Entry and exit in the post-liberalized German airline market

This paper examines actual entry and exit dynamics in the German airline market in light of the European liberalization process. For this purpose, flight schedules data was used to derive entry and exit statistics for a set of chosen inner-German routes over a period of thirteen years. Data was analysed in cross-sectional and longitudinal manner while identifying different entry waves and examining incumbent behaviour with respect to entry deterrence strategies. Furthermore, the market’s concentration and structure was tracked to allow a conclusion with respect to the market’s contestability after liberalization. The results suggest the existence of remaining structural barriers to entry and depict a rather unsatisfactory overall state of the post-liberalized market.

3

TABLE OF CONTENTS

  List of Illustrations  5

  Abbreviations  6

  Acknowledgements  7

1  INTRODUCTION  8

2  THE LIBERALIZATION PROCESS OF EUROPE’S SKIES  12

2.1  Pre-Liberalization Status Quo  13

2.2  Liberalization of Intra-European Air Transport  14

2.2.1  The three Liberalization Packages  15

2.2.2  Cabotage Rights  18

2.2.3  The Single European Sky Initiative 2004  19

2.3  Liberalization of International Air Transport  20

2.4  Liberalization Review  22

3  THEORY  23

3.1  What do we know about entry?  23

3.2  Entry deterrence and dynamic competition  25

3.2.1  A model of limit capacity  26

3.2.2  Excess capacity as prospect for collusion  27

3.3  A four-stage game of airline market entry  28

3.4  Airline capacity utilization  30

4  THE GERMAN AIRLINE MARKET  33

4.1  Infrastructure  33

4.1.1  Airports  34

4.1.2  Slots, slot pools and slot allocation  38

4.2  Demand for air travel  40

4.3  Supply of air travel  44

4.4  Ground transport  47

4.4.1  Railway system  47

4.4.2  Road network  48

4.5  Modes of transport compared  49

4.6  The short-haul problem  51

4

5  ANALYSIS OF ENTRY AND EXIT  52

5.1  Purpose  53

5.2  Data  53

5.2.1  Data construction  54

5.2.2  Strengths and weaknesses of the data set  56

5.3  Methodology  56

5.4  Entry, exit and welfare effects  58

5.5  Entry and exit over time  59

5.5.1  The 1996 entry wave  63

5.5.2  The 2004 entry wave  64

5.5.3  Airlines 1993 2006  66

5.6  Entry, exit and market concentration  67

5.7  Entry, exit and incumbent behaviour  74

5.8  Entry, exit and prices  80

5.9  The German Airline Market in 2006  83

5.9.1  Route network  85

5.9.2  Market shares  86

5.9.3  Brand competition  88

5.10  Summary of Results  90

6  IMPEDIMENTS TO COMPETITION  93

6.1  Barriers to entry  93

6.1.1  Strategic barriers to entry  94

6.1.2  Structural barriers to entry  94

6.2  Alliances an economic necessity?  96

7  SUMMARY  99

Appendix 101   

  A Glossary  101

  B Freedoms of the Air  103

  C Germany’s International and Regional Airports  104

  D IATA Airport and Airline Codes  106

  E Additional Tables and Statistics  107

  References  116

5

  LIST OF ILLUSTRATIONS  

  FIGURES  

  Figure 2 1: Multi-level air transport  

  Figure 3 1: Passenger load factor on inner-German routes  

  Figure 4 1: German airport network  

  Figure 4 2: German airports according to PAX  

  Figure 4 3: Demand for air travel and GDP  

  Figure 4 4: World passenger development 1996 2005  

  Figure 4 5: Air traffic growth until 2015.  

  Figure 4 6: Development of LCC route network.  

  Figure 4 7: Capacity growth of Low Cost Carriers for Europe and Germany 2001 2006  

  Figure 5 1: Operated routes with no entry 1993 2006  

  Figure 5 2: Aggregated entry, exit and net entry rates ( ) 1993 2006  

  Figure 5 3: Airlines 1993 2006  

  Figure 5 4: Number of routes (total: 25) with given number of operating airlines 1993 2006  

  Figure 5 5: Number of LON routes (total: 8) with given number of airlines 1993 2006  

  Figure 5 6: Distribution of non-monopolistic routes 1993 2006  

  Figure 5 7: Entry, exit and output  

  Figure 5 8: Development of frequency on route BERDUS  

  Figure 5 9: Price indices Germany 2000 2006 (2000 100)  

  Figure 5 10: Operating carriers on inner-German routes 2006  

  Figure 5 11: Airline route networks 2006  

  Figure 5 12: Airline ownership and control 2006  

  TABLES  

Table 2 1:  Overview of deregulatory measures  

Table 4 1:  Comparison of transport modes for short-haul inner-German routes  

Table 5 1:  Entry and exit activity 1993 2006  

Table 5 2:  Route entry and exit in 1996  

Table 5 3:  Route entry and exit 2004 2005  

Table 5 4:  Average number of airlines per operated route  

Table 5 5:  Herfindahl index and numbers-equivalent of firms  

Table 5 6:  Number of different airlines on inner-German and LON routes 1993 2006  

Table 5 7:  Effects of firm entry and exit on capacity and frequency  

Table 5 8:  Correlation of Lufthansa output with entry and exit  

Table 5 9:  Airline route networks 2006  

Table 5 10:  Market shares and concentration ratios 1993 2006  

Table 5 11:  4U versus X3 competition  

Art. – Article (law) ASA – air service agreement ATC – air traffic control ATM – air traffic management BKartA – Bundeskartellamt / Federal cartel office

BMVBS – Bundesministerium für Verkehr, Bau und Stadtentwicklung / Federal Ministry of

Transport, Building and Urban Affairs CRS – computer reservation system

DFS – Deutsche Flugsicherung GmbH / German air traffic controller DLR – Deutsches Zentrum für Luft- und Raumfahrt / German Aerospace Center cp. – compare

EC – Regulation of the European Commission [law] EC DG – European Commission Directorate-General ECJ – Court of Justice of the European Communities EEC – Council Regulation (of the Council of the European Union) [law] FFP – frequent flyer program H – Herfindahl index

IATA – International Air Transport Association ICAO – International Civil Aviation Organisation ICE – Inter City Express (German high speed train) LCC – Low Cost Carrier

LuftVG – Luftverkehrsgesetz / Air Traffic Act of Germany LuftVZO – Luftverkehrs-Zulassungs-Ordnung / Air Traffic Authorisation Ordinance NEF – numbers-equivalent of firm (inverse of Herfindahl index) OAG – OAG Worldwide Ltd. (global travel and transport information company) p.a. – per annum para. – paragraph [law] PAX – passengers resp. – respectively RPK – revenue passenger-kilometres TACO – travel agent commission override program

ACKNOWLEDGEMENTS

I would like to thank Prof. Benjamin Bental, Ph.D., for giving me the opportunity to conduct my thesis with him and for providing me with freedom in regard to the choice of topic. Special thanks go to the German Aerospace Center, Insitute of Air Transport and Airport Reseach, in particular to Prof. Dr. Hansjochen Ehmer and Dr. Peter Berster who gave me access to their data base without which this thesis would not have been possible. I am especially grateful to Dr. Peter Berster who supported my first steps in the field of airline economics with a great patience for the questions arising. Last but not least, I am gratefully acknowledging the scholarship from the Stiftung der Deutschen Wirtschaft which gave me the opportunity to conduct my studies in a qualitatively high manner, including semesters in Moscow and Paris and additional education by means of seminars and workshops.

1 Introduction

“The airline industry is an enigma.”

(Rigas Doganis, 2006) 1

The airline industry is an industry of contradictions: international by its nature yet bound to national, bilateral or at best to multilateral economic regulations and legal restrictions. It is an industry of dynamic growth and change, greatly influenced by new technologies and innovations. ² It is highly cyclical while strongly correlating with GDP and frequently put into turbulences via internal and external shocks. ³ Despite its dynamic growth, the airline industry turns out to be of only marginal profitability. ⁴ Trends towards “deregulation” (U.S.) or “liberalization” (EU) have fostered the dynamics of change that did not only produce the suspected outcomes of competition and increased social welfare but that shook airlines, markets and resulted in periods of high losses with airlines filing for bankruptcy, the emergence of stronger networks, new business models and a new structure of the industry as a whole. The Federal Aviation Act of 1958 of the United States of America, still the basic governing statute, declared a policy to adapt the air transportation system “to the present and future needs of the domestic and foreign commerce of the United States, the Postal Service, and the national defense.” ⁵ These three aspects of public interest in air transport did and still do motivate both domestic and international regulation on the aspects of market access, designation, flight frequency and capacity as well as fares. In 1978, however, the Airline

¹ Rigas Doganis is non-executive director of South African Airways and former CEO of Olympic Airways, Greece. He is lecturing in air transport and author of books on the international airline business. See Doganis (2006), p. xii.

² Growth is measured in multiple ways. Literature usually refers to (revenue) tonne-kilometres resp. (revenue) passenger-kilometres (see Glossary). Growth may also be approximated via the total number of passengers (PAX) of an airline or on (a) route(s) in a period of time, usually referring to either weekly or annual data. Growth in the airline industry is said to be “dynamic” because it is forecasted to be strong -ICAO predicts a growth rate of 5.8% (5.6%) in scheduled airline passenger traffic for 2007(2008).

³ Worldwide air traffic (measured in passenger-kilometres) is said to have an income elasticity of 2. Thus, changes in GDP affect the airline industry strongly although there might be a time lag. See Doganis (2006), p. 17. External shocks such as the terrorist attacks of 2001 (U.S.) and 2006 (UK) have a strong dampening affect on traffic although the overall growth rate recovers and returns to the original trend. See Airbus (2002), p. 12 and ICAO (2005), p. 6.

⁴ Net profit of all ICAO airlines was 1% of revenue in 2005; see ICAO (2005), p. 6. Even in the profitable years of 1997-99, net profit was only between 2-3% and negative net profit usually exceeds (in absolute terms) positive net profits. See Doganis (2006), pp. 4-8.

⁵ In O’Connor (1995), p.13.

INTRODUCTION 9

Deregulation Act became law and with it entry and exit controls were abolished (with some exceptions) and regulation of air fares was virtually eliminated. ⁶ As a result the shape of the U.S. domestic market changed dramatically: many new small firms evolved with the number of operating carriers nearly doubling whereas existing airlines expanded into new routes. ⁷ Route density increased as did the rate of competition while air fares dropped substantially and with it airline profitability. Incumbent carriers began to set up hubs and controlled feeder flights on spoke routes by acquiring commuter airlines, which resulted in the reduction of direct flights and an increase in the average fare on hub routes. ⁸ Moreover, the mid-1980s brought “a series of mergers and a general trend toward consolidation into fewer but larger airlines” whereas the 1990s forced many airlines into bankruptcy. ⁹ Only Southwest Airlines, the first carrier using the new no frills, low cost business model, prospered. Legacy carriers, in response, continued to strengthen their networks and started to engage heavily in alliances, the latter remain until today a controversy on whether being a way to undermine competition or an economic necessity. All the same, the upshot of U.S. deregulation remains positive, leaving the industry with a better cost structure and a higher operating efficiency. Even though many new entrants and some former large carriers disappeared, the remaining airlines that managed to survive do now show a strong capability to compete. 10

As the incidents from the U.S. domestic airline market suggest, “deregulation” or “liberalization” is not a process that can be described by a simple equation where reducing regulatory barriers equals market entry equals more competition equals lower fares and higher welfare. Instead, regulatory barriers to entry tend to be replaced by competitive barriers to entry, consolidation in the competitive market might lead to a market concentration even higher than under regulation and state subsidies might be the last possibility to save large carriers that are of

⁶ Regulation continues with respect to “minimal essential services at small points”. Moreover, the U.S. domestic market remains closed for nonnational (“substantially owned and effectively controlled by their own nationals”) carriers. In fact, the U.S. domestic market is with respect to rights granted for international air traffic still very restrictive. Cp. O’Connor (1995).

⁷ Cp. Gillen (2001), p. 42.

⁸ Morrison and Winston (1995) give an analysis on the negative welfare effects of the hub-and-spoke system due to fewer (direct) routes and a “hub-premium”, however the hub system also increases the frequency of flights on spoke routes.

⁹ O’Connor (1995), p. vii. Apart from Southwest Airlines all new interstate entrants of the 1978 to the early 1980s disappeared again in the late

1980s via acquisitions (cp. Morrison and Winston [1995]). Among the airlines that went into bankruptcy was also the trunk carrier Pan American, the former monopolist on U.S. international routes.

¹⁰ Morrison and Winston (1995) find that despite the various strategies of airlines to limit competition passengers reaped gains from deregulation that remain considerable. Campbell (1999) also argues that deregulation led to the elimination of weaker competitors with a more efficient industry by means of lower costs and an improved yield management.

INTRODUCTION 10

national importance not withstanding the somewhat symbolic role of national flag carriers. In the end, as the U.S. experience shows, deregulation posed to airlines the question of how to survive and of whether and how to regulate deregulation to officials. Nevertheless, with the largely positive effect in mind, Europe liberalized its airline industry to transfer the single European market also to its skies. 11

The European liberalization process went even beyond U.S. deregulation boundaries and results for the European economic area seem to be positive: the number of scheduled airlines in the European Union increased steadily as did the number of intra-community routes. ¹² However, many entrants exited shortly after entry, consolidation is strongly evidenced, stronger networks developed and alliances undermine competition. ¹³ Contrary to expectations, the average price level of Europe’s air fares did not drop; in fact it even increased despite the existence of more promotional tariffs. ¹⁴ The result is surprising since new entrants, mostly low cost carriers, seem to offer endless possibilities for passengers to fly to any destination in Europe for a fare even below “cab prices”. Current literature capitalizes on this “low cost revolution”. Yet, when focusing on low cost carriers, which frequent secondary airports rather than primary airports, large routes like those connecting hubs remain mostly out of sight. Moreover, to the best of my knowledge, there is little information on whether and how a former closed national market has changed within its boundaries. ¹⁵ Yet, to evaluate the liberalization’s aftermath it is these routes that need to be looked upon since remaining oligopolistic or even monopolistic national (hub) routes that unite a large share of passenger traffic would challenge the achievements. Accordingly, for the analysis of entry and exit in the post-liberalized German airline market this paper singles out the eight largest airports of Germany and analyses all routes between these airports for a period of 13 years from 1993, when the Third Liberalization

¹¹ Art. 2 EC Treaty [Treaty establishing the European Community]: „The Community shall have as its task, by establishing a common market and an economic and monetary union…”.

¹² Cp. EC DG Energy and Transport (2007/02) Website and OAG (2006).

¹³ O’Connor (1995), pp. 32-34 and Campbell (1999), p.73 both pronounce the impediments to competition via hub-and-spoke systems, merger and acquisitions as well as alliances. Doganis (2006) even devotes a whole chapter to the topic of alliances. Yet, the American approach to alliances seems to be more “laissez-faire” than the European, with Americans rather viewing alliances as an economic necessity instead of an impediment to competition. (cp. Campbell [1999]). Supposedly, whether to sentence an alliance or not is a case decision: if the existence of an alliance on a certain route deters entry or unites former competitors, than it is certainly true that competition is hampered.

¹⁴ Cp. Jung (1999), p.31. and EC DG Transport and Energy (2006), p.117.

¹⁵ Gillen (2001) analyses the effects of liberalized international bilaterals for the Northern German region resp. Hamburg airport. The most recent study on entry and exit after liberalization in the UK market focuses on routes between UK and EU airports. See Gil-Moltó and Piga (2006).

INTRODUCTION 11

Package was enacted, until 2006. The objective is to observe whether the legislative liberalization of the European sky had an effect on the entry and exit dynamics on large national routes as significant sub-markets. Beforehand, the European liberalization process will be summarized in Chapter 2. Beginning with an introduction to the different levels of air transport, a description of the non-liberalized European air transport market follows. The Chapter then continues by introducing both the legal and economical motivations for the liberalization process to begin and summarizes the important legislative measures that lead to the shape of the industry as existent today. Chapter 3 then deals with the role of entry and exit as a proxy for the level of competition and the contestability of a market from the theoretical point of view. Thereby a brief summary on entry research as well as a set of models that deal with excess capacity as entry deterrence will be introduced. However, as the airline industry is somewhat special neither model perfectly fits and a simple proposal for a model of airline entry is made. Subsequently, in order to gain further insight into the industry, the German airline market is defined and characterized according to its infrastructure for both air and ground in Chapter 4. Further attention is given to the characteristics of demand and supply, thereby outlining special challenges faced by airlines operating within the German regional market. With the legislative, theoretic and economic background in mind, the actual analysis of entry and exit is presented in Chapter 5. Thereby, entry and exit statistics are presented over time and routes. Moreover, market concentration and incumbent behaviour is analysed. Also, the effects on consumer welfare and prices are commented upon in short. Furthermore, an overview on the structure of the German airline market in 2006 gives further insight by covering the incumbents, their route network, market shares and interdependences. As both entry dynamics and entry deterring strategies turn out to be rather weak or insignificant, Chapter 6 devotes further attention to airline-market specific barriers to entry of structural kind. Also, the role of alliances as either impediment to competition or economic necessity will be looked upon shortly. Finally, the paper is summarized in Chapter 7.

2 The Liberalization Process of Europe’s Skies

“Nothing like the system of government imposed impediments to economic decision-making

As mentioned above, air transport is despite its purely international sphere bound to rather national legislation. Subsequently, aviation is confronted with several levels of either national or supranational law. Likewise, air transport can be classified into either domestic or international traffic not withstanding the economic differentiation of traffic according to length of haul. ¹⁷ However, with the foundation of the European Community that later became the European Union, a third level took shape: the intra-European traffic. Figure 2-1 presents the different organizational levels of air transport schematically.

¹⁶ Jeff Shane is Under Secretary for Policy of the Department of Transportation (U.S.A.) in the Bush Administration since March 2003. [In Doganis (2006), p. 27.]

¹⁷ Length of haul is the distance of freight resp. passengers carried, which can be of short, medium or long type. See Glossary.

THE LIBERALIZATION PROCESS OF EUROPE’S SKIES 13

Generally speaking, domestic air transport (a) is all intra-national air traffic between a country’s airports no matter the type of airport. International air transport (c) is all air traffic between two countries, usually via primary airports only. Intra-European air transport (b) is all air traffic between the airports of the member states of the European Union while traffic frequents both primary and secondary airports directly. 18

2.1 Pre-Liberalization Status Quo

Without the European liberalization, all domestic air traffic fell only under national laws and control, regulated in the case of Germany via the basic governing statute, the Basic Law of the Federal Republic of Germany, in combination with the Air Traffic Act (“Luftverkehrsgesetz” [LuftVG]). Competition within this domestic market is, next to being subject to federal policy, controlled by the Bundeskartellamt [BKartA], the German federal cartel office. All international transport to and from Germany is regulated via air service agreements [ASAs]. These air service agreements are usually bilateral and each uniquely agreed upon between the two countries of origin-destination relationship. As of December 2006, Germany held ASAs with 138 different countries. ¹⁹ Intra-European traffic, without the EU liberalization, consequently is also regulated by bilateral agreements between each of the member states of the European Union, while each agreement is unique of its type and with respect to the rights granted. 20

It is worth mentioning that bilateral air service agreements are trade agreements between governments with typically both parties to the agreement trying to strengthen their national flag carriers on the route. Moreover, market access (what airports a foreign airline may serve) and

¹⁸ Airports are classified in multiple ways, which can be technical, economical or organizational-based; thus, there is no unique classification for civil airports. Literature most commonly refers to primary and secondary airports. This paper follows the terminology. One may think of primary airports as large airports that handle both domestic and short- to long-haul international traffic, while secondary airports handle domestic to medium-haul international traffic only. Consequently, primary airports are also of larger size in terms of passengers handled. Section 4.1.1 will treat the distinction of airports in more detail.

¹⁹ This includes countries that are now members of the European Union with whom Germany however had agreed upon air service agreements already before. Source: Luftrecht-Online (2007/02) Website.

²⁰ See Appendix B – Freedoms of the Air. Only transit rights (First and Second Freedom) are granted on a multilateral basis via the “Transit Agreement” of 1944. The granting of Third and Fourth Freedom Rights (the rights of an airline of one country to carry traffic to and from another country) are subject to bilateral negotiation so are the Fifth to Seventh Freedoms but as special case since a third country is involved. The Eight and Ninth Freedoms (cabotage rights) have not been granted yet within bilateral ASAs, however, as we will see, the European liberalization went further.

THE LIBERALIZATION PROCESS OF EUROPE’S SKIES 14

designation (how many airlines may serve a route) as well as even frequencies (how often a foreign airline may serve a certain route) and capacities (with what seat capacity this route may be served by the foreign airline) are agreed upon. ²¹ Furthermore, most bilateral ASAs support the respective airlines to agree on air fares, with airlines also being “encouraged to use the tariff-fixing machinery of the International Air Transport Association (IATA) […]. However, both governments must approve such fares and tariffs. This is the so-called ‘double approval’ regime.” ²² Obviously a system of bilateral ASAs within the European Union with their regulatory character to predetermine market access, designation, frequency, capacity and fares would not allow for “an internal market characterised by the abolition, as between Member States, of obstacles to the free movement of goods, persons, services and capital”. 23

2.2 Liberalization of Intra-European Air Transport

The liberalization of the European market for air transport is a process of more than 20 years of continuous proposals, guidelines and ordinances. ²⁴ Both the European Commission’s Directorate-General for Competition and the Directorate-General for Energy and Transport followed two complementary lines of approach for a liberalized European air transport system: on the one hand, to ensure competition that will not be distorted via uncompetitive practices imposed by individual governments or the industry itself in order to be in line with the goals set in the “Treaty establishing the European Community” (EC DG Competition); on the other hand, to foster an airline industry that will be able to capitalise on the effects of liberalization through improvements in the industry’s cost structure, gains in efficiency and a stronger competitiveness (EC DG Energy and Transport). In particular the DG Transport and Energy’s initiative was not only incited by internal but also by external factors: the deregulation of the U.S. domestic market, despite and because of its turbulences, lead to highly competitive U.S. carriers. In addition, U.S. public policy began to promote a further liberalization of international air service agreements in 1978 that aimed not only at allowing direct competition on Trans-Atlantic routes

²¹ Cp. Doganis (2006), pp. 27-31.

²² Doganis (2006), p. 28.

²³ Art. 3 para. 1.c EC Treaty [Treaty establishing the European Community]

²⁴ Cp. Piepelow (1997) pp. 94-110; Gillen (2001), pp. 29-40; Jung (1999), pp. 13-32 and Doganis (2006), pp. 31-66.

THE LIBERALIZATION PROCESS OF EUROPE’S SKIES 15

but also tried to achieve better market access for U.S. carriers into the European market. ²⁵ Nonetheless, following the “Treaty establishing the European Community”, which was signed on March 25, 1957, the liberalization process of the European airline market was no phenomenon that swapped over the Atlantic Ocean but that, not withstanding influences through the experiences of U.S. deregulation, has its own raison d’être. The initiative to liberalize the European airline industry commenced as early as in the 1970s with the European Commission trying to develop a common air transport policy that was submitted to the Council of the European Union. Yet, success failed to appear although the Court of Justice of the European Communities [ECJ] has ruled in 1974 that all principal articles of the EC Treaty do also apply to the transport industry. ²⁶ Furthermore, although the legal postulates were set ever since 1974 and despite the Council of the European Union’s action plan in 1978, two memoranda by the European Commission in 1979 and 1984 as well as another sentence of the ECJ in 1986, which (among other things) laid down that competition law of the EC Treaty fully applies to air transport, liberalization did not gather way. It was not until 1988, when the first out of three Liberalization Packages became effective, that the liberalization of the European airline market did accelerate and began to transform the regulatory framework to what it has become today.

2.2.1 The three Liberalization Packages

The First Liberalization Package, which was enacted on January 1, 1988, reinforced the ECJ ruling of 1986 by explicitly acknowledging that all competition rules of the EC Treaty do apply to air transport but most importantly, it enacted a package of measures concentrating on the core aspects of regulation: market access, designation control, capacity approval and fare agreements. As a consequence, the fares regime became more liberal by adjusting fares to longterm costs, yet still under the double approval regime. Market access was facilitated by allowing airlines to open new routes between their home country’s primary airports and the foreign country’s secondary airports. Moreover, the capacity rule, which assigned capacities to the

²⁵ U.S. “open skies” is somewhat asymmetric: it does not grant foreign airlines full access to the U.S. market by limiting market access to few gateway airports only, while U.S. carriers may fly to Europe from any U.S. airport. Cp. Campbell (1999), p. 96. Section 2.3 will treat the international air traffic liberalization, characterized by U.S. “open-market” and “open-skies” agreements, in more detail.

²⁶ Art. 2 EC Treaty, Art. 3 EC Treaty [common market] and Art. 81-89 EC Treaty [rules on competition and state aid].

THE LIBERALIZATION PROCESS OF EUROPE’S SKIES 16

designated airlines typically by an equal share of 50:50, was relaxed to 60:40. Also, the single designation regime, which authorized one airline per country only to operate a route, was mitigated by permitting multiple designations on routes if the annual passenger volume exceeded a predefined amount. Last but not least, the Fifth Freedom of the Air, the right granted by a state to another state to land and to take on passenger traffic to or from a third state in the territory of the foreign state, was enhanced but remained bound to capacity restrictions and airport access.

The Second Liberalization Package was ratified on November 1, 1990 according to plan. Its main issues were a further liberalization of the tariff-fixing regime that established zones around fares, still adjusted to long-term costs of an airline, within which fares were automatically approved. Fares below this zone fell into the double approval, fares above into the new double disapproval regime. ²⁷ Airlines were also allowed to adjust their fares to approved fares of competing airlines provided that service with respect to quality was equipollent. Moreover, the capacity rule was further softened by allowing designated carriers to raise capacity over time (max. 7.5 percentage points per flight schedule period) as long as not exceeding 60% of total route capacity. Routes connecting secondary airports only were fully freed from this regulated capacity allocation. In addition, route passenger volumes to allow for multiple designations were lowered. With respect to the Freedoms of the Air, routes connecting primary airports of one country with secondary airports of another were granted full Third and Fourth Freedom Rights without any further exemptions. Likewise, exemptions to the Fifth Freedom Right with respect to the airports served were basically abolished while capacity constraints were reduced. Regulation remained, similar to U.S. deregulation, for small routes between secondary airports. The Second Liberalization Package installed measures that both protected small regional carriers from competition by larger carriers as well as promoted airlines to operate routes that would otherwise not be operated as a result of an airline’s economic considerations. 28

The Third Liberalization Package, enacted on January 1, 1993, was the last Liberalization Package of the series and incorporated an extensive set of prescriptions. Among them were the

²⁷ Both countries to a route need to disapprove a fare for it to be rejected (double disapproval) vs. approving it (double approval).

²⁸ Regional airlines that operated routes with small propeller aircraft were protected from competition by airlines with jet airplanes. Routes that would not be operated as a result of rational-economical decision-making could be promoted by allowing an airline to monopolise this route for a certain time period.

THE LIBERALIZATION PROCESS OF EUROPE’S SKIES 17

directions to grant open market access on routes, to fully liberalize the fare regime from state exertion of influence, regulations for slot allocation, by-laws that allow for an efficient use of the European Cartel Law as well as detailed regularisations considering computer reservation systems [CRS]. ²⁹ Specifically, both the restrictions on multiple designations as well as on capacity were fully eliminated. ³⁰ The tariff-fixing machinery was completely abandoned and fares were left subject to the market mechanism only. ³¹ Furthermore, competition rules of the EC Treaty were set to be fully applicable to both air transport between the member states and within a member state. As a result, the control of competitive practices further falls into the scope of national cartel authorities as well as into the Commission’s. ³² Moreover, native discrimination concerning the establishment and operation of airlines within the European Union but outside their home country was eliminated by relaxing the ownership role. Before, every member state applied different rules for the granting and maintenance of operating licences, which certainly implied advantages for own nationals. Also, the so-called national ownership rule allowed only airlines that were “substantially owned and effectively controlled by their own nationals” to operate certain routes. ³³ The Third Liberalization Package established common criteria European airlines have to comply with in order to be granted with an operating license by a member state of the European Union, and the community ownership rule was introduced, which widened the “substantial ownership” and “effective control” to any member state of the EU, not necessarily the state of registration. ³⁴ Yet, for the time being, this last large Liberalization Package remained imperfectly on cabotage rights, which stayed limited to consecutive cabotage (Eight Freedom of the Air) until April 1, 1997. 35

²⁹ CRS have been used by the owning airline to establish a competitive advantage by biasing query results. See 6.1.1 Strategic barriers to entry.

³⁰ Council Regulation (EEC) No 2408/92 of 23 July 1992 on access for Community air carriers to intra-Community air routes.

³¹ Council Regulation (EEC) No 2409/92 of 23 July 1992 on fares and rates for air services. However, there are some limited safeguards to prevent predatory or excessive pricing. Cp. Doganis (2006), p 47.

³² Council Regulation (EEC) No 2410/92 of 23 July 1992 amending Regulation (EEC) No 3975/87 laying down the procedure for the application of the rules on competition to undertakings in the air transport sector.

³³ This still applies to most international routes, although international routes to and from the EU are renewed to adopt the community ownership rule. See 2.3 Liberalization of International Air Transport and cp. Doganis (2006), pp. 54-59.

³⁴ Council Regulation (EEC) No 2407/92 of 23 July 1992 on licensing of air carriers.

³⁵ Art. 3 para. 2 Council Regulation (EEC) No 2408/92 of 23 July 1992 on access for Community air carriers to intra-Community air routes.: “Community air carriers shall be permitted by the Member State(s) concerned to exercise traffic rights on routes within the Community.” However, this regulation is offset until April 1, 1997 unless a) “the traffic rights are exercised on a service which constitutes and is scheduled as an extension of a service from, or as a preliminary of a service to, the State or registration of the carrier” and b) “the air carrier does not use, for the cabotage service, more than 50 % of its seasonal capacity on the same service of which the cabotage service constitutes the extension or the preliminary.” Thus, cabotage right is limited to a certain capacity and route choice.

THE LIBERALIZATION PROCESS OF EUROPE’S SKIES 18

2.2.2 Cabotage Rights

Cabotage is the right of an airline of foreign nationality to transport domestic traffic of the granting country: a) entirely within the territory of that state with consecutive airports in the airline’s home country or a foreign country (consecutive cabotage, Eight Freedom of the Air) or

b) entirely within the territory of that state without consecutive airports in the airline’s home country or a foreign country (unlimited cabotage, Ninth Freedom of the Air). With regard to the scheme displaying the levels of air transport introduced in the beginning of this Chapter, cabotage refers to all traffic on routes (a) for carriers of non-German nationality and applies for all carriers of European Union state of registration.

Article 3 paragraph 1 of Council Regulation (EEC) No 2408/92 of 23 July 1992 on the access for community air carriers to intra-community air routes states that “Community air carriers shall be permitted by the Member State(s) concerned to exercise traffic rights on routes within the Community”. However, the second paragraph of the same regulation states: “Notwithstanding paragraph 1, before 1 April 1997 a Member State shall not be required to authorize cabotage traffic rights within its territory by Community air carriers licensed by another Member State”. Thus, the last stronghold, namely that of intra-state traffic, fell on April 1, 1997 to be fully opened for competition by allowing any European carrier to enter and operate these respective routes.

Hence, for the region of the European Union, air transport and thus the airline industry as such are since then organized within a single European market with no further distinction of intra-national or cross-border traffic between the member states. Henceforth, domestic air transport denotes all traffic between any airport of primary or secondary type of any member of the European Union (denoted in Figure 2-1: Multi-level air transport by lines a and b as well as by the dashed line b’) while international transport marks any traffic from any European airport to a third country’s airport (represented by line c for the case of Germany). Summarizing the legislative liberalization measures from 1988 to 1997 with respect to intra-European transport, the following Table 2-1 gives an overview of the most significant deregulatory changes shaping the industry with regard to entry and competition. Thereby, the first column lists all pre-liberalization conditions and contrasts them with the measures taken in each of the three Liberalization Packages. The comparison is differentiated according to the four core aspects that regulation focused upon: market access, designation, capacity and fares.

THE LIBERALIZATION PROCESS OF EUROPE’S SKIES 19

However, it is noteworthy that European liberalization did neither conclude in 1997 nor stay restrained on topics regarding airlines directly only. Liberalization went further to areas such as slot allocation at airports (EEC 95/93 and EC 793/2004), ground handling services (EEC 96/97) and computer reservation systems (EEC 3089/93 and EEC 323/99), by passing directives or codes of conduct. Furthermore, liberalization also took on progress with respect to air traffic management, which cumulated in the 2004 single European sky initiative. Although this initiative does not directly affect entry and competition between airlines it is yet a step that will be shortly introduced since it is supported by airlines on the one hand and demonstrates the profoundness of the European Union skies’ restructuring very descriptively on the other hand.

2.2.3 The Single European Sky Initiative 2004

The European skies’ traffic is secured by means of 47 different air traffic control organisations, which are basically nationality bounded. ³⁶ This system proves to be only

³⁶ Data for 2004. Cp. Fischer (2004), p. 23.

THE LIBERALIZATION PROCESS OF EUROPE’S SKIES 20

somewhat efficient, especially when comparing it to the single ATC organization operating the U.S. air traffic on the basis of ATC costs per flight or flights handled per air controller. Moreover, “[i]mplementation of the common transport policy requires an efficient air transport system allowing safe and regular operation of air transport services, thus facilitating the free movement of goods, persons and services.” 37

Consequently, the European Commission plans to reduce the fragmentation of the European sky with respect to the organisation and use of air space as well as air navigation services to “create a more integrated European air space: the single European sky”. ³⁸ As Regulation (EC) No 551/2004 states, an “[e]fficient airspace management is fundamental to increasing the capacity of the air traffic services system, to providing the optimum response to various user requirements and to achieving the most flexible use of airspace.” The key adjustments are of structural (consolidation and integration), operational (interoperability) and technological (new technology to be commonly adopted) kind. However, air traffic services remain a continuous monopoly where benchmarking and best practice shall improve the performance within the non-competitive environment. ³⁹ Nevertheless, the single European sky initiative is a further step towards a more competitive European airline industry as a whole.

2.3 Liberalization of International Air Transport

International air transport is mostly regulated, until today, by means of air service agreements between each state of origin-destination relationship. These ASAs are typical trade agreements, as described earlier. ⁴⁰ However, there were two trends that commenced to transform these agreements. First, U.S. public policy started already in 1978 to push for an advanced

³⁷ Regulation (EC) No 549/2004 of the European Parliament and of the Council of 10 March 2004 laying down the framework for the creation of the single European sky.

³⁸ Regulation (EC) No 550/2004 of the European Parliament and of the Council of 10 March 2004 on the provision of air navigation services in the single European sky (the service provision Regulation).

Regulation (EC) No 551/2004 of the European Parliament and of the Council of 10 March 2004 on the organisation and use of the airspace in the single European sky (the airspace Regulation).

Regulation (EC) No 552/2004 of the European Parliament and of the Council of 10 March 2004 on the interoperability of the European air traffic management network (the interoperability Regulation).

³⁹ Cp. EC DG Energy and Transport (2004).

⁴⁰ See 2.1 Pre-Liberalization Status Quo.

THE LIBERALIZATION PROCESS OF EUROPE’S SKIES 21

liberalization of international air service agreements, which occurred in two phases. ⁴¹ This first generation of “open market” agreements (1978-91) broadened market access by allowing a wider range of airports to be frequented on international routes, permitted multiple designation, abandoned capacity and frequency controls, withdrew the anti-trust immunity for IATA air fare conferences and introduced the double-disapproval regime. The successive second-generation “open skies” agreements, which evolved since 1992, went further and granted open market access by allowing an unlimited number of airports to serve as point of departure/arrival for international air transport. Furthermore, unlimited Fifth Freedom Right was granted and fares were fully deregulated by allowing free pricing. However, U.S. “open skies” agreements do not extend to domestic cabotage and apply the national ownership rule. Thus, airlines permitted to operate a route between the U.S. and the designating state need to be “substantially owned” and “effectively controlled” by nationals of the respective state. ⁴² Second, as a result of the European liberalization, European carriers were granted unlimited traffic rights while ownership and nationality constraints were removed. However, these rights apply only for inner-European traffic. Most ASAs, just like the U.S. “open skies” agreements, underlie the national ownership rule and thus distort the common European airline market by discriminating European carriers on grounds of their nationality. ⁴³ To answer this problem, the European Commission received, after a case file in front of the Court of Justice of the European Communities [ECJ] and the respective ruling of the ECJ in November 2002, two mandates: one to negotiate directly with the United States and the other to negotiate so-called “horizontal agreements” with other third countries in order to correct the legal deficiencies within all ASAs between the members of the EU and any other third country. This will allow European carriers to take off to a third country from any European airport. ⁴⁴ Consequently, liberalizing international bilateral air service agreements is another important step for a fully integrated common European airline industry with no discrimination within domestic and no further discrimination on international air traffic.

⁴¹ Cp. Doganis (2006), pp. 32-45.

⁴² Moreover, U.S. “open skies” agreements are somewhat asymmetrical to the advantage of U.S. carriers. See Footnote 25.

⁴³ Doganis (2006), p.53. For example, only German airlines may operate the Trans-Atlantic routes between a German and an U.S. airport while U.S. carriers may enter the European market by transporting traffic from Germany to any other member state of the EU. All other bilateral ASAs also discriminate European carriers on the ground of their nationality.

⁴⁴ It was the November 2002 ECJ ruling that also made cross-border mergers and acquisitions feasible within former national flag carriers. M&A between EU airlines were factually already possible after the Third Liberalization Package though.

THE LIBERALIZATION PROCESS OF EUROPE’S SKIES 22

2.4 Liberalization Review

Just like U.S. deregulation, EU liberalization is no closed construct that eliminated all impediments to competition with no single future call for action on further liberalization to remain. In fact, some deficiencies still linger yet those mostly rest on legal intricacies or other areas within this wide and complex industry. ⁴⁵ With respect to competition and market access, the possibility for a state to intervene in regional markets as introduced within the scope of the Second Liberalization Package is still extant. Furthermore, states may still set up autonomous rules on how to split transport within an airport system or even limit the exercise of traffic rights in case of infrastructural or ecological dilemmas. Especially with respect to legal intricacies, for instance considering the supervisory and exercise power of the European Commission in order to enforce European Cartel Law, there is certainly a lot to be said, which will nevertheless be neglected here, due to the focus of this paper. Nonetheless, all regulatory pre-conditions that suggest to set off dynamics on firm entry and exit were met by passing the Third Liberalization Package and, most importantly, with the enforcement of unlimited cabotage in 1997. That is, the market has been fully opened and competition is feasible: the predetermination of designation, capacity and frequency was abolished and fares are left to be determined by market forces only. Moreover, the community ownership rule allows cross-border financial investments without restrictions, even including former national flag carriers due to the November 2002 ECJ ruling and unlimited cabotage gives airlines the opportunity to operate whatever route desired. Consequently, the European airline industry now operates in a European domestic market, thus possibly showing entry dynamics of former “foreign” airlines into another country’s formerly fully closed national market. Therefore, all regulatory preconditions for the analysis of entry and exit within this paper are fulfilled.

However, it is important to understand that the development of the European air transport system, in both sky (with respect to air traffic management and international transport) and ground (with respect to ground service, airport capacities, slot allocations and computer reservation systems) is not complete, although rulings and initiatives have been already started.

⁴⁵ For a detailed analysis of the deficiencies of European Liberalization regarding legal entities see Jung (1999), pp. 21-32.

3 Theory

“The importance of firm entry and exit as determinants of market performance

is well recognized.” 46

This paper analyses the entry and exit dynamics of a former highly regulated and closed national market with regard to the legislative liberalization process that intended to open the market for competition. Apart from traditional measures of industry structure such as concentration ratios, which are static concepts, an industry’s performance can be determined and competition proxied via the dynamic analysis of entry and exit. ⁴⁷ Therefore, actual firm entry and exit within an industry or market is tracked over a period of time and allows drawing several conclusions as a result. First, especially with regard to liberalization, market accessibility is evidenced when actual entry occurs. Second, the threat of entry and actual entry in combination with incumbent behaviour allow to draw conclusions on both strategic entry barriers, anticipated behaviour and remaining structural barriers to entry. Third, entry in combination with exit, thereby resulting in failure rates and within-industry firm turnover, are a proxy for “barriers to survival” which differ across as well as within industries over time. 48

3.1 What do we know about entry?

Theoretical and empirical research has already long analysed the drivers, effects and deterrents of entry. Moreover, actual patterns of entry, survival and exit have been studied and stylized facts could be derived from both theoretical and empirical work. Geroski (1995) provided a brief survey on entry research within which he summarized empirical work as a series of several stylized facts that will be recapitulated briefly:

⁴⁶ Dunne, Roberts and Samuelson (1988), p. 495.

⁴⁷ Dunne, Roberts and Samuelson (1988) were first to analyse the patterns of firm entry and exit across industries over time.

⁴⁸ Cp. Dunne, Roberts and Samuelson (1988) and Geroski (1995).

THEORY 24

As a result of the research conducted so far, it has become common knowledge that market entry is no rare phenomenon, in fact, “large numbers of firms enter most markets in most years”. ⁴⁹ However, entry rates highly exceed market penetration rates. Moreover, entry and exit rates are highly positively correlated, as for example Dunne, Roberts and Samuelson (1988) have shown within their analysis of firm entry and exit patterns across a large sample of plantlevel data, aggregated to the firm-level, for several U.S. manufacturing industries. Consequently, as net entry rates are only modest fractions of gross entry rates, the survival rate of most entrants is low. One reason is the fact that “de novo entry”, which is more common, is less successful. Dunne, Roberts and Samuelson found that entrants, differentiated into new entrants or diversifying entrants, which, in order to enter, do either need to invest in new inventory or not, differ substantially with respect to their probability of survival and the size of the surviving firms. Thereby, “small, young firms” have the highest failure rates. ⁵⁰ However, entry on a large-scale as well as very rapid post-entry penetration rates also seem to be penalized by costs of adjustment. With respect to entry drivers and deterrents, empirical studies allowed for a derivation of stylized results, some of which show that entry seems to react slowly to high profits, which might be explained partly by the fact that “economic estimates of the height of entry barriers suggest that they are high”. ⁴⁹ By contrast, Hilke (1984) proposed that growth is a better driver of entry. Hence, entry rates seem to be “hard to explain using conventional measures of profitability and entry barriers”, especially when trying to explain variations of entry over time. ⁴⁹ Yet, it is known that entry rates come in different waves, frequently containing different entrants, which often peak in the early phase of the life of a market. As for incumbent behaviour, empirical results show that incumbents tend to respond selectively to entry. Thereby, prices are rather unusually employed as entry-deterring strategy. In fact, the range of methods used to prevent entry such as limit or predatory pricing, advertising, capacity extensions or learning curve strategies as well as economies of scale and scope do not only appear to vary across industries but seem to be rather limited in use. Main empirical findings on the determinants of firm entry and exit over the last thirty years, notably within the last decade, showed that limit pricing is among the least used entry deterrence strategies whereas advertising and R&D intensity are strong respectively moderate

⁴⁹ Geroski (1995), pp. 421-440.

⁵⁰ Dunne, Roberts and Samuelson (1988), p. 496.

THEORY 25

barriers to entry, as well as widely used. ⁵¹ Also, economies of scale seem to constitute a significant barrier to entry thereby acting as substitute for entry deterrence strategies. As the required scale of entry increases, the rate of entry decreases. Subsequently, with a high minimum efficient size of entry, entry deterrence is less frequently necessary. ⁵¹ Excess capacity, however, gives raise to divergent concepts. While empirical research mostly finds that excess capacity is both an insignificant barrier to entry and rarely used as entry deterrence strategy, several theoretical studies provided the basis for “the use of excess capacity as a successful barrier to entry.” 52

3.2 Entry deterrence and dynamic competition

Theoretical research proposed that excess capacity can act as a barrier to entry. Thereby, excess capacity is seen as a pre-emptive credible commitment to an aggressive post-entry competition. Accordingly, incumbent firms signal their ability to increase production and engage in price wars in case of entry so as to threaten the possible entrant with lower profits in the post-entry equilibrium and thus discouraging entry. ⁵³ Spence (1997), who employed Cournot competition in the post-entry subgame, suggested that potential entrants believed incumbents to expand output to the installed capacity level following entry and therefore refrained from entry. However, this model is not subgame perfect since Cournot competition would always lead to oligopoly output rather than the expanded (monopoly) output threatened via excess capacity. ⁵⁴ Thus, excess capacity would not be a credible commitment but incumbents would carry idle capacity. Dixit (1980), however, showed that even if subgame perfection is imposed, the choice of capacity can be used strategically to deter entry and Bulow, Geanakoplos and Klemperer (1985) demonstrated that idle capacity can indeed be found in a perfect equilibrium. Given a constant elasticity of demand, an incumbent can build up enough capacity to create zero profits for an entrant but would restrict output if entry did not occur in order to maximize profits and

⁵¹ Cp. Cincera and Galgau (2005), pp. 81-83.

⁵² Mathis and Koscianski (1996), p. 264.

⁵³ Cp. Benoît and Krishna (1991), Allen (1993) and Mathis and Koscianski (1996) for the summary on excess capacity research.

⁵⁴ Note that in Cournot competition firms are price-takers and demand is assumed to be linear in the models under study thereby a two-firm case is assumed (entry into a monopoly market). As prices fall as output is increased, in order to maximize profits, the former monopolistic firm needs to decrease output rather than expand it given entry.