Professional Documents
Culture Documents
Damien Geradin
Tilburg University and Howrey LLP
Article 23(2) and 23(5) of Council Regulation 1/2003 of 16 December 2002 on the
implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty,
(2003) O.J. L 1/1.
(ii)
(iii)
(iv)
(v)
(vi)
(vii)
(viii)
(ix)
(x)
(xi)
3. State the number of loyalty discounts and rebate programs that your agency
found to be unlawful over the past ten years (1999 to date); include cases
resolved informally as well as those that led to a formal decision. If your
agency has found any loyalty discounts and rebate programs to be unlawful,
2
3
Kerse & Khan, EC Antitrust Procedure, (2005, Fifth Edition), para. 7-008.
Replacing Soda Ash ICI and Soda Ash Solvay (respectively [1991] OJ L152/40 and 152/21,
[1994] 4 CMLR 645 annulled for procedural reasons).
Case T-219/99, British Airways v. Commission [2003] ECR II-5917 and Case C-95/04 P,
British Airways v. Commission, 15 March 2007
Case T-203/01, Manufacture Franaise des Pneumatiques Michelin v. Commission [2003]
ECR II-04071
Case T-228/97, Irish Sugar plc v Commission [1999] ECR II-2969, confirmed on appeal by
Case C-497/99 P, Irish Sugar plc v Commission [2001] ECR I-5333.
EVALUATIVE CRITERIA
5. In your jurisdiction, is the term single-product loyalty discounts and rebates
used in a manner different from the definition in the first paragraph above?
If so, how?
In this Questionnaire, loyalty discounts and rebates are defined as discounts or
rebates on units purchased of a single product, conditioned upon the level or share
of purchases.
Under the current case law of the Community Courts, the definition of loyalty
discounts and rebates is different from the definition given in this Questionnaire.
8
9
10
11
12
13
14
15
16
Ibid. 84.
See Case T-219/99, British Airways v. Commission [2003] ECR II-5917, 248.
See Case T-203/01, Manufacture franaise des pneumatiques Michelin v. Commission, [2003]
ECR II-4071 (Michelin II) at 239.
(ii)
rebate
is
objectively
justified
by
economic
Finally, it should be noted that when assessing the legality of a rebate, the
Community Courts have constantly taken into account the level of foreclosure:
only rebate regimes which foreclosed a substantial portion of customers on the
market have been found illegal. Indeed, if a loyalty rebate is available to only a
small number of customers, then the rebate cannot restrict competition since there
is a viable access to the market for competitors of the dominant firm. This
approach is also reflected in other areas of Article 82 EC case-law. As regards
exclusive dealing, which generally has greater anti-competitive potential than
17
18
See notably the speech delivered by Philip Lowe (Director General of the Directorate General
for Competition of the European Commission) at the Fordham Antitrust Thirtieth Annual
Conference on International Antitrust Law and Policy Conference in Washington 23 October
2003: We, I think as you, were slightly surprised at the Court of First Instances analysis in
Michelin II, that it placed so great an emphasis on per se rules and on certain types of conduct
and did not go into any further economic analysis of the case.
Or, in the words of the European Court of Justice, a firm granting such rebates is at liberty to
demonstrate that its bonus system producing an exclusionary effect is economically justified.
(Case C-95/04 P, British Airways plc v Commission, [2007] ECR I-2331, 69 and 86). See
also Conclusions of Advocate General Juliane Kokott in British Airways, C-95/04P, 23
February 2006, pt. 56 to 59 Not all rebates and bonuses which a dominant undertaking grants
to its contractual partners and produce a foreclosure effect are necessarily abusive and
therefore prohibited under Article 82 EC. According to consistent case-law, such rebates and
bonuses are to be regarded as abusive only if they are not based on an economic transaction
which justifies them. If there is a discernible objective economic justification for the rebates or
bonuses, they are not to be regarded as abusive, despite their foreclosure effect. (emphasis
added)
10
19
20
21
22
(i)
The incidence of the rebate on the market, i.e., whether the rebate
affects a substantial part of market demand or whether, on the
contrary, the rebate is only offered to a few selected customers of the
dominant firm, in which case the (negative) incidence of the rebate
may be minimal; 20 and
(ii)
See Case T-65/98, van den Bergh Foods [2003] ECR I-4653, upheld by the ECJ in Case C552/03 P, Unilever Bestfoods [2006] ECR II-9091, and Commission Decision of 11 October
2007 in Case COMP/37966 Distrigas.
See Discussion Paper, 59 and 162(d).
Ibid. 162(e).
See also Discussion Paper 55 which provides that Article 82 prohibits exclusionary conduct
which produces actual or likely anti-competitive effects in the market and which can harm
consumers in a direct or indirect way. The longer the conduct has already been going on, the
more weight will in general be given to actual effects.
11
(ii)
Ibid. 172-176.
Ibid. 84.
See Case 85/76, Hoffman-La Roche v. Commission [1979] ECR 461, 91.
12
26
27
28
Where the effective price (i.e., the standard list price minus the rebate) is
lower than average avoidable costs (AAC) 27 , the rebate is presumed
predatory. In this case, the European Commission examines whether the
rebate can be objectively justified; and/or generates pro-competitive
efficiencies. 28
Where the effective price is above average total costs (ATC), the rebate is
presumed non predatory.
When the effective price is above AAC but below ATC, unlawful predation
can only be established on the basis of additional evidence of a plan or a
strategy to predate. There is proof of such strategy where there is direct
evidence of intent; evidence that the pricing only makes commercial sense as
part of a predatory strategy; the actual or likely exclusion of the prey; whether
certain customers are selectively targeted; whether the dominant undertaking
actually incurred specific costs in order for instance to expend capacity;
whether there is recoupment etc.
13
Incremental rebates under the Discussion paper are subject to a predation test
relatively similar to the one applicable to unconditional rebates. Thus, if the
effective price is above ATC, the rebate is presumed non predatory. If, on the
contrary, the effective price is below ATC, the European Commission will
examine whether the rebate has foreclosure effects (looking at, for instance, the
market coverage of the rebate, whether and the extent to which rivals have
counterstrategies at their disposal etc.). If the rebate has foreclosure effects, the
European Commission will examine whether the rebate can be objectively
justified and/or generates pro-competitive efficiencies.
The European Commission also indicates that incremental rebates with a
standardized threshold are less likely to produce anticompetitive foreclosure
effects than incremental rebates with an individualized threshold or where the
threshold is set in terms of a percentage of total requirements of the buyer.
b) Framework of analysis of conditional incremental rebates under the 2008
Contribution
In the 2008 Contribution, the European Commission continues to apply a
predation test to incremental rebates. The 2008 Contribution introduces the long
run average incremental cost (LRAIC) 29 as the relevant cost benchmark (instead
of the ATC cost benchmark used in the 2005 Discussion Paper).
The effective price that the customer faces in deciding from whom to purchase,
and which the rival will thus have to match, is the standard (list) price less the
rebate it loses by switching, calculated over the incremental purchases
considered. 30 Once the effective price is calculated, the European Commission
compares this price with the dominant undertakings costs.
29
30
LRAIC is the average of all the (variable and fixed) costs that a company incurs to produce a
particular product. Average total cost (ATC) and LRAIC are good proxies for each other, and
are the same in the case of single product undertakings. If multi-product undertakings have
economies of scope, LRAIC would be below ATC for each individual product, as true
common costs are not taken into account in LRAIC (2008 Contribution, footnote 10).
2008 Contribution, 16.
14
Where the effective price is lower than the average avoidable costs (AAC),
the rebate is presumed predatory. In this case, the European Commission will
examine whether the rebate can be objectively justified; and/or generates procompetitive efficiencies.
Where the effective price is above LRAIC, the rebate is presumed non
predatory.
When prices are above AAC but below LRAIC, the European Commission
will investigate whether other factors point to the conclusion that entry or
expansion by as efficient competitors is likely to be affected. In this context, it
will review the market coverage of the rebate, whether and the extent to which
rivals have counterstrategies at their disposal etc. Where competitors do not
have sufficient possibilities to match the effective price, the European
Commission will conclude that the rebate is capable of foreclosing as efficient
competitors. 31 In this case, the European Commission will examine whether
the rebate can be objectively justified; and/or generates pro-competitive
efficiencies.
Like in the Discussion Paper, the European Commission notes that incremental
rebates with a standardized threshold are less likely to produce anticompetitive
foreclosure effects than incremental rebates with an individualized threshold.
3) Conditional all-unit rebates
a) Framework of analysis of conditional all-unit rebates under the
Discussion Paper
Under the Discussion Paper, all-unit rebates are subject to a complex suction
effect test described hereafter. The suction effect test is a tool to assess when a
rebate scheme can exclude efficient rivals from the contestable part of a
customers demand. The suction effect test involves the following scenario.
A dominant supplier sells to a particular company. The supplier has an assured
base of sales to that customer because, for a portion of the customers demand,
there are no proper substitutes. 32 These sales represent the non-contestable share
of that companys demand. However, the portion of the customers demand for
which substitutes are available is the commercially viable share, or the
contestable share, of that customers demand. 33 The dominant supplier offers the
company an all-unit rebate. This gives the customer a rebate if it purchases more
than the rebate threshold level within the reference period. 34 The quantities
highlighted in bold are shown schematically below.
31
32
33
34
15
The competition concern is that when the non-contestable part of the customer in
question is large compared to the contestable part, the all-unit rebate may allow
the dominant supplier to leverage its power from the non-contestable part to the
contestable part. Indeed, while the dominant supplier can recoup the rebate on its
overall sales including both contestable and non-contestable parts, competing
suppliers will have to recoup the rebate over a smaller base represented by the
contestable part. This all-unit rebate scheme could thus have the effect of
excluding equally efficient rivals from that part of the customers sales that would
otherwise be contestable. This would happen if the rebate scheme means that the
dominant supplier is selling units in the contestable part of demand at an effective
price that does not cover the costs of supplying them.
When implementing the test for a single product rebate the Discussion Paper
specifies that the relevant cost standard is to compare incremental revenues to
average total costs (ATC).
A simple numerical example will make the leverage mechanism clear. Suppose
that Customer A will always buy 100 units that are available only from the
dominant supplier, so the assured base or the non-contestable share is 100 units.
But the customers total demand is for 200 units, and the remaining 100 units
could be satisfied by products sold by either the dominant supplier or one of its
competitors. Thus the commercially viable share, or contestable share, is 100
units. The ATC is $ 1 per unit.
The dominant supplier offers the following pricing scheme. The customer pays $
2 per unit if they buy any quantity less than 200 units. But if they buy 200 units
they are given a rebate worth $ 120 in total, or $ 0.6 for each of the 200 units
bought in total. To determine whether there is a suction effect, the Discussion
Paper required the calculation of the effective price for the units that belong to the
contestable share and see whether this price is inferior to the dominants supplier
ATC.
16
If the effective price is above ATC, the European Commission considers that it is
unlikely to have foreclosure effects. Indeed, the European Commission has
considered that in some exceptional circumstances, pricing above ATC could be
predatory. This includes cases of collective dominance where collectively
dominant companies apply a clear strategy to collectively exclude or discipline a
competitor by selectively undercutting the competitor and thereby putting pressure
on its margins, while collectively sharing the loss of revenues. 35 It also include
the situation where a single dominant company operates in a market where it has
certain non-replicable advantages or where economies of scale are very important
and entrants necessarily will have to operate for an initial period at a significant
cost disadvantage because entry can practically only take place below the
minimum efficient scale. 36 .
If the effective price is below ATC, the European Commission will examine
whether there are other factors point to the conclusion that entry or expansion by
as efficient competitors is likely to be affected. In this context, it will review the
market coverage of the rebate, whether there has been aggressive and significant
entry and/or expansion by competitors and/or switching competitors etc. Where
European Commission concludes that the rebate is capable of foreclosing as
efficient competitors, it will examine whether the rebate can be objectively
justified; and/or generates pro-competitive efficiencies.
b) Framework of analysis of conditional all-unit rebates under the 2008
Contribution
In the 2008 Contribution, when assessing the legality of conditional all-unit
rebates (or retroactive rebates), the European Commission continues to use a
suction effect test to determine the effective price. Regrettably, the European
Commission gives only very general indication as to the determination of the size
of the contestable share (so-called relevant range in the 2008 Contribution).
Thus, it notes that If customers are likely to be willing and able to switch large
amounts of demand to a (potential) rival relatively quickly, the relevant range is
likely to be relatively large. If on the other hand customers are likely only to want
or be able to switch small amounts incrementally, then the relevant range will be
relatively small. 37
35
36
37
17
8. Has your jurisdiction developed any safe harbors governing loyalty discounts
or rebates? Yes/No. If yes, please explain the terms of the safe harbor.
The European Commission has not developed clear safe-harbors with regard to
loyalty discounts and rebates. As explained in our answers to Question 6, if the
effective price of a loyalty rebate (such as an all-unit rebate) is above ATC,
according to the Discussion Paper, the European Commission states that it is
unlikely to conclude there is foreclosure. It further states that under exceptional
circumstances, such as when the dominant firm has non-replicable advantages, the
European Commission may conclude that, even though the effective price above
ATC, the rebate in question may create foreclosure. The implication is that
dominant firms granting rebates can draw no comfort from this statement. Even
when their effective price is above ATC, they might still be challenged by the
European Commission.
18
POLICY
10. What policy considerations does your jurisdiction consider with respect to
loyalty discounts and rebates?
You may wish to address the following sorts of issues: Are loyalty discounts and
rebates common? Does your jurisdiction generally consider them to be procompetitive? Does your answer depend on whether the firm offering the
discounts is dominant? Does your jurisdiction view loyalty discounts and
rebates by a dominant firm as generally anticompetitive? What competitive
38
39
40
19
41
42
43
2008 Contribution, 2.
Idem. 3.
Case C-163/99, Portugal v. EC Commission [2001] ECR I-2613, 52.
20
45
See Selective Price Cuts and Fidelity Rebates, Economic Discussion Paper, July 2005,
OFT804 (the OFT report), 2.29 where the OFT states Theory does not suggest that
dominant firms would usually use discount schemes to harm competition., a requirement
that discounts must be justified [in order not to be considered anti-competitive] could chill
price competition where firms are discouraged from employing beneficial discounts due to the
burden of having to justify them to the authorities.
See OFT Report, at 2.9 et seq.
21
47
For instance, the threshold is set at a threshold above the level that the buyer would purchase
from the dominant company in the absence of any loyalty enhancing obligation or rebate, the
non contestable share is larger than the contestable share, the respective size of these shares do
not fluctuate, etc.
United Kingdom, Roundtable on Bundled and Loyalty Discounts and Rebates,
DAF/COMP/WP3/WD(2008)46, 10 June 2008 at 38.
22
The application of a price-cost test raises the complicated question of the selection
of the appropriate cost standard.
The approach of the European Commission has evolved since its 2005 Discussion
Paper with regard to the selection of the appropriate cost benchmark. Thus, in the
2008 Contribution, the Commission uses the LRAIC cost benchmark instead of
the ATC cost benchmark. 51 This evolution is, however, of limited importance as
48
49
50
51
This has been recognized by the DOJ in its Competition and Monopoly: Single-Firm
Conduct Under Section 2 of the Sherman Act, 2008, available at
www.usdoj.gov/atr/public/reports/236681.htm, (see p .117 where the DOJ states: the
Department believes that an approach requiring courts to determine whether a portion of a
market is uncontestable and to quantify that portion, as well as to analyze whether a discount
deprived the plaintiff of efficient scale, would be difficult to administer. More importantly,
such an approach would not provide much clarity to firms deciding whether to offer discounts
and likely would chill desirable price competition.
U.S. Department of Justice, Competition and Monopoly: Single-Firm Conduct Under
Section 2 of the Sherman Act, 2008, at p.111.
U.S. Department of Justice, Competition and Monopoly: Single-Firm Conduct Under
Section 2 of the Sherman Act, 2008, at p.116.
As noted above, when implementing the test for a single product loyalty rebate, the
Discussion Paper requires a comparison of the dominant firms incremental revenues to its
23
52
53
54
55
ATC. In its 2008 Contribution, the European Commission considers that the LRAIC is instead
the applicable cost standard.
In practice the incremental cost of offering a new product might be less than the average total
cost when production of the two shares some facilities or inputs. An example would be if both
products were developed using shared research facilities and if research done to develop one
meant that it was cheaper to develop the other. In cases where the incremental cost of offering
contestable products is lower than the average total cost, a rebate is less likely to be found
abusive, under the tests in the Discussion Paper, where it is characterised as a multi-product
rebate rather than a single product rebate
See Discussion Paper, 154.
See, e.g. Ahlborn and Bailey, supra note 17, at 140; Simon Bishop and Philip Marsden,
Editorial - The Article 82 Discussion Paper: A Missed Opportunity, April 2006, European
Competition Journal, 1-7 at p.5.
See Discussion Paper, 111.
24
56
In the US, federal courts and the DOJ have made abundantly clear that single-product rebates
will not be subject to antitrust liability when the effective price is above AVC. See U.S.
Department of Justice, Competition and Monopoly: Single-Firm Conduct Under Section 2 of
the Sherman Act, 2008, available at www.usdoj.gov/atr/public/reports/236681.htm
25