Ontario Superior Court of Justice (Divisional Court)

Court File No. 367/04

Date: 2006-02-08

Epstein and Lax JJ.

Counsel:

Gordon McKee and Jill Lawrie, for appellants, St. Jude Medical, Inc. and St. Jude Medical Canada, Inc.

Gavin MacKenzie and James Newland, for respondents, Erik Andersen, Yvonne Andersen and Sharon Frost.

The judgment of the court was delivered by

[1] Lax J.:—The only issue on this appeal is the quantum of costs awarded to the plaintiffs on a successful certification motion under the Class Proceedings Act, 1992, S.O. 1992, c. 6. Cullity J. fixed fees of the certification motion at $396,859.80 and disbursements at $213,841.05, for a total award of $610,700.85 plus applicable GST. The defendants have paid to the plaintiffs the amount of $156,000 comprising fees of $100,000 and disbursements of $56,000. They contend that this reflects an appropriate costs award for the motion.

Background to the Certification Motion

[2] The defendants are alleged to have manufactured, distributed and sold a defective heart valve that was released onto the market and implanted in patients without adequate testing and then failed to recall the product when its adverse consequences became apparent. In Canada, over 2000 Silzone valves were implanted in patients and over 1000 patients are included in the class that was certified by Justice Cullity. Parallel class proceedings are pending in Quebec, British Columbia and the United States. The plaintiffs' position is that the aggregate claims of class members are in an amount substantially in excess of $100,000,000.

[3] The plaintiffs' certification motion was originally supported only by affidavits sworn by the proposed representative plaintiffs and by a lawyer who had reviewed medical records and medical and scientific literature and other publicly-available information concerning the Silzone valve's release into the marketplace and its recall. The material filed by the defendants in response consisted of eight affidavits including affidavits of five expert witnesses whose specialties included cardiology, cardiac surgery, toxicology, psychiatry and haematology. The plaintiffs replied with extensive expert evidence.

[4] The defendants brought a motion to strike most of the plaintiffs' expert evidence and delivered two affidavits of a sixth expert, a pathologist. The plaintiffs' pathologist swore two affidavits in response. After the motion to strike was disposed of, the defendants delivered five further affidavits. The plaintiffs responded to this evidence and delivered further affidavits of a professor of clinical epidemiology and biostatistics and a specialist in bioengineering and biomedical research. The defendants responded by delivering five further affidavits from experts who had previously sworn affidavits on the certification motion and the plaintiffs delivered five expert affidavits in reply. In all, 14 experts swore 32 affidavits.

[5] Based on an extensive record that included 20 volumes of evidence and approximately five days of submissions in total (3 1/2 days for the certification motion), Justice Cullity ordered that this action be certified as a class proceeding [see 2003 5686 (ON SC), 38 C.P.C. (5th) 122]. The Divisional Court dismissed a motion for leave to appeal the certification order.

[6] The parties delivered extensive written submissions on costs of the certification motion. The successful plaintiffs sought an award of $997,825.55 for fees and $468,969.82 for disbursements of which $409,938.56 was for expert witness fees. In awarding a total amount of $610,700.85 plus applicable GST, the motion judge is said to have erred in three ways: (1) by not applying or not properly applying the principles of fairness, reasonableness, predictability and expectations; (2) by failing to have regard for the existing range of costs awards on certification motions; and (3) by awarding an amount that is so excessive as to require appellate intervention.

The Principles to be Applied in Reviewing a Costs Award

[7] The general principle for review of a discretionary decision was articulated by Chief Justice Lamer in Canadian Pacific Ltd. v. Matsqui Indian Band, [1995] 1 S.C.R. 3 at 32, 122 D.L.R. (4th) 129, quoting with approval the dictum of Lord Diplock in Hadmor Productions Ltd. v. Hamilton, [1982] 1 All E.R. 1042 at 1046, that an appellate court "must defer to the judge's exercise of his discretion and must not interfere with it merely on the ground that the members of the appellate court would have exercised the discretion differently."

[8] In Hamilton v. Open Window Bakery Ltd., 2004 SCC 9, [2004] 1 S.C.R. 303, [2003] S.C.J. No. 72 (QL), 235 D.L.R. (4th) 193, Arbour J. applied the principle to the review of a costs award. In that case, the Court of Appeal had set aside the trial judge's order for solicitor and client costs. She stated at para. 27:

[a] court should set aside a costs award on appeal only if the trial judge has made an error in principle or if the costs award is plainly wrong (citation omitted). ... In light of the privileged position of the trial judge to assess firsthand the credibility of witnesses, and given the highly fact-driven nature of the analysis that was required here, the costs order made by Wilkins J. must be restored.

[9] After the costs endorsement in this case was released on May 21, 2004, the Court of Appeal released three decisions that discuss the principles that are in issue on this appeal: Boucher v. Public Accountants Council for the Province of Ontario (2004), 71 O.R. (3d) 291, [2004] O.J. No. 2634 (QL) (C.A.) (released June 22, 2004); Moon v. Sher (2004), 246 D.L.R. (4th) 440, [2004] O.J. No. 4651 (QL) (C.A.) (released November 16, 2004); Coldmatic Refrigeration of Canada Ltd. v. Leveltek Processing LLC (2005), 75 O.R. (3d) 638, [2005] O.J. No. 160 (QL) (C.A.) (released January 24, 2005). Molloy J. considered these appellate decisions and granted leave to appeal on the basis that there was good reason to doubt the correctness of the decision of Cullity J. [138 A.C.W.S. (3d) 622].

[10] In Boucher, Armstrong J.A. concluded that the costs award was so excessive as to call for appellate intervention and reduced the total amount awarded from $187,682.51 to $63,000. In doing so, he found that the motion judge failed to address the overriding principle of reasonableness and committed an error in principle because (1) the amount of costs awarded on a partial indemnity scale was virtually the same as the amounts claimed on a substantial indemnity scale; and, (2) the amount was significantly greater than the costs awarded in an earlier proceeding before a different motion judge involving the same parties, the same arguments and similar issues.

[11] In Moon, the court recognized at para. 29 that Armstrong J.A. in Boucher, after considering a "number of factors unique to the case", reduced the award because he concluded it was not a fair and reasonable sum to award in the circumstances of the case. Similarly, in Moon, the court found the award to "greatly exceed what is fair and reasonable" because the motion was not complex, yet counsel spent almost 270 hours or seven full weeks in preparation for it and his law clerk and student spent almost the same amount of time. The motion judge fixed the total costs for one party at $85,000 and for the other party at $46,000. The Court of Appeal, at para. 44, varied the quantum to $40,000 and $27,500, amounts that were "more in keeping with what a losing party would reasonably expect to pay the winning party in this type of proceeding."

[12] The legal issues in Coldmatic were similar to those in Moon. The court, at para. 7, found that the motion judge erred in principle by awarding fees of $36,000 (the equivalent of three weeks of work charged at the highest hourly rate) for a two-hour motion with one day of cross-examination because "the case was not complex factually and the argument proceeded on the basis of settled principles of law." The award was reduced to $15,000.

[13] Although it may appear that these cases stand for the proposition that "excessive" awards constitute an error in principle, it is our view that the court interfered because the motion judge, in exercising his or her discretion, had failed to properly apply the principles of reasonableness, proportionality or consistency. These are related principles. A failure to apply proper principles may lead to an excessive award or one which is "plainly wrong", but excessiveness is an elusive and inherently subjective concept unless it is measured against something.

[14] In Moon and Coldmatic, the costs awards were measured against the complexity (or more accurately, lack of complexity) of motions of that kind. In Boucher, the award in question was measured against the amount of an award on the higher scale as well as against the amount of the earlier costs award. However, the costs in the earlier proceeding were determined prior to the introduction of the costs grid. It is generally recognized that the costs grid, which came into effect on January 1, 2002 and disappeared on July 1, 2005, produced costs awards that were significantly higher than they were before. As Nordheimer J. observed in Gariepy v. Shell Oil Co., [2002] O.J. No. 3495 (QL), 116 A.C.W.S. (3d) 495 (S.C.J.), at para. 17; affirmed [2004] O.J. No. 5309 (QL) (Div. Ct.), direct comparisons with pre-costs grid awards are neither possible nor particularly useful. We agree.

The Principles to be Applied in Fixing Costs

[15] The principles that should guide the exercise of discretion in fixing costs are reviewed in Boucher at paras. 24, 37 and 38:

[24] ... While it is appropriate to do the costs grid calculation, it is also necessary to step back and consider the result produced and question whether, in all the circumstances, the result is fair and reasonable. This approach was sanctioned by the court in Zesta Engineering Ltd. v. Cloutier, [2002] O.J. No. 4495 (QL), 21 C.C.E.L. (3d) 161 (C.A.) at para. 4 where it said:

In our view, the costs award should reflect more what the court views as a fair and reasonable amount that should be paid by the unsuccessful parties rather than any exact measure of the actual costs to the successful litigant.

[37] The failure to refer, in assessing costs, to the overriding principle of reasonableness, can produce a result that is contrary to the fundamental objective of access to justice... However, in my view, the chilling effect of a costs award of the magnitude of the award in this case generally exceeds any fair and reasonable expectation of the parties.

[38] In deciding what is fair and reasonable, as suggested above, the expectation of the parties concerning the quantum of a costs award is a relevant factor [citations omitted]. I refrain from attempting to articulate a more detailed or formulaic approach. The notions of fairness and reasonableness are embedded in the common law. Judges have been applying these notions for centuries to the factual matrix of particular cases.

[16] As Boucher observes, the notions of fairness and reasonableness are notions with which judges have been long familiar, although it is not evident that they are necessarily familiar with how to determine the reasonable expectations of the parties. In fixing costs, judges have traditionally had regard to the factors set out in Rule 57.01(1). This Rule includes the factors with which we are all familiar as well as the following two additional factors which came into effect on July 1, 2005:

57.01(1)(0.a)

the principle of indemnity, including where applicable, the experience of the lawyer for the party entitled to the costs as well as the rates charged and the hours spent by that lawyer;

(0.b)

the amount of costs that an unsuccessful party could reasonably expect to pay in relation to the step in the proceeding for which costs are being fixed;

[17] Although the costs grid has been revoked, judges who are tasked with fixing costs must continue to consider as a factor the experience of the lawyer, the rates charged and the hours spent in light of subrule (0.a) and in light of the information to the profession from the Costs Subcommittee of the Civil Rules Committee, which has published a guideline with rates and years of experience on a partial indemnity scale that is identical to the former costs grid. Costs on a substantial indemnity scale are to be 1.5 times higher. Thus, an hours and rates calculation will continue to be performed and serve as a guideline for allowable costs.

[18] Subrule 57.01 (0.b) reflects the dictum from the brief endorsement of the Court of Appeal in Zesta Engineering Ltd. v. Cloutier (2002), 21 C.C.E.L. (3d) 161, quoted with approval in Boucher and Moon. As stated in Boucher at para. 25, Zesta and Stellarbridge Management Inc. v. Magna International Canada Inc. (2004), 71 O.R. (3d) 263"confirmed a well-settled approach to the fixing of costs prior to the establishment of the costs grid as articulated by Morden A.C.J.O. in Murano v. Bank of Montreal (1998), 41 O.R. (3d) 222 (C.A.) at p. 249...". This approach emphasizes the overriding principle of reasonableness.

[19] The amount of costs that an unsuccessful party could reasonably expect to pay in relation to a step in the proceeding for which costs are being fixed is, in view of subrule 57.01(1)(0.b), one of the factors to be considered in fixing costs. Although Boucher does not explain how this is to be determined, the introduction of the Costs Outline in rule 57.01(6) will provide some measure of what the unsuccessful party expected as the court will now have available to it the hours, rates and amount claimed by the unsuccessful party who, if successful, would be seeking this amount for costs.

[20] An award of costs is a matter in the discretion of the judge by virtue of s. 131 of the Courts of Justice Act, R.S.O. 1990, c. C.43, and in light of the factors set out in rule 57.01(1) of the Rules of Civil Procedure, R.R.O. 1990, Reg. 194. The principle of deference is firmly established in our law and particular deference is due to an exercise of discretion by a judge who is case-managing a proceeding in light of the familiarity that case management judges acquire not only in relation to particular motions, but also with the proceedings as a whole: Khan v Metroland Printing, Publishing & Distributing Ltd. (2003), 68 O.R. (3d) 135, [2003] O.J. No. 4261 (QL) (Div. Ct.), at para. 5.

[21] This statement has even greater force in class proceedings. The motion judge not only develops familiarity with the particular proceeding, but also acquires unique expertise in a relatively new, different and complex area of law. This should be recognized and respected by appellate courts: Carom v. Bre-X Minerals Ltd. (2000), 51 O.R. (3d) 236 at 247, 196 D.L.R. (4th) 344 (C.A.), para. 36.

[22] The award in issue here is very large. It is probably the largest award for a class certification motion in Ontario, but in our view the magnitude of a costs award by itself, does not constitute a principled basis upon which to interfere. Appellate intervention based solely on quantum is problematic because there is no meaningful way to determine when a number is too high. However, an award of this magnitude warrants careful scrutiny because in exercising the discretion to award a significant amount for costs, we must be satisfied that the judge has exercised the discretion on proper principles and that the award is not "plainly wrong". We understand these principles to be:

1. The discretion of the court must be exercised in light of the specific facts and circumstances of the case in relation to the factors set out in rule 57.01(1): Boucher, Moon and Coldmatic.

2. A consideration of experience, rates charged and hours spent (formerly a costs grid calculation) is appropriate, but is subject to the overriding principle of reasonableness as applied to the factual matrix of the particular case: Boucher. The quantum should reflect an amount the court considers to be fair and reasonable rather than any exact measure of the actual costs to the successful litigant: Zesta Engineering.

3. The reasonable expectation of the unsuccessful party is one of the factors to be considered in determining an amount that is fair and reasonable: rule 57.01(1)(0.b).

4. The court should seek to avoid inconsistency with comparable awards in other cases. "Like cases, [if they can be found], should conclude with like substantive results": Murano at p. 249.

5. The court should seek to balance the indemnity principle with the fundamental objective of access to justice: Boucher.

6. A discretionary decision of a case-management judge in a class proceeding is entitled to a very high level of deference: Khan and Bre-X.

[23] With these principles in mind, we turn to the issue on appeal.

The Costs Decision

Fees

[24] The fees sought by the successful plaintiffs were $997,825.55 and they were awarded $396,859.80, including counsel fees. The plaintiffs' submissions were supported by time dockets and details of disbursements. However, the defendants did not put their own time dockets and expert witness expenses in evidence, nor provide the court with information that would have allowed the court to consider the relative expenditures by parties on both sides of the motion.

[25] An attack on the quantum of costs based on excess in circumstances where the court does not have before it the bills of all counsel has been the subject of judicial comment. In Risorto v. State Farm Mutual Automobile Insurance Co. (2003), 64 O.R. (3d) 135, 32 C.P.C. (5th) 304 (S.C.J.), which is a decision that pre-dates Boucher, Winkler J. observes at p. 307, that this "is no more than an attack in the air".

[26] Nordheimer J. echoes this comment in Hague v. Liberty Mutual Insurance Co., [2005] O.J. No. 1660 (QL), 138 A.C.W.S. (3d) 804 (S.C.J.), which is a post-Boucher decision. In asking exactly how the reasonable expectations of the parties are to be found out as part of the costs process, he states at para. 15:

...it would appear that the expectation of the parties will fall to be determined in one of two ways. It may be determined by the unsuccessful party revealing what his/her/its costs were on the same matter as some measure of what was expected. The unsuccessful party is, of course, not required to reveal that information, but, if they choose not to do so, they may impair their ability to make any meaningful submissions on this aspect of the process. (emphasis added). The other method and these are not mutually exclusive, is for the court to discern the likely expectation of the parties through the use of the factors to which courts have historically had reference in determining the appropriate amount of costs, some of which factors are set out in Rule 57.01. These include the nature of the motion, the importance of the issues raised in relation to the overall proceeding, the complexity of the issues, the conduct of the parties, and the actual costs incurred by the successful party.

[27] The recent amendments to rule 57.01(1) now require that a party seeking costs disclose the hours claimed and the actual rate charged by the lawyer, but where a party does not do this, we agree with the submission of the plaintiffs that the inference must be that the defendants devoted as much or more time and money in an attempt to defeat the motion for certification as the plaintiffs did in seeking certification. However, Cullity J. recognized that this is not determinative. It can only be some measure of what was expected and where both sides engage in overkill, as Cullity J. found was the case here, the court must search beyond this.

[28] He found that there was duplication of effort by lawyers in the three different firms representing the plaintiffs, as well as within the same firms and that a significant amount of time was concerned more with the merits of the issues to be tried than with issues on the certification motion. He states at para. 14:

A further relevant consideration is that the evidence filed by the plaintiffs extended well beyond what in my opinion was required to establish "some factual basis" for the certification requirements... The parties' willingness to engage in a battle of experts resulted in overkill on each side. I do not question the quality of the work done and of the services rendered by plaintiffs' solicitors for their clients. The case for certification was thoroughly and expertly prepared and presented. It does not follow that the defendants should reasonably be expected to pay for all the time expended. (emphasis added)

[29] He then went on to quote the following comment from the decision of the Divisional Court in Gratton-Masuy Environmental Technologies Inc. (c.o.b. Ecoflow Ontario) v. Building Materials Evaluation Commission, [2003] O.J. No. 1658 (QL), 122 A.C.W.S. (3d) 645

[16] ... [the award must reflect] more what the court views as a fair and reasonable amount that should be paid by the unsuccessful parties rather than any exact measure of the actual costs to the successful litigant. This is a fundamental concept in fixing or assessing costs.

[17] The amount at which costs are to be fixed is not simply an arithmetic function dependent on the number of hours worked and the hourly rates employed but rather, the party paying the cost should be subjected to an award which is fair and predictable. In other words, the party required to pay costs must not be faced with an award that does not reasonably reflect the amount of time and effort that was warranted by the proceedings.

[30] Gratton-Masuy is cited with approval in Moon at paras. 30 and 31 and encapsulates the principles discussed in the trilogy of costs decisions from the Court of Appeal. Thus, although Cullity J. did not have the benefit of these decisions, there can be little question that he was familiar with and understood the governing principles to be applied in fixing costs.

[31] He noted that the existence of the costs grid had brought about a "marked inflation" of the amounts, but it did not remove the discretion of the court under section 131 of the Courts of Justice Act and rule 57.01(1). He states in para. 16:

The hourly rates in the grid are maximum rates and the time to which they are applied must not be outside the range that the court considers was reasonably warranted by the facts and issues in the particular case.

[32] Finally, in para. 17, he concludes:

The question, again, is not whether over-lawyering and an excessive expenditure of time produced benefits for the clients. I do not believe the defendants should be required to pay for more than 40 per cent of the time charged.

[33] In the result, he fixed the fees, exclusive of counsel fees, at $381,659.80.

[34] The defendants submit that Cullity J. erred in failing to apply two overlapping principles: first, that he failed to "step back" after a costs grid calculation and consider whether the result was a fair and reasonable amount for the defendants to pay; second, that he failed to apply the principle that a fair and reasonable result must reflect the reasonable expectations of the parties. The defendants submit that the range of awards in other certification motions informs reasonable expectations.

[35] We do not accept the first submission. In the passages we have referred to, it is apparent that Cullity J. did "step back". He did not accept the costs grid calculation at face value. He did not treat the fixing of costs as simply a mechanical exercise of multiplying hours by specified rates. He accepted the hourly rates, but took into account that there was duplication of time and work and that the hours docketed in two of the three firms was excessive and that the work in connection with the expert evidence went beyond that which was warranted for a certification motion.

[36] He found that in the circumstances of this case, the plaintiffs should receive a substantial award of costs, but one that was significantly less than the one requested. He considered the time spent in light of the factors set out in rule 57.01(1), but then asked what amount was reasonable for the defendants to pay. He concluded that this was a reduction of 40 per cent of the bill. In our view, Cullity J. expressly considered reasonableness from the point of view of the paying party and this satisfies the "stepping back" requirement.

[37] The second submission raises the use of comparators as a measure of reasonable expectations. It is submitted that Cullity J. erred in failing to award an amount that was within a range of awards on certification motions and therefore did not consider reasonableness in the context of predictability. This was the main reason that Molloy J. granted leave to appeal.

[38] In his reasons, Cullity J. says the following:

[5] I accept that, in exercising the discretion of the court to award costs ... the court should seek to avoid clear inconsistency with awards in other cases. However, a determination of consistency, or inconsistency, requires a judgment with respect to the degree of similarity with the previous decisions.

[7] I have no difficulty in accepting that the amounts claimed in this case are sufficiently greater than those awarded in earlier decisions to give rise to at least an appearance of inconsistency. I do not however find it useful to attempt to extrapolate from those decisions even a range within which an award of costs in this case should appropriately be made. Class actions are case-managed for a good reason. Some cases move smoothly and relatively quickly to the hearing of the motion to certify the proceedings. Others, however, require numerous case conferences to deal with productions, representation and other procedural matters that lead up to the hearing of the certification motion. I agree with Nordheimer J. in Gariepy v. Shell Oil Co., [2002] O.J. No. 3495 (QL) (S.C.J.), (at para 17) that, without an intimate understanding of the course of proceedings in other cases, it is difficult – and I think – dangerous to attempt to make comparisons that will provide firm guidance in determining an appropriate level of costs in respect of a particular motion.

[12] I accept [the plaintiffs'] submission that the discretion of the court must be exercised in the light of the specific facts and circumstances of this case and that the assistance to be derived from the amounts awarded – as distinct from the principles applied – in other cases is limited. ...

[39] The defendants submit and Molloy J. accepted that the statement of Nordheimer J. in Gariepy upon which Cullity J. relied, does not support the proposition for which it was cited. We respectfully disagree.

[40] Gariepy was a case where the plaintiffs were not successful on a certification motion and resisted an award of costs on the basis that it would impede access to justice. In that case, the plaintiffs argued that if costs were awarded, it should not exceed the range of $15,000. In his reasons, Nordheimer J. referred to the awards of Winkler J. in Lau v. Bayview Landmark Inc., [1999] O.J. No. 4385 (QL), 92 A.C.W.S. (3d) 752 (S.C.J.), and Brockenshire J. in Nantais v. Telectronics Proprietary (Canada) Ltd., [1996] O.J. No. 5205 (QL) (Ont. Ct. (Gen. Div.)), where the costs of certification motions were respectively fixed at $52,831 and $96,113.66. After observing that"[i]n either case, the amount awarded for costs represents the costs that were appropriate for that motion", Nordheimer J. says the following in para. 17:

In any event, the difficulty with making comparisons between cases is readily apparent. I have no information on what was involved in those cases in terms of the number of counsel, the issues raised, the amount of costs sought, the time spent, the length of argument, etc. Of equal importance is the fact that each of those cases was decided before the corning into force of the cost grid. It is generally recognized that the cost grid has had a significant impact on the issue of costs in a number of respects, not the least of which is the quantum of costs being awarded. Direct comparisons are therefore neither possible nor particularly useful. I would observe in a very general way, however, that the amount of costs fixed in the cases to which the plaintiffs referred me appear to be very modest given the complexity that is normally involved in a certification motion.

[41] While we accept that in the above passage Nordheimer J. gives an additional reason for rejecting comparators (the inability to directly compare pre-costs grid awards), his comments are of a more general nature and are repeated in Hague at para. 9:

Put shortly, comparisons to other cases are of very limited use. I appreciate that one should strive for some measure of consistency in terms of costs awards but the more complicated and fact specific any step in an action is, then the more reason there is for differentiation to occur between cases in terms of the amount of costs that are ultimately fixed. Certification motions, like summary judgment motions and motions for interlocutory junctions, tend to be more complex and time intensive. They also tend to involve matters that are of greater overall significance to the parties. All of these factors not only generally result in costs awards that are larger in amount, they also make comparisons with other decisions more difficult, and arguably unsuitable.

[42] We find no inconsistency between the statements of Nordheimer J. in Gariepy and Hague nor any error in Cullity J.'s reliance on Gariepy for the proposition that without an intimate understanding of the course of proceedings in other cases, it is difficult to attempt to make comparisons that will provide firm guidance in determining an appropriate level of costs in respect of a particular motion, (emphasis added). In Mandeville v. Manufacturers Life Insurance Co., [2002] O.J. No. 5388 (QL), 133 A.C.W.S. (3d) 406 (S.C.J.), Nordheimer J. also states, at para. 28"comparisons are probably not very useful".

[43] The defendants concede that the certification motion was complicated and required considerable effort and time, warranting a costs award at the "high end of a reasonable range of costs awards on certification motions". They rely on Nantais, where Brockenshire J. awarded costs of $96,113.66 to the plaintiffs on a motion for certification in a case involving an allegedly defective medical device. As this is a 1996 pre-costs grid decision where only one expert affidavit was filed, it is not a useful comparator.

[44] One can find post-costs grid certification motions where costs were awarded in amounts in the range of $100,000. In Gariepy, the two defendants were awarded a total of $175,000. In Pearson v. Inco Ltd., [2002] O.J. No. 3532 (QL), 116 A.C.W.S. (3d) 634 (S.C.J.), aff'd, [2004] O.J. No. 317 (QL), 128 A.C.W.S. (3d) 875 (Div. Ct.), rev'd on other grounds, 2006 913 (ON CA), [2005] O.J. No. 4918 (QL), 261 D.L.R. (4th) 629 (C.A.), the four defendants were awarded a total of approximately $185,000. However, one can also find certification motions or motions that are comparable in terms of their length, complexity and importance where the costs that were allowed are significantly greater.

[45] In Mandeville, which was a Rule 21 motion and a certification motion, both argued over three days, Nordheimer J. awarded fees to the plaintiff (in part on a substantial indemnity scale) of $324,000 and total costs of $430,350. In Delrina Corp. (c.o.b. Carolian Systems) v. Triolet Systems Inc., [2002] O.J. No. 3729 (QL), 117 A.C.W.S. (3d) 31 (C.A.), the costs of the appeal including a three day hearing that disposed of the action were fixed on a substantial indemnity scale in the amount of $526,464.42, with fees of $477,325 without reference to comparators.

[46] The defendants sought to distinguish Mandeville on the basis that the defendants in that case had agreed to an award in the range of $400,000 and that otherwise, Nordheimer J. would have fixed the costs in a lesser amount. Although Nordheimer J. was reluctant to go outside a range of costs suggested by counsel without good reason, he specifically found that he was not bound to fix the costs in an amount that is at or above the amount the unsuccessful party submits is a fair amount any more than he was bound by the amount suggested by the successful party.

[47] Before fixing the costs in the total amount of $430,350, an amount he arrived at following a careful analysis, Nordheimer J. referred to the costs award in Delrina where the amount was $100,000 higher and to the awards in Gariepy and Pearson which were substantially less. In the result, he fixed the costs in an amount that he viewed as appropriate to the motions that were before him notwithstanding that this amount was more than the twice the amount he had awarded in Gariepy and Pearson only a few months earlier and these were comparable motions in terms of their length and complexity.

[48] It is perhaps useful at this juncture to be reminded that the awarding and fixing of costs is an exercise of discretion in which a judge attempts to do procedural and substantive justice between the parties. Cullity J. was well aware that the costs sought (and ultimately awarded) were sufficiently greater than those awarded in earlier decisions so as to give an appearance of inconsistency. From his reasons, it is evident that he did not consider the earlier decisions to be useful as comparators and he explains why. The effect of the defendants' submission is that the costs awarded to a successful party on a certification motion would never exceed the range of $100,000 notwithstanding that the proceedings may be entirely different. We agree with the comments of Nordheimer J. in Hague and with Cullity J. in this case that it is difficult and of limited value to make comparisons with other costs awards in very complex, high stakes, fact specific cases without an intimate understanding of the other cases. These are not routine motions.

[49] Moreover, to inflexibly fix costs in an amount that is within a range of earlier decisions may satisfy the consistency principle, but it would seriously undermine the discretion given to a judge by the legislature in s. 131 of the Courts of Justice Act. As part of the exercise of discretion, the judge must decide whether the case at bar can usefully be compared with other cases and this requires a judgment with respect to the degree of similarity with other cases. It would also impair the ability of the judge to do procedural and substantive justice between the parties in determining a fair and reasonable amount in the particular circumstances of the case. The result would be consistent, but artificial and arbitrary.

[50] We of course accept and are bound by the principle articulated by Morden A.C.J.O. in Murano to which we have earlier referred that "like cases should conclude with like substantive results". We also agree that it is a desirable objective to have some measure of consistency in awards of costs on similar motions, but this presupposes that like cases can be found. In our view, an experienced class proceedings judge who has case-managed the proceeding is in the best position to determine this. The determination is essentially a factual one in light of the judge's experience with the particular proceeding leading to the certification motion and in light of his or her experience more generally with class proceedings. It is entitled to considerable deference.

Counsel Fees

[51] For the sake of completeness, we mention that Cullity J. declined to follow the reasoning in Banihashem-Bakhtiari v. Axes Investments Inc. (2003), 66 O.R. (3d) 284, [2003] O.J. No. 3071 (QL) (S.C.J.), where Lane J. concluded that fees for multiple counsel are permissible under the costs grid, but that the aggregate of counsel fees cannot exceed the maximum permitted. Cullity J. awarded three counsel fees for three and one-half days, which, in aggregate, exceed the maximum. Subsequently, in Celanese Canada Inc. v. Canadian National Railway Co., [2005] O.J. No. 1122 (QL), 138 A.C.W.S. (3d) 23 (C.A.), Borins J.A. at paras. 45-51, analyzed the conflicting jurisprudence and followed Lane J.'s approach. See also, Walker v. Ritchie, [2005] O.J. No. 1600 (QL), 138 A.C.W.S. (3d) 1156 (C.A.), at paras. 101 and 102. As this point was not argued, we do not propose to vary this aspect of the award, although counsel may agree to do so.

Disbursements

[52] The plaintiffs sought $468,969.82 for disbursements ($409,938.56 for expert witness fees) inclusive of GST and were awarded $213,841.05 ($192,781.61 for expert witness fees) plus GST. The defendants submit that Cullity J. erred in failing to reduce the amount claimed for expert witness fees on the basis that the expert opinions went to the ultimate merits and were of substantial value to the plaintiffs' preparation for trial. There is no merit in this submission.

[53] Cullity J. observed that in fixing the costs of a certification motion, it is difficult to distinguish the time required in preparing for a motion to certify from the preparation of the trial of common issues. He correctly noted that the motion is essentially procedural and not concerned with the merits of the issues raised by the pleadings, although evidence going to the merits may be relevant. He said that the plaintiffs allowed themselves to be drawn into issues that related increasingly to the merits of the action and that a contest between the experts on each side ensued.

[54] He carefully considered the amounts paid to each of the plaintiffs' expert witnesses. He did not allow any of the fees paid to one of the plaintiffs' witnesses and substantially reduced the amounts paid to the others because they were not "reasonably necessary for the purposes of the motion". He found that their evidence "exceeded to a significant extent that which ... was reasonably necessary for the purpose of certification". Although he did not approach it in the same way, the reduction in the amount paid to the expert witnesses is roughly proportionate to the reduction of 40 per cent that he found appropriate for fees.

The Total Award

[55] A final submission advanced by the defendants is that an award of this magnitude will have a chilling effect on class proceedings. We do not find this submission compelling in circumstances where the defendants, at least initially, drove the plaintiffs into a game of high stakes poker, sparing no expense in marshalling evidence and then declined to put their own costs before the court. Having lost a very expensive and important motion, it is disingenuous for the defendants to now claim that the costs award is outside the range of what they reasonably expected. If the plaintiffs had lost the motion, it similarly would not lie in their mouths to make this submission.

[56] A similar argument was advanced and rejected in Gariepy, Pearson and Fehringer v. Sun Media Corp., [2002] O.J. No. 5514 (QL) (S.C.J.), where the plaintiffs who were unsuccessful on certification motions resisted an award of costs in part on the basis that this would deter the bringing forward of class actions. The Court of Appeal in Pearson did not address this point. We do not need to elaborate on this further except to agree with the observation of Cullity J. in the case at bar at para. 15:

In cases other than class proceedings, the clients' obligation to pay their lawyers' fees may assist solicitors to resist a temptation to engage in overkill. In class proceedings – where contingency fees are the norm and costs awards are only in a strained sense reconcilable with the concept of a partial, or substantial indemnity for fees to be charged to the client – the temptation may be less easily resisted.

[57] Finally, we do not accept the defendants' submission that this award will set a new threshold for future costs awards in class proceedings. There is no evidence that the large award in Mandeville in 2002 ushered in a new higher range of awards on certification motions. Subsequent awards are in lesser amounts and vary according to the circumstances of each case.

[58] In reaching the conclusion that we should not interfere with this award, we were greatly assisted by the careful and detailed reasons of Cullity J. In exercising his discretion to award a significant amount for costs, he recognized that a comprehensive analysis was required. We have earlier said that the magnitude of an award is not determinative of its correctness, but there is a high obligation to adequately explain the basis for it.

[59] For these reasons, the appeal is dismissed. If costs are not agreed, the parties are to file Costs Outlines with the Registrar of the Divisional Court within 30 days, in accordance with rule 57.01(6).

[60] Appeal dismissed.