Habilitation thesis
LOYALTY REBATES ANALYSIS IN COMPETITION LAW
Rastislav Funta
2017
Issued by
IRIS - Vydavateľstvo a tlač, s.r.o. Bratislava, in cooperation with Danubius University, Faculty of Law,
in 2017. First Edition.
© Rastislav Funta, 2017
Cover Design © IRIS - Vydavateľstvo a tlač, s.r.o.
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted without the prior written permission of the author.
ISBN 978-80-8200-004-0
3
Content ... 03
Abstract... 07
Introduction and definition of loyalty rebates ... 09
1. The prevailing legal doctrine regarding loyalty rebates in the EU and US law ... 13
1.1. Ordoliberal influence in the EU and the view from behind the Atlantic ... 13
1.2. American law and Loyalty rebates ... 15
1.3. American Law and Section 2 of the Sherman Act ... 20
2. The role and importance of economic analysis in EU and US law ... 25
2.1. Economic analysis as a means of understanding of business behavior ... 25
2.2. Economic analysis as a means of formulating competition rules ... 26
2.3. Economic analysis as a means of defining the boundaries of administrative measures ... 29
3. European Commission Guidelines on the application of Art. 102 TFEU ... 30
3.1. Price-cost tests ... 33
3.1.1. Price arrangements ... 34
3.1.2. Average total cost test ... 35
3.1.3. Average avoidable cost test ... 35
3.2. The first part of the equally effective competitor test ... 35
3.3. The second part of the equally effective competitor test ... 37
3.4. Critical assessment of the equally effective competitor test . 42 4. Efficiency requirements under Art. 102 TFEU ... 45
4.1. Overview ... 45
4.2. The meaning of Art. 102 TFEU ... 48
4.3. Efficiency according to Art. 102 TFEU ... 51
4.4. Structure of Art. 102 TFEU vs. Art. 101 (3) TFEU ... 52
4.5. Objective Reason vs. Efficiency - should they serve the same goal? ... 53
4.6. Efficiency assessment ... 55
4
4.6.1. Efficiency of profit and necessity ... 55
4.6.2. Benefit for consumers ... 56
4.6.3. Non-Exclusion of effective competition ... 56
4.7. The role of efficiency under Art. 102 TFEU ... 56
5. Assessment of loyalty rebates prior to the Intel judgment ... 57
5.1. The case law of EU Courts before issuing guidelines of the European Commission ... 57
5.1.1. Hoffmann-La-Roche ... 57
5.1.2. Michelin I ... 59
5.1.3. Michelin II ... 60
5.1.4. British Airways ... 61
5.2. Guidelines of the European Commission ... 63
5.3. Decision of the European Commission (2006) ... 67
5.3.1. Economic assessment by Tomra ... 69
5.3.2. Economic assessment by the European Commission ... 71
5.3.3. Critical assessment ... 71
5.4. Judgment of the General Court (2010) ... 74
5.4.1. The intention of closing the market vs. the intention to cause damage ... 75
5.4.2. The equally effective competitor test ... 75
5.4.3. Formal approach vs. Efficiency based approach ... 78
5.4.4. Pro-competitive vs. Anti-competitive conduct (the theory of consumer harm) ... 80
5.4.5. Serious infringement vs. very serious infringement ... 81
5.5. CJEU Judgment (2012) ... 82
5.5.1. Objective intent vs. Subjective intent ... 82
5.5.2. Market foreclosure threshold ... 83
5.5.3. Form based analysis vs. Efficiency based analysis ... 83
5.6. Perspective of further development ... 84
5.7. How American courts analyze loyalty rebates ... 85
5.7.1. American courts judgments ... 91
5.7.2. American federal court ruling (2014): loyalty rebates do not violate US antitrust laws unless they are below cost ... 95
6. General Court judgment in the case Intel ... 99
6.1. Analysis of the judgment ...101
5
6.2. Criticism of the judgment ...104 6.3. Advocate General Wahl and his opinions ...110
7. How to correctly analyze loyalty rebates? ...119 7.1. Can rebates exclude equally effective competitors
from customers of the dominant company? ...121 7.1.1. The equally effective competitor test ...121 7.1.2. The view on "all the circumstances" of the case ...125 7.1.3. Concurrent application of the "equally effective
competitor test" and the "test under all circumstances" ...126 7.2. Does the excluded customers have a significant share
on the relevant market? ...127 7.3. Are the anti-competitive effects of rebates balanced
by efficiency? ...129
8. Conclusion ...131 8.1. The correction of misconceptions about loyalty rebates ...131 8.1.1. First misconception: loyalty rebates do not represent
competition ...133 8.1.2. Second misconception: loyalty rebates are not justified in favor of competition ...133 8.1.3. Third misconception: loyalty rebates act as a tax on smaller sellers and distort competition ...134 Literature ...143
6
Abstract
The General Court of the European Union (GCEU) confirmed in mid 2014 the decision of the European Commission in which the European Commission condemned Intel Corporation for breaching Art. 102 of the Treaty of the Functioning of the European Union (TFEU) due to application of loyalty rebates with the largest European distributor of personal computers. In the following habilitation thesis I will discuss not only general topics about loyalty rebates in EU and US law, but will also analyze deficiencies of the present judgment and take the view, that it is time for the Court of Justice of the European Union (CJEU) to rethink the application of the quasi per se rule regarding the exclusive dealings and loyalty rebates in favor of the rule of reason (a rule, that allows the Court to decide on the basis of specific circumstances of the case). In my view, such an approach would be beneficial because it would create greater consistency in the case law of the CJEU on unilateral pricing behavior.
Keywords
EU, Loyalty rebates, TFEU, USA
7
Introduction and definition of loyalty rebates
Generally speaking, loyalty discounts and rebates can be defined as a reduction in price of the product concerned that the seller or supplier provides to the buyer or distributor in exchange for a substitute for such an exclusive relationship. As a result, the loyalty program is structured to provide significant benefits to the customer if it maintains or increases its purchasing costs towards particular suppliers and, on the other hand, imposes sanctions on customers if they turn their purchases towards competitors.
Because in the normal course of business, the supplier does not provide the customer with a price reduction, if only a limited part of its purchase needs are realized from the competition, both types of loyalty structures can have an effect similar to exclusive trading agreements which as we know know force the customer to realize all or a substantial part of its purchase needs from a particular supplier.
As practice shows, these practices often occure on the market.
However, loyalty discounts and rebates are not always problematic.
If competitors are able to compete under the same conditions and if customers are able to actively respond on these developments, loyalty programs are unlikely to be anti-competitive, but instead can be an instrument to promote price competition, which can ultimately increase the overall level of social well-being . On the other hand, in the case of a dominant company, such loyalty programs may cause major competition policy concerns.
Although in everyday life there is often no distinction between discounts and rebates, the main difference between them is that their acquisition is conditional on reaching a certain volume of purchases within a certain period. While the discount is seen as a strategy to increase sales volume, the rebate is a special type of discount for those customers whose volume of purchase reaches a certain level (threshold or threshold level). However, although in the case of the existence of a dominant company, it is not sufficient to examine a discount or rebate in order to properly assess their loyalty effects from the point of view of competition law.1 It is necessary to
1 O'DONOGHUE, R., PADILLA, J. (2013): The Law and Economics of Article 102 TFEU, p. 461 and follows; in regards to abuse of dominant position in EU
8
consider how the structure of discounts or rebates may change with respect to the following three features:
Firstly, depending on the threshold, it is possible to distinguish between loyalty2 and targeted discounts and rebates. In the former case, the threshold is set by the customer to be reached, defined as the percentage of the customer expenditure growth compared to the previous period (discount or rebate on expenditure growth) or based on the exclusivity obligation (discount or rebate on the basis of exclusivity). In the latter case, the threshold to be reached by the customer is based on individualized or standardized volume (e.g.
quantity discounts or rebates). This threshold is typically set to match all or a substantial portion of the customer's demand. This means that both loyalty programs tend to have essentially the same economic consequences.
Secondly, depending on the scope, it is possible to distinguish between incremental discounts and rebates applied in the future (also known as potential discounts or rebates) and retroactive discounts and rebates that are applied retroactively (also known as repeated discounts or rebates).
Thirdly, depending on the range of products, it is possible to distinguish between individual items discounts and rebates (also known as discounts or rebates for individual products) and coupled discounts and rebates (also known as discounts or rebates for more products).
Multiple (or volume) discounts or rebates are provided to buyers who purchase a certain volume of products (e.g. 10 toys for the price of 9).
Law see: SVOBODA, P. (2010): Úvod do evropského práva, p. 246 and follows; TICHÝ, L., ARNOLD, R., ZEMÁNEK, J., KRÁL, R., DUMBROVSKÝ, T. (2010): Evropské právo, p. 535 and follows.
2 Loyalty rebates are a reward for distorting market conditions in favor of a dominant company. see: MUNKOVÁ, J., SVOBODA, P., KINDL, J. (2006):
Soutěžní právo, p. 178.
9
Targeted rebates are rebates granted to customers who buy more products than the specified threshold for a certain period (reference time). The threshold is adjusted according to the customer's ability to absorb the products and is often consistent with its previous purchases. Targeted rebates may be retroactive or incremental;
standardized or individualized. Retroactive targeted rebates are those that are provided for all purchases once the customer has reached a threshold. If the threshold is, for example, 100 pieces of toys, the customer get a discount for the first 100 pieces, as well as a discount when the next threshold is exceeded. Standardized targeted rebates are those where all customers have the same threshold and receive the same rebate. Personalized targeted rebates are those where customers have different thresholds and/or different rebates.
Selective rebates are rebates granted only to specific customers or groups of customers who have been identified by the supplier as passing from the competition.
Loyalty rebates/discounts should be distinguished from the so-called best price clauses. They differ from them because the loyalty rebates (discounts) do not include the seller's commitment to keep his price at a lower competitive price. On the contrary, it should be seen that the anti-competitive loyalty rebate mechanism is the opposite: it discourages the company from price alignment with competitors.
Loyalty rebates also differ from the so-called most-favored-nation clauses where the seller commits that the agreed buyer will receive lower prices that are offered to other buyers. Unlike the so-called most-favored-nation clauses loyalty rebates (1) include the seller's commitment to keep the price difference between the agreed and non-agreed buyers, and (2) often involve the buyer's obligation to buy all or the high volume of the goods from the seller. Only when none of these factors is present, we would have the equivalent of so- called most-favored-nation clauses.
The distinction between European and American case law regarding loyalty rebates is greater than in almost any other area of international anti-competitive law. While US case law traditionally sees loyalty rebates as beneficial and supportive of competition practices, the Court of Justice of the European Union (CJEU) has repeatedly ruled that loyalty rebates are unlawful and distort
10
competition. In fact, the difference in the assessment of rebates under
§ 2 of the Sherman Act and Art. 102 TFEU may not be even greater.
In the US, the presumption of legality prevails in the case of single- product rebates unless they are predatory, because "price cuts to increase business activity are often understood as the very essence of the competition for which laws were designed to promote competition."3
In Europe, the CJEU stated that conditional rebates "do not constitute a way of doing normal competition."4 A much less tolerant European approach to the issue of loyalty rebates has a significant impact not only on European companies but also on many of their US competitors. In the area of anti-competitive law, where there is no shortage for controversial decisions, the assessment of loyalty rebates under Art. 102 TFEU is probably the most controversial area of European competition policy. In particular, based on the assumption that rebates are incompatible with the competition process, the European courts have consistently forbidden pricing programs of dominant companies (concerning loyalty rebates) when they are based on costs.
In terms of economics, rebate systems represent a form of price discrimination because they allow sellers to charge different prices for different units. Since any rebates on purchases above a certain threshold can be interpreted as a charge for purchases below a certain threshold, the existence of a rebate does not automatically mean that prices are higher in the absence of rebates. The application of loyalty rebates by large companies is commonly understood to have a negative effect, as they could exclude competing manufacturers from the market. For this reason, the aim of this work is to clarify loyalty rebates as well to guide through critical and extensive comparative analysis.
3 Concord Boat Corp. v. Brunswick Corp., 207 F.3d 1039 (8th Cir. 2000).
4 T-219/99, British Airways v. Commissionn, 2003 E.C.R. II-5917, 291.
11
5. Assessment of loyalty rebates prior to the Intel judgment
In order to assess the Intel judgment in its broader context, it is first necessary to summarize the main findings of the EU Courts' case-law about loyalty rebates (discounts) before the European Commission's guidelines have been published. We will then have a look at the opinion of the Competition Policy Advisory Group and the European Commission view in its guidelines. Furthermore, this part of the work describes several key elements of the GCEU and the CJEU judgment in Tomra,5 which were adopted following the European Commission's guidelines. It also refers to the judgments in TeliaSonera6 and Post Danmark,7 which, although not related to loyalty rebates, deal with potential exclusionary pricing practices of companies in a dominant position. The final part of this chapter describes the discussion between US lawyers on competition law, the courts case law by assessing the lawfulness of loyalty rebates.
5.1. The case law of EU courts before issuing guidelines of the European Commission
5.1.1. Hoffmann-La-Roche
Over the past thirty-five years, the EU judges have adopted a number of important judgments in regards to loyalty and its compliance with Art. 102 TFEU. The Hoffman-La Roche case concerned loyalty agreements where the company granted a rebate (discount) to its customers who have fulfilled all or most of their purchase demands from this company. The European Commission has decided about an abuse of a dominant position by granting loyalty rebates to 22 large customers. The CJEU addressed three questions:8 whether such reductions constitute abuses of a dominant position; what was the nature of the rebates granted in this case and whether they were exempted from the provisions of Art. 102 TFEU because of the so-
5 T‑155/06, Tomra Systems and others v. Commission [2010] ECR II‑4361 and C-549/10 P, Tomra Systems and others v. Commission [2012] ECR I-0000.
6 C-52/09, Konkurrentsverket v. TeliaSonera Sverige AB, [2011] E.C.R. I-527.
7 Ibit. 47. C-209/10.
8 DABBAH, M. M. (2004): EC and UK Competition Law: Commentary, Cases and Materials, p. 160.
12
called english clause9 contained in the contract. The CJEU found that what appeared at first glance as a quantitative rebate was in fact a loyalty rebate.
The CJEU has held that, when a company in a „dominant position ties purchasers-even if it does so at their request-by an obligation or promise on their part to obtain all or most of their demands exclusively from the said undertaking abuses its dominant position within the meaning of Art. 102 TFEU."10 This is also the case when a dominant company (without the existence of contracts) introduce a loyalty rebate system in the sense that the customer will satisfy all (or most of) its demands through a company in a dominant position.
The CJEU examined the fact that those rebates were „incompatible with the objective of undistorted competition within the Common Market, because they are not based on an economic transaction which justifies this burden or benefit but are designed to deprive the purchaser of or restrict his possible choices of sources of supply and to deny other producers access to the market."11
The CJEU, distinguishing between loyalty rebates and quantity rebates, stated that the former, unlike the quantity rebates, were
"designed through the grant of a financial advantage to prevent customers from obtaining their supplies from competing producers."12 In connection with loyalty rebates, the CJEU stated that their result is the application of different conditions for the same or comparable performance. This is because different customers pay different prices for the same amount of goods, depending on whether they are all made through the company in a dominant position or whether they have multiple delivery options.
9 So called english clause provides that the buyer must notify any better offer and allows him to accept such an offer if the supplier can not offer him with a comparable offer.
10 Ibit. 85. C-85/76, point. 89.
11 Ibit. 85. C-85/76, point. 90.
12 Ibit. 85. C-85/76, point. 90.
13
5.1.2. Michelin I
In the case of Michelin I, the CJEU stated that the rebate scheme which was the subject to an assessment and which was characterized by the application of sales targets did not fall within the Hoffman-La Roche differentiation method (loyalty and volume discounts) because it required from sellers to enter into exclusive sales agreements or to obtain a proportion of its supplies from Michelin.
For the purposes of determining whether Michelin committed an abuse in providing these rebates, it was necessary to "consider all the circumstances, particularly the criteria and rules for the grant of the discount, and to investigate whether, in providing an advantage not based on any economic service justifying it, the discount tends to remove or restrict the buyer's freedom to choose his sources of supply, to bar competitors from access to the market, to apply dissimilar conditions to equivalent transactions with other trading parties or to strengthen the dominant position by distorting competition."13
The company in a dominant position has a special responsibility not to interfere into an effective and undistorted competition on the EU internal market. The real objection to the scheme created by Michelin is that it required from customers to realize more purchases than in the past period, set different purchasing targets for each customer by overcoming previous purchases, have a long reference period (one year), and was not transparent. On this basis, it can be argued that, in the case of a more transparent scheme, a shorter reference period and less discriminatory purchasing objectives would unlikely breach the Art. 102 TFEU.
5.1.3. Michelin II
Several years later, the CJEU dealt with the Michelin II case. Again, it was an assessment of the company's conduct with EU competition law. In keeping with the approach of the CJEU in Michelin I, the GCEU stated that it will follow up all the relevant circumstances when investigating whether the rebates were exclusionary. In its
13 C-322/81, Nederlandsche Banden-Industrie Michelin NV v. Commission (Michelin I), [1983] ECR 3461, point. 73.
14
decision, the European Commission came to the conclusion that the rebates generated loyalty. This was because the rebate was calculated from the sellers total turnover with Michelin (so-called retroactive/
loyalty rebates) and because the reference period for rebates was one year.
The CJEU made clear that "if a discount is granted for purchases made during a reference period, the loyalty-inducing effect is less significant where the additional discount applies only to the quantities exceeding a certain threshold than where the discount applies to total turnover achieved during the reference period."14
In this case, Michelin has implemented a wide range of pricing schemes. These pricing terms included quantitative discounts, premium rebates/discounts (later replaced by rebates/discounts to reach the target), and a bonus system based on services and agreements marked as Michelin Friendship Club (provisions where larger dealers could enter into a closer relationship with Michelin).
These rebates/discounts, in addition to the quantity rebates/discounts, were based on individualized annual volumes. The European Commission has questioned the legality of these rebates by being unfair, encouraging loyalty and divide the market.
These quantitative rebates/discounts differ from all other rebates/discounts systems in a way, that they were based on standardized (rather than individualized) volume targets. Going out from the Michelin II case, loyalty rebates/discounts are abusive under the Hoffmann-La Roche rule. Quantitative rebates/discounts that have no proven economic justification may require loyalty. With regards to rebates the CJEU stated that for the „purposes of establishing an infringement of Art. 102 TFEU, it is sufficient to show that the abusive conduct of the company in a dominant position tends to restrict competition or, in other words, that the conduct is capable of having that effect."15
14 T-203/01, Manufacture française des pneumatiques Michelin v. Commission (Michelin II), [2003] ECR II-4071, point. 85.
15 Ibit. 105. T-203/01, point. 239.
15
5.1.4. British Airways
Finally, in British Airways (BA), the CJEU had to assess the compatibility of BA's agreements concluded with the British travel agency, with Art. 102 TFEU. These entitled the travel agency to a basic commission from the BA ticket sales turnover and three different financial incentive schemes: marketing contracts, global contracts and a performance reward plan. The BA judgment stemmed from the action brought by a competing airline Virgin regarding the advantages granted by BA to travel agencies. The European Commission has found that BA is in a dominant position.
BA was penalized with a fine of 6.8 million EUR for providing travel agencies with a commission on ticket sales.
These agreements included forms of financial incentives in the form of the three different schemes of financial incentives mentioned above. First, they allowed travel agencies that have exceeded the agreed annual ticket sales target to gain another benefit in the form of premium for the result. The second was a way to obtain additional commissions calculated on the BA's share of global sales. The third was based on the results achieved by the travel agencies regarding the sale of BA tickets, comparing the results of a particular calendar month with the results in the same month of the previous year. In the view of the European Commission, BA introduced a system of trade agreements and premiums, thereby abusing a dominant position. The consequence was the suppression of BA's competition from the air transport market.
BA contravened "Art. 102 TFEU by applying to air travel agents in the United Kingdom performance reward schemes that were both discriminatory against some of their beneficiaries in relation to others and had as their object and effect, without any economically justified consideration, the reward of the loyalty of those agents to BA and thereby the ousting of rival airlines both from the United Kingdom market for air travel agency services and, as a necessary consequence, from the United Kingdom air travel markets."16 BA has appealed against the GCEU judgment. During the 7 years since the original decision was issued, the European Commission started to
16 T-219/99, British Airways v. Commission, [2003] ECR II-5917, point. 299.
16
comment on the need to apply the notion based on the impact, consumer benefit analysis, and set up an innovative approach in assessing the negative effects in the case of rebates/discount schemes. The judgment of the GCEU was confirmed by the CJEU.
Referring to Michelin I the CJEU17 found that the exclusionary effect may result from discounts or bonuses for the performance. The CJEU has found that this effect is particularly strong if the dominant company holds a market share that is subsequently higher than the share of the competitors, which was the case of BA.
5.2. Guidelines of the European Commission
The case-law of the EU courts on rebates, but also other forms of unilateral conduct of companies in a dominant position, is criticized as overly formalistic and not in accordance with the knowledge of the economy. The approach based on efficiency can be found in the July 2005 report by the EAGCP18 about an economic approach to Art. 102 TFEU. This report, written by leading economists, highlighted the benefits of an economic approach focused on improving the well-being of consumers. In this regard, the key element of this report was that:
"An economic approach by the application of [Art. 102 TFEU]
indicates that the assessment of each particular case will not be carried out on the basis of the fact that a company, for example, is engaged in tied transactions, etc. but will rather be based on an assessment of the anti-competitive effects resulting from such behavior. This suggests that competition authorities will have to identify distortions of competition and assess to what extent such a negative impact on consumers will potentially be balanced by the increase in efficiency, as well as an effective justification based on an economic analysis based on the facts of each case."19
17 C-95/04 P, British Airways plc v. Commission, [2007] ECR I‑2331, point. 71.
18 Commission report on the interpretation of Art. 82 ZES titled „An economic approach to Article 82“, July 2005.
19 Ibit. 109. Commission report on the interpretation of Art. 82 ZES, p. 3.
17
Concerning rebates/discounts, the report of the European Commission's Advisory Group on Competition Policy states that "in case of rebates/discounts in the form of net quantitative rebates/discounts is more likely the question of efficiency than in case of loyalty rebates/discounts. It is hard to prove that efficiency considerations affect the rebate system. This is because the form of rebate itself is not a clear indicator in this case."20 This shows that it is time for the CJEU to pass from the per se rule of illegality which is contained in the Hoffman-La Roche case in favor of the rule of reason analysis.
In line with the philosophy of the report of the European Commission's Advisory Group on Competition Policy, former EU Commissioner Neelie Kroes stated during her speech at an international conference in September 2005 that she was "convinced that the impact on the market needs to be assessed primarily on the basis of market effects, although there may be exceptions ... [...] The enforcement of Art. 102 TFEU should focus on real competition concerns, in other words, conduct which has actual or probable restrictive effects on the market that harm consumers. We should generally welcome the rebates because they are beneficial for consumers."21
This led to the publication of a discussion paper in 2005 outlining the contours of the European Commission's new approach and adopting its guidelines three years later. The European Commission's guidelines have sought to move away from the terminology traditionally used by the European Commission and the EU courts as they relied on concepts of loyalty rebates, quantity rebates or targeted rebates towards a more neutral concept of "conditional discounts or rebates" defined as "rebates granted to customers to reward them for a particular form of purchasing behaviour."22 While the European Commission's guidelines have accepted that conditional rebates are not unusual in practice, it is noted that these rebates (if granted by a company in a dominant position) may also
20 Ibit. 109. Commission report on the interpretation of Art. 82 ZES, p. 36.
21 KROES, N. (2005): Preliminary Thoughts on Policy Review of Article 82. pp.
1-3.
22 Ibit. 52. Guidance on the Commission's enforcement priorities, point. 37.
18
lead to actual or potential market closures similar to the exclusive purchasing obligation.
The European Commission's guidelines highlight a number of factors that are considered as important in the assessment if a certain system of conditional rebates can lead to anti-competitive market foreclosure. This means that an anti-competitive foreclosure is likely to occur when competitors of a company in a dominant position are not able to compete under the same conditions for the entire demand of an individual customer.23 The reason is that a conditional rebate granted by a company in a dominant position may allow it to use the incontestable share of the demand of each customer as an advantage in order to reduce the price paid for the contestable share of the demand (the volume for which the customer is able to find an alternative).
When departing from the formal approach followed by EU courts, the European Commission's guidelines states that the European Commission intends to examine "to what extent the data are available and reliable, whether the rebate system is capable of hindering expansion or entry even by competitors that are equally efficient by making it more difficult for them to supply part of the demands of individual customers. The Commission will estimate what price a competitor would have to offer in order to compensate the customer for the loss of the conditional rebate if the latter would switch part of its demand (‘the relevant range’) away from the dominant undertaking. The effective price that the competitor will have to match is not the average price of the dominant undertaking, but the normal (list) price less the rebate the customer loses by switching, calculated over the relevant range of sales and in the relevant period of time. The Commission will take into account the margin of error that may be caused by the uncertainties inherent in this kind of analysis."24
23 Ibit. 52. Guidance on the Commission's enforcement priorities, point. 39.
24 Ibit. 52. Guidance on the Commission's enforcement priorities, point. 41.
19
The European Commission's guidelines state that for incremental rebates, such an appropriate range is the incremental purchase (purchase made above the threshold established by the company in a dominant position for the grant of a rebate). In case of retroactive rebates, such an appropriate range is determined on the basis of an assessment of how much customer's purchasing demands can be realized from the competitor (contestable share of the demand). The methodology of the European Commission's guidelines focuses on the total conditional rebate that is provided by a company in a dominant position vis a vis the customer's contestable share of demand and assesses whether the effective price is above or below the costs of the company in a dominant position (that is why it can be referred to as a discount or rebate test).25 As it will be mentioned below, this methodology was used by the European Commission in the Intel judgment.
European Commission guidelines distinguish between 3 scenarios.
First, when the effective price (the list price minus rebate) will remain permanently above LRAIC's of a company in a dominant position. This should allow an equally efficient competitor to compete on the market with profit regardless of the rebate. On the other hand, in cases where the effective price is below AAC, the rebate system is capable of pushing as efficient competitors out of the market. Finally, if the effective price is between AAC and LRAIC, the European Commission will examine whether other factors refer to the possibility of excluding an equally effective competitor. In this case, the European Commission will examine whether and to what extent competitors have realistic and effective counter strategies.
The European Commission's guidelines state that the European Commission will be willing to consider the claims of companies in a dominant position that rebate systems create benefits that are being passed on to customers.26 In regards to other effects, the European Commission's guidelines state that, in general, incremental rebates will more likely encourage resellers to produce and sell a higher
25 Ibit. 52. Guidance on the Commission's enforcement priorities, point. 44.
26 Ibit. 52. Guidance on the Commission's enforcement priorities, point. 24.
20
volume than is the case of rectroactive rebates but without adequate explanation why this happen.
5.3. Decision of the European Commission (2006)
In 2006, the European Commission imposed a fine of 24 million € on the transnational company Tomra for violating the EU's antitrust rules in relation to the abuse of a dominant position (Art. 102 TFEU) when excluding competitors from the market for empty beverage packaging machines, usually installed in stores and supermarkets, to facilitate the collection of empty and used beverage package for recycling purposes. The violations caused by Tomra and identified by the European Commission, following a complaint lodged by the German Prokent, consisted of a system of commercial contracts, including practices known as exclusivity agreements and individual retroactive rebates, applied to the sale of these devices for large-scale retail chains operating on 5 national markets (Austria, Germany, the Netherlands, Norway, Sweden) where the Norwegian group was active through its local subsidiaries.27
The European Commission found that, throughout the duration of the infringement of EU law (1998-2002) and in countries under investigation, the average market share of Tomra was about 80% and the practices involved 40% of the total demand. The European Commission's investigation considered the fact, that Tomra's unilateral conduct prevented, respectively at least made it difficult for new entrants to enter the market, although in some cases competing suppliers were totally eliminated through acquisitions or insolvency. In 2006, Tomra filed an appeal with the GCEU claiming that the European Commission's decision was void. In its appeal, Tomra complained that the decision was based on facts that could not satisfactorily demonstrate the exclusionary intention of such practices. The GCEU has in essence confirmed the findings made by the European Commission.
27 European Commission Decision of 29 March 2006 relating to a proceeding under Article 82 [EC] and Article 54 of the EEA Agreement (COMP/E-1/38.113 – Prokent/Tomra).
21
The importance of the Tomra case is also because it repreesents the first proceeding where the European institutions have had to deal with the assessment of the impact of competition practices described in the Commission's discussion paper and included in its guidelines.
However, in the light of the critical assessment of the decision issued by the European Commission and the decisions of the GCEU and the CJEU, it can be stated that the test designed to verify the existence of exclusionary behavior was only partially implemented.
In an article published in the European Commission Policy Newsletter28 the members of the team working on the Tomra case provided a clearer formulation and have reformulated the economic analysis developed by the European Commission. The joint analysis of the documents will result in an assessment of the approach based on assessing the impact of competition practices on the rebates. The following sections will provide an insight into the economic assessment of the case that has been implemented by the European Commission, while in the part that follows this one the emphasis will be put on the legal assessment confirmed by the GCEU.
5.3.1. Economic assessment by Tomra
The report submitted by Tomra envisaged the following scenario.
The company in a dominant position sells to the customer a product unit for a catalog price of 1 €. In addition, the customer's total demand is equal to 120 units and the company in a dominant position will provide a 10 % rebate/discount if the customer reaches a 100- point threshold. In addition, an extreme situation is assumed when the suction effect produced by a retroactive rebate is at maximum, corresponding to the last unit before the threshold (i.e. the amount of units purchased by the customer is 99). For this reason, if the customer increases the number of units from 99 to 100, he will pay a negative price of 9 € from the last 100th unit and the total revenue for the company drops from 99 € to 90 €.
Therefore, if we only calculate the contestable share of the demand that equals to 21 out of 120 units, the competing company would be
28 RIGAUD, F. P., VAIGAUSKAITE, D. (2006): Prokent/Tomra, a textbook case? Abuse of dominance under perfect information, pp. 19-24.
22
obliged to offer the customer an effective cost of 0.43 € (when applying the formula above: 1 [1 - (10% / 17.5%)] equivalent to the share of the contestable share of the demand, 21/120). An equally effective price 0.43 € can be calculated as a difference between the total sum the customer would have paid if he would cover his total demand from the dominant company (120 x 0.90 € = 108 €) and the total sum the customer would have paid if he only cover the incontestable share of the demand and the conversion of the contestable part on the competing company (120 - 21 x 1 € = 99 €), all divided by the contestable share of the demand (9/21 = 0.43 €).
The report indicates that the average total cost of Tomra's production is less than 0.43 €).
A competing company must provide an effective price to the customer in the event of loss of rebates/discount granted by a company in a dominant position. In the event that a competing company only competes for a number of units above the threshold, e.g. 20 units, the customer may still purchase 100 units and benefit from the rebate granted by the company in a dominant position. This means that the price that a competing company has to match equals to 0.90 €. On the other hand, if a competing company is able to compete for the whole purchase needs of the customer, 120 units, it results to the fact that a rebate/discount granted by a company in a dominant position loses its loyalty effect and leads to a reduction in prices. Here again, the price with which a competing company must match to distract all customers from the dominant company amounts to 0.90 €.29
In regards to these two cases, a competing company must offer the customer an effective price below 0.90 €. The more units the competitor sells, the higher the price it can charge. The claims made by Tomra were essentially based on an analysis of the average price which a competing company must offer in order to compete with the rebates/discounts granted by the company in a dominant position.
The ultimate aim of the arguments put forward by Tomra was to
29 RBB Economics (2007): Tomra: Rolling Back Form-Based Analysis of Rebates?, pp. 1-4.
23
show that an equally efficient competitor would be able to compete with profit, to cover the cost of production even in the worst case, when it sells only the necessary quantity of units that prevent a customer from benefiting from a rebate/discount and which forces the competing company to offer the lowest possible effective price to compensate the customer for the lost of rebate/discount.
5.3.2. Economic assessment by the European Commission
The European Commission rejects the analysis proposed by Tomra because the behavior of a competing company would not be linked to maximizing profits (although it is important to emphasize that it would achieve a positive profit if it could sell 21 units and thus obtain a rebate granted by a company in a dominant position). In fact, it would be more rational for a competing company to give up the last unit and sell exclusively incremental units above the threshold (20 units instead of 21). It would be limited only to offset the discounted price of the company in a dominant position (0.90 €) but not to offer a much lower effective price that would compensate the customer for the lost of the rebate/discount (0.43 €).
As a result, the competing company would achieve total sales of 18 € (total sales of units above the treshold x discounted price, i.e. 20 x 0.90 €), instead of 9 € (total unit sales above the threshold + last unit before the threshold x effective price i.e. 21 x 0.43 €). This would lead to the fact that a company in a dominant position would refrain from granting a rebate, which would result in an increase in the income gap between the two competitors.30
5.3.3. Critical assessment
The decision issued by the European Commission in the Tomra case is largely based on the observation that the rebate granted by a dominant company would potentially force a competing company to adopt irrational behavior in the sense that it would bear an unnecessary loss and renounce higher levels of profit. However, the approach adopted by the European Commission seems to focus exclusively on the last unit before the threshold.
30 Ibit. 118. European Commission Decision of 29 March 2006, points. 388-389.
24
As noted above in regards to economic analysis of loyalty rebates, the incremental price to be paid indirectly by the customer for the marginal units required to reach the threshold is in most cases negative. The European Commission's guidelines in this respect state that "the potential foreclosing effect of retroactive rebates is in principle strongest on the last purchased unit of the product before the threshold is exceeded. However, what is in the Commission's view relevant for an assessment of the loyalty enhancing effect of a rebate is not simply the effect on competition to provide the last individual unit, but the foreclosing effect of the rebate system on (actual or potential) competitors of the dominant supplier."31
A retroactive rebate causes anticompetitive foreclosure only if the dominant company is confronted with a rigid part of demand that prevents a competing company from achieving its minimal effective output (the contestable share of the customer's demand forces the competing company to bid below cost). In determining the anti- competitive effect of rebates, it is therefore extremely important to demonstrate the presence of a certain sales base, which however may not exist simply because there is a dominant company on the relevant market.
As a consequence, two critical considerations are taken into account when speaking about the example referred to by Tomra and rejected by the European Commission. First, for the purpose of our critical assessment, let's assume that the average total cost per unit is equal to 0.30 € if the same sales base is at least 86 units. A rebate/discount would force the competing company to offer a price below the cost of production (we apply the formula 1 above: 1 [1 - (10% / 14%)] = 0.29 €, while the contestable share of the demand is equal to 14/100).
It should be noted that in our analysis we only count the contestable units below the threshold (14 units) and not, as claimed by Tomra, also units above the threshold (20 units, while the total customer demand is being equal to 120 units and the threshold established by the company in a dominant position is equal to 100 units, for a total
31 Ibit. 52. Guidance on the Commission's enforcement priorities, point. 40.
25
of 34 contestable units, where a competing company would be able to get retroactive rebates).
Secondly, taking into account a more realistic scenario when an equally efficient competitor could compete for at least 32 units, the result would be a higher level of profit for a competing company if it offset the rebates/discounts of the company in a dominant position than in the case of sales of units above the threshold level. In fact, a competing company would have a total profit of 12.4 € (all units below the threshold x effective price, 32 x 0.69 €, all minus relative costs, 32 x 0.30 €; when applying formula 1, the effective price is equal to: 1 [1 - (10% / 32%)] = 0.69 €, while the contestable share of the demand is equal to 32/100).
It is important to mention that, in view of the above conservative scenario we have only counted the contestable units (the dominant company and the competing company will be able to compete for units above the threshold). In case of a more realistic scenario, a competing company, through the margin generated by the expansion of production, would have the opportunity to compensate the loss of rebate/discount. Even in the extreme scenario where the dominant undertaking sets a threshold value on the level of the customer's demands, a competing company could achieve a rebate/discount granted by a company in a dominant position. If the average total cost is less than 0.43 €, a competing company (selling 21 units) would make a profit, despite the fact that the cost of the first units would be negative.
In conclusion, the main problem in the economic consideration developed in order to reject Tomra's argument is to generalize and overstate the results so that it can be said that the rebate/discount system provided by a dominant company can deter existing or potential competitors. However, if the price for which a competing company has to compete is counted only for a very limited part of the contestable share of the demand, it would be risky to protect inefficient companies from healthy price competition.
26
5.4. Judgment of the General Court (2010)32
In the Tomra case the GCEU, by not tapping to apply the economic analysis of the unilateral conduct of a dominant company, confirmed the settled case-law in the light of the unnecessary assessment of the actual effects on the relevant market through its alleged abuse.
Overall, in line with the positions adopted by the European Commission in relation to the key points of this case, the GCEU based its argument on the strategies implemented by the Norwegian group in order to exclude its competitors. In the light of these arguments, the GCEU did not carry out a thorough examination of the economic assessment carried out by the European Commission in order to establish the unlawful nature of the practices attributed to Tomra.
5.4.1. The intention of closing the market vs. the intention to cause damage
The GCEU judgment begins with the following argument: "A rebate system which has a foreclosure effect on the market will be regarded as contrary to Art. 102 TFEU EC if it is applied by an undertaking in a dominant position."33 In order to assess whether the rebate scheme is to be regarded as abusive, it is necessary to ascertain whether it is capable or intends to restrict the level of competition on the market when assessing all the circumstances.34 Therefore, even if it appears that the GCEU and the European Commission recognize the rule of reason than the per se rule, in regards to the need to assess all the circumstances and the context in which it is used, its decision does not explain the elements which have to be taken into account when assessing all the circumstances of the case. The GCEU does not provide any further guidance in this respect. The GCEU states that
32 Ibit. 96. T-155/06.
33 Ibit. 96. T-155/06. point. 211.
34 Ibit. 96. T-155/06. point. 215.
27
the European Commission has not made its decision against Tomra based on its internal documentation.35
5.4.2. The equally effective competitor test
In its appeal Tomra argued that the European Commission, in order to demonstrate the existence of anti-competitive market foreclosure, has wrongly focused on the content of the agreements rather than on the market as such.36 As a result, it refers to the economic argument that the coverage of the contracts was not sufficient to have a negative effect on an equally efficient competitor, demonstrating that only a limited part of the market was affected, while the other part remains open (on average 61% in case of five national markets).
In addition, Tomra emphasizes the fact that the European Commission, opposite to what it advocates in its discussion paper, did not find out whether the market situation was such where one or more competitors would have the opportunity to compete effectively with profit nor did the European Commission estimate the contestable share of the customer's demands, nor did it carry out a price-cost test to empirically prove the alleged exclusionary practices and harm to consumers. On the other hand, the European Commission counted only the incremental price that a competing company would have to offer to match the retroactive rebate offered by the company in a dominant position. In Tomra's opinion, the European Commission decision has not shown the ability of retroactive rebates to force competitors to set the price below cost.37
Despite the fact that Tomra's proposal was based on solid justification in accordance with the provisions of the European Commission's guidelines, the GCEU completely rejected it claiming that the closure of a substantial part of the relevant market can not be justified by the fact that the size of the market in question is sufficient to allow other companies to compete. In this regard, the three main points of the GCEU arguments are:
35 Ibit. 52. Guidance on the Commission's enforcement priorities. points. 39 and 40.
36 Ibit. 96. T-155/06. point. 200.
37 Ibit. 96. T-155/06. points. 247-249.
28
Customers on the excluded part of the market should be able to benefit from any competition on the market; competitors should be able to compete on a performance-driven basis across the market and not just on a certain part of the market;38 it is difficult to agree with the claimant's allegations that a competitor can offset lower prices.
Due to the presence of the retroactive rebate, if the customer demand is higher than the threshold, all units below the treshold are excluded.
Similarly, if customer demand is below the threshold, then the entire customer demand is excluded.39
However, it is important to remember that these three ways of thinking, which were mentioned above, are showing many mistakes.
In the first point, the GCEU did not focus on the closure of the market but on the exclusion. Therefore, it may be assumed that the exclusion of the customer itself constitutes an exclusionary action. In regards to the second point, it seems that any retroactive rebate involves an abusive foreclosure of the market, although it should only be assumed if it is allocated to the total demand. Despite the fact that the loyalty program creates a market foreclosure, this does not mean that it is anti-competitive and that it can restrict the entry of competing companies in the market. Concerning the third point, which seems to be the most problematic, we could assume that given the existence of the retroactive rebate and the fact that the price provided by the company in the dominant position for the incremental units will be very low, the compensatory price of the competing company will not be attractive.
The European Commission's guidelines explicitly state that, in order to properly assess the negative nature of the rebates, it is not sufficient to focus exclusively on the incremental units above the threshold, but rather to make a full assessment of the rebate as such.40 To this end, it is necessary to calculate the actual price which
38 Ibit. 96. T-155/06. point. 241.
39 Ibit. 96. T-155/06. points. 365-377.
40 Ibit. 52. Guidance on the Commission's enforcement priorities. point. 40.
29
a competing company would have to provide in order to match the rebate scheme of the dominant company.41
Although the average price that a competing company has to offer to a customer is below the discounted price proposed by a company in a dominant position it does not mean that this competing company should not be profitable using the contestable share of the customer's demand. Therefore, the assumption on which the third point is based can not be justified from an economic point of view. Only after application of the price-cost test that was used by the European Commission in the Intel case we may come to a similar conclusion (i.e. that the effective price that a competing company would have to provide is lower than the cost).42 Consequently, this part of the analysis (because of the contradiction with the new approach set out in the European Commission guidelines) is considered to be the least satisfactory.
In addition, the GCEU has provided another argument in its judgment based on which it may appear that one of the main objectives of Art. 102 TFEU is the protection of competitors in parallel with consumer protection.43 As a consequence, according to the GCEU, the protection of competitors or customer protection is a sufficient reason to establish the presence of abuse, while the European Commission in its guidelines confirms also the need to identify anti-competitive foreclosure leading to consumer harm.
5.4.3. Formal approach vs. Efficiency based approach
In the European Commission decision it should be recognized that the economic analysis of actual effects is not fully missing. Indeed, the European Commission analysis appears to be narrowly linked to the comparison of the contestable shares and the market share of the Norwegian group as well as on the fact that the prices, due to the rebate provided by Tomra, did not decrease, but have increased. In its proposal, Tomra challenged these findings and pointed out that the graph illustrating the suction effect contained mathematical
41 Ibit. 52. Guidance on the Commission's enforcement priorities. point 41.
42 See Chapter 6: General Court judgment in the case Intel.
43 Ibit. 96. T-155/06. point. 206.
30
errors; most of the results provided by the European Commission were inconsistent with empirical evidence; prices did not increase and were not negative.
The GCEU explicitly acknowledges that the European Commission has not substantiated its allegations of abuse but merely complemented it by a brief examination of the effects on national markets (point 288, which is contrary to point 219, where the GCEU states that, even the settled case-law does not request it, the European Commission has also analyzed the actual effects of such practices in terms of market conditions). In the light of the settled case-law the GCEU points out, that it is not necessary to show that the abusive conduct may effect competition on the relevant market, but it is sufficient to show that it is intended to restrict competition, or it is capable of having such an effect.44
In the view of the GCEU, the analysis carried out by the European Commission, which merely complements its findings, contains some mistakes. It seems that the GCEU appears to wish to express the fact that a bad analysis of actual effects may not be used to rebut the conclusions of the European Commission in relation to the effect of the market foreclosure by Tomra caused through retroactive rebates.
In fact, it confirmed that contrary to "the applicants contention, the fact that the retroactive rebate schemes oblige competitors to ask negative prices from the applicants’ customers who are benefiting from rebates, cannot be regarded as one of the fundamental bases of the contested decision in showing that retroactive rebate schemes are capable of having anti-competitive effects."45 It seems that the GCEU is giving more importance to loyalty effect than its ability to exclude competitors from the market, as evidenced by the fact that the GCEU rejects Tomra's proof of the existence of the actual market foreclosure and instead focuses only on the issue of the alleged restriction of competition.
44 Ibit. 96. T-155/06. points. 288-289.
45 Ibit. 96. T-155/06. point. 258.
31
5.4.4. Pro-competitive vs. Anti-competitive conduct (the theory of consumer harm)
Consequently, it seems clear that the GCEU did not fully address the issue of the objection raised by Tomra that correct and properly analysis should necessarily be based on a more detailed economic assessment, in line with the test proposed by the European Commission in its guidelines. However, the GCEU confirmed in general that a constraint of about 40% of the total demand have a result that such practices being able to reduce the number of competitors on the market. The GCEU considered also the possible assumption of negative rebate effects, taking into account the set of features typical for rebate structures, such as their individualized nature, as well as their use for the majority of customers.46
The main objection is based on the focus of such practices to reduce the number of competitors, which should in fact be regarded as an inevitable but not sufficient condition for assessing such behavior as anti-competitive. In fact, there exists the possibility of wrong definition when setting up the difference between competitive and anti-competitive practices because any change is intended to cause harm to consumers.
In the light of the critical assessment from the previous section (where it was provided that the standard predatory test would not be appropriate for loyalty rebates), it must be borne in mind that, in its judgment, the GCEU has correctly held that the rebate granted by a company in a dominant position must not be linked to profit sacrifice in order to be considered anti-competitive. According to the GCEU, Tomra has already excluded part of its main competitors by implicitly setting very low incremental prices and, on the other hand, imposed retroactive rebates on customers, thereby depriving them of the opportunity to benefit from offers proposed by other competitors.
In this respect, the main criticism is that it is true that if a company in a dominant position provides a retroactive rebate on the contestable and incontestable share of the customer's demand, the competing company must match only with the contestable share in question. In
46 Ibit. 96. T-155/06. points. 260-261.
32
any case, it would not mean that a company in a dominant position has an absolute advantage over the contestable share in question, where both companies have set the same effective price and that the resulting market foreclosure does not result in a profit-sacrifice. The main problem is the impossibility of an equally effective competitor to achieve a minimal effective scale of production due to the presence of a monopolized portion of demand that can result in anti- competitive market foreclosures, even if the company in a dominant position does not bears profit-sacrifice.
5.4.5. Serious infringement vs. very serious infringement
Although the amount of the fine is relatively low compared to the fines imposed in other cases, if the amount is calculated in proportion to the turnover of the company, it is the highest sanction imposed in case of abuse of a dominant position in the EU. Of particular importance is the fact that the company, which ranks 57th in size and the company's field of activity, the European Commission decided to impose a fine equal to 8% of Tomra's turnover for such a serious infringement, while in earlier cases it decided to impose for very serious infringements a fine of only 1.5% of Microsoft's turnover.47
Although the GCEU recalled that fines imposed in other cases can not be considered as a source of information having binding effect,48 it can be argued that, following the criticism of the formal approach previously used in loyalty rebate and discount cases, the Tomra case was the main intention for the European Commission to create a classic case. The aim was therefore to verify what was previously proposed by the academic community, then processed in a discussion paper and finally adopted in the European Commission's guidelines (although its application was not entirely satisfactory due to the lack of access to the analytical approach).
47 Ibit. 96. T-155/06. point. 305.
48 Ibit. 96. T-155/06. point. 314.
33
5.5. CJEU Judgment (2012)49
In 2012, the CJEU confirmed the administrative sanction against Tomra, rejecting the appeal filed by the Norwegian transnational company against the judgment delivered by the GCEU in 2010, which in turn confirmed the 2006 European Commission's decision.
For this reason, thanks to the third and final instance proceedings before the CJEU, the dispute initiated a decade ago have finally reached the final stage.
5.5.1. Objective intent vs. Subjective intent
In its appeal, Tomra challenged the fact that the European Commission has failed to demonstrate the anti-competitive intention of the disputed practices, nor did it take into account the draft evidence submitted by the Scandinavian group in relation to its intention of healthy competition in the market. The CJEU recalls that the notion of abuse of a dominant position has an objective content.50 In fact, the European Commission did not base its decision solely on the evidence relating to Tomra's subjective intent. However, it is legitimate for the European Commission to refer to subjective factors, namely the motives on which the business strategy is based.
This means that the presence of an anti-competitive intention represent one of many factors identified for the purpose of demonstrating the existence of abuse.51 Therefore, according to the CJEU,52 the European Commission was not obliged to verify the anti-competitive intention but rather to prove, in the light of all relevant circumstances, that Tomra's conduct had the capacity to create restriction on competitors. According to the CJEU, if the conduct of a dominant company is capable of restricting competition, it is not necessary to analyze the concrete effects of such practice.
49 Ibit. 96. C-549/10 P.
50 Ibit. 96. C-549/10 P. point. 17.
51 Ibit. 96. C-549/10 P. point. 20.
52 Ibit. 96. C-549/10 P. point. 21.
