книги / 704

2. RESEARCH REPORT

In order to advance towards the Action Plan’s goal of a single ASEAN aviation marketplace, ASEAN members adopted two agreements: the 2008 Multilateral Agreement on Air Services (MAAS) and the 2010 Multilateral Agreement on the Full Liberalisation of Passenger Air Services (MAFLPAS), each composed of a number of implementing protocols. The agreements and their implementing protocols allow for progressive liberalisation in an effort to increase their acceptance by the ASEAN States.

ASEAN States, not unlike their EU counterparts, took a phased-in approach to liberalisation with important milestones coming into effect over time in order to give the market time to adapt to a new liberalised environment and for regulatory convergence within member countries.

The 2008 MAAS agreement gradually introduces market liberalisation up to the fifth freedom through a series of six protocols. The protocols divide intra-ASEAN routes into three groups of growing importance: intra sub-region, inter sub-regions and between national capitals. In each case, a first protocol liberalises third and fourth freedoms, while a second one liberalises fifth freedoms. This segregation reflects the fact that most national primary markets in the regions are concentrated national capitals.

The agreement has been slow to implement and the initial deadline for ratifying all six protocols by 31 December 2010 was missed. Eight of the ten ASEAN countries have ratified the agreement. Indonesia, the largest ASEAN member, has yet to ratify the protocols. However, Indonesia has expressed concern that its high fuel prices and tariffs on aircraft and aircraft parts may put it at a disadvantage compared to other ASEAN countries. Indonesia also faces a situation of high congestion and challenges in expanding capacity at Jakarta’s Soekarno-Hatta International Airport, the nation’s busiest airport by far. Meanwhile, the Philippines have ratified the first four protocols but have yet to liberalise access to Manila’s Ninoy

Aquino International Airport, citing high congestion suffered at that airport (Hanaoka et al., 2014).

While ASEAN governments attempt to liberalise air services, local LCCs have been aggressively developing their business plans. Malaysian carrier Air Asia, for example, established six foreign jointventure affiliates within ASEAN countries, majority owned and controlled by local nationals to comply with ownership and control legislation in the various ASEAN nations. Meanwhile, Singapore’s Tiger

Airways established locally owned and controlled subsidiaries in Indonesia and the Philippines. In Indonesia, Lion Air established joint ventures in Malaysia (Malindo Air) and Thailand (Thai Lion Air) while

Australia’s JetStar, itself a subsidiary of Qantas, established Jetstar Asia Airways in Singapore and Jetstar Pacific Airlines in Viet Nam50. Low-cost Southeast Asian carriers have thus been at the forefront of liberalisation in the region, finding innovative ways to mitigate existing restrictions in air services. In some cases though, it has been critical of local governments, who often are both the regulator and the owner of the local full service carrier, especially on the question of interpreting effective control which is often a question of interpretation by the regulator.

Hanaoka et al. (2014) have described this corporate structure model as “branchising” and can be useful in obtaining regulatory permission to operate a common brand under multiple joint ventures. However the strategy is not always successful. For example Air Asia had to abandon its planned joint venture with VietJet Air as Vietnamese authorities would not approve the name VietJet Air Asia in 2011 since they do not allow Vietnamese carriers to adopt names and logos of foreign airlines after having made an exception once for Jetstar in 2007. Air Asia also struggled in establishing Air Asia Japan because of strategic divergences with its Japanese partners, All Nippon Airways (ANA), which led Air Asia to pull out of the joint venture and ANA to rebrand the carrier as Vanilla Air. Since then, Air Asia has attempted to relaunch the Air Asia Japan with new partners and expects to start operations in 2016.

Tan (2010) examined these foreign affiliates and found that they respect the letter of the law in terms of local ownership, with each entity being majority owned by nationals from the country where the airline is based. He explains that these agreements introduce the concept of an ASEAN community carrier,

2. RESEARCH REPORT

which could be majority owned by interests from various ASEAN member countries. He also finds that both the CEO and a majority of board members are local nationals, thus meeting national airline control requirements. However, he is less clear on effective control since aviation management expertise, brand management and marketing are all vested with the parent carrier.

Hanaoka et al. (2014) point out that foreign ownership limits continue to exist with the ASEAN community and vary between 40% in the Philippines and 49% in Indonesia, Malaysia, Thailand and Viet Nam, although in the case of the latter, no single foreign investor may own more than 30% of a Vietnamese carrier. Only Singapore allows for 100% foreign ownership, but in ASAs, the other country may not recognise an airline with less than 50% Singapore ownership as being entitled to Singapore’s traffic rights under that ASA. In fact, Singapore Airlines limits its foreign ownership to under 50% to make sure it can take advantage of any traffic rights granted to Singapore by a third country through an ASA.

The ASEAN agreement is actually quite liberal when it comes to the matter of ownership and control as it distinguishes between regulatory control and economic control. Airlines based in one ASEAN country must be under full regulatory control of that country on matters such as safety, security or the environment. But economic control must be in majority vested with nationals from ASEAN, a concept which resembles the Community carrier present in the EU. However, unlike in the EU, the wording of the agreement says that states may accept the designation of such a carrier rather than must accept, which creates a risk for airlines with minority local ownership that their application to be designated could be turned down, as was the case with Air Asia Pacific.

One of the clearest impacts of liberalisation can be observed on the Kuala Lumpur-Singapore route. For 34 years the route had been a shared duopoly between Singapore Airlines and Malaysia Airlines. Once the route was liberalised, first in February 2008 to allow limited new services by LCCs and then in December 2008 to fully align with MAAS, Air Asia, JetStar and Tiger air joined the two incumbent carriers. The results were immediate. Traffic on the route jumped from 1.7 million passengers in 2007, a level it had sustained for most of the decade, to 2.5 million passengers in 2009 and 2.7 million passengers in 2010. Within three years, flights between both cities had grown by 27% while tourism to Malaysia from Singapore rose 24%. This surge in demand was in response to the average ticket price dropping from USD 180 to USD 30 (Hanaoka et al., 2014). The previous example is an empirical reminder that traffic stimulation best occurs when a market is both liberalised and subject to competition between full service carriers and LCCs.

In 2010, ASEAN concluded its first ASA as a trading block. The agreement, which was signed with China, is a milestone in that for the first time there is the recognition of the concept of an ASEAN carrier. The agreement, which provides for unlimited third and fourth freedoms between both parties, cannot be called an open skies agreement in the contemporary sense of the word, but it does represent a significant improvement in a part of the world that is still not liberalised. The agreement is silent on the issue of airline competition, reflecting the absence of an ASEAN-wide competition framework and the lack of a uniform regulatory system for the entire ASEAN block (Tan, 2010).

As we have seen, the ASEAN agreement is silent beyond fifth freedoms and thus no seventh freedoms were granted for ASEAN carriers from one country to operate stand-alone flights from another ASEAN country to China. Tan (2014) argues that this creates an uneven playing field for ASEAN carriers as Chinese carriers could, in theory connect any point in China to any point within ASEAN countries, something no ASEAN carrier can do. While Chinese carriers do tend to focus on their own respective hubs, negating somewhat this argument, they also operate point-to-point services from secondary cities to ASEAN countries.

2. RESEARCH REPORT

Considering the economic potential of the region, the continued emergence of a middle class and the success LCCs have had in penetrating and reshaping this market, one can expect that a future fully liberalised ASEAN single aviation market, combined with adequate infrastructure and sensible aviation policies, could lead to very strong traffic growth in the region and a multiplication of liberal ASAs between ASEAN countries and other states or trading blocs.

The Association of Southeast Asian Nations–European Union air services agreement

In 2014, ASEAN officials and their EU counterparts began discussing the possibility of crafting an open skies type of ASA between the two blocs. This would be in line with previous EU open skies agreements, much as with the United States or Canada but would be a first for ASEAN in negotiating as a bloc with another bloc. The impetus for such an agreement came both from the fact that traffic between both regions doubled in the last 15 years and that this traffic is increasingly being routed through the Gulf countries or Turkey, although the latter’s ASEAN network is far less developed (Tan, 2014).

The EU and ASEAN vary greatly in their level of integration. While the EU has truly formed a single aviation market in every practical sense, ASEAN countries continue to operate their own aviation markets with some regional integration. For example, EU carriers enjoy all nine freedoms of the air when operating within the EU, whereas ASEAN carriers are limited to unlimited third, fourth and fifth freedom rights within the ASEAN region. The right of establishment also differs significantly. EU nationals can establish an EU carrier in any EU country, while in ASEAN, the designation of “Community carrier” is subject to the acceptance of the member state receiving the application of a designated airline (Lykotrafiti, 2015).

A key value for the EU is the concept of horizontal agreements, where rights granted to one EU country can be exploited by any EU carrier. Tan (2014) points out that only Indonesia, Malaysia, Singapore and Viet Nam allow for horizontality for all EU ASAs, while Brunei, Cambodia, the Philippines and Thailand have agreed to EU designation in a more limited way, i.e. through amendments of bilateral agreements with individual EU states, and Laos and Myanmar do not allow it.

The challenge in negotiating such an agreement will be on the issue of fifth freedoms and beyond, as unlimited third and fourth freedoms seem to be pretty much a given within the current context. However, the great distances between both markets and relatively low point-to-point demand except in certain key markets could limit the impact of a fully open EU-ASEAN market. Aside from the largest markets, cities on both sides would have to be served via a hub. This lack of direct connectivity will make it challenging for EU and ASEAN carriers to wrestle some traffic away from sixth freedom carriers. Hence,

“unlimited traffic rights alone will not help the EU and ASEAN carriers fill up their flights, particularly on the thinner routes” (Tan, 2014).

Tan (2014) also raises the issue of the lack of a single ASEAN sky as another challenge, pointing to the fact that EU carriers enjoy seventh freedom rights anywhere in the EU while the converse does not apply to ASEAN countries. However, this may be a non-issue since EU network carriers have shown little or no interest in availing themselves of such rights. However, it does open up the possibility that a low-cost, leisure EU carrier could cherry-pick the most profitable ASEAN-EU routes, something an ASEAN carrier would be unable to do.

In light of this, EU and ASEAN airlines would benefit from an ASA that allows them to co-operate more closely and compete more effectively against Gulf carriers. This includes the possibility of operating joint ventures, unrestricted fifth freedom rights and liberalising ownership requirements. The three Gulf carriers now offer 52 flights a day from Southeast Asia with connections to Europe via Gulf hubs

2. RESEARCH REPORT

compared to only 28 in 2010. In comparison, the number of direct flights between Southeast Asia and Europe only increased from 57 a day in 2010 to 61 in 2015 (CAPA, 2015a).

Pursuing increased co-operation between ASEAN and EU carriers could lead to better services and lower fares for passengers and shippers and help this market grow. It can also help EU and ASEAN carriers better compete on this market against carriers based outside of those two trading blocks. To that effect, the European Commission will be seeking the authority to negotiate an ASA with ASEAN in the context of an aviation package, which will include an "Aviation Strategy" communication identifying challenges and measures for improving the competitiveness of EU aviation. This inclusion was proposed by the European Commission to the European Parliament in December 2015 and aligns with the Kuala Lumpur declaration creating the ASEAN Community single market effective 31 December 2015.

Alliances and joint ventures: Industry’s response to competition and ownership and control policies

Competition policy and aviation liberalisation

The financial gains of airline consolidation can carry the risk of social welfare losses as a result of a less competitive marketplace. In the United States, for example, domestic deregulation was followed by reductions in air fares, which pushed uncompetitive operators out of the marketplace and brought on several waves of industry consolidations, first into six large full-service network legacy carriers and now just three, plus Southwest Airlines, which is currently the largest domestic operator. Concentration of air carriers invites scrutiny by competition authorities to assess if market power results and ensure the remaining airlines do not abuse any potential monopolistic powers. Thus, throughout the merger process, the US government closely monitored how the reduction in competition affected social welfare. In parallel LCCs, such as Southwest Airlines, Jet Blue and Spirit Airlines have emerged as a viable and profitable alternative to full service carriers.

In 2001, USDOT evaluated how successful competition law enforcement had been in the wake of domestic US deregulation. In its report, “Enforcement Policy Regarding Unfair Exclusionary Conduct in the Air Transportation Industry”, DOT critically examined economic theories related to contestable markets brought forward by the Chicago School. These theories postulated that the mere threat of competition was enough to force a monopolist to keep prices at a competitive level to thwart future entry. This was thought to be inherently true in the airline industry where capital, namely aircraft, are mobile by definition and thus an airline could easily re-allocate an aircraft from a route with competitive price to one with a monopolistic (higher) price.

However, this theory ignored how airline networks were actually constructed. Since deregulation, the US domestic network has been constructed as spokes around key hubs. Competition takes place at spokes, where network carriers try to attract traffic to travel through their respective hubs, but the hubs themselves remained monopolistic. Network carriers had no interest in offering a flight from a competitor’s hub to a spoke airport; the opportunity cost would be too high, they would compete from a position of weakness, traffic would only be point-to-point and such an action would surely invite the other carrier to retaliate.

LCCs would try to penetrate these markets but the home carrier would retaliate by flooding the market with discounted capacity and multiple new frequencies, forcing the new entry out of the market. Incumbents also monopolised gates and held onto existing slots on unprofitable routes to ensure neither could be used by a new entrant. With most major markets being a hub51, small LCCs struggled to

2. RESEARCH REPORT

establish themselves against these large network carriers and would often fail. The report points to 130 US carriers that were established after deregulation and had folded by the turn of the century.

During the first two decades of deregulation, numerous law suits were brought against the network carriers by new entrants claiming to be victims of predatory pricing. However, US courts set a high standard of proof for predatory behaviour, repeatedly finding that incumbent airlines had the right to defend their market with aggressive tactics, such as undercutting prices. In 1986, the US Supreme Court stated in the Matsushita Electric Industrial Co. v. Zenith Radio Corp case that “predatory pricing is rarely tried and even more rarely successful” (USDOJ, 2008). This sentiment is understandable. Identifiable standards do exist in statute to identify predatory pricing but are very difficult to prove.

In the years following the DOT report, thinking began to evolve on this issue. While airlines could still engage in very aggressive practices to prevent entry by new players, the courts and the government closely monitored the recovery phase of price predation. Thus incumbents were still free to sell below cost to block the entry of new players, but would have to maintain these lower prices once the threat of new competition dissipated. The EU has adopted similar competition policies focused on price recovery to mitigate the monopolistic effects of predatory pricing while still enabling consumers to benefit from lower prices.

National competition authorities and policies thus have an important role to play in ensuring that sufficient competition exists within the marketplace. In their analysis of the marketplace, they endeavour to ensure that consumers enjoy low fares while the system remains economically sustainable.

The global marketplace has also experienced a significant wave of consolidations. While national ownership and control restrictions have acted to prevent mergers and acquisitions, airline alliances have enabled airlines to organise global networks and better manage capacity. The emergence of metalneutral joint ventures (where aircraft are deployed without regard to ownership) introduces a new dimension. In effect, these create an airline within the airline, or alliance within the alliance. They offer carriers many of the benefits of consolidation while avoiding the legal ramifications of changing the equity structure.

Relaxation of the nationality restrictions would facilitate airline access to foreign capital markets. It would also be expected to lead to industry rationalisation and consolidation through mergers and acquisitions. This scenario of consolidations through mergers and acquisitions has already materialised in both the US market, following the adoption of the Airline Deregulation Act (1978), and in the EU market, following the adoption of the three air liberalisation packages (1987, 1990, 1992) and the creation of a single air transport market (1997). This was not accompanied by relaxed nationality restrictions in the US, while in the EU nationality restrictions were exchanged for community restrictions. Industry consolidation and rationalisation is meant to dispense with excess capacity and enhance efficiency. The net result is a more concentrated market, with fewer airlines and fewer hubs.

This effect raises two types of concerns for governments. First, more concentration might allow airlines to exploit market power, both at the airport and route level, as discussed in the previous paragraphs. Secondly, governments may have concerns over the future of their national airlines and hub airports.

The fear that foreign takeover of ownership and control might result in changes in the airlines’ strategies and have negative economic effects nationally at least partly explains why some governments refrain from relaxing ownership and control conditions.

In a fully liberalised environment, airlines can more easily join forces and then seek opportunities to streamline their costs by consolidating their hubs. Airline network reorganisation may have severe implications for the hubs within these networks. Airlines operating hub-and-spoke networks may go

2. RESEARCH REPORT

bankrupt, close down duplicating hubs or specialise in certain markets. In all cases the implications for the affected airports and their respective regions can be substantial.

This is because hub airports provide employment and, most importantly, connectivity. Several definitions for the latter exist. ICAO defines it as “the movement of passengers, mail and cargo involving the minimum of transit points, making the trip as short as possible, with optimal user satisfaction, at the minimum price possible” (Abdennebi, 2014). For a more detailed assessment of connectivity and its measures see Airport Council International - Europe(ACIEurope, 2014), which distinguishes direct, indirect, total and hub connectivity, as well as the closely related onward connectivity. Important is that hub operations allow regions to benefit from a larger number of directly served destinations than comparable regions without a hub operation, at higher frequency and with more opportunities for sameday return and more long-haul connections.

Having a broad portfolio of direct routes resulting from hubbing activities delivers economic benefits, in particular for the business community. The direct benefits of reduced generalised travel costs for consumers “ripple” through the rest of the economy, for example in the form of agglomeration effects, an improved business climate, regional employment and inbound tourism. Controlling for the reverse causality between employment and traffic, according to Brueckner (2003) a 10% increase in passenger enplanements in a US metropolitan area leads to a 1% growth in employment in service-related industries. Bèl and Fageda (2008) find that a 10% increase in intercontinental direct routes results in a 4% growth in international headquarters in European metropolitan areas. According to a study on the economic impact of US hubs, the presence of a hub operation in a region increases high-technology employment by over 12 000 (Button et al., 1999). Using a Granger causality test, they find that hubs create employment rather than airlines selecting cities as hubs simply because they are important economic centres.

As the markets of the United States and Europe have matured, the number of international hubs has decreased. According to Redondi et al. (2012), 37 worldwide airports have been dehubbed over the period 1997-2009. These airports did not recover their original traffic within five years and moreover the dehubbing process seems to be irreversible as all their examples show that once a network carrier stopped operating hub activities at an airport, it would not be replaced by another network carrier, but by LCCs at best (Redondi et al., 2012). This leaves the airports with expensive infrastructure suited for transferring passengers while hosting mainly LCCs operating solely point-to-point flights and demanding low airport charges or network carriers connecting the airport to their own hubs.

A number of examples of this situation have surfaced over the years. In the United States, airline mergers and acquisitions have led to the dehubbing of St. Louis (American), Pittsburgh (US Airways), Cincinnati (Delta), Memphis (Delta) and Cleveland (United) to name but a few. In Europe, dehubbing of national airports mainly took place in the wake of the national carrier’s bankruptcy (Sabena, Malèv, etc.). However, the three dominant airline groups have elected to operate multi-hub systems, despite the loss of connectivity that this entails. This is a reflection of a number of factors, including being able to protect a given country market, addressing labour issues and air traffic rights in markets where an EU-wide ASA is not in place.

Antitrust considerations

When deregulation and the ensuing liberalisation began in the USA in the mid to late 1970s, the focus was on domestic services operated by domestic carriers alone. As the concept matured and expanded beyond national borders, starting with the US-Netherlands Open Skies agreement of 1992, alliances of airlines with different nationalities became a prominent feature of international aviation as a response to

2. RESEARCH REPORT

the restrictions on the flow of airline traffic in the web of inconsistent bilateral agreements, as well as restrictions on the flow of capital in the ownership and control provisions of the bilateral agreements and national laws. The rules governing ownership under ASAs, even when routes and volumes are fully unrestricted, remain subject to strict limitations. Alliances between airlines involve co-ordinating services and prices and therefore are subject to scrutiny by the national competition authorities of each airline’s home country. Approval or clearance of an alliance can involve several authorities, each applying different rules. Individual states grant antitrust immunity or clearance, conditional or not, to airline alliances based upon their own national competition laws, except within the European Union where antitrust immunity review may occur at the EU level is granted in line with EU law.

The 11th Meeting of ICAO’s Air Transport Regulation Panel (ATRP/11) concluded that ICAO should consider measures to foster co-operation between competition authorities towards regulatory convergence. The panel recommended that ICAO explore the possibility of developing a set of core principles for fair competition in international air transport. It also suggested that ICAO should consider, in preparation for ATConf/6, collecting information regarding competition policies and practices developed or applied by other organisations and that ICAO should explore the possibility of developing a set of core principles on fair competition in international air transport.

Meanwhile, national competition regulators have been working together to provide rulings that are consistent with each other and do not create a distortion that would favour one member of an alliance over another. The US-Aruba ASA of 18 September 1997 is accompanied by a memorandum stating: “Lack of regulatory convergence adds an added layer of complexity, but in practice competition rulings have avoided creating significant market distortions.”

Institutional arrangements for competition regulation differ between jurisdictions. In some jurisdictions, aviation is subject to the decisions of the general competition authority; in others the Ministry of Transport has been assigned authority to confer antitrust immunity. In many cases, the substantive standards and procedural rules being applied differ. This can result in differences in the way issues of competition are viewed but in general all competition authorities seek outcomes that improve efficiency overall and will consider a similar range of factors in making decisions on mergers and alliances.

In the United States, the Department of Transport (USDOT or DOT) has sole jurisdiction over granting antitrust immunity for airline alliances, although the US Department of Justice does provide some input. USDOT pursues a two-step analysis, taking into account impacts on both city-pairs and the airline’s network. In the first instance, it determines whether the alliance merits approval by evaluating if:

1.it is not adverse to the public interest or,

2.in the event it substantially reduces or eliminates competition,

it is necessary to meet a serious transportation need or to secure important public benefits and,

the benefits thereof cannot be secured by reasonably available alternative means having materially less anti-competitive effects.

Thereafter, if the agreement is approved, the authority examines whether immunity from antitrust laws should be granted. This is granted when:

1.the parties would not otherwise proceed with the agreement

2.the grant of ATI is necessary to benefit the public interest.

In the European Union, as of 1 May 2004, following the advent of Regulation 1/200352 and Regulation 411/200453 it is the alliance parties’ own responsibility to safeguard compliance with Article 101 of the

2. RESEARCH REPORT

Treaty on the Functioning of the European Union (TFEU). Article 101 TFEU prohibits agreements, decisions or practices which may prevent restrict or distort trade between member states, such as setting fares or capacity, unless the Commission issues an exemption, which is the EU equivalent of antitrust review.

The Directorate General for Competition (DG COMP) is entrusted with the enforcement of EU competition law and general transport policy lies with the Directorate General for Mobility and Transport (DG MOVE). This division of competences entails that DG COMP, when reviewing airline co-operation agreements, is only entitled to apply the legal tests provided for by Article 101 and 102 of the TFEU, i.e. the potential of consumer welfare being harmed.

In Japan, antitrust immunity is granted by the Ministry of Land, Infrastructure, Transport and Tourism, while in South Korea it is granted jointly by the Ministry of Land, Infrastructure and Transport and the Korean Fair Trade Commission. In Canada it is granted by the Competition Bureau, an independent law enforcement agency. In Australia, authorisation is granted by the Australian Competition and Consumer Commission.

Awarding of antitrust immunity is usually done in response to a request from national carriers or alliances containing a national carrier and determinations usually consider whether or not an open skies agreement is in place with the state to which the foreign partner airline belongs. Antitrust immunities are usually granted in the context of decisions to liberalised ASAs (OECD, 2014). In that respect,

Singapore’s case is somewhat different as it has granted immunity to alliances that do not include Singaporean carriers and its decisions do not take into account the existence of an open skies agreement. The latter part has permitted Singapore Airlines to operate with antitrust immunity with fellow Star Alliance members based in countries that do not have an open skies agreement with Singapore.

Competition authorities will have concerns about market concentration in traffic between two hubs of a same alliance. To alleviate potential harm to consumers in that market, regulatory authorities have sometimes resorted to carve-outs, where carriers are free to co-operate fully except on hub-to-hub routes. For example, the antitrust immunity granted to Lufthansa and United to co-operate on flights across the North Atlantic specifically excludes the Frankfurt to Chicago or Washington markets. These carve-outs enable benefits to be realised from the liberalisation provided for by the US-EU open skies agreement and from the co-ordination provided by the Star Alliance, while mitigating potential negative impacts. However, carve-outs have a significantly negative impact on joint ventures. USDOT considers them incompatible with joint ventures and they are usually not used or retained when a joint venture is in place. In the case of metal-neutral joint ventures, partner carriers are able to co-operate on planning, share revenues, and consolidate their flights, thereby deriving economies of density. In the first case, despite being in an alliance, each carrier will have an incentive to attract passengers to its own flights in order to obtain all the passenger revenues. In the second case, carriers are indifferent on the passenger’s choice of carrier, since revenues and cost are shared, hence the metal neutrality.

In a joint venture, partners may have a strong economic incentive to consolidate flights in some markets by using larger aircraft, which normally have lower operating costs and lower fuel burn per passengerkm, thus improving the efficiency of air transportation and presenting an opportunity to lower fares and increase profitability. Thus from a broad public policy perspective, in the case of a joint venture, regulatory authorities will need to weigh the trade-offs in efficiency to be gained from greater competition and from economies of density. Brueckner and Proost (2009) provide a mathematical demonstration of this issue above. They modelled an airline profit function and show that the effect of carve-outs for non-joint ventures is clearly positive as it prevents anti-competitive behaviour. However,

2. RESEARCH REPORT

in the case of joint ventures, when economies of traffic density are strong, the loss in consumer surplus from potentially lower fares outweighs the gains derived from a more competitive marketplace.

Airline ownership and control

The genesis of the ownership and control requirement must be sought in the International Air Services Transit Agreement (IASTA) and the International Air Transport Agreement54. These Agreements constitute ancillary accords produced by the Chicago Conference, next to the Chicago Convention. IASTA, ratified by 129 state parties, provides for the multilateral exchange of the first two technical freedoms of the air (Two Freedoms Agreement). The International Air Transport Agreement, ratified by a mere eleven states (excluding the United States), provides for the multilateral exchange of the first five freedoms of the air (Five Freedoms Agreement). Both Agreements provide that “[e]ach contracting State reserves the right to withhold or revoke a certificate or permit to an air transport enterprise of another State in any case where it is not satisfied that substantial ownership and effective control are vested in nationals of a contracting State”55.

This clause found its way into the first bilateral ASA between the United States and United Kingdom, Bermuda I56. The parties to Bermuda I, for the first time in a bilateral ASA, reserved their right, under

Article 6 of the agreement, not to allow the exercise of traffic rights by carriers in cases they were “not satisfied that substantial ownership and effective control of such carriers are vested in nationals of either

Contracting [State]”.

The multilateral nature of the Two Freedoms Agreement and the Five Freedoms Agreement implies that the term "Contracting State" refers to a multitude of states, in fact, to all parties to the agreement. In contrast, the reference to "either Contracting [State]" in Bermuda I was aimed at the United States and the United Kingdom exclusively, Bermuda I being a bilateral agreement. What is common in both cases is the discretion of the parties to block the designation of an airline, where the ownership and control criteria are not met ("[e]ach Contracting [State] reserves the right"), rather than any kind of obligation to do so.

The requirement that airlines be “substantially owned and effectively controlled by a State or its nationals” was first set out in 1944 during the Chicago Conference. It was proposed by the United States and the United Kingdom as a means to keep “enemy” states and their airlines outside the framework of

Chicago (Haanappel, 1980). From a national security perspective, “allied” and “neutral” states sought to ensure that no “enemy” state would ever be able to take control over their airlines and exploit their traffic rights. From a commercial point of view, the ownership requirement was conceived as an effective way of preserving access to valuable commercial rights for the airlines of contracting states. The proposal was opposed by several smaller states because of their dependence on foreign capital for the provision of air services (Hörstke, 2003).

Determining true ownership and control can be rather complex. On the question of ownership, if the carrier is a public company, its shares are constantly trading hands in the secondary market making it impossible for regulators to accurately and constantly monitor the nationality of the shareholder. Various solutions have been put in place to counter that, such as limiting the voting rights for share classes available to foreigners (i.e. have one class of shares available only to domestic investors with 75% of the voting rights and have another class available to all with 25% of voting rights) or having investors self-declare their nationality and the carrier monitoring that it remains within the legislated limits. In Germany, airline shares must be registered with restricted transferability, instead of anonymous bearer shares.

2. RESEARCH REPORT

Control is even more challenging to monitor as it holds some element of subjectivity as to what effective control actually is. There is no globally accepted empirical measure of control. This criterion looks at factors that confer the possibility of exercising a decisive influence on the airline, such as significant foreign minority shareholdings; whether there is one large foreign shareholding with the rest of the capital being divided into small shares; whether members of the board of directors have foreign citizenship, especially its president, etc. It is more ambiguous than the criterion of substantial ownership since it emphasises substance over form (qualitative criterion). The fact that it is rarely defined in law leaves room for varying interpretations on a case-by-case basis according to the aeropolitical interests in play on each occasion. As will be explained further in this section, EU Regulation 1008/2008 defines effective control in Article 2 (9) as exercising a direct influence on the use of an asset by an air carrier and on the carriers’ governing bodies and management. By contrast, no definition is provided for in US statutory law.

According to ICAO, some of the main reasons for the ownership and control criteria today are that they allow a state to:

refuse to authorise air services by air carriers owned or controlled by certain other states

establish a link between the air carrier using international commercial rights and the state to which these rights pertain, thereby preventing a situation of potentially non-reciprocated benefits when an air carrier from one state uses another state's rights;

implement a ‘balance of benefits’ policy in terms of the air carriers of the state involved

ensure, in certain circumstances, that national air carriers do not use the rights of a foreign state to serve their own state57.

The first three reasons stated by ICAO constitute “the external aspect” of the ownership and control issue, as opposed to the fourth reason, which constitutes “the internal aspect” of the issue. The external aspect has strong economic overtones and reflects protectionist bilateralism. The internal aspect is more

“strategic” and reflects national security and defence considerations.

The first reason (refusal to authorise air services by air carriers owned or controlled by certain other states) enables states to keep airlines registered in states with which they do not maintain good aeropolitical or other relations out of their market. It is a reflection of the meaning that was ascribed to national sovereignty in the era of nation states, that is to say, before the establishment of “communities of states”.

The second reason (prevention of unreciprocated benefits) aims at addressing the issue of free-riding. At the Chicago Conference states failed to agree on the multilateral exchange of traffic rights. As a result, market access in air transport is regulated bilaterally via ASA. The ownership and control clauses, by establishing a link between the air carrier and the designating state, prevent non-party states from indirectly gaining a free ride through their carriers.

The third reason (balance of benefits) has a strong protectionist colouring. States exchange mostly equivalent commercial rights in order to receive reciprocal access to the respective aviation markets. By creating a link between the air carrier and the designating state, nationality clauses enable states to identify the airlines that may operate and to specify the rights that may be exercised by those airlines. This allows states to implement a “balance of benefits” policy (Hörstke, 2003). By keeping airlines tethered to their home states and to the citizen-investors of those states, the nationality rule contributed to maintaining a degree of parity in the international marketplace; no single carrier could