Economics on Demand The role of efficiencies in mergers and Art. 101: Good things come to those who wait?

Economics on Demand The role of efficiencies in mergers and Art. 101: Good things come to those who wait? 15 November 2017 Urs Haegler and Sergey Khodjamirian

OUTLINE Why are efficiencies important? Economics of efficiencies European Commission s framework for assessing efficiencies European Commission cases involving efficiencies Airline JVs UPS/TNT COMPASS LEXECON 1

EFFICIENCY IS INTEGRAL TO COMPETITION POLICY Competition policy Unrestricted competitive process More efficient use of scarce resources Consumer welfare improved COMPASS LEXECON 3

EFFICIENCY DEFENCE INCREASINGLY BEING CONSIDERED No room for efficiency defence Merger test = dominance test Efficiency offence Efficiencies can increase dominance (e.g. GE/Honeywell) Efficiencies can be taken into account 2001 Green Paper 2004 Horizontal merger guidelines 2008 Non-horizontal merger guidelines e.g. TomTom/TeleAtlas Consideration of efficiencies becoming more common DB/NYSE UPS/TNT Orange/Jazztel UPP analysis 1990s 2000s 2010s EC yet to clear an otherwise anti-competitive merger explicitly on efficiency grounds COMPASS LEXECON 4

PRE-MERGER: BERTRAND DUOPOLY (HOMOGENOUS GOOD) Price of bananas Consumer surplus Price C MC Allocative efficiency D Quantity of bananas COMPASS LEXECON 6

POST-MERGER: MONOPOLY (NO EFFICIENCIES) Price of bananas Price M Price C Consumer surplus Deadweight loss / allocative inefficiency MC D MR Quantity of bananas COMPASS LEXECON 7

POST-MERGER: MONOPOLY (WITH EFFICIENCIES) Price of bananas Consumer surplus Price C Price M Cost reduction offsets price increase MC H L D MR Quantity of bananas COMPASS LEXECON 8

SOURCES OF EFFICIENCIES Supply-side Firm-level efficiencies Transaction efficiencies Mergers Restrictive agreements Demand-side Network efficiencies Complement pricing One-stop shopping Dynamic efficiencies COMPASS LEXECON 9

SUPPLY-SIDE: FIRM-LEVEL EFFICIENCIES ECONOMIES OF SCALE Achieved by expanding scale of production and spreading costs over a larger output, which leads to lower costs per unit. Dodson & Fogg Bendini, Lambert & Locke + COMPASS LEXECON 10

SUPPLY-SIDE: OTHER FIRM-LEVEL EFFICIENCIES Source of efficiency Economies of scope Description Produce different goods under same roof Example Publisher producing both hardcover and paperback editions from same manuscript Rationalisation Synergies Purchasing economies Move production from high- to low-cost facility Integrate hard-to-trade assets Obtain bulk discounts on raw materials European TV manufacturer buys Chinese factory Combining complementary patents Airline buys more airline food in bulk COMPASS LEXECON 11

SUPPLY-SIDE: TRANSACTION EFFICIENCIES When firms adopt contracts or organisational forms that minimise certain market failures that arise in the open marketplace Double mark-up Free-riding Hold-ups Particularly relevant for vertical mergers and agreements (e.g. MFNs and RPMs) COMPASS LEXECON 12

DEMAND-SIDE: NETWORK EFFICIENCIES In network industries, the larger the network, the more valuable the product can be for users More customers More apps COMPASS LEXECON 13

DYNAMIC EFFICIENCIES When firms invest and innovate to create better products or more efficient production methods R&D Learning-by-doing Entrepreneurial creativity New, cheaper and/or better products COMPASS LEXECON 14

CRITERIA FOR ACCEPTING EFFICIENCIES Horizontal merger guidelines ( HMG ) Verifiability Merger-specificity Pass-on Art. 101 (3) Efficiency gains Indispensability Fair share CRITERIA ARE CUMULATIVE COMPASS LEXECON 16

VERIFIABILITY / EFFICIENCY GAINS reasonably certain that the efficiencies are likely to materialise, and be substantial enough ( 86 EC HMG) Typical evidence: Internal documents on or modelling of efficiencies related to the merger or agreement COMPASS LEXECON 17

MERGER SPECIFICITY / INDISPENSABILITY no less anticompetitive, realistic and attainable alternatives of a nonconcentrative nature [ ] than the notified merger which preserve the claimed efficiencies. ( 85 EC HMG) Economies of scale: Why need a merger/jv? Merger/JV might be preferable: Quicker If industry is static or declining COMPASS LEXECON 18

PASS-ON / FAIR SHARE to be passed on, to a sufficient degree, to the consumer. ( 84 EC HMG) Transaction efficiencies Demand-side efficiencies Dynamic efficiencies Firm-level efficiencies Clawback? Fixed-cost reductions Variable-cost reductions? COMPASS LEXECON 19

PASS-ON OF VARIABLE COST REDUCTIONS Price of bananas Price C Price M Consumer surplus MC L Variable cost reductions normally partially passed on to consumers even under monopoly! MR D Quantity of bananas MC H Degree of pass-on may depend on: Elasticity of demand Curvature of demand (elasticity of elasticity of demand) Elasticity of supply Number of competitors COMPASS LEXECON 20

JOINT VENTURES BETWEEN MEMBERS OF ALLIANCES CONTEXT Transatlantic JV between Oneworld members American Airlines, British Airways and Iberia Commitments on 6 transatlantic routes accepted on 14 June 2010 Transatlantic JV ( A++ ) between Star Alliance members Air Canada, Continental, Lufthansa and United Airlines Commitments on 1 transatlantic route accepted on 23 May 2013 Transatlantic JV between Skyteam members Air France-KLM, Alitalia and Delta Commitments on 3 transatlantic routes accepted on 12 May 2015 COMPASS LEXECON 22

COMBINING AIRLINE NETWORKS DEN Airline 1 AMS Pre-JV, Airlines 1 and 2 compete on NYC FRA. With the JV, there is seamless interlining NYC FRA Behind/ Beyond Airline 2 Trunk route Trunk route Behind/ Beyond AMS NYC NYC FRA FRA Airline 2 Behind DEN NYC FRA NYC FRA AMS DEN NYC FRA AMS COMPASS LEXECON 23

Fare OUT-OF-MARKET EFFICIENCIES (REDUCING DOUBLE MARK-UP) NYC Airline 1 FRA Airline 2 AMS Example: NYC-AMS market Pre-JV: Airline 2 buys a NYC FRA ticket from Airline 1 Airline 2 sells a NYC AMS ticket to the passenger Results in two mark-ups With the JV: Pre-JV Mark-up (Airline 2) Marginal cost (Airline 2) Mark-up (Airline 1) Post-JV Reduction in fare Mark-up Mark-up (JV) (JV) Marginal cost (Airline 2) 2) Airlines jointly price the entire journey as a single entity. Results in a single mark-up and a fare reduction This in turn leads to higher passenger volumes (also on the trunk route) Rely on econometric analysis of O&D-specific fares data to estimate the extent of the fare reduction Marginal cost (Airline 1) Marginal cost (Airline 1) 1) Passenger volume (Pre-JV) Passenger volume (Post-JV) COMPASS LEXECON 24

ELIGIBLE BENEFICIARIES OF OUT-OF-MARKET EFFICIENCIES The extent to which out-of-market efficiencies are credited was an important point of contention In the end, the Commission required: Considerable commonality between the potentially harmed and the benefitting passengers (not substantially the same!). The benefitting passenger s itinerary must flow over the route of concern. Tracking individuals travel histories: Benefit credited? Miss Smith NYC FRA FRA AMS NYC Nov 2015 [Harm] May 2016 [Benefit] Mr Darwin NYC FRA AMS Jan 2015 [Benefit] Mrs Allen DEN MIA FRA NYC FRA Jul 2015 [Benefit] Feb 2016 [Harm] COMPASS LEXECON 25

Departure time from New York Departure time from New York SCHEDULE DELAY REDUCTIONS Schedule delay reductions occur on return trips on the trunk route (i.e., in-market ) Ideally, Mr Johnson would like to depart from Frankfurt at 7:00 AM and depart from New York at 5:30 PM next day ( X ) Pre-JV Mr Johnson normally takes the same airline to and fro Airline 1 (8:00AM and 5:30PM) is preferred based on Mr Johnson s schedule Minimum schedule delay of 60 minutes With JV Schedule combination: 30minutes Schedule coordination: 20 minutes Increased frequencies: 0 minutes 8PM 6:30AM, 7:30PM 8PM 6:30AM, 6:40AM, 7:30PM 7:30PM 8:00AM, 7:30PM 7PM 7PM 6PM 7:00AM, 5:30PM 60 mins 8:00AM, 5:30PM 5PM 6AM 7AM 8AM 9AM Departure time for the return flight from Frankfurt 6PM 7:00AM, 6:00PM 30 20 mins 0 mins 6:30AM, 6:40AM, 7:00AM, 8:00AM, 5:30PM 5:30PM 5:30PM 5PM 6AM 7AM 8AM 9AM Departure time for the return flight from Frankfurt Airline 1 Airline 2 Mr Johnson Airline 1 Airline 2 New JV options Mr Johnson COMPASS LEXECON 26

ECONOMIES OF DENSITY Economies of density in the airline industry can rise from the ability of an operator to : Carry more passengers on the same aircraft (i.e., to increase load factor) Optimise the size of aircraft deployed ( up-gauging ) Increased frequencies of service on a given route. Variable cost Slope: AVC Post Increasing the number of passengers on existing aircrafts and optimising the size of new aircrafts reduces the average variable costs per passenger. Slope: AVC Pre Reductions in the costs feed into reduced fares for passengers depending on the pass-through rate. Pre - JV JV Number of passengers COMPASS LEXECON 27

INDISPENSABILITY Is an in-depth JV necessary to generate schedule delay reductions or economies of density, i.e., is there not a realistic and less restrictive alternative agreement ( RALRA ) delivering the same efficiencies? Standard code-sharing is less restrictive but does not generate a significant reduction in double marginalisation, nor does it deliver schedule delay reductions or economies of density to the same extent. Code-sharing with two-part tariffs: Operating carrier charges a fixed fee + marginal cost for each unit purchased by the marketing carrier. Arrangement does not exist in practice Not commercially realistic, because demand and cost uncertainty exposes airlines to risk Carve-out of O&D passengers does not deliver the same level of efficiency gains Carve-out of all passengers does not deliver the same level of efficiency gains either. COMPASS LEXECON 28

OUTLINE Why are efficiencies important? The economics of efficiencies European Commission framework for assessing efficiencies European Commission cases involving efficiencies Airline JVs UPS/TNT COMPASS LEXECON 29

UPS/TNT CONTEXT The relevant market International express deliveries of small packages in the EEA. Focus on global integrators: Global integrators logistics services combining trucks and planes, which can guarantee overnight express deliveries. Extensive aircraft fleet to send packages quickly over long distances and ground networks of road vehicles and sorting centres. There are 4 operators (DHL, UPS, TNT and FedEx): DHL is leader in EEA; UPS makes a strong second; TNT is a maverick ; and FedEx is a smaller competitor. Significance of the case First merger with efficiencies crucial in eliminating competition concerns. Balancing of efficiencies with the anticompetitive effects. COMPASS LEXECON 30

UPS/TNT COMMISSION S ANALYSIS Three-step balancing analysis: 1. Price-Concentration Analysis ( PCA ) estimate likely impact of the merger on prices (before efficiencies); 2. Efficiencies estimate likely efficiencies and assess whether these efficiencies outweigh the likely price effects; and 3. Assess market conditions in countries where the efficiencies did not clearly outweigh the estimated price increases. COMPASS LEXECON 31

UPS/TNT STEP 1: LIKELY PRICE EFFECTS (BEFORE EFFICIENCIES) PCA How prices of intra-eea express services varied on number of competitors across lanes ( O&D pairs ). Econometric analysis considers following factors as potentially influencing price per kg: Cost; Distance; Market size; Customer size; and Presence of competitors. Price effects predicted for 25 EEA countries: Light red (12) price increase >5%; Light orange (13) price increase <5% [NB. No PCA in remaining 5 EEA countries (orange)] COMPASS LEXECON 32

UPS/TNT STEP 2: EFFICIENCIES Criteria for efficiencies to be credited: Cost savings in intra-europe air network and ground handling costs. Effective within 3 years after the merger. Allocated overall cost savings to individual lanes: Origin and destination within the EEA; Proportional to each lane's share of UPS total European air transport costs; Applied a pass-through rate of [60-70]% (from PCA). In green countries efficiencies outweigh price effects. In these 10 countries with likely net price reductions, the Commission found that FedEx was a significant player. These intra-eea express markets (including previously light-red Greece) were found not to be problematic. The remaining red countries (15 in total) required detailed country-level analysis. COMPASS LEXECON 33

UPS/TNT STEP 3: DETAILED COUNTRY-LEVEL ANALYSIS In 9 countries, the Commission found a clear positive net price effect, i.e., it deemed the merger to be problematic without further analysis. (DARKER RED) For the other 6 countries (FI, DK, NL, CZ, HU, SL) net price effect was found to be negligible, neutral or even negative (LIGHTER RED). Performed overall assessment of market conditions Found that other market factors pointed towards a likely SIEC. The Commission considered that in these markets FedEx would not exert a significant market pressure to counteract the price increases. COMPASS LEXECON 34

Concluding remarks COMPASS LEXECON 35

CONCLUDING REMARKS Efficiency benefits can be used to offset restrictive effects identified in merger control or in the review of agreements: These efficiencies need to be verifiable, merger-specific/indispensable, and to a significant extent passed on to consumers Variable cost reductions can be expected to be at least partly passed on in most cases Efficiencies can be of a non-cost nature (e.g., improved product quality) this is particularly relevant in network industries, or where the merger/agreement improves investment incentives. Efficiencies as part of the business rationale can positively affect authorities view of a merger or an agreement COMPASS LEXECON 36

Thank you Dr. Urs Haegler, Vice President, London Sergey Khodjamirian, Senior Economist, London COMPASS LEXECON 37