Canada (Director of Investigation and Research) v. Southam Inc., [1997] 1 S.C.R. 748

Southam Inc.,

Lower Mainland Publishing Ltd.,

RIM Publishing Inc.,

Yellow Cedar Properties Ltd.,

North Shore Free Press Ltd.,

Specialty Publishers Inc.,

and Elty Publications Ltd. Appellants

v.

Director of Investigation and Research Respondent

Indexed as: Canada (Director of Investigation and Research) v. Southam Inc.

File No.: 24915.

Hearing and judgment as to appeal on the remedy: November 25, 1996.

Reasons and judgment as to appeal on the merits: March 20, 1997.

Present: Lamer C.J. and La Forest, L’Heureux‑Dubé, Sopinka, Gonthier, Cory, McLachlin, Iacobucci and Major JJ.

on appeal from the federal court of appeal

Administrative law ‑‑ Statutory appeals ‑‑ Tribunal composed of economic, commercial and legal experts -- Tribunal’s decision not protected by privative clause ‑‑ Standard of appellate review -- Competition Act, R.S.C., 1985, c. C-34, s. 92(1) -- Competition Tribunal Act, R.S.C., 1985, c. 19 (2nd Supp.), ss. 12, 13.

Competition law -- Remedy -- Tribunal investigating substantial lessening of competition -- Tribunal ordering divestment of one business at owner’s option -- Whether remedial order should be set aside -- Competition Act, R.S.C., 1985, c. C-34, s. 92(1) -- Competition Tribunal Act, R.S.C., 1985, c. 19 (2nd Supp.), ss. 12, 13.

Vancouver’s two daily newspapers (owned by Southam Inc.) were less successful, when compared with daily newspapers in other regions of Canada, relative to the many smaller community newspapers circulating in their distribution area. The community newspapers differed from the dailies in that they served a smaller area, were distributed free of charge, and were printed from one to three times a week. In 1989, Southam Inc. began to acquire community and specialized newspapers in the area, and one year later had obtained a controlling interest in 13 community newspapers (including the two strongest ones, the North Shore News and the Vancouver Courier), a real estate advertising publication, three distribution services and two printing concerns. Southam Inc. also established a local supplement to one of its dailies but eventually discontinued it.

The respondent applied for an order requiring Southam to divest itself of the North Shore News, the Vancouver Courier, and the Real Estate Weekly, alleging that the concentration of these properties in the hands of one publisher was likely to lessen competition substantially in the retail print advertising and real estate print advertising markets in the Lower Mainland. The Competition Tribunal found a substantial lessening in competition in the real estate print advertising market in the North Shore. It ordered Southam to divest itself, at its option, of either the North Shore News or the Real Estate Weekly. It rejected Southam’s proposal that it sell the real estate section of the North Shore News. The Director of Investigation and Research appealed the Tribunal’s decision on the merits and Southam appealed the Tribunal’s decision on the remedy. The Federal Court of Appeal allowed the first appeal and dismissed the second.

This appeal raises two issues. The first is whether the Federal Court of Appeal erred in concluding that it owed no deference to the Tribunal’s finding about the dimensions of the relevant market and in subsequently substituting for that finding one of its own. The second is whether the Federal Court of Appeal erred in refusing to set aside the Tribunal’s remedial order.

Held: The appeal on the merits should be allowed; the appeal on the remedy should be dismissed.

Merits

The standard of review is a function of many factors and may fall between correctness, at the more exacting end of the spectrum, and patently unreasonable, at the more deferential end. In the absence of a privative clause, the reviewing court may review decisions taken by the tribunal even within its own jurisdiction. Accordingly, the task for the reviewing court in a statutory appeal is more akin to appellate review than to judicial review. Nevertheless, the reviewing court must look to several factors to determine what limits it should observe in exercising its statutorily mandated appellate function. Among the factors to be considered are the nature of the problem before the tribunal, the applicable law properly interpreted in the light of its purpose and the area of the tribunal’s expertise.

The problem before the Tribunal in this case was a problem of mixed law and fact. Questions of law are questions about what the correct legal test is; questions of fact are questions about what actually took place between the parties; and questions of mixed law and fact are questions about whether the facts satisfy the legal tests. The distinction between questions of law and questions of mixed law and fact will sometimes be difficult to make. In theoretical terms, the rule is that as the level of generality of the challenged proposition approaches complete particularity, the matter approaches unqualified application of law and draws away from the forging of new law, and hence draws nigh to being an unqualified question of mixed law and fact.

The Tribunal did not fail to consider relevant items of evidence and so did not err in law by failing to consider them. To suggest that it erred in law by failing to accord adequate weight to certain factors is inimical to the very notion of a balancing test, which is a kind of legal rule whose application should be subtle and flexible, but not mechanical. As a matter of law, the Tribunal should consider each factor, but the according of weight to the factors should be left, at least initially, to the Tribunal. The Tribunal forged no new legal principle and so any error it might have made can only have been one of mixed law and fact. This suggests that some measure of deference accordingly is owed to the Tribunal’s decision. Appellate courts should be reluctant to venture into a re‑examination of the conclusions of the Tribunal on questions of mixed law and fact.

The absence of a privative clause counsels a less deferential posture for appellate courts than would be appropriate if a privative clause were present. The Tribunal, however, has been recognized as being especially well‑suited to overseeing a complex statutory scheme whose objectives are peculiarly economic. Because an appellate court is likely to encounter difficulties in understanding the economic and commercial ramifications of the Tribunal’s decisions and consequently to be less able to secure the fulfilment of the purpose of the Competition Act, the purpose of the Act is better served by appellate deference to the Tribunal’s decisions.

Expertise, which in this case overlaps with the purpose of the statute that the Tribunal administers, is the most important of the factors that a court must consider in settling on a standard of review. The Tribunal’s expertise lies in economics and in commerce, and these are matters concerning which the members of the Tribunal are likely to be far more knowledgeable than the typical judge will be. The particular dispute in this case is one that falls squarely within the area of the Tribunal’s expertise.

A standard more deferential than correctness but less deferential than “not patently unreasonable” is required. Because several considerations, including particularly the expertise of the Tribunal, counsel deference while others suggest a more exacting form of review, the proper standard of review falls somewhere between the ends of the spectrum.

The need for a third standard of review is especially clear in cases, like this one, in which appeal from a tribunal’s decision lies by statutory right. The presence of the statutory right of appeal obviates the need to find a jurisdictional error. Because the standard of patent unreasonableness is principally a test for determining whether a tribunal has exceeded its jurisdiction, it will rarely be the appropriate standard of review in statutory appeals. However, because tribunals typically enjoy some expertise and deal with problems of a difficult and intricate nature, a standard more deferential than correctness is needed. This third standard should be whether the decision of the Tribunal is unreasonable. This test is to be distinguished from the most deferential standard of review, which requires courts to consider whether a tribunal’s decision is patently unreasonable. An unreasonable decision is one that, in the main, is not supported by any reasons that can stand up to a somewhat probing examination. The difference between “unreasonable” and “patently unreasonable” lies in the immediacy or obviousness of the defect. If the defect is apparent on the face of the tribunal’s reasons, then the tribunal’s decision is patently unreasonable. But if it takes some significant searching or testing to find the defect, then the decision is unreasonable but not patently unreasonable.

The “clearly wrong” test is close to the standard of reasonableness simpliciter. Many things are wrong that are not unreasonable, but when “clearly” is added to “wrong”, the meaning is brought much nearer to that of “unreasonable”. Consequently, the clearly wrong test represents a striking out from the correctness test in the direction of deference. But the clearly wrong test does not go so far as the standard of patent unreasonableness. The clearly wrong test, because of its familiarity to Canadian judges, may serve as a guide in applying the standard of reasonableness simpliciter.

In the final result, the standard of reasonableness simply instructs reviewing courts to accord considerable weight to the views of tribunals about matters with respect to which they have significant expertise. While it is convenient to put the matter in terms of a standard of review, at bottom the issue is the weight that should be accorded to expert opinions.

The Tribunal did not act unreasonably when it decided that Southam’s daily newspapers and community newspapers are in different product markets.

That the Tribunal discounted evidence of functional interchangeability between the dailies and the community newspapers was reasonable on the facts and was not without foundation or logical coherence. It is reasonable, if only reasonable, to suppose that advertisers are sufficiently discerning about the media they employ that they are unlikely to respond to changes in the relative prices of the two kinds of newspapers by taking their business from the one to the other.

The Tribunal also discounted evidence that Southam regarded the community newspapers as competitors for its dailies. This discounting is perhaps unusual given that Southam’s expert identified this competition with community newspapers as the source of the dailies’ difficulties. The Tribunal’s findings, however, were not unreasonable and they did not need to be correct. Judicial restraint is needed if a cohesive, rational and sensible system of judicial review is to be fashioned.

Remedy

Because the Competition Act addresses the problem of substantial lessening of competition, the appropriate remedy is to restore competition to the point at which it can no longer be said to be substantially less than it was before the merger. The test that the Tribunal has applied in consent cases should be applied in all cases.

The Tribunal’s choice of remedy is a matter of mixed law and fact and the standard of review is one of reasonableness.

Southam’s proposed remedy of selling the real estate section of the North Shore News fails because it would not likely be effective in eliminating the substantial lessening of competition. This decision was not unreasonable and should be allowed to stand. The remedy chosen by the Tribunal is not punitive, because the Tribunal found that it was the only effective remedy. If the choice is between a remedy that goes farther than is strictly necessary to restore competition to an acceptable level and a remedy that does not go far enough even to reach the acceptable level, then surely the former option must be preferred. The Tribunal did not wrongly require the appellants to demonstrate the effectiveness of their proposed remedy; the person who asserts should prove.

Cases Cited

Considered: Pezim v. British Columbia (Superintendent of Brokers), 1994 CanLII 103 (SCC), [1994] 2 S.C.R. 557; referred to: R. v. Nova Scotia Pharmaceutical Society, 1992 CanLII 72 (SCC), [1992] 2 S.C.R. 606; Bell Canada v. Canada (Canadian Radio‑Television and Telecommunications Commission), 1989 CanLII 67 (SCC), [1989] 1 S.C.R. 1722; Chrysler Canada Ltd. v. Canada (Competition Tribunal), 1992 CanLII 68 (SCC), [1992] 2 S.C.R. 394; United Brotherhood of Carpenters and Joiners of America, Local 579 v. Bradco Construction Ltd., 1993 CanLII 88 (SCC), [1993] 2 S.C.R. 316; Canadian Union of Public Employees, Local 963 v. New Brunswick Liquor Corp., 1979 CanLII 23 (SCC), [1979] 2 S.C.R. 227; Canada (Attorney General) v. Public Service Alliance of Canada, 1993 CanLII 125 (SCC), [1993] 1 S.C.R. 941; National Corn Growers Assn. v. Canada (Import Tribunal), 1990 CanLII 49 (SCC), [1990] 2 S.C.R. 1324; Toronto (City) Board of Education v. O.S.S.T.F., District 15, 1997 CanLII 378 (SCC), [1997] 1 S.C.R. 487; Stein v. “Kathy K” (The Ship), 1975 CanLII 146 (SCC), [1976] 2 S.C.R. 802; The Queen v. J. W. Mills & Son Ltd., 1968 CanLII 1279 (CA EXC), [1968] 2 Ex. C.R. 275; Canada (Director of Investigation & Research) v. Air Canada (1989), 1989 CanLII 10325 (CT), 27 C.P.R. (3d) 476.

Statutes and Regulations Cited

Competition Act, R.S.C., 1985, c. C‑34, ss. 1.1 [ad. R.S.C., 1985, c. 19 (2nd Supp.), s. 19], 92(1) [rep. & sub. R.S.C., 1985, c. 19 (2nd Supp.), s. 45].

Competition Tribunal Act, R.S.C., 1985, c. 19 (2nd Supp.), ss. 3(2), (3), 4(1), 8(1), 10(1), (2), 12(1), (2), 13(1), (2).

Authors Cited

Kerans, R. P. Standards of Review Employed by Appellate Courts. Edmonton: Juriliber, 1994.

APPEAL from a judgment of the Federal Court of Appeal allowing an appeal on the merits 1995 CanLII 3523 (FCA), [1995] 3 F.C. 557, (1995), 127 D.L.R. (4th) 263, 185 N.R. 321, 63 C.P.R. (3d) 1, 21 B.L.R. (2d) 1, from a judgment of the Competition Tribunal (1992), 1992 CanLII 15430 (CT), 43 C.P.R. (3d) 161, with additional reasons (1993), 1993 CanLII 1509 (CT), 48 C.P.R. (3d) 224, and from a judgment of the Federal Court of Appeal (1995), 1995 CanLII 11081 (FCA), 127 D.L.R. (4th) 329, 185 N.R. 291, 63 C.P.R. (3d) 67, 21 B.L.R. (2d) 68, dismissing an appeal as to remedy from a judgment of the Competition Tribunal (1992), 1992 CanLII 15495 (CT), 47 C.P.R. (3d) 240. Appeal on the merits allowed; appeal on the remedy dismissed.

Neil Finkelstein, Glenn Leslie and Mark Katz, for the appellants.

Stanley Wong, André Brantz and J. Kevin Wright, for the respondent.

//Iacobucci J.//

The judgment of the Court was delivered by

1Iacobucci J. -- The principal question raised by this appeal is whether a decision of the Competition Tribunal (the “Tribunal”) is entitled to curial deference. Following the approach outlined by this Court in its recent jurisprudence, I conclude that the particular decision of the Tribunal here at issue is entitled to deference.

1. Facts

1. Facts

2Two daily newspapers serve the region in and around Vancouver. They are the Vancouver Sun and the Vancouver Province. The appellant Southam Inc., through its subsidiary Pacific Press Limited, owns both.

3In addition to the two dailies, many smaller community newspapers circulate in the Lower Mainland of British Columbia. These community newspapers differ from the daily newspapers in a few respects: they serve smaller regions, they are distributed free of charge to all households in the regions they serve, and they are published only once, twice, or at most three times weekly. Community newspapers have been more successful in the Lower Mainland than in any other comparable region of Canada. Daily newspapers, by contrast, have been less successful in Vancouver than in other major Canadian cities.

4In 1986, Southam consulted Dr. Christine Urban, an American expert, about the problems its Vancouver dailies were facing. Dr. Urban identified Vancouver’s strong community newspapers as the cause of the dailies’ malaise. She advised Southam to act to stem the growing power of the community newspapers.

5In September, 1986, Southam introduced a flyer delivery service to the Lower Mainland. Known as Flyer Force, the new service offered delivery of flyers to even the households that did not receive a Southam newspaper. In 1988, several community newspapers, whose business included the delivery of flyers, joined to form a group whose geographic reach would rival Flyer Force’s. This group was initially called the MetroVan Group. Later in 1988, the MetroVan Group expanded and changed its name to MetroGroup.

6In September, 1988, Southam began to publish the North Shore Extra. This was a bi-weekly publication whose editorial focus was on the North Shore district of the Lower Mainland. The Extra was inserted as a supplement into copies of the Vancouver Sun bound for households in the North Shore. Additionally, the Extra was delivered to North Shore households that did not receive the Sun.

7In January, 1989, Southam began to acquire community and specialized newspapers in the Lower Mainland. By May, 1990, the company had acquired a controlling interest in 13 community newspapers, a real estate advertising publication, three distribution services, and two printing concerns. Among its acquisitions were the Lower Mainland’s two strongest community newspapers, the North Shore News and the Vancouver Courier, as well as the Real Estate Weekly.

8In April, 1990, Southam discontinued the North Shore Extra.

9On November 20, 1990, the respondent, the Director of Investigation and Research, applied for an order requiring Southam to divest itself of the North Shore News, the Vancouver Courier, and the Real Estate Weekly. The Director’s reason for taking this step was that Southam’s acquisition of these publications was likely to lessen competition substantially in the retail print advertising and real estate print advertising markets in the Lower Mainland.

10In early 1991, Southam shut down Flyer Force.

2. Relevant Statutory Provisions

11 Section 92 of the Competition Act, R.S.C., 1985, c. C-34 addresses the problem of mergers that are likely to lessen competition substantially:

92. (1) Where, on application by the Director, the Tribunal finds that a merger or proposed merger prevents or lessens, or is likely to prevent or lessen, competition substantially

(a) in a trade, industry or profession,

(b) among the sources from which a trade, industry or profession obtains a product,

(c) among the outlets through which a trade, industry or profession disposes of a product, or

(d) otherwise than as described in paragraphs (a) to (c),

the Tribunal may, subject to sections 94 to 96,

(e) in the case of a completed merger, order any party to the merger or any other person

(i) to dissolve the merger in such manner as the Tribunal directs,

(ii) to dispose of assets or shares designated by the Tribunal in such manner as the Tribunal directs, or

(iii) in addition to or in lieu of the action referred to in subparagraph (i) or (ii), with the consent of the person against whom the order is directed and the Director, to take any action, . . .

12 Various sections of the Competition Tribunal Act, R.S.C., 1985, c. 19 (2nd Supp.), create and provide for the constitution of the Tribunal:

3. . . .

(2) The Tribunal shall consist of

(a) not more than four members to be appointed from among the judges of the Federal Court -- Trial Division by the Governor in Council on the recommendation of the Minister of Justice; and

(b) not more than eight other members to be appointed by the Governor in Council on the recommendation of the Minister.

(3) The Governor in Council may establish an advisory council to advise the Minister with respect to appointments of lay members, which council is to be composed of not more than ten members who are knowledgeable in economics, industry, commerce or public affairs and may include, without restricting the generality of the foregoing, individuals chosen from business communities, the legal community, consumer groups and labour.

4. (1) The Governor in Council shall designate one of the judicial members to be Chairman of the Tribunal.

5. (1) Subject to section 11, every application to the Tribunal shall be heard before not less than three or more than five members sitting together, at least one of whom is a judicial member and at least one of whom is a lay member.

(2) The Chairman shall designate a judicial member to preside at any hearing or, if the Chairman is present at a hearing, may preside himself.

13 Sections 12 and 13 divide questions before the Tribunal into questions of law, questions of fact, and questions of mixed law and fact, and assign responsibility for resolving those questions, both in the first instance and on appeal:

12. (1) In any proceedings before the Tribunal,

(a) questions of law shall be determined only by the judicial members sitting in those proceedings; and

(b) questions of fact or mixed law and fact shall be determined by all the members sitting in those proceedings.

(2) In any proceedings before the Tribunal,

(a) in the event of a difference of opinion among the members determining any question, the opinion of the majority shall prevail; and

(b) in the event of an equally divided opinion among the members determining any question, the presiding member may determine the question.

13. (1) Subject to subsection (2), an appeal lies to the Federal Court of Appeal from any decision or order, whether final, interlocutory or interim, of the Tribunal as if it were a judgment of the Federal Court ‑- Trial Division.

(2) An appeal on a question of fact lies under subsection (1) only with the leave of the Federal Court of Appeal.

3. Judgments in Appeal

A. Competition Tribunal

(i) On the merits (1992), 1992 CanLII 15430 (CT), 43 C.P.R. (3d) 161, with additional reasons (1993), 1993 CanLII 1509 (CT), 48 C.P.R. (3d) 224

3. Judgments in Appeal

14 Following 40 days of hearings, the Tribunal found that the acquisition by Southam of the community newspapers and affiliated businesses did not substantially lessen competition in the market for retail print advertising in the Lower Mainland. The Tribunal did find, however, that Southam’s purchases had substantially lessened competition in the market for real estate print advertising in the North Shore region. After hearing argument on the issue of remedies, the Tribunal ordered Southam to divest itself, at its option, of either the North Shore News or the Real Estate Weekly. The Tribunal rejected Southam’s proposed remedy, which was to sell the real estate section of the North Shore News.

15 During the hearing, the Tribunal heard from 50 witnesses and received literally volumes of documents in evidence. That the Tribunal paid heed to this prodigious body of evidence is clear from its written reasons, which occupy some 147 pages in a law report. Fortunately, it is not necessary for purposes of this appeal to reproduce the Tribunal’s reasons in any detail.

16 The principal underlying question for the Tribunal was whether Southam’s daily newspapers and its newly acquired community newspapers are in the same market. Its approach to this problem was to ask whether the two kinds of products are close substitutes for one another. The traditional economic measure of substitutability is cross-elasticity of demand, which is the extent to which consumers will switch from one product to another in response to slight changes in their relative prices. However, the Tribunal recognized that direct statistical evidence of cross-elasticity of demand will rarely be available. Accordingly, the members determined that recourse should be had to “indirect evidence” of substitutability. Indirect indicia of substitutability include (at p. 179) “the physical characteristics of the products, the uses to which the products are put, and whatever evidence there is about the behaviour of buyers that casts light on their willingness to switch from one product to another in response to changes in relative prices”. Also relevant are “[t]he views of industry participants about what products and which firms they regard as actual and prospective competitors”.

17 Almost 100 pages of the Tribunal’s decision are taken up with a painstaking review and evaluation of the evidence. On the strength of this, the Tribunal concluded that daily newspapers and community newspapers, though remarkably similar at first glance, serve different retail print advertising markets. Daily newspapers, which circulate widely but reach only a relatively small percentage of households, appeal to the advertising needs of large national firms that serve customers throughout a metropolitan region. Community newspapers, by contrast, circulate only within small communities but typically reach all of the households within those communities. These newspapers appeal to local advertisers whose customers live only within a certain district. In support of this conclusion, the Tribunal presented an informal survey of the behaviour of selected advertisers in the Lower Mainland.

18 The Tribunal also cited considerable evidence to suggest that Southam regarded the community newspapers as its chief competitors. In one document, Dr. Christine Urban, an American newspaper consultant retained by Southam, identified strong community newspapers as the root of Southam’s problems in the Lower Mainland. In another document quoted in the Tribunal’s decision at p. 195, an official of Southam warned against the danger of conceding forever to the community newspapers “a substantial portion of what is normally daily newspaper business”. However, the members did not regard this evidence of what they called “inter-industry competition” as decisive. In their view, it showed that Southam believed that it was competing with the community newspapers. But simply to state that something is believed does not guarantee that it is so, and in this case the Tribunal found that Southam’s belief was unfounded. “With their present product configurations”, concluded the Tribunal at p. 277, “the dailies and community newspapers are at best weak substitutes for some advertisers”.

19 Because the two kinds of newspapers were at best only weak substitutes, the Tribunal concluded that they were not in the same relevant product market and therefore that the acquisition by Southam of several community newspapers and affiliated businesses did not substantially lessen competition in the market for retail print advertising in the Lower Mainland.

20 However, the Tribunal did find that the acquisition by Southam of both the North Shore News, with its weekly real estate supplement, and the Real Estate Weekly, with its North Shore edition, gave Southam monopoly power over the market for real estate print advertising on the North Shore. The result was to lessen competition substantially in that market. The Tribunal ordered the parties to appear at a later date to consider the question of the remedy.

(ii) As to remedy (1992), 1992 CanLII 15495 (CT), 47 C.P.R. (3d) 240

21 Having heard argument on the question, the Tribunal found that the test of a proposed remedy in contested proceedings is whether it will restore the competitive situation as it existed before the merger and is not, as Southam submitted, whether it will eliminate any substantial lessening of competition that the merger may have produced. However, the Tribunal found that, even accepting Southam’s proposed test, Southam’s proposed remedy of selling the weekly real estate supplement to the North Shore News still would not be effective. The Tribunal thought it likely that the real estate supplement would founder on its own; certainly it would not be as substantial a presence in the North Shore as a stand-alone publication as it had been as part of the North Shore News. The Tribunal noted that Southam had offered to reach an accommodation with any prospective buyer concerning the continuation of the supplement’s association with the North Shore News. The Tribunal members concluded, however, that they lacked the jurisdiction to order Southam to reach an accommodation. And in any event, the Tribunal doubted whether such a negotiated association would be conducive to the fostering of a competitive environment. Accordingly, the Tribunal ordered Southam to divest itself, at its option, of either the North Shore News or the Real Estate Weekly.

B. Federal Court of Appeal

(i) On the merits, 1995 CanLII 3523 (FCA), [1995] 3 F.C. 557

B. Federal Court of Appeal

22 The Director of Investigation and Research appealed the Tribunal’s decision on the merits and Southam appealed the Tribunal’s decision on the remedy. The Federal Court of Appeal allowed the first appeal and dismissed the second.

23 Robertson J.A., writing for the court, concluded that the Tribunal, though it had stated the correct formula, had nonetheless applied the wrong legal test. He accepted the Tribunal’s account of the kinds of evidence that it had to consider, but stated that the Tribunal had failed to consider all of these. He found, in particular, that the Tribunal had not considered evidence that daily newspapers and community newspapers are functionally interchangeable and evidence that the owners of the daily newspapers considered themselves to be in competition against the community newspapers. Failure to consider relevant factors, he said, is an error of law. And to his mind, the Tribunal is entitled to no deference on a question of law.

24 By way of buttressing this conclusion, he emphasized that the Competition Tribunal Act mandates an unusual division of labour among the members of the Tribunal. Each panel of the Tribunal, he observed, must have at least one judicial member and the judicial members of any panel are entirely responsible for the settling of such legal questions as may arise in the course of a proceeding. Section 12 of the Act provides:

12. (1) In any proceedings before the Tribunal,

(a) questions of law shall be determined only by the judicial members sitting in those proceedings; and

(b) questions of fact or mixed law and fact shall be determined by all the members sitting in those proceedings.

Consequently, an appeal from the Tribunal on a question of law is akin to an appeal from the Trial Division of the Federal Court. What is more, an appeal lies from any decision of the Tribunal on a question of law, and no privative clause protects the Tribunal’s decisions. The Competition Tribunal Act provides:

13. (1) Subject to subsection (2), an appeal lies to the Federal Court of Appeal from any decision or order, whether final, interlocutory or interim, of the Tribunal as if it were a judgment of the Federal Court ‑- Trial Division.

Robertson J.A. further stressed that the judicial members of the Tribunal are not more expert in matters of law than are judges of the Federal Court of Appeal.

25 Invoking the power of the Federal Court of Appeal to substitute its own findings for those of a tribunal, Robertson J.A. held that the evidence before the Tribunal of the functional interchangeability of daily and community newspapers and of inter-industry competition was more than sufficient to show that the two kinds of newspapers are in the same market. Accordingly, he remitted the matter back to the Tribunal with instructions that it should inquire whether the acquisition of the North Shore News, the Vancouver Courier, and the Real Estate Weekly had resulted in a substantial lessening of competition in the market for retail print advertising in the Lower Mainland of British Columbia.

(ii) As to remedy (1992), 1992 CanLII 15495 (CT), 47 C.P.R. (3d) 240

26 Turning to Southam’s appeal of the remedy, Robertson J.A. declined to decide what the appropriate test for a remedy is, because Southam’s proposed remedy failed regardless of the test applied. In answer to Southam’s protest that the Tribunal had imposed a penalty on it, Robertson J.A. observed that the Tribunal had sought only to impose an effective remedy. To his mind, this way of proceeding could not be objectionable. Against the complaint that the Tribunal had wrongly placed the burden of proving the effectiveness of its proposed remedy on Southam, Robertson J.A. invoked the maxim that he who asserts must prove. To Southam’s argument that the Tribunal had wrongly dismissed its proposed remedy as ineffective, he said that curial deference was due to the Tribunal on this, a finding of mixed law and fact.

4. Issues

4. Issues

27 This appeal raises two issues. The first is whether the Federal Court of Appeal erred in concluding that it owed no deference to the Tribunal’s finding about the dimensions of the relevant market and in subsequently substituting for that finding one of its own. The second is whether the Federal Court of Appeal erred in refusing to set aside the Tribunal’s remedial order.

5. Analysis

5. Analysis

28 The principal question in this appeal concerns the limits that an appellate court should observe in deciding a statutory appeal from a decision like the one that the Tribunal reached in this case. Ultimately, this comes down to a question about the standard of review that an appellate court should apply in a case such as this one. In the reasons that follow, the answer given is that the Tribunal should be held to the standard of reasonableness simpliciter. In other words, a court, in reviewing the Tribunal’s decision, must inquire whether that decision was reasonable. If it was, then the decision should stand. Otherwise, it must fall.

29 The secondary question is whether the Tribunal chose an appropriate remedy. My conclusion is that, even though the Tribunal imposed too strict a test, its chosen remedy is appropriate.

A. Statutory Right of Appeal

A. Statutory Right of Appeal

30 In Pezim v. British Columbia (Superintendent of Brokers), 1994 CanLII 103 (SCC), [1994] 2 S.C.R. 557, a decision which, like this one, concerned a decision of an expert tribunal that was subject to a statutory right of appeal, the Court declared that the standard of review is a function of many factors. Depending on how the factors play out in a particular instance, the standard may fall somewhere between correctness, at the more exacting end of the spectrum, and patently unreasonable, at the more deferential end. See pp. 589-90.

31 An appellate court must consider the factors with a view to determining the approach that it should take as a court sitting in appeal of the decision of the tribunal. There is no privative clause, and so jurisdiction is not at issue. The tribunal enjoys jurisdiction by virtue of its constating statute and the appellate court enjoys jurisdiction by virtue of a statutory right of appeal. The legislative intent is clear. The question is what limits an appellate court should observe in the exercise of its statutorily mandated appellate function.

32 I wish to emphasize that in cases like the instant appeal no question arises about the extent of the tribunal’s jurisdiction. Where the statute confers a right of appeal, an appellate court need not look to see whether the tribunal has exceeded its jurisdiction by breaching the rules of natural justice or by rendering a decision that is patently unreasonable. The manner and standard of review will be determined in the way that appellate courts generally determine the posture they will take with respect to the decisions of courts below. In particular, appellate courts must have regard to the nature of the problem, to the applicable law properly interpreted in the light of its purpose, and to the expertise of the tribunal.

33 I propose to consider each of the relevant factors in turn.

B. The Nature of the Problem Before the Tribunal

B. The Nature of the Problem Before the Tribunal

34 The parties vigorously dispute the nature of the problem before the Tribunal. The appellants say that the problem is one of fact. The respondent insists that the problem is one of law. In my view, the problem is one of mixed law and fact.

35 Section 12(1) of the Competition Tribunal Act contemplates a tripartite classification of questions before the Tribunal into questions of law, questions of fact, and questions of mixed law and fact. Briefly stated, questions of law are questions about what the correct legal test is; questions of fact are questions about what actually took place between the parties; and questions of mixed law and fact are questions about whether the facts satisfy the legal tests. A simple example will illustrate these concepts. In the law of tort, the question what “negligence” means is a question of law. The question whether the defendant did this or that is a question of fact. And, once it has been decided that the applicable standard is one of negligence, the question whether the defendant satisfied the appropriate standard of care is a question of mixed law and fact. I recognize, however, that the distinction between law on the one hand and mixed law and fact on the other is difficult. On occasion, what appears to be mixed law and fact turns out to be law, or vice versa.

36 For example, the majority of the British Columbia Court of Appeal in Pezim, supra, concluded that it was an error of law to regard newly acquired information on the value of assets as a “material change” in the affairs of a company. It was common ground in that case that the proper test was whether the information constituted a material change; the argument was about whether the acquisition of information of a certain kind qualified as such a change. To some extent, then, the question resembled one of mixed law and fact. But the question was one of law, in part because the words in question were present in a statutory provision and questions of statutory interpretation are generally questions of law, but also because the point in controversy was one that might potentially arise in many cases in the future: the argument was about kinds of information and not merely about the particular information that was at issue in that case. The rule on which the British Columbia Securities Commission seemed to rely -- that newly acquired information about the value of assets can constitute a material change -- was a matter of law, because it had the potential to apply widely to many cases.

37 By contrast, the matrices of facts at issue in some cases are so particular, indeed so unique, that decisions about whether they satisfy legal tests do not have any great precedential value. If a court were to decide that driving at a certain speed on a certain road under certain conditions was negligent, its decision would not have any great value as a precedent. In short, as the level of generality of the challenged proposition approaches utter particularity, the matter approaches pure application, and hence draws nigh to being an unqualified question of mixed law and fact. See R. P. Kerans, Standards of Review Employed by Appellate Courts (1994), at pp. 103-108. Of course, it is not easy to say precisely where the line should be drawn; though in most cases it should be sufficiently clear whether the dispute is over a general proposition that might qualify as a principle of law or over a very particular set of circumstances that is not apt to be of much interest to judges and lawyers in the future.

38 Part of the confusion in this case arises from the fact that the parties are arguing about two different questions. On the surface, it appears that the parties agree about the law: both say that, in determining the dimensions of the relevant market, the Tribunal must consider indirect evidence of cross-elasticity of demand. No one quarrels with the Tribunal’s understanding of the kinds of indirect evidence it should consider.

39 However, the respondent says that, having informed itself correctly on the law, the Tribunal proceeded nevertheless to ignore certain kinds of indirect evidence. Because the Tribunal must be judged according to what it does and not according to what it says, the import of the respondent’s submission is that the Tribunal erred in law. After all, if a decision-maker says that the correct test requires him or her to consider A, B, C, and D, but in fact the decision-maker considers only A, B, and C, then the outcome is as if he or she had applied a law that required consideration of only A, B, and C. If the correct test requires him or her to consider D as well, then the decision-maker has in effect applied the wrong law, and so has made an error of law.

40 The appellants, for their part, maintain that the Tribunal considered all the relevant kinds of indirect evidence, including the kinds that the respondent says it ignored. Accordingly, the appellants argue that if the Tribunal erred, it can only have been in applying the correct legal test to the facts. Such an error, say the appellants, is an error of fact. As authority for their position, they cite a passage from the decision of this Court in R. v. Nova Scotia Pharmaceutical Society, 1992 CanLII 72 (SCC), [1992] 2 S.C.R. 606, at p. 647:

In the context of s. 32(1)(c), the process followed and the criteria used to arrive at a determination of “undueness” are questions of law and as such are reviewable by an appellate court. The application of this process and these criteria, that is the full inquiry, often involving complicated economic issues, into whether the impugned agreement was an undue restriction on competition, remains a question of fact. The general rule that appellate courts should be reluctant to venture into a re-examination of the factual conclusions of the trial judge applies with special force in a complex matter such as here.

41 Both positions, so far as they go, are correct. If the Tribunal did ignore items of evidence that the law requires it to consider, then the Tribunal erred in law. Similarly, if the Tribunal considered all the mandatory kinds of evidence but still reached the wrong conclusion, then its error was one of mixed law and fact. The question, then, becomes whether the Tribunal erred in the way that the respondent says it erred.

42 Even a cursory reading of the Tribunal’s reasons discloses that the Tribunal did not fail to consider relevant items of evidence. The respondent charges -- and the Federal Court of Appeal agreed with him on this point -- that the Tribunal ignored evidence of functional interchangeability and of inter-industry competition. But this overlooks the 14 pages that the Tribunal devoted to functional interchangeability, and the 28 pages that the Tribunal devoted to inter-industry competition. See pp. 191-218 and pp. 225-38. A great part, if not actually the bulk of the Tribunal’s decision is taken up with an examination of the very factors that the respondent says it ignored. Therefore, the Tribunal did not err in law by failing to consider relevant factors.

43 The suggestion remains, however, that the Tribunal might have erred in law by failing to accord adequate weight to certain factors. The problem with this suggestion is that it is inimical to the very notion of a balancing test. A balancing test is a legal rule whose application should be subtle and flexible, but not mechanical. It would be dangerous in the extreme to accord certain kinds of evidence decisive weight as, for example, by saying that evidence of inter-industry competition should always be sufficient to prove that two companies are operating in the same market. A test would be stilted and impossible of application if it purported to assign fixed weights to certain factors as, for example, by saying that evidence of inter-industry competition should weigh 10 times as heavily in the Tribunal’s deliberations as does evidence of physical similarities between the products in question. These sorts of things are not readily quantifiable. They should not be considered as matters of law but should be left initially at least to determination by the Tribunal. The most that can be said, as a matter of law, is that the Tribunal should consider each factor; but the according of weight to the factors should be left to the Tribunal.

44 It seems, then, that if the Tribunal erred, it was in applying the law to the facts; and that is a matter of mixed law and fact. This is especially so if, as here, the legal principle being applied involves a balancing test, because with a typical multi-factored balancing test so many factors weigh in the balance that a duplication of any one set of relevant circumstances in the future is unlikely. At the outside, the decision of the Tribunal in this case stands for the proposition that a large daily newspaper does not compete for retail advertising business with small community newspapers though probably it does not stand even for so general a proposition as that, because the Tribunal’s decision rested in part on its assessment of the behaviour of these parties. Depending as it does so fully on the facts and circumstances of the case, the decision is too particular to have any great value as a general precedent.

45 In short, the Tribunal forged no new legal principle, and so its error, if there was an error, can only have been of mixed law and fact. It should be noted that no one has suggested that the Tribunal erred in its findings of fact. All of this tends to suggest that some measure of deference is owed to the decision of the Tribunal because, to paraphrase what Gonthier J. stated in Nova Scotia Pharmaceutical Society, supra, appellate courts should be reluctant to venture into a re-examination of the conclusions of the Tribunal on questions of mixed law and fact.

C. The Words of the Tribunal’s Constating Statute

C. The Words of the Tribunal’s Constating Statute

46 Section 13 of the CompetitionTribunal Act confers a right of appeal from orders and decisions of the Tribunal:

13. (1) Subject to subsection (2), an appeal lies to the Federal Court of Appeal from any decision or order, whether final, interlocutory or interim, of the Tribunal as if it were a judgment of the Federal Court -- Trial Division.

(2) An appeal on a question of fact lies under subsection (1) only with the leave of the Federal Court of Appeal.

That Parliament granted such a broad, even unfettered right of appeal, as if from a judgment of a trial court, perhaps counsels a less-than-deferential posture for appellate courts than would be appropriate if a privative clause were present. However, as this Court has noted several times recently, the absence of a privative clause does not settle the question. See Pezim, supra, at p. 591; Bell Canada v. Canada (Canadian Radio-Television and Telecommunications Commission), 1989 CanLII 67 (SCC), [1989] 1 S.C.R. 1722, at p. 1746.

D. The Purpose of the Statute that the Tribunal Administers

47 Parliament has described the purpose of the Competition Act in the following terms:

1.1 The purpose of this Act is to maintain and encourage competition in Canada in order to promote the efficiency and adaptability of the Canadian economy, in order to expand opportunities for Canadian participation in world markets while at the same time recognizing the role of foreign competition in Canada, in order to ensure that small and medium-sized enterprises have an equitable opportunity to participate in the Canadian economy and in order to provide consumers with competitive prices and product choices.

Competition Act, s. 1.1, as am. by R.S.C., 1985, c. 19, s. 19 (2nd Supp.).

48 The aims of the Act are more “economic” than they are strictly “legal”. The “efficiency and adaptability of the Canadian economy” and the relationships among Canadian companies and their foreign competitors are matters that business women and men and economists are better able to understand than is a typical judge. Perhaps recognizing this, Parliament created a specialized Competition Tribunal and invested it with responsibility for the administration of the civil part of the Competition Act. See Competition Tribunal Act, s. 8(1).

49 This Court has said in the past that the Tribunal is especially well-suited to the task of overseeing a complex statutory scheme whose objectives are peculiarly economic:

Section 8(1) [of the Competition Tribunal Act] confirms the jurisdiction of the Tribunal over Part VIII. The civil part of the [Competition Act] therefore falls entirely under the Tribunal’s jurisdiction. It is readily apparent from the [Competition Act] and the [Competition Tribunal Act] that Parliament created the Tribunal as a specialized body to deal solely and exclusively with Part VIII [of the Competition Act], since it involves complex issues of competition law, such as abuses of dominant position and mergers.

Chrysler Canada Ltd. v. Canada (Competition Tribunal), 1992 CanLII 68 (SCC), [1992] 2 S.C.R. 394, at p. 406.

Because an appellate court is likely to encounter difficulties in understanding the economic and commercial ramifications of the Tribunal’s decisions and consequently to be less able to secure the fulfilment of the purpose of the Competition Act than is the Tribunal, the natural inference is that the purpose of the Act is better served by appellate deference to the Tribunal’s decisions.

E. The Area of the Tribunal’s Expertise

E. The Area of the Tribunal’s Expertise

50 Expertise, which in this case overlaps with the purpose of the statute that the tribunal administers, is the most important of the factors that a court must consider in settling on a standard of review. This Court has said as much several times before, though perhaps never so clearly as in the following passage, from United Brotherhood of Carpenters and Joiners of America, Local 579 v. Bradco Construction Ltd., 1993 CanLII 88 (SCC), [1993] 2 S.C.R. 316, at p. 335:

. . . the expertise of the tribunal is of the utmost importance in determining the intention of the legislator with respect to the degree of deference to be shown to a tribunal’s decision in the absence of a full privative clause. Even where the tribunal’s enabling statute provides explicitly for appellate review, as was the case in Bell Canada . . ., it has been stressed that deference should be shown by the appellate tribunal to the opinions of the specialized lower tribunal on matters squarely within its jurisdiction.

51 As I have already said, the Tribunal’s expertise lies in economics and in commerce. The Tribunal comprises not more than four judicial members, all of whom are judges of the Federal Court -- Trial Division, and not more than eight lay members, who are appointed on the advice of a council of persons learned in “economics, industry, commerce or public affairs”. See Competition Tribunal Act, s. 3. The preponderance of lay members reflects the judgment of Parliament that, for purposes of administering the Competition Act, economic or commercial expertise is more desirable and important than legal acumen.

52 The particular dispute in this case concerns the definition of the relevant product market -- a matter that falls squarely within the area of the Tribunal’s economic or commercial expertise. Undeniably, the determination of cross-elasticity of demand, which is in theory the truest indicium of the dimensions of a product market, requires some economic or statistical skill. But even an assessment of indirect evidence of substitutability, such as evidence that two kinds of products are functionally interchangeable, needs a variety of discernment that has more to do with business experience than with legal training. Someone with experience in business will be better able to predict likely consumer behaviour than a judge will be. What is more, indirect evidence is useful only as a surrogate for cross-elasticity of demand, so that what is required in the end is an assessment of the economic significance of the evidence; and to this task an economist is almost by definition better suited than is a judge.

53 All of this is not to say that judges are somehow incompetent in matters of competition law. Significantly, Parliament mandated that the Tribunal should include judicial members, and that the Chairman should always be a judge. See Competition Tribunal Act, s. 4. Clearly it was Parliament’s view that questions of competition law are not altogether beyond the ken of judges. However, one of the principal roles of the judicial members is to decide such questions of pure law as may arise before the Tribunal. Over those questions they have exclusive jurisdiction. See supra at s. 12(1)(a). But over questions of fact and of mixed law and fact, the judicial members share their jurisdiction with the lay members. See, supra, at s. 12(1)(b). Thus, while judges are able to pronounce on questions of the latter kind, they may do so only together with the lay members; and, in a typically constituted panel, such as the one that sat in this case, the lay members outnumber the judicial ones, so that in the event of a disagreement between the two camps, the lay members as a group will prevail. This makes sense because, as I have observed, the expertise of the lay members is invaluable in the application of the principles of competition law.

F. The Standard

F. The Standard

54 In my view, considering all of the factors I have canvassed, what is dictated is a standard more deferential than correctness but less deferential than “not patently unreasonable”. Several considerations counsel deference: the fact that the dispute is over a question of mixed law and fact; the fact that the purpose of the Competition Act is broadly economic, and so is better served by the exercise of economic judgment; and the fact that the application of principles of competition law falls squarely within the area of the Tribunal’s expertise. Other considerations counsel a more exacting form of review: the existence of an unfettered statutory right of appeal from decisions of the Tribunal and the presence of judges on the Tribunal. Because there are indications both ways, the proper standard of review falls somewhere between the ends of the spectrum. Because the expertise of the Tribunal, which is the most important consideration, suggests deference, a posture more deferential than exacting is warranted.

55 I wish to emphasize that the need to find a middle ground in cases like this one is almost a necessary consequence of our standard-of-review jurisprudence. Because appeal lies by statutory right from the Tribunal’s decisions on questions of mixed law and fact, the reviewing court need not confine itself to the search for errors that are patently unreasonable. The standard of patent unreasonableness is principally a jurisdictional test and, as I have said, the statutory right of appeal puts the jurisdictional question to rest. See Canadian Union of Public Employees, Local 963 v. New Brunswick Liquor Corp., 1979 CanLII 23 (SCC), [1979] 2 S.C.R. 227, at p. 237. But on the other hand, appeal from a decision of an expert tribunal is not exactly like appeal from a decision of a trial court. Presumably if Parliament entrusts a certain matter to a tribunal and not (initially at least) to the courts, it is because the tribunal enjoys some advantage that judges do not. For that reason alone, review of the decision of a tribunal should often be on a standard more deferential than correctness. Accordingly, a third standard is needed.

56 I conclude that the third standard should be whether the decision of the Tribunal is unreasonable. This test is to be distinguished from the most deferential standard of review, which requires courts to consider whether a tribunal’s decision is patently unreasonable. An unreasonable decision is one that, in the main, is not supported by any reasons that can stand up to a somewhat probing examination. Accordingly, a court reviewing a conclusion on the reasonableness standard must look to see whether any reasons support it. The defect, if there is one, could presumably be in the evidentiary foundation itself or in the logical process by which conclusions are sought to be drawn from it. An example of the former kind of defect would be an assumption that had no basis in the evidence, or that was contrary to the overwhelming weight of the evidence. An example of the latter kind of defect would be a contradiction in the premises or an invalid inference.

57 The difference between “unreasonable” and “patently unreasonable” lies in the immediacy or obviousness of the defect. If the defect is apparent on the face of the tribunal’s reasons, then the tribunal’s decision is patently unreasonable. But if it takes some significant searching or testing to find the defect, then the decision is unreasonable but not patently unreasonable. As Cory J. observed in Canada (Attorney General) v. Public Service Alliance of Canada, 1993 CanLII 125 (SCC), [1993] 1 S.C.R. 941, at p. 963, “[i]n the Shorter Oxford English Dictionary ‘patently’, an adverb, is defined as ‘openly, evidently, clearly’”. This is not to say, of course, that judges reviewing a decision on the standard of patent unreasonableness may not examine the record. If the decision under review is sufficiently difficult, then perhaps a great deal of reading and thinking will be required before the judge will be able to grasp the dimensions of the problem. See National Corn Growers Assn. v. Canada (Import Tribunal), 1990 CanLII 49 (SCC), [1990] 2 S.C.R. 1324, at p. 1370, per Gonthier J.; see also Toronto (City) Board of Education v. O.S.S.T.F., District 15, 1997 CanLII 378 (SCC), [1997] 1 S.C.R. 487, at para. 47, per Cory J. But once the lines of the problem have come into focus, if the decision is patently unreasonable, then the unreasonableness will be evident.

58 The standard of reasonableness simpliciter is the same standard that was applied in Pezim, and for good reason: the parallels between this case and that one are obvious. Pezim involved the decision of a securities commission, one of whose tasks was to be sensitive to and enhance capital market efficiency; this appeal involves the decision of the Tribunal, one of whose tasks is to recognize and in its own way to promote the efficiency of the Canadian economy. In Pezim, appeals from decisions of the securities commission lay as of right; in this case, appeals from decisions of the Tribunal lie as of right. The questions in Pezim were entirely within the competence of the commission to answer; the question in this appeal is entirely within the competence of the Tribunal to answer. The principal difference between Pezim and this case is that Pezim involved what were called questions of law. However, as I have already explained, the questions in that case were questions of law only in a somewhat attenuated sense. The difference between the questions in the two cases is therefore not as great as it might at first seem.

59 The standard of reasonableness simpliciter is also closely akin to the standard that this Court has said should be applied in reviewing findings of fact by trial judges. In Stein v. “Kathy K” (The Ship), 1975 CanLII 146 (SCC), [1976] 2 S.C.R. 802, at p. 806, Ritchie J. described the standard in the following terms:

. . . the accepted approach of a court of appeal is to test the findings [of fact] made at trial on the basis of whether or not they were clearly wrong rather than whether they accorded with that court’s view of the balance of probability. [Emphasis added.]

60 Even as a matter of semantics, the closeness of the “clearly wrong” test to the standard of reasonableness simpliciter is obvious. It is true that many things are wrong that are not unreasonable; but when “clearly” is added to “wrong”, the meaning is brought much nearer to that of “unreasonable”. Consequently, the clearly wrong test represents a striking out from the correctness test in the direction of deference. But the clearly wrong test does not go so far as the standard of patent unreasonableness. For if many things are wrong that are not unreasonable, then many things are clearly wrong that are not patently unreasonable (on the assumption that “clearly” and “patently” are close synonyms). It follows, then, that the clearly wrong test, like the standard of reasonableness simpliciter, falls on the continuum between correctness and the standard of patent unreasonableness. Because the clearly wrong test is familiar to Canadian judges, it may serve as a guide to them in applying the standard of reasonableness simpliciter.

61 Putting all of the foregoing considerations into the balance and taking my cue from this Court’s decisions on the subject, including particularly relatively recent decisions, I am of the view that decisions of the Tribunal should be subject to review on a reasonableness standard. That this standard is appropriate and sensible becomes clear when one considers the complexity of economic life in our country and the need for effective regulatory instruments administered by those most knowledgeable and informed about what is being regulated. It bears noting, however, that the standard I have chosen permits recourse to the courts for judicial intervention in cases in which the Tribunal has been shown to have acted unreasonably.

62 In the final result, the standard of reasonableness simply instructs reviewing courts to accord considerable weight to the views of tribunals about matters with respect to which they have significant expertise. While a policy of deference to expertise may take the form of a particular standard of review, at bottom the issue is the weight that should be accorded to expert opinions. In other words, deference in terms of a “standard of reasonableness” and deference in terms of “weight” are two sides of the same coin. In this respect, I agree with Kerans, supra, at p. 17, who has described deference to expertise in the following way:

Experts, in our society, are called that precisely because they can arrive at well-informed and rational conclusions. If that is so, they should be able to explain, to a fair-minded but less well-informed observer, the reasons for their conclusions. If they cannot, they are not very expert. If something is worth knowing and relying upon, it is worth telling. Expertise commands deference only when the expert is coherent. Expertise loses a right to deference when it is not defensible. That said, it seems obvious that [appellate courts] manifestly must give great weight to cogent views thus articulated. [Emphasis added.]

G. Application of the Standard

G. Application of the Standard

63 The question, then, is whether the Tribunal acted unreasonably when it decided that Southam’s daily newspapers and community newspapers are in different product markets. I conclude that it did not.

64 The Federal Court of Appeal identified what it thought were two defects in the Tribunal’s decision. The first is that the Tribunal failed to consider evidence that daily newspapers and community newspapers are functionally interchangeable. The second is that the Tribunal failed to consider evidence that Southam considered the community newspapers to be its principal rivals in the Lower Mainland.

65 By “functional interchangeability”, the Federal Court of Appeal apparently meant “end use” or “purpose”. See pp. 636-37. The Tribunal, for its part, elaborated (at pp. 225-38) at great length on the use to which advertisers put daily and community newspapers. At the end of 14 pages, it came to the conclusion with which the Federal Court of Appeal would later take issue: that advertisers use daily newspapers to reach consumers throughout the entire Lower Mainland and use community newspapers to reach smaller, “local” audiences.

66 The Federal Court of Appeal quarrelled with this conclusion on several grounds. Its first, and most general objection, was to the weight that the Tribunal assigned to the criterion of functional interchangeability. In the court’s view, at p. 635, the Tribunal gave this important criterion short shrift: “the Tribunal clearly failed to consider the importance of functional interchangeability, which is not simply one of many criteria to be considered but a central part of the framework”. However, as I have already noted, the weighing of criteria in a balancing test must be largely a matter of discretion. The very purpose of a multi-factored test, such as the one that the Tribunal used to determine the dimensions of the relevant product market, is to permit triers of fact to do justice in diverse particular cases.

67 As a general matter, in cases like this one, the aims and objectives of the statute may not be served by assigning principal or overriding importance to any one factor. It cannot be said as a matter of law that evidence of functional interchangeability should weigh more heavily in the balance than other kinds of evidence. The question therefore must be whether the Tribunal’s attention to functional interchangeability was reasonable on the facts of this case.

68 For my part, I cannot say that the Tribunal acted unreasonably to discount the evidence of functional interchangeability. It had its reasons for doing so, and those reasons cannot be said to be without foundation or logical coherence. In particular, the Tribunal seems to have thought that daily newspapers and community newspapers serve different purposes. The former appeal to large advertisers who wish to convey their message throughout a metropolitan region. The latter appeal to smaller advertisers, who wish to reach all or many of the consumers living in a particular neighbourhood or district of a city. See the Tribunal’s decision at p. 238. While I might not agree, as a matter of empirical “fact”, that this description of the purposes of the respective kinds of newspaper is exhaustive, I think that it is not without its reasons. It is reasonable, if only reasonable, to suppose that advertisers are sufficiently discerning about the media they employ that they are unlikely to respond to changes in the relative prices of the two kinds of newspaper by taking their business from the one to the other. Fortunately for the Tribunal, its decision need only be reasonable and not necessarily correct.

69 However, that does not finish the matter. The Federal Court of Appeal had two other difficulties with the Tribunal’s approach, and they appear to go to the reasoning that underlies the Tribunal’s conclusion. The first is that it is inconsistent to lump together daily newspapers and community newspapers for purposes of distinguishing them from broadcast media but then to separate the two kinds of newspapers for purposes of distinguishing them from one another. The second is that the Tribunal’s conclusion confuses geographical scope with purpose. Both alleged difficulties turn out on closer inspection not to be troubling.

70 The Federal Court of Appeal, at p. 636, described the first alleged difficulty in these terms: “If ‘multiple price/product’ advertising is a relevant purpose for distinguishing between print and electronic media then it must also be relevant as between advertising in daily and community newspapers”. But, with respect, this conclusion does not follow. It is perfectly consistent to distinguish between the broadcast media and the print media on one ground and to distinguish further between two kinds of print media on another ground. Broadcasters attract advertisers who want to convey an “image”. See the Tribunal’s decision at p. 221. Newspapers attract advertisers who want to convey a great deal of specific information about a variety of products all at once. Accordingly, the two kinds of media serve different markets. However, from the fact that newspapers in general serve a certain broad class of advertiser, it does not follow that all newspapers serve precisely the same particular advertisers, or the same relevant advertising markets. Further division of the market is possible. Thus, daily newspapers serve advertisers who wish to reach even a relatively small proportion of people throughout a large region. Community newspapers serve advertisers who wish to reach a large proportion of people in a small region. See, supra, at p. 238. These markets are at least possibly, and therefore reasonably, different.

71 If the identification of an overarching, broad purpose that two kinds of products serve were sufficient to place those products in the same market, then all products could be placed in the same market, because all products serve the general purpose of satisfying consumers’ needs. Certainly, following the Federal Court of Appeal’s reasoning it would be possible to argue that broadcast media and print media are in the same market because both kinds of media serve advertisers. But it is not so, and the Federal Court of Appeal admitted at p. 636 that it is not so. The trick is to settle on the correct level of generality. Canadian courts have recognized as much in the past:

. . . speaking generally, it is of importance to bear in mind that the term “market” is a relative concept. In one sense, there is only one market in an economy since, to some extent, all products and services are substitutes for each other in competing for the customer’s dollar.

In another sense, almost every firm has its own market since, in most industries, each firm’s product is differentiated, to some extent, from that of all other firms.

Defining the relevant market in any particular case, therefore, requires a balanced consideration of a number of characteristics or dimensions to meet the analytical needs of the specific matter under consideration.

The Queen v. J. W. Mills & Son Ltd., 1968 CanLII 1279 (CA EXC), [1968] 2 Ex. C.R. 275, at p. 305.

72 What has to be kept in mind is that purposes are as various as markets, and both come in different sizes. Consequently it is unhelpful to suggest that once a purpose has been identified, all those products that serve that purpose should be considered to fall within a single market. It is the correct or relevant purpose that must be found, which is to say the broadest purpose that is consistent with a high cross-elasticity of demand. For example, cars and tanks both serve the general purpose of conveying people from place to place. But no one would suggest that cars and tanks are in the same market. The reason is that consumers do not modify their car-purchasing behaviour in response to slight changes in the price of tanks, and governments do not modify their tank-purchasing behaviour in response to slight changes in the price of cars. A person who is in the market for a station wagon does not shop with an eye on the price of armaments. Again, the Minister of National Defence does not check prices at local car dealerships before announcing an acquisition of new military hardware.

73 The relevant purpose is a function of the psychology of consumption or preference. Consequently, in order to choose the relevant purpose, the adjudicator must possess in advance some idea about the behaviour of consumers. In this way, the purpose inquiry is a little circular. Tribunals inquire into purpose in order to get a grip on the tendency of consumers to substitute one product for another, but they will not hit on the right purpose unless they already have a notion of what consumers will substitute for what. This circularity does not, however, alter the fact that more is needed to establish functional interchangeability than citation of a common purpose. That daily newspapers and community newspapers both seek the trade of “multiple price/product” advertisers does not show, without more, that they are competing in the same market. It was open to the Tribunal to conclude, after consulting evidence of the behaviour of advertisers, that purchasing decisions in the real world are taken on the basis of some more particular purpose than to convey information about several products at once.

74 The Federal Court of Appeal at p. 636-37 also took issue, at a theoretical level, with the Tribunal’s attention to the geographic scope of the different kinds of newspapers:

But the fact that the community newspapers are more local in nature does not go to the question of functional interchangeability, but to the behaviour of buyers as to preference for geographical scope. This latter subjective factor should not be mingled with the purely objective factor of functional interchangeability which focuses on use or purpose.

Immediately, any argument that depends on a classification of purpose as “objective” is suspect. Purpose is at least, in part, a matter of intention and so is at least, in part, “subjective”. Presumably, almost any object can be put to a multitude of uses. An axe handle, for example, can serve as a bludgeon or as an axe handle. The purpose it serves depends on the intention of the person in whose hand it is. In like manner, the purposes daily newspapers and community newspapers serve depend on the intentions of their users.

75 In the right hands, both could function as birdcage liners or as wrapping for fish and chips. At times, both probably do. However, those functions are uninteresting because they are atypical, and the Tribunal was right not to mention them. But in order to exclude those purposes and settle on the relevant ones, the Tribunal had to consider, at least implicitly, the intentions of the users of the two kinds of newspaper. Therefore, it was not illegitimate for the Tribunal to look to what the Federal Court of Appeal at p. 636 called “preference for geographical scope”. Reaching consumers throughout a large region is one purpose. Reaching consumers in a neighbourhood is another purpose. It does not matter that the difference between them is in the intention of the advertiser. Intention is a component of purpose. Of course, “objective” considerations also play a part. A newspaper cannot be an aircraft, however much someone might wish that it could be. And this is reflected in the Tribunal’s distinction. A community newspaper cannot reach a large audience, however much an advertiser might wish that it could, and a daily newspaper cannot reach only the consumers in a small locality.

76 It appears, then, that the Tribunal considered at length, at much greater length than did the Federal Court of Appeal, whether daily newspapers and community newspapers serve the same purpose. It concluded that they do not, and gave reasons for its conclusion. The reasons that the Federal Court of Appeal offered for questioning that conclusion are, with respect, unconvincing. Accordingly, failing the appearance of some other basic objection to the Tribunal’s conclusion about functional interchangeability, that conclusion should stand.

77 The Federal Court of Appeal also found fault with the Tribunal’s treatment of evidence that Southam regarded the community newspapers as its chief competitors. In particular, it objected to the Tribunal’s preference for a “more focused analysis” of the evidence of inter-industry competition. In the court’s view at p. 638, “[t]he evidence of broad competitiveness is sufficient to show that there is competition in fact between the Pacific Dailies and the community newspapers”. It was error, said the Federal Court of Appeal, for the Tribunal to ignore that evidence.

78 In fact, the Tribunal devoted 28 pages of its reasons (at pp. 191-218) to the question of inter-industry competition. The Tribunal did not “ignore” evidence of broad inter-industry competition. It simply did not regard that evidence as decisive (at pp. 191-92):

. . . determining that Pacific Press regarded the community newspapers as “competitors” is not by itself enough to place them in the same market. Competition means many things to many people. What the tribunal must establish is whether dailies and the community newspapers are in the same product market for the purposes of assessing the implications of the acquisitions in question in this case. As discussed above in general terms, that exercise involves resolving whether dailies and community newspapers are effective substitutes for newspaper retail advertising services. The actions taken and the views expressed by participants in the alleged market are recognized by both parties and by expert witnesses as an important source of information in trying to answer this question. [Emphasis added.]

In short, the Tribunal found that although evidence of inter-industry competition suggests a certain conclusion, it is not sufficient by itself to establish that conclusion. In this it relied on the elementary principle that thinking something is so does not make it so. A company can believe that it is competing with another company without it actually (or legally) being so.

79 It is possible that if I were deciding this case de novo, I might not dismiss so readily as the Tribunal did what is admittedly weighty evidence of inter-industry competition. In my view, it is very revealing that Southam’s own expert, an American newspaper consultant, identified the community newspapers as the source of Southam’s difficulties in the Lower Mainland. To find, in the face of such evidence, that the daily newspapers and the community newspapers are not competitors is perhaps unusual. In that sense, the Tribunal’s finding is difficult to accept. However, it is not unreasonable. The Tribunal explained that, in its view, Southam was mistaken about who its competitors were; and though I may not consider that reason compelling, I cannot say that it is not a reason for which there is a logical and evidentiary underpinning. More generally, I notice that the Tribunal seems to have been preoccupied with the definition of the relevant market. It is possible that the members may occasionally have lost sight of the ultimate inquiry, which is whether the acquisition of the community newspapers by Southam substantially lessened competition. But again, I cannot say that the Tribunal’s approach was unreasonable. Definition of the relevant market is indeed a necessary step in the inquiry; and the fact that the Tribunal dwelled on it is perhaps understandable if, as seems to have been the case, the bounds of the relevant market were not clear.

80 I wish to observe, by way of concluding my discussion of this issue, that a reviewer, and even one who has embarked upon review on a standard of reasonableness simpliciter, will often be tempted to find some way to intervene when the reviewer him- or herself would have come to a conclusion opposite to the tribunal’s. Appellate courts must resist such temptations. My statement that I might not have come to the same conclusion as the Tribunal should not be taken as an invitation to appellate courts to intervene in cases such as this one but rather as a caution against such intervention and a call for restraint. Judicial restraint is needed if a cohesive, rational, and, I believe, sensible system of judicial review is to be fashioned.

81 Accordingly, the Tribunal’s conclusion must stand.

H. Remedy

H. Remedy

82 Having found that Southam’s acquisitions had produced a substantial lessening of competition in the market for real estate print advertising on the North Shore, the Tribunal ordered Southam to divest itself, at its own option, of either the Real Estate Weekly or the North Shore News. The Federal Court of Appeal declined to disturb this remedy. I agree with the Federal Court of Appeal that the remedy settled upon by the Tribunal should be allowed to stand.

83 The appellants submit that the correct test for a remedy under the Competition Act is whether it eliminates any substantial lessening of competition that the merger may have caused. The appellants observe that this is the standard that has been applied in cases under s. 92(1)(e)(iii) of the Competition Act, in which the parties have consented to the remedy. See, e.g., Canada (Director of Investigation & Research) v. Air Canada (1989), 1989 CanLII 10325 (CT), 27 C.P.R. (3d) 476 (Comp. Trib.), at pp. 513-14. They observe also that substantial lessening of competition is the evil that Parliament has sought to address in the Act. Mergers themselves are not considered to be objectionable except in so far as they produce a substantial lessening of competition. Therefore, restoration to the pre-merger situation is not what is wanted. Indeed, presumably some lessening of competition following a merger is tolerated, because the Act proscribes only a substantial lessening of competition. The appellants object further to what they see as the punitive quality of the remedy that the Tribunal imposed, and to what they regard as the illicit shifting to them of the burden of showing that the proposed remedy would be effective.

84 The respondent, for his part, says that the test of a remedy is whether it restores the parties to the pre-merger competitive situation. I believe that the appellants’ test is the better one.

85 The evil to which the drafters of the Competition Act addressed themselves is substantial lessening of competition. See Competition Act, s. 92(1). It hardly needs arguing that the appropriate remedy for a substantial lessening of competition is to restore competition to the point at which it can no longer be said to be substantially less than it was before the merger. This is the test that the Tribunal has applied in consent cases. The Tribunal attempted to distinguish this case from those cases on precisely the ground that here the Director did not consent to the appellants’ proposed remedy. But the distinction is not a sensible one. I can think of only two reasons why the test should be more forgiving where the parties have consented to a remedy. The first is that parties who have not consented should be punished for their obduracy. The second, which is related to the first, is that the law should provide parties with an incentive to come to a consensual arrangement. Neither reason is valid on closer analysis. The burden of a harsh standard falls entirely on one of the parties: the company. No punishment falls on the Director when he or she is obdurate, and the harsh standard gives him or her no incentive to consent to a remedy. Therefore, even if there is a policy of encouraging consent and punishing obduracy, it is not well served by the imposition of a more stringent standard in cases in which the parties have not consented. The better approach is to apply the same standard in contested proceedings as in consent proceedings.

86 However, the appellants do not benefit by their proposed standard. The reason is that the Tribunal expressly found that, even accepting that the appropriate standard is the one used in consent proceedings, Southam’s proposed remedy fails because it would not likely be effective in eliminating the substantial lessening of competition. Robertson J.A. accepted this finding, saying that it was entitled to deference. I agree.

87 The Tribunal’s choice of remedy is a matter of mixed law and fact. The question whether a particular remedy eliminates the substantial lessening of competition is a matter of the application of a legal standard to a particular set of facts. Therefore, for reasons I have already given, the Tribunal’s decision must be reviewed according to a standard of reasonableness.

88 Because the Tribunal did not decide unreasonably when it decided that Southam’s proposed remedy would not be effective, its decision should be allowed to stand. What Southam proposed was that it should sell the real estate supplement that appears weekly in the North Shore News. But, as the Tribunal very properly pointed out, it is not clear that the supplement would prosper or even survive on its own. Even if the supplement continued to enjoy the advantages of a close association with the North Shore News, the closeness of the association would not tend to foster competition. See the Tribunal’s decision, supra, at p. 252.

89 The appellants’ other objections to the remedy are unconvincing. The remedy is not punitive, because the Tribunal found that it was the only effective remedy. If the choice is between a remedy that goes farther than is strictly necessary to restore competition to an acceptable level and a remedy that does not go far enough even to reach the acceptable level, then surely the former option must be preferred. At the very least, a remedy must be effective. If the least intrusive of the possible effective remedies overshoots the mark, that is perhaps unfortunate but, from a legal point of view, such a remedy is not defective. As for the claim that the Tribunal wrongly required the appellants to demonstrate the effectiveness of their proposed remedy, no more need be said than that he who asserts should prove, as Robertson J.A. so aptly put it ((1995), 1995 CanLII 11081 (FCA), 127 D.L.R., (4th) 329) at p. 337.

90 Therefore, I would dismiss the appeal of the remedy.

6. Conclusion

6. Conclusion

91 The Tribunal decided that the acquisition by Southam of several community newspapers did not substantially lessen competition in the market for retail print advertising in the Lower Mainland of British Columbia. That decision is entitled to deference. Because it is not unreasonable, it must be allowed to stand.

92 Accordingly, I would allow the appeal on the merits with costs throughout, set aside the judgment of the Federal Court of Appeal, and restore the order of the Tribunal. I would dismiss the appeal on the remedy with costs.