers Get Burned by Low Oil Prices 1. Wu, B., Wang, L., Lv, S.X., Zeng, Y.R., 2021. Effective crude oil price forecasting using new text-based and big-data-driven model. Meas. J. Int. Meas. Confed. 168, 108468. https://doi.org/10.1016/j.measurement.2020.108468. Yu, L., Zhao, Y., Tang, L., 2014. A compressed sensing based AI learning paradigm for crude oil price forecasting. Energy Econ. 46, 236e245. https://doi.org/10.1016/ j.eneco.2014.09.019.

This page intentionally left blank

CHAPTER 18

The effect of accidents on aircraft manufacturers’ competition Wouter Dewulf1, Silke Forbes2 and Yufei Li2 1

Department of Transport and Regional Economics, University of Antwerp, Antwerp, Belgium; Department of Economics, Tufts University, Medford, MA, United States

2

1. Introduction Safety has always been an essential element to the business success of the passenger airline industry (Chang and Yeh, 2004). One of the first heavily publicized accidents in the media related to air travel was the German-built airship LZ 129 Hindenburg disaster originating from Frankfurt on May 6, 1937 in New Jersey. This tragic event resulted in the loss of trust in air travel by zeppelin airships. One of the main reasons that airships were abandoned for air travel shortly after the disaster is that the public confidence in this type of airships never recovered after this fatal accident. Airships were replaced by fixed-wing airplanes, which increased technical reliability, enhanced technology, and improved range after World War II. The British built de Havilland DH.106 Comet was the world’s first commercial jet airliner, launched in 1952. It flew faster and higher than the previous propeller-powered aircraft and offered a comfortable and relatively quiet passenger cabin. It attracted much interest from airlines all over the world. However, within a year of entering into service, three Comets were involved in highly publicized fatal accidents in the media after suffering catastrophic in-flight airframe break-ups. These fatal accidents were caused by structural failure resulting from metal fatigue or overstressing the airframe during a storm. This aircraft type’s sales never fully recovered due to public distrust in the Comet, even when the newly released versions of the Comet were much safer. The Boeing Company took over the commercial lead of airframe manufacturers with the Boeing 707 launched in 1958 by Pan Am airlines which offered better operations characteristics such as a more extended range, a higher seat capacity and lower fuel burn.

The Air Transportation Industry ISBN 978-0-323-91522-9 https://doi.org/10.1016/B978-0-323-91522-9.00015-4

© 2022 Elsevier Inc. All rights reserved.

411

412

The Air Transportation Industry

McDonnell Douglas launched in 1971 the DC-10, as a three-engine wide-body successor of the DC-8. Several incidents and accidents happened in early operations of the aircraft due to a design flaw in locking mechanism of the cargo doors. Following the American Airlines flight 191 crash on May 25, 1979, the FAA decided to ground all DC-10s based in the USA. In August 1983, McDonnell Douglas announced that production would end due to a lack of orders, as the aircraft type had a widespread public fear and distrust after the accidents. This 1979 event contributed to McDonnell Douglas’s financial difficulties, which failed to develop and sell new products to catch up with the intensified competition in the United States. The Concorde, jointly built by a British-French consortium between BAC and Aérospatiale, is a supersonic passenger airliner operated from 1976 until 2003 when British Airways and Air France jointly decided to retire the Concorde. Low passenger numbers following the July 25, 2000 crash in Paris and the slump in air travel following the 9/11 attacks in 2001 were given as the main reasons to withdraw the Concorde from the airlines’ operations. In the last decades the number of fatal aviation accidents has decreased significantly. The fast expansion of the low cost carriers and the increase of available seats on the market did not lead to more accidents. The public generally perceived air travel as a very safe and convenient way for longer distance travel. However, two recent aircraft accidents, Lion Air Flight 610 on October 29, 2018 and Ethiopian Airlines Flight 302 on March 10, 2019, both with a Boeing 737 MAX eight aircraft changed this perception for some time, especially for this particular aircraft type. The authorities grounded the 737 MAX aircraft worldwide shortly after the second crash due to a software design fault with the Maneuvering Characteristics Augmentation System (MCAS). Boeing’s hasty design and testing work, under time pressure due to the commercial success of the re-engined Airbus 320neo aircraft, is said to have led to insufficient testing and faulty systems programming on the Boeing 737 MAX. These errors led to the Lion Air and Ethiopian Air flights’ fatal crashes, with huge impact on Boeing’s stock price, company image and order book, and distrust from the general public as serious consequences. As mentioned earlier, these events demonstrate that safety and public perception of the particular aircraft type are paramount to an airframe manufacturer’s success. This chapter analyzes the effect of accidents and fatal

The effect of accidents on aircraft manufacturers’ competition

413

accidents on aircraft manufacturers competition, and aims to answer the following research questions: - Do aviation accidents have an impact on aircraft manufacturers’ competition? - Are airlines devoted to aircraft manufacturers, even when accidents with the specific manufacturer’s aircraft type occur? These research questions will be answered in the rest of this chapter, starting with a brief literature review and overview of the aircraft manufacturers market to set the scene. The fourth part discusses the evolution and severity of aircraft accidents over the last 70 years, followed by an analysis of the impact of accidents on the competition between aircraft manufacturers. The final part of this chapter investigates the reasons why airlines often tend to stick to a specific aircraft manufacturer.

2. Aircraft accidents, aircraft safety, and airline stock prices: a literature review There is a vast amount of academic literature related to aircraft accidents and aviation safety. The literature mainly focuses on research areas such as the causes of aviation accidents, risk assessments, safety-enhancing investments, pilot fatigue, and human errors. The most commonly used measurement in safety performance is accident rates, including either all accidents or focusing on fatal accidents only. However, these accidents are only the tip of the iceberg and over 90% of “latent” events, often called “incidents” in the literature, are not reflected by these indicators (Liou et al., 2007). Hwang and Yoon (1981) therefore propose to use multiple criteria decision-making (MCDM) which is characterized by multiple conflicting criteria. Loeb, Talley, and Zlatoper investigated and classified the causes of aviation accidents. Milan Janik (Janik, 2000) built on this work by adding an assessment of risk and safety in civil aviation and proposing a methodology for quantifying risk and safety risk assessment. Commercial airlines are regulated for safety, but they are also incentivized by customer responses to their safety reputation. Aircraft accidents, especially fatal accidents, can potentially have a severe impact on an airline’s reputation. Several authors have examined the impact of aircraft accidents on economic indicators such as share value. Chance and Ferris (1987) estimated the effect of crashes involving domestic airlines with at least 10

414

The Air Transportation Industry

fatalities on stock prices. They found a negative 1.2% statistically significant average unexpected return for the airlines involved in the crashes on the day of the crash. However, they did not find evidence of an impact on either other non-related airlines or the impacted aircraft manufacturer. Walker et al. (2005) estimated the effect of 138 US aircraft crashes on the airlines operating the aircraft and found a decreasing trend of the airline’s stock price, which lasted for the first trading week after the crash. After grouping the accidents by different causes of the crash, they concluded that airlines experienced the hardest times if the crash was caused by terrorist activity. In a similar study, Kaplanski and Levy (2010) found evidence of a significant adverse event effect with an average market loss of more than 60 billion USD per aviation disaster, whereas the estimated actual loss of the aircraft crash is no more than 1 billion USD. Ho et al. (2013) quantified the negative influence of an aviation disaster on airlines’ post-crash equity value while the magnitude of the influence is closely related to the level of fatality of the accident. They concluded that both the directly impacted airline and the non-impacted airlines experienced negative post-crash stock returns. However, the directly impacted airlines’ shareholders experienced a larger short-term value loss than those of the non-impacted airlines. The crashes with a larger number of fatalities resulted in larger losses of the directly impacted airlines’ equity values. Airlines face a trade-off between the costs of additional safety-enhancing investments and the benefits of reducing accident or accident rates. Rose (1990) estimated the impact of financial indicators on the carrier’s safety performance by analyzing data from 35 sizable scheduled passenger airlines over 1957 to 1986. She used financial indicators such as the operating margins, interest coverage, working capital and current ratios, system average length, and cumulative airline operating expenses as variables to explain the accident rate. She concluded that an increase in the operating margins would on average, decrease the expected accident rate, especially for smaller carriers. In another paper, Rose (1992) also studied the effect of the airline industry’s deregulation on airline safety performance. She concluded that airline safety performance did not decrease due to airline deregulation. While extensive research was done on the impact on the airline’s stock value following an aircraft crash, limited research has been performed on the actual impact on the value of aircraft manufacturers following a disaster with an aircraft of the related aircraft manufacturer. An exception is Chalk (1987) who showed that on average the impacted aircraft manufacturer

The effect of accidents on aircraft manufacturers’ competition

415

suffers a loss in value of 21.3 million USD as the result of a fatal crash in which the structural integrity of the aircraft was at issue. In this chapter, we provide new insights into this research topic by examining the impact of accidents on aircraft orders.

3. The aircraft manufacturers market: the story of a continuous consolidation In the beginning of the 20th century, the aircraft manufacturers market was very diverse. However, most of these companies have since been acquired and merged by competitors or went out of business. The Lockheed Corporation, an American aerospace company, founded in 1926, merged with Martin Marietta to form Lockheed Martin in 1995. After launching several civil aircraft, of which the Lockheed TriStar L1011 was the final developed civil product, Lockheed withdrew in 1984 from the commercial aircraft business due to the below-target sales of this aircraft model. The Lockheed Corporation, but also its competitor McDonnell Douglas were weakened by the research and development costs related to the development of a long haul wide-body aircraft to compete with the Boeing 747 and the planned Boeing 777. The management of the Lockheed Corporation decided to focus on the defense industry instead. McDonnell Douglas, a major American aerospace manufacturing corporation and defense contractor, formed by the merger of McDonnell Aircraft and the Douglas Aircraft Company in 1967, was acquired and merged in 1997 by Boeing, its main competitor. The exit of Lockheed from the civilian airliner market and the acquisition of McDonnell Douglas by Boeing lead to a strong consolidation of the aircraft manufacturers in the United States. Fokker, founded in 1912, was a Dutch aircraft manufacturer. During its most successful period from the mid-1920s up to mid-1930s, it dominated the civil aviation market with the F.VIIa-3m trimotor passenger aircraft. In the 1970s, Fokker launched several new aircraft, like the F-28, which was upgraded in 1986 into the larger F-100, seating up to 109 passengers. However, regional aircraft manufacturers Bombardier from Canada and Embraer from Brazil entered the regional jet market by upgrading and enlarging their regional turboprop variants moving toward the 70e100 seater products of Fokker. The Fokker aircraft products were squeezed between Embraer and Bombardier’s upgraded regional jet products and the Boeing 737 and Airbus 320 series, which could offer a 100e150 seater

416

The Air Transportation Industry

product range. Fokker went into bankruptcy in 1996, caused by weak sales of its largest and newest aircraft model, the Fokker 100, and hence a strategic refocusing attempt from its main shareholder DASA. Montreal-based Bombardier, Brazil-based Embraer, and European joint venture ATR focus on the production of small-capacity (50 seats) and medium-capacity (50e100 seats) turboprop and jet aircraft for commercial use, but have traditionally played no role in the market for planes with more than 125 seats. Airbus and Bombardier signed a strategic partnership on the production and marketing of the Airbus A220 (previously CS100 and CS 300 series) aircraft in 2017, followed by a full transfer of the assets related to the A220 R&D and production to Airbus and the government of Quebec in 2020 (Airbus, 2020a,b,c). Boeing and Embraer also announced a partnership in 2019, approved by the respective authorities in 2020. Boeing was in 2019 the world’s largest aerospace company manufacturing commercial aircraft, defense, space and security systems. It was founded in 1917 and delivered its first commercial jet aircraft, the Boeing 707, in 1958. Since it has launched an extensive range of successful airliners like the B727, B737, B747, B757, B767, B777, and B787. The Boeing 737 series is the most successful aircraft ever made by an aircraft manufacturer. Apart from the subprime crisis in 2008 and 2009, Boeing enjoyed stable and increasing net earnings from 2001 till 2018. However, the Boeing 737 MAX’s worldwide grounding and engine issues with the Boeing 787 lead to an all-time high loss of 636 million USD in 2019. The total revenues in 2019 were 76.6 billion USD which is 24.3% less than in 2018, mainly caused by reduced sales of civil aircraft (Boeing, 2020a). Boeing published a net loss of 3.5 billion USD in the first three quarters of 2020, caused by the same reason (Boeing, 2020b). Airbus was founded in 1970, with the aim of consolidating the fragmented commercial jetliner manufacturing industry in Europe, mainly situated in France, Germany, the United Kingdom and Spain, into a strong counterweight and international competitor, mainly for the US-based aircraft manufacturers. In the 1960 and 1970s the global market for commercial jetliners was overwhelmingly dominated by American companies. Boeing, Lockheed and Douglas Aircraft controlled about 90% of the market (Olienyk and Carbaugh, 2011). The first Airbus commercial aircraft, the wide-body A300, was delivered in 1974. Since its inception, Airbus successfully gained market share from Boeing by expanding its product range step by step with a series of narrow and wide-body aircraft. The last couple of years Airbus was at par with Boeing when considering the

The effect of accidents on aircraft manufacturers’ competition

417

number of aircraft delivered yearly. In 2019, Airbus even exceeded Boeing in the total number of aircraft delivered. Airbus achieved 70.5 billion EUR revenues and a net loss of 1.4 billion EUR with a free cash flow of 3.5 billion EUR in 2019 (Airbus, 2020b). Like Boeing, the third quarter 2020 results are not positive, however less negative than Boeing’s financial figures. The consolidated revenues decreased to 30.2 billion EUR (9m 2019: 46.2 billion EUR), driven by the challenging market environment impacting the commercial aircraft business with around 40% fewer deliveries year-on-year. Comparing the Boeing and Airbus aircraft types, there are only limited product feature differences between the two companies’ major aircraft types, although the design philosophy is entirely different. Airbus starts from the idea that technology should prevail over the pilots’ commands, while Boeing gives the pilot in principle the final command in operating the aircraft. Airbus was an early adopter of the “fly-by-wire” principle, while Boeing adhered for a long time to the mechanical cables principle for steering and maneuvering capabilities. However, comparing seating capacity, performance, fuel efficiency and range, the Airbus A320 versus the Boeing 737 series, the Airbus A350 versus the Boeing 787, and even the Airbus A380 and the Boeing 747-8 are very similar products. Only the wide-body twin-engine Boeing 777 has up to a certain extent distinct product features like a large size and more extended range. However, the costs per seat mile of the Boeing 777 are very comparable to the Airbus A330. Given that both Airbus and Boeing offer very similar products, airlines should be rather indifferent, apart from the purchase price, about what type of products to acquire. These products should be easily substituted by the buying airlines, depending on factors like price and availability. However, the number of potential airline customers is rather limited, and the yearly demand for new aircraft is limited. Consequently, the competitive rivalry between the two companies could be fierce to grasp the maximum of the orders available in the market. This substitution effect could be facilitated if a product from one manufacturer suffers an accident caused by a product fault, leading to distrust from passengers or airlines. However, companies develop loyalties toward their customers to successfully compete and safeguard their respective market shares (Besanko et al., 2017). Airlines standardize their fleets to economize on pilot and cabin crew training, operational and maintenance processes, spare part stocks and enhanced network flexibility in the operations. For instance, Air Asia group is a

418

The Air Transportation Industry

typical Airbus customer operating both the A320 and A330 series. Southwest Airlines is a typical Boeing operator flying different models of the 737 series. Through this aircraft choice, these airlines achieve economies of scale through fleet harmonization and standardization. Russian aircraft manufacturers Sukhoi and Irkut, and the Chinese aircraft manufacturer COMAC have mainly local and state-owned customers. When considering the aircraft manufacturers segment of 125þ seater aircraft, which is the largest aircraft product segment, Airbus and Boeing now effectively constitute a duopoly. Considering the Airbus and Boeing’s strategic partnerships with Bombardier and Embraer respectively, the 50 to 125 seater regional jet aircraft market is now also aligned with the Airbus-Boeing duopoly. Therefore, the global and competitive 125þ seater aircraft market is currently neatly divided between Airbus and Boeing.

4. Aircraft accidents: a historic overview of air travel from safe to safest way of travel Although air transport enjoys an excellent safety record, public perception often focuses excessively on accidents (Liou et al., 2007). The general public tends to pay significantly more attention to small probability events with a high impact than vice versa (Camerer and Howard, 1989). Aircraft accidents with casualties are nowadays very infrequent. However, when these fatal accidents occur, they are often catastrophic, spectacular, mysterious, and highly publicized. Many travelers still have a fear of flying, while air transport is statistically by far the safest way of transport compared to other transport modes. The index of aviation safety can be expressed in “number of fatal aircraft accidents,” “number of fatal aircraft accidents per million flights,” or “number of passenger casualties per billion miles flown.” These indicators are applied to compare and rank the accident rates per airline, per geographic region, or make a time series to analyze these indicators’ evolution. Fig. 18.1 gives an overview of the number of fatal aircraft accidents from 1946 to 2019, and includes all aircraft types, both for civil and military use. The 5-year moving average line shows that, although the number of flights increased vastly during that period, the total number of aircraft accidents dropped dramatically from about 70 per year in the late 1940s to about 18 in 2019. Fig. 18.2 demonstrates the significantly improved safety record of aircraft for the last 60 years. In 1958 the aviation world still had 11 fatal aircraft accidents per million flights. This indicator has been reduced to 0.2

The effect of accidents on aircraft manufacturers’ competition

419

Figure 18.1 Total number of fatal aircraft accidents from 1946 to 2019. (Source: https:// aviation-safety.net/graphics/infographics/Fatal-Accidents-Per-Year-1946-2019.jpg.)

Figure 18.2 Number of fatal aircraft accidents per million flights 1958e2018. (Source: A Statistical Analysis of Commercial Aviation Aircraft 1958e2018.)

fatal aircraft accidents per million flights in 2018. Apart from some hick-ups, the trend has been continuously decreasing since. 1989 was the last tragic year with more than one fatal aircraft accident per million flights, due to several high impact crashes with Korean Air, China Airlines, PIA, UTA, and Cubana, among others. The above accident indicators also contain smaller aircraft types and military aircraft. As mentioned before, the current aircraft manufacturers market is a duopoly between Airbus and Boeing. Based on data from the Aviation Safety Network, the below graphs give an overview of the

420

The Air Transportation Industry

Figure 18.3 Number of accidents/fatal accidents with Boeing manufactured aircraft 1978e2020. (Source: Authors’ calculations based on ASN Aviation Safety Database (https://aviation-safety.net/database/).)

number of accidents and fatal accidents with at least one casualty for Boeing and Airbus airframes respectively over the period 1978 to 2020. Fig. 18.3 shows the stable, yet a slowly declining number of accidents and fatal accidents with Boeing aircraft. The number of yearly accidents varies between 25 and 35, and the trendline is slightly downward sloping. However, the number of yearly fatal accidents is decreasing significantly over the reference period. From around 15 yearly accidents with at least one casualty in the mid-1980s, the indicator decreased to almost 10 in the mid-1990s and stabilized below five from mid-2000 onward. In 2018 and 2019, the number of accidents with at least one casualty stabilized around 5. This increase is due to the two Boeing 737 MAX accidents and several other accidents like the Cubana 737-200 and the Amazon Air 767 accidents during these two years. The Airbus figures show a different outlook compared to the abovedescribed figures of Boeing. While the absolute number of accidents is steadily increasing over time, mainly due to the higher number of Airbus aircraft in service, the rate of accidents with fatalities remains low, with zero fatalities in 2018 and 2019. Boeing and Airbus have more than 10,000 aircraft in service currently, half of these are narrow-body aircraft like the Boeing 737 series and the Airbus 320 series. At first glance, one could derive that the last couple of years, given the current similar aircraft fleets in

The effect of accidents on aircraft manufacturers’ competition

421

Figure 18.4 Number of accidents/fatal accidents with Airbus manufactured aircraft 1978e2020. (Source: Authors’ calculations based on https://aviation-safety.net/database/.)

operation, Airbus performs better in the accident and fatal accident indicators than its main competitor Boeing. However, a full analysis of thednot publicly availableddata on aircraft incidents is needed to give substance to this interpretation (Fig. 18.4).

5. The impact of accidents on aircraft manufacturers’ competition Given the high growth figures for aviation in the last decades and the declining rate of accidents per million flights, one can conclude that air transport is a very safe way of flying. A. Barnett and M.K. Higgins (Barnett, 1990) concluded that, throughout the world, the mortality risk of air travel in 1977e86 was lower by a factor of four than in the early 1970s, and by a factor of nine lower than that of the early 1960s (Barnett, 1990). When the yearly aviation-related casualties are put into perspective in regard to other accident, homicide or disease rates, the aviation-related casualties are very low. However, according to Barnett (1990), the New York Times includes more page-one stories of commercial jets’ accidents than other mortal danger like cancer, AIDS and homicide from October 1, 1988 to September 30, 1989. In 2020, the air transport sector counted 423 casualties (Aviation Safety Network, 2021). More than half of these casualties were caused by the impact of two large accidents: the Ukraine International Airlines 737 NG that crashed after a rocket attack on the January 8, 2020 suffering 176 casualties, and the Pakistan International Airlines A320 that

422

The Air Transportation Industry

crashed on the May 27, 2020 leading to 98 casualties. Both accidents were not aircraft manufacturer related. The Ukraine International Airlines accident was caused by a tragic mistake done by the Iranian Army. A succession of fatal pilot errors caused the Pakistan International Airlines accident. The majority of the aircraft accidents from 2000 up to 2020 were caused by a problem outside of the airframe manufacturers’ control. The accidents’ leading causes were pilot errors, severe weather conditions, or unintended attacks by armed groups. However, the situation with the 737 MAX where two aircraft crashed due to a technical failure, was a gamechanger in the current aircraft manufacturers’ competitive environment. The aviation authorities grounded the aircraft worldwide for over a year. 387 Boeing 737 MAX aircraft were grounded, and about 450 produced aircraft were stored at Boeing facilities, awaiting deliveries to their respective customers. The general public suddenly lost faith in the Boeing 737 MAX product, and Boeing’s reputation suffered severe damage. The last real public faith loss in an aircraft type was with the DC-10 in the late 70s and early 80s, as described above. This phenomenon and several accidents with the MD-11, the successor of the DC-10, lead to the financial difficulties, decline and take over by Boeing of McDonnell Douglas. Boeing did not deliver 737 MAX aircraft for more than a year between the second half of 2019 and the first half of 2020 to its customers, while the 737 MAX series is by far the most successful product in the Boeing order book with more than 4 500 737 MAX aircraft on order. On November 18, 2020, following a comprehensive and methodical safety review process that took 20 months to complete, the FAA ungrounded the 737 MAX. Boeing delivered 118 aircraft from January till November 2020, down from 380 in 2019 and 806 in 2018. Boeing’s commercial aircraft unit has been seriously impacted by the combined impact of the March 2019 grounding of the 737 MAX and more recently the coronavirus pandemic, which has seriously reduced the demand for new planes. Besides, Boeing has recorded by the end of 2020 more than 1000 order cancellations or adjustments, of which 641 are Boeing 737 MAX jets (Reuters, 2021). In the meantime, Airbus was close to delivering 560 planes to customers as of December 31, 2020. The 2020 total is well short of the record 863 aircraft that Airbus handed over to customers in 2019, and 800 in 2018, but it still would be considered a success taking into account the widespread groundings of fleets while COVID-19 wiped out demand for travel (Bloomberg, 2021). These figures demonstrate that the grounding of the 737 MAX had a significant effect on Boeing’s aircraft deliveries in 2019 and 2020, while Airbus suffered less.

The effect of accidents on aircraft manufacturers’ competition

423

Airbus’ aircraft deliveries in 2020 decreased by 35% and 30% respectively compared to 2019 and 2018, while Boeing is forecast to deliver in 2020 67% and 84% less aircraft compared to 2019 and 2018. It must be noted that Boeing had at the end of 2020 450 737 MAX aircraft produced at a slower pace and stored at their facilities. However, if these stored aircraft were delivered to their customers once the respective local authorities had released the ban on the 737 MAX, these deliveries would not make up for the 2019 and 2020 figures. It is evident that the 737 MAX problems had a considerable impact on Boeing’s production output, financial situation and stock value. However, there is no evidence that Airbus picked up many of Boeing’s canceled orders with their competing A320 product. In order to investigate how aviation accidents affect airlines’ aircraft purchases, we have collected data on accidents from the Aviation Safety Network and commercial aircraft orders from Boeing (see Li, 2020 for more details). The data cover the time period 1978e2019. We regress monthly aircraft orders on lagged monthly numbers of Boeing accidents and present the results in Table 18.1. In all regressions, we include month fixed effects to account for seasonality in orders (not separately reported). We find that Boeing typically receives most of its orders in June and July and in November and December, likely coinciding with the end of the fiscal year for most airlines. In the first column of Table 18.1, we report results from a regression of Boeing orders on monthly lags of all types of accidents. We find that almost all coefficient estimates are negative, and they are jointly highly significant. Lags 5 and 8e11 are also individually significant, suggesting that the effect of accidents on orders is strongest close to a year after the accident. The largest individual effects are for months 9e11, with a reduction of 3.7e4.3 orders per month. The cumulative effect in the year after the accident is 23 orders. We find no statistically significant effect more than 12 months after the accident. We also examine the effect of fatal accidents only on orders. While one would expect a stronger effect on aircraft orders in this case, these types of accidents are quite rare which limits the statistical power of the test. Our results show negative coefficients for most lags, but they are no longer jointly significant. We find the largest effects for lags 9 and 10 with reductions of 7.9 and 5.7 orders per month, respectively, and these effects are statistically significant. We estimate that the cumulative reduction in orders over the full year after a fatal accident is 27.2.

424

The Air Transportation Industry

Table 18.1 The effect of accidents on Boeing aircraft orders. All Fatal All Boeing orders (monthly) accidents accidents accidents

Boeing accidents t-1 Boeing accidents t-2 Boeing accidents t-3 Boeing accidents t-4 Boeing accidents t-5 Boeing accidents t-6 Boeing accidents t-7 Boeing accidents t-8 Boeing accidents t-9 Boeing accidents t-10 Boeing accidents t-11 Boeing accidents t-12 Airbus accidents t-1 Airbus accidents t-2 Airbus accidents t-3 Airbus accidents t-4 Airbus accidents t-5 Airbus accidents t-6 Airbus accidents t-7 Airbus accidents t-8 Airbus accidents t-9

0.6126 (1.4353) 0.2424 (1.4302) 1.9071 (1.4306) 1.2586 (1.4337) 2.9938 (1.4313)b 1.9801 (1.4316) 0.9169 (1.4321) 2.5773 (1.4302)a 3.6839 (1.4292)b 4.2792 (1.4276)c 3.6669 (1.4320)b 0.6975 (1.4348)

1.5835 (3.1942) 1.4269 (3.1886) 5.4785 (3.1751)a 0.1285 (3.1773) 2.6626 (3.1687) 2.5849 (3.1668) 0.3949 (3.1669) 0.6326 (3.1658) 7.8589 (3.1741)b 5.7491 (3.1737)a 3.3772 (3.2006) 1.4712 (3.2049)

0.3619 (1.4213) 0.0701 (1.4151) 1.4275 (1.4152) 1.0528 (1.4207) 2.4792 (1.4186)a 1.4922 (1.4214) 1.2127 (1.4214) 2.2898 (1.4200) 3.4187 (1.4244)b 3.7156 (1.4240)c 2.9654 (1.4293)b 0.2262 (1.4318) 1.4237 (2.7278) 2.6462 (2.7218) 5.1754 (2.7055)a 4.0407 (2.7222) 2.8474 (2.7789) 0.4028 (2.7889) 7.0786 (2.7914)b 1.1185 (2.7961) 0.6611 (2.7754)

Fatal accidents

1.6127 (3.2695) 1.6859 (3.2635) 5.0110 (3.2471) 0.4252 (3.2422) 2.8748 (3.2217) 2.5803 (3.2219) 0.6996 (3.2228) 0.4037 (3.2285) 7.9007 (3.2430)b 6.3142 (3.2386)a 3.8529 (3.2663) 0.9510 (3.2714) 3.1941 (7.9033) 0.0084 (7.9315) 5.6692 (7.9582) 2.0192 (7.9851) 1.7646 (7.9840) 11.3415 (7.9831) 3.2140 (7.9831) 3.6878 (7.9856) 1.2216 (7.9970)

The effect of accidents on aircraft manufacturers’ competition

425

Table 18.1 The effect of accidents on Boeing aircraft orders.dcont'd Boeing orders (monthly)

All accidents

Fatal accidents

All accidents

Fatal accidents

Yes 0.0001

Yes 0.1350

0.6417 (2.7472) 1.1531 (2.7668) 0.1376 (2.7756) Yes 0.0075

1.3473 (7.9946) 6.1869 (7.9592) 3.7970 (7.9309) Yes 0.1388

e

e

0.0124

0.9631

0.186 492

0.149 492

0.229 492

0.158 492

Airbus accidents t-10 Airbus accidents t-11 Airbus accidents t-12 Month fixed effects Joint F-test on Boeing variables (p-value) Joint F-test on Airbus variables (p-value) R-squared Observations

Notes: The table shows results from regressions of monthly orders of civilian Boeing aircraft on lagged accidents of Boeing and Airbus planes for the last 12 months. Standard errors are in parentheses. a p < .1. b p < .05. c p < .01.

We investigate whether accidents on one manufacturer’s planes lead airlines to switch to the other manufacturer. Given that the market for large commercial aircraft is a duopoly, this would mean that accidents on Airbus planes would have airlines switch from Airbus to Boeing. We estimate how Boeing’s orders are affected by Airbus accidents in the last two columns of Table 18.1. We again consider all accidents (Column 3) and fatal accidents only (Column 4). In both cases, we control for the effect of Boeing’s own accidents and for month fixed effects. We find only weak evidence that accidents lead to a substitution in aircraft orders to the other manufacturer. In the case of all accidents, the majority of coefficients are positive, suggesting that a recent Airbus accident increases demand for Boeing aircraft. However, only lags 3 and 7 are statistically significant and an F-test rejects joint significance of all 12 lags. The magnitudes of the effects are similar in size as the effects of Boeing’s own accidents. When we investigate the impact of fatal Airbus accidents, our results in Column 4 show no indication of any effect of these accidents on Boeing orders. However, we have to contend again with the very small overall number of these accidents resulting in low statistical power. The next part of this chapter looks into some reasons why airlines tend to stick to either Airbus or Boeing for future orders.

426

The Air Transportation Industry

6. Why are airlines so faithful to their chosen aircraft manufacturer? As discussed in the previous parts of this chapter, airlines tend to hold on to either Airbus or Boeing airframes for future orders when renewing or expanding their fleet. Aircraft accidents hardly influence the orders of the impacted or competitive airframe manufacturer. However, the Boeing 737 MAX disasters lead to several order cancellations. Boeing reported that by the end of November 2020, 538 Boeing 737 MAX aircraft were canceled, most of these cancellations were 737 MAX aircraft (ABCnews, 2020). However, it is not entirely clear if the airlines canceled these aircraft due to the aircraft’s long-time grounding and loss of faith in the aircraft type, or whether they took advantage of contractual possibilities to cancel the orders given the dire financial situation at many passenger airlines. When analyzing the worldwide airlines’ fleets, one can observe two significant trends that favor airlines to stick to one airframe manufacturer, that is, high switching costs and cockpit commonality. One trend favors operating both aircraft manufacturers’ models, that is, negotiation leverage. These trends will be discussed next. Low cost airlines tend to standardize on one aircraft type and stick to that aircraft type for fleet renewal and expansion. Standardization enhances economies of scale in the aircraft’s purchase and operations, creates simplicity of the company processes, enhances flexibility in the daily operations, and reduces costs for training, maintenance, and ground handling. Both Ryanair and Southwest Airlines started their LCC business model with a small fleet of 737-200 aircraft, but gradually expanded their fleet with 737 CG and 737 NG aircraft. These airlines also ordered a large number of 737 MAX aircraft. Easyjet started with some Boeing 737s but switched gradually to a full Airbus A320 series fleet. Asian LCC Air Asia and Hungarian based Wizzair also operate a unique fleet of only Airbus 320 series. Although these LCC airlines go onto the market to get the best prices, they tend to stick to Boeing or Airbus depending on their current fleet composition. The aforementioned reasons prevail to stick to one single aircraft type. The costs and complexity to operate aircraft from two manufacturers simultaneously during the transition period and in parallel train the crew, change processes, schedule the aircraft and reduce flexibility make this dual operation too costly for an LCC. The costs to switch from Airbus to Boeing or vice versa are too high to offset potential purchase price

The effect of accidents on aircraft manufacturers’ competition

427

advantages. Even in case of some severe technical dysfunctions as witnessed with the Boeing 737 MAX, Ryanair, Southwest Airlines, and many other airlines stuck with these aircraft’s orders. Boeing launched the first cockpit commonality with the wide-body Boeing 767 in 1981 and the narrow-body Boeing 757 launched in 1982. Pilots could, after a short conversion course, fly both aircraft models. Since the launch of its first fly-by-wire aircraft, the narrow-body A320, in 1987 Airbus has prioritized commonality throughout its product lines. Identical cockpits and operating procedures were applied to each of the A320 Family’s versions, the A318, A319, A320, and A321, allowing pilots to fly all these aircraft with a single type rating. By creating a standard process, Airbus streamlined and simplified many aspects of flying and aircraft maintenance. Their goal was to make operations, training and maintenance simpler and less expensive for customers by applying a high interchangeability of processes, systems, and parts. The narrow-body A320, the twin-engine A330 (including its A330-900 and A330-800 variants), the A350 wide-body and double-deck A380 feature nearly identical flight decks and similar handling and flying characteristics. This commonality enables Cross-Crew-Qualification of pilots, which reduces pilot training time while bringing significant savings through reduced maintenance and streamlined procedures. Another advantage of this Airbus commonality is Mixed Fleet Flying, which is a pilot’s ability to be current on more than one Airbus fly-by-wire aircraft type at a time. For example, this capability enables a pilot rated on an A330 to switch from long-haul operations to short- or medium-haul flights in an aircraft of the A320 Family. This feature is especially interesting for start-up or smaller airlines who like to combine two aircraft types for their operations and achieve economies of scale, even when the number of aircraft in the fleet is limited to achieve economies of scale with only one aircraft type. Boeing has the same cockpit commonality, however, at a lesser level. The Boeing 737 MAX was designed so that Boeing 737 NG pilots could with minimal training fly on a 737 MAX. The MCAS simulating similar flying characteristics as the Boeing 737 NG would facilitate this. The Boeing 777 and 787 have the same type rating for pilots, allowing pilots to interchange, with minimal training, between both aircraft types. These features make airlines reluctant to switch from one airframe manufacturer to another. The costs of change are very high during the

428

The Air Transportation Industry

transition process and are hardly ever compensated by the new aircraft manufacturer’s better purchase conditions. This fact is one major reason why airlines tend to stick to the aircraft manufacturer of their current fleet. Another trend is that legacy carriers like to keep their future aircraft purchase options open and operate both extensive Airbus and Boeing fleets. These airlines tend to opt either for Boeing 737 or Airbus 320 to operate the short-haul flights, Boeing 777 of Airbus 380 for the long-haul flights, and either Boeing 787 or Airbus 330 or 350 to operate thinner or highfrequency routes. Airlines often differentiate within these aircraft types to opt for variants with a smaller or larger capacity, shorter or longer-range and lower or higher MTOW to enhance network flexibility. The fleet size within the respective segments is considered to be large enough to achieve economies of scale. The airlines are striving to avoid to be too dependent on one single aircraft manufacturer. British Airways, previously an almost entirely Boeing aircraft operator, ordered in 1998,188 Airbus 320 aircraft for its short-haul operations, while the market expected British Airways to order Boeing 737 aircraft. Delta Airlines, a longstanding Boeing customer, ordered in 2014 50 Airbus 350 aircraft for its long haul operations to replace its aging Boeing 767 and 747 aircraft fleet. The market expected Delta Airlines to opt for the Boeing 787 aircraft instead as Delta Airlines was already operating 18 Boeing 777 aircraft. Another way to avoid being too dependent on one aircraft manufacturer is by ordering aircraft from another manufacturer for another airline in the same airline holding company. Both the Air France KLM and IAG group operate similar aircraft models but from another aircraft manufacturer in the same airline group. Air France focuses on the Airbus products A320 for short-haul operations, A330 for medium-haul operations and A350 and the recently retired A380 for long-haul destinations. However, sister company KLM operates mainly Boeing products like Boeing 737 for shorthaul, Boeing 787 for medium-haul and Boeing 777 and the recently retired Boeing 747 for long-haul operations. Within the IAG group, Iberia and Aer Lingus operate a full Airbus fleet of A320 and A330, while British Airways operates a mixed Boeing and Airbus types fleet. Even LCC Ryanair operates a small fleet of A320s in its Austrian subsidiary Laudamotion and gets in this way familiar with the Airbus 320 aircraft’s operational characteristics. This information can be used by Ryanair as negotiation leverage during purchase negotiations with Boeing. This fact reinforces the above-described trends that airlines do not switch easily from one aircraft manufacturer to another. Even in case of an

The effect of accidents on aircraft manufacturers’ competition

429

accident or a series of accidents, airlines do not easily switch their current aircraft manufacturer to the competitive manufacturer. Economic calculations and deliberations based on recurring and one-off operating costs are primarily driving the future fleet acquisition strategy of airlines.

7. Conclusion Safety has always been paramount to the success of an airline’s business model. Aircraft accidents, in particular fatal accidents, can impact not only the airline’s reputation and profitability but also airframe manufacturers. While previous work has mostly focused on the impact of accidents on airlines, this chapter provided new evidence on the effect of accidents and fatal accidents on aircraft manufacturers. Between 1950 and 80, several severe aviation accidents lead to the distrust of the passengers, resulting in the production stop of some aircraft models. However, the fast rollout and expansion of air transport and the rapidly improving safety indicators to an almost zero fatality rate for air travel in the last 40 years lead to growing public confidence in all Westernbuilt aircraft models. This chapter concluded that there is little evidence that, in the past, aircraft accidents have had a long-term impact on aircraft orders with the competing and non-impacted aircraft manufacturer. It remains to be seen whether orders for this aircraft type recover as the economy improves or if some airlines will permanently shift to competing products such as the Airbus A320. This chapter discussed airlines’ tendency to hold on to either Airbus or Boeing airframes for future orders when renewing or expanding their fleets. Two significant trends were observed that explain why airlines favor holding on to one airframe manufacturer, i.e., high switching costs and cockpit commonality. The costs and complexity of switching to another aircraft manufacturer and operating two aircraft manufacturers simultaneously during the transition period and in parallel train the crew, change processes, schedule the aircraft, and reduce flexibility make this dual operation very expensive for an airline. Besides, cockpit commonality facilitates standardization. This standardization enhances economies of scale in the aircraft’s operations, creates simplicity of the company processes, enhances flexibility in the daily operations, and reduces costs for training, maintenance, and ground handling. The costs to switch from Airbus to Boeing or vice versa are therefore too high to offset potential purchase price advantages.

430

The Air Transportation Industry

An opposing factor, however, is negotiation leverage which favors operating both aircraft manufacturers’ models, mainly in airlines with a large fleet. The fleet size of large airlines within the respective segments is large enough to achieve economies of scale. The airlines are striving to avoid being not dependent on one single aircraft manufacturer. This chapter concludes that there is little evidence that aircraft accidents, including fatal accidents, have an impact on aircraft manufacturers’ competition. However, this may change as the Boeing 737 MAX disasters in 2018 and 2019 have led to considerable passenger distrust toward this aircraft model and several order cancellations. We have yet to see whether the combined impact of the Boeing 737 MAX accidents and the COVID19 pandemic on the international air transport will change the preCOVID-19 equilibrium in the duopoly between Airbus and Boeing in the future.

References ABCnews, 2020. https://abcnews.go.com/Business/wireStory/boeing-suffers-canceledorders-737-max-plane-74605476#:w:text¼Boeing%20Co.%20reported%20more%20 cancellations,to%20536%20for%20the%20year. Airbus, 2020a. https://www.airbus.com/aircraft/passenger-aircraft/a220-family.html. Airbus, 2020b. https://www.airbus.com/investors/financial-results-and-annual-reports.html. Airbus, 2020c. https://www.airbus.com/aircraft/passenger-aircraft/commonality.html. Aviation Safety Network, 2021. https://aviation-safety.net/database/dblist.php?Year¼2020& lang¼&page¼1. Barnett, A., 1990. Air safety: end of the golden age? Chance 3 (2), 8e12. Besanko, D., et al., 2017. Economics of Strategy, seventh ed. John Wiley & Sons Inc. Bloomberg, 2021. https://www.bloomberg.com/news/articles/2021-01-02/airbus-said-todeliver-about-560-planes-last-year-with-late-push. Boeing, 2020a. https://investors.boeing.com/investors/investor-news/press-release-details/ 2020/Boeing-Reports-Fourth-Quarter-Results/default.aspx. Boeing, 2020b. https://investors.boeing.com/investors/investor-news/press-release-details/ 2020/Boeing-Reports-Third-Quarter-Results/default.aspx#:w:text¼The%20Boeing%20 Company%20%5BNYSE%3A%20BA,flow%20of%20(%244.8)%20billion. Camerer, C.F., Howard, K., 1989. Decision processes for low probability events: policy implications. J. Pol. Anal. Manag. 8 (4), 565e592. Chalk, A., 1987. Market forces and commercial aircraft safety. J. Ind. Econ. xxxvi (1), 61e82. Chance, D., Ferris, S., 1987. The effect of aviation disasters on the air transport industry: a financial market perspective. J. Transport Econ. Pol. 21 (May 87), 151e165. Chang, Y.-H., Yeh, C.-H., 2004. A new airline safety index. Transp. Res. Part B 38, 369e383. Hwang, C.L., Yoon, K.S., 1981. Multiple Attribute Decision Making: Method and Applications. Springer, New York. Ho, J.C., et al., 2013. Investor sentiment risk factor and asset pricing anomalies. Journal 2013 New Zealand Finance Colloqium.

The effect of accidents on aircraft manufacturers’ competition

431

Janic, M., January 2000. An assessment of risk and safety in civil aviation. J. Air Transport. Manag. 6 (1), 43e50. Kaplanski, G., Levy, H., 2010. Sentiment and stock prices: the case of aviation disasters. J. Financ. Econ. 95 (2), 174e201. Li, Y., 2020. “The Impact of Aviation Accidents on Aircraft Manufacturers”. Thesis Submitted. Tufts University. Liou, J.H., Tzeng, G.-H., Changa, H.-C., July 2007. Airline safety measurement using a hybrid model. J. Air Transp. Manag. 13 (4), 243e249. Olienyk, J., Carbaugh, R.J., 2011. Boeing and Airbus: Duopoly in Jeopardy? Global Econ. J. 11 (1). Article 4. Reuters, 2021. https://www.reuters.com/article/us-boeing-deliveries/boeing-limps-into2021-with-more-737-max-cancellations-delayed-787-deliveries-idUSKBN29H28M. Rose, N.L., 1990. Profitability and product quality: economic determinants of airline safety performance. J. Polit. Econ. 98 (5), 944e964. Rose, N.L., 1992. Fear of flying economic analyses of airline safety. J. Econ. Perspect. 6 (2), 75e94. Walker, T.J., et al., 2005. On the performance of airlines and airplane manufacturers following aviation disasters. Can. J. Adm. Sci. 22 (1), 21e34.

This page intentionally left blank

CHAPTER 19

How strategy can influence the market: recommendations and conclusions Rosário Macário1, 2 and Eddy Van de Voorde2 1

CERIS, Instituto Superior Técnico, Universidade de Lisboa, Lisbon, Portugal; 2Department of Transport and Regional Economics, University of Antwerp, Antwerp, Belgium

1. Introduction The aviation sector comprises a variety of actors, many of whom have conflicting interests. Typical examples of such conflicts occur within the context of negotiations between airlines and ground handling companies with regard to price, volume, and quality. In order to understand how the various aviation actors position themselves within a competitive market, this book analyzes the various market segments in detail. What is the industrial-economic structure driving the decisions? What are the main economic problems? What are the consequences for the negotiations among the various actors? What impact does this have on the global aviation market? Is the aviation market evolving toward more horizontal and vertical cooperation? This book provides insight into the relationship between the various economic agents of the air transport industry, the possible strategic choices, and the mutual competitive strength that is likely to exist within the future aviation market. Our aim is to use the various contributions to highlight the industrial-economic and strategic struggle that underlies the current market structure, for aviation as a whole, as well as for the constituent actors (e.g., carriers, authorities, handlers). In other words, we seek to describe the specific market (and sub-markets) and identify the main economic problems. In the following sections, we examine the theoretical underpinnings of the existing market forms within the aviation sector. Despite the high level of heterogeneity within this sector, it is still possible to identify the industrial-economic variables that drive each of the actors. We illustrate this according to what could be considered an average aviation model. We then refer to the chapters in this book to address the current market The Air Transportation Industry ISBN 978-0-323-91522-9 https://doi.org/10.1016/B978-0-323-91522-9.00009-9

© 2022 Elsevier Inc. All rights reserved.

433

434

The Air Transportation Industry

(and sub-market) forms for each actor, as well as the developments that may be imminent. Finally, we present a synthesis of future strategies, focusing on who does what and when.

2. Market structures The global aviation industry has undergone major changes worldwide since deregulation. The global phenomenon of liberalization has combined with increasing competition to force airlines around the world to seek ways to operate more efficiently. This has sparked an ongoing transformation within the industry, involving a large number of actors (Belobaba et al., 2016, p. 487). Airports have sharpened their focus on business operations and developed new non-aeronautical activities. Ground handlers have become global operators in many parts of the world. The result is a global aviation industry operating within an international regulatory environment that ranges from strict regulation and protectionism in some countries or regions to near-complete deregulation in others (Odoni, 2016, p. 45). This situation has resulted in an uneven playing field. There are obviously limits to growth. In many parts of the world, capacity restrictions at airports are becoming increasingly severe. At the same time, most airports have opted for strategies of expansion. This can lead to serious problems, due to flight delays, cancellations and the resulting low reliability of flight schedules and the tremendous associated costs (Odoni, 2016, p. 45). It is therefore not surprising that various international and national institutions and organizations, both public and private, are responding to the need for policymaking on economic, regulatory, and technical matters. Relevant questions concern the extent to which certain decisions will be guided by current and future market forms within the global aviation sector. Conversely, it is important to consider the extent to which existing market forms are affected by exogenous and endogenous decisions. Any research in this regard must begin by examining the strong link that connects economic activities and economic growth to the aviation sector as a whole. The next phase involves investigating the various relationships between the actors within the system. This is because the actors do not all contribute equally to success (or failure). As correctly argued by Odoni (2016, p. 45), the investigation of any aspect of the global aviation industry must proceed from the complex regulatory, legal, and institutional context within which the industry

How strategy can influence the market: recommendations and conclusions

435

operates. For just about any airline, success and survival depends on identifying an operational and financial model that will ensure sustainable profitability and allow the airline to continue to finance the necessary new investments. The airline industry is both capital-intensive and laborintensive, and it is subject to the cyclicity of macroeconomic forces (Belobaba et al., 2016, p. 500), as well as to the consequences of severe shocks, such as the COVID-19 pandemic. The prevailing market forms are inextricably linked to the form of competition. Neoclassical theory distinguishes four theoretical market structures: perfect competition, monopolistic competition, oligopoly, and monopoly. Perfect competition is based on the presence of six important conditions.1 If these conditions are met, there will be a competitive equilibrium in which all companies generate a normal profit. Perfect competition also imposes discipline, in the sense that all companies present are forced to produce as efficiently as possible, given the current state of technology. In practice, competition often results in a market or industry structure consisting of a limited number of large companies. In this situation, each company has sufficient market power to set its own pricing, with some or all companies being able to realize abnormal profits in the long run. One possible reason why competition within a market can ultimately lead to a decline in the number of companies is that, as companies grow, they achieve economies of scale and average costs decrease. In the most extreme case, this can produce a natural monopoly, in which a single company can produce at a lower average cost than any combination of competing companies. A monopoly is not necessarily negative, if the cost savings that could be realized from scaling up are passed along to consumers in the form of lower prices. In many situations, however, a monopolist will use its power to limit output and raise prices, thus generating abnormal profits for itself, while creating harmful consequences for consumer welfare. In imperfect competition, a distinction is made between monopolistic competition and oligopoly. Monopolistic competition is based on the assumption that the number of companies will remain large, and it 1

These six conditions are as follows: (1) there is a large number of buyers and sellers; (2) producers and consumers have perfect market knowledge; (3) the products sold are identical; (4) the companies are independent of each other and aim to maximize profit; (5) companies can enter and exit freely; and (6) companies can sell as much output as they wish at the current market price.

436

The Air Transportation Industry

emphasizes both price and non-price forms of competition. An oligopoly assumes a small number of companies (but more than one). The companies are aware of their interdependence. Changes in price or output by one company have an effect on the profit generation of competing companies, causing them to adjust their own prices and production levels. Within an oligopoly, competition ranges from fierce price competition, which can often lead to significant losses, to collusion, in which the companies make joint decisions about their prices and production levels. The neoclassical theory outlined above is based on a static conception of competition, with an emphasis on long-term equilibrium. Some researchers (e.g., Schumpeter and the Austrian School), however, have opted for a more dynamic approach (see, e.g., Lipczynski et al., 2009, pp. 74e75). The fact that a company (monopoly) makes an abnormal profit does not constitute evidence that the company is guilty of abusing its market power (cf. monopoly position) at the expense of the consumer. Monopoly profits are also good for motivating and guiding entrepreneurs in making decisions that ultimately improve the allocation of scarce resources. One important principle in this regard is that competition is driven by innovation. This dynamic view of competition argues that monopoly status is only a temporary phenomenon and that is incapable of maintaining a stable, long-term equilibrium, as is argued by neoclassical theory.2 The characteristics of the market structure of the aviation sector are important in this regard. The aviation chain as a whole is profitable, but not all actors within the chain are able to optimize their profits and share in the profitability of the chain. This is undoubtedly due to the fact that some actors (e.g., airport operators) have more power (e.g., in negotiations) than do others (e.g., handling companies). In the following section, we discuss the current market structure. We do not attempt to provide a detailed explanation of variations in performance between sub-sectors and actors, as often measured by the profitability of the business. We also do not go into detail when identifying structural variables at the industry level (e.g., concentration, economies of scale, and entry and exit conditions) or the main determinants that actually drive business performance.

2

The static and dynamic theories have been interpreted empirically within the research field of industrial organization.

How strategy can influence the market: recommendations and conclusions

437

3. Current market structure This section is limited to a simple exercise that demonstrates the operation of the current market structure for a single airport (i.e., Brussels). For each sub-actor, we focus on such aspects as the number and size of buyers and sellers, entry and exit conditions, product differentiation, vertical integration, and diversification. At this point, therefore, we do not consider the actual behavior or specific performance of companies. Following this relatively static exercise, we provide a more detailed discussion of the main conclusions of the various chapters in this book, which can be used to identify potential new developments and determine possible strategies for the future. The discussion proceeds from a figure presented in Chapter 1, which illustrates the high level of heterogeneity existing within the aviation landscape, drawing from an overview of the mutual relationships between actors, based on pricing and invoices. The central actor is the airport manager or operator. Important customers include the carriers that bring passengers and cargo. Although in some cases, a single carrier provides the greatest share of the landings, relative market power is usually limited, partly because highly competitive airports are coordinated and the limited capacity is distributed through slot allocation. In most cases, the government is responsible for regulation. Airport operators and/or regulators often grant licenses for handling aircraft to handling companies based on an auctioning process.3 Several service providers are licensed according to an agreement with the airport provider, concluding contracts directly with the carriers. Shops and parking operators conclude agreements with the airport operator. Although we currently have no figures that allow the quantification of the interrelationships between actors, based on who pays what amount to which actor and at what time, the mapping of relationships between actors seems to be valid and scalable for all airports. The most important actors are identified in Fig. 19.1, with the limitation that the figure is based on economic operations, in which the airport is central. As a result, some important aviation actors (e.g., aircraft manufacturers) are not explicitly included, but embedded within the microeconomy of the carriers.

3

For example, in EU regulation, the number of licenses for ground handlers increases in proportion to the number of aircraft movements and passengers.

438

The Air Transportation Industry

FREIGHT

PAX

freight

taxes

AIRPORTS

CARRIERS

Ticket + taxes +sales on board fees

Handling fees

Government

landing fee + taxes pax

i

concession

HANDLERS SERVICE PROVIDERS F U E L

CATE RING

MAIN TEN ANCE

PARKING



concession

RENT-A-CAR

… .

sales

SHOPS

Figure 19.1 Pricing and payment of airport bills. (Own composition).

In Table 19.1, we translate the existing market structure to the environment as we know it today, as specifically applied to the case of Brussels Airport. We describe the current market structure in terms of the number of players present, the possible mutual relationships (e.g., who grants which license to whom), the form of regulation that applies, and the possible barriers to entry and/or exit. As indicated in Table 19.1, there are many clear relationships between the various market actors. The airport operator plays an important role in this regard, and therefore has distinct market power. Although regulation is in place, its scope and impact are limited. One important question concerns the extent to which the model applied to Brussels Airport can be considered within a more generic context. For regional airports (which are usually much smaller), it is known that, in many cases, such airports (both hubs and non-hubs) are used by only one airline, which can hold a considerable amount of bargaining power. This is because both parties (i.e., the airport and the airline) are fully aware that, if negotiations fail to produce a workable agreement, the loss of the airport’s only important customer greatly increases the likelihood of closure for that airport.

How strategy can influence the market: recommendations and conclusions

439

Table 19.1 Airport actors: a case study of Brussels Airport. Actor Players Important features

Government (1)

Regulator

Government (2)

Civil Aviation Authority

Government (3)

Slot coordinator

ATC provider

Skeyes

The following activities are currently regulated in Brussels: Landing and take-off of aircraft; parking of aircraft; use of terminal facilities by passengers; kerosene deliveries to aircraft through centralized infrastructures; operations aimed at ensuring the safety and security of passengers and airport facilities. The Civil Aviation Authority grants traffic rights based on EU and Belgian law, as well as on bilateral agreements. The allocation of traffic rights depends on the airline’s credentials and the type of flight (commercial, noncommercial, planned, unplanned). Brussels Airport is a coordinated airport, and airlines must therefore request slots in order to operate. The slots at the airport are coordinated by the Belgium Slot CoordinationdBSC vzw. The ATC provider for Brussels Airport is skeyes, an autonomous public company that is responsible for flights over Belgium from the ground level to flight level 245. Terminal navigation costs for takeoff and landing at Brussels Airport are charged by skeyes, and they are independent of the rates and fees of Brussels Airport. Continued

440

The Air Transportation Industry

Table 19.1 Airport actors: a case study of Brussels Airport.dcont’d Actor

Players

Important features

Airport operator

Private company (Brussels Airport Company) Capital composition: 39% Canadian pension fund OTPP; 36% consortium of the Dutch pension fund APG, the Australian investor QIC, and insurer Swiss Life; 25% þ 1 share of the Belgian federal government. At the beginning of 2020, approximately 60 continental and inter-continental carriers were flying to Brussels. Different sizes. Hub for Brussels Airlines, which accounts for approximately 40% of all passenger flights at Brussels Airport.

Regulation based on European Directive 2009/12/EC, which states that there must be an independent supervisory authority. There is no risk of any abuse of market power (cf. existing regulation and competition linked to location).

Airlines

Ground handlers

Platform handlingb: Two passenger licenses and three cargo licenses awarded by Brussels Airport.c

Catering

Licenses granted by Brussels Airport Company.e

Independent slot allocation Regular consultation between airport operator and users with regard to airport charges, transparency, non-discrimination and independent supervision Five-year periods of regulation, with annual consultation Extinguishing “adapted single till” systema Benchmark exercise required, as tariffs are geared to a set of reference airports (þadditional control by regulator based on a financial model). Access to the ground handling market at Brussels Airport is regulated in the Royal Decree of November 6, 2010 (the “Royal Decree for ground handling”), which is a transposition of European Directive 96/67/EC. The licenses are valid for a period of seven (7) years.d The catering licenses at Brussels Airport are held by two companies: Newrest Servair and LSG SkyChefs.

How strategy can influence the market: recommendations and conclusions

441

Table 19.1 Airport actors: a case study of Brussels Airport.dcont’d Actor

Players

Important features

Fuel providers

Two companies and one provider

Maintenance

Several companies

Shops

Many companies

Parking

Concession with one operator on airport territory; many alternative parking lots in the vicinity of the airport.

Skytanking and Brussels Airfuels Services (B.A.S.) as handler with license, for transport on tarmac. Airlines can contract directly with fuel suppliers (e.g., Total, Q8, .) and they are supplied by HRS, the concession operator for the infrastructure. Number of players not limited (current suppliers include Brussels Airlines, Sabena Aerospace, Lufthansa Technik, and TUI). Commercial agreements between Brussels Airport Company and shop operators. Pricing concession fee depending on location, size, and other factors. Concession with the interparking company

In an “adjusted single till” system, a part of the proceeds from commercial activities can be used to finance regulated activities. This system is ultimately replaced by a “dual till” system, in which there is no cross-subsidization of regulated activities by commercial activities. b Platform handling includes the handling of baggage in the airport building and on the tarmac, as well as the de-icing of aircraft and the management of lost baggage. c For passenger handling, the two licenses of Aviapartner and Swissport were renewed in February 2018. Swissport went bankrupt in June 2020. Following a temporary appointment, the French company Alyzia received the second license effective December 2020. For freight, the quantity threshold was exceeded and a third license was added and assigned to the dnata company. d The number of licenses to be awarded was originally two for passenger handling and two for cargo handling. A third cargo handler was provided by Royal Decree, as more than 140,000 tons of cargo had been transported by full-cargo aircraft in the year before the tender. For passengers, the threshold of 24 million passengers was not reached. e Catering transport includes the transportation, loading on and unloading from the aircraft of food and beverages (excluding duty-free products for customers). The current license period is from October 28, 2018 to October 27, 2025. Source: Table composed by authors. a

442

The Air Transportation Industry

4. What will the future bring? A detailed comparison of other airports to the model of Brussels Airport (as described in Table 19.1) could be quite informative. Not every airport must necessarily follow the same model structure. Differences in operation often begin with differences in the ownership structure of the airport operator, which is just as likely to involve full government ownership as it is to involve the private corporate ownership described in the table (with the government being an important minority shareholder). This difference alone could have an important effect on the market structure and the rest of the aviation chain. One important question at this point concerns the extent to which future developments are likely to have an influence on the existing market structure. Relationships between the various market actors revolve around relative bargaining power, and future developments could potentially change these mutual relationships. Each of the sub-sectors of aviation activities analyzed in this book will be confronted with major changes within their own environments, as well as throughout the aviation chains within which they operate. It is interesting to note that all actors function within a broad framework of continuing economic conflict and competition. Rather than providing an overview of the main conclusions of each chapter separately, we position these conclusions within the broad competitive framework of the global aviation industry. In doing so, we consider several crucial questions. To what extent will the expected future developments change the current market structure, as reflected in Table 19.1? What are the consequences for the global aviation industry? Will the market and power relationships between the various actors change? How will shareholders respond to these developments? In light of these questions, we aim to develop a synthesis of the most important expected developments within the various links of the aviation chain. Such considerations can in themselves serve as input for estimating several possible future strategies, proceeding from the central question of who or what will become a valuable catalyst for the air transport industry (Table 19.2).

How strategy can influence the market: recommendations and conclusions

443

Table 19.2 Airport actors: potential future developments. Actor What the future will bring

Government (1)

Government (2)

Government (3)

ATC provider

At the policy level, it will be important to learn that policy (by definition, a purpose of action) should develop into a more proactive and strategic formula, leaving enforcement to agents who have sufficient proficiency in that function. In each country, the existence of clear strategic options for the future of air transport will be a fundamental asset in the industry’s development, and it will provide indispensable guidance for planning agencies, and even for operators. As societies become more demanding in general, both policy makers and policy takers will be required to enhance the quality of their decisions by enhancing transparency and providing deep scientific support. In democratic states, legislation will increasingly emphasize the independence of regulators as a basic requirement, while simultaneously ensuring that their accountability is more visible to the public. Regulators will bear the greatest obligation to ensure the development of adequate research in order to ensure their own competence in guiding the market toward the strategic objectives that have been defined, taking into account a good balance between challenges relating to sustainable development and the economic feasibility of the industry. Slot coordinators will have responsibilities extending beyond the current allocation of slots, recognizing the existence of a variety of options for the fair allocation of slots, each of which will produce different results and have a different impact within the competitive environment. Even though slot allocation is an instrument of airport competition, slot coordinators must be independent from airports and airlines. The extent to which slot coordinators will move closer to regulators remains to be seen. ATC will inevitably undergo a shift toward private ownership. This will require strong, effective regulation from the economic perspective, as well as in terms of safety and security. This structural shift will be one of the most challenging developments within the air transport industry. Many potential shareholders will show interest in this business, some from within the sector, and others from external environments. The independence of ATC providers will definitely be an indispensable requirement for success. Continued

444

The Air Transportation Industry

Table 19.2 Airport actors: potential future developments.dcont’d Actor

What the future will bring

Airport operator

Airport operators will be more actively engaged in international networks. More than ever, airports will be conscious that they have no monopolistic power. One major question will concern the relative resilience of airlines and the strategies that they will need to pursue in order to recover from the dramatic crisis imposed by the pandemic. As in previous crises, the industry will be dynamic. Bankruptcy will be observed in some cases. The exceptional situation resulting from the pandemic has led to a process of flexibilization within the EU and the US with regard to government aid. This development will open the door new discussions on regulation. As commercial airlines seek to secure markets, some regenerated form of interline agreements is likely to emerge as a means of ensuring specific levels of demand. Network economies will play a role in the regeneration of commercial instruments. Ground handlers will consolidate, and we are likely to observe a process in which a limited number of global operators stabilize within the market and enter into alliances. Small operators will tend to be franchised by global operators. Planning activities will benefit from aggregation, and technology will enable remote provision (e.g., flight operations, planning). Catering products will be minimized for economy and touristic classes, while becoming more sophisticated as a relevant frill for business and first class, especially on longhaul flights. Fuel will remain a critical input for airlines. In the past 20 years, oil prices have been erratic and subject to many geopolitical dynamics. They will nevertheless continue to be a major cost component for airlines, in addition to being a key to their success or failure. The link between the fundamentals and the price of kerosene is fragile. Oil has become a financial speculative product. New rules are likely to emerge for forward-looking and future markets. One crucial question will concern how the kerosene supply at airports is to be organized. Maintenance will be characterized by independent companies, often with capital ownership from lessors and/ or airlines. Maintenance will tend toward specialization, in order to obtain economies of scale and scope that would otherwise be difficult to achieve.

Airlines

Ground handlers

Catering

Fuel providers

Maintenance

How strategy can influence the market: recommendations and conclusions

445

Table 19.2 Airport actors: potential future developments.dcont’d Actor

What the future will bring

Shops

Airports will continue to invest in the provision of a favorable level of diversity to their final customer (i.e., the customer of the customer). The non-aeronautical commercial activities of airports will be an important means of revealing the core purpose of an airport from the perspective of the customer. Such diversity will continue to part of the competitive attraction of an airport to transit passengers. Parking will continue to constitute one of the most important businesses in airports, with rentals accounting for a significant share of the space and entering into partnerships with airline operators. The strengths of these activities are quite visible in airports in the US, and parking-related activities are likely to spread to other continents as a means of generating profits.

Parking

Table composed by authors.

5. Conclusions The aviation sector has several specific characteristics that make it unique. First, the sector is heterogeneous by nature, with wide differences between market actors, each with its own market structure. Despite the crucial importance of studying the interrelationships existing between actors to understanding the global market structure, the data that are currently available do not always allow such research. This situation is complicated by the volatile nature of the global aviation sector, which is driven by both exogenous factors (e.g., kerosene prices) and endogenous factors (e.g., the pursuit of economies of scale through mergers and acquisitions). This book represents an attempt to enhance the scientific understanding of the market structure of the aviation industry. The COVID-19 pandemic has caused the most severe crisis that the aviation sector has ever experienced. The levels of air traffic observed in 2019 are not projected to return before 2023 at the earliest. Government support has played a particularly fundamental role in avoiding bankruptcies, while simultaneously serving as an important source of conflict, particularly for low-cost airlines. This situation could potentially result in the reformulation of economic regulation within the sector.

446

The Air Transportation Industry

Issues relating to labor have experienced some of the most profound effects of the crisis, extending across all agents within the air transport industry. Such issues constitute a heterogeneous area that will require an international approach in order to avoid the undue exploitation of labor by some specific segments. Airline pricing appears to be a crucial instrument within the highly competitive development of the aviation business. At the same time, aggressive pricing can pose a barrier to entry within the market. Given that the aviation market is in constant development, very little generalization is possible. Nevertheless, each case provides an opportunity to learn about the behavior of agents and the effectiveness of incentive signals. This highlights the need for a pricing laboratory within the industry, which would make it possible to re-consider regulation and obtain more sophisticated performance-measurement mechanisms, thereby enhancing transparency within the aviation market. Fuel prices constitute a key element for the industry, and they are surrounded by a high level of speculation. In the past, fuel games (e.g., hedging) have been a common practice, resulting in significant losses for many airlines. Lessons were learned, leading to the recommendation to discontinue such practices. Given that airlines will continue to try to reduce their exposure to the volatility of fuel prices, new forms of fuel hedging are likely to emerge. In this area as well, research could potentially contribute to the development of more sophisticated algorithms for reducing the risks associated with fuel-hedging practices. The crisis resulting from the pandemic has also revealed the importance of maintaining strategic planning and well-considered scenarization, particularly for agents that have no direct influence on demand (e.g., airports). The crisis will teach airports that proactive contingency planning is a crucial to their resilience. Environmental concerns have also created additional challenges for the air transport industry. Despite several policies and measures that have been adopted by the sector, a large amount of knowledge is still needed with regard to the inducement of behaviors directed toward sustainability. In times of uncertainty, the domain that is likely to face the greatest challenges is that of regulation. These challenges include security and pricing, as well as issues relating to capital ownership and government aid. In all fields, there is a clear need to obtain better insight into the mechanisms that dominate markets and the relationships between agents. As revealed throughout the chapters in this book, observation of industry

How strategy can influence the market: recommendations and conclusions

447

service chain as a whole suggests that the process of liberalization is far from complete and/or fair across agents. Conflicts persist, and the uneven regulatory frameworks within which the air transport industry is embedded require a strong capacity to develop policies that will be able to compensate for a wide range of completive advantages and disadvantages.

References This concluding chapter draws on all of the chapters in this book. In addition, the following references were used. Belobaba, P.P., Swelbar, W.S., Odoni, A.R., 2016. Critical issues and prospects for the global airline industry,. In: Belobaba, P., Odoni, A., Barnhart, C. (Eds.), The Global Airline Industry, second ed. Wiley, Chichester. Lipczynski, J., Wilson, J.O.S., Goddart, J., 2009. Industrial Organisation. Competition, Strategy, Policy, third ed. Pearson, Harlow. Odoni, A.R., 2016. The international institutional and regulatory environment,. In: Belobaba, P., Odoni, A., Barnhart, C. (Eds.), The Global Airline Industry, second ed. Wiley, Chichester.

Further reading Starkie, D., 2008. Aviation Markets. Studies in Competition and Regulatory Reform. Ashgate, Farnham.

This page intentionally left blank

Index Note: ‘Page numbers followed by “f ” indicate figures, “t” indicate tables.’

A Accounting standards, 393 AerCap, 85 Agents’ behavior, 205e206 Air cargo industry air freight forwarding industry, 369e374 business model, of air freight forwarder, 364e369 freight forwarders, at major European cargo airports, 374e377 freight forwarders, in literature review, 362e364 Aircraft accidents, 413e415, 418e421 Aircraft manufacturers, 20e21, 82e83 competition, 411e413, 426e429 aircraft accidents, 413e415, 418e421 aircraft manufacturers market, 415e418 aircraft safety, and airline stock prices, 413e415 impact of accidents on, 421e425 market, 415e418 Aircraft-related emissions, 280e281 Aircraft safety, 413e415 Aireon, 350 Air freight forwarder, business model of, 364e369 Air freight forwarding industry, 369e374 Airline Deregulation Act, 153, 226e227 Airline operations, 37 Airline pricing, recent trends in, 156e161 Airlines, 19e20, 87e88 Airline stock prices, 413e415 Airline strategic behavior phases in, 318e324 Air Navigation Service Provider (ANSP), 37, 335e336

ATM/CNS profits, 355e357, 355t business model asset choices, 351e352 constructs, 354, 354t governance choices, 352 and its impact on ATM/CNS profits, 345e357, 347t strategic choices, 348e351 strategy outcomes, 352e353 commercialization and competition, 338e340 effects of, 340e345 timeline, 337f Air Operator Certificates (AOCs), 67, 186e187 Airport actors case study of Brussels Airport, 439te441t potential future developments, 443te445t Airport bills, pricing and payment of, 438f Airport handling, 86 Airport industry, regulation on, 125e126 airport regulation, 126e133 data, 137e138 empirical analysis, 133e137 empirical results, 138e145 Airport operators, 21 Airports environmental practice and carbon reduction initiatives, 278e281 Airport slot allocation, 196f, 202e210 approaches, 196e200 ASAM auction mechanism, 207e208 ASAM model architecture, 203e207 conceptual structure of model and auction mechanism, 208e210

449

450

Index

Airport Slots Auction Model (ASAM), 196, 202e203 airline agents’ bidding behavior in, 210e214 airline agents in, 212e214 auction mechanism, 207e208 model architecture, 203e207 synthetic auction market of Heathrow Airport, 215e220 Airport strategic planning, 225e226 absorbing rare, high-impact shock events, in airport strategic planning, 236e242 high-impact shock events and deep uncertainty, 233e235 year-to-year traffic volatility at airports, 226e233 Air traffic and airline financial results, 45e52 Air traffic controller (ATCO), 335e336 definitions, 336 training school, 348 Air transport accessibility daily accessibility indicator, 298e299 location indicator, 299 potential indicator, 297e298 and related concepts accessibility and connectivity, 301e303 accessibility, resilience, criticality, and vulnerability, 303e305 relative network efficiency indicator, 299e300 Air transportation security, 248e251 Air transportation vertical channel, 79e81, 79f Air transport business, 99e100 conflict situations within and between actors, 18e23 COVID-19 on airline business, 100e112 evolution toward new business models, 15e18 interdependence and market power, 12e15 market structure, 2e15 mid-term issues, 117e120

new era of nationalization, 112e117 reactions of low-cost carriers, 120e121 size order of market parties, 5e11 Air transport CO2 emissions airline operations, 37 air navigation service provider’ operations, 37 comparison of measures, 39 market-based measures, 39 new standards, 36e37 sustainable aviation fuels, 37e38 technological developments, 35e36 Air transport industry traffic, 30e34 Air transport markets, 227e231 Air transport networks, 295e296 air transport accessibility daily accessibility indicator, 298e299 location indicator, 299 potential indicator, 297e298 relative network efficiency indicator, 299e300 air transport accessibility and related concepts accessibility and connectivity, 301e303 accessibility, resilience, criticality, and vulnerability, 303e305 tentative future research agenda big data and open sources, 309e310 equity, 308e309 ICT, 307e308 intermodality, 306e307 short distances and greener modes, 305e306 Air transport security costs, 257 Air transport system, 195, 231e232 Air transport traffic and impacts worldwide, 274e277 American National Air Traffic Controllers Association, 341 Amsterdam Airport Schiphol, 178e187 Artificial intelligence (AI) models, 394e395 Artificial markets, 155 Auction market, 215e217 Average cost function, 134e136 Aviation history, fault line in, 152e153

Index

Aviation industry, employment in, 56e58 Aviation market, competition in, 66e68 Aviation security framework, 255, 258e260

B Best Alternative to a Negotiated Agreement (BATNAs), 288 Bid increment decision, 213 Bid price determination, 210e212 Black Swan-concept, 234 Boeing aircraft orders, 424te425t Branded-fare families, 156e157 Brent oil price, 384f Business model of air freight forwarder, 364e369

C Carbon footprint, 273e274 airports environmental practice and carbon reduction initiatives, 278e281 air transport traffic and impacts worldwide, 274e277 environmental sustainability practice, 281e286 good practice recommendations and opportunities, 290e292 negotiation, 286e289 Carbon Offsetting and Reduction Scheme for International Civil Aviation (CORSIA), 39e42 Cathay Airlines annual fuel hending losses, 402t Cathay Airlines fuel price, 402f Cathay Pacific Airways, 398e401 Climate change, 28e29 global policies to address, 29e30 Cobb-Douglas functional form, 136 CO2 emissions trading systems on air traffic and airline financial results, 45e52 air transport industry traffic and, 30e34 analysis of supply and demand for carbon offsets, 42e45

451

Carbon Offsetting and Reduction Scheme for International Civil Aviation (CORSIA), 39e42 climate change and, 28e29 global policies to address climate change, 29e30 from international air transport, 34e35 reduce air transport airline operations, 37 air navigation service provider’ operations, 37 comparison of measures, 39 market-based measures, 39 new standards, 36e37 sustainable aviation fuels, 37e38 technological developments, 35e36 Collaboration strategy of ANSP, 349e350 Commercialization and competition, 338e340 effects of on ANS prices, 343e344 on costs, 342e343 on customer relationships, 344e345 on government relationship, 345 on labor and capital, 345 safety and security, 341e342 on service quality, 344 Company valuation, 389e390 Concentration Ratio Four (CR4), 371e372 Confirmation Bias (CB), 391e392 Contractual costs, 392 CORSIA eligible emissions units, 42 Cost structure, 353 COVID-19, 173e174 demand growth and the airport capacity crunch, 174e176 excess demand for slots, 176e177

D Daily accessibility indicator, 298e299 Deviant behavior, 161e165 Difference-in-differences analysis (DID), 136

452

Index

E

G

EasyJet, 116 Elasticity of demand, 63e65 Elasticity of supply, 65e66 Employment in aviation industry, 56e58 Engine manufacturers, 84 Entry Point North (EPN), 350 Environmental sustainability practice, 281e286 European Cockpit Association (ECA), 67e68 European Satellite Services Provider (ESSP), 350 EU Slot Regulation, 172

GDS, 86e87 General Electric Commercial Aviation Services (GECAS), 85 Generally Accepted Accounting Principles (US GAAP), 393 Grandfather rights, deficiencies of, 185e186 GroupEAD, 350

F

Financial support to traditional flag carriers, 112e113 Fleet and capacity reductions, 108e112 Flexible master planning, 239e240 Flight Calibration Services, 350 Freight forwarders, at major European cargo airports, 374e377 Frequentis DFS Aerosense, 350 Fuel costs relevance in aviation, 383e384 Fuel hedging, 395e396 cases, 396e401 challenges, 392e395 drivers company valuation, 389e390 management competence, 391e392 mitigate volatility, 390e391 fuel costs relevance in aviation, 383e384 fundamentals, 384e387 in reality, 388e401 recent developments, 402e404 regional perspective, 395e396 Fuel prices, 446 Full freighter traffic, 182e183 Full-time equivalent (FTE), 135

H Handling companies, 21e22 Herfindahl-Hirschman Index (HHI), 370 Hurricane Pam, 235 Hybrid business model, 163

I ICAO Technical Advisory Body (TAB), 42 Identification approach, 136e137 Inelastic supply of labor, 65f Information and Communication Technologies (ICT), 307e308 Innovation strategy of the ANSP, 350e351 International air transport, 34e35 International Air Transport Association, 71 International Civil Aviation Organization, 249e250 International Transport Workers’ Federation(ITF), 67e68 Into wing fuel price, 401 Italian regulatory framework, 133

L Labor in aviation industry bargaining power competition in aviation market, 66e68 elasticity of demand, 63e65 elasticity of supply, 65e66 economic shocks, 71e72 employment in aviation industry, 56e58

Index

industrial action, 68e71 monopoly (the power of the unions), 61e62 monopsony (the power of the employers), 59e60 wage determination, 58e59 Landing andtake-off (LTO) cycle, 280e281 Leasing companies, 84e86 Location indicator, 299 Long-haul, low-cost (LHLC) operations, 238e239 Long versus short hedging, 387 Low-cost carrier Norwegian’s continued growth strategy, 327e330 Low-cost carriers (LCCs), 67e68, 115e117, 160 Low-cost market, 120

M Management competence, 391e392 Market-based measures, 190e192 Market structures, 434e438 Mitigate volatility, 390e391 Monopsonistic labor market, 65 Montreal Convention 1971, 249e250 Multiple correspondence analysis (MCA), 346e348

N National Air Traffic Services (NATS), 339 Nationalized airlines, new model of, 113e114 Nationally Determined Contributions (NDCs), 30 Neoclassical theory, 436 Norwegian Airlines, 315e316 airport competition, 324e327 EU deregulation, 316e318 low-cost carrier Norwegian’s continued growth strategy, 327e330 phases in airline strategic behavior, 318e324 Norwegian Air Shuttle (NAS), 322

O Operational scope of ANSP, 348e349

453

P Potential indicator, 297e298 Predictive modeling, 394e395 Principal component analysis (PCA), 346e348 Public-private partnership (PPP), 337

Q Quantum artificial intelligence (QAI), 394e395

R

Regional financial performance, 6t Regulatory framework, 251e257 European Framework, 255e257 ICAO, 253e255 Relative network efficiency indicator, 299e300 Revenue management, 155 Revenue structure, 353 Risk based security, 260e262 Robust airport strategic planning, 239e240 Ryanair, 115e116

S Security in airports criticisms to aviation security framework, 258e260 risk based security, 260e262 SeMS-security management system, 262e263 Security management system (SeMS), 262e263 Shock events and economic characteristics of airports, 232e233 Short take-off and landing airports (STOL), 317 Slot allocation mechanism, 197 Slot allocation scheme in EU, 199f Slot allocation solution alternatives, 201f Slot Coordinator, 206 Small and medium-sized airports, 247e248 air transportation security, 248e251 air transport security costs, 257 proportionality of security in airports

454

Index

Small and medium-sized airports (Continued) criticisms to aviation security framework, 258e260 risk based security, 260e262 SeMS-security management system, 262e263 regulatory framework, 251e257 European Framework, 255e257 ICAO, 253e255 security in network of airports, 263e269 Southwest Airlines, 115, 397e398 Strategic jet fuel hedging, 389f Super-congested airport, 189e190 Synthetic mini auction market, 215

T Tentative future research agenda big data and open sources, 309e310 equity, 308e309

ICT, 307e308 intermodality, 306e307 short distances and greener modes, 305e306 Transfer traffic, 181

U United Nations Framework Convention on Climate Change (UNFCCC), 29 US domestic market, 227

W Wage determination, 58e59 in competitive market, 58f with inelastic demand, 64f Wage rates with monopsonistic demand, 60f with union influence, 62f West Texas Intermediate (WTI), 384f Wizz Air, 116e117