Revised abuse of dominance test stands as Supreme Court refuses leave to appeal to Canada Pipe
May 10th, 2007
On May 10, 2007, the Supreme Court of Canada denied leave to appeal to Canada Pipe Company Ltd. to appeal from a decision of the Federal Court of Appeal that substantially revised the test to be applied in cases of abuse of dominance. In this decision, the Federal Court of Appeal allowed the Commissioner of Competition’s appeal from the Competition Tribunal’s 2005 finding that Canada Pipe Company’s rebate program was not anti-competitive, on the basis that the Tribunal applied the wrong test. The case will now go back to the Tribunal for re-determination based on test articulated by the Federal Court of Appeal.
Key aspects of the new test for abuse of dominance are:
- Competitors, not competition: Past decisions of the Tribunal have defined anti-competitive acts in terms of an intended predatory, exclusionary or disciplinary effects on a competitor. The court emphasized that in identifying anti-competitive acts, one must look at the conduct’s effects on a competitor, not on competition.
- Centrality of intention: The court emphasized the centrality of intention, holding that anti-competitive acts are identified by reference to their purpose.
- Business justification neutralizes intent; not a defence: a valid business justification is not an absolute defence. Rather, evidence of a business justification can only be used to counterbalance or neutralize evidence of anti-competitive intent. It must provide a “credible efficiency or pro-competitive explanation, unrelated to an anti-competitive purpose”.
- But for test for substantial lessening of competition: the Tribunal must in all cases undertake a comparison between market conditions with the impugned practice, and likely market conditions in the absence of the impugned practice to determine whether the practice substantially lessens or prevents competition. The fact that the market is competitive and experiences entry, even with the impugned practice, is not relevant.
The court did not intervene in the Tribunal’s findings on market definition and dominance. The standard of review applicable to these findings is that of “reasonableness”. The majority found that these findings were reasonable. Pelletier J.A., however, dissenting, would have reversed the Tribunal’s finding that Canada Pipe had market power in several geographic markets.
Canada Pipe’s stocking distributor program
Canada Pipe’s Bibby division manufactures cast iron drain, waste and vent pipes (“DWV”) for use in buildings. It markets most of its products through distributors. Distributors who stock Bibby products exclusively and purchase a minimum amount pay a much lower price than non-exclusive distributors (in one example, 40% lower) and also receive quarterly rebates under Bibby’s “Stocking Distributor Program” (“SDP”).
Bibby’s market share is between 80-90%. The only other manufacturer of cast iron drain, waste and vent pipes in Canada, Vandem Industries, was founded in 1997 and has a much smaller market share. It complained to the Competition Bureau that Bibby was abusing its dominant position through its SDP.
Abuse of dominant position
Section 79 of the Competition Act allows the Tribunal to make a remedial order where (1) a firm “substantially or completely controls” a market (dominance) (s. 79(1)(a)), and (2) engages in a practice of anti-competitive acts (s. 79(1)(b)), which (3) substantially lessens or prevents competition or is likely to do so (s. 79(1)(c)).
The Tribunal decision
The Tribunal found that Bibby dominates the market for cast iron DWV pipes, but that its SDP did not constitute a practice of anti-competitive acts and did not lessen or prevent competition substantially.
Bibby’s SDP operates as an inducement to exclusive dealing, the Tribunal found. However, several of its key features led the Tribunal to conclude that it is not anti-competitive:
- The terms of the program are clear and transparent. The program operates on a calendar year basis, and distributors renew their commitment to it on a quarterly basis.
- There are no contractual constraints preventing distributors from terminating their participation in the program.
- There are no heavy penalties for leaving the program. Distributors that leave the program do not have to pay back the discounts they obtained. At worst, they may not receive the rebate for purchases already made.
The Tribunal distinguished these features from practices that have in the past been declared “anti-competitive” by the Tribunal. Bibby’s program lacked the offensive characteristics found in past cases.
The Tribunal found that although the discounts and rebates are attractive to distributors, and a distributor would weigh them carefully before leaving the program, it did not impose significant “switching costs” on distributors. Indeed, some distributors did leave the program in order to stock competing products. Consequently, the Tribunal found, the SDP does not have an exclusionary effect and is not anti-competitive.
Bibby had a reasonable business justification for the Stocking Distributor Program, the Tribunal also found. The program kept Bibby’s volumes of its standard products high enough that it could offer a complete array of “exotic”, or specialized, cast iron fittings.
Both the Commissioner and Canada Pipe appealed. On June 23, 2006, the Federal Court of Appeal allowed the Commissioner’s appeal and dismissed Canada Pipe’s cross-appeal.
Discrete statutory tests
Central to the court’s reasoning is the fact that s. 75 is divided into three elements – dominance, practice of anti-competitive acts, and substantial lessening of competition. Each of these three elements must give rise to a distinct legal test, Desjardins J.A. stated. Conflating the elements would make one of them redundant, violating the principle of statutory interpretation that “no legislative provision should be interpreted so as to render it mere surplusage”.
Competitors, not competition
As Desjardins J.A. notes, the Competition Act does not contain a definition of “anti-competitive act”. Rather, it contains a non-exhaustive list of anti-competitive acts. In its very first case under the abuse of dominance provisions, the Tribunal had to develop a test for determining whether conduct not listed in the Act could be classified as anti-competitive. The Tribunal discerned a common thread running through the conduct in the list, a common purpose, namely, “an intended negative effect on a competitor that is predatory, exclusionary or disciplinary”.
Having adopted this test, Desjardins J.A. noted the Tribunal’s use of the word “competitor”, and held:
The paragraph 79(1)(b) inquiry is thus focused upon the intended effects of the act on a competit or. As a result, some types of effects on competit ion in the market might be irrelevant for purposes of paragraph 79(1)(b), if these effects do not manifest through a negative effect on a competitor. It is important to recognize that “anti-competitive” therefore has a restricted meaning within the context of paragraph 79(1)(b). While, for the Act as a whole, “competition” has many facets as enumerated in section 1.1, for the particular purposes of paragraph 79(1)(b), “anti-competitive” refers to an act whose purpose is a negative effect on a competit or. [Emphasis in orginal]
Desjardins J.A. therefore held that evidence concerning competition in the market is irrelevant to the question of whether conduct constitutes an “anti-competitive act”:
One must remember, however, that in the context of paragraph 79(1)(b), evidentiary factors are relevant only insofar as they shed light upon the paragraph 79(1)(b) statutory test, that is upon the purpose of the act vis-à-vis competitors. Evidence concerning other types of effects of the impugned act that are not related to competitors – while perhaps pertinent in respect of the paragraph 79(1)(c) assessment of competition – are not directly relevant for paragraph 79(1)(b). Similarly, evidence concerning the general competitive state and structure of the relevant market, and whether such features can be causally attributed to the impuged act, are not the direct focus of the paragraph 79(1)(b) analysis… [Emphasis in original]
Effects upon consumers are thus also irrelevant to s. 79(1)(b), Desjardins J.A. held.
The Tribunal erred in approaching the SDP from the perspective of its effects on competition generally and on consumers, Desjardins J.A. held. It adopted an “unwarrantedly and incorrectly narrow test”, she stated, because the Tribunal’s reasoning
would imply that unless an impugned act prevents entry of competitors or otherwise prevents competition, or unless it is the (predominant) cause of the uncompetitive attributes observed in the market, the act cannot be considered anti-competitive. This result is clearly incorrect. [Emphasis in original]
Desjardins J.A. attributed the Tribunal’s error to a conflation of the tests under s. 79(1)(b) and s. 79(1)(c).
Centrality of intent
Desjardins J.A.’s decision is remarkable for its emphasis on purpose as the key indicator of whether an act is anti-competitive. Interpreting the Tribunal’s definition of anti-competitive act in Nutrasweet, she observed: “an anti-competitive act is identified by reference to its purpose”.
Desjardins J.A. qualified this, noting that evidence of subjective intent, while probative if it exists, is not required. Intention can be deemed from reasonably foreseeable consequences of the acts. It seems, however, that for Desjardins J.A., the anti-competitive purpose must be the overriding purpose of the act, even though this purpose can be inferred from the act’s consequences.
Legitimate business justification neutralizes intent
The so-called “legitimate business justification defence” is not a complete defence, Desjardins J.A. held. Rather, it only applies at the stage of determining whether an anti-competitive act has been committed, and it operates to negative evidence of anti-competitive purpose.
The Tribunal erred in accepting Canada Pipe’s business justification for the SDP program, which was that it enabled Canada Pipe to stock a complete line of exotic pipe fittings, which benefits consumers, Desjardins J.A. held. This is “actually a self-interest argument based on selling more product” that does not qualify as a reasonable business justification, she stated. Moreover, benefits to consumers cannot establish a reasonable business justification.
Desjardins J.A. then held that, to be accepted, a valid business justification must:
- “provide a credible efficiency or pro-competitive explanation”;
- which is not related to an anti-competitive purpose; and
- be attributable to the respondent.
But for test for substantial lessening of competition
The major error in the Tribunal’s reasoning, according to Desjardins J.A., was its approach to the test for the third element, substantial lessening of competition.
Section 79(1)(c) requires a showing that the practice of anti-competitive acts “has had, is having or is likely to have the effect of preventing or lessening competition substantially in a market”. The word “substantial” demands a relative and comparative assessment. The Tribunal must compare the competitiveness of the market with the impugned practice with the competitiveness that would exist without the impugned practice to see whether any lessening or prevention is “substantial”. This comparison must be carried out for the past, present, and future. This, Desjardins J.A. indicated, is equivalent to a “but for” test.
Desjardins J.A. then retreated slightly from this reasoning. The “but for” test is but one comparative approach; there may be others. However, in every case, the Tribunal must consider the “but for” test.
Addressing the application of the “but for” test, Desjardins J.A. said it is up to the Tribunal to determine the appropriate methodology. She suggested that it could involve a hypothetical model. Or a comparison of the competitiveness of the market before the introduction of the impugned practice with its competitiveness afterwards might be used as a proxy.
The Tribunal erred, Desjardins J.A. held, because it applied an absolute, rather than a comparative analysis. That is, the Tribunal relied heavily on the fact that the SDP did not prevent entry. It did not, however, consider whether it made entry substantially more difficult, thus substantially lessening competition; nor did it consider whether the market would become substantially more competitive without the practice.
The panel unanimously held that the Tribunal’s conclusions on product market definition were reasonable. The majority also held that it was open to the Tribunal to conclude that Canada Pipe possessed the requisite degree of market power. Accordingly, it did not disturb this finding.
Pelletier J.A. dissented, however, on this point. He noted that suppliers had entered several geographic markets and were competing with Canada Pipe. He said that if Canada Pipe really had market power, it would have been able to maintain supra-competitive prices in the face of entry, but, in fact, had dropped its prices in response to competition. Thus, he held, Canada Pipe did not possess market power in the geographic markets that had experienced entry.
It is most unfortunate that the Supreme Court has denied leave to appeal, as several aspects of the Federal Court of Appeal’s decision are problematic.
The court’s statement that for purposes of determining whether an act is anti-competitive, effects on competitors are relevant, and effects on competition are not, is problematic. First, the statutory text itself does not limit the test to effects on a competitor. True, the examples of anti-competitive acts given in s. 78(1) are phrased in terms of effects on competitors. This, presumably, is what led the Tribunal to use the term “competitor” in Nutrasweet. Nevertheless, the list in s. 78(1) is expressly declared to be non-exhaustive. Desjardins J.A. appears to have treated the test in Nutrasweet as though it were statutory language.
Desjardins J.A.’s approach to the Nutrasweet test represents a departure from the Tribunal’s reasoning in past abuse of dominance cases, including Nutrasweet itself. In Nutrasweet itself, the Tribunal considered the effects of the impugned conduct on both competitors and competition. In Laidlaw, Nielsen, and Teledirect, the Tribunal refined the Nutrasweet test in a way that makes it clear that effects on competition generally are relevant to determining whether an act is anti-competitive or not.
Desjardins J.A.’s primary concern appears to have been that if effects on competition are relevant to determining whether an anti-competitive act has been committed, the substantial lessening or prevention of competition element becomes redundant, which offends the rules of statutory interpretation. However, her interpretation requires her to hold (explicitly) that “competition” means something different in s. 79(1)(b) than it does in s. 79(1)(c). This offends another canon of statutory interpretation: the presumption of consistent expression, that is, that words used in a statute mean the same thing every time they are used. Moreover, there is an obvious link between the two elements. Anti-competitive, that is, exclusionary, disciplinary or predatory, conduct by a dominant firm is, generally speaking, likely to result in a substantial lessening or prevention of competition.
Desjardins J.A.’s interpretation of the test raises interpretive problems. For instance, must the effects be on an identifiable competitor, as opposed to competitors in general, or a hypothetical competitor? It would seem that they must, since effects on competitors in general, or on hypothetical or potential competitors, would really amount to an effect on competition, which, according to Desjardins J.A., cannot be considered at the s. 79(1)(b) stage. This leads to a further problem. If there must be an effect on an identifiable competitor, then it may be that a monopolist could not engage in anti-competitive acts, since there are no competitors to suffer these effects.
Finally, Desjardins J.A.’s focus on effects on competitors is incongruous, given the oft-stated maxim that competition law protects competition, not competitors.
Desjardins J.A.’s emphasis on purpose in identifying anti-competitive acts is also problematic. Her reasoning gives effects only a supporting role: they are relevant to establishing purpose. This emphasis on purpose is inconsistent with recent thinking on abuse of dominance, which promotes an effects-based analysis. As well, reliance on intention is not consistent with the remedial nature of s. 79. That is, s. 79 is concerned with remedying anti-competitive effects caused by dominant firms. While intention is a good indicator of a guilty mind, it provides no indication about whether an act has anti-competitive effects. It is at least arguable, therefore, that the intention element, as expressed by the Tribunal in past cases, is designed to help distinguish between vigorous competition on the merits and abuse of dominance. The Tribunal has, in past cases, commented on the difficulty of distinguishing between competition on the merits and anti-competitive conduct.
Desjardins J.A.’s approach to the business justification defence may also be internally inconsistent. On the one had, she deprecates Canada Pipe’s business justification as a “self-interest argument”. On the other, she states that a benefit to consumers does not count; the business justification must be “attributable to respondent”. It is hard to square these two statements. To be attributable to the business itself, the business justification must necessarily be in that business’ self-interest. It would not make sense as a business justification otherwise. We expect businesses to act in their own self-interest. Indeed, we are suspicious when businesses embark on conduct that is not profitable.
Finally, while the “substantial lessening or prevention of competition” test used by the Competition Act does connote a comparative analysis, as Desjardins J.A. points out, her insistence on the “but for” test in every case may lead to a standard of perfection. That is, it is not enough, as she holds, to find that the market is competitive and that the impugned practices are not having exclusionary effects. The Tribunal must go further as ask whether there might be even more competition if the respondent were forced to stop the impugned practice. This is, arguably, inconsistent with the approach of the Supreme Court in Southam. The court held that, in choosing a remedy in a merger case, the Tribunal must chose a remedy that eliminates the substantial lessening of competition, not a remedy that restores the parties to the pre-merger competitive situation.
Published May 10, 2007
Canada (Commissioner of Competition) v. Canada Pipe Co.,  F.C.J. No. 1027 (“Appeal decision”)
Canada (Commissioner of Competition) v. Canada Pipe Co.,  F.C.J. No. 1028 (“Cross-appeal decision”)
Competition Act, R.S.C. 1985, c. C-34.
 At ¶26, citing R. v. Proulx,  1 S.C.R. 61 at ¶28. This is known as the presumption against tautology. See Sullivan, Driedger on the Construction of Statutes, 3d ed, 1994, at p. 159ff.
Competition Act, R.S.C. 1985, c. C-34, s. 78.
Canada (Director of Investigation and Research) v. Nutrasweet Co. (1990), 32 C.P.R. (3d) 1 at 34 (Comp. Trib.)
 Appeal decision, ¶68
 Appeal decision, ¶78
 Appeal decision, ¶79
 Appeal decision, ¶82
 Appeal decision, ¶66.
 Appeal decision, ¶66, 71-73.
 Appeal decision, ¶73.
 Appeal decision, ¶87-88
 Appeal decision, at ¶89-90
 Appeal decision, ¶37-38
 Appeal decision, ¶45-46.
Canada (Director of Investigation and Research, Competition Act) v. Laidlaw Waste Systems Ltd. (1992) 40 C.P.R. (3d) 289.
Canada (Competition Act, Director of Investigation and Research) v. The D & B Companies of Canada Ltd. (1995) 64 C.P.R. (3d) 216 at 257
Canada (Competition Act, Director of Investigation and Research) v. Tele-Direct (Publications) Inc.,  C.C.T.D. No. 8 at ¶540-542.
 See Driedger at p. 163ff.
 See, for instance, the December 2005 EU DG Competition discussion paper on the application of Article 82 of the Treaty to exclusionary abuses.
Canada (Director of Investigation and Research) v. Southam Inc.,  S.C.J. No. 116, ¶83-85.