Federal Court of Appeal Decisions

Date: 20010404

Docket: A-533-00

Neutral citation: 2001 FCA 104

CORAM:        STONE J.A.

LÉTOURNEAU J.A.

EVANS J.A.

BETWEEN:

                                       THE COMMISSIONER OF COMPETITION

                                                                                                                                            Appellant

                                                                         - and -

                               SUPERIOR PROPANE INC. and ICG PROPANE INC.

                                                                                                                                      Respondents

                               Heard at Ottawa, Ontario, on January 9, 10 and 11, 2001.

                               Judgment delivered at Ottawa, Ontario, on April 4, 2001.

REASONS FOR JUDGMENT BY:                                                                             EVANS J.A.

CONCURRED IN BY:                                                                                                 STONE J.A.

DISSENTING REASONS IN PART BY:                                                     LÉTOURNEAU J.A.

Date: 20010404

Docket: A-533-00

Neutral citation: 2001 FCA 104

CORAM:        STONE J.A.

LÉTOURNEAU J.A.

EVANS J.A.

BETWEEN:

                                 THE COMMISSIONER OF COMPETITION

                                                                                                                                Appellant

                                                                   - and -

                         SUPERIOR PROPANE INC. and ICG PROPANE INC.

                                                                                                                          Respondents

                                              REASONS FOR JUDGMENT

EVANS J.A.

A.        INTRODUCTION

[1]                This is an appeal from a decision of the Competition Tribunal ("the Tribunal"), dated August 30, 2000, dismissing an application by the Commissioner of Competition ("the Commissioner") for an order to dissolve the merger of the respondents, Superior Propane Inc. and ICG Propane Inc., or otherwise to remedy the lessening of competition likely to occur in the propane delivery market in Canada as a result of the merger.

[2]                The appeal raises a question of fundamental importance to the administration of the Competition Act that has been the subject of vigorous debate among economists and lawyers in Canada and elsewhere. Indeed, the issue is one on which the Commissioner, and his predecessor, the Director of Investigation and Research, Bureau of Competition Policy, have at different times propounded more than one view. However, the volume of the literature to which it has given rise far exceeds that of the jurisprudence and, prior to the decision under appeal, the question had been the subject of judicial comment in only one case.

[3]                The question concerns the scope of the so-called "efficiency defence". Under this statutory defence, a merger must be permitted, even though it is likely to prevent or substantially lessen competition in a particular market, if the efficiency gains flowing from the merger are greater than, and offset, the effects of the lessening of competition.

[4]                The precise issue raised by this appeal is whether, for the purpose of the efficiency defence, the "effects" of an anti-competitive merger are limited as a matter of law to the loss of resources to the economy as a whole (the deadweight loss), or whether they include a wider range of the effects of a substantial lessening of competition. The latter would include the wealth transfer from consumers to producers that occurs when the merged entity exercises its market power to increase prices above competitive levels, the elimination of smaller competitors from the market, and the creation of a monopoly.

[5]                The Tribunal held that the merger would substantially lessen or prevent competition in nearly all local propane markets in Canada, as well as in the market for national account coordination services associated with the delivery of propane. The total divestiture by Superior of all of ICG's shares and assets was found to be the only appropriate remedy to prevent this. However, by a majority the Tribunal also concluded that, since the merger was likely to result in efficiency gains of $29.2 million, and would result in only $3.0 million of quantitative deadweight loss and $3.0 million of qualitative deadweight loss, the merger was saved by the efficiency defence contained in the Competition Act.

[6]                Using the "total surplus standard", the Tribunal concluded that the deadweight loss was the sole "effect" of the lessening of competition that must be balanced against the efficiency gains. Accordingly, the Tribunal treated as irrelevant all other effects, including the size of the wealth transfer from consumers to Superior as a result of the higher than competitive market prices that Superior was likely to charge for propane as a result of the merger.

B.        LEGISLATIVE FRAMEWORK

[7]              The statutory provisions relevant to this appeal are as follows:

Competition Act, R.S.C. 1985, c. C-34

Competition Tribunal Act, R.S.C. 1985, c. 19 (2^nd Supp.)

C.        THE TRIBUNAL'S DECISION

[8]                The Tribunal heard this matter over 48 days; 39 days were devoted to hearing from 91 witnesses, including 17 experts, at least 10 of whom have a Ph.D. in economics, while submissions from counsel took another 9 days. The reasons for decision of the majority of the Tribunal (the presiding judicial member, Nadon J., and one of the lay members, Dr. Schwartz, an economist) run to some 469 paragraphs. There are also substantial dissenting reasons by the second lay member, Ms. Lloyd, covering, in part, issues that lie at the heart of this appeal.

[9]                The first 317 paragraphs of the majority's reasons, written by Nadon J., deal at length with whether the merger would prevent or substantially lessen competition within the meaning of section 92 of the Competition Act. The Tribunal was unanimous in concluding that it would and, since the Tribunal's conclusion on this is not the subject of appeal, I can deal with its findings relatively briefly.

[10]            First, the Tribunal found that the merger would not substantially lessen competition in only 8 of 74 local markets for the supply of propane: paragraph 307 of the Reasons. At the other extreme, in 16 markets the merged entity would have a monopoly or near monopoly, that is, a market share ranging from 97% to 100%: paragraph 306. And, in another 16 markets, where there was already substantial market concentration, the merger would remove healthy competition: paragraph 308. The remaining 33 markets were in an intermediate category in that, while Superior and ICG were the largest sellers of propane, and the merger was likely to lessen competition substantially, competition from other suppliers would continue after the merger: paragraph 309. Finally, the Tribunal found that the merger would lessen competition substantially in the coordination services offered to national account customers, leaving the merged entity as the only firm in Canada serving this market: paragraph 310.

[11]            Second, the demand for propane is fairly inelastic, that is, consumers are relatively insensitive to price increases. Although some consumers purchase propane for less than essential purposes, such as heating their swimming pools, most purchase it for home heating, automotive fuel and industrial purposes. Consequently, propane is not a discretionary item that most consumers can choose to forego.   

[12]            Moreover, the cost of switching from propane to an alternative form of fuel is relatively high. For example, consumers who purchase propane to heat their homes will normally be deterred from substituting oil as a heating fuel by the considerable expense of converting to an oil burning furnace unless, for instance, their furnace is at the end of its useful life: paragraphs 24-25.

[13]            Third, relatively high barriers to entry face potential competitors in the market and hence increase the ability of the merged entity to raise prices above the competitive level. For example, consumers are often required to sign exclusive supply contracts stipulating that for five years they will purchase propane exclusively from the supplier and that, in the case of Superior, when the contract expires, consumers will give the supplier the right of first refusal. These supply contracts often contain lengthy notice of termination clauses that, in the case of ICG, require consumers to give 180 days notice prior to the termination date of the contract. In the absence of such notice, the contract is automatically renewed: paragraphs 132-146.

[14]            Another factor that makes switching suppliers difficult and costly is that the supplier, rather than the consumer, typically owns the propane tank: paragraph 147. In addition, a reputation for reliable delivery is an important factor in this market and consumers are therefore reluctant to switch to a new supplier with no established reputation: paragraph 154. Finally, new entrants are also likely to be discouraged by the maturity of the market; that is, there is little potential for growth in the demand for propane: paragraph 158.

[15]            In support of these findings on market entry barriers, the Tribunal noted that, when Imperial Oil Limited, a very large corporation, entered the market for propane distribution in 1990, it withdrew after nine years because market barriers made the venture uneconomic. Since then, no other entrants of comparable size or stature have materialized: paragraph 153.

[16]            On the basis of considerations of the kind noted above, the Tribunal concluded that, as a result of the merger, the merged entity was likely to increase the price of propane by an average of 8%: paragraphs 252-253. Having found that the merger would lead to a substantial lessening of competition contrary to section 92, the Tribunal concluded that only a total divestiture by Superior of all ICG's assets and shares would restore competition to the pre-merger level: paragraphs 314 and 316.

[17]            The Tribunal then proceeded to a consideration of the efficiency defence under section 96. It held that the merging parties had the burden of proving the efficiencies that would not have been generated but for the merger, while the Commissioner bore the burden of proving the anti-competitive effects, since he was in the better position to do so by virtue of the investigative powers conferred on him by the Act: paragraph 403. The merging parties had the burden of establishing that the resulting efficiencies would be greater than and offset the anti-competitive effects of the merger.    

[18]            The majority calculated the net efficiency savings that would result from the merger, and could not have been achieved by other means, to be $29.2 million in each of the next ten years: paragraph 383. Ms. Lloyd dissented from the majority's view on this issue and held that the evidence before the Tribunal was insufficient to support this figure: paragraph 470. However, there is no appeal from this aspect of the Tribunal's decision and it is unnecessary to say more about it here.

[19]            Having made its entry on the "efficiency" side of the ledger, the Tribunal then considered the "effects" that would result from the "lessening or prevention of competition" if the merger was approved. The submissions and evidence before the Tribunal on this question went to two issues: the definition of "effects" for the purpose of section 96 and their quantification. The principal question in this appeal concerns the Tribunal's conclusion on the first of these issues.

[20]            The Tribunal had before it evidence describing various methodologies developed by economists for determining the effects of an anti-competitive merger. I should make it clear that the various standards considered by the Tribunal are, for the most part, the work of economists in the United States, and have been used as a basis for competition policy prescriptions. However, antitrust law in the United States does not have an efficiency defence comparable to section 96, although efficiencies are taken into consideration by the Federal Trade Commission when scrutinising a merger, along with other factors, including the likely wealth transfer from consumers to producers likely to result from it.

[21]            Two of the methodologies for determining when efficiency gains offset the adverse effects of an anti-competitive merger are likely to give a narrow scope to the efficiency defence. For example, under the "price standard" efficiencies will only justify an anti-competitive merger if they result in price decreases or, at least, do not increase prices. This is the most difficult standard for the parties to a merger to satisfy, and is the standard normally applied by the Federal Trade Commission as the basis for approving an anti-competitive merger: Horizontal Merger Guidelines (Department of Justice and Federal Trade Commission; April 2, 1992, revised April 8, 1997), pages 148-50.

[22]            The "consumer surplus standard" posits that a merger should be permitted only if the resulting efficiency gains exceed the sum of the wealth transferred to the producers and the deadweight loss occasioned by increases in price charged by the merged entity. In practice, this standard will also be difficult to establish and consequently will tend to narrow the availability of the efficiency defence.

[23]            On the basis of a report prepared for the Commissioner by an expert witness, Dr. Peter Townley, a professor of economics at Acadia University, counsel for the Commissioner submitted that the Tribunal should adopt a "balancing weights standard" as the basis for determining whether the efficiency gains from the merger of Superior and ICG were greater than, and offset, its anti-competitive effects.

[24]            Using this methodology, the Tribunal would determine the anti-competitive effects of a merger by taking into account a range of factors, but would not assign to each a fixed, a priori weight. The factors include: the deadweight loss; the wealth transfer from consumers resulting from the increase in prices through the exercise of market power; the loss of product choices and services currently associated with the product; and the prevention of competition and the creation of a monopoly or near monopoly in some or all of the relevant markets: paragraphs 386-387 and 431.

[25]            The Tribunal rejected this approach in favour of the "total surplus standard" which looks only at the overall loss to the economy as a result of the fall in demand for the merged entity's products following a post-merger increase in price, and the inefficient allocation of resources that occurs when, as prices rise, consumers purchase a less suitable substitute. The resulting loss of resources to the economy constitutes deadweight loss.

[26]            The Tribunal relied on the analyses of leading economists and of "law and economics" scholars, mainly from the United States, but also from Canada, in support of the "total surplus standard". Under this standard, an anti-competitive merger is allowed to proceed when efficiency gains exceed deadweight loss. Its rationale is that this standard measures the net increase or loss in general welfare as a result of the merger. In addition, it provides a predictable standard for merger review, and hence firms are not deterred from effecting mergers that will increase total economic resources by an inability to predict whether their merger will receive regulatory approval.

[27]            Under the total surplus standard, the wealth likely to be transferred from consumers to producers as a result of the merger is not considered to be an anti-competitive effect, because such a transfer is neutral: that is, it neither increases, nor decreases total societal wealth. Proponents of the total surplus standard argue that there is no economic reason for favouring a dollar in the hands of consumers of the products of the merged entity over a dollar in the hands of the producers or its shareholders, who are, after all, also consumers. Moreover, in the absence of complete data on the socio-economic profiles of the consumers and of the shareholders of the producers, it would be impossible to assess whether the redistributive effects of the wealth transferred as a result of the higher prices charged by the merged entity would be fair and equitable: paragraphs 423-425.

[28]            The Tribunal concluded that, properly interpreted, section 96 of the Competition Act mandates a methodology for determining the effects to be balanced against efficiency gains that ignores wealth transfers, or distributive effects, and focusses exclusively on the extent to which the merger increases net wealth in the economy as a whole. The reasons that the Tribunal advanced for its conclusion can be summarized as follows.

[29]            First, even if the necessary data were available, an assessment of the merits, or otherwise, of the distributive effects of a merger is a political task best performed by elected politicians, not by members of the Tribunal, who are appointed for their expertise in economics or commerce: paragraphs 431-432 and 438.

[30]            Second, since section 96 allows an anti-competitive merger where the efficiencies gained thereby are greater than, and offset, the effects of the lessening of competition, efficiency "was Parliament's paramount objective in passing the merger provisions of the Act": paragraph 437. Therefore, "effects" in section 96 should be interpreted in a way that best attains that objective. This excludes an interpretation that requires, or permits, distributive or other effects of a merger to be considered that are unrelated to the maximization of total societal wealth: paragraphs 411-413, 426 and 432.

[31]            Third, if business people are unable to predict whether the Commissioner or the Tribunal is likely to conclude that the efficiencies to be gained from a proposed merger will exceed, and offset, the adverse effects of the merger as calculated by the balancing weights standard, they will be deterred from merging, to the detriment of the economy as a whole: paragraph 433.

[32]            Accordingly, in the Tribunal's view, the difficulty of applying the balancing weights standard advanced by the Commissioner militates against its adoption. Indeed, even though Professor Townley favoured this approach he conceded in his evidence that, as an economist, he could not advise the Tribunal what weights to assign to the various factors to be considered. Hence, he could not say whether the efficiency gains from the merger of Superior and ICG were greater than and offset its effects.

[33]            Fourth, the Tribunal noted that in the Merger Enforcement Guidelines ("MEG") (Director of Investigation and Research, Information Bulletin No. 5, March 1991 (Supply and Services Canada, 1991)), which had been in force since 1991, the Commissioner had indicated that the effects of an anti-competitive merger were to be assessed for the purpose of section 96 by the total surplus standard.

[34]            Indeed, even after the appropriateness of the total surplus standard had been questioned by Reed J. when sitting as the judicial member of the Tribunal in Canada (Director of Investigation and Research) v. Hillsdown Holdings (Canada) Ltd. (1992), 41 C.P.R. (3d) 289 (Comp. Trib.), the predecessor of the current Commissioner publicly reaffirmed the position taken in the MEG. In Hillsdown, supra, Reed J. had doubted (at page 339) whether an interpretation of "effects", such as that contained in the MEG, that omitted from consideration the wealth transferred from consumers to producers was consistent with the purposes of the Act.

[35]            Fifth, the Tribunal stated that the purpose and objectives section of the Competition Act, section 1.1, should not be read as requiring each of the objectives listed in it to be considered in the context of identifying the effects of a merger for the purpose of section 96. Rather, the references in section 1.1 to the Act's objectives, such as promoting "the efficiency and adaptability of the Canadian economy", ensuring that "small and medium-sized enterprises have an equitable opportunity to participate in the Canadian economy", and providing "consumers with competitive prices and product choices" should be regarded as no more than statements of the beneficial results of attaining the stated purpose of the Act, namely "to promote and encourage competition in Canada". Further, to the extent that there was a conflict between the general provision, section 1.1, and the specific, section 96, the latter should prevail: paragraphs 408-410.

[36]            The dissenting member of the Tribunal took issue with much of the majority's reasoning on the meaning of "effects" in section 96. In Ms. Lloyd's view, any interpretation of section 96 that excluded from "effects" the wealth transfer from consumers to producers likely to result from an anti-competitive merger was inconsistent with the objectives of the Act: paragraph 506.

[37]            She concluded that a flexible approach that enabled the Tribunal to take into account, along with other factors, the wealth transfer, both quantitatively and qualitatively, was more compatible with the statutory scheme, particularly in so far as its objectives include "to provide consumers with competitive prices and product choices": paragraph 511. Ms. Lloyd summarized (at paragraph 506) her position as follows:

While I recognize that efficiencies are given special consideration under section 96 and may constitute a defence to an otherwise anti-competitive merger, it appears to me that section 96 is an exception to the application of section 92 of the Act and not an exception to the Act itself. (Emphasis added)                                                           

D.        THE ISSUES

[38]                        The appeal raises three issues for the Court to decide.

           (1)        What standard of review is applicable to the Tribunal's determination of the "effects" of a merger to be considered under section 96?

           (2)        Did the Tribunal err in law when it interpreted "effects" as limited to those identified by the total surplus standard?

           (3)        Did the Tribunal err in law when it imposed on the Director the legal burden of proving the effects of the merger?   

E. ANALYSIS

Issue 1:           The Standard of Review

[39]            It was common ground between counsel that, in view of the Supreme Court of Canada's decision in Canada (Director of Investigation and Research) v. Southam Inc., [1997] 1 S.C.R. 748, if the Tribunal's interpretation of the word, "effects", was entitled to any deference, the less deferential standard of reasonableness simpliciter would be appropriate.

[40]            The disputed question was, of course, whether any deference was due at all. In my view, the answer is to be found, for the most part, in the reasoning in Southam, supra, which also concerned the Tribunal, and in Pushpanathan v. Canada (Minister of Citizenship and Immigration), [1998] 1 S.C.R. 982, which is an important, comprehensive and general elaboration of the pragmatic and functional analysis for determining the standard of judicial review of administrative action.

[41]            I turn, then, to an examination of the elements of the pragmatic and functional analysis as they apply to the case before us. A consensus seems to have emerged in the jurisprudence that the expertise of the tribunal under review, and the relevance of that expertise to resolving the issues in dispute, will normally be the most important of the pragmatic or functional factors considered in determining the standard of review: Pushpanathan, supra, at pages 1006-07, paragraph 32. I deal first with the nature of the issue decided by the Tribunal in this case.

(i) The nature of the issue decided by the Tribunal

[42]            In holding that the meaning of the word, "effects", in section 96 is limited to the deadweight loss resulting from an anti-competitive merger, the Tribunal was clearly interpreting the Act and was thereby deciding a question of law.

[43]            This is because the Tribunal's ruling purports to be of general application to all cases in which the efficiency defence is invoked. The Tribunal did not confine itself merely to identifying the factors to be considered or not considered in this case, nor to prescribing a methodology for determining only the "effects" of Superior's merger with ICG. Instead, the Tribunal makes it abundantly clear in its reasons that, as a matter of interpretation, the word, "effects", means only deadweight loss and that the efficiency defence is, in all cases, simply a codification of the total surplus standard.

[44]            For instance, based on its conclusion that section 96 encapsulates the total surplus standard, the Tribunal made the following findings with respect to the meaning of "the effects" of an anti-competitive merger:

[423]      The economic effects of an anti-competitive merger are the effects on real resources, that is, the changes in the way the economy deploys those resources as the result of the merger.

[427]      Assessing a merger's effects in this way is generally called the "total surplus standard". ...transfers from consumers to shareholders are not counted as losses under the total surplus standard. The anti-competitive effect of the merger is measured solely by the deadweight loss ... Under the total surplus standard, efficiencies need only exceed the deadweight loss to save an anti-competitive merger.

[430] ... The only standard that addresses solely the effects of a merger on economic resources is the total surplus standard.

[447]      The Tribunal further believes that the only effects that can be considered under subsection 96(1) are the effects on resource allocation, as measured in principle by the deadweight loss which takes both quantitative and qualitative effects into account. Accordingly, the Tribunal believes that the total surplus standard is the correct approach for analysing the effects of a merger under subsection 96(1).                                                                                                             

Since none of these statements was made with reference to the particular facts of the case, the Tribunal must have intended its view of the meaning of the word, "effects", to apply whenever the section 96 efficiency defence is raised.

[45]            Referring to the task of distinguishing the interpretation of a statutory standard (normally a question of law) from its application to the facts of a case (often a question of mixed fact and law), Iacobucci J. said in Southam, supra (at page 768, paragraph 37):

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.                                                                                                                             

In a similar vein (supra, at page 767, paragraph 36), he had characterised a question as one of law "because the point in controversy was one that might potentially arise in many cases in the future."

[46]            Applying these observations to this case, I am of the view that, since the Tribunal's determination of what can be considered as an "effect" of the merger of Superior and ICG was intended to be of general application, it would be of "much interest to judges and lawyers", because other panels of the Tribunal will regard it as a legal proposition having considerable persuasive authority whenever they have to consider the efficiency defence under section 96. To use another of Iacobucci J.'s felicitous phrases (Southam, supra, at page 771, paragraph 45), the Tribunal in this case clearly "forged new legal principle".

(ii) The expertise of the Tribunal

[47]            Since the ultimate issue in determining the standard of review is whether the legislature should be taken to have intended the specialist administrative tribunal or the courts to bear primary responsibility for determining the question in dispute, it must be understood that "expertise" is a relative, not an absolute concept:United Brotherhood of Carpenters & Joiners of America, Local 579 v. Bradco Construction Ltd., [1993] 2 S.C.R. 316, at page 335. In assessing the relative expertise of the Tribunal and the Court, I have had regard to the following considerations.

[48]            First, the Tribunal is an adjudicative body. Just as it has done with the administration of human rights legislation, Parliament has divided responsibility for administering the Competition Act between the Competition Bureau, the policy-making, investigative and enforcement agency, headed now by the Commissioner, and the Tribunal, the adjudicative agency. In this respect, the Tribunal is different from multi-functional administrative agencies, such as securities commissions in many provinces, which typically have wide powers that match their regulatory mandate. The absence of broad policy development powers is a factor that limits the scope of the Tribunal's expertise: Pezim v. British Columbia (Superintendent of Brokers), [1994] 2 S.C.R. 557, at page 596.

[49]            Second, expertise may be assessed by reference to the composition of an administrative tribunal. Hearings of the Tribunal are conducted before three to five members, at least one of whom must be a judicial member and one a lay member: Competition Tribunal Act ("CTA"), subsection 10(1). This case was heard by three members: the presiding judicial member and two lay members.

[50]            The judicial member is one of the maximum of four judges of the Trial Division of the Federal Court whom the Governor in Council may appoint to the Tribunal on the recommendation of the Minister of Justice: CTA, paragraph 3(2)(a). In addition to presiding at hearings of the Tribunal, the judicial member alone decides any questions of law that arise before the Tribunal: CTA, paragraph 12(1)(a).

[51]            I note that in the Hillsdown case (supra, at page 337, note 21), Reed J. made it clear that the validity of the definition in the MEG of "effects" involved the interpretation of section 96, and was thus a question of law alone. Hence, the Tribunal's reasons on this issue expressed her view as the judicial member of the Tribunal.

[52]            In contrast, Nadon J. does not state that his determination of the meaning of "effects" is solely his decision. However, since the Act gives to the judicial member sole responsibility for deciding questions of law, the standard of review cannot depend on whether, in a particular case, the lay member's participation in the decision on the legal issue extended beyond consultation.

[53]            A maximum of eight lay members are also appointed to the Tribunal by the Governor in Council on the recommendation of the Minister of Industry: CTA, paragraph 3(2)(b). No qualifications are prescribed for lay members. However, before making a recommendation, the Minister must consult with an Advisory Council comprising not more than ten members, who, the CTA, subsection 3(3) provides, are appointed from those

[54]            It is reasonable to infer from this provision that the Council was expected to recommend the appointment of lay members with a breadth of experience similar to that of the Advisory Council members themselves. Thus, members' fields of expertise need not be limited to economics, but may extend more broadly to public affairs. Further, their perspectives may include not only those of the business communities, including small and medium-sized business, but also of consumer groups and labour.

[55]            Questions of fact, and of mixed fact and law, are decided by all of the members of the panel of the Tribunal hearing a matter: CTA, paragraph 12(1)(b). In addition, even though the judicial member alone decides questions of law, the judicial member may well make his or her rulings after discussing the issues with the lay members and benefiting from whatever contribution they are able to make to the resolution of the legal issue from their perspective and on the basis of their expertise. After all, questions of law are rarely decided in the abstract, and generally require that careful consideration be given to the likely practical consequences and implications of deciding them one way rather than another.

[56]            In short, the composition of the Tribunal indicates a considerable level of expertise. This Court does not defer to decisions of the Trial Division of this Court on questions of law: President and Fellows of Harvard College v. Canada (Commissioner of Patents), [2000] 4 F.C. 528, at paragraph 180 (F.C.A). However, the fact that no more than four members of the Court may be appointed as judicial members suggests that, when sitting as the judicial member of the Tribunal and having the assistance of the lay members, a judge of the Trial Division can be expected to have a level of expertise or experience in this area of the law over and above that acquired by a judge in the ordinary course of judicial work. Nor do I disregard the importance of the understanding of the issues in dispute in this case that the Tribunal would have obtained after conducting 48 days of hearings.

[57]            Indeed, on more than one occasion, the Supreme Court of Canada has recognized (Southam, supra, at pages 772-73, paragraph 49) that the Tribunal

is especially well suited to the task of overseeing a complex statutory scheme whose objectives are peculiarly economic.                                                                               

Iacobucci J. also noted in that case that, since the aims of the Competition Act are "more ‘economic' than ‘legal'" (supra, at page 772, paragraph 48), it was appropriate to conclude that "the Tribunal's expertise lies in economics and commerce" (supra, at page 773, paragraph 51).      

(iii) A question of law within the Tribunal's expertise?

[58]            Counsel for the respondents submitted that characterising a question decided by an administrative tribunal as one of statutory interpretation, and therefore one of law, is not necessarily determinative of the standard of review: see Pushpanathan, supra, at page 1008, paragraph 34. However, it seems to me an obvious inference from the reasons for judgment of Iacobucci J. in Southam, supra, that, when all the factors in the pragmatic or functional mix are weighed together, the fact that the Tribunal in the case before us was deciding a question of law with a high degree of generality tips the scale in the direction of correctness as the applicable standard of review.

[59]            Thus, speaking at the level of principle, Iacobucci J. said (supra, at page 769, paragraph 39) that, if a decision-maker fails to consider all the factors that the legislature required to be considered, "then the decision-maker has in effect applied the wrong law, and so has made an error of law." And, turning to the Tribunal in particular, he said (supra, at page 769, paragraph 41): "If the Tribunal did ignore evidence that the law requires it to consider, then the Tribunal erred in law."

[60]            In my view, there is nothing about the word, "effects", to exclude the general principle that, in the absence of indicators to the contrary, statutory interpretation is a question of law that is reviewable on a standard of correctness. As Bastarache J. said in Pushpanathan (supra, at page 1012, paragraph 38):

Without an implied or express legislative intent to the contrary ... legislatures should be assumed to have left highly generalized propositions of law to courts.                    

[61]            Thus, as a linguistic matter, the word, effects, does not suggest an implicit delegation of authority to the Tribunal to determine what factors must, and must not, be considered in determining what they are. If, as seems to be the case on the basis of the reasoning in Southam, supra, Iacobucci J. would have regarded a general proposition advanced by the Tribunal about the meaning of the word, "market", as subject to review for correctness, the same would seem equally true of the phrase, "the effects of any prevention or lessening of competition". Nor am I persuaded by counsel for the respondents that in Southam (supra, at pages 789-90, paragraphs 83-85) Iacobucci J. applied a standard other than correctness to the Tribunal's determination that the test for the remedy was the restoration of the parties to the pre-merger competitive position.

[62]            Moreover, an important element of the Tribunal's reasoning was its view of the statutory objectives provision of the Competition Act, section 1.1, and the relationship of that section to section 96. This is an issue of statutory interpretation of a kind with which courts are accustomed to dealing in the course of their ordinary work.

[63]            In short, I am not satisfied that Nadon J.'s expertise in competition law in general, and in the complexities of the merger of Superior and ICG in particular, gave him such a significant interpretative advantage over members of this Court as clearly to indicate Parliament's intention that the standard of review on the issue in dispute here should be that of unreasonableness. At the end of the day, the question of what count as "the effects of any prevention or lessening of competition" must be decided within the parameters of the Act, including its stated objectives. While economic expertise undoubtedly elucidates the strengths, weaknesses and consequences of the various choices available, it cannot be determinative of which of them, if any, is compatible with the Competition Act.

(iv) The Tribunal's constitutive statute and the scope of judicial review

[64]            Finally, the provisions of an administrative tribunal's constitutive statute respecting the grounds of judicial review, or the existence and scope of any right of appeal, may give some indication of the legislature's intention on the standard of review to be applied by a court to the tribunal's decisions.

[65]            At the one extreme, a strong preclusive clause, such as the bundle of exclusive jurisdiction, finality and "no certiorari" clauses typically found in the statutory schemes administered by labour relations boards, is indicative of a legislative intent to keep judicial review to a minimum. Hence, patent unreasonableness is generally the standard of review applied to labour boards' interpretation of the legislation that they administer.

[66]            At the other end of the spectrum are statutory rights of appeal that empower the appellate court to exercise any of the powers of the tribunal, direct the tribunal to take any action that the court considers proper and, for this purpose, to substitute its opinion for that of the tribunal. Rights of appeal from decisions of discipline committees of professional regulatory bodies are often of this kind.

[67]            There is a right of appeal from any decision of the Tribunal to this Court "as if it were a judgment of the Federal Court – Trial Division", except that, when the appeal is on a question of fact, leave of the Federal Court of Appeal is required: CTA, subsections 13(1) and (2). Section 27 of the Federal Court Act, R.S.C. 1985, c. F-7, imposes no limitations on the scope of the right of appeal from final judgments of the Trial Division to the Court of Appeal.

[68]            In my opinion, although expeditious decision-making is undoubtedly important in the review of mergers, the existence of an unrestricted right of appeal on questions of law, and of a modified right of appeal on questions of fact, must be entered as a factor indicative of Parliament's intention that the Tribunal's determinations of questions of law should be reviewable on appeal on a correctness standard.

(v) Conclusion

[69]            After weighing the factors to be considered in the pragmatic or functional analysis, and carefully examining the reasons for judgment in Southam, supra, I have concluded that it is the Court's function to determine whether the Tribunal was correct to decide that the effects of an anti-competitive merger that may be considered under section 96 are limited to the loss of resources to the economy as a whole resulting from the merger, to the exclusion of effects that relate to other statutory objectives, such as the wealth transfer from consumers to producers as a result of price increases, and the impact on competing small and medium sized businesses. A proposition of such generality is, to my mind, clearly a question of law.

[70]            I am not persuaded that, on an appeal to this Court, either the expertise of the Tribunal, or the degree of indeterminacy inherent in the word, "effects", indicates that the Court should review the Tribunal's decision on this issue on a standard other than that of correctness.

[71]            As Iacobucci J. noted in Southam, (supra, at pages 774-75, paragraph 53) with respect to the statutory requirement for, and to the role of, a judicial member of the Tribunal:

Clearly it was Parliament's view that questions of competition law are not altogether beyond the ken of judges.                                                                              

This comment seems applicable also to the judges of this Court.

[72]                        The composition of the Tribunal, and the rights of appeal from its decisions, reflect a carefully constructed compromise between assigning competition law exclusively to the domain of the judiciary, and entrusting it to a "non-judicial" regulatory agency, such as the Federal Trade Commission of the United States, which would operate subject to minimal judicial supervision: Canada (Director of Investigation and Research) v. Southam Inc., [1995] 3 F.C. 557, at page 604 (F.C.A.) (per Robertson J.A.).

Issue 2:            The Meaning of Effects in Section 96

[73]            The issue here is whether the Tribunal was correct in its interpretation of the phrase, "the effects of the prevention or lessening of competition ...", when it limited the relevant effects of the anti-competitive merger to those determined by the application of the total surplus standard. In my view, by so limiting the factors to be considered as "effects", the Tribunal erred in law because it failed to ensure that all the objectives of the Competition Act, and the particular circumstances of each merger, could be considered in the balancing exercise mandated by section 96.

[74]            With respect, I do not agree with the Tribunal's view that the list of objectives in section 1.1 of the Competition Act is merely a legislative rationale for the statutory purpose of maintaining and encouraging competition or that, if it is more than that, it should be read subject to the specific and contrary provisions of section 96. My reasons for these conclusions are as follows.

(i) The statutory text

(a) subsection 96(1)

[75]            Subsection 96(1) directs the Tribunal to consider whether the efficiencies produced by an anti-competitive merger are greater than, and offset, its anti-competitive effects. This is, in substance, a balancing test that weighs efficiencies on one hand, against anti-competitive effects on the other.

[76]            Writing of another provision in the Competition Act that called for the balancing of various factors, namely the determination of the scope of the relevant market, Iacobucci J. said in Southam (supra, at page 770, paragraph 43):   

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. [...] A test would be stilted and impossible of application if it purported to assign fixed weights to certain factors.                                               

Hence, since the efficiency defence requires the Tribunal to balance competing objectives, its operation should remain flexible and not stilted by an overarching and restrictive interpretation.

[77]            In referring to "the effects of any prevention or lessening of competition", subsection 96(1) does not stipulate what effects must or may be considered. When used in non-statutory contexts, the word, "effects", is broad enough to encompass anything caused by an event. Indeed, even though it does not consider the redistribution of wealth itself to be an "effect" for the purpose of section 96, the Tribunal recognizes, as all commentators do, that one of the de facto effects of the merger is a redistribution of wealth: paragraph 446.

[78]            In addition, section 5.5 of the MEG explicitly recognises that a merger may have more than one effect:

Where a merger results in a price increase, it brings both a neutral redistribution effect and a negative resource allocation effect on the sum of producer and consumer surplus (total surplus) within Canada.                                                           

The MEG concluded, however, that:

The efficiency gains described above are balanced against the latter effect, i.e., the deadweight loss to the Canadian economy.                                                                    

[79]            Thus, it is not doubted that the redistribution of resources is an effect of an anti-competitive merger, in the sense that it is caused by the exercise of market power created by the merger. Nevertheless, the Tribunal's interpretation of the word, "effects", as it is used in section 96, narrows it to a single effect, namely the loss or inefficient allocation of resources in the economy as a whole as measured by the deadweight loss.

[80]            Moreover, the statutory requirement that, for the section 96 defence to succeed, the efficiency gains must be greater than, and offset, the effects of a lessening of competition suggests a more judgmental assessment than is called for by the largely quantitative calculation of deadweight loss that the Tribunal held was statutorily mandated.

[81]            Of course, the precise meaning to be given to a word when it appears in a statute, especially if it is commonly used in everyday speech, must be determined by reference to its context. Hence, it was not necessarily an error of law for the Tribunal in this case to give to the word, "effects", a narrower meaning than would normally be ascribed to it in other contexts. The pertinent enquiry is whether, in the context of the Competition Act, the Tribunal was correct to narrow its meaning to the single effect of deadweight loss.

(b) subsection 96(3)

[82]            I attach some weight to subsection 96(3) of the Competition Act, which provides that the Tribunal shall not find that a merger or a proposed merger "is likely to bring about gains in efficiency by reason only of a redistribution of income between two or more persons." Hence, subsection 96(3) expressly limits the weight accorded to redistribution in assessing the efficiencies generated by a merger.

[83]            No similar limitation is imposed by the Act on the effects side of the balance. If Parliament had intended redistribution of income to be excluded altogether from the "effects" of an anti-competitive merger, as the Tribunal held, the drafter might well have been expected to have made an express provision, similar to that contained in subsection 96(3) with respect the efficiencies side of the balance. The absence of such a provision suggests that, contrary to the Tribunal's conclusion, Parliament did not intend to impose such a limitation on the "effects" side.

(ii) Statutory purposes and objectives

(a) section 1.1

[84]            I turn now to section 1.1 of the Competition Act which, for convenience's sake, I set out again.

[85]            I see nothing in the wording of this provision to indicate that it is anything other than a typical statutory purposes clause, and should be construed accordingly. As is not uncommon in such clauses, not all of the stated purposes or objectives can be served at the same time, nor are all necessarily consistent.

[86]            For instance, the objective of expanding "opportunities for Canadian participation in world markets" may be irrelevant when the merged entity is unlikely to compete abroad. Further, as is the case here, there may be a conflict between the aim of promoting "the efficiency and adaptability of the Canadian economy" and providing consumers with "competitive prices and product choices." In addition, of course, the wording of a particular provision in a statute may be so clear and precise that it must be regarded as overriding an ambiguous purpose clause.

[87]            Nonetheless, despite the typically indeterminate quality and inherent inconsistencies of purpose or objectives clauses, including section 1.1, statutory provisions containing general statements of legislative purpose are integral to the statute and can carry as much weight as its other sections: Ruth Sullivan, Driedger on the Construction of Statutes, 3^rd edition (Butterworths Canada Ltd. 1994), pages 263-68.

Thus, a purpose clause serves as a guide to the court or tribunal in its interpretation of other statutory provisions: R. v. T. (V.), [1992] 1 S.C.R. 749, at page 765, and may establish the parameters within which it must interpret the provisions of the statute: CAIMAW, Local 14 v. Paccar of Canada Ltd, [1989] 2 S.C.R. 983, at page 1028.

[88]            In my view, section 1.1 suggests that an interpretation of "effects" should not focus exclusively on one of the objectives of promoting competition, namely, promoting the efficiency and adaptability of the economy. Rather, the "effects" to be considered under section 96 should also include the other statutory objectives to be served by the encouragement of competition that an anti-competitive merger may frustrate, such as the ability of medium and small businesses to participate in the economy, and the availability to consumers of a choice of goods at competitive prices.

[89]            Indeed, in moving the second reading of Bill C-91, An Act to Establish the Competition Tribunal and to amend the Combines Investigation Act and the Bank Act and other Acts in consequence thereof, 1^st Session, 33^rd Parliament, 1984-85-86, which became the Competition Act and Competition Tribunal Act, the Minister of Consumer and Corporate Affairs and Canada Post noted (House of Commons Debates (April 7, 1986) at 11927):

The fourth but not the least objective is to provide consumers with competitive prices and product choices. As such, this objective becomes the common denominator in what we are trying to achieve. This is the ultimate objective of the Bill. (Emphasis added)                                                                                     

[90]            In spite of the existence of the multiple and ultimately inconsistent objectives set out in section 1.1, in certain instances the Act clearly prefers one objective over another. Thus, section 96 gives primacy to the statutory objective of economic efficiency, because it provides that, if efficiency gains exceed, and offset, the effects of an anti-competitive merger, the merger must be permitted to proceed, even though it would otherwise be prohibited by section 92. In this sense, the Tribunal was correct to state that section 96 gives paramountcy to the statutory objective of economic efficiency.

[91]            However, it does not follow from this that the only effects to be weighed against efficiency gains are limited to potential losses to the economy as a whole. Indeed, in the same Parliamentary speech referred to above, the Minister indicated (Debates, supra, at 11928) that the question posed to the Tribunal is:

Would a particular merger result in efficiency gains which would offset any negative effects on competition? (Emphasis added)                                                      

[92]                        Thus, although section 96 requires the approval of an anti-competitive merger where the efficiencies generated are greater than, and offset, its anti-competitive effects, the ultimate preference for the objective of efficiency in no way restricts the countervailing "effects" to deadweight loss. Instead, the word, "effects", should be interpreted to include all the anti-competitive effects to which a merger found to fall within section 92 in fact gives rise, having regard to all of the statutory purposes set out in section 1.1.

(b) "economic" purposes

[93]            In support of the position that the only effects of a merger that can be considered under section 96 are the resources lost to the economy as a whole, the respondents argued that the Supreme Court of Canada in Southam (supra, at page 772, paragraphs 48 and 49) authoritatively characterized the aims and objectives of the Competition Act as "more ‘economic' than strictly ‘legal'" and as "peculiarly economic". In my opinion, however, these statements are not dispositive of the issue under consideration here, namely, whether the Tribunal's interpretation of "effects" was correct.

[94]            First, while these statements were clearly directed to the purposes of the Competition Act administered by the Tribunal, they were made in the context of the pragmatic or functional analysis conducted to determine the appropriate standard of review. When he used the words quoted above, Iacobucci J. was characterising the purpose of the Act in order to delineate the areas of expertise of the Court and the Tribunal respectively. Hence, they are not decisive in the context of the issue at stake here, namely, determining which effects of an anti-competitive merger may be considered as "effects" under section 96.

[95]            Second, a characterisation of the objectives of the Competition Act as economic does not necessarily lead to the conclusion that it is only permissible to consider as "effects" under section 96 the resources likely to be lost to the economy as a whole. I would have thought that the extent to which a merger is likely to result in the elimination of small and medium sized businesses from a market, or to cause consumers to pay more than competitive prices, are sufficiently "economic" to fall within Iacobucci J.'s characterisation of the aims and objectives of the Act.

[96]            Third, I have already noted the inclusion of persons with a wide range of backgrounds on the Advisory Council that the Minister of Industry must consult before making recommendations to the Governor in Council on the appointment of lay members to the Tribunal. The statutory inclusion of Council members from a wide range of backgrounds, including consumer groups and labour, suggests that the perspectives of those appointed are likely to extend beyond general welfare economics. This, in turn, is an indication that the Act itself is not concerned with "economics" so narrowly conceived as to exclude from consideration under section 96 the redistributive effects of higher prices that consumers will have to pay as a result of the merger, or its impact on small and medium sized businesses.

[97]            The Tribunal stated that taking into account a broader range of anti-competitive effects of a merger than the deadweight loss would license members of the Tribunal "to advance their views on the social merit of various groups in society" or "to achieve the proper distribution of income in society". These "political" tasks, the Tribunal stated, cannot be regarded as mandated by the Act, because they are not within the expertise of the members of the Tribunal, who "are selected for their expertise and experience in order to evaluate evidence that is economic and commercial in nature": paragraph 431.

[98]            In my view, this conclusion gives insufficient weight to the range of experience and perspectives that the Act contemplates that the members of the Tribunal may possess, and overstates the degree of "social engineering" involved in considering a broad range of anti-competitive effects under section 96. Like other regulatory administrative tribunals, the Tribunal is charged with the responsibility of protecting the public interest, which it does by striking a balance among conflicting interests and objectives in a manner that respects the text and purposes of the legislation, is informed both by technical expertise and by the judgment that comes from its members' varied experiences, and is responsive to the particularities of the case.

[99]            Of course, balancing competing objectives in order to determine where the public interest lies in a given case requires the exercise of discretion. However, the procedure and composition of the Tribunal equip it for this task no less well than those of other independent, specialized, administrative tribunals that are required to perform similar balancing exercises in the discharge of their regulatory functions.

[100]        Finally, I also find it difficult to accept the Tribunal's interpretation of the Act for the following two reasons. First, when Bill C-91 was introduced in Parliament it was widely regarded as a consumer protection measure. Thus, the Minister responsible stated in the House of Commons (Debates, supra, at 11927) that the Consumers' Association of Canada saw the Bill as promising "real progress for consumers". Indeed, the guidebook introduced when the legislation was first tabled states (Consumer and Corporate Affairs Canada, Competition Law Amendments: A Guide (December, 1985), page 4):

Consumers and small business are among the prime beneficiaries of an effective competition policy.   

[101]        In addition, the background document released when the amendments were previously tabled (Consumer and Corporate Affairs Canada, Combines Investigation Act Amendments 1984: Background Information and Explanatory Notes (April, 1984), page 2), states that:

the Bill is concerned with fairness in the functioning of markets – fairness between producers and consumers, fairness between businesses and their suppliers, and suppliers and their customers.                                                                                         

[102]        It thus seems to me unlikely that Parliament either intended or understood that the efficiency defence would allow an anti-competitive merger to proceed, regardless of how much the merged entity might raise prices, provided only that the efficiencies achieved by the merger exceeded the resulting loss of resources in the economy at large. As Reed J. noted in the Hillsdown case, supra, at pages 337-38, differences in the drafting of the efficiency defence in the precursors to Bill C-91, which were not enacted, point in the same direction, and are considered in paragraphs 129-131, post.

[103]        Second, the result of applying the total surplus standard has some consequences that are so paradoxical in light of the consumer protection objectives of the Act that Parliament should not be regarded as having intended to limit the "effects" of the merger for the purpose of section 96 to deadweight loss. For example, use of the total surplus standard for calculating the anti-competitive effects of a merger makes it easier to justify a merger between suppliers of goods for which demand is relatively inelastic than of goods for which demand is relatively elastic.

[104]        This is because, where the demand for particular goods is inelastic, as it is for propane, the goods cannot be substituted as cost-effectively as where the demand is elastic. Hence, price increases that result from the exercise of market power are tolerated more by purchasers of goods for which the demand is inelastic than by purchasers of those where the demand is elastic. Thus, since purchasers of goods for which demand is inelastic are relatively insensitive to price, fewer will purchase substitute goods despite increases in price. Therefore, a significant price increase will result in a smaller deadweight loss in a product where demand is inelastic than where it is elastic.

[105]        Thus, on the Tribunal's interpretation of section 96, the more inelastic the demand for the goods produced by the merged entity, the smaller will be the efficiencies required from the merger in order to offset its anti-competitive effects. It follows on this reasoning that, for the purpose of balancing efficiencies and effects, a potentially large wealth transfer from consumers of goods for which demand is inelastic to producers is to be ignored.

[106]        It is certainly not obvious how an interpretation of "effects" that creates a differential treatment of mergers by reference to the elasticity of demand for the goods produced by the merged entity is rationally related to any of the statutory aims of the Competition Act.

[107]        Another consequence of limiting the anti-competitive "effects" of a merger to deadweight loss is that it is irrelevant that the merger results in the creation of a monopoly in one or more of the merged entity's markets. According to the Tribunal, the fact that the merged entity of Superior and ICG will eliminate all consumer choice, and remove all competition, in the propane supply market, as it is likely to do in Atlantic Canada, for example, is not an "effect" that legally can be weighed under section 96 against the efficiency gains from the merger.

[108]        Again, such a conclusion seems to me so at odds with the stated purpose of the Act, namely "to maintain and encourage competition", and the statutory objectives to be achieved thereby, as to cast serious doubt on the correctness of the Tribunal's interpretation of section 96.

[109]        Given the purposes historically pursued by competition legislation and, in particular, the expressly stated purpose and objectives of the Competition Act, it is reasonable to infer from Parliament's failure to state expressly that only deadweight loss is to be considered as an "effect" of a merger for the purpose of section 96, that other effects related to the statutory purpose and objectives, including the interests of the consumers of the merged entity's products, must also be taken into account when the trade-off is made between efficiencies and anti-competitive effects.

(iii) Predictability

[110]        It was strenuously argued by counsel for the respondents that, since one of the objectives of the Competition Act set out in section 1.1 is to "promote the efficiency and adaptability of the Canadian economy", it was important for business people to be able to predict whether or not a proposed merger was likely to receive regulatory approval. Otherwise, they might be deterred from entering into a merger that would violate section 92 by substantially lessening competition, but would increase wealth in the Canadian economy as a whole by producing substantial efficiency gains.

[111]        Hence, it was argued, it is consistent with the purpose of section 96 to interpret the efficiency defence as requiring the use of the total surplus standard to determine the anti-competitive effects of a merger, because the use of this standard makes the result of the section 96 balancing exercise much more predictable. While far from self-applying, the total surplus standard will generally make it much easier than the balancing weights approach favoured by the Commissioner to predict what will be the "effects" of a merger.

[112]        While not without some attraction, this argument when considered alone is far from dispositive in a regulatory context. And, when assessed with the stronger arguments pointing in the opposite direction, it does not in my view significantly buttress the Tribunal's interpretation of section 96.

[113]        First, discretionary decision-making in the regulation of economic activity is commonplace and predictability of outcome is a matter of degree. Indeed, since discretion is essential to the efficacy of most regulatory regimes, the interest of individuals in being able to arrange their affairs in the more or less certain knowledge of how they will be regarded by agencies of the state is not so highly valued as in other areas (such as taxation or criminal law) where the state impinges on individual conduct.

[114]        Hence, even if true, the submission that the total surplus standard may make the result of the balancing exercise more predictable than the balancing weights approach must be assessed in the context of the administration of a public programme of economic regulation.

[115]        Second, one should not exaggerate the differences in the degrees of predictability inherent in the total surplus and balancing weights standards for determining the "effects" of an anti-competitive merger. Given the difficulties of, for example, assessing both the relative elasticity of demand for the goods produced or supplied by a merged entity, and the qualitative aspect of deadweight loss, the application of the total surplus standard is far from mechanical. Indeed, while section 5.5 of the MEG has adopted the total surplus standard, it also states that the "calculation of the anti-competitive effects of mergers is generally very difficult to make." See also Roy M. Davidson, "When Merger Guidelines Fail to Guide" (1991) 12 Canadian Competition Policy Record 44, at 46-47.

[116]        Conversely, it is in my view far from a fatal objection to the balancing weights approach that its proponent at the hearing before the Tribunal, Professor Townley, testified that, as an economist, he was unable to determine what were the effects of the merger of Superior and ICG and whether the efficiencies likely to be produced thereby were greater than, and offset, them. I take his point simply to have been that he was called as a witness expert in economics and that the balancing exercise called for by section 96 required broader public policy judgments that were outside his area of expertise, but were for the Tribunal to make as it thought would best advance the public interest within the parameters of the Act.

[117]        Third, there are various tools available to administrative agencies that enable them to give more precision, and hence predictability of application, to the discretionary statutory standards that they must apply to particular fact situations: speeches by members of the administrative agency detailing agency thinking on an issue, and more formal published policy guidelines that can be elaborated and tailored from time to time to take account of agency experience with administering the regulatory scheme, for example. I discuss below the MEG issued by the Commissioner, in so far as they deal with the Competition Bureau's view of the interpretation of section 96.    

[118]        In addition, parties contemplating a merger may submit details to the Commissioner at an early stage of the process in order to obtain an initial indication of whether approval is likely to be forthcoming and, if the Commissioner thinks that there may be problems, what they are and how they may be addressed. Administrative adjudication is only the rarely seen, though important, tip of the regulatory process iceberg.

[119]        Hence, even if the total surplus standard provides more predictability to prospective merging parties, when compared, for instance, to the balancing weights approach, the predictability argument is not sufficiently compelling to persuade me that it is the methodology mandated by section 96 for determining the "effects" of an anti-competitive merger in all cases.

(iv) Merger Enforcement Guidelines

[120]        Both the Tribunal and, on appeal, counsel for the respondents, gave considerable weight to the MEG, issued in 1991 by the Director of Investigation and Research, Bureau of Competition Policy.

[121]        Section 5.5 of the MEG state that efficiency gains are to be balanced only against "a negative resource allocation effect on the sum of producer and consumer surplus (total surplus within Canada)"; in other words, "the deadweight loss to the Canadian economy." It also states that the redistribution of wealth as a result of price increases stemming from the merger is "neutral", noting in footnote 37 that:

When a dollar is transferred from a buyer to a seller, it cannot be determined a priori who is more deserving or in whose hands it has a greater value.                        

[122]        In a speech given in Toronto on June 8, 1992 to the Canadian Institute, the then Director of Investigation and Research responded to the doubts expressed by Reed J., as the judicial member of the Tribunal that decided the Hillsdown case, supra, about whether the MEG were consistent with the Competition Act to the extent that they adopted total surplus standard as the basis for determining the "effects" of an anti-competitive merger. The Director saw no need to amend the MEG at that time, since economists advocated that wealth transfers to producers from consumers should be treated as a neutral effect of a merger, Reed J.'s expressions of doubt were only obiter and the Tribunal endorsed no other methodology for determining the "effects" to be taken into account under section 96.

[123]        In 1998, the approach to the determination of the anti-competitive effects of a merger adopted in the MEG was essentially endorsed in the Competition Bureau's publication, The Merger Guidelines as Applied to a Bank Merger.

[124]        The simple answer to the respondents' reliance on the MEG is that they are not law because they are not made under a grant of statutory authority, and cannot determine the meaning of the Act. Indeed, to the extent that they are inconsistent with the Act, they should be ignored. Further, the limited nature and intent of the MEG is clearly set out at the beginning of the document under the heading "Interpretation":

This document is intended solely to provide enforcement guidelines. As such, it sets forth the general approach that is taken to merger review, and is not a binding statement of how discretion will be exercised in a particular situation. Specific guidance regarding a specific merger may be requested from the Bureau through its program advisory opinions. The Guidelines are not intended to be a substitute for the advice of merger counsellors. They do not represent a significant change in enforcement policy or restate the law. Final interpretation of the law is the responsibility of the Competition Tribunal and the courts. (Emphasis added)            

[125]        Of course, it may do little to inspire public confidence in the administration of the Competition Act that, in the context of the merger of Superior and ICG, the present Commissioner has apparently disavowed the interpretation of section 96 advanced in the MEG, which have still not been replaced. However, there was no allegation by the respondents that they had relied to their detriment on the MEG when they agreed to merge. While there was no evidence in the record about any discussions that may have taken place between the merging parties and the Bureau, it would not be surprising if such discussions had occurred and it had been indicated to the respondents that the Commissioner no longer thought that deadweight loss, measured both quantitatively and qualitatively, was the only "effect" that could ever be taken into account under section 96.

[126]        In addition, the possibility that a reviewing court may not agree with an agency's view of the law is an inevitable risk associated with the administrative practice of issuing non-binding guidelines and other policy documents to shed light on agency thinking and to assist those subject to the regulatory regime that it administers. This risk should deter neither the courts from deciding what the law is, nor agencies from engaging in the often useful exercise of administrative rule-making.

(v) The authorities

[127]        Finally, I consider whether existing authorities demonstrate the correctness of the Tribunal's interpretation of section 96. I turn, first, to the only other judicial pronouncement on the issue, namely, the decision of Reed J. in the Hillsdown case, supra. I agree with the respondents' position that what Reed J. said in that case is not dispositive of this case: not only is it, like the case before us, a decision of the Tribunal, but Reed J.'s statements did not form part of the ratio and, in some respects at least, she expressed herself more in the form of a doubt than of a definitive assertion that the interpretation in the MEG of "effects" was wrong in law.

[128]        Nonetheless, I find myself largely in agreement with the reasons given by Reed J. for querying whether the Tribunal was permitted to look only at deadweight loss when determining the effects to be balanced against any efficiency gains that, without the merger, were unlikely to be achieved.

[129]      In particular, I adopt her analysis of the legislative history of section 96: Hillsdown, supra, at pages 337-39. She observed that, unlike the present section 96, the previous, unenacted versions of the efficiency defence contained in both Bill C-42, An Act to amend the Combines Investigation Act and to amend the Bank Act and other Acts in relation thereto or in consequence thereof, 2^nd Session, 30^th Parliament, 1976-77, and in Bill C-29, An Act to amend the Combines Investigation Act and the Bank Act and other Acts in consequence thereof, 2^nd Session, 32^nd Parliament, 1983-84, did not require that the efficiencies gained from an anti-competitive merger be balanced against its effects.

[130]        Thus, Bill C-42 would have permitted an anti-competitive merger to proceed, provided only that substantial efficiency gains could be proved "by way of savings of resources for the Canadian economy" that would not otherwise have been attained: clause 31.71(5). Bill C-29 called for a determination of whether "the gains in efficiency would result in a substantial real net saving for the Canadian economy": clause 31.73(c). Neither of these provisions calls for a balancing of efficiencies against effects. Instead they focus on resource maximization in the economy as a whole in the same way as the total surplus standard.

[131]        I agree with Reed J.'s conclusion that, seen against this background, the more open-ended direction given to decision-makers by section 96, namely to balance the efficiency gains against the "effects" of an anti-competitive merger, should not be interpreted in substantially the same manner as the above clauses, which explicitly permitted anti-competitive mergers when the resulting efficiency gains produced net savings of resources for the Canadian economy. While the earlier bills seem clearly to have encapsulated the total surplus standard in the efficiency defences, section 96 does not.

[132]        I note, too, that, even though she may not have been entitled as a lay member of the Tribunal to express a view on an issue that I have held to be a question of law alone, Ms. Lloyd did not agree that "effects" were confined to deadweight loss to the exclusion of effects relating to the other objectives of the Act: paragraph 506.

[133]        In a word, views expressed by Tribunal members who have considered the issue are about evenly split. I draw some comfort from the existence of this division of opinion both between the judicial members who have considered the issue (Reed J. and Nadon J.), and between the lay members of the Tribunal in this case, if, as I understand it, Dr. Schwartz agreed with Nadon J. Thus, in disagreeing with the Tribunal's interpretation of section 96, I cannot be said to have gone against the unanimous view of those more expert than I in this area of the law.

[134]        Finally, it was suggested in argument that the Tribunal's interpretation had the support of all economists who had studied the issue. I do not dispute that an impressive array of economists, and law and economics specialists, both in Canada and the United States, have argued that the total surplus standard is the appropriate basis for determining whether an anti-competitive merger that produces efficiency gains should be permitted.

[135]        Nonetheless, the Horizontal Merger Guidelines, supra, in the United States continue to treat the exercise of market power leading to an increase in price above the competitive level as the most important anti-competitive effect of a merger, and the resulting wealth transfer from the consumers to the producers, as a misallocation of resources: see P.T. Denis, "Advances in the 1992 Horizontal Merger Guidelines in the Analysis of Competitive Effects" (Symposium on the New 1992 Merger Guidelines), Antitrust Bulletin, Fall 1993, pages 479-515.

[136]        Of course, as I have already noted, since there is no specific efficiency defence in the United States' legislation, the approach of the Federal Trade Commission to efficiency gains when considering the approval of anti-competitive mergers has limited relevance to the problem before us. Nonetheless, it is interesting to note that efficiency gains are generally most likely to make a difference in merger review when the likely adverse effects of the merger are not great, and will almost never justify a merger to monopoly or near monopoly: Horizontal Merger Guidelines, supra, at page 150.

[137]        In addition, some commentators in the United States have expressed surprise at the interpretation of section 96 adopted in the MEG. See, for example, J.F. Brodley, "The Economic Goals of Antitrust: Efficiency, Consumer Welfare, and Technological Progress, (1987) 62 N.Y.U. L. Rev. 1020, at 1035-36; S.F. Ross, "Afterword – Did the Canadian Parliament Really Permit Mergers That Exploit Canadian Consumers So the World Can Be More Efficient?" (1997) 65 Antitrust Law Journal 641. Thus, Ross writes (supra, at 652, note 41):

As Professor Brodley has observed, the logical extension of competition policy based solely on societal wealth maximization would be to prefer a monopolist that was able to perfectly price discriminate (charge each consumer the maximum amount each consumer was willing to pay) to the typical Canadian industry with a relatively few number of firms, which would not produce at a single competitive price.                                                                                                                              

[138]        Hence, it is clear that there is more than one view among commentators on whether competition policy should disregard a priori transfers of wealth and other "effects" of anti-competitive mergers, and consider only whether the merger has the effect of increasing or decreasing the resources in the economy as a whole. Nonetheless, when the issue arises in the legal context of a section 92 proceeding instituted by the Commissioner, it must be answered by reference to the Competition Act, and Parliament's stated purpose and objectives in enacting it. In my view, the narrow reading that the Tribunal gave to the word, "effects", in section 96 cannot be justified by reference to the views of lawyer-economists in the United States, no matter how eminent.

(vi) Conclusions

[139]        Having concluded for the above reasons that the Tribunal erred in law when it interpreted section 96 as mandating that, in all cases, the only effects of an anti-competitive merger that may be balanced against the efficiencies created by the merger are those identified by the total surplus standard, this Court should not prescribe the "correct" methodology for determining the extent of the anti-competitive effects of a merger. Such a task is beyond the limits of the Court's competence.

[140]        Whatever standard is selected (and, for all I know, the same standard may not be equally apposite for all mergers) must be more reflective than the total surplus standard of the different objectives of the Competition Act. It should also be sufficiently flexible in its application to enable the Tribunal fully to assess the particular fact situation before it.

[141]        It seems to me that the balancing weights approach proposed by Professor Townley, and adopted by the Commissioner, meets these broad requirements. Of course, this approach will no doubt require considerable elaboration and refinement when it comes to be applied to the facts of particular cases.

[142]        Further, while the adoption of the balancing weights approach is likely to expand the anti-competitive effects to be considered, and hence to narrow the scope of the defence, I see no reason why it should, as the respondent submitted, practically write section 96 out of the Act.

Issue 3:            The Burden of Proof

[143]        The Tribunal held (at paragraph 403) that, since section 96 constitutes a defence to an infringement of section 92, the merging parties bear the burden of proving each of its elements on the balance of probabilities, except the existence or scale of the effects that must be balanced against the efficiency gains.

[144]        The Commissioner submitted that the Tribunal had erred in law in holding that he had the legal onus of proving anything at all under section 96. While the evidential burden may shift as a case unfolds, the legal burden throughout, counsel for the Commissioner argued, remains with the respondents.

[145]        Apart from setting out the parties' contentions, the Tribunal gave little clue about the reasons for its conclusion on the burden of proof issue. One obvious possibility is that the Tribunal endorsed the submissions made by the party in whose favour it decided particular issues. Thus, the Tribunal stated that the respondents had submitted that the Commissioner bears the burden of proving the scale of the "effects" of an anti-competitive merger (that is, on the Tribunal's interpretation of section 96, the deadweight loss), because the Commissioner's investigative powers put him in a better position than the respondents to obtain the necessary information from third parties.

[146]        On the hearing of the appeal, counsel for the respondents added that, if merging parties were to attempt to obtain the kind of information required to establish the effects of the merger for the purpose of section 96, including information on competitors' pricing and costs, they would run the risk of being accused of conspiring to restrain competition contrary to section 45 of the Act.

[147]        Counsel for the Commissioner, on the other hand, relied on statements made by officials of the Ministry of Consumer and Corporate Affairs, when appearing before the Legislative Committee on Bill-91. They advised the Committee that, once the Commissioner had proved a substantial lessening of competition under section 92, the burden of proving any defence was borne by the merging parties.

[148]        On the facts, the burden of proof did not have to be decided when the present case was before the Tribunal. However, since the effects that must be considered by the Tribunal include facts not taken into account when it first made the decision under appeal, it is necessary for the Court to determine the issue, which is largely one of first impression.

[149]        It seems clear that deciding which party bears the burden of proving what elements of the efficiency defence is a pure question of law that is not confined to the particularities of this case. For the reasons given earlier, I am of the view that the standard of review of the Tribunal's conclusion on this issue is correctness.

[150]        Two general principles would seem to support the Commissioner's position.    First, the party who asserts, must prove the assertion. Since it is the respondents who assert that the efficiency gains of the merger are likely to exceed, and to offset, its anti-competitive effects, this principle indicates that the respondents should be required to prove each and every aspect of the assertion. The second general principle is that the burden of proving a defence generally rests with the defendant.

[151]        However, the principle that the party who asserts must prove is not absolute: Sopinka, Lederman and Bryant, The Law of Evidence in Canada, 2^nd edition, (Butterworths Canada Ltd.; 1999), page 89. In addition, in the absence of authority, considerations of fairness, probability and policy would seem to be important determinants of the legal burden of proof: Sopinka, Lederman and Bryant, supra, at pages 86-90.

[152]        It would be somewhat odd, as counsel for the respondents argued, to place the legal onus of proving the anti-competitive effects of a merger on the party whose interest it is to deny that they exist or to minimise them. In addition, in the process of establishing a substantial lessening of competition, the Commissioner will often have gathered evidence on the effects of the merger that will also be relevant to the section 96 defence, including evidence on likely price increases following the merger and the impact of the merger on inter-related businesses.

[153]        These are matters on which the Commissioner is in a better position than the respondents to gather evidence by virtue of the investigative powers conferred on him by statute. Indeed, as Sopinka, Lederman and Bryant note (supra, at page 89), if "one party is peculiarly situated to prove a fact" a court may reverse the burden and place it on that party.

[154]        Accordingly, I have concluded that the Tribunal was correct to distribute the legal onus of proof as it did, so that the respondents bear the onus of proving every aspect of the section 96 defence, save for the anti-competitive effects of the merger.

F. CONCLUSIONS

[155]        In summary, I would allow the appeal, set aside the decision of the Tribunal with respect to the interpretation of section 96 of the Competition Act and remit the matter to the Tribunal for redetermination in a manner consistent with these reasons.

[156]        The Tribunal need only identify and assess "the effects of the prevention or lessening of competition" for the purpose of section 96 and decide whether the efficiencies that the Tribunal has already found to have been proved by the respondents are likely to be greater than, and to offset, those effects.

[157]        The Commissioner has the legal burden of proving the extent of the relevant effects, while the respondents have the burden, not only of proving the scale of the efficiency gains that would not have occurred but for the merger, but also of persuading the Tribunal on the ultimate issue, namely, that the efficiency gains are likely to be greater than, and to offset, the effects.

[158]        The appellant should have his costs, but because the respondents were successful on the burden of proof issue, I would reduce the costs awarded by 20% of those otherwise allowable.

                                                                                   "John M. Evans"              

                                                                                                      J.A.                      

"I agree

A.J. Stone J.A."

Date: 20010404

Docket: A-533-00

Neutral citation: 2001 FCA 104

CORAM:          STONE J.A.

LÉTOURNEAU J.A.

EVANS J.A.

BETWEEN:

               THE COMMISSIONER OF COMPETITION

                                                                                              Appellant

AND:

       SUPERIOR PROPANE INC. and ICG PROPANE INC.

                                                                                         Respondents

                            REASONS FOR JUDGMENT

LÉTOURNEAU J.A.

[159]        I have had the benefit of reading the reasons for judgment issued by my colleague, Evans J.A.. I agree with him that the interpretation of the word "effects" in section 96 of the Competition Act (Act) R.S.C. 1985, c. C-34 involves a pure question of law that falls to be decided on a standard of correctness.

[160]        I also agree with my colleague that the word "effects" in section 96 of the Act ought not to be limited, as the Tribunal did, to the effects identified by the total surplus standard. As my colleague has pointed out, the interpretation of section 96 of the Act involves balancing market power and efficiency gains. The approach taken in this matter both in the United States and in Canada is by no means free from ambiguity and harsh criticism: see Robert H. Lande, The Rise and (Coming) Fall of Efficiency as the Ruler of Antitrust (1988) 33 Antitrust 429; David B. Andretsch, Divergent Views in antitrust Economics (1988) 33 Antitrust Bull. 135; Alan A. Fisher, Frederick I. Johnson and Robert H. Lande, Price Effects of Horizontal Mergers (1989) 77 Calif. L.R. 777; Lloyd Constantine, An Antitrust Enforcer Confronts the New Economics (1989) 58 Antritrust L.J. 661; Roy M. Davidson, When Merger Guidelines Fail to Guide (1992), Canadian Competition Policy Record 44, at page 46; Stephen F. Ross, Afterword - Did the Canadian Parliament Really Permit Mergers that Exploit Canadian Consumers so the World can be More Efficient? (1997) 65 Antitrust Law Journal 641, at pages 643-646; Tim Hazledine, Rationalism Rebuffed? Lessons from Modern Canadian and New Zealand Competition Policy (1998) Review of Industrial Organization 243; Jennifer Halliday, The Recognition, Status and Form of the Efficiency Defence to a Merger: Current Situation and Prospects for the Future (1999) World Competition 91. A review of these authorities reveals that the provision is at best confusing and puzzling. At worst, it can defeat the very purpose of the Act. I reproduce sections 96 and 1.1 for convenience:

96. (1) The Tribunal shall not make an order under section 92 if it finds that the merger or proposed merger in respect of which the application is made has brought about or is likely to bring about gains in efficiency that will be greater than, and will offset, the effects of any prevention or lessening of competition that will result or is likely to result from the merger or proposed merger and that the gains in efficiency would not likely be attained if the order were made.

(2) In considering whether a merger or proposed merger is likely to bring about gains in efficiency described in subsection (1), the Tribunal shall consider whether such gains will result in

(a) a significant increase in the real value of exports; or

(b) a significant substitution of domestic products for imported products.

(3) For the purposes of this section, the Tribunal shall not find that a merger or proposed merger has brought about or is likely to bring about gains in efficiency by reason only of a redistribution of income between two or more persons.

96. (1) Le Tribunal ne rend pas l'ordonnance prévue à l'article 92 dans les cas où il conclut que le fusionnement, réalisé ou proposé, qui fait l'objet de la demande a eu pour effet ou aura vraisemblablement pour effet d'entraîner des gains en efficience, que ces gains surpasseront et neutraliseront les effets de l'empêchement ou de la diminution de la concurrence qui résulteront ou résulteront vraisemblablement du fusionnement réalisé ou proposé et que ces gains ne seraient vraisemblablement pas réalisés si l'ordonnance était rendue.

(2) Dans l'étude de la question de savoir si un fusionnement, réalisé ou proposé, entraînera vraisemblablement les gains en efficience visés au paragraphe (1), le Tribunal évalue si ces gains se traduiront:

a) soit en une augmentation relativement importante de la valeur réelle des exportations;

b) soit en une substitution relativement importante de produits nationaux à des produits étrangers.

(3) Pour l'application du présent article, le Tribunal ne conclut pas, en raison seulement d'une redistribution de revenu entre plusieurs personnes, qu'un fusionnement réalisé ou proposé a entraîné ou entraînera vraisemblablement des gains en efficience.

                                 Purpose

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.

                                    Objet

1.1 La présente loi a pour objet de préserver et de favoriser la concurrence au Canada dans le but de stimuler l'adaptabilité et l'efficience de l'économie canadienne, d'améliorer les chances de participation canadienne aux marchés mondiaux tout en tenant simultanément compte du rôle de la concurrence étrangère au Canada, d'assurer à la petite et à la moyenne entreprise une chance honnête de participer à l'économie canadienne, de même que dans le but d'assurer aux consommateurs des prix compétitifs et un choix dans les produits.

[161]        First, section 96 is broadly worded and provides no guidance as to the meaning of efficiency found in the section, the effects of the lessening of competition which are to be weighed against the efficiency gains, and the test, standard or trade-off to be applied in weighing the elements.

[162]        For example, what kind of economic efficiency does section 96 of the Act refer to? Allocative efficiency which is achieved when the existing products at the allocated prices satisfy the consumers' want? Or productive efficiency which is obtained when output is produced with the most cost-effective combination of productive resources available under present technology? Or technological or dynamic efficiency which is achieved through better industrial research development and a better rate of technological progress?

[163]        What are the anti-competitive effects of the merger that are to be weighed? Is it limited to deadweight loss which occurs when, because of higher prices, consumers choose an alternative and less appropriate substitute for the product that they would have other bought? Does the trade-off analysis include anti-competitive effects likely to arise in other related markets which would be affected by the merger? Does it include wealth transfers from consumers to producers that result from an exercise at market power? Are all the effects of the merger to be weighed and what weight should be given to them? Are they all of the same significance and value? On what basis is one effect to be preferred over the other? On what basis should some effects, if any, be ignored or discarded?

[164]        What standard should be applied to the trade-off analysis required by the application of section 96? The total surplus standard chosen by the Tribunal in this case which considers only the deadweight loss and none of the redistributive effects involved in the wealth transfer from consumers to producers? Or the price standard under which efficiencies allow for mergers only if prices are to be maintained or reduced? Or the consumer surplus standard which disallows a merger where the loss of consumer surplus exceeds the efficiency gains?

[165]        Second, the relationship of section 96 with section 1.1, which states the purpose of the Act, is not defined and, in fact to many, section 96 contradicts section 1.1 and defeats the purposes contained in that purpose clause. When weighing the efficiency gains of a merger against the lessening of competition, what should be done of the stated objectives of ensuring that small and medium-sized enterprises have an equitable opportunity to participate in the Canadian economy as well as providing consumers with competitive prices and product choices?

[166]        Third, section 96 poses no limits to the kind of mergers that can be effected and authorized as long as the efficiency gains will be greater than, and will offset, the effects of the lessening of competition and as long as these gains cannot be obtained by other means than the merger. This means that the creation of a monopoly or near monopoly through mergers could be authorized even though it would eliminate competition altogether, discourage competitive prices for consumers and would undermine, to the point of eradication, the development of small and medium-size enterprises, all these effects contrary to the purposes stated in section 1.1 of the Act.

[167]        Fourth, the problems created by section 96 are compounded by the fact that the provision is mandatory. The Tribunal shall not make an order preventing a merger where the undefined and elusive balancing test of section 96 is met.

[168]        Fifth, section 96 appears to have no geographical scope or limit so that efficiency gains made for the benefit of foreign corporations to the detriment of Canadian workers and consumers could be counted in the trade-off analysis that the provision requires. Or are mergers to be approved only if the efficiency gains in Canada exceed the losses in Canada? In the increasing context of globalization of trade and commerce, not to mention international trade treaties such as the North American Free Trade Agreement Implementation Act, S.C. 1993, c. 44 and the World Trade Organization Agreement Implementation Act, S.C. 1994, c. 47, Schedules I-IV, the issue of whether the balancing test applicable under section 96 is global or limited in its application to Canada becomes crucial. Yet, the provision still provides no guidance in this respect.

[169]        It is no wonder that conflicting views on the scope of section 96 have emerged and that the section, in search of predictability and workability, has been read down by eliminating some of the significant effects of the lessening of competition. It also comes as no surprise that many, influenced as they were by the Chicago school of thought in antitrust matters, concluded, as the Tribunal did in the present case, that efficiency of the economy overrides competition even with respect to an Act designed to maintain and promote competition.

[170]        It is true as Mr. Justice Iacobucci said in Dir. Inv. and Res. v. Southam Inc. [1997] 1 S.C.R. 748, at page 772 that the focus of the Act is on economy rather than law or, if one prefers, that the "aims of the Act are more economic than they are strictly legal". But section 96 really begs the question: what kind of economy? Monopolistic, competitive or a proper balance between these two poles?

[171]        The Tribunal found that the merger was likely to prevent competition substantially in Atlantic Canada and to lessen competition substantially in coordination services offered to national account customers: see decision, paragraphs 310 and 313. There was also conclusive evidence that, in many large areas of the country, the merger would not merely lessen competition, but would in fact eliminate it and create monopolies. The following Chart illustrates the impact of the merger with respect to monopolies or near monopolies: see Compendium of the appellant, page 001327:

                                                                                  Table 4

                                                                Geographical Markets with Merger-to-Monopoly

Pre-Merger               Post-Merger

Market                                                     SPI           ICG               SPI

%             %                   %

Val d'Or                                                                   74            23                  97

Sept Iles/Baie Comeau                                                             55            45                 100

Bancroft/Pembroke/Eganville                                   92            5                   97

Dryden/Fort Frances/Kenora/Ignace                                        47            52                  99

Echo Bay/Sault Ste Marie                                                        55            44                  99

Hearst/Wawa/Manitouwadge/Marathon                                 43            53                  96

Little Current/Sudbury                                                            51            48                  99

North Bay                                                                                81            16                  97

Thunder Bay                                                                            46            54                 100

Fort McMurray                                                                       32            67                  99

Whitecourt                                                                               55            45                 100

Burns Lake/Terrace/Smithers/Prince Rupert                           62            37                  99

Fort Nelson                                                                              44            56                 100

Valemont                                                                  43            57                 100

Watson Lake                                                                            25            75                 100

Whitehorse                                                                               33            67                 100

[172]        The Tribunal, in view of its conclusion that efficiency is the paramount objective of the Act, ignored as an effect of the merger the fact that monopolies in certain product markets would ensue and failed to give any weight to that effect in its analysis under section 96. The Act maintains and promotes competition. It assumes that economic efficiency will generally and primarily develop through competition. It also accepts in section 96 that, in some cases, a reduction in competition can and will produce more efficiency than competition as it existed before merger.

[173]        In my respectful view, however, section 96 was not meant to authorize the creation of monopolies since it would defeat the purpose of section 1.1. The section was not intended to authorize mergers resulting in monopolies whereby, contrary to section 1.1, competition is eliminated, small and medium-sized enterprises are not able to enter or survive in the market and consumers are deprived of competitive prices.

[174]        As the Supreme Court of the United States has asserted repeatedly with respect to the U.S. antitrust laws, "Congress was dealing with competition, which it sought to protect, and monopoly, which it sought to prevent: Standard Oil v. Federal Trade Comm'n, 340 U.S. 231, at pages 248-49 (1951) quoting A.E. Staley Mfg. Co. v. Federal Trade Comm'n, 135 F. 2d 453, at page 455 (7^th Cir. 1943)". As my colleague pointed out, a similar expression of intent can be found in the Minister's (Minister of Consumer and Corporate Affairs and Canada Post) statement in the House of Commons where he reasserted in presenting the Bill that the ultimate objective of the Act was to provide consumers with competitive prices and product choices.

[175]        I agree with my colleague that the application of a balancing test requires a flexibility that the total surplus standard does not provide. It is true that a flexible approach may not yield the predictability that the assumptions and presumptions underlying the total surplus standard afford. However, if predictability is the preferred option, Parliament is at liberty to revisit section 96 and say so.

[176]        Finally, contrary to my colleague, I believe the Tribunal erred when it put on the Commissioner the legal burden (i.e., the burden of persuasion) of proving the effects of the lessening of competition. In practice, the merging parties will lead evidence of efficiency gains and of some of the effects of the lessening of competition. This is the evidential burden. They need to do that to establish that the gains offset the effects. Of course, the tendency for the merging parties might be to increase the amount of gains and downplay the effects of the lessening of competition. This is why, as we have seen in this case, the Commissioner also bears in practice an evidential burden, that is the burden of leading evidence as to both components of the efficiency defence to alert the Tribunal to what the real, as opposed to the alleged, gains and effects are. In the end, however, the legal burden is on the merging parties to convince the Tribunal, first, that the efficiency gains are of the amount that they have contended, second, that the effects of the lessening of competition are those that they have identified and not those submitted by the Commissioner, and third, that the efficiency gains are greater than, and will offset, the effects.

[177]        I agree with the respondents that the Commissioner, with his statutory investigative powers, may be in a better position to gather information relevant to the effects and, indeed, that it would have done so in the context of the application of section 92 to which section 96 is a defence. The availability of statutory investigative powers will, indeed, enable the Commissioner to assume his evidentiary burden of gathering and filing relevant evidence to counter and rebut the allegations and evidence of the merging parties as to the effects of the lessening of competition. However, this is not sufficient to transfer the legal burden of proving these effects on the Commissioner. Indeed, there is no rationale and justification for putting on the Commissioner the burden of persuasion on one of the three components of the efficiency defence.

[178]        In conclusion, I would dispose of the matter as proposed by my colleague, except as to costs where I would make no apportionment in view of my conclusion that the Tribunal also erred on the issue of the legal burden of proof.

                                                                                                                               "Gilles Létourneau"                

                                                                                                                                                      J.A.