17 October 2008
On 27 June 2007 the European Commission (the “Commission”) decided to prohibit the proposed merger between Ryanair and Aer Lingus.1 This decision consolidates the approach that the Commission has taken for previously notified airline transactions. However, certain elements of the Commission’s decision merits mention, as identified in this article.
Introduction and Background
Many commentators, including the Commission itself, have been predicting consolidation in the airline industry for a long time.2 On the face of it, from a European perspective, that consolidation has yet to occur in earnest, although the ball is now rolling. Since the terrorist attacks of 2001, we have seen two large-scale European airline mergers, KLM/Air France in 20043 and Lufthansa/Swiss in 2005.4 Further consolidation seems likely, despite the new opportunities presented by the EU-US Open Skies Agreement.5 Recent oil price hikes have had an adverse effect on the cost of aviation fuel, and the airlines’ hedging activity in this respect will end shortly, which is likely to trigger substantial costs increases.6 Even prior to this, several relatively large European airlines are publicly willing (or obliged) to consider consolidation.7 Consolidation is already gathering pace in the United States, with the recently announced Delta/Northwest and United/US Airways mergers.
Against this backdrop, this article looks at whether the approach of the Commission to mergers in the sector has changed as a result of the Commission’s most recent EC Merger Regulation (ECMR) decision in the sector, Ryanair/Aer Lingus. Prohibition decisions in ECMR cases are rare — there have only been 20 since the ECMR came into force 18 years ago — but quickly gain notoriety. The Ryanair/Aer Lingus case obliged the Commission, in the context of a hostile takeover, to examine difficult circumstances that differed greatly from any previous airline transaction that it had scrutinized. However, the Commission’s approach cannot be characterized as different to that taken in its previous decisions.
Why Ryanair/Aer Lingus is different
To understand that Ryanair/Aer Lingus presented distinct circumstances, we need to look at some key similarities between the merging parties. These include the respective airlines’ common levels of service, their common operating models and, most importantly, their shared use of the same base airport.
Common use of a base airport. To date, the Commission’s decisions have usually concerned two airlines whose main centres of operations are based out of different airports, usually in different countries.8 In these cases, the merging networks were largely complementary, and any overlaps, while intense (because there is a congested base/hub airport at one or both ends), were limited.9 In contrast, Ryanair and Aer Lingus used the same base airport (Dublin), which resulted in a unusually large number of overlaps.
Common operating models. In addition, past decisions have usually concerned companies with distinct operating models. Most have involved “network carriers” taking over other types of air passenger transport activity.10 These include examples of:
Network carriers merging with tour operators;11
Network carriers taking over smaller regional airlines;12 and
Network carriers of uneven sizes merging.13
The business model run by both Ryanair and Aer Lingus is known as “point-to-point.” Like network carrier operations, a point-to-point venture concentrates traffic at a particular airport. In this case, each party’s point-to-point routes had Dublin airport, or the “base” airport, at one end. Unlike network carrier operations, point-to-point implies that each individual route is offered independently from the others, and connecting flights are discouraged. A point-to-point model is unlikely to cover long-haul flights, which depend upon several feeder flights within a hub and spoke operation for maximum seat utilization. The distinction between network carriers and point-to-point operators is an important one for the competitive assessment. The Commission views point-to-point operations as offering more supply-side flexibility in terms of how the fleet is used, how flights are scheduled and which destinations are offered.14 In this case, and in the absence of similar competitors, it resulted in each party being the other’s closest competitor (see below).
Perception of common levels of service. In this case, the Commission characterized the level of service offered by each company as “the typical attributes of low-frills point-to-point operators.”15
In summary, Ryanair/Air Lingus was the first time that the Commission had considered similar companies operating from the same base airport. In making its determination, the similarities between the parties were relevant to the Commission’s competitive assessment of the merger, while use of the same base airport presented particular difficulties that could not be overcome.
The market definitions in this sector are already relatively settled: the basic premise is demand-side oriented — i.e., that relevant markets will comprise routes between a given point or region of origin and a given point or region of destination.16 Ryanair/Aer Lingus does not present any new market definitions for scheduled air passenger transport mergers going forward, although several were considered and rejected during the investigation. Nonetheless, the decision does detail the substitutability of airports at many locations within the European Union, and could be said to provide more certainty, in terms of concrete results (e.g., whether all London airports are substitutable) and established method, for future transactions. One note of caution is that the airport substitutability analysis did not take into account possibilities for connecting flights, which was not relevant to this transaction but would be relevant to transactions involving a network carrier. Furthermore, the distinctions between time-sensitive and non-time sensitive passengers and business and leisure passengers appear not to be appropriate to low-cost airlines offering a limited service.
The arguments for what constitute a given point or region of origin/destination have been well-documented in previous decisions. If merging airlines run scheduled flights on a route between the same airports, then these activities fall into the same market. However, some decisions, including Ryanair/Aer Lingus, have given the Commission an opportunity to decide whether two airports serving the same city or region should be considered as the same point of origin or destination.17 In Ryanair/Aer Lingus, the Commission recognized that many destinations within the EU are served by more than one airport, and examined routes to what it termed “secondary” airports (usually smaller, more remote airports) to work out if these were substitutable with routes into primary airports.
According to the Commission, the primary concerns affecting a customer’s choice of airport are:(i) travel time to the airport; (ii) the total cost of the journey (flight plus travel to airport, plus any parking fees); (iii) flight timetable/frequency; and, to a lesser extent, (iv) the quality of service offered by the airline at each location. Therefore, in determining which Ryanair/Aer Lingusroutes were in the same market, the Commission took into account the following factors:
Catchment areas. The Commission used the 100km distance/one-hour travel time benchmark(as measured from the relevant city centre) that was suggested by the airlines.18 Thus, where both airports are within this radius, this suggests that they are each substitutable. If the percentage of leisure passengers on a given route is high, then the catchment area for the airport can be widened, due to this customer group being less sensitive as regards travel time.
Views of individual customers. The Commission conducted an independent survey of individual customers at Dublin airport. If a customer had considered flying to a different airport as an alternative, the Commission took this as indirect evidence that the two routes in question exerted a competitive pressure on one another.
Views of competitors, airports and Member State civil aviation authorities. The Commission took note of the substantiated views of these parties (and the results of any independently commissioned reports).
However, some decisions, including Ryanair/Aer Lingus, have given the Commission an opportunity to decide whether two airports serving the same city or region should be considered as the same point of origin or destination.
Marketing efforts. The Commission decided that if an airport was presented in marketing publications as a substitute route to a named city (particularly in Ryanair marketing), then the customer would be more likely to think of it as an alternative.
Transport. Also considered was the availability of transport between the secondary airport and the city centre.
Price correlation analysis. Where available, the Commission’s price correlation analysis, which analysed the impact of one party’s pricing behavior on one route on the other party’s prices for another, potentially substitutable route.
Other forms of transport. These are generally taken into account, to the extent that they are a viable alternative.19 In the case at hand, the Commission rejected their inclusion due to Ireland’s island geography. The Commission’s approach confirms what was already known: that time is of the essence. Routes from the catchment areas served by Dublin, Cork and Shannon to other cities, even those within the United Kingdom, using ferry and train/road transport, take so long that most passengers booking flights to and from Ireland would not consider these as alternatives. Contrast this situation with that under earlier decisions: British Airways/TAT,20 in which the competitive constraint presented by the forthcoming Eurostar train on routes between London and Paris was taken into account, and KLM/Air France, in which the high-speed train between Amsterdam and Paris was taken into account for the competitive assessment and had an effect on the scope of commitments (see below).21
Seats available for purchase on charter flights to the same destinations. The Commission decided that these could affect market definition and/or constitute a competitive constraint, to the extent that they are available. The extent to which these would be considered turned on the number of seats available on a given route, although the limited frequency of these flights, their seasonal nature and the conventional requirement of a minimum period of seven days between legs, led the Commission to consider them as irrelevant in this case.
Sales of package holidays. These were considered with regard to certain routes on the basis that both parties offered accommodation options via their online booking services. However, the effect of package holidays on the demand for scheduled services was judged to be limited to those passengers seeking accommodation as well as flights, and of negligible impact.
The Commission’s examination delivered some interesting results. The product markets for point-to-point operations were wider than would have been anticipated. London’s five airports are substitutable between each other. As are Brussels Zavantem and (Brussels) Charleroi, ParisCharles de Gaulle and (Paris) Beauvais, and Frankfurt-Am-Main and (Frankfurt) Hahn. The following pairs were not substitutable: Amsterdam Schipol and Eindhoven, and Dublin and Belfast.
Separate markets for “time-sensitive” and “non-time sensitive” passengers. This was relevant in KLM/Air France,22 but not in Ryanair/Aer Lingus, due to the parties’ business models:neither airline offers distinct services for time-sensitive passengers (such as unrestricted tickets, which have much greater flexibility).
Other considerations. Likewise, distinctions between: (i) business and leisure passengers; (ii) late booking passengers and others; and (iii) “price-sensitive” and “price-insensitive” customers were considered irrelevant. The airlines business models did not distinguish between business and leisure customers, late-booking passengers were not obviously offered differing prices and in the Commission’s view, it is difficult to see how the latter distinction could be used to affect market definition, not least because of the problems of defining such a customer group (the Commission wondered whether any customer could be said to be truly “price insensitive”).
The high number of overlapping routes, together with the parties operating from the same base airport, using similar business models and offering similar levels of service, all led to significant concerns in the competitive assessment. These similarities between the parties were also detrimental to the Commission’s assessment of the effect that a merger would have on routes on which there was no current overlap, because the threat of entry by the other party currently acted as a significant competitive constraint.
To make matters worse, a survey of potential new entrants found that there was no appetite for entry, due in part to Ryanair’s aggressive market tactics. Thus, Ryanair’s unusually assertive approach to the market appears to have played a role in determining barriers to entry, which impacted negatively on the Commission’s assessment. Does this imply that commitments for mergers involving more than averagely aggressive players would have to be more onerous than would otherwise be the case?
Ryanair/Aer Lingus also suggests that upfront buyers might be harder to find for slots becoming available as a result of a conditional clearance, if the airport in question is in too peripheral a location within the European Union. If this is correct, then slot remedies relating to such airports may have to be more generous in order to attract new entrants.
The competitive assessment first focused on the parties’ market shares and HHI scores, for a given route. The percentage of transfer passengers traveling on a given route is usually also taken into account, but it was not relevant in the context of this (point-to-point) analysis. The Commission found that the merger would create a monopoly on 22 routes and market shares of more than 60 percent on all other routes where there was overlap.
On those routes where a duopoly existed, the Commission noted that the parties would already be each other’s closest competitor. However, in relation to the other routes on which there was competition, the parties were found to be each other’s closest competitor by reference to a number of criteria:
- Whether the other players operated similar business models – broadly, they did not;
- The similarity of the parties’ size (including the size of fleet), business model, costs structure and market position;
- The difference in average fare between the parties – judged here to be minimal; and
- Evidence of the effect of one’s presence and pricing behavior on the other (with regard to pricing conduct, this was relatively easy to monitor due to high price transparency).
In relation to buyer power, the Commission noted that post-merger the threat to the merged entity posed by switching would be negligible because of the non-existent or marginal role played by competitors on many of the affected routes. Therefore, customers had little or no buyer power.
The Commission then looked at whether there was an actual and immediate threat of entry by a third party on any of the affected markets, and, in particular, the prospects of another airline starting operations from Dublin airport. The opinions canvassed from other airlines confirmed not only that there was no threat of entry, but that there was a widespread aversion to it. Besides the considerable financial risk posed by building a brand to rival either Aer Lingus or Ryanair in Ireland, many airlines cited the high likelihood of aggressive retaliation by Ryanair as a significant disincentive, even if entry were possible. These opinions were corroborated by an analysis of Ryanair’s aggressive behavior towards past entrants, who had since withdrawn services (e.g., Easyjet and GoFly). GoFly described the prospect of starting an airline operation out of Dublin as “commercial suicide.”23 In addition, Dublin as a potential entry point, was seen by many as unattractive because of its size, relative congestion and location on the outskirts of the European Union.
With regard to efficiencies, it was not apparent that Ryanair would pass the benefits to customers.In KLM/Air France, the Commission acknowledged that costs savings and service improvements would flow from the combined networks, benefiting consumers and the economy at large. No such network-related benefit was available here, because the parties were point-to-point operators. This would suggest that mergers of network carriers stand a greater chance of running efficiencies arguments successfully than do mergers of point-to-point players.
The Commission’s Notice on remedies states that commitments “have to eliminate the competition concerns entirely and have to be comprehensive and effective from all points of view.”24 The remedies resulting from airline mergers concentrate on sufficient surrender of landing and take-off slots at either end of an affected route to enable another player to operate an effective competing service.
The Commission’s approach to remedies in Ryanair/Aer Lingus was no different than in previous airline mergers, and underlined the general point that remedies must sufficiently address concerns. The main difficulty was that Ryanair failed to surrender sufficient slots under conditions that would make entry viable for a competitor, and which would replace the competitive pressure of Aer Lingus. The Commission concluded that Ryanair’s slot-based remedies fell way short of what was required. Additional (and somewhat novel) commitments proposed by Ryanair, and rejected by the Commission, included:
- A ten percent reduction in Aer Lingus’ fares on all short-haul routes for one year following the merger — rejected on the grounds that it would be almost impossible to monitor, and that even if it were, Ryanair offered no commitment relating to its own prices, or those of Aer Lingus after one year had passed;
- An offer to operate Ryanair and Aer Lingus separately — rejected on the basis that it did not address any identifiable concern; and
- An offer not to increase the frequency of its flights on routes where a new entry occurs for a period of six IATA seasons — also rejected on the basis that it did not address any identifiable concern.
With regard to acceptable remedies, KLM/Air France is a useful, relatively recent decision to examine. In that matter, the parties agreed to surrender slots at Amsterdam Schipol and Paris (Orly or Charles de Gaulle) and Milan Malpensa airports, to make available a sufficient number of frequencies for all problematic routes, in order to alleviate the Commission’s concerns.25 Previously, in the Commission’s assessments of airline alliances under Article 81 EC Treaty,26 undertakings relating to slots usually had a duration of six years. KLM/Air France broke new ground because the slot undertakings given were of unlimited duration, meaning that underused or misused slots would be returned to a slot coordinator rather than the airlines that previously owned them.
Slot undertakings are usually accompanied by the parties agreeing not to increase the frequency of their flights on the problematic routes, to enable a competitor to establish itself. As discussed above, KLM/Air France also resulted in the parties agreeing to enter into internodal arrangements, by which the Paris-Amsterdam high-speed train link could offer return trips by air as a means of increasing frequency, and thereby increasing its appeal to customers. Furthermore, and as in Lufthansa/Swiss, the relevant national regulatory authorities (in France and the Netherlands) agreed to give traffic rights to other carriers wishing to use Paris or Amsterdam as potential stopovers en route to non-EU destinations, including the United States, and to refrain from imposing price regulation on long-haul routes.
It could be said that Ryanair/Aer Lingus was notified at an unfortunate time for the Commission. Following several years of significant challenges in this narrow-profit-margin sector, the Commission has appeared keen to demonstrate to the industry that consolidation and merger control clearance are not mutually exclusive. The characteristics of the Ryanair/Aer Lingus overlap, arising in the context of a hostile takeover, and the aggressive stance of the potential purchaser (both in the market and in its offer of remedies to the Commission), resulted in a prohibition decision. Nonetheless, the decision itself demonstrates that the Commission has been consistent in its application of the ECMR to the airline sector, in terms of market definitions, competitive assessment and acceptable remedies. Certain minor new insights are provided, but these are complementary to previous decisions. The most interestingquestion raised by the case is whether the involvement of an aggressive player affects the outcome of a decision. Ryanair has since lodged an appeal with the Court of First Instance, but on the basis of the text of the non-confidential version of this decision, it seems unlikely that this will be successful.
1 Case No. IV/M.4439, Ryanair/Aer Lingus, Commission Decision dated 27 June 2007 (Ryanair/Aer Lingus).
2 “The long-awaited consolidation of the European airline sector can be done in full respect of the competition rules,” Mario Monti in 2004, the then Competition Commissioner, on the occasion of the conditional clearance of the Air France/KLM merger in April 2004 (quotation from Commission press release ref. IP/04/04 dated 11 April 2004).
3 Case No. IV/M.3280, KLM/Air France, Commission Decision dated 11 February 2004 (KLM/Air France).
4 Case No. IV/M.3770, Lufthansa/Swiss, Commission Decision dated 4 July 2005 (Lufthansa/Swiss).
5 This agreement came into effect on 30 March of this year, and one of its most important provisions allows European airlines to operate flights from anywhere in the EU to anywhere in the US, thereby ending the system of exclusive rights for certain airlines in relation to specific airports.
6 “European airlines facing profit squeeze,” article in the Times (UK edition) dated 25 April 2008, page 51.
7 Examples include:
• Italian flag-carrier Alitalia, which faces imminent liquidation or takeover. Alitalia has attracted interest from Air France/KLM, Russian flag-carrier Aeroflot and most recently, a consortium including Italian discount carrier, Air One; and
• UK operator BMI. According to press reports (see for example “Slot strategies” article in Aviation Week and Space Technology dated 7 April 2008), the founder of BMI seems likely to dispose of his 50% (plus one share) holding this year. The German flag carrier, Lufthansa (a 30% shareholder in BMI) has first refusal. Nordic airline SAS, the remaining BMI shareholder, is also contemplating sale of its 20% stake, also this year. Virgin Atlantic has stated publicly its interest in acquiring BMI.
8 Ryanair/Aer Lingus, para. 335.
9 Past decisions that fit into this category include the two largest: Lufthansa/Swiss (the concerns focused on two such routes: Munich - Zurich, and Frankfurt - Zurich) and KLM/Air France (routes between Amsterdam and five French cities, four Italian ones, two US ones and Lagos). mayer brown 25
10 Network carriers are airlines that operate what is known as a “hub and spoke” system of scheduled (usually full-service) flights. A network carrier’s flights (the “spokes”) feed into its “hub” in time for passengers to catch connecting flights elsewhere.
11 Such as Case No. IV/M.1354, SAIR/LTU, Commission Decision dated 21 December 1998, in which network carrier Swissair took over tour operator LTU.
12 Such as Case No. IV/M.259, British Airways/TAT, Commission Decision dated 27 November 1992, in which network carrier British Airways took a 49% stake in French regional operator TAT.
13 Such as Case No. IV/M.1855, Singapore Airlines/Virgin Atlantic, Commission Decision dated 21 December 1998.
14 Ryanair/Aer Lingus, para. 43.
15 Ryanair/Aer Lingus, para. 50 (these attributes include very high direct distribution over the internet, only one-way restricted fares, baggage fees, single economy-type cabin service and no complimentary meals on board).
16 Ryanair/Aer Lingus, para. 332. Though not considered in Ryanair/Aer Lingus, it might be useful to add here that indirect flights only appear to be relevant to market definition in the context of long-haul journeys, where they may be considered as competitive constraints to direct flights providing the connection time is no greater than 150 minutes. - Air France/KLM, para. 22.
17 See for example, Case No. IV/M.259, British Airways/TAT, Commission Decision dated 27 November 1992, para. 21; Case No. IV/M.1494, SAIR/AOM, Commission Decision dated 3 August 1999, para. 16; and Air France/KLM, para. 24.
18 While understanding that all passengers within a given catchment area would not necessarily emanate from the city centre, the Commission stressed the need to apply some form of criteria. The Commission took into account that destinations such as Tenerife, where there is no single destination but a number of resorts, needed to be looked at more comprehensively.
19 See, for example, discussion of the role of the Zurich-Frankfurt high speed train in Lufthansa/Swiss, para. 56.
20 Case No. IV/M.806, British Airways/TAT, Commission Decision dated 26 August 2996, para. 20.
21 KLM/Air France, paras. 70-72 (competitive assessment), para. 165 (internodal agreement commitment). Inter alia the Commission accepted that the slots given up by the parties at Amsterdam and Paris airports could be used for other destinations once this link was operative. Commission press release IP/04/194 dated 11 February 2004.
22 KLM/Air France, para.19.
23 Ryanair/Aer Lingus, para. 654, quoted the chief executive of GoFly: “If anyone starts an airline out of Dublin, all I would say to them is: make sure you have a lot of money. It would be commercial suicide[to go against Ryanair].”
24 Commission Notice on remedies acceptable under Council Regulation (EEC) No. 139/2004 and under Commission Regulation (EC) No. 802/2004, para. 9.
25 For example, the parties undertook to make available six frequencies per day for a competing Paris-Lyon flight service.
26 These are cooperation agreements between airlines in relation to certain routes, which fall short of the change in control required for an arrangement to be considered instead as a concentration under the ECMR. Prior to modernisation, these were often notified to the Commission They can include any of the following: code-sharing, prices and commission, joint sales and marketing, revenue and profit-sharing schemes, commonality of frequent flyer programmes, network planning and inventory/yield management).