Allstate Insurance v. ING Insurance et al., 2015 ONSC 4020
CITATION: Allstate Insurance v. ING Insurance et al. 2015 ONSC 4020 COURT FILE NO.: CV-14-504938 DATE: 20150623
SUPERIOR COURT OF JUSTICE - ONTARIO
RE: ALLSTATE INSURANCE COMPANY OF CANADA Appellant
-AND-
ING INSURANCE COMPANY OF CANADA AND AVIVA CANADA INC. Respondents
BEFORE: F.L. Myers J.
COUNSEL: Eric Grossman and Michael Warfe the Appellant David Murray for the respondent
HEARD: June 18, 2015
ENDORSEMENT
[1] Allstate appeals from the arbitral award delivered by Arbitrator Vance Cooper dated May 1, 2014. The Arbitrator decided that under s.268(2) of the Insurance Act, R.S.O. 1990, c.I.8, Allstate was the insurer required to pay the claim for accident benefits made by an insured whom I will identify as “R” for the purposes of this ruling. The finding that led to this holding was that R was “principally dependent” upon her mother and her mother’s spouse for financial support and therefore, as a result of the application of s.2(6) of Statutory Accident Benefits Schedule as incorporated into s.268(5) of the Insurance Act, s.268(2)1 of the Act applies and requires the mother’s insurer, Allstate, to pay R’s accident benefits.
[2] In my view, the Arbitrator made no reviewable error in his decision. The appeal is therefore dismissed.
[3] One point deserves note at the outset. There are some mathematical calculations required to consider aspects of the applicable legal test. The case law recognizes that if a person has the means to provide for more than 50% of her own needs, then she cannot be dependent. Conversely, where one person provides more than 50% of the needs of another, then the other is financially dependent upon the provider. See The Dominion of Canada General Insurance Company v. Intact Insurance Company, 2015 ONSC 3689 at para. 10. The Arbitrator found that even using inputs favourable to Allstate, the relevant mathematical calculation came out against it – but just barely. Therefore, on the appeal, counsel for Allstate made a number of points which resolved to arguing, that had the Arbitrator changed just one input number, even by a very small amount, the outcome would have been different. This approach, while compelling to read and hear, is based upon two false premises.
[4] First, it assumes that the mathematical result necessarily and solely determines the outcome. In my view, the math is just a part of the test that has arisen out of the seminal decision of Miller v. Safeco, 1985 2022 (ON CA), 50 O.R. (2d) 797 (C.A.). I agree with the insightful comments of Corbett J. in State Farm v. Bunyan, 2013 ONSC 6670, at paras. 19 to 22 to the effect that while the math is an important factor it is not the only factor. The legal issue is whether R was principally dependent on her mother and her mother’s spouse. In Miller, the Court of Appeal approved four factors to consider dependency. Even those four are not necessarily the exclusive considerations. A change in the math from 50.001% dependency to 49.999% dependency may or may not overcome other aspects of the factual dependency between the relevant parties. All of the accountants before the Arbitrator agreed that the math that they were performing was artificial. I would say highly artificial and necessarily inaccurate is a better description. A change of $8, while perhaps crossing a magical mathematical line, does not alter the “big picture” on the facts in the context of this specific case as found by the Arbitrator. See Security National v. The Personal, April 1, 2011, Arbitrator Bialkowski, at page 9, 4^th^ para.
[5] The second faulty premise in the appellant’s argument on math was the notion that this court was entitled to look at each individual mathematical input into the equations. The equations accepted by the Arbitrator came from evidence given by experts. The Arbitrator accepted the evidence of expert witness Olsen on R’s means and expert witness Smith on R’s needs. It is the ratio of means to needs that is the calculation in issue. To alter either of the input numbers (or the assumptions and inputs that lay beneath those numbers) would be to undermine the Arbitrator’s credibility findings and findings of fact. Although the arbitration agreement in this case allows appeals on issue of fact, that does not, in my view, alter the applicable standard of review. Perell J. reiterated the applicable standard just this month in The Dominion of Canada supra. A deferential reasonableness standard applies to mixed fact and law. I am prepared to consider the possibility that pure issues of fact might also be subject to a reasonableness standard although the respondent argues that a “no evidence” standard should apply to findings of fact. In either case, the standard is deferential and cannot be met in this appeal.
Earning Capacity
[6] The appellant rightly argues that in assessing R’s means – her ability to contribute to her own support - the Arbitrator was required to consider not just what R actually earned, but her earning capacity. The parties agreed that the Arbitrator correctly identified that his assessment of R’s means and needs required him to choose a period of time “which fairly reflects the status of the parties at the time of the accident.” See p. 27, 3^rd^ para. Hence the Arbitrator looked at the one year period preceding the accident to assess R’s means and needs. The Arbitrator also rightly noted that each case is fact specific. See page 28, 5^th^ para. The Arbitrator was alive to the point that a claimant’s means are not necessarily limited to just her earnings. In the 1st para. on p. 32 of his decision, he cited Perell J. in Gore Mutual v. Co-operators (2008), 2008 46914 (ON SC), 93 O.R. (3d) 234, at paras. 20-21 describing the many inputs to determine an individual’s earning capacity. On page 34, the Arbitrator cited Arbitrator Samis, in St. Paul Travellers v. York Fire & Casualty Insurance (11 August 2011) quite expressly distinguishing an individual’s earnings from her capacity to support herself. In the end, on page 37, 3^rd^ para., the Arbitrator accepted the highest amount put forward by any of the three accounting experts as his assessment of R’s “earnings and means” to support herself of $10,2385.18. This is the amount that was calculated by Allstate’s expert Ms Olsen.
[7] As the party trying to prove that R was self-sufficient (and hence not principally dependent on her mother) it was in Allstate’s interest before the Arbitrator to maximize her means. The Arbitrator accepted Allstate’s expert’s calculation despite having some concerns that it might be a little too high. Implicit in the Arbitrator’s finding is that he rejected Mr. Grossman’s argument that despite his own expert’s evidence, the Arbitrator should increase his assessment of R’s means because there was no evidence of any reason why she was not working more. Allstate did not lead any evidence of jobs that were available to R by her training and education in her location over the relevant year. There was some evidence that R’s employer would not give her further hours. In any event, in light of the Arbitrator’s clear understanding of the need to fairly reflect the status of the parties at the time of the accident and his recitation of the distinction between earnings and capacity (and his repetition of that distinction at the top of page 38) the finding that her “earnings and means” was $10,385.18 as testified by Ms Olsen, implicitly rejected Mr. Grossman’s argument that her capacity to earn was even higher than his own expert said.
[8] The appellant argues that the Arbitrator failed to include in the calculation of R’s ability to support herself the value of housekeeping services and chores that she did for the family. Had she been paid for this work, her earnings would have risen and, as noted above, just a few dollars one way or the other could change the outcome of the calculation. However, all three of the accountant expert witnesses, including Allstate’s expert, agreed that there should be no implication of value for these services as they are offset by other benefits that R received at home for which she did not have to pay. They all agreed that the services “are a wash.” Before the Arbitrator, Mr. Grossman conceded it was a wash. Now that sees that a small move could have tipped scales, he wants the court to consider the issue on appeal.
[9] In my view, it is not open to a party to re-consider on appeal a concession made before the Arbitrator. The appellant provided no law suggesting otherwise. An adjudicator does not err in failing to rule on matters that counsel agree are not in issue. Moreover, as noted above, for me to change the value of R’s means by an amount for services would be tantamount to rejecting the evidence of Allstate’s expert whose evidence the Arbitrator accepted. There is no basis to do so. Even if errors of fact were readily reviewable on appeal, there was no error made by the Arbitrator on this issue.
R’s Needs
[10] The Arbitrator identified that the case law provides that there are two ways to calculate a person’s financial needs. One can look at the person and try to calculate what she spends and others spend to provide for her basic necessities or one can look at government statistics for a given location and determine what the poverty rate is or use another acceptable proxy for the cost of self-sufficiency in that location.
[11] In this case, the Arbitrator made a clear choice to adopt an objective statistical approach to the assessment of needs. Prior case law has not been definitive as to whether one approach is necessarily better than the other. However, in Miller itself, the judge at first instance relied solely upon objective government statistics to determine the claimant’s needs. The Court of Appeal dismissed the appeal without comment on this issue. The Court of Appeal expressly approved four of the five factors that the lower court judge had used to assess the claim of dependency. Allstate argues that the factor rejected by the Court of Appeal, “the general standard of living within the family unit” amounts to an express rejection of the use of statistics concerning the general standard of living as an approach to the assessment of individual’s needs. Rather, the appellant argues that in all cases it is necessary to obtain evidence as to the particular facts concerning the individual, and the spending by and for the individual on housing, food, transportation, necessities of life, and other basic costs required for self-sufficiency. Were this so, the Court of Appeal would have had to overrule the decision of the lower court judge who did not engage in such an individualized assessment. As the Court of Appeal only rejected one factor, it is incomprehensible that it would not have commented upon the trial judge having wrongly used statistics to determine needs if that is what the rejection of that factor was meant to cover.
[12] It is apparent however, that the rejected fifth factor does not preclude the determination of need by statistics concerning the general standard of living. The Court of Appeal rejected the use of evidence concerning the general standard of living “within the family unit.” Subsequent case law has been clear that the assessment of needs concerns basic self-sufficiency. The fact that a person may be in a wealthy family and have an extravagant lifestyle is not taken into account in the dependency analysis. Similarly, the willingness of an individual to live in penury would not allow for a reduction in assessing her needs for self-sufficiency. To ignore the words “within the family unit” used by the Court of Appeal leads to an interpretation that is inconsistent with the outcome of Miller. I do not accept the appellant’s argument.
[13] Moreover, I wholly agree with the outcome reached by the Arbitrator. As this case so amply demonstrates, the assessment of needs based on approximations of household expenses is always going to be both inaccurate and expensive. It is intrusive and results in extended hearings on expert evidence in which the experts themselves concede that their calculations are inaccurate. There is no accurate way to measure with hindsight an individual’s hypothetical needs sometime in past. Considering just one example, there was a debate among the experts as to how to deal with the cost of housing for R. She lived at home with her mother, stepfather and two siblings. At times, the stepfather’s two children stayed over. Assuming that parents kept absolutely accurate records of their household expenditures, how does one convert a gross spending figure to an individualized cost for one member of the household? Should the total amount be divided by five or seven? Should an assessment be made as to whether individual members of the household consume more than others? If an average cost is taken, what does it mean? For R to be self-sufficient, she would require and have to pay for accommodation. The fact that her stepfather and mother pay $X per person in the household has no bearing on the cost of R obtaining self-sufficiency. If one neighbouring household has a mortgage and the other neighbour does not have a mortgage, all other things being equal, the parents in the latter household will pay less per person to maintain their family. But any individual family member’s needs in order to move out and be self-sufficient will not change. Looking at average cost per person therefore looks at the perspective of the payor (i.e. the parent’s cost). It is not necessarily an assessment of an child’s needs. Yet, the proper way to divide or allocate estimated gross household spending so as to reflect a figure for R was a topic of great debate at the hearing (and in many such hearings). It is a proxy but not necessarily a particularly meaningful one. It is certainly no better than looking at government statistics to determine the cost of housing in a locale.
[14] Using government statistics also has the benefit of simplicity, low-cost, proportionality, and it was accepted in Miller. While Allstate may be unhappy bearing the substantial accident benefits paid in this catastrophic personal injury case, I dare say that over time, it (and all insurers) will see a substantial savings in the cost of lawyers and accountants for these types of hearings by using an objective statistical approach to the analysis of needs instead of hiring experts to create hypothetical models of a claimant’s needs based on estimated and guesstimated historical spending and meaningless averages.
[15] Therefore the Arbitrator was correct both in law and policy on this issue.
[16] Allstate argues however that the Arbitrator erred in his choice of the statistic that he found reflected R’s needs. In particular, Allstate argues that the Arbitrator should not have used the cost of an individual household but should have assumed that as a teenager embarking upon independence, R would have shared accommodation with one or more roommates. The per person cost of shared accommodations is, of course, substantially less than individual costs. However, the issue is the ability and cost of R being self-sufficient. To assume a roommate means that essentially the roommate is contributing to R’s benefit by accepting shared accommodation with her so as to bring R’s cost down. Rather than treating R as being self-sufficient, adding a hypothetical roommate is just substituting a new provider of support instead of the parents. Moreover, once again, Allstate is asking the court to substitute its view on a finding of fact (the cost of self-sufficiency for R) made by the Arbitrator who accepted the evidence of expert witness Smith on this issue.
[17] In any event, recognizing that Allstate may take issue with his findings on statistics, the Arbitrator also made findings on R’s needs based upon the individualized analyses of costs and estimated costs conducted the various expert witnesses. At the top of page 44, he found R’s needs to be $21,988. This amount is higher than the value ascribed by the government statistics. Accordingly, even if the Arbitrator was wrong in using statistics, the calculation of financial dependence using the individualized evidence would have resulted in a finding that R was even less able to meet her own needs. That is, even if Allstate were to succeed on this issue, R would still be found to be dependent upon her mother and stepfather and therefore Allstate would be required to respond under the mother’s policy of insurance.
Did R’s mother and her spouse contribute more than R?
[18] The case law is clear that not only must a dependent be unable to provide for 50% of her own needs, but another person must provide more for the dependent than the dependent provides for herself. That is, if several people each provide a small amount to assist the claimant, it is unlikely that she will be found to be dependent on any of the providers. The Arbitrator understood this to be the rule and made specific reference to it in the first paragraph on page 37 as follows:
If her needs are more than twice her means and if that shortfall is comprised of financial support from her mother… and her stepfather… Then [R] is principally dependent for financial support on her mother and stepfather… [Emphasis added]
[19] Allstate argues that although the Arbitrator stated the right test, he failed to apply it. There was some evidence before the Arbitrator that R’s father and grandmother each provided small amounts of support in money or money’s worth to R. The grandmother bought an occasional meal. The father provided R with $20 “here and there.”[^1] Allstate argues that because the numbers are so close, attributing just one meal to the grandmother or one $20 payment to the father would be sufficient to reduce the amount of support attributable to the mother and her spouse to just below the amount of R’s means to provide for herself. Allstate says that the Arbitrator erred in law when he failed to recognize that small amounts of support were provided for R by her father and grandmother so that the shortfall in her ability to meet her needs is not comprised just of financial support from the mother and her spouse. Once the inputs of the father and grandmother are excluded from the calculation, the amount left attributable to the mother and stepfather is less than the 50% cutoff.
[20] Among other things, this argument demonstrates the artificiality of the mathematical process. At the hearing before the Arbitrator, the respondent argued repeatedly for recognition of a $12 item. It turns out that that miniscule factual issue might have mattered. Yet so many of the household expenses were estimated, averaged, or imputed that a suggestion that the entire outcome may turn on a $12 item or a $20 gift here and there, is no sensible.
[21] On the question of the parents’ relative contributions, the following exchange occurred between the Arbitrator and counsel for Allstate:
The Arbitrator: But just to stop you there for a second. And I guess this is the tricky part of this case is: isn’t the test not whether the claimant provides for 50 percent plus a penny of their own needs, but rather, are they principally dependent on one person for 50 percent plus one penny of their needs… In the case that I have to decide it is been agreed that [R] was looking to the joint support, if you will. I’m not saying more or less than 50 because I have to decide that, but she was looking to her mother and stepfather for financial support.
Mr. Grossman: Yes…
The Arbitrator: You say less than 50 percent and the respondent insurers say for more than 50 percent.
Mr. Grossman: Right. And I guess in response to your point, luckily, it’s not your problem today. If [R] had less than 50 percent of the resources to provide for her needs you would likely in other cases have to look at where the needs were being met additional to her own resources. Here it’s binary, it’s one or the other.
[22] That is, the parties and the Arbitrator were very alive to the issue of whether the support provided by the mother and her spouse exceeded 50 percent of R’s needs. Mr. Grossman was express in recognizing that in other cases, once it is determined that a claimant does not provide for 50 percent of her own needs, there is a remaining issue as to whether any of the providers of support exceed the 50 percent mark. But, in this case, it was clear that the parties determined that the contributions of others were not an issue. They were trivial and the timing of payment was uncertain. Again, Mr. Grossman was express on the point by noting the binary nature of the Arbitrator’s decision in this case. Once again, although it now appears that very small differences in numbers could have made a difference below (subject to my discussion of faulty premises at the outset), an arbitrator does not err for the purpose of appeal when he or she resolves the issues as put forward by the appellant’s counsel and accepts counsel’s strategic determinations of which issues to raise and which issues not to raise. It is frequently the case in litigation that there can be a number of issues that never get pressed. Counsel have any number of good, valid strategic reasons for determining to refrain from raising issues. I was cited no law in which a court allowed an appeal on the basis that an arbitrator accepted a strategic concession by the appellant’s counsel.
[23] In my view the Arbitrator was correct in accepting for the purpose of the case argued before him that the decision was binary in the sense that the mother and her spouse were the providers of whatever portion of R’s needs that R did not provide for herself.
Summary
[24] I do not agree with Allstate’s submission that the Arbitrator set out to change the law and in doing so ignored law that was binding upon him. In my view he clarified competing strands of case law and did so cogently, logically, and practically.
[25] The Arbitrator also made clear that there was room to reduce the quantum he found to best represent R’s means and to increase the amount he found to best represent her needs. Once he was below the magic 50 percent line, there was no need for any greater analysis or specificity as the test was met or failed (as the case may be) once the bright line was crossed. It is perhaps unfortunate that the numbers use produced a result that was so close to the line although it highlights the artificiality of the mathematical piece of the process. Nevertheless, it is clear that on reading the Arbitrator’s decision, he never lost sight of his ultimate goal being the determination of whether R was “principally dependent” on the insured mother and her spouse.
[26] Accordingly the appeal is dismissed. The parties agreed that costs fixed at $15,000 should follow the event. I find the sum to be very reasonable. Therefore, the appellant shall pay the respondent $15,000 all-inclusive for costs.
[27] I wish to express my appreciation to counsel for both parties for their excellent written and oral submissions.
F.L. Myers J.
Date: June 23, 2015
[^1]: There was some evidence before the Arbitrator that R’s father introduced her to an accountant to have her taxes done. However, there was no evidence as to who paid the accountant or how much. Allstate asks me to infer that the father paid. However this is not a trial de novo. Absent evidence, argument, and a primary finding below, I do not see how it is appropriate for me to consider drawing such an inference on appeal.