DATE: 20030408
DOCKET: C37858
COURT OF APPEAL FOR ONTARIO
McMURTRY, C.J.O., CATZMAN and ROSENBERG JJ.A.
B E T W E E N :
SEHDEV KUMAR
Paul J. Pape for the appellant
Appellant
- and -
THE MUTUAL LIFE ASSURANCE COMPANY OF CANADA, and PRUDENTIAL ASSURANCE COMPANY LIMITED
F. Paul Morrison and Dana M. Peebles for the respondents
Respondents
Heard: November 5 & 6, 2002
On appeal from the Order of the Divisional Court (McRae, Then and Day JJ.) dated December 14, 2001, dismissing the appeal of the appellant from the Order of Cumming J. dated October 18, 2000.
ROSENBERG J.A.:
[1] This is one of two appeals heard together by this court concerning the Class Proceedings Act, 1992, S.O. 1992, c. 6. Both are appeals from refusals by motions judges to certify, as class actions under the CPA, claims originating from the sale of so-called “premium offset” or “vanishing premiums” participating whole life insurance policies.
[2] The appellant and another person brought a motion for certification on behalf of a proposed class of persons who had allegedly received premium offset representations from the respondent in a “common manner.” Cumming J. dismissed the motion and the Divisional Court dismissed the appellant’s appeal from the order of Cumming J. as well as the companion appeal, Zicherman v. The Equitable Life Assurance Company of Canada.
[3] In my view, the motions judge and the Divisional Court were correct in denying certification. For the following reasons, I would dismiss the appeal.
THE FACTS
[4] The appellant Sehdev Kumar and the other proposed representative plaintiff Rosel Williams brought actions against Mutual Life Assurance Company of Canada and Prudential Assurance Company Limited for having deceptively sold them life insurance policies with a premium offset feature. The Canadian business of Prudential was acquired by Mutual on March 1, 1995. Mutual’s name has since been changed to Clarica Life Insurance Company.
[5] The appellant is a semi-retired university professor. He is married and has two children. In 1991, when he was 49 years old, he purchased a $500,000 whole life insurance policy from the Prudential. When Ms. Williams purchased her policy in 1989, she was a 23-year-old single mother living with her parents. She purchased a Prudential whole life policy with $100,000 coverage. She has since settled with the respondents. I will, however, make brief reference to the facts of her case since they demonstrate some of the problems with the proposed class action.
[6] The appellant’s policy was a participating permanent whole life policy. As a participating policy holder the appellant was entitled to dividends in any year in which the company declared them. The size of the dividends depended upon the performance of the company’s investments. The theory behind a premium offset or vanishing premium policy is that at a certain point in time the accumulated dividends would be sufficient to pay or “offset” future premiums as they came due. Whole life policies with vanishing premiums pay dividends to policyholders that purportedly cover the cost of premiums within a number of years, upon a specified "cross-over" date. These policies apparently accomplish their objectives so long as interest rates are high. If rates decline, the dividends fail to produce sufficient income to pay the remaining premiums, and insurers move back the "cross-over" dates by several years. Using the premiums to offset future premiums was not the only way that participating policies were sold. For example, the policyholder could use the dividends to buy additional term insurance.
[7] The appellant claims that before he bought the policy in 1991 he was told that after nine years he would no longer have to pay premiums. As a university professor, this representation was important to him because he knew that he would be required to retire at age 65 and live on a reduced income.
[8] The appellant claims that the two Prudential agents with whom he dealt, the third party defendants Chagger and Piruchta, showed him a computer generated illustration for a Life 2,000-PEP Series A insurance policy. The appellant wrote across the top of this illustration, “In this you pay premiums for nine years until age 58”. The illustration itself included a disclaimer:
The Paid-up additions and their Cash Values are based on the Company’s current dividend scale and ARE NOT GUARANTEED. In our present environment, current dividend scales are expected to change more frequently, thus long term projections should be viewed with caution.
[9] The policy itself also stipulated that the annual premiums are payable for the duration of the contract, that is until death or the surrender of the policy. The appellant claims that in 1995, Piruchta told him he would have to pay premiums for a total of twelve years, not nine years and for the first time explained to him that the duration of premium payments was dependent upon the dividends declared by the company. The appellant accordingly cancelled the policy and complained to Mutual and expressed his intent to make a complaint to the Life Insurance Commission of Ontario.
[10] He then commenced a class action for damages for, inter alia, negligent misrepresentation. He also brought claims for breach of contract, failure to warn, breach of fiduciary duty, unjust enrichment, tort of deceit and breach of s. 52(1) of the Competition Act, R.S.C. 1985, c.C-34. On behalf of the class, he sought damages equivalent to (a) the premiums payable after the specified "vanishing premium" dates and to (b) the reduced amount of the policies' cash surrender value.
[11] Ms. Williams met once with the agent that sold her the policy. The agent proposed a particular type of policy and she agreed. She did not receive any illustration when she purchased the policy in 1989. She was subsequently sent two illustrations from a rival company and later two further illustrations from her agent. Ms. Williams alleged that she was told that she would only have to pay premiums for ten years. She was subsequently told that the premiums would not “vanish” after ten years and she stopped paying the monthly premiums as of January 1997.
Prudential’s use of vanishing premium policies
[12] In response to demands from its agents and to remain competitive in the market place where the premium offset feature had become popular, Prudential made this feature available in 1982. Between 1982 and 1986, premium offset was not actively promoted by Prudential. It was however, available at the request of its agents. Prudential began actively marketing this feature of its policies from July 1986 to March 1995. Over 120,000 whole life policies were sold during this period. There is apparently no way to tell how many whole life policyholders were told about the premium offset feature, were shown illustrations, or chose to use this feature to reduce their annual premiums.
[13] Prior to 1988, there was no formal training structure in place for Prudential sales agents. Rather, the managers of each office trained agents. In 1988, Prudential developed a training programme for new agents to educate them about specific insurance products. As early as July 1982, Prudential had developed software for generating computer illustrations of premium offsets. These illustrations came with pre-printed warnings. In 1986, Prudential began disseminating sales bulletins to its sales agents promoting the premium offset concept.
[14] In 1990, Prudential produced a Life 2000 Series Marketing Guide. The Guide illustrated premium offset as an available feature for several policies. However, it did not explicitly direct sales agents to warn prospective customers of the contingency that premium offset is dependent on dividend performance.
[15] Prudential’s first dividend scale decrease occurred in 1992 as a result of declining interest rates. In 1993, Prudential began advising its agents to present multiple scale illustrations to customers showing future values based on three projections: dividends higher than, equal to and lower than the current rate. Prudential maintains that these were not a guarantee or a prediction by Prudential of future developments regarding dividend scales or interest rates. Rather, the illustrations merely set out the current dividend scales and interest rates.
[16] The motions judge described some of the potential problems with the premium offset feature:
In theory, if not always in practice, the client’s choice of policy would depend upon the individual’s needs and ability to pay. It is apparent that there were problems throughout the life insurance industry with the so-called “premium offset” feature to some whole life policies. Perhaps it is not too cynical to imagine that many agents might over-emphasize and over-simplify this optional feature given the attractive dividends being paid. Perhaps many people tended to expect that the then-prevailing economic conditions would persist into the future. The certainties of the present are more apparent than the uncertainties of the future. The “premium offset” feature is dependent, of course, upon the payment of dividends and the quantity thereof. Thus, the superficial appeal of a projected future “premium offset” date is fraught with all the uncertainties attendant upon the political economy and the reality that unfolds when the future becomes the present (at para. 7).
[17] After this action was commenced Clarica established an ADR programme to resolve complaints such as the appellant’s. The respondents received 239 complaints over the years about premium offset policies. About half of the complaints were received after the ADR programme was established.
THE DEFINITION OF THE CLASS AND COMMON ISSUES BEFORE THE MOTIONS JUDGE
[18] The appellant proposed that the class be defined as follows:
All owners of Class Policies purchased from Prudential. Class Policies being defined as
Any participating whole life policy issued by Prudential between January 1, 1980 and December 31, 1995 which is in force as of August 31, 1998 (or “Current Class Policy”) or which has become a Lapsed Policy between January 1, 1990 and August 13, 1998 (a “Lapsed Class Policy”) except those policies in respect of which the owners have released Prudential from claims related to premium offset or to the sale of the policies.
[19] The appellant proposed the following common question:
Did the use of illustrations and/or representations, in writing or verbal, create an obligation on the part of Prudential with respect to a specified offset date despite the terms of the policy and the terms of any illustration?
THE REASONS OF CUMMING J.
[20] The motions judge found that there was no evidence to support “the plaintiffs’ bald allegation of uniformity in Prudential’s sales techniques and materials over the 1980-1995 period”. Rather, the “particular circumstances of the individual client and the oral statements and explanations of the particular agent were of fundamental importance”. This would involve examination of the unique sales experience of each policyholder. The motions judge then examined the elements of negligent misrepresentation as set out in Queen v. Cognos Inc., 1993 146 (SCC), [1993] 1 S.C.R. 87 at 110. Those elements are:
(1) there must be a duty of care based on a "special relationship" between the representor and the representee; (2) the representation in question must be untrue, inaccurate, or misleading; (3) the representor must have acted negligently in making said misrepresentation; (4) the representee must have relied, in a reasonable manner, on said negligent misrepresentation; and (5) the reliance must have been detrimental to the representee in the sense that damages resulted.
[21] The motions judge had the greatest concern with elements (2) to (5). In summary, he found that an individual determination would have to be made with respect to each policyholder for each of those elements. For example, as to element (2) there would have to be evidence as to whether any representation as to premium offset was made and then, if so, whether or not it was accurate. In summary, he concluded that, “While the theories of liability can be phrased commonly the actual determination of liability for each class member can only be made upon an examination of the unique circumstances with respect to each class member’s purchase of a policy”. Accordingly, he found that the claims of the class members do not raise common issues.
[22] Even if there was a common issue, the motions judge found that a class proceeding would not be the preferable procedure as required by s. 5(1)(d) of the Act. He reviewed the three policy objectives underlying the Act of access to justice; judicial economy; and behaviour modification. He was also of the view that the court must consider whether certification would be a “fair, efficient and manageable way of advancing the claims” referring to the decision of the Divisional Court in Carom v. Bre-X Minerals Ltd. (1999), 1999 19916 (ON SCDC), 46 O.R. (3d) 315. [That decision was reversed in part by this court in reasons reported at (2000), 2000 16886 (ON CA), 51 O.R. (3d) 236. Leave to appeal to the Supreme Court of Canada was denied, [2000] S.C.C.A. No. 660.]
[23] The motions judge considered that the necessary inquiry into individual issues “would significantly increase the time, cost and complexity of the proceedings such that the CPA’s objective of access to justice and judicial efficiency would be impeded if not frustrated”. He therefore found that a class proceeding was not the preferable procedure.
THE REASONS OF THE DIVISIONAL COURT
[24] The Divisional Court dealt with the appeal from the judgment of Cumming J. and the appeal from the decision of Ferrier J. in the matter of Zicherman v. The Equitable Life Assurance Company of Canada. By the time the cases came before the Divisional Court, this court’s decision in Bre-X had been released. Writing for the Divisional Court, McRae J. held that the Bre-X decision merely stood for the proposition that “[W]here there is certification for a number of common issues, judicial expediency is best served if all issues are canvassed in the same action” (at para 23).
[25] Before the Divisional Court, Mr. Pape, who had not appeared before the motions judge, argued that if the motions judge was not satisfied with the common issue proposed by the plaintiff, he should have reformulated the issue. Mr. Pape suggested that the issue might be defined as, “[W]as there an organized and systematic marketing of premium offset policies by the insurance company which was misleading?” The Divisional Court held that there was no such obligation on the motions judge and, in any event, even if the common issue was reformulated as suggested, there was no evidence of common complaint with respect to this issue:
The issue is not with respect to the use of illustrations or the systematic marketing of "premium offset" policies by the insurance companies, but rather, some individual complaints by some clients about the sales approaches of some agents. Many tens of thousands of policies were sold by hundreds of agents, but a relatively small number of purchasers complained about representations allegedly made to them by agents at the time of sale. These transactions do not present common issues but, rather, individual representations (at para. 11).
[26] The Divisional Court also held that the motions judge made proper use of the evidentiary record. He recognized that he should not examine the facts with a view to determining the strength of the allegation. Rather, his responsibility was to “look to the allegation to determine if there was an identifiable class and common issues which would merit certification” (at para. 10). The Court agreed with the motions judge’s decision and dismissed the appeal.
THE ISSUES
[27] The appellant raises three issues:
(1) Did the Divisional Court err in principle in finding that the appeal must be approached solely on the basis of the issues as presented to the motions judge?
(2) Did the courts err in principle in their use of the evidentiary record to determine there were no common issues?
(3) Did the motions judge and the Divisional Court err in principle in finding there were no common issues to be certified?
[28] It seems to me that the appellant cannot succeed without also showing that the motions judge erred in finding that a class proceeding was not the preferable procedure for the resolution of the common issues.
THE LEGISLATION
[29] The relevant parts of the CPA are ss. 5 and 6, which provide as follows:
5 (1) The Court shall certify a class proceeding on a motion under section 2, 3 or 4 if,
(a) the pleadings or the notice of application discloses a cause of action;
(b) there is an identifiable class of two or more persons that would be represented by the representative plaintiff or defendant;
(c) the claims or defences of the class members raise common issues;
(d) a class proceeding would be the preferable procedure for the resolution of the common issues; and
(e) there is a representative plaintiff or defendant who,
(i) would fairly and adequately represent the interests of the class,
(ii) has produced a plan for the proceeding that sets out a workable method of advancing the proceeding on behalf of the class and of notifying class members of the proceeding, and
(iii) does not have, on the common issues for the class, an interest in conflict with the interests of other class members.
(2) Despite subsection (1), where a class includes a subclass whose members have claims or defences that raise common issues not shared by all the class members, so that, in the opinion of the court, the protection of the interests of the subclass members requires that they be separately represented, the court shall not certify the class proceeding unless there is a representative plaintiff or defendant who,
(a) would fairly and adequately represent the interests of the subclass;
(b) has produced a plan for the proceeding that sets out a workable method of advancing the proceeding on behalf of the subclass and of notifying subclass members of the proceeding; and
(c) does not have, on the common issues for the subclass, an interest in conflict with the interests of other subclass members.
The court shall not refuse to certify a proceeding as a class proceeding solely on any of the following grounds:
The relief claimed includes a claim for damages that would require individual assessment after determination of the common issues.
The relief claimed relates to separate contracts involving different class members.
Different remedies are sought for different class members.
The number of class members or the identity of each class member is not known.
The class includes a subclass whose members have claims or defences that raise common issues not shared by all class members.
ANALYSIS
(1) The role of the motions judge and the Divisional Court
[30] If the reasons of the Divisional Court stand for the proposition that the motions judges should not modify the definition of the class or the common issues as presented by the plaintiff then, in my view, this was an error. I can see no reason in principle why the motions judge cannot modify the definition of the class or the common issues if the judge is of the view that such modification is required to accord with the Act. Further, in my view, it was open to the Divisional Court, in this case, to modify the definition of the class or the common issue. In Anderson v. Wilson (1998), 1998 18878 (ON SC), 37 O.R. (3d) 235, the Divisional Court amended both the class and the common issues as certified by the motions judge (1997), 1997 12104 (ON SC), 32 O.R. (3d) 400. On further appeal, (1999), 1999 3753 (ON CA), 44 O.R. (3d) 673, this court again varied both the class and the common issues. Similarly, in Hollick v. Toronto (City), 2001 SCC 68, [2001] 3 S.C.R. 158 at para. 21, McLachlin C.J.C., speaking for the court, held that where the class proposed by the plaintiff could be defined more narrowly, “the court should either disallow certification or allow certification on condition that the definition of the class be amended… .”
[31] While I would not go so far as to suggest that there is a duty on the motions judge to modify the definitions, the class or the common issue, it is certainly open to the judge to do so. It should be borne in mind that the judges hearing these motions are experienced with managing CPA cases. They are entitled to bring their experience to bear in formulating the class and the common issue, provided, of course, that the parties are not unfairly prejudiced.
[32] That said, in this case the issue is of no real consequence in this appeal. As I have pointed out, the Divisional Court did consider the reformulation of the common issue suggested by counsel for the appellant. The court held that even as reformulated, the proceeding should not be certified as a class action.
[33] In this case, I can see no prejudice to the respondents in considering the reformulated question. In fairness to the appellant, the class and common issue as initially proposed were drawn from the decision in Dabbs v. Sun Life Assurance Co. of Canada (1998), 1998 14855 (ON SC), 40 O.R. (3d) 429 (Gen. Div.)). Dabbs was a motion for certification and approval of a settlement in a vanishing premium case. Thus, the defendant was consenting to the certification, provided the motion judge also approved the settlement. The motions judge in Dabbs found that the proposed class represented an identifiable class within the meaning of s. 5(1)(b) of the Act and that the statement of claim raised a common issue.
[34] Further, when the instant matter was before the motions judge in September 2000, neither he nor the parties had the benefit of the decisions of the Supreme Court of Canada in Hollick and Rumley v. British Columbia, 2001 SCC 69, [2001] 3 S.C.R. 184. It is apparent that in reformulating the common issue the appellant is attempting to bring his case within Rumley. In this case, in view of the extensive record before the motions judge, I can see no unfairness to the respondents in considering the reformulated common issue.
(2) Use of the evidentiary record
[35] The appellant submits that the motions judge erred in using the evidentiary record provided by the respondents to determine the merits of the proposed claim. In Hollick, the court dealt with the use of evidence on the certification motion. The court held at para. 16 that “the certification stage is decidedly not meant to be a test of the merits of the action”. Thus, the “question at the certification stage is not whether the claim is likely to succeed, but whether the suit is appropriately prosecuted as a class action”.
[36] The Supreme Court then considered the proper use of the evidentiary record. McLachlin C.J.C. held at para. 25 that the class representative “must show some basis in fact for each of the certification requirements set out in s. 5 of the Act, other than the requirement that the pleadings disclose a cause of action”. Section 5(1)(a) requires that the pleadings disclose a cause of action. This requirement is governed by the rule that pleadings should not be struck for failure to disclose a cause of action unless it is “plain and obvious” that no claim exists.
[37] The motions judge made a number of findings of fact based upon the material submitted by the respondents. For example, he stated that the “documentation produced does not support the plaintiffs’ allegation that Prudential engaged in any organized and systemic marketing of premium offset insurance policies” (at para. 17). He also stated that the “evidentiary record establishes that Prudential … did not require or encourage its agents to use the concept, did not train its agents to use the concept in any specific or common way and provided cautions and warnings about premium offsets in its directions to its agents and in the illustrations they employed” (at para. 18). Further, he stated that the “evidence establishes that agents gave unique and individually targeted presentations to each different prospective policy purchaser” (at para. 19).
[38] As I understand it, the respondents’ evidence was adduced to support its position that there was no common issue of fact raised by the pleadings. The evidence was also relevant to the proposed class definition. As held in Hollick at para. 20, there must “be some rational relationship between the class and common issues”. The respondents’ evidence was relevant to this issue as well. The appellant proposed a class comprising over 100,000 policyholders. The respondents were entitled to show that many of those potential class members never received any illustrations and thus that there was no relationship between the class and the proposed common issue. Indeed, the other original plaintiff, Ms. Williams, did not receive any written illustration before buying her policy.
[39] In my view, the motions judge did not err. He accurately summarized the respondents’ evidence and the appellant did not contradict that evidence. The legal consequences flowing from that evidence are another matter and involves consideration of the finding by the motions judge that there was no common issue and that, in any event, the class proceeding would not be the preferable procedure for the resolution of the common issue.
(3) The common issues
[40] Relying on Dabbs, the appellant proposed the following common issue before the motions judge:
Did the use of illustrations and/or any representations, in writing or verbal, create an obligation on the part of Prudential with respect to a specified offset date despite the terms of the policy and the terms of any illustration? (at p. 433)
[41] In light of Rumley, the appellant refined the common issue and framed it as follows:
Was there an organized and systematic marketing of premium offset policies by Prudential which was misleading?
I do not see a radical difference between these two proposals.
[42] In the course of oral argument in this court, Mr. Pape seemed to suggest a different common issue:
Was Prudential negligent in failing to properly train its agents in the use of the premium offset feature?
[43] In my view, however the proposed common issue is framed, the motions judge and the Divisional Court were correct in refusing to certify this class action. For the purposes of this discussion I will assume in the appellant’s favour that there is an identifiable class of two or more persons within the meaning of s. 5(1)(b). The respondents’ evidence shows that Clarica has received several hundred complaints about premium offset policies. This evidence is similar to the evidence relied upon in Hollick showing that of the potential class of 30,000 persons who lived in the vicinity of the Keele Valley landfill, there had only been between 150 and 500 complaints. The Supreme Court nevertheless held that the plaintiff had shown a sufficient basis in fact to satisfy the commonality requirement.
[44] The Supreme Court of Canada has recently looked at the question of common issues in three cases, Hollick, Rumley and Western Canadian Shopping Centres Inc. v. Bennett Jones Verchere (2001), 2001 SCC 46, 201 D.L.R. (4th) 385 under three different statutory regimes. Those cases suggest a number of principles that the court must apply in determining whether there are one or more common issues. It is not necessary that the common issues predominate over individual issues or that the resolution of the common issues be determinative of each class member’s claim (Western Canadian at para. 39). The underlying question is a practical one based on issues of fairness and efficiency in the sense that allowing the action to proceed as a class proceeding “will avoid duplication of fact-finding or legal analysis” (Western Canadian at para. 39). While not stated exactly in those terms it seems to me that this is similar to Campbell J.’s statement in Anderson v. Wilson (1998), 1998 18878 (ON SC), 37 O.R. (3d) 235 (Div. Ct.) at 243 that the “common issues need only be issues of fact or law that move the litigation forward”.
[45] The Supreme Court decisions do, however, show that the court is required to make a realistic appraisal of the place of the proposed common issue in the litigation. McLachlin C.J.C. expressed this requirement in different ways:
In Rumley at para. 29:
It would not serve the ends of either fairness or efficiency to certify an action on the basis of issues that are common only when stated in the most general terms. Inevitably such an action would ultimately break down into individual proceedings. That the suit had initially been certified as a class action could only make the proceeding less fair and less efficient.
And in Western Canadian at para. 39 (adopted in Hollick at para. 18):
[T]he class members’ claims must share a substantial common ingredient to justify a class action. Determining whether the common issues justify a class action may require the court to examine the significance of the common issues in relation to individual issues.
[46] In my view, the problem with the appellant’s application is that any common issue, however phrased, cannot meet the requirements that the issue be necessary to the resolution of each class member’s claim and a substantial ingredient of each of the class members’ claims. A proposed common issue that does not meet these requirements is not a common issue for the purposes of s. 5 of the CPA.
[47] In this case, establishing that Prudential was negligent in any of the ways suggested by the appellant would not represent a substantial ingredient in each of the class members’ claims. It would not, to use Campbell J.’s phrase in Anderson, “move the litigation forward”. As the motions judge pointed out, since Prudential had no direct dealings with any of the class members at the time the policies were sold, the class members would still at least have to show that the agents with whom they dealt made representations about premium offset, that those representations constituted negligent misrepresentations about the premium offset feature, and that the prospective policyholder reasonably relied upon the representation. As was stated in Rumley at para. 29, “Inevitably such an action would ultimately break down into individual proceedings.”
[48] I agree with the motion judge’s conclusion on this issue:
While the theories of liability can be phrased commonly the actual determination of liability for each class member can only be made upon an examination of the unique circumstances with respect to each class member’s purchase of a policy (at para. 39).
[49] The appellant points out that in finding that there were no common issues, the motions judge relied upon the decision of the Divisional Court in Bre-X, which decision was subsequently overturned by this court. In Bre-X, the motions judge identified some 15 common issues relating to the torts of conspiracy and fraudulent misrepresentation and breach of the Competition Act. He held that a class proceeding was the preferable procedure for the resolution of those common issues. However, he refused to certify the class action in negligent misrepresentation and the Divisional Court upheld this result. This court held that given the definitions of fraudulent and negligent misrepresentation, there was no logical or principled basis for treating them differently for the purposes of certification. Since the parties accepted that the action based on fraudulent misrepresentation was properly certified there was no reason not to certify the negligent misrepresentation claim. As MacPherson J.A., speaking for the court, said at para. 42:
I could understand an order certifying, or refusing to certify, both claims. I do not, however, understand why opposite orders were considered appropriate for the two claims.
[50] He went on to point out that several of the common issues related directly to the negligent misrepresentation claim and that the two core issues in the litigation—whether there was gold in the Busang mine and, if not, what was the various defendants’ knowledge of that fact—were common to the claims of misrepresentation whether made fraudulently or negligently. Moreover, Bre-X is factually a different case. The foundations of the plaintiffs’ case are the public statements made by the Bre-X insiders to sell a single product, Bre-X shares, and the insiders’ knowledge that those claims were untrue. This case, on the other hand, will always come down to individual representations made by hundreds of different agents selling a variety of whole life insurance products in widely different circumstances. It has not been suggested that the premium offset feature is per se illegal or misleading. Only if the feature is not properly used does it become misleading.
(4) Preferable procedure
[51] Even if the appellant were able to get over the common issue problem, I agree with the motions judge that the class action would not be the preferable means of resolving the common issues. The question of preferability should be examined “through the lens of the three principal advantages of class actions – judicial economy, access to justice, and behaviour modification” (Hollick at para. 27).
[52] Many of the comments made by the court in Hollick are applicable to this case. Although class actions will be allowable even where there are substantial individual issues, preferability “must take into account the importance of the common issues in relation to the claims as a whole” (Hollick at para. 30). Resolution of the proposed common issues would, in my view, have almost no impact on the claims for the reasons set forth above. In terms of judicial economy, as was said in Hollick at para. 32 “any common issue here is negligible in relation to the individual issues”. Thus, “[o]nce the common issue is seen in the context of the entire claim, it becomes difficult to say that the resolution of the common issue will significantly advance the action”.
[53] It seems to me that the comments of Winkler J. in Mouhteros v. DeVry Canada Inc.(1998), 1998 14686 (ON SC), 41 O.R. (3d) 63 (Gen. Div.) at 73 apply to this case: “[C]ertification in this case will result in a multitude of individual trials, which will completely overwhelm any advantage to be derived from a trial of the few common issues”.
[54] I am not persuaded that the appellant has shown that allowing a class action would serve the interests of access to justice. In this respect, the fact that Clarica has established an ADR programme to deal with policyholders’ complaints about the premium offset is a relevant, although probably minor, consideration. See Hollick at paragraphs 33-5. More importantly, it seems to me that since resolution of the common issue would play such a minimal role in resolution of the individual claims, the potential members of the class would be faced with the same costs to litigate their claim as if they were bringing the claims as individuals and not members of the class.
[55] I acknowledge that the class action could serve the interests of behaviour modification by exposing the vanishing premium/premium offset sales technique to public scrutiny through a class action, assuming that a court were to find that the technique amounted to negligent misrepresentation. However, the other considerations relating to judicial economy and access to justice so far outweigh this consideration that a class action should not be considered a preferable procedure.
[56] It is apparent that the appellant has reformulated the common issue in an attempt to bring his case within the holding of the Supreme Court of Canada in Rumley. The Rumley plaintiffs brought an action against the government of British Columbia for compensatory and punitive damages based upon abuse at a residential school for children with disabilities. The Supreme Court found that the commonality and preferability requirements under the British Columbia Class Proceedings Act, R.S.B.C. 1996, c. 50 were met. Fundamental issues in that case were the duty of care and whether the government’s conduct fell below an acceptable standard. Resolving those issues was necessary to the resolution of each class member’s claim. As Nordheimer J. observed in Gariepy v. Shell Oil Co., [2002] O.J. No. 2766 (S.C.J.) at para. 73:
In Rumley, the determination of systemic negligence would have left the members of the class only with the requirement of establishing the fact of the abuse and the injuries that flowed from it. In that sense, the members of the class in Rumley would be left, in essence, with only having to prove their damages.
[57] As I have explained, in this case resolution of the proposed common issue in this case would contribute little, if anything, to the resolution of each class member’s claim. McLachlin C.J.C. held in Rumley at para. 30 that plaintiffs are entitled to restrict the grounds of negligence they wish to advance to make the case more amenable to class proceedings. Thus, it was open to the appellant in this case to attempt to establish liability based on the company’s systemic negligence rather than vicarious liability for the actions of the agents. However, that does not advance the appellant’s position in this case since whether the company is liable on a theory of systemic negligence or vicarious liability, the class members would still have to prove the elements of negligent misrepresentation as set out in Cognos.
[58] It may be that a common issue could be framed to assist in litigation of the first element from Cognos, “a duty of care based on a ‘special relationship’ between the representor and the representee”. In other words, a common issue might be framed to address the question of whether Prudential owed a duty of care to the individual policyholders. However, given that life insurance policies are sold by individual agents having regard to the particular circumstances of the client, resolving that issue would not sufficiently advance the policyholders’ cases to make class proceeding the preferable procedure. The centrality of the relationship between the agent and the client dictates that there would have to be an individual inquiry as to whether the premium offset representation was made, how it was made and whether it had any impact in the particular case.
[59] In Hollick at para. 19 the court found that “some aspect of the issue of liability is common within the meaning of s. 5(1)(c)” since “[f]or any putative class member to prevail individually, he or she would have to show, among other things, that the respondent emitted pollutants into the air”. The court went on to find that there was a rational connection between the class as defined and the asserted common issues. I have some concern that this element can be satisfied in this case since the class as defined embraces all persons who purchased participating whole life policies issued by Prudential during the specified time period. The difficulty with that definition is that it embraces many policyholders who would not have purchased policies with the premium offset option. Nevertheless, I will put that concern to one side since, as I have pointed out, McLachlin C.J.C. held in Hollick at para. 21 that courts can allow certification on condition that the definition of the class is amended.
[60] In Hollick, the Supreme Court refused to certify the class action because it would not be the preferable procedure. The reasoning of the court in Hollick at para. 32, concerning judicial economy, applies in this case:
I am not persuaded that the class action would be the preferable means of resolving the class members' claims. Turning first to the issue of judicial economy, I note that any common issue here is negligible in relation to the individual issues. While each of the class members must, in order to recover, establish that the Keele Valley landfill emitted physical or noise pollution, there is no reason to think that any pollution was distributed evenly across the geographical area or time period specified in the class definition. On the contrary, it is likely that some areas were affected more seriously than others, and that some areas were affected at one time while other areas were affected at other times. As the Divisional Court noted: "[E]ven if one considers only the 150 persons who made complaints -- those complaints relate to different dates and different locations spread out over seven years and 16 square miles" (p. 480). Some class members are close to the site, some are further away. Some class members are close to other possible sources of pollution. Once the common issue is seen in the context of the entire claim, it becomes difficult to say that the resolution of the common issue will significantly advance the action [emphasis added].
[61] My earlier discussion concerning access to justice and behaviour modification applies to this theory as well.
DISPOSITION
[62] Accordingly, I would dismiss the appeal. The respondents shall have ten days from release of these reasons to provide the Senior Legal Officer with submissions on costs and their bill of costs. The appellant may file his submissions as to costs within seven days after receipt of the respondents’ submissions. The respondents may respond within 7 days thereafter.
Signed: “M. Rosenberg J.A.”
“I agree R.Roy McMurtry C.J.O.”
“I agree M. A. Catzman J.A.”
RELEASED: APRIL 8, 2003