Airport Economics, Policy and Management: The European Union David Gillen Sauder School of Business University of British Columbia Vancouver, BC Canada V6T 1Z2 Hans-Martin Niemeier University of Applied Sciences Bremen Bremen, Germany Rafael del Pino Foundation Comparative Political Economy and Infrastructure Performance: The Case of Airports Madrid, September 18 th & 19 th 2006. * We are indebted to Christiane Müller-Rostin, Vanessa Kamp and Helen Huang for excellent research assistance as well as to students in Transport Economics, Bremen University of Applied Science for assisting with gathering regulation and institutional information..
Table of Contents INTRODUCTION...3 EVOLUTION OF TRANSPORTATION POLICY IN EUROPE: AVIATION REFORM...6 STATISTICAL PROFILE OF THE EU AIRPORT BUSINESS...8 LEGISLATION AFFECTING AIRPORTS...12 PRIVATIZATION: CHANGING OWNERSHIP, INCENTIVES AND PERFORMANCE OF EUROPEAN AIRPORTS...15 PRIVATIZATION OF CENTRAL EUROPEAN AIRPORTS...15 CHANGING NATURE OF PUBLIC AIRPORTS IN CENTRAL EUROPE...17 REGULATION OF EUROPEAN AIRPORTS...18 INSTITUTIONAL SETTING...18 HIGH OR LOW POWERED REGULATION?...19 SCOPE OF REGULATION AND THE SINGLE TILL VERSUS DUAL TILL DEBATE...20 PRICE CAP VERSUS COST PLUS AND SLIDING SCALE REGULATION...22 Cost based regulation....22 Pure and hybrid Price caps...23 Revenue sharing agreements...24 REGIONAL AIRPORTS IN THE EU...27 AIRPORT COMPETITION AND REGIONAL AIRPORTS...31 CAPACITY CONSTRAINTS AND SLOT ALLOCATION...33 SUMMARY...35 REFERENCES...38-2 -
Introduction The airport business has emerged over the last two decades as a thriving commercial enterprise being transformed from government owned and managed entities with public utility outlooks. Nowhere has this been more evident than in the European Union. While it lagged behind the US and many other countries in deregulating its airline industry, it was one of the leaders in reforming policy and regulations affecting its airport infrastructure. Airport policy as a progressive move to improving economic efficiency and liberalizing access in the EU eclipsed most other parts of the world. 1 The BAA was the first fully privatized airport system in the world. This has been a model studied and adapted by many jurisdictions inside and beyond the EU. As well there has been minority selling of airports in Hamburg, Frankfurt, Düsseldorf, Rome, Copenhagen, Zurich and. most recently Paris. The sale of a majority share of Bratislava and Budapest airport was completed in early 2006. 2 Airports are often seen as 'public assets' suitable for promoting economic development. It is therefore not surprising that many airports in Europe are owned by a level of government; national, state or local government. Spain, Portugal, Sweden, Ireland and Greece own all airports in their countries while Germany s airports (larger ones) are owned by state or local rather than national governments. French regional airports are owned by the central government but are managed by local governments. The shift to more market oriented policies has reflected a position, belief perhaps, that airports are not natural monopolies across all sources of revenue so abuse of monopoly 1.Only the Australians and New Zealanders have been more aggressive in moving to privatization and the reliance on market forces to deliver airport services. 2. Infratil has recently purchased Lubeck Airport in northern Germany and other regional airports will move to full or partial privatization as the EU expands. In the new states of the EU, Malta International is 40 percent state owned and Slovenia s Ljubljana is 49 percent state owned. Similar trends are taking place in many countries around the world, developed and developing. - 3 -
power was considered less an issue for concern. 3 Secondly, there has been a shift to considering regulation or semi-regulatory processes as an alternative means of governance for the airport system. There has, for example, been substantial debate of the type of regulation price cap versus cost based regulation with single versus dual till systems and whether there needs to be regulation at all. The most recent experiences have been to consider the role of ancillary revenue in complement with aviation revenue in providing an incentive structure to not exercise monopoly power on aviation revenue. This view has been taken up by airports that depict themselves as an ordinary competitive industry. However airlines, on the other hand, having substantially cut their costs and fares in the last fire years criticize airports for not having achieved similar cuts. They see airports as natural monopolies which are not regulated effectively by an independent regulator. Shifts in airport ownership also provide access to private capital and to private sector managerial, technical and operating skills. This evolution to privatization and commercialization may well have changed the nature of the airport industry. With a view that airports are a modern business leads to changes in incentives as well as the attraction of more entrepreneurial oriented personnel. The EU aviation market is the second largest common market in the world and growth within and between the EU and other parts of the world have been well beyond forecasts. In 2004, the total number of passengers transported by air in the EU rose by 8.8%, to 650 million; 24% were carried on national flights, 42% on intra-eu flights and 34% on extra-eu flights. 4. Growth has been above forecasts at the five or six largest European airports and significant growth has also occurred in the new EU states to the east as a torrent of new low cost carriers take to the skies. These growth rates in traffic 3. This is questionable as airports have continuously earned relatively high rates of return which might be interpreted as monopoly and or location rents. 4. Source: Eurostat - 4 -
(see Table 1 as an example), have lead to growth in airport use. This tremendous traffic growth has resulted in strains on many airports which are reaching capacity limits. This in turn has led to delays, grey markets for slots and pressure to expand capacity, e.g. in Frankfurt. 5 In the east part of the EU the traffic growth is being fuelled for the most part by startup and current EU low cost carriers, this in many cases shifts traffic to regional airports. The major issues facing airports in Europe include: pressures for reductions in aviation rates and charges, the changing demands from LCCs versus traditional carriers and the strategies for dealing with delay. The two additional matters that are driving airport policy are slot allocation and how to internalize environmental costs in airport charges. Europe unlike other parts of the world has started to impose both noise and air pollution charges by either levying them directly or placing the charge on the passenger ticket; e.g. Schiphol Airport and Geneva. Noise charges are most prominent and the Netherlands, Germany and Belgium. 6 In this paper we examine four key issues of airport policy and management; airport privatization, airport regulation, regional airports and airport capacity and slot allocation. The paper begins with a description of the evolution of transport policy in the Europe and its integration with broader transport policy reform, Next we provide a statistical profile of a sample of airports in Europe and include a description of legislation affecting airports. Following this is an assessment of privatization of airports in Europe and its affects on incentives and performance. Should airports be privatized and what is the trend in the EU and in member states? Have private airports become more business orientated? How does privatization change the incentives and performance of privatized airports? 5. The treatment of secondary slot trading in the UK and the EU are quite different and discussed later in this chapter. 6. The TRL reports that noise charges are becoming increasing ineffective as the chapter classification is outdated. This has the implication that noise charges reflect to a lesser extent externalities and noise restricted airports are led to use economically inefficient instruments such as noise restrictions. - 5 -
The regulation of airports is investigated next, how are airports regulated in the EU and what implication has this for incentives and efficiency? After that, we analyze the growing role of regional airports which benefit from the structural changes in the aviation market and highlight some potential for competition. Finally we look at the growing capacity problems of European airports and how efficiently scarce slots are allocated. Will the reform process of slot allocation rules lead to a better utilization of airports and effective incentives for investment in additional capacity? Evolution of Transportation Policy in Europe: Aviation Reform Airlines in the EU operated until the late 1980s under a set of regulations which determined fares, frequency, routes and carrier access to markets. The EU now is the largest open aviation market in the world. With the freedom to compete and develop new products and services the airline industry grew rapidly and hence did the demand for airport services. The European Union's Common Transport Policy (CTP) strongly emphasizes intermodal transport solutions in which different transport modes complement each other. Europe is distinctive in that most states, such as France and Germany, for example, have excellent high speed rail systems and most other European countries have well developed rail networks and services. These can increase the threshold of substitutability to 400 km. 7 The Community policy on liberalizing air transport covers four main areas: market access, capacity control, fares and the issue of operating licenses for companies. It was launched in 1980 and has been implemented in three stages, with Stage 3, the third air transport package, coming into force on 1 January 1993. A transitional period was laid down for air cabotage within the EU, which became reality only on 1 April 1997 7 The evolution of low cost carriers has had a profound affect on this in the last 5 years. - 6 -
The central motive of the three air transport market liberalization packages was to bring the single market to the field of air transport which was (and to some extent still is) marked by state's furthering the commercial interests of their national flag carrier. The opening of the market to all Community carriers which was completed in 1997 when restrictions on full cabotage rights disappeared has led to more equitable conditions for access, thereby increasing competition and efficiency and, to some extent, lowering prices for consumers. However, the air transport market is far from the ideal of perfect competition. At busy airports, the continued existence of restrictive slot allocation practices still acts as a barrier to entry to new carriers. However, the abundance of regional airports and ex-military airfields in most EU countries has tempered this constraint to some degree, enabling LCCs to enter on a large scale. Certainly LCCs have been able to exploit this entry to a large degree. The outcome of European deregulation is in line with the US experience (see Morrison and Winston (2000)). Arndt (2004) shows for the period of 1989 to 2000 that liberalization resulted in fares falling by 31 % to 35 %, flights increased by 20 % to 54 % and consumer surplus rose by 311 US $ per passenger for 1999 in 1989$s. Since 2000 LCC have intensified competition, gaining larger market shares which in turn has lead to further price reductions. Liberalization has not only led to dramatic changes in the airline industry but it has also led to significant changes in the airport industry as competition downstream has put pressure to improve efficiency upstream in the provision of airport services. Yet, airports in the overall supply chain have maintained their profitability from both location and, in some cases, monopoly rents while airlines as a distribution system for airports earn competitive or lesser returns. Liberalization has also resulted in changes in network structure leading to a change in demand for airport services. Brussels with Sabena, Zurich with Swiss and AMS with the merger of KLM/Air France have affected the demand for airport services. Yet, these - 7 -
airline failures are not the trend, the shift in business models with LCCs increasingly dominating domestic markets and legacy carriers more focused on international markets has affected the demand for airport services. With these changes, the lack of efficient prices for runway and other airport services has led to a crisis in investment in capacity as the price signals provide the wrong incentives. This means there will be increased slot scarcity which increases the importance of the investment issue. Statistical Profile of the EU Airport Business Europe has two distinguishing features from airports elsewhere in the world, there are an enormous number of airports and, the airports are connected with and compete with a well-developed rail system. The rail system has reduced the distinction between airports and insofar their substitutability. On the other hand hubs like Schiphol and Frankfurt have increased their catchment areas and gained market power. While there are essentially 5 large hubs in the EU (LHR, FRA, CDG, AMS, MAD, MUN) there are a number of regional airports varying in size from 24 million passengers at Barcelona to 10 million at Hamburg and 5 million at Prague. Like the US there are multiple airport groups but unlike the US this can sometimes be an entire country s airports, e.g. AENA, a division of the Spanish government, owns and operates all airports in Spain, Finnish Airports group (CAA) owns and operates 25 airports in Finland, LFV (Swedish state enterprise) owns and runs 19 airports in Sweden. Airport groups include Aeroports de Paris, Airports de Roma, and the ANA group, which manages all major airports in Portugal. Airports in a number of countries are quite close together which creates benefits and costs. 8 In Table 1, 20 of Europe s top airports are listed for passenger and freight transport, including the growth in traffic from 2003 to 2004. 9 It is clear from this table that the growth in passenger traffic has been remarkable and has been spread across Europe and not 8 Regional airports are discussed at length in a later section. 9 UK airports are included for a more complete comparison. - 8 -
concentrated at a few airports. Freight traffic growth has been less evenly distributed reflecting the relative growth in freight carried by integrators; the significant growth at Cologne-Bonn reflects the developments of UPS at that airport, for example. INSERT TABLE 1 ABOUT HERE In the following tables, we provide a profile of revenues and costs for a sample of airports and airport groups in Europe (more complete data for a number of airports is contained in an appendix). These data are designed to illustrate the differences in revenues, costs and profitability of the sample of airports and airport groups. This diversity reflects the differing approaches to the airport business as well as the type of traffic an airport serves. Table 2 (A and B) provide an illustration for a sample of airports and the airport groups the total aircraft movements; the five large hubs plus Munich are an order of magnitude larger than the remaining larger regional airports; about 25 percent larger. Airport groups represent traffic for most in not all of a country. INSERT TABLE 2A AND 2B ABOUT HERE Table 3 and 4 illustrate for a sample of airports total operating revenues and costs while Table 5 and 6 provide the same information for airport groups in Europe. Among airports LHR dominates in both revenues and costs. Looking only at mainland Europe MUC is noticeably larger than the other major and regional airports but in all cases these airports have in excess of 5 million passengers; the level at which most scale economies are exhausted. There is much more variation in costs than in revenues; the standard deviation is 55% of costs while only 36% of revenues. Looking at the airport groups, both revenues and costs are dominated by BAA, AENA, the Fraport group and ADP, the remaining groups are relatively small as they have no - 9 -
major hub or major regional airports in their groups. Interestingly the airport groups exhibit the same amount of variation in revenues and costs, about 80%. INSERT TABLES 3, 4, 5 AND 6 ABOUT HERE Table 7 exhibits the proportion of aeronautical revenues in total revenues for both airports and airport groups. 10 This figure indicates differences in business strategies as well as incentives provided by the form of regulation of airports. CPH which is regarded as well run and having an excellent retail strategy draws just over 50% of total revenues from airside charges but they draw another 20% from ground handling. HAM and DUS high proportion of aviation revue also reflects contributions from ground handling. The UK airports are less than 50% and the average for the mainland European airports (in the sample) is 57%. INSERT TABLE 7 ABOUT HERE Profitability as measured by EBITDA per passenger for our sample of airports and for the airport groups are represented in Table 8 and 9 respectively. The BAA dominates in overall EBITDA and on a per passenger basis as well. VIE and CHP are both reputed to have excellent retail plans, and this is reflected in their EBITDA values. VIE has an EBITA of $11 per passenger with their strong revenue strategy overcoming somewhat higher costs than average (see Table 5). CHP s lower cost despite lower revenue contributes to its EBITA performance. Airport groups earn a much lower EBITDA than individually managed airports. The average for the groups is $7.64. INSERT TABLE 8 AND 9 ABOUT HERE Table 10 provides information on the relative cost efficiency of a sample of airports and the airport groups. Total operating costs for airports vary over a significant range, from just over $5 to near $23 and an average of $13.47 while airport group costs per 10 Aviation revenues include ground handling, aeronautical revenues in the definition of ATRS does not. - 10 -
passenger have a similar variance but a higher mean, $15.87 per passenger. To provide a different perspective on relative cost efficiency Table 11 presents costs relative to the low operating cost airport; for airports this is Tallin-Ulemiste Airport in Estonia and for airport groups, ADP (Portugal Airports group). For airports, TLL is highly efficient, this is in part a result of very low wages; average employee compensation was approximately $13,000 and compare this with HAM at $68,000 and AMS at $82,000. TLL is apparently more than 4 times more efficient than CGN, but this is misleading since CGN is a large cargo airport (with a UPS base). However for passenger oriented airports including VIE and MUC, there is a significant cost difference. There is an unclear relationship between cost efficiency (as measured by cost per passenger) and EBITDA. Examining Table 12 airports such as MAN have high EBITDA but are very costly, yet LGW is both profitable and relatively cost efficient while MUC which is high cost ($22 per passenger) yet earns a relatively low EBITDA of $4 per passenger,. This brief summary provides a glimpse into the relative size and performance of the airports in Europe. There is significant variation across both cost and revenue performance. Some airports such as FRA have high costs but also high operating revenues resulting in a high EBITA performance whereas CPH is cost efficient and generates approximately $15 per passenger in operating revenue, yielding a below average EBITA. Efficient airports are those which provide economically efficient levels of output, but output is a combination of aircraft movements and passengers handled. A different ranking of cost efficiency occurs when comparing cost per passenger versus cost per movement yet there is a high correlation, 0.75 with much of the difference being accounted for my cargo oriented airports. However, economic efficiency requires all of allocative, productive and dynamic efficiencies; most all studies have focused on productive efficiency. - 11 -
Legislation Affecting Airports Countries that wish to participate in the European Common Aviation area (ECAA) must adhere to EU law in the filed of aviation. Prior to December 2005, member countries in the EU had bilateral agreements with the most recently added member states. These states must alter their national law to become part of the EUAA at which time their bilateral agreements become obsolete. The Guide to Community Legislation in the Field of Aviation provides a detailed list of all legislation, which applies to aviation. 11 Such legislation covers eight areas: economic policy, air traffic management, safety, security, environmental matters, social matters, passenger protection, and external relations. The Guide makes the point that such legislation is mandatory for member countries, that there is no possibility of filing differences and perhaps most interestingly, that economic aspects were of particular importance in developing the policy. In effect the policy seems to focus on consumer sovereignty rather than producer protection. The EU air transport policy is a policy for the whole of the transport sector meaning all air carriers, airports and air traffic control services. All eight areas have application for airports. Economic policy covers licensing, insurance requirements, access to air routes, slot allocation and prices for services. Also included in this section would be competition rules. Annex III and article 5 of the ECAA Agreement specify the set of competition rules applicable and make particular reference to Articles 81 and 82 of the treaty of European Communities; Article 81 refers to actions by economic agents which have as their object or effect the prevention, restriction or distortion of competition within the common market while article 82 focuses upon abuse of dominant position. 11 See, Guide to Community Legislation in the Field of Aviation, Report Issued for Director General Energy and transport, European Commission, May 2006-12 -
Environmental matters focus predominately on noise emissions and noise management. More recently air pollution has become a major issue. There are no directives aimed specifically at aviation covering air pollution but there are proposals to include aviation in the tradable permits scheme. Article 87 of the Treaty has particular application to airports, as well as carriers. Article 14 and Annex III of the ECAA agreement set out the conditions of state aid, specifically such aid was not allowed to distort or threaten to distort competition by favouring agents, undertakings or products. The form of aid may be grants, interest relief, tax relief, state guarantees and preferential access to state provision of goods and services or purchasing. It could also include restructuring aid and exclusive rights concessions. Aid is allowed to be provided under some conditions, such as a regional development program, but such aid must be available to all parties. This was the key issue in the Charleroi Airport-Ryanair case; aid was made available only to Ryanair. The EU has made a distinction between capital and operating subsidies but they have not balanced this distinction across modes or between service and infrastructure. 12 In aviation subsidies can take a number of forms; in Spain (Aena Spanish Airports) lossmaking airports are assisted by profitable airports. State governments have provided assistance to airports; Schiphol, Charleroi and Strasbourg are examples. Other means of state aid include capital subsidies, route support in the form of landing charges, Ryanair has been active in seeking (and receiving) support from Strasbourg Airport (ruled illegal by a French court). Morrell (2006) cites Pau, Stockholm Skavsta, Klagenfurt, Girona, Birmingham, Stansted and Aarhus as other airports that have supported Ryanair. There are some pieces of legislation as well as Directives aimed specifically at airports. A Directive covering ground handling was made in 1996 (October) which mandated that 12.The EU has implicitly approved operating and capital subsidies for rail - 13 -
airports provide access to third parties in the ground handling market. The purpose of the Directive was to ensure all carriers had equal access to efficient ground handling services. Historically, ground handling was a monopoly provided by the airport (as in Germany) or the airline (national carrier), as in Spain. The impetus for opening ground handling markets to competition came from carriers. The airports were slow to implement it as in many cases they were earning sizable profits. The EU wanted to ensure there were no barriers to entry or expansion of existing and entrant carriers. The outcomes from the directive differed across countries. In the UK the market became completely open with independent ground handlers providing service, in Germany the airports lobbied to have the national government pass legislation (in spite of the EU directive) to have the airport plus one independent provider; that is, a duopoly with the airport being guaranteed a position. In Spain the outcome was also a duopoly but with the airline and one independent, Italy had the airport and 2 independents, which were carriers. In France the outcome was mixed as Airport de Paris simply kept a monopoly while other airports in France had open access to ground handlers. As we can see there were not a variety of market structures that evolved, essentially all were duopolies (except in the UK and Netherlands), albeit with different composition of players. Slot trading is also an area covered by EU legislation. Under current EU rules slots must not be traded and unused slots must be returned to the pool for redistribution. However, it is well known there is an active grey market for slots in mainland Europe and an active trading market for slots at busy UK airports. 13 The UK has been pushing for a formalized secondary trading market while other member countries are less enthusiastic about formalizing the market. 13 This issue is discussed in detail later in the paper. - 14 -
Privatization: Changing ownership, incentives and performance of European Airports In this section we first analyse the effects of privatisation which has led to variety of different forms of private airports. As the public airports will play a major role we look secondly at the variety of public forms of ownership and organization. Privatization of Central European Airports In 1987 the British government privatised the three London airports together with the BAA s Scottish airports thereby triggering a trend of privatisation of airports in the midst 90s which has transformed the UK airport industry. The majority of UK airports are privatised and the industry is subject to UK style public utility regulation (Graham, 2004). While BAA s performance and its rising share prices were widely seen as a success making it a kind of role model for the privatisation of airports most European governments were rather reluctant to privatise their airports as fast as the UK and on such a large scale. Airport privatisation gained momentum in the 90s with the first wave of privatisation of Vienna (share of 27 %) in 1992, Copenhagen (25%) in 1994, Athens (45%) in 1996, Dusseldorf (50%), Rome (45.5%) and Naples (65%) in 1997, Skavsta Stockholm (90 %), Florence (39%), Turin (41 %), Hamburg (36%) and Zurich (50%) in 2000 and finally Fraport (29%) in 2001. The crisis of aviation from 2001 onwards more or less broke this wave and only recently has the process seems to start off again with the partial privatisation of Brussels, Budapest, Bratislava, Lübeck, Malta and Paris in 2006. Privatized airports may be listed on the stock exchange with a diverse or concentrated ownership structure, or they might be sold to a strategic investor, other airport operators or financial institutions. Very often full privatization is restricted as the former public owners want to secure certain political interests to be guarantied by a golden share or a wide ownership clause. Currently at most airports the private owner has only a minority - 15 -
stake either in form of a stake of up to 49 % signalling private public partnership in roughly equal terms or in form of a minority of less that 25 %. Only the airports of Bratislava, Brussels, Copenhagen, Malta, Vienna (50 % plus 10% employee foundation) are privately owned by a majority share. No major airport in Continental Europe has been fully privatised without any ownership restrictions. Only for Ireland, Netherlands and Malta are there strong expectations that airports will be partially or completely privatized. To this point privatisation has not changed the nature of the industry as in the UK, but it has made the airports more profit orientated. This has contributed to some cost cutting efforts especially in ground handling and to the development of non aviation business. ADP is a typical example of partial privatisation to develop retail and retail business which have been neglected by the public utility type of airports management. Privatisation has not lead to a strong tendency to increase aviation revenues by differentiated price structures and peak pricing. The typical private airport in Europe is not a fully but a partial privatized airport. Such an airport has with a wider range of motives like regional development, job creation and son than the purely profit maximizing airport. Privatisation has also not lead to more airport competition as the examples of BAA and ADP sold under common ownership and the takeovers of Frankfurt with Hahn and Vienna with Bratislava 14 show. Privatisation has certainly led to a greater variety of governance structures than ever before. In this respect Europe is certainly different since within the same liberalized aviation market the provision of airport services are governed by different airport models sometimes even within a member state. 14 Erste Bank commented that profits from Bratislava/Kosice will be low but it is still better to pay a high price and receive low contributions that let a strong competitor grow next door (2006, p. 1). Erste Bank estimated that the purchasing price of 525.7mn includes a premium of 359mn - 16 -
INSERT TABLE 12 ABOUT HERE Changing nature of public airports in Central Europe Parallel to the privatisation process public European airports have changed their governance structure as well. The federal government of Austria for instance has not only sold its 50% stake in Vienna airport but also the 50% stake in regional international airports such as Graz, Innsbruck, Linz to regional and local administrations (Schneider, 2004, 150). Other countries like the Slovak Republic have done the same. In the public sphere we observe different owners at the various levels of the state from the level of the municipality, to the region and up to the central government (see Table 13). At the lowest level, the city government is typically the owner of the airport. As many airports are located at the outskirts of a city and serve a whole region, very often the city governments together with the governments of surrounding cities own the airport. In federal states such as Austria and Germany the regional state also has a share in the airport. In other countries airports are owned by the central government either fully, or partially. While there seems to be less change within the ownership of public airports the owners have adopted new organizational structures over the past three decades. INSERT TABLE 13 ABOUT HERE Traditionally a public airport was run as a public bureau. This, however, caused major management problems as investment had to be financed out of the public budget and depreciation of assets was not taken into account. As a result an increasing number of public owners corporatised their airports by transforming them into firms with commercial accounting methods. More commercial freedom was given to the management and - 17 -
airports become more commercialised. They adopted private management tools for cost control and began to market themselves. Profit making becomes an objective in airport management. Regulation of European Airports In this section we provide an overview how European airports are regulated and we assess whether regulation increases efficiency and economic welfare. Only in Switzerland the partially privatized Zürich airport is not ex ante regulated and as in New Zealand only subject to competition law. We focus on two central questions; are the airports regulated by an institutional setting which guaranties a fair process, and second, how low or high powered are the regulatory systems in Europe. What incentives for productive and allocative efficiency do the different systems set for airport management? Institutional setting Regulatory economics agrees on certain principles, namely that the regulator should be independent, but accountable to democratic bodies and that regulation should be a fair, accessible and open process. While these principles are self evident they nevertheless are not so evidently practiced in Europe. Our survey shows that an independent regulator is established in only a few European states, namely in the United Kingdom, Ireland and the Netherlands. Interestingly very often countries are privatizing their airports without avoiding conflicts of interest. In Austria Vienna airport was privatized in three steps in 1992, 1995 and 2001. Up to the last step the central government held a major share and regulated the airport charges. German airports are regulated by the federal states that have a minority or majority share in the partial privatized airports of Frankfurt, Hahn, Hamburg and Hannover (for further discussion see Niemeier, 2002). A fair, accessible and open process requires as a minimum a consultation process. In the past 15 years more and more European states have implemented a consultation on - 18 -
airport charges and it has become standard practice today as the ACI survey (2003) and our survey illustrates. Still there is room for improvement as in most consultation processes the airports do not provide the necessary information to make a decision on airport charges transparent or plausible to the airlines. The standards of UK regulation which are even open for the general public are hardly met in continental Europe with the exception of Brussels airport in which the regulator demands a consensus among the airport and its main users This is clearly indicated by the recent decision of Lufthansa to acquire a 9.1 % share in their main hub Frankfurt, so that Lufthansa can obtain a seat on the board of directors in order to be better informed. Another example is the recent price cap regulation of Aeroports de Paris in which the value of the regulated asset base and the percentage return on capital are not disclosed by ADP or the French government (Morgan Stanley, 2006, p. 4). While it might be very complicated to completely avoid the risk of capture many of the European regulatory systems which do not separate the functions of ownership and regulation and lack transparency and fairness seem to be perfect for regulatory capture. It gives the management of the airport the rare opportunity to influence through the owner the regulator in various ways which in turn lessens incentives for efficiency and might create rents for management and employees. High or low powered regulation? Regulation has changed in the last 15 years and today European airports are regulated under systems with very different regulatory power (see ). The central problem for regulation is the regulator has asymmetric information about the demand and cost functions and that the regulator must design a contract to set incentives for the regulated firm. While high powered regulation sets incentives for cost reductions and productive efficiency, an efficient price structure low powered regulation does not. This power is largely determined by whether the regulated prices are cost based or price capped, but also by the decisions on the scope of regulation as well as risk sharing arrangements. As - 19 -
with other regulated industries, practice, which is evolving rapidly, continues to outstrip theory (Newberry, 2003), we confine ourselves to an evaluation of the central points 15 of European airport regulation to get a view on how effectively the incentives for efficient performance are set. Scope of regulation and the single till versus dual till debate Regulation should be confined to those activities in which the airport has persistent monopoly power. This is the case where the airport services are essential for downstream users and cannot be duplicated without substantial costs. Traditionally the airside system with the apron and the passenger and freight handling terminals are regarded as such services. Non aviation activities and ground handling are seen as activities in which the airport might have some market power, but at least potential competition might discipline the airport (Templin 2006). Ground handling was liberalized in principle, in 1996 so that self handling and third party providers could enter the market. Ground handling services should not be regulated, but central infrastructure services such as package handling systems should be part of the regulated activities. This is not an issue in countries such as the UK where airports do not offer ground handling, but it is of importance in countries such as Austria, Germany, Italy, Spain and France where airports provide ground handling services. In Germany the regulation of charges does not include regulation of the central infrastructure fee offering airports an easy way to shift costs to their users. At Amsterdam, Brussels and Rome Fiumicino the central infrastructure fee is regulated while in Paris Charles De Gaulle and Madrid Barajas this is not the case (Templin, 2006). Besides the direct regulation of certain activities of an airport all activities can be regulated indirectly through the single till principle under a price cap as well as under a 15 We had to leave out such topics as price basket versus average revenue cap and cost pass through mechanisms. See Graham (2003), Kunz (1999 and 2000) - 20 -
cost based regulation. This is one of the most debated issues among airports and airlines (for a good overview see Starkie and Yarrow, 2001). The single till principle was recommended by ICAO and has been widely used in Europe, but this long tradition is slowly breaking up. The UK was obliged to regulate UK airports by a bilateral air service agreement, but this is no more the case. The CAA (2001 and 2002) recommended a dual till in 2001, but surprisingly the Competition Commission rejected the proposal. However, the debate is not over because some airports have moved to the dual till principle. In Europe the price cap for Hamburg Airport was the first one to be set on a dual till in 2000 because regulation should be confined to the monopolistic bottleneck and incentives for developing the non-aviation business should not be lessened (Niemeier, 2002). In 2001 Malta airport followed suit with a dual till price cap. Brussels airport which is regulated on a rate of return basis with some yardstick elements has defined a stepwise move from the single to the dual till over the next 20 years. The regulatory framework for ADP sets some mixed incentives to develop non aviation business. The retail and real estate business are, relative to other European Hubs, underdeveloped and ADP has roughly the size of London Heathrow to develop this business. One might think that the French government may have an interest to develop this business as this increases the value of the airport. However, the chosen regulatory framework is a single till with a vaguely defined option to take part of real estate and retail income out of the till in the next regulation period from 2011 to 2015. 16 The differences are mainly due to different degrees of non-aviation business left out of the till. 16 Morgan Stanley (2006) values the ADP in different scenarios between 38.1 and 127.1 per share. - 21 -
Overall the European regulation should be better targeted at the monopolistic bottleneck. In ground handling it is not regulating enough and in the non aviation business to much leading to distortions in these competitive markets and in the core business of airports. Price cap versus cost plus and sliding scale regulation In Europe airport charges have traditionally been regulated on a rate of return or cost plus basis (see table 13). This is a rather low powered type of regulation as it sets no incentive for cost reduction. While the following discussion showed that many of the original claims of superiority were exaggerated and that there might be better regulatory systems there remain important differences among the practical regulatory regimes for airports. We differentiate between cost based regulation, pure price caps, hybrid price caps and revenue sharing agreements. Cost based regulation. Many authorities in Europe regulate airport charges according to principles of cost relatedness. The charges should create just enough revenues to cover total costs including the depreciation of capital and a normal rate of return on capital. The structure of charges should also be cost related, namely each charge should reflect its costs. In Europe many of the public airport systems like Greece, Poland and Finland set their charges in this way. Charges are supposed to be set according to ICAO principles of cost relatedness. CAA s and Departments of Transport which operate and manage airports directly follow this principle. In the case of formally privatized airports such as most German airports the regulator approves charges only if they are cost related. The problems with cost based regulation are twofold. Firstly, the incentives are set for an inefficient choice of inputs (Sherman, 1989). Secondly, cost based regulation leads to an inefficient price structure (Sherman, 1989). Under cost based regulation the airport has no incentive to adopt peak pricing, but rather to lower the price of capital intensive peak demand in order to justify more capital assets, and charge a monopoly price at off-peak - 22 -
times to realize a profit that greater capital will justify (Sherman, 1989, p 241). The incentives at cost based regulated airports lead to average pricing without peak and congestion pricing. The scarcity of airport infrastructure is not currently correctly evaluated by a functioning price mechanism as airports do not charge efficiently and slot allocation is not reformed. Peak pricing is rarely practiced. Weight related charges are kept even if demand outstrips capacity at peak times or during the whole day. This leads to false prices since slots are not allocated to those with the highest willingness to pay. In a situation of excess demand charges should theoretically be a fixed charge otherwise small aircrafts pay less than large aircrafts. From the six busiest European airports only London Heathrow and Gatwick are levying a fixed charge per movement in 1994 while Düsseldorf, Frankfurt, Madrid and Paris Orly have weight related charges. The ratio between a large aircraft like a B 747 with 396 passengers and a small regional jet like then CRJ with 38 passengers is about 4.9 in London. In contrast to this the two German airports have a ratio of 8.1 to 8.8, Madrid and Orly of 13.5 and 13.7. From this price structure small aircrafts benefits leading to relative high percentage of aircrafts in the size of up to 49 seats (Forsyth & Niemeier). When excess demand exists, it is not rationed away efficiently, but rather, expensive additional capacity is provided which raises environmental and political concerns. Thereby the question of airport expansion becomes more and more of a political question. Pure and hybrid Price caps Unlike cost based regulation price caps do not regulate profits, but set incentives for cost reduction. The gains from cost reduction can be kept by the regulated airport within the regulation period and might then be passed on to the users via lower charges in the next - 23 -
period. Quality might be monitored or regulated since the airport might try to achieve cost reductions by lowering quality 17. Pure and hybrid price caps differ in the way by how the X in the price cap formula is set. Pure price caps set the X without reference to the costs of the regulated firm by benchmarking while hybrid price caps set the X with reference to the regulated cost base. Hybrid price caps provide fewer incentives for cost reductions. Only a few regulators of public utilities use pure price caps. Hybrid price caps are the standard of UK regulation. For European airports none of the regulators have developed a pure price capping system. The price caps at ADP, Copenhagen and Dublin are based on costs. Hybrid price cap regulation is superior to cost based regulation because it is forward looking while cost plus regulation relies on historic costs. The regulatory lag of typically 5 years is sufficient to set al least some incentives towards cost reduction. Most important price cap regulation does not regulate the charges structure according to arbitrary cost allocations based on historic costs. A well defined price cap sets incentives for Ramsey prices as well as for peak pricing. However, these incentives have been not strong enough to convince airport managers to implement such price structures which are heavily opposed by airlines (Forsyth & Niemeier 2003 and 2004). Revenue sharing agreements Revenue sharing agreements in the European airport industry often relate the level of charges to the passenger growth over a certain period. These so called sliding scales can be combined with price cap regulation as in the case of Hamburg and Vienna (), but do not have to be. At the German airports Frankfurt and Düsseldorf they are the result of Memorandum of Understanding between the airports and its users legalized as a public 17 On quality standards monitoring of UK airports see CAA, 2001, of Hamburg airport see Niemeier, 2002, of ADP see Morgan Stanley, 2006-24 -
contract between the airport and regulator (Klenk, 2004). In case of disagreement the charges would be fixed according to cost based The core of these contracts is a revenue sharing agreement. The average charge per passenger will be determined by the future passenger growth rate. At Frankfurt airport both parties agreed that with a projected growth rate of 4% average charges could be raised by 2%cent. Note that these are nominal prices as the agreement is not related to the price level. In the case of a higher growth rate airlines participate with a 33% share in additional revenues. With lower growth rates the airport cannot fully compensate revenue losses through higher charges. Only 33% of the loss can be compensated. The agreement results in a sliding scale of airport charges which is related to passenger growth. Such agreements have the important advantage that they break with the tradition of low powered cost plus regulation. Within the contract period, the airport may behave as though it is subject to a price cap, though not of the CPI-X form. Furthermore, at first sight they offer some stability if demand fluctuates. A demand shock leads to higher charges so that the airport can cover average costs as to avoid bankruptcy which would undermine the political stability of regulation as well. However, there are disadvantages. Firstly, in the cases of Frankfurt and Düsseldorf the incentives for cost reduction and for traffic increase are rather mild as the level of charges is stabilized at a high level. Very often the agreement seems to reflect more the limited bargaining power of airlines and the absence of an active regulator acting in the interest of the consumer and the public. It could be that for example Fraport prefers the contract to the cost plus regulation because the rate of return on the aeronautical assets is higher than the normal rate of return accepted by the cost related regulation. Secondly, a flat linear sliding scale guarantees the airport nearly the same revenues irrespective of output. This reduces the incentives to differentiate charges and increase output. Thirdly, it usually creates an inefficient price - 25 -
structure. Fast rising demand leads to lower charges and lower demand to higher charges. The effects of the different regulatory systems for airports have not been assessed. Unlike other regulated industries (for the positive effects of price cap regulation see Ai and Sappington (2001)) there are no studies on the cost savings and productivity effects of price cap regulation or on gold plating effects of cost based regulation. The effects of price cap regulation on price structure are even less well empirically studied. It should be recognized that there are limits on the strength of incentives and that implementing peak pricing has a political dimension as the majority of airlines which practice peak pricing oppose peak pricing by airports because it reduces their rent from scarce slots are hostile to them. INSERT TABLE 14 ABOUT HERE Another less well-researched effect is the effect of different regulatory systems on investment behavior of airports. Even though the current slot allocation system and the lack of peak and congestion pricing at European airports makes it difficult to estimate the excess demand European airports will have to invest in new runway and terminal capacity. While rate of return regulation might lead to overinvestment price cap regulation has been criticized by Helm and Thompson (1991) of leading to underinvestment. The regulated airport might not invest because it fears that after it has invested it will not recover its sunk costs since the regulator might opportunistically set the price cap too low in the next regulatory period. However, the empirical support for this theory has been weak and Starkie (2006) found no evidence for the BAA airports. An asymmetry exists as prices which are too low lead to unprofitable investment, whereas prices which are too high need not necessarily lead to higher investment especially if - 26 -