Competition Policy Newsletter

Consolidation of the EU airline industry: MERGER CONTROL How the Commission kept seatbelts fastened in the 2009 airline merger wave

Lucia Bonova, Dagmara Koska, Axel Specker (1)

1. Introduction 1 tering the route as an operating carrier. The Com- mission therefore concluded that the transaction The year 2009 was marked by a wave of mergers would not significantly impede effective competition in the European airline industry, leading the Com- on any of these routes. The same conclusion was mission to review a number of cases, namely / reached for all other routes on which the parties’ Vueling/, /bmi, Lufthansa/SNAH activities overlapped — one reason being that the () and Lufthansa/. combined entity would continue to face sufficient In all but the Lufthansa/bmi case, the Commission competition on these routes. The transaction was identified competition concerns with respect to a thus approved in phase I on 14 May 2009. number of short-haul routes. To address these con- cerns, the parties submitted commitments in phase As regards case M.5335 — Lufthansa/SNAH (Brus- I (Iberia/Vueling/Clickair) and phase II (Lufthansa/ sels Airlines) ( 4), which was conditionally approved in SNAH (Brussels Airlines) and Lufthansa/Austrian Air- phase II on 22 June 2009, the in-depth investigation lines). confirmed that the transaction would have raised competition concerns with respect to four routes: In case M.5364 — Iberia/Vueling/Clickair ( 2), on 9 Brussels-Munich, Brussels-Hamburg, Brussels- January 2009, after a phase I investigation, the Com- Frankfurt, and Brussels-Zurich. On the Brussels- mission approved the proposed acquisition of the Frankfurt route, the Commission found that the two Spanish low-cost airlines, Vueling and Clickair, merger would lead to a monopoly for time-sensitive by the Spanish flag-carrier Iberia. The Commission’s passengers. It would also eliminate close competi- investigation found that the merger as initially no- tion between Lufthansa and Brussels Airlines for tified would have restricted competition in the air non-time-sensitive passengers, since the competitive transport of passengers or even led to a monopoly pressure exercised by the train service on this route on seven European routes (namely -Ven- was found not to be sufficient to compensate for the ice, Barcelona-, Barcelona-Nice, Barcelona- loss of competition from an air carrier. As concerns Athens, Madrid-Venice, Madrid-Naples and Ibiza- the Brussels-Munich and Brussels-Hamburg routes, ) as well as on twelve Spanish domestic routes the merger would lead to a monopoly. On Brussels- (Barcelona-Málaga, Barcelona-Santiago de Compos- Zurich, where Lufthansa’s subsidiary Swiss is the tela, Barcelona-, Barcelona-Granada, Barcelo- operating carrier and Brussels Airlines is Swiss’s na-Oviedo, Bilbao-Málaga, Bilbao-Ibiza, Bilbao-Se- code-share partner, the Commission’s investigation ville, Ibiza-Seville, Ibiza-Valencia, -Ibiza and led to the conclusion that the merger would elimi- Seville-Valencia). nate the significant likelihood of Brussels Airlines

3 entering the route with own services, and potential In case M.5403 — Lufthansa/bmi ( ), the Commis- competition between the parties would thus be lost. sion examined the proposed acquisition of UK The transaction would, therefore, significantly im- carrier British Midland (bmi) by Germany’s largest pede effective competition on the Brussels-Zurich airline Lufthansa in terms of its impact on the air route. transport of passengers on a number of short- and long-haul routes. On many of these routes, the par- On 28 August 2009, the Commission conditionally ties were already cooperating with each other pre- approved the notified transaction in case M.5440 — transaction by way of code-sharing, i.e. one party Lufthansa/Austrian Airlines ( 5), after the phase II in- was actually operating the route while the other par- vestigation confirmed competition concerns on five ty merely marketed flights on this route. The Com- routes: Vienna-Frankfurt, Vienna-Munich, Vienna- mission’s investigation showed that, on routes where Stuttgart, Vienna-Cologne and Vienna-Brussels. the parties were already cooperating, the party doing The Commission found that Lufthansa’s subsidiary the marketing was unlikely to start competing by en- Germanwings and Austrian Airlines were the only competitors currently active on the Vienna-Stuttgart (1) The content of this article does not necessarily reflect the and Vienna-Cologne routes and thus the transaction official position of the European Commission. Responsi- bility for the information and views expressed lies entirely would create a monopoly on these routes. Similar- with the authors. (2) OJ C 72, 26.3.2009, p. 23. (4) OJ C 295, 4.12.2009, p. 11-13. (3) OJ C 158, 11.7.2009, p. 1. (5) OJ C16, 22.1.2010, p. 11-16.

Number 3 — 2009 53 Merger control ly, on the Vienna-Brussels route, Austrian Airlines assessing the competitive effects of mergers: the faced competition from Brussels Airlines (which Commission takes assesses all effects of a particular was then being acquired by Lufthansa). The transac- transaction, including those that may beyond the tion would therefore lead to competition concerns, impact on individual city pairs. Thus the Commis- given that the only remaining competitor, SkyEur- sion examined the effects of the Lufthansa/Austrian ope, did not exert sufficient competitive pressure on Airlines merger on a bundle of flights to Central and the parties and had already filed for insolvency. On Eastern which, individually, might not have the Vienna-Munich and Vienna-Frankfurt routes, been considered a substantial part of the common the transaction would eliminate at least potential market given the low number of passengers flying competition between the parties. Existing competi- on these routes annually. tors were found not to exercise enough competitive constraint on the parties to prevent them from rais- The Commission has traditionally assessed the im- ing prices post-merger. pact of airline mergers on two main categories of passengers, defined in the past as time-sensitive and To remove the abovementioned competition con- non time-sensitive passengers. The market investiga- cerns, the parties involved offered to transfer their tion in the abovementioned cases provided evidence slots (i.e. landing and take-off rights at specific that this distinction remains important when assess- times) at the airports where slots were found to ing the substitutability of airports (7), or substitut- present a barrier to entry. Slots are thus to be of- ability between passengers’ preferences for air and fered free of charge at Barcelona, Madrid and other train services with respect to the of service, the airports (Iberia/Vueling/Clickair), Brussels, Frankfurt, type of carriers (low cost vs. network carriers) and Munich, Hamburg and Zurich (Lufthansa/SNAH) schedules. In recent cases, and in particular in the and Vienna, Frankfurt, Munich, Stuttgart, Cologne Lufthansa/SNAH (Brussels Airlines) case, the Com- and Brussels (Lufthansa/Austrian Airlines). These slot mission conducted a detailed qualitative and quanti- remedies were complemented by ancillary remedies tative analysis before concluding that the distinction such as interlining, prorate and code sharing agree- remains relevant. ments, as well as access to the incumbents’ frequent flyer programmes. In principle, when dealing with short-haul routes, The aim of this article is to discuss the Commis- the Commission takes into consideration only direct sion’s approach to market definition and the relevant services offered on a given route. An exception to counterfactual in the airline cases. We shall also ex- this rule might be warranted if factual examination amine recent developments in the way the parties’ of the traffic shows that a substantial proportion alliance partners are treated when assessing the ef- of passengers travel indirectly. This is true notably fects of airline mergers, and the investigative tools in cases where a direct service allows for a one-day that have been applied by the Commission in these return trip. In these particular circumstances, the cases. The article also highlights the main features Commission recognized that indirect and direct of the commitments made in the most recent airline service belong to the same market. mergers. A further typical element investigated in airline cas- 2. Market definition es is the substitutability of airports at the point of in airline merger cases origin and destination. The aim is to determine the catchment area of adjacent airports and to estab- In airline cases, the typical distinction between lish whether customers consider air services offered product and geographic market definition is only from these airports as substitutes for one another. remotely relevant, as the markets are defined on The Commission analyses this substitutability on a the basis of demand and supply on routes between case-by-case basis, using qualitative and quantitative individual city pairs, i.e. taking into account bidi- tools. rectional traffic flow between point A and point B. Demand-side substitution is crucial in these cases, as Finally, the Commission investigated substitutability passengers usually do not readily substitute destina- between rail and air services on short-haul routes tions in response to significant non-transitory price where a high-speed train is operating. This issue increases. (6) All recent airline mergers were there- has arisen recently in connection with the Brussels- fore assessed on the basis of city pairs where the Frankfurt route (Lufthansa/SNAH Brussels Airlines) parties’ activities overlapped. However, the question and the Vienna-Munich route (Lufthansa/Austrian of market definition is only a starting point when Airlines).

(6) By way of exception, destination substitutability could be (7) Indeed, non time-sensitive passengers are more inclined higher for holidaymakers: see COMP/M. 5141 — KLM/ to travel from secondary airports while primary airports . are the only option for most time-sensitive passengers.

54 Number 3 — 2009 Competition Policy Newsletter

3. The counterfactual most likely alternative counterfactual scenario if the MERGER CONTROL proposed transaction did not go ahead. The Com- Under its Horizontal Merger Guidelines, when assess- mission found that, without the merger, Austrian ing the competitive effects of a merger, the Commis- Airlines — which was then being privatised — was sion compares the competitive conditions that would likely to be acquired by another airline, namely Air result from the notified merger with the conditions France-KLM. In this alternative counterfactual sce- that would have prevailed without it – which, in most nario, Austrian Airlines and Lufthansa were likely cases, means the competitive conditions at the time to terminate their joint venture and become actual of the merger. However, in some circumstances, the competitors on the relevant routes. However, in the Commission may take into account market changes end, the Commission did not need to conclude on that can reasonably be predicted. (8) the relevant counterfactual as such conclusion would not alter its assessment of the effects of the case. One of the main features of airline cases is that, quite commonly, at the time of the proposed merger there is already a network of cooperation agreements 4. Treatment of alliance partners between the parties, ranging from different types of The Commission has traditionally given its investi- code-share agreements to common bottom-line joint gations a broad scope, analysing not only overlaps ventures. Frequently, therefore, the parties have ar- between the merging parties but also overlaps be- gued that, given these agreements, the transaction tween one party and the other party’s alliance part- would not materially alter the competitive conditions ner. When assessing the affected markets, the Com- already prevailing on the relevant markets. mission used to aggregate the parties’ market shares The Commission, as a matter of principle, does not with those of the alliance partners with which a accept such an argument prima facie and is commit- merging party was closely cooperating. ted to investigating the merger-specific effects of In older merger cases, therefore, the Commission each transaction, since the merger will make con- investigated not only the routes where the parties’ tractual cooperation structural and hence perpetual. activities overlapped but also the routes where the The Commission therefore analyses, amongst other services of one of the parties overlapped with things, the level of actual competition between the those of the other party’s alliance partner. If the parties as well as the likelihood of future competi- combined market share of the respective players tion between them on the routes where they were was significant, these routes were considered as af- cooperating before the merger. By way of example, fected markets. In the latest cases discussed in this in the Lufthansa/bmi transaction where, before the article, however, the Commission decided to restrict merger, Lufthansa and bmi were code-sharing on its analysis to overlaps between the merging parties a number of routes, the Commission concluded and to consider other markets to be affected only that the parties were not likely to enter the routes if there is solid evidence that the cooperation be- where they currently code-share with each other. (9) tween one merging party and a third party (its alli- However, a similar analysis of the Lufthansa/SNAH ance partner) will be extended to the other merging (Brussels Airlines) merger produced different results party. In other words, a market is considered as af- as regards the Brussels-Zurich route. Here, the fected only if there is evidence that one party will Commission found that, if the merger did not go be integrated into the other party’s partnership with ahead, SN was the most likely entrant on this route another airline on this market, or that competition and the transaction therefore eliminated this poten- will otherwise be reduced as a result or as a foresee- tial competition. able consequence of the merger. Similarly, in the Lufthansa/Austrian Airlines case, where the parties were already involved in a profit- 5. Investigation methods applied sharing joint venture, the Commission analysed the In addition to standard investigative tools such as transaction in the light of two counterfactual sce- questionnaires sent to relevant market players (10), narios. One was what would happen if the trans- the Commission used several other tools. In par- action turned the parties’ contractual cooperation ticular, during the phase II of the Lufthansa/SNAH into a permanent structural link. The other was the (Brussels Airlines) case, a customer survey was used to gather direct evidence of passenger behaviour, (8) Guidelines on the assessment of horizontal mergers under such as the degree of substitution between Brus- the Council Regulation on the control of concentrations between undertakings, OJ C 31, 05.02.2004, p. 5-18. sels National (Zaventem) and Antwerp airports. The (9) In that case, the Commission’s investigation did not in- dicate that the relevant code-share agreements would be (10) Such as questionnaires to competing airlines and train op- contrary to Article 101 TFEU. The relevant counterfactu- erators, corporate customers, travel agents, consumer as- al was therefore a situation in which the parties cooperate sociations, airports, slot coordinators and Member States’ under the code-share agreements. civil aviation authorities.

Number 3 — 2009 55 Merger control survey also helped measure the degree of competi- At the core of the remedies is a commitment to give tive constraint exercised both between the parties new entrants (or competitors willing to expand their and on the parties by their competitors, including existing services) slots on the routes where compe- train services. The survey was run by an independ- tition concerns were identified and where slots are ent contractor, selected on the basis of a call for scarce. The underlying rationale is to lower barriers tender, and the parties had the opportunity to com- to entry in situations where there are slot constraints ment on the relevant questions before the customer at (at least) one end of the problematic route. New survey was launched. or existing competitors willing to enter or expand on such routes would normally find it difficult to In phase II of both Lufthansa/SNAH (Brussels Air- obtain slots from the slot coordinator. Slot rem- lines) and Lufthansa/Austrian Airlines, the Commission edies thus aim to lower entry barriers so that en- also carried out a thorough econometric analysis of try or expansion is more likely to occur. The slots daily pricing data for several booking classes obtained are normally offered at airports at both ends of the from the parties and some of their competitors. The route (11) for a specified number of flights, i.e. daily Commission was thus able to reconstitute monthly or weekly frequencies. The number of slots made average net fares per class to analyze, for example, available generally corresponds to the overlap be- how the entry of competitors on the parties’ routes tween the parties. This solution aims to replicate the had, in the past, affected the parties’ fares. The Com- competitive pressure that the parties exercised on mission was also able to analyze the prices charged each other before the merger. by the parties to non time-sensitive passengers (as proxied by non-flexible economy class fares) com- According to the recent commitments, on short pared with their pricing to time-sensitive passengers, haul routes, the parties have to grant slots that are who primarily purchase flexible fares. within +/- 20 minutes (12) of the time requested by As regards potential entry, the total passenger de- the applicant and that are in line with the applicant’s mand on the route in question served as a starting business model as regards, for example, turn-around point for assessing whether the route would eco- time. (13) The commitments generally do not contain nomically sustain operations by an additional op- any limitation on the number of slots to be released erating carrier. The Commission also examined the at peak hours. parties’ internal documents in order to assess their entry strategies both in general and for the particu- The new commitments simplify the slot allocation lar routes at issue. The parties’ past entries were also procedure to help new entrants plan their schedules, analyzed to determine entry patterns on the basis and introduce a fast-track dispute resolution proce- of route characteristics (e.g. total demand on the dure that can be invoked by any slot applicant/new route, passenger mix, connection to existing hubs/ entrant who has reasons to believe that the parties bases, profitability and yields etc.). Similar criteria are failing to comply with their commitments. were used to assess whether or not a competitor was likely to enter the route in question. To this end, An important element of the most recent commit- the Commission also requested internal documents ments is the prospect of obtaining grandfathering form third party carriers. rights to the slots transferred. As a rule, slots are earmarked for the route where competition con- cerns were identified. However, once these slots 6. Remedies have been used on a given route for a number of The commitments accepted in recent cases contain full consecutive IATA seasons (four as a rule (14)), a number of improvements over previous packages, the slot transfer becomes definitive and the new en- reflecting the lessons the Commission has learnt trant is able to use these slots on any other route. from past experience with airline mergers and the In the Iberia/Vueling/Clickair case, this standard way remedies were implemented in those cases. period was reduced to two full consecutive IATA Broadly speaking, the recent commitments reduce seasons for applicants who request slots to operate, the burden imposed on the new entrant by the slot application procedure. They also facilitate slot trans- (11) Unless these slots can be obtained from slot coordinators fer from the parties and give the applicant more vis- through a normal procedure. (12) This is an improvement on previous cases where slots ibility in the future operation of these slots. Fur- granted by the parties could deviate from the appli- thermore, the enhanced commitments are designed cant’s request by 30 minutes; see (for example) case to encourage competitors to enter those routes COMP/M.3280 — / KLM. where competition concerns were identified. Incen- (13) The commitments foresee that ‘arrival and departure slots tives include, for example, the prospect of obtaining shall be such as to allow for reasonable aircraft rotation taking into account the Prospective New Entrant’s business model’. grandfathering rights to slots at congested airports. (14) However, the specificities of a case may require that this Last but not least, the commitments are designed to period be shortened to two IATA seasons or pro-longed ensure effective monitoring by the Commission. to six or eight IATA seasons.

56 Number 3 — 2009 Competition Policy Newsletter

from one single airport, a significant number of city behind ancillary remedies of this kind is to ensure MERGER CONTROL pairs (15) with respect to which competition concerns that the parties’ competitors have access to connect- have been identified. Market players regarded such ing traffic, i.e. they can fill their planes with connect- grandfathering rights as an additional incentive for ing passengers and thus achieve/increase profitabil- airlines to enter these routes. This is because slots ity for the services that they add on a given route. are particularly valuable assets at some European air- The Commission market-tested each set of com- ports where there are considerable slot constraints. mitments with the customers and competitors of Moreover, grandfathering rights to slots allow for a the respective parties and with other market partici- flexible allocation of aircraft and strengthen the vi- pants. Most of the feedback it received was positive, ability of services offered by competitors. i.e. respondents felt that the commitments would Where new or existing competitors are willing to sufficiently facilitate entry or expansion on the rem- operate additional frequencies on these routes, the edy routes and would solve the competition con- commitments enable these competitors to enter into cerns raised by the proposed concentrations. Fur- special prorate and code-share agreements with the thermore, in each case, respondents to the market parties, allowing them to place their codes on the test expressed interest in taking up the remedy slots parties’ flights. In addition, new entrants may con- despite the current economic crisis. It remains to clude interlining and intermodal agreements, and be seen how these improved remedy packages will they may be granted access to the parties’ frequent work in practice: their full positive impact may not flyer programme for their passengers. The rationale be felt until after the industry has recovered.

(15) If an airline operates a significant number of city pairs, it is likely to (have) set up a base at the airport.

Number 3 — 2009 57