Zurich Insurance Co. v. Ontario (Human Rights Comm.) (No. 2)

1985-08-02

Ontario Board of Inquiry

CHRR Doc. 85-069

Michael G. Bates Complainant

v.

Zurich Insurance Company of Canada Respondent

Date of Decision: August 2, 1985

Place: Toronto, Ontario

Before: Frederick H. Zemans

Appearances by: Janet Minor, T. H. Wickett, Counsel for the Ontario Human Rights Commission and Michael G. Bates

Neil Finkelstein and J.F. Howard, Counsel for Zurich Insurance Company of Canada

AGE DISCRIMINATION — FAMILY STATUS — SEX DISCRIMINATION — PUBLIC SERVICES AND FACILITIES — automobile insurance contract discriminates against young, single males — INSURANCE — automobile insurance contract employs discriminatory classification system — BUSINESS NECESSITY — insurance rates discriminatory for economic reasons

HUMAN RIGHTS — right to contract on equal terms — EXEMPTIONS — gender, marital status — DISCRIMINATION — unacceptable risk to insurance company as reasonable cause — EVIDENCE — availability and admissibility of statistical evidence — BONA FIDE OCCUPATIONAL QUALIFICATION — Etobicoke test

Summary: The Board of Inquiry finds that Zurich Insurance Company of Canada discriminated against Michael G. Bates, an under-25, unmarried male driver, because it charged him a higher premium than it charges over-25 single males, or married males, or females.

The Board of Inquiry finds that Zurich has offered no evidence that unmarried males under 25 constitute a higher risk. On the contrary, evidence was presented that the factors of sex, marital status, and age are probably not causally related to the differentials in the claims experience demonstrated for the various groupings.

In the over-25 category, Zurich relies on non-discriminatory causally connected factors, such as driving record, driving experience, and vehicle use to establish rate differentials. The Board finds that there is no evidence offered that these same factors cannot be successfully used with under-25 insured persons.

The defence available to Zurich under Section 21 of the Ontario Human Rights Code is to show that the rates are reasonable and bona fide in the circumstances.

The Board applies the Borough of Etobicoke test of a bona fide occupational qualification. It finds that there is no statistical or scientific base for the classifications based on age, sex, and marital status and there is no statistical base for the view that administration costs are higher for the under-25 age group.

In the circumstances, Zurich must show that there are no non-discriminatory group factors which can be used to determine premiums for insured drivers. Neither the ease of establishing insurance rates, nor industry tradition justifies discrimination. The Board finds that Zurich has failed to establish that no non-discriminatory rate classifications are available for the under-25 group

In addition, Zurich has failed to establish that the essence of its business will be undermined if it cannot use rate classifications based on sex, age, and marital status.

The Board finds that Zurich Insurance has discriminated against Michael G. Bates and orders the hearing resumed to address the issue of an appropriate remedy.

Preliminary Matters

1On April 30th, 1984, I was appointed to serve as a Board of Inquiry pursuant to the Ontario Human Rights Code, 1981, (the "Code") by the Minister of Labour, the Honourable Russell H. Ramsay, to hear and decide the complaint made by Michael G. Bates (the "Complainant") against Zurich Insurance Company of Canada, ("Zurich") 188 University Avenue, Toronto, Ontario. This complaint alleges that the Complainant has been denied the right to contract on equal terms without discrimination contrary to sections 3 and 8 of the Code. The complaint also alleges that the Complainant's right to equal treatment in services, goods and facilities has been infringed contrary to sections 1 and 8 of the Code.

2I had previously been appointed on January 23rd, 1984, by the Minister of Labour to hear and decide a complaint made by Brian B. Hope against The Royal Insurance Company of Canada, 10 Wellington Street, East, Toronto, Ontario. This complaint alleged that Mr. Hope had been denied the right to contract on equal terms without discrimination contrary to sections 3 and 8 of the Code.

3In essence, the two complaints concern allegations that young, single male drivers must pay automobile insurance rates which exceed those of young, single female drivers, as well as those charged to young married male drivers and that this discrepancy contravenes the provisions of the Code.

4A single hearing into both these complaints was commenced on May 22, 1984 with counsel for both respondents present. At the May 22nd hearing, all parties agreed to have both complaints heard together, and the Respondent, Royal Insurance Company of Canada, brought an application to have the complaint against it dismissed. I dismissed this preliminary application in my interim decision of June 22, 1984. When we reconvened to hear the evidence on September 18th, 1984, counsel for the Ontario Human Rights Commission indicated that my preliminary decision against the Royal Insurance Company of Canada was before the courts and recommended that this complaint be adjourned sine die and that the complaint against Zurich proceed. This recommendation was agreed to by counsel for both Respondent insurance companies and I ordered that the complaint of Brian Hope against The Royal Insurance Co. and the complaint of Michael Bates against Zurich be heard separately. The Hope hearing was adjoured sine die and I proceeded to hear the evidence in the matter of Michael G. Bates and Zurich.

5At the outset of these proceedings, counsel for the Complainant asked to amend Mr. Bates' complaint (Exhibit one). This amendment was made on consent and added the further allegation that Mr. Bates was discriminated against on the basis of "marital status" contrary to the Code.

6Counsel for Zurich agreed that the complaint filed by Mr. Bates and the response to the complaint by Zurich did establish a prima facie case to which Zurich was prepared to respond at this hearing without the Complainant first calling evidence.

7Counsel also indicated at the outset of the hearing that there had been an agreement that the hearing would be divided into two parts. The first part would involve the hearing of evidence dealing with the question of whether or not there had been a breach of the Code, and the second part would hear further evidence, if necessary, with respect to the appropriate remedy in this matter.

8The Complainant alleges that he was discriminated against on or about the 2nd day of March, 1983 with respect to his age, sex and marital status. He specifically alleges in his complaint that he is a 20-year-old single male who, on or about February 3rd, 1983, was informed by his insurance brokers, Hunter, Rowell and Company Limited, that his automobile insurance premium would be $1002.00 to provide him with $200,000.00 public liability and property damage coverage by Zurich. The Complainant further alleges that he was not to be provided with collision insurance for the 1976 Chevrolet Camaro that he was driving at all relevant times.

9Mr. Bates' complaint states:

4. It is my understanding that a single female of the age of 20 with the same automobile coverage for a 1976 Chevrolet Camaro would have to pay a premium of approximately $434.00 to the above-named Respondent.

5. It is also my understanding that a 25-year-old, single male with the same automobile coverage for a 1976 Chevrolet Camaro would have to pay a premium of approximately $321.00 to the above-named Respondent.

6. I further understand that a 20-year-old male who is married with the same automobile coverage for the 1976 Chevrolet Camaro would pay a premium of approximately $606.00 to the above-named Respondent.

7. I have no driving accidents nor have I ever received any demerit points. However I was informed by Hunter, Rowell and Company that I would pay higher rates than women, married men of my age, and single men who were 25 years of age or older.

8. I am a 20-year-old, single male and believe that my right to contract on equal terms has been infringed by the above-named Respondent because of my age, sex and marital status contrary to Sections 3 and 8 of the Ontario Human Rights Code, 1981, Statutes of Ontario 1981, Chapter 53.

9. l am a 20-year-old, single male and I believe that my right to equal treatment and services, goods and facilities has been infringed by the above-named Respondent because of my age, sex and marital status, contrary to Sections 1 and 8 of the Ontario Human Rights Code, 1981, Statutes of Ontario 1981, Chapter 53.

10In his opening remarks, Mr. Howard, counsel for Zurich agreed that differentiation or distinction between individual insureds in the circumstances enumerated in Section 21 of the Code does exist in the context of automobile insurance and that the Complainant had established a prima facie case. He stated that Zurich's evidence would seek to support the proposition that the differentiation for single, male drivers under the age of twenty-five was arrived at in good faith and was based on reasonable grounds within the provisions of section 21 of the Code. Mr. Howard further asserted that Zurich would attempt to prove that 20-year-old, single males get reasonable insurance rates when a comparison is made of the loss experience for this age grouping as compared to the premium charged.

The Evidence

11Mr. Lee Marshall Alexander was the first witness to testify for Zurich. Mr. Alexander has been the Vice-President and Chief Actuary of the Insurance Bureau of Canada (the IBC) for approximately five years. Mr. Alexander testified that the current Statistical Plan for Automobile Insurance in Ontario (the "Statistical Plan") is based on a statistical model developed in 1926 by the Canadian Automobile Underwriters' Association's first Casualty Actuary, Mr. C.H. Frederickson, F.C.A.S. In early 1929, an Ontario Royal Commission was appointed with the Honourable Mr. Justice Frank E. Hodgins as Commissioner. His Royal Commission on Automobile Insurance Premium Rates was submitted in 1930 and concluded with six recommendations, the fifth of which stated, at p. 78:

That the loss costs of insurance in Ontario in the future should be established by the combination of the experience of all companies, and that such experience should be developed on the statistical plan prescribed pursuant to Section 69(a) of the present Insurance Act.

Exhibit 4 quotes one paragraph of the conclusion of Mr. Justice Hodgins' report, which gives the 1985 reader some insight into his thinking:

I am deeply impressed with the difficulties encountered in the earlier years of automobile insurance by those who carried it on. It is a comparatively new form of insurance, and has not the stable elements which make life insurance and fire insurance less difficult, while it has a great variety of risks and hazards on different makes and styles of motors. It has not yet reached the point in organization which should have produced lessened expenses, and not very much attention has been paid to this question; nor has it had until now, any trustworthy record of loss experience. It has had to deal with a rate war, and strong competition, increasing even now in its own field. It has not yet overcome the difficulties of fleet-rating or the hazards of collision due to recklessness, youth, want of experience and there are other elements apparent to me which make it difficult to forecast the changes which are inevitable in the business.

12Although he did not discuss the historical development of the current Statistical Plan in his oral testimony, the summary of Mr. Alexander's testimony (Exhibit 4) indicates that upon receiving the Royal Commission report, the Ontario Superintendent of Insurance, R. Leighton Foster, appointed an Advisory Committee to recommend a compulsory standardized statistical plan. During 1930, this Committee recommended the original, all-industry uniform statistical plan. The statistics department of the Canadian Automobile Underwriters' Association was appointed the Automobile Insurance Statistical Agency by the Government of Ontario and, subsequently, by the other provinces as similar provisions were included in their Acts.

13The Statistical Plan that is in effect in the Province of Ontario is approved by the National Statistical Advisory Committee and by the Committee of the Superintendents of Insurance. The statutory requirement for such statistical compilation is found in sections 80(1) and 80(2) of the Ontario Insurance Act, R.S.O. 1980 c. 218 which reads as follows:

80(1) Every licensed insurer that carries on in Ontario the business of automobile insurance, fire insurance, property damage insurance or sprinkler leakage insurance shall prepare and file, when required, with the Superintendent, or with such statistical agency as he may designate, such statistical return of the experience of such business as a Superintendent may require and in such form and manner and according to such system of classification as he may approve.

(2) The Superintendent may require any agency so designated to compile the data so filed in such form as he may approve, and the expense of making the compilation shall be apportioned among the insurers whose data is compiled by such agency by the Superintendent who shall certify in writing the amount due from each insurer and it is payable by the insurer to such agency forthwith.

14Mr. Alexander testified that the 1982 Automobile Insurance Experience, as of December 31, 1982, familiarly called the "Green Book" (Exhibit 6), was compiled by the IBC) under his direction as the Chief Actuary. (Evidence, Volume III, p. 27) Accordingly, it is his responsibility to supervise the collection and publication of annual loss experience figures. The underlying purpose of this collection and publication is:

to define individual groups of insured in such a manner that we have a predictive value from that grouping of insureds. The actual losses that we would expect to incur under each of the cells that we define through the class plan should ideally be equal to the provision for losses that we made in our rates. This must operate in a competitive environment. (Evidence, Volume III, p. 30)

15The statistical returns required by the Province of Ontario have been restricted to automobile insurance. The governmental function of the Statistical Agency was carried out by the Canadian Automobile Underwriters' Association for thirty-eight years until January, 1968, when the Statistics Division was transferred to the IBC) Since that time, the Statistics Division of the IBC has acted as the automobile insurance statistical agency for all provinces, except Quebec which, in January, 1978, established the Groupment des Assureurs Automobiles as its statistical agency and those provinces where a government monopoly provides basic automobile insurance coverage (British Columbia, Manitoba and Saskatchewan).

16Mr. Alexander testified that the entire structure of automobile insurance rate-setting in provinces other than Quebec, British Columbia, Manitoba and Saskatchewan flows from automobile insurers' collective loss experience compiled within the statistical framework. The basic document is the Statistical Plan. Mr. Alexander testified that the process of implementing changes to the Statistical Plan is a lengthy one. It should be noted that the current Statistical Plan is not relevant to this hearing as it did not become effective until January 1st, 1985.

17Mr. Alexander testified at length about the classification coding used in the current Statistical Plan and it is necessary for me to summarize in some detail this evidence as it is fundamental to Zurich's position. The Statistical Plan in use at the time of the complaint consists of three categories: i) individually rated vehicles; ii) fleets and iii) non-private passenger vehicles rated on other than a per vehicle basis. Of the individual classifications set forth under specific cases, the single, male, 20-year-old Complainant would fall under Class 11, while a 20-year-old married male would fall under Class 08 and a 20-year-old single female would fall under Class 18 (Evidence, Volume III, p. 38). In addition to the age, sex and marital status classifications, there is a classification for private passenger automobiles having regard to years of claims free driving. But the most significant classifications for the purpose of this hearing are the "Type of Use" classifications which are schematically illustrated at page 11 of Exhibit 4 as follows:

25 or over no driving to work to work less than 10 miles to work more than 10 miles business use Class 01 Class 02 Class 03 Class 07 ages 23 and 24 Class 13 or 06 ages 21–24 ages 21–24 Under 25 ages 21 and 22 Class 09 Class 19 Class 12 or 06 ages 19 and 20 Class 11 or 06 ages 16–20 ages 16–20 ages 16–18 Class 08 Class 18 Class 10 or 06 unmarried males married males females

This schematic illustration of the rate categories graphically illustrates the issues raised by the Complainant during these proceedings. For those drivers over the age of 24, the categories are determined according to the use and mileage of vehicles, while for those drivers under 25 years of age, factors such as sex, marriage and age are determinative of the categories in which insureds are placed within the Statistical Plan.

18All insurers in Ontario are required to provide information on the basis of the Statistical Plan to the IBC) which, subsequently, collates and compiles the industry's statistics. The annual publication of the statistical experience is the previously mentioned Green Book which was filed as Exhibit 6. Since 1970, the Green Book has been expanded to include a section showing loss ratios based upon actual premiums and losses reported to the IBC) for the previous five policy years. The five year historical profile is employed by the IBC to arrive at a factor to apply to current year losses. (Evidence, Volume III, pp. 54–56) Mr. Alexander testified that the Complainant's Class 11 has a comparative loss cost that is roughly two and one half times higher than the average cost for all classes and all driving records. The loss frequency for all classes and all driving records in 1982 in Canada was 6.46 claims per 100, while the loss frequency for unmarried males ages 19 and 20 in Class 11 was 15.74 claims per 100. The average cost of the claims for all vehicles was $2,728.00 per claim. The average cost of a Class 11 claim was $3,352.00 per claim. (Evidence, Volume III, p. 41) When asked by his counsel if, in his opinion as an actuary, the Class 11 experience would at least support higher charges for liability insurance for Class vehicles, Mr. Alexander indicated that it would. (Evidence, Volume III, p. 71)

19Mr. Alexander testified that premium rating is based on information concerning both the vehicle and the principal operator of the vehicle, but that the fundamental determinant is who is the principal operator. Mr. Alexander explained that the IBC is the statutory agent for the Superintendent of Insurance of Ontario with respect to collection of statistical information and the presentation of the automobile insurance industry's (the "industry's") perspective on various issues. The manner in which the statistics obtained by individual insurance companies from the IBC) are employed is left entirely to the discretion of the individual insurance companies. Mr. Alexander was not aware of any provisions which require insurance companies to use the IBC) statistics; whether the companies utilize these statistics in fixing their rates is an independent decision which will generally be dictated by market forces. Mr. Alexander agreed that the Superintendent of Insurance of Ontario has no authority with respect to fixing of rates, although the Insurance Act requires that rates be filed with the Superintendent prior to their being utilized. (Evidence, Volume III, p. 84) Mr. Alexander also indicated that the IBC) does not collect or compile any other statistics for the industry in Ontario, other than the information contained in the Statistical Plan. (Evidence, Volume III, p. 89)

20Mr. Alexander acknowledged that during the past five years, the Superintendent of Insurance of Ontario has asked the IBC) to study the possibility of reducing the impact of or replacing age, sex and marital status as rating categories for the purchase of automobile insurance. He also acknowledged that the industry has not moved as quickly or as far as the Superintendent has suggested, even though the industry has studied the problem and has introduced changes to the Statistical Plan to allow for further analysis of it. (Evidence, Volume III, pp. 98–100) Mr. Alexander agreed that the industry makes no distinctions with respect to sex, age, or marital status for drivers over-25. Rather, the criteria for rating are determined on the basis of the use of the vehicle. The four categories of use (not driving to work, driving less than 10 miles to work, driving more than 10 miles to work and business use) are relevant only to the risks involved in the classification system. (Evidence, Volume III, pp. 103–104)

21With respect to drivers under 25, Mr. Alexander gave the following testimony:

Q. Is there any component of mileage involved in the definition in any of those classes, Sir?

A. Not directly, no.

Q. Where is the indirect component, Sir?

A. To the extent that it would be imputed, if one knew, or one might surmise that there would be a different mileage exposure to females than males, or ...

Q. When you say it could be imputed and it might be surmised, Sir, are those statistics kept?

A. Not directly, I said.

Q. Well, are they kept indirectly, Sir; the statistics?

A. No.

Q. So it is purely a question of surmise or assumption?

A. Yes. (Evidence, Volume III, p. 110)

22Mr. Alexander indicated that he understood that it was during the 1960s that the insurance categories for drivers under the age of 25 years of age were developed and followed by further refinements with respect to the age and sex distinctions. Mr. Alexander believed that the sex distinction was the most recent, but was not certain whether the marital distinction was, in fact, the most recent. As well, he believed that the current rate structure for under-25 year-olds reached its current form in the mid-1970s and that there had previously been a category of unmarried males between 25 and 29 which has been continued in some jurisdictions, but not in Ontario.

23When asked for the statistical basis for dividing the under-25 drivers in this manner, rather than the same manner as the drivers over the age of 25, Mr. Alexander stated:

A. Such a change would have come from individual company classification efforts. As a result of such changes, it would perhaps spread in practice and become a norm for the industry. This in essence is how many classification changes take place.

Q. But you are not giving us any indication of a statistical basis for it?

A. Some companies would have had to establish such statistics.

Q. Well, again, sir, you are really speculating. You say that is what probably happened?

A. Yes.

Q. You have not obviously looked into the historical origin of that classification?

A. That is correct. (Evidence, Volume III, pp. 108–109)

24Mr. Alexander was unable to say why mileage was not introduced into consideration in the under-25 year-old group, but was professionally satisfied that the categories in the under-25 group realistically reflect risks and, therefore, are proper classifications. He admitted that the IBC) does not "require or feel it is even necessary to have a causal relationship in designing the classifications that are used in rating". In his view: "That is not a necessary ingredient ... The causal relationship is more of a perceived relationship. Statistically it is very difficult, if not impossible, to show true causal relationship in any of our classifications". (Evidence, Volume III, p. 111) According to Mr. Alexander's experience, rate distinctions are developed and used "because they have meaning for the industry and its competitive view of setting rates." (Evidence, Volume III, p. 112) When asked whether a more closely defined causal criterion such as mileage should be incorporated into the rating system, if such a causal connection was proven, Mr. Alexander stated that, if such were the case, the industry would adopt such a separation given the industry's "competitive environment". (Evidence, Volume III, pp. 113–114)

25In the under-25 category, while there is a distinction between married males and unmarried males, there are no tests to support such a distinction based on mileage. Rather, according to Mr. Alexander, the present distinction is based on a "perceived causal relationship" which is used because it is easier for the industry. "I am not looking for a causal relationship ... I am looking for a separation based on factors that are easily interpretable, easily understood, to create these statistical cells". (Evidence, Volume III, p. 121) Mr. Alexander stated that he did not see it as his responsibility as a statistician – and the IBC) does not endeavour – to go beyond arbitrary classification to try to isolate the actual factors which are responsible for different statistical results. Mr. Alexander agreed that the miles driven per year, the number of years of driving experience and the driving record of an individual were factors which are subject to statistical verification and which are more likely to affect the risk of the individual driver or the rest of the class. He also indicated that these items are isolated to the extent that the industry feels they need to be considering the cost of such further isolation. Though he expressed some concern with respect to the introduction of a mileage factor for persons under the age of 25, Mr. Alexander indicated that there had been no studies that would support the proposition that it would be difficult to implement such a criterion due to difficulty in obtaining accurate information. He also agreed that mileage is a criterion currently being used in the over-25 category and that it did not create an insurmountable problem.

26With respect to the differential for married males under-25, Mr. Alexander testified that it was his speculation that married status in the under-25 is not the norm; whereas, one would perhaps expect to find marriage as normative behaviour for drivers over-25. He further stated that there is no information for Ontario dealing with the differences between married and unmarried men over-25 and their respective loss ratios. However, to Mr. Alexander, this distinction is still relevant for the under-25 group because "it has been continued as a class distinction". (Evidence, Volume III, pp. 130–132) Mr. Alexander could not give the Board any information as to why it the question of marital status has any significance, it is only used with respect to males and not to females. Mr. Alexander testified that there have not been any studies which try to ascertain other factors that may be more useful in classifying persons under age 25. "We generally function from the bottom up in terms of companies doing their analysis and erecting their classes in order to maximize competition among companies". (Evidence, Volume III, pp. 130–131) Subsequently, Mr. Alexander did acknowledge that a further separation of rates – along mileage lines – would more accurately reflect the factors which lead to risk and would be more fair and equitable for individual drivers. He nonetheless stated that he was satisfied that the present rate system incorporates a fair and proper classification for those persons over-25 and those under-25.

27Mr. Alexander testified that there were a number of aspects of automobile insurance which were not statistically based. He also testified that he was not aware of any statistical analysis of the loss ratio for multi-vehicle owners; or of any statistical base, outside of the Facility Association¹, for the setting of rates based on driving convictions; or of any statistical base for the "good student" discount used in the industry. (Evidence, Volume III, pp. 143–146) He further testified that the justification for driver training discounts given to young drivers who have taken a driver programme is a "business justification ... and hopefully it can be shown to be a statistical justification". (Evidence, Volume III, p. 152) According to Mr. Alexander, data was being collected with respect to driver training discounts and is expected to be published shortly. He further indicated that there was no statistical basis for the forgiveness practice which some insurance companies use. Mr. Alexander stated that most of these decisions were made by insurance companies on the basis of industry conditions and their perception of the insurance market. (Evidence, Volume III, p. 156–159)

28Mr. Alexander was asked to compare specifically the 1982 information with respect to single, male drivers and married male drivers. He referred to Exhibit 6B which states that married males between the ages of 16 and 20 (Class 08) in 1982 had a claim frequency per hundred cars insured of 16.15 and an average cost per claim of $3,500.00. The ratio of loss expenses to premiums earned for Class 08 in 1982 was $132.00 which means that for every one hundred dollars ($100.00) of premiums earned by the insurance companies, one hundred and thirty-two dollars ($132.00) were paid out per claim. these figures were compared to those for Class 11 drivers in 1982. This category is the Complainant's class and contains only unmarried males ages 19 and 20. Exhibit 6B states that the claim frequency per one hundred cars insured for unmarried, male drivers ages 19 and 20 in 1982 was 15.74 and that the average cost per claim was $3,352.00 with a loss ratio of 72. (Evidence, Volume IV, p. 10–11) This figure indicates that, in 1982, for every hundred dollars of premiums collected, $72.00 was paid out.

29Mr. Alexander pointed out that a comparison of drivers in Class 08 of married males aged-20 and under is more appropriately made with classes 10 and 11, which include all unmarried males from age 16 to 20. Refering to Exhibit 6B, he indicated that the average claim frequency per 100 cars insured countrywide, in an urban setting, was 17.79. The average cost per claim was $3,308.00 and the loss ratio was 79 for insured unmarried males under twenty. Mr. Alexander summarized the comparison of the under-20 married and unmarried drivers for 1982, countrywide – urban, as indicating that the claim frequency was 10 percent higher for unmarried males; that the average cost per claim was slightly lower; and that the loss ratio was very substantially lower. (Evidence, Volume IV, p. 15) When asked whether, in his opinion, it would be fair to charge unmarried males a surcharge or more than the married male for motor vehicle insurance, Mr. Alexander indicated that it probably was and that it was difficult to generalize with respect to frequency in claim costs for only one year. He indicated that more weight should be given to the frequency, rather than to the average cost of claims as lost costs can vary from $500.00 to $1,000,000.00. Mr. Alexander testified that the average loss for class 10 and 11 drivers would be $588.00 and that the average loss incurred under Class 08 would be $566.00 which would indicate that the average loss would be 4 percent higher for classes 10 and 11 than Class 08. Mr. Alexander agreed that a 4 percent differential was not statistically significant. (Evidence, Volume IV, p. 19)

30Mr. Alexander testified that it was not possible to make a comparison between unmarried, male drivers under-25 and unmarried male drivers over-25, nor was it possible to compare unmarried, male drivers of the Complainant's age (20) and unmarried, female drivers of the same age. During his re-examination, Mr. Alexander stated that one of the difficulties in considering the Green Book statistics with respect to younger drivers is the limited volume of data. He indicated that, in many of the categories, the numbers were so small that it was difficult to get accurate information and that this dearth of information is one of the reasons why his organization publishes three year comparisons. Many companies have charged the same premium for Class 10 and Class 11 drivers because of the limited volume of data. (Evidence, Volume IV, pp. 22–24) In conclusion, I found the testimony of L.M. Alexander to be both helpful and credible.

31Mr. J. Scott Bradley also testified for Zurich. He is currently Zurich's Superintendent of Actuarial Services and has been in their employ since 1980. Mr. Bradley trained in mathematics at the University of British Columbia and at McMaster University where he obtained his Ph.D in 1979. His responsibilities with Zurich include rate-making, loss reserve analysis, adequate testing, experience rating analysis, statistical analysis and special mathematical or statistical studies. Mr. Bradley described in detail the actuarial procedures used by Zurich in setting its automobile insurance rates. Zurich uses three primary sources in developing its suggested car insurance rates and considers the previous year's experience of its own company and the industry as a whole. Since becoming a member of the Insurers Advisory Organization (the "IAO") Zurich also considers the IAO statistics. To the extent that the company considers that its own experience is statistically reliable, Zurich relies on its own data.

32According to Mr. Bradley, Zurich's objective is to calculate a projected loss cost that will be used for an average vehicle across Ontario to pay expected claims. This calculation is the basis on which Zurich distributes costs by territories, regions, driver classification, and, finally, by driving record in accordance with the present Statistical Plan. The final calculation is ultimately made by the management of the company and includes commission and expense factors. (Evidence, Volume IV, p. 38)

33The actuarial department prepares a proposed driving premium that is the product of four items: the base rate, the class differential, the driving record differential, and the increased limits factor. I will not lengthen this judgment by describing the mathematical evidence of Mr. Bradley, not because I did not find it informative but because I cannot do justice to its mathematical sophistication. For Mr. Bradley's testimony, refer to:

Evidence of Scott Bradley, Volume IV, especially pp. 37–89, 92–99; Volume V, pp. 72–73;

Exhibits 8 and 8A, Summary of Evidence of Scott Bradley; and Exhibits 16–28, sources of data and documentation used in Zurich rate-setting process

34I have carefully considered Mr. Bradley's evidence and find his testimony to have been presented in a forthright and believable fashion with every intention of being of benefit to this Inquiry.

35Mr. Bradley testified that the process Zurich uses to calculate base rates was developed for fourteen classes, five driving records and twelve territories. The techniques which he described in detail are standard actuarial techniques which are generally accepted by the profession. The only other calculation that Mr. Bradley would undertake after calculating the differentials would be to calculate the average effect of the rate change to provide the company with a 25percent margin of profit. The data that Mr. Bradley and the actuarial department compiles is passed on to the underwriting, sales and marketing departments which then decide when, if, and how much the rate changes would be made.

36With respect to Mr. Bates' complaint, Mr. Bradley testified that Mr. Bates would have been classified by Zurich, as a Class 11 driver, being a 19 or 20-year-old, single male with a driving record 3. Though he had possessed his license for less than three years, he was given a three year accident free driving record because he had successfully completed a driver's education programme for the Ontario Motor League, (Evidence, Volume IV, p. 91) A comparison of Zurich's rates would indicate that, while Mr. Bates in Class 11 with a driving record 3 would be charged an annual premium of $1,002.00, a female, aged 20 with the same driving record 3 would be charged $522.00. A 25 year-old, single male with the same coverage, the same vehicle and the same driving record would be charged $448.00 and a 20-year-old, married male with the same coverage, the same vehicle and the same driving record would be charged $818.00. (Evidence, Volume IV, p. 92)

37Mr. Bradley referred to the information provided by Zurich in response to the Human Rights Commission's questionnaire in these proceedings and which forms Exhibit 2. This information indicates the following:

Class Information – Ontario – BI/PD

1981

1. Age 20

– single male – statistical class 11 claim frequency per 100 cars = 13.48 average claim cost = $5,588.61 – actual loss ratio 125.93%

2. Age 20

– single female – statistical class 18 claim frequency per 100 cars = 11.88 average claim cost = $2,485.15 – actual loss ratio 103.27%

3. Age 25

– single male – statistical class 01 claim frequency per 100 cars = 5.11 average claim cost = $2,006.95 – actual loss ratio 77.6%

4. Age 20

– married male – statistical class 08 claim frequency per 100 cars = 14.89 average claim cost = $1,688.21 – annual loss ratio 64.71%

1982

1. Age 20

– single male – statistical class 11 claims frequency per 100 cars = 13.0 average claim cost = $6,383.00 -actual loss ratio 119.0%

2. Age 20

– single female – statistical class 18 claims frequency per 100 cars = 8.0 average claim cost = $2,816.80 – actual loss ratio 65%

3. Age 25

– single male – statistical class 01 claims frequency per 100 cars = 4.9 average claims cost $2,484.20 – actual loss ratio 75.9%

4. Age 20

– married male – statistical class 08 claims frequency per 100 cars = 10.0 average claims cost $1,931.00 – actual loss ratio 37.3%

38Mr. Bradley emphasized that Exhibit 2 contains the Zurich experience and indicates that the actual loss ratio for Class 11 drivers in 1981 was 125 percent. The break even point for the company is approximately 75 percent of premiums paid. With a loss ratio of close to 126 percent of Class 11 premiums, the difference would have to be paid out of the company's surplus. For Class 01 drivers, the loss ratio was 77.6 percent which is within the company's range of expectations. Similarly, in 1982, the loss ratio for Class 11 was 119 percent while, for Class 18, the loss ratio dropped to 65 percent and for Class 08 to 37.3 percent. Exhibit 2 also indicates the results of the 1981 and 1982 Zurich Green Book and the numbers of policyholders in the book so that it is possible to better understand the specific calculations and, particularly, the limited number of Zurich policyholders in Classes 08, 11 and 18.

ZURICH INSURANCE COMPANY

Policy Year 1981 Results From 1981 Green Book for Ontario

Class	Exposures	Claims	Loss Amount	Premium
1	49,254	2,517	$5,051,485	$6,509,660
2	41,813	2,741	6,062,625	7,286,007
3	7,478	454	1,037,857	1,437,741
4	–0–	–0–	–0–	–0–
6	3,230	143	340,455	619,351
7	3,291	242	437,159	711,555
8	94	14	23,635	36,524
9	2,158	173	266,582	528,103
10	320	86	194,071	222,604
11	1,076	145	824,849	654,981
12	1,247	137	1,136,630	501,854
13	1,418	144	653,089	435,256
18	985	117	290,763	281,547
19	3,661	277	515,277	769,794
Total	116,025	7,190	$16,834,477	$19,994,977

Policy Year 1982 Results from 1982 Green Book for Ontario

Class	Exposures	Claims	Loss Amount	Premium
1	45,673	2,236	$5,554,702	$7,316,897
2	36,000	2,112	7,310,771	7,272,602
3	7,272	404	900,744	1,520,804
4	–0–	–0–	–0–	–0–
6	3,285	160	582,298	662,203
7	3,152	221	520,663	767,943
8	60	6	11,586	31,032
9	1,950	147	1,076,464	538,212
10	248	41	74,905	215,520
11	939	122	525,702	661,563
12	1,547	151	1,222,923	669,480
13	1,662	119	287,270	531,511
18	854	68	450,537	294,943
19	3,509	197	480,172	844,124
Total	106,151	5,984	$18,998,737	$21,326,834

Mr. Bradley stated that, in 1982, Class 11 was the only one of the four classes being analyzed which was for Zurich appreciably above the 75 percent breakeven point. He indicated that one of the possible reasons was that the premiums that had been set and approved by the underwriting department were too low in relation to the claims frequency that developed during that year. Despite the high premiums, they had not been sufficient to pay the losses that were incurred in that particular class of drivers during that particular year.

Q. So you are saying that on the basis of this information that even though the Class 11 premiums were considerably higher, it wasn't high enough?

A. Based on my experience, that is true. (Evidence, Volume IV, p. 98)

39In cross-examination, Mr. Bradley was asked why Zurich assigns the previously mentioned ratio of 60 percent for the current year loss costs and 40 percent for the preceding year loss costs. Mr. Bradley acknowledged that the literature pertaining to the weighting of loss costs indicates that some writers recommend the weighting of loss costs at 60 and 40 percent, while others recommend a weighting of 70 and 30 percent. The 70 and 30 percent weighting system is more commonly used in the United States where statistical plans are based on an "accident year", rather than a policy year. In the American situation, the statistics are more developed at the time of calculation, and, consequently, there are fewer developments. Mr. Bradley stated that it was his opinion that a 70 percent weighting factor was too high a percentage for Ontario, but he left me with the impression that the determination of weighting factors is a matter of individual discretion that can vary from company to company and from actuary to actuary. (Evidence, Volume IV, p. 111)

40Mr. Bradley acknowledged that the figures that he calculates indicate what the company "should" be charging as premiums. The actual premiums as printed in Zurich's rate manual are determined by Zurich's underwriting, sales and marketing departments. As to how differentials are set, Mr. Bradley stated that the following factors are taken into account:

A. Obviously we are going to look at our own experience. Alright? The point in question, if, for example, an existing differential is 2.8, the indicated differential was 2.5, and we had 130% loss ratio on that class, and, you know, the overall loss ratio was, say, 80 %, I would definitely suggest they increase the differential for that cell.

Now, whether they do or not is their decision, but I would suggest that they increase it because obviously we are not charging enough money. (Evidence, Volume IV, p. 124)

41Mr. Bradley explained that, with a sample of 1,084 claims, his figures would be 90 percent certain and his predictions with respect to the number of claims presented would be within a range of plus or minus five percent. If the sample dropped appreciably to, for example, half – the range of predictability would drop correspondingly to approximately 80 percent certainty or remain at 90 percent, but with a 10 percent margin of error. Mr. Bradley acknowledged that the smaller the sample, the less the predictive value of the data obtained.

42Mr. Bradley compared the loss ratio figures for Zurich with those in Exhibit 6B provided by the IBC) on a country-wide basis. The relevant data are as follows:

	IBC (Country Wide)	Zurich (Ontario)
1981
Class 01 (Over 25-Doesn't Drive to Work)	78%	77%
Class 08 (Married Male of 20)	139%	64%
Class 11 (Single Male of 20)	76%	125%
Class 18 (Single Female of 20)	94%	103%

The 1982 data indicated that, in Classes 18 and 01, Zurich's and the IBC's loss experiences were similar while in Class 11, the Zurich loss experience was substantially higher, and in Class 18 was substantially lower than the IBC) Mr. Bradley objected to these comparisons for the reason that it opposed acutal Zurich premiums with artificially constructed industry premiums. He indicated that the IBC) Green Book figures are comparing IBC) exposure to an IAO adjusted premium. Mr. Bradley explained that the IBC) loss ratios are computed by dividing the premiums earned in the latter half of the year by the IAO exposures and then by multiplying the result by the industry exposures. In his opinion, a comparison is inappropriate as the Zurich loss ratio deals only with Zurich premiums and claims. (Evidence, Volume V, p. 18)

43Mr. Bradley acknowledged that there had not been any kind of independent investigation into the question of rates or premium structures of other companies. In Mr. Bradley's opinion, the Green Book figures coud not be adjusted to give any meaningful comparison with the Zurich results. The loss ratios shown in Exhibit 6B of the industry Green Book "do not mean anything to me. I make absolutely no use of them ... I would agree they probably do represent some sort of industry trend, I do not know what". (Evidence, Volume V, p. 20)

44At the time of Mr. Bates application for insurance in February 1983, the available information would have been the 1981 Green Books for both Zurich and IBC) These books would have contained information with respect to the years 1979, 1980 and 1981 inclusive. According to Mr. Bradley, when Zurich sets its rates, it looks at country-wide figures for the industry and for Zurich, as well as provincial figures for Zurich and for the IBC) The Ontario results for Zurich constitute 50 to 66 percent of the country-wide figures. All rates are set at the Head Office but differentials vary from province to province.

45Mr. Bradley stated that, in 1981, there were only 145 claims by Zurich policy holders in Class 11 in 1981 and 122 claims in the same Class in 1982. Mr. Bradley agreed that he would put very little weight on the data obtained from evaluating these limited claims. He testified that a figure of $145.00 was the loss cost per car insured by Zurich in 1981 was determined by dividing the total premiums ($16,834,477.00) by the total number of exposures (116,025). In other words, in 1981, Zurich insured 116,025 persons and for that insurance year incurred claims in the amount of $16,834,477.00. The company generated premiums in the amount of $19,994,977.00 for the 1981 business year. In response to a question from myself as to how differentials would be calculated if age, sex and marital status were no longer acceptable categories for under 25 year-old drivers, Mr. Bradley stated that the company would still commence with the same loss costs figure but its actuarial analysis and calculations would be applied to the existing classes. (Evidence, Volume V, p. 44)

46Mr. Bradley produced a schedule of surcharges from Zurich's rate manual dated October 1982 which was effective in February, 1983 and it was marked as Exhibit 10. This schedule explained Zurich's calculation of additional or surcharge premiums for policyholders who had committed highway traffic offences or who had been involved in accidents. Mr. Bradley indicated that he was not aware of any statistical basis for the various surcharges. (Evidence, Volume V, p. 35) It was agreed that no surcharge was applied to Mr. Bates, although if a surcharge had been added to his premium, it would have been on a percentage basis of his Class 11 premium. If Mr. Bates was paying a surcharge, it would have amounted to a considerably larger sum than a surcharge for a person in another rating class, being a percentage of his existing premium.

47Mr. Bradley testified that, where a father who is in Class 02 owns a car and has a son under-25 as an occasional driver, the rate differential would be 1.57. However, if the son was 22 or 23 years of age and owned a car which he drove to work, then the father would be considered the occasional driver and the differential would be lower – 1.45. Thus, there is no additional differential charge when a father is an occasional driver.

48Mr. Manley Pryce, Deputy Manager for Canada for Zurich, testified on its behalf. Mr. Pryce is a graduate of the University of British Columbia, Faculty of Law and has been employed by Zurich since 1953. He testified that the work of the actuarial department was given to him as a member of management responsible for automobile rates. The automobile rates are jointly reviewed by Mr. Pryce and the manager who is responsible for automobile rate lines. Mr. Pryce reviews the trends with his manager and considers whether it would be appropriate to introduce rate changes. When questioned whether rate changes are affected by the market place, Mr. Pryce responded that market forces do play a role: where there is a strong indication of a need for a rate adjustment, particularly upwards, market forces may very well prevent rate increases. (Evidence, Volume V, p. 79) If considering recommending a rate adjustment, Mr. Pryce writes to Zurich's branch managers indicating the actuarial data and his perception of the need for a rate adjustment. He would ask for the branch managers' business assessment of current market conditions and whether and to what extent rate changes would be appropriate. He would examine the loss ratio on an overall basis, by classes and overall of the total Zurich portfolio and then examine the actuarial indications. One aspect of the situation may affect, influence, soften or harden another issue. Mr. Pryce stated that: "when you look at the premiums related to a loss ratio in Class 11 of 125.9, it is obvious that we did not have adequate premiums ... from a rather small group of policy holders". (Evidence, Volume V, p. 81) From a pure business point of view, Mr. Pryce stated that on Class 11 Zurich had lost money. He further stated that, in his opinion, Zurich's automobile insurance business in Ontario was unprofitable during 1981 and 1982.

49Mr. Pryce emphasized during his cross-examination that his company considers average loss frequencies and loss ratios by class with respect to their total portfolio. When asked what relative weight is given to the average loss frequency and loss ratio, Mr. Pryce stated that: "I think we have to give it reasonable consideration, if you like, equal consideration to all of them to try and evaluate them in terms of trends or changes that have occurred". (Evidence, Volume V, p. 104) The factors that he would consider when setting rates are the rating of competing companies in the industry; the impact of individual rate changes on individual risk; the serious changes in loss ratio within a province; other changes within a Province; or changes in a premium tax. Mr. Pryce acknowledged that Zurich has accepted common law marriages as an indication of marriage for males under-25. He further indicated that Zurich's business surcharges would amount to less than 1 percent of their automobile premiums. Generally, surcharges with respect to convictions would not be added until the time of renewal. With respect to the issuance of a new policy, Zurich might become aware of convictions and the company would amend the endorsement and the originally quoted premium to include appropriate surcharges.

50Mr. Brian Robert Newton testified for the Complainant. Mr. Newton is the senior actuary in the office of the Superintendent of Insurance for Ontario in the Ministry of Consumer and Commercial Relations. This Office provides actuarial advice in the course of the Superintendent's regulatory function; the investigation of consumer complaints, such as unfair rating practices; the regulation of the. Facility Association and automobile insurance rates; and the review of rate filings by a number of automobile insurance companies. (Evidence, Volume VII, pp. 66–67)

51Mr. Newton testified that there is no code or codified principles or standards which govern the method of classification of risk systems, although there is now a movement towards one. There are, however, certain principles or standards which the industry follows in establishing classification of risk systems. These standards are stated in the draft report, Statement of Risk Classification Principles, a 1984 document of the Canadian Institute of Actuaries (Exhibit 31). This draft Report contains a proposed set of classification principles which is waiting to be adopted or amended by the Institute for the guidance of its membership. At present, this Report has not been accepted. The Report describes a grouping of the risks as being a necessary part of the insurance process as a means of determining averages and applying those averages to rate individuals. The Report then describes considerations which must be used to determine these specific classes:

1. Each class should exhibit different risk expectations;

2. Significantly dissimilar risks should be assigned to different classes;

3. Each class should be large enough to allow credible statistical predictions about that class;

4. There should be enough classes to avoid extreme discontinuities between adjacent classes. If the difference between two classes is very large, by reference to a particular variable, further subdivisions should be introduced to smooth out the difference;

5. Ambiguity should be avoided. Particular individuals should only be assignable to one class. There should be no doubt into which class they fall;

6. The variables should be capable of being measured reliably;

7. Where possible, there should be a "cause and effect" relation ship perceived by the insuring public. This however, is not always possible. For example, regarding life insurance, it is known that women live longer than men. The reason why is not known. Thus, sex is not really a cause yet it is a factor;

8. The classification should be publicly acceptable. (Evidence, Volume III, pp. 76–81)

52These considerations or principles can be summarized as separation, homogeneity, stability, verifiability, causality and acceptability. (Evidence, Volume VIII, p. 18)

53Mr. Newton testified that he personally agrees with these principles although he would place a greater emphasis on causality – the factor most clearly related to the degree of risk. This emphasis, he believes, would make the system more readily understood and accepted by the general public and by agents underwriting the losses. It would also encourage people to take greater control of their losses. For example, if people knew that their driving record influenced their rates, they would try to improve their driving record. (Evidence, Volume VII, p. 86 and Volume VIII, p. 18) In cross-examination, Mr. Newton also reiterated his preference for emphasizing causality:

Q. And it is ... you say you would like to emphasize causality.

A. I would say the factor ... if you are using a factor to cause this sub-division, which is generally acceptable by most people to be related to the risk, then you feel comfortable with the factor you are using for separation ...

... And you immediately start looking further into the matter to see if you can find some differences in related factors which will explain why you have observed the differences. And if you can discover that underlying, more, let us say, causal factor, or more closely related to causal factor, then it appears to be preferable to use the intuitively acceptable causal factor rather than the unexplainable classification factor. (Evidence, Volume VIII, pp. 20–21)

54Mr. Newton further testified that indirect factors or proxies were used by the industry, instead of the more direct but less tangible factors: "Instead of rating by a factor which you think is the real measure of the risk, there may be some difficulties, so you attempt to find another factor which will produce a similar sort of separation". (Evidence, Volume VIII, p. 75) Mr. Newton gave several examples of factors that are used as proxies. The first was discounts for students based on grades in school. The assumption is that those students with high grades study more and drive less. High grades, therefore, are used as a proxy or substitute for mileage. A second example given by Mr. Newton was the male classification factor, which is the subject of this inquiry. It is argued, Mr. Newton testified, that an average male will drive twice as far a year as the average female. Thus, "by simply dividing people into two classes, males and females, one, automatically, has identified a group, which on average, has a higher automobile insurance risk". (Evidence, Volume VII, p. 76)

55Mr. Newton also gave testimony concerning the origins of the Statistical Plan. He testified that it has its origins in the hearings in Ontario conducted by the Hodgson Commission in the late 1920s. This Commission recommended a uniform statistical plan. The initial plan was quite limited as it dealt with: 1) only two territories – Toronto and the rest of the province – and 2) the cost of cars, dividing them into expensive and less expensive. Over time, especially during the 1940s, 1950s and including the 1960s, the classifications were increased and the system developed to its present stage. The classification system is a national, industry-led system. (Evidence, Volume VII, pp. 87–89)

56Mr. Newton also testified about The Report of the Select Committee on Company Law of the Provincial Legislature, which dealt with motor vehicle insurance. This Report stated that there was a tendency by the public to question the basis of the rating system, especially regarding the number of classifications and the factors on which the classifications were based, such as charging higher rates for young drivers, and the use of conviction surcharges.

57In addition, Mr. Newton testified that he was asked to do a study of the Statistical Plan. He was instructed to look into the question of uniformity – did all companies have the same practice regarding driving records, such as the use of driver training discounts. (Evidence, Volume VII, pp. 91–92) Mr. Newton testified that, during the study, he experienced several problems. First, he had difficulty in addressing the issues on the basis of information from the Statistical Plan. The young male/female differentiation, the young driver surcharge, and the structure of the statistical categories meant that it was impossible to undertake comparisons.

A. ... It seemed to be a somewhat disorderly structure for the statistical categories. There was no sense of, say, statistical controls of holding everything else equal and measuring one variable, like age, across the system. (Evidence, Volume VII, p. 93)

He went on to testify that because the age variable stopped at 25, and the male/female differentiation only existed under 25, or under 30 for the unmarried males, it was "extremely difficult to draw conclusions about the particular variables that were in question". In Mr. Newton's opinion:

It was more, what I call, a self-fulfilling system. The divisions had been divided up to reflect rating classifications and differences appeared there ... there was not really convincing evidence in the structure of the statistics, to say that this particular structure was the right one. (Evidence, Volume VII, p. 93)

58The second problem that Mr. Newton identified was the lack of uniformity among insurance companies in the factors which they used to determine driving records. At that time, the system included the number of claim-free years a driver had. The categories were 5, 3, 2, 1 and 0 claim free years. However, the problem was that companies had different policies concerning this factor. Some companies had forgiveness plans which allowed persons in the five year claim free category one claim which would not put them in the 0 category. Other companies would place those drivers in the five year category into the three year, rather than the 0 year category if they made a claim and only if a driver made a second claim would that driver be placed into the 0 category. Furthermore, according to Mr. Newton, the 0 category was not a homogeneous group since the number of accidents in it could range from one to several. In short, Mr. Newton described these categories as "impure". He did say, however, that the statistics did appear to show that the risk of a person in a claim-free category 0 might be two to three times higher than a person in claim-free category 5. He further testified that, while the classifications of drivers by their driving record seemed to show meaningful differences, there was no uniformity in the coding. This problem, he stated, would cause some doubt, given the lack of uniformity of company practice. (Evidence, Volume VII, pp. 94–95)

59According to Mr. Newton's testimony, he participated in meetings held between the Superintendent's Office and the industry during 1977 and 1978. The reports from the industry supported the continued use of the age, marital status and sex factors, although they also stated that there was room for improving the statistical system. Despite this industry position, Mr. Newton testified that the series of meetings did produce a number of recommendations. One such recommendation called for the elimination of the 0–4 category for unmarried males between the ages of 25–29, on the basis that there were no categories for married or single females in that age range. Thus, unmarried males were not necessarily being compared with their peers in age.

60This recommendation was accepted and, consequently, all provinces, with the exception of Quebec, no longer have the "unmarried male between 25–29" category. Mr. Newton noted that one 1982 industry study under the Chairmanship of Don Wagner, the Vice-President of Zurich, recommended that the "Type of Use" categories be extended to the under-25 category. (Evidence, Volume VII, p. 100)

61Mr. Newton discussed the three specific categories at issue in this inquiry: the age, sex and marital status classifications for drivers under 25 years of age. Concerning the male/female differentiation, Mr. Newton testified that:

there seems to be nothing inherent in being male or female, of itself, which would cause a person to be a high risk for automobile insurance or a low risk for automobile insurance. (Evidence, Volume VII, p. 101)

Mr. Newton stated that, while differences do appear in the male and female categories, the structures of the Statistical Plan make it difficult to actually compare like males with like females. He further added that, since there does seem to be a difference in the industry's treatment of these two groups this distinction must be used as a proxy for other things.

62Mr. Newton testified that the male/female distinction is actually a proxy for 1) mileage; and 2) social habits. Some of the literatures suggests that young males drive twice as far as females and are more likely to drink and drive. (Evidence, Volume VII, p. 102) Mr. Newton objected to this use of male/female differentiation. He stated that males and females are not homogeneous, as there is considerable variation among young males and females in their driving habits. He went on to say that he felt "less confident about using a proxy which is not directed to the average exposure factor, and I would also try and seek out a better measure of the exposure". (Evidence, Volume VII, p. 102) Mr. Newton also stated that he felt that, by wrongly regarding the young males as a homogeneous group, the present system was unfair to the low mileage, careful, young male drivers and that it was under-charging high mileage female drivers whose social habits might not be so acceptable. (Evidence, Volume VII, p. 104)

63Mr. Newton also testified that a better alternative to the use of sex as a proxy for mileage would be to use the mileage factor itself. The insurance industry employs this mileage factor for drivers over the age of 25 through the "Type of Use" categories, and there is no reason why this determination could not be done for the under-25 category. This employment of the mileage factor, according to Mr. Newton, would reduce the variability in the male/female classes and would make the data more homogeneous with respect to risk. (Evidence, Volume VII, pp. 102–103) Furthermore, Mr. Newton testified that the sex factor should not be used as a proxy for social habits. According to Mr. Newton, the industry uses driver convictions and claim records as classification factors for those drivers over the age of 25; therefore, there is no reason why the industry could not do the same for the under 25 year-old category. In short, Mr. Newton testified that sex should not be used as a factor for classifying drivers.

64Concerning the second area at issue, the age classification, Mr. Newton testified that:

there doesn't seem to be any intuitive reason to suppose that because a person is one year-older that they are any more likely or less likely to be involved in an automobile accident the following year, all other things being equal. (Evidence, Volume VII, p. 104)

He went on to state that age is being used as a proxy for driver experience because young drivers obviously have less years' experience in driving and that, as drivers get older, they learn skills necessary to avoid accident situations. According to Mr. Newton, age is also used as a proxy for social habits since the presumption is that, as people get older, they drive less. Thus, age itself is not seen by Mr. Newton as a factor directly related to the risk of accident. (Evidence, Volume VII, p. 105) Mr. Newton further testified that there are studies which relate mileage by age, but he stated that they are not completely reliable. For example, the Ontario Ministry of Transportation's statistics do show a reduction in the accident frequency by age. They do not, however, differentiate according to fault. Mr. Newton stated that it is reasonable to believe "that much of this dropping off of accident rate with age, reflects a drop in the mileage driven, by age." He also emphasized that, within an age group, there would be considerable variation in individual habits. (Evidence, Volume VII, p. 105)

65Mr. Newton next addressed the issue of marital status. He began by stating that:

Again, it is difficult to see how the act of becoming married has any direct relationship to the automobile insurance risk. (Evidence, Volume VII, pp. 105–106)

In Mr. Newton's view, marital status, like age and sex, is a proxy for other factors: 1) social habits – marriage indicates an increase in responsibility and maturity, a presumption Mr. Newton stated was difficult to measure; and 2) mileage – married people drive less. Mr. Newton stated that these assumptions are questionable as it can be said that some married couples will drive and socialize as much after marriage as before marriage. Mr. Newton testified that the statistics in the IBC) Green Book indicate that, for urban areas, there is very little difference in loss experience between Class 11 unmarried males age 19–20 – and Class 08 – married males between the ages of 16–20. For Mr. Newton, this absence of difference indicates that:

for the urban areas, that possibly the distinction of marital status does not seem to be separating these groups into a different risk group. It could possibly be because the definition of who is married and who is not, is not really a workable classification difference. (Evidence, Volume VII, p. 107)

66As well, Mr. Newton testified that there are several companies which keep statistics over and above those required for the Statistical Plan, and which have different rating systems. For example, one company in Ontario allows a rate differential in the under 25 category for driving above or below 7,500 miles a year. Another company has occasional driver categories for males and females under 25. In cross-examination, he stated that this "seems to be a slightly more logical system of rating vehicles". He testified that he did not know the basis for these exceptions. (Evidence, Volume VIII, p. 41)

67During cross-examination, Mr. Newton was asked if he agreed with the statement that age, sex and marital status were factors which were more verifiable in the automobile context than the anticipated mileage factor. He replied:

A. I would have some qualification in claiming that age, sex and marital status are verifiable in the automobile insurance context.

He continued by saying that:

when one starts trying to categorize the vehicle by the age, sex, or marital status of the driver, the problem is there is not an absolute, direct link saying definitely that this person is the driver of the particular vehicle that is finding its way into the class statistics. (Evidence, Volume VIII, p. 28)

He testified that, while it is not possible to verify mileage for the next year at the time that a policy is issued, it is possible to get some admittedly imperfect estimate by reference to a previous period. He further stated that, while these three factors may be easier to measure than social habits or driving patterns, there is a problem of establishing the relationship of these three factors to the driving risk. (Evidence, Volume VIII, p. 29)

68Since social and driving habits are difficult to measure, it is apparent that insurance agents and brokers may make little effort to test the veracity of factors such as marital status. Mr. Newton pointed to the difficulty of defining the concept of marriage in this context: should it be construed in the sense of marriage ascending from law or religion or should the intricacies of the Family Law Reform Act concerning common-law marriage be considered. (Evidence, Volume VIII, p. 30) He further added that he has had people tell him that their insurance rates had increased because agents told them that they were not married although they were living in various relationships which they believed did qualify under the law as marriage. According to Mr. Newton, some people were also upset with agents giving them what they felt were moralistic sermons. Of the 4,000,000 insured in Ontario, the Superintendent's Office has received about 2,000 complaints concerning such moralizing in the industry.

69Furthermore, Mr. Newton testified that Mr. Alexander's statement that the average loss rate for Class 11 drivers was three times as high as the loss rate for the average car does not address the matter at issue in this Hearing, namely, the fairness of charging single males under 24 higher insurance premiums than married persons, females and those single males over 25 years of age. In fact, he stated that "every class has experience which is higher or lower than the average". (Evidence, Volume VII, p. 111) He further stated that the statistical evidence supports the view that there is little significant difference between Class 08 and Class 11 drivers. (Evidence, Volume VIII, p. 32)

70During cross-examination, Mr. Newton was presented with the data from Exhibit 22A. The first set of data compares Claim Frequency in Ontario between Zurich and the industry:

	Zurich	Industry
Class 08	14.89	14.24
Class 10	26.88	23.16
Class 11	13.48	15.82

(Evidence, Volume VIII, p. 34)

Mr. Newton objected to combining Classes 10 and 11 and comparing them with Class 08. He testified that Class 08, consisting of married males, may contain older, more experienced drivers than those drivers in Class 10. The second set of figures which Mr. Newton was presented with concerned Loss Cost Per Car:

	Zurich	Industry
Class 08	$251.00	$382.85
Class 10	$606.00	$732.58
Class 11	$766.00	$525.14

Mr. Newton also objected to the use of Zurich's statistics here for, as he stated, Zurich's statistics were derived from a small number of cars. The industry data, on the other hand, was derived from a larger data base and is probably more reliable. He pointed out that Zurich's loss per car was much lower than the industry's figure in Class 11. Concerning these statistics, Mr. Newton stated that one must be sure that the figures are reliable and he questioned whether they represented one year's experience or three years' average experience.

71The next set of figures in Exhibit 22A concern Average Claim Cost:

	Zurich	Industry
Class 08	$1688.21	$2688.32
Class 10	$2256.64	$3163.07
Class 11	$5688.61	$3318.85

72Mr. Newton testified that both Zurich's statistics and the industry's statistics were unreliable in that they were based on inadequate populations. He stated that, therefore, accurate conclusions about differences in risk over time could not be made. He pointed out that Zurich's figures were based on 94 policies issued in Class 08; 320 policies in Class 10 and 1076 policies issued in Class 11. (Evidence, Volume VIII, p. 38) His second objection over the use of these average claim cost statistics to prove that discrimination on the basis of marital status was reasonable was that they did involve the isolation of the marital status factor.

A. The second point I would like to make, Sir with respect, is that we are comparing Class 08 with Class 11. We are not looking into the possible differences in driver experience in those two categories; we have not standardized the data for the differences in mileage exposure, which is regarded as a rating factor for the older groups. Therefore, it is difficult, at this stage, to draw a valid conclusion that the marital status, by itself, is responsible for the differences observed. (Evidence, Volume VIII, p. 38)

73Mr. Newton was then asked to comment on the fact that, in the industry figures for the Average Claim Cost, the Class 08 figure was $2688.32, while the Class 11 figure was $3318.85. He testified that "I am not necessarily certain in my mind that the difference justifies the differential being charged between those classes ... I would say, on a simple observing of these classifications, taking them with their faults, there is a difference in the numbers". (Evidence, Volume VIII, pp. 39–40)

74Mr. Newton was then cross-examined on the data in Exhibit 32D which compared the experience of the Facility Association with that of the industry for the years 1980 and 1981. The frequency and average cost per car, rural and urban, was as follows: Class 08 – Facility 750, Green Book 425; Class 11 – Facility 943, Green Book 533. Mr. Newton stated that, in his opinion, this comparison is a "classic example of the dangers of company frequencies and average loss costs in cells without taking into account the underlying variations". (Evidence, Volume VIII, p. 45) As Facility Association drivers are high risk drivers, there is a strong bias in that population compared to the population on which the Green Book is based. For example, according to Mr. Newton, there is almost a negligible proportion of drivers with a five-year claims-free record in the Facility Association; whereas, the Green Book has a heavy concentration of drivers with record 5. When comparing Facility Association experience with the experience of the Green Book, the excess in the Facility Association could be attributed to the excess population of drivers with record 0 whom one would expect to have higher claims. According to Mr. Newton:

it is not a valid method of drawing conclusions from numbers of this sort until you know that the underlying data has been, let us say, standardized, or the effects of other distortions have been removed from it as far as you can. And this data is very badly distorted. (Evidence, Volume VIII, p. 45)

Mr. Newton stated that this criticism applies to the Green Book data as well because of the driver experience, mileage and other factors involved. (Evidence, Volume VIII, p. 46)

75Mr. Newton also discussed other anomalies which exist because of: 1) the structure of the existing Statistical Plan and the companies' use of that Plan; and 2) the lack of certainty in the definitions of the statistical categories. One such anomaly concerns the fact that a principal driver over 25 in combination with an occasional driver under 25 may pay more than a principal driver who is under 25. (Evidence, Volume VII, p. 113) Mr. Newton testified that another reason why the system must be improved is that it is compulsory by law. The government now has a greater duty to ensure that rate classifications are not unfair or questionable.

76During cross-examination, Mr. Newton was asked about the financial state of the automobile insurance industry. He explained that it was a "mixed-bag": some companies are financially weak, while others, despite the fact that they are underwriting losses, are reporting a profit, after taking into account investment income. The rate of profit varies from company to company with some companies turning in "substantial profits". He continued that "some companies are not earning as much as they used to be accustomed to". (Evidence, Volume VIII, p. 8)

77Mr. Newton agreed that the Superintendent's Office has the power to order unfair discrimination to cease under Part 18 of the Insurance Act. Section 393(b)iii describes an unfair practice as:

Any unfair discrimination in any rate or schedule of rates between risks in Ontario of essentially the same physical hazards in the same territorial classifications. (Evidence, Volume VIII, p. 11)

Section 394 gives the Superintendent the right "to issue an order that such practices shall cease if he finds them". When asked if he would commence proceedings if unfair discrimination is discerned, Mr. Newton testified that his office takes an individual case approach.

Q. I am sure you would agree with me that, if as a result of those reviews, you came to the conclusion that there was unfair discrimination, then proceedings would be commenced, would they not?

A. As a matter of practice, we proceed more by way of individual complaints that consumers raise questions with the Office whether certain rating practices, in their particular case, are fair, or not. And we evaluate the situation and if we feel that there is a possibility that Part 18 is being transgressed, then we discuss the matter with the insurance company. But we work, generally, on an individual case basis rather than on the sort of looking at a rate manual and saying we are going to start a hearing under Part 18 on the basis of the rate manual. (Evidence, Volume VIII, pp. 12–13)

Mr. Newton also stated that the Office of the Superintendent has rarely had a rate manual submitted which contained a clear-cut section 18 violation.

78However, Mr. Newton testified that if the Office of the Superintendent found evidence that companies were not reporting information correctly to the Statistical Plan to a degree that would negate the usefulness of the information, it would take corrective action. Furthermore, if the structure of the Plan was faulty, due to either bad faith or good faith, the Office would "try and negotiate with the industry to have the structure altered". Consequently, if a situation was seen as unreasonable and negotiations failed, then his office might take action on its own.

79Mr. Newton admitted that, in 1982, he approved Facility Association rates which distinguished between Class 08 and Classes 11 and 10, with the latter classes being charged more. This approval was because, according to Mr. Newton, "that was in line with normal industry practice". According to Mr. Newton, the philosophy of the Superintendent's Office in 1980 was that, unless there was very clear evidence that a particular factor was grossly unfair, the Facility Association would not be used as a means to lead the industry. Rather, the Facility Association would follow the industry. (Evidence, Volume VIII, p. 48) Finally, Mr. Newton testified that negotiations between the Superintendent's Office and the industry will continue in the future with the Superintendent's Office putting forward a list of additional reforms to the Statistical Plan. The Office, ultimately, wishes to move towards "a rating system which conforms to the law and is statistically based" and wishes to design a "statistical plan which supports such a rating system". (Evidence, Volume VIII, p. 51) I found that Mr. Newton's evidence was given in a straightforward and helpful fashion and that his opinions were of considerable assistance to me in understanding and assessing the Statistical Plan.

80The next witness for the Complainant was Mr. Murray Alan Thompson, the Superintendent of Insurance for the Province of Ontario. Mr. Thompson, as Superintendent, is the public official responsible for the conduct of the insurance business in Ontario. Mr. Thompson was involved with the Select Committee of the Legislature on Company Law (the Committee) and preparation of its two reports, issued in 1977 and 1978. The Superintendent's Office provided an extensive briefing to the Committee and provided detailed information to the Committee's consultants. As stated earlier, the Committee was involved in the issues of compulsory automobile insurance, and, in particular, the examination of the Statistical Plan and the high risk pool mechanism.

81With regard to the Statistical Plan, the Committee was concerned with the complexity of the Plan and the public perception and understanding of it which, Mr. Thompson testified, was notably lacking. Mr. Thompson testified that, more particularly, the Committee was concerned with the Plan's apparent absence of objective factors for the determination of risk classification; the non-binding status of the Plan for rating purposes; and, the potential for distortion from the risk point of view which is caused by the categorization of individuals and which may lead to the public perception that insurance rates are not standardized. (Evidence, Volume IX, p. 10) Mr. Thompson testified that the Committee urged the Superintendent and the industry to develop a different Statistical Plan using factors other than age, sex and marital status.

82Mr. Thompson also stated that, as a result of the work of the Committee, steps were taken by the Office of the Superintendent. In 1980, automobile insurance was made compulsory and work was continued on the Statistical Plan in the area of classification of risk. Ongoing discussions with the industry continued in an effort to develop alternatives to the existing Plan. The goal of the Office was to develop more objective criteria:

to reduce classifications, if possible, to make it simpler and to, basically, fulfill the wish of the Select Committee, is that this become more understandable to the public, and, if possible, a form of encouragement to the public to get through the system quicker and into the proper risk category (Evidence, Volume IX, pp. 12)

More specifically, according to Mr. Thompson, a principal objective was to develop alternatives to the age, sex and marital status classifications. The Superintendent's Office also wanted to impress upon the industry that an alternative system was needed because the law was changing and placing a greater emphasis on equality and against discrimination, as witnessed by human rights legislation and the Canadian Charter of Rights and Freedoms. With these changes, the age, sex and marital status classifications would fall into the jurisdiction of the courts. Mr. Thompson testified that the industry did not recommend any changes to the system: "there seemed to be a reluctance to introduce any change into the system in a quick enough time ..." (Evidence, Volume IX, p. 16)

83However, specific changes to the Plan were made. The category 04, unmarried males between the ages of 25 and 29, was removed, as well as other changes effective January 1, 1985. When asked if he was satisfied with these changes, Mr. Thompson testified:

I am satisfied to the extent that I think that they have gone as far as they can at the present time. I am still concerned about the fact that a decent period of time is necessary for the development of credible statistics to show that the alternate factors that are selected will, in fact, stand up and will, in fact, produce, the type of statistical base that the industry requires to change its system. (Evidence, Volume IX, p. 18)

84Mr. Thompson testified that no proceedings had been commenced pursuant to the discrimination provisions of the Insurance Act because the Superintendent's Office wished to allow for "orderly change" of the rate structure. Without an alternative system, moving against the present classifications could have a chaotic effect on the market place. (Evidence, Volume IX, p. 20) I found Mr. Thompson to be an informed and concerned public servant whose testimony was both forthright and credible.

The Issue and Discussion

85The main issue in this case is whether Zurich has discriminated against the Complainant, Mr. Bates, by charging him an automobile insurance premium based on the factors of age, sex and marital status. As both the Respondent and the Complainant agree that a prima facie case of discrimination has been established, the only question to be answered is whether Zurich's rate classification system for automobile insurance does not contravene sections 1 and 3 of the Ontario Human Rights Code because it satisfies the criteria of section 21 of the Code. The relevant sections of the Code are set out below:

1. Every person has a right to equal treatment with respect to services, goods and facilities, without discrimination because of race, ancestry, place of origin, colour, ethnic origin, citizenship, creed, sex, age, marital status, family status or handicap.

2. Every person having legal capacity has a right to contract on equal terms without discrimination because of race, ancestry, place of origin, colour, ethnic origin, citizenship, creed, sex, age, marital status, family status or handicap.

3. The right under sections 1 and 3 to equal treatment with respect to services and to contract on equal terms, without discrimination because of age, sex, marital status, family status or handicap, is not infringed where a contract of automobile, life, accident or sickness or disability insurance or a contract of group insurance between an insurer and an association or person other than an employer, or a life annuity, differentiates or makes a distinction, exclusion or preference on reasonable and bona fide grounds because of age, sex, marital status, family status or handicap.

The issue, then, is whether Zurich's rate classification system for automobile insurance with its reliance on the group factors of age, sex, and marital status is "reasonable and bona fide"?

86This determination is a difficult problem in that this case appears to be one of first instance. There are no Canadian or American cases in which a similar exemption provision has been applied in the context of insurance contracts. It is necessary, therefore, and quite reasonable, to turn to cases which deal more generally with the same situation that this case presents: a limited exception to a general prohibition against discrimination. These cases have arisen in the employment discrimination context under human rights legislation in both Canada and the United States.

87The first and leading case in Canada on the bona fide occupational qualification exemption for discrimination in the employment context is Ontario Human Rights Commission v. Borough of Etobicoke (1982), 1982 CanLII 15 (SCC), 132 D.L.R. (3d) 14 [3 C.H.R.R. D/781] (S.C.C.). In this case, the Supreme Court of Canada dealt with the issue of compulsory retirement at age 60 for firemen. It held that compulsory retirement did not constitute a "bona fide occupational qualification and requirement" within the meaning of subsection 4(6) of the old Ontario Human Right Code, R.S.O. 1980, c. 340 (the "1980 Code"). Paragraph 4(1)(b) of the 1980 Code prohibited a refusal to employ or to continue to employ any person because of age. However, subsection 4(6) provided an exemption for discrimination based on age, sex or marital status where "age, sex or marital status is a bona fide occupational qualification and requirement for the position of employment".

88The Etobicoke case set out a two-part test of what constitutes a "bona fide occupational qualification and requirement (a "BFOQ"). McIntyre J. held:

To be a bona fide occupational qualification and requirement a limitation, such as a mandatory retirement at a fixed age, must be imposed honestly, in good faith, and in the sincerely held belief that such limitation is imposed in the interests of the adequate performance of the work involved with all reasonable dispatch, safety and economy, and not for ulterior or extraneous reasons aimed at objectives that would defeat the purpose of the Code. (19–20)

The Complainant in these proceedings does not allege any dishonesty, bad faith or malice on the part of the Respondent, and no ulterior motive has been shown or even suggested. As for the second part of the Supreme Court's test, McIntyre J. continued:

In addition, it must be related in an objective sense to the performance of the employment concerned, in that it is reasonably necessary to assure the efficient and economical performance of the job without endangering the employee, his fellow employees and the general public. (20)

89On the facts of the Etobicoke case, the Supreme Court found there was a lack of sufficient scientific evidence to establish a danger to public safety by the continued employment of persons over the mandatory retirement age. While the Court did not hold that scientific evidence was necessary in all cases, it did conclude:

In dealing with the question of a mandatory retirement age it would seem that evidence as to the duties to be performed and the relationship between the aging process and the safe, efficient performance of their duties would be imperative. (22)

90Counsel for Zurich objected to my relying upon the Etobicoke case stating that the "requirement" element of the "bona fide occupational qualification and requirement" criteria was essential to the Supreme Court's thinking that the discriminatory limitation must be "reasonably necessary" to assure the safe, efficient and economical performance of the job. (Argument, p. 85)

91However, the more recent decision of the Supreme Court of Canada in Caldwell v. Stuart (1984), 1984 CanLII 128 (SCC), 15 D.L.R. (4th) 1 [5 C.H.R.R. D/2483] (S.C.C.), strains this submission. In Caldwell, the Supreme Court applied the Etobicoke test, including the "reasonably necessary" component, to the "bona fide occupational qualification" provision of the British Columbia Human Rights Code, R.S.B.C. 1976, c. 186, (the "B.C. Code") despite the fact that this provision referred only to a "bona fide qualification" and did not include the term "requirement". On the facts, the Supreme Court held that the dismissal of a Roman Catholic school teacher by a separate school board – by reason of her failure to observe the requirements and practices of her church in marrying a divorced man – did not constitute discrimination in employment under the B.C. Code. The Court found that the employer had established the BFOQ defence to the allegation of discrimination on the basis of marital status and religion. The relevant section of the B.C. Code provides that:

Every person has the right of equality of opportunity based on bona fide qualifications in respect of his occupation or employment ... and ...

(a) No employer shall refuse to employ or continue to employ ... or discriminate against that person in respect to employment or a condition of employement ...

McIntyre J. in Caldwell reaffirmed the Etobicoke test for a "bona fide occupational qualification" defence, despite the fact that it was faced with statutory language that did not refer to a "requirement", but only to "bona fide qualifications". The Court held:

The test employed in the Etobicoke case has two branches. The first is subjective ... In addressing the second requirement of religious conformance was not reasonably necessary to assume the efficient performance of the teaching function in any objective sense. Considerations of economy and safety are not involved. However, the essence of the test remains applicable and may be phrased this way: ”˜Is the requirement of religious conformance by Catholic teachers, objectively viewed, reasonably necessary to assure the accomplishment of the objectives of the Church in operating a Catholic School with its distinct characteristics for the purposes of providing a Catholic education to its students?"

It is my opinion that objectively viewed ... the requirement of religious conformance ... is reasonably necessary to assure the achievement of the objects of the school. It is my view that the Etobicoke test is thus met ... (at 16–18)

It is true that the Court in Etobicoke indicated, as a general rule, that persons should be dealt with and assessed as individuals in the employment context without regard to group characteristics. It is also true that the nature of the insurance industry requires that persons be treated in groupings of some kind. It would not seem possible, then to deal with all insured persons or applicants for insurance solely on an individual basis in terms of risk classification or rate setting. For example, annual driving tests for Ontario's four million insured would not appear feasible.

92Nevertheless, the nature of the insurance industry does not necessarily lead to the conclusion that the "reasonably necessary" test in Etobicoke is inapplicable to the case at hand. While no perfect analogy exists for the application of Etobicoke in the present circumstances, it is reasonable to utilize the basic principles of the Etobicoke test. Both Etobicoke and the present case deal with a limted exception to a general prohibition against discrimination. Furthermore, the thrust of the Etobicoke case is that when interpreting and applying these limited exceptions of the Code, recognition should be given to the basic purpose and policy of the Code which, as the Code's preamble states, is the recognition of the inherent dignity and rights of the individual. It is apparent that this basic policy of recognizing the worth of the individual without discrimination is stated even more strongly in the present Code than in the predecessor legislation which was before the Court in Etobicoke. The present Code states the desire "to revise and extend the protection of human rights in Ontario". Persons in this province, then, should be dealt with and assessed, as far as possible, on an individual basis, without regard to stereotypical group characteristics.

93Just as employers are required to demonstrate that a discriminatory limitation is reasonably necessary to assure the safe, efficient and economic performance of the job in order to establish a "bona fide occupational qualification", so too insurers should be required by section 21 of the Code to demonstrate that a rate classification system, which discriminates on the basis of prohibited group characteristics, is reasonably necessary to ensure the efficient operation of the insurance system. Although counsel for Zurich maintained that such a burden of proof should not be required of insurance companies, counsel has implicitly accepted this burden in that, throughout the hearing, it has been maintained that the evidence collected pursuant to the Statistical Plan is the "objective" evidence that proves that these prohibited group characteristics are both reasonable and necessary. (Respondent's Statement of Fact and Law, paras. 19–20)

94How is one to decide what is "reasonably necessary" with respect to automobile insurance? In the recent decision of Carson v. Air Canada (February 15, 1985) (F.C.A.), the Federal Court of Appeal upheld a decision of a Review Tribunal constituted under the Canadian Human Rights Act, which had extended the basic two-part test of Etobicoke. The Review Tribunal had upheld the decision of an inferior Tribunal, holding that an employer's policy against hiring new pilots over the age of 27 was a discriminatory employment practice contrary to sections 7 and 10 of the Canadian Human Rights Act and that the employer had not satisfied the requirements of a "bona fide occupational requirement" defence provided in paragraph 14(a) of that Act (Carson v. Air Canada (1984), 5 C.H.R.R. D1857 (Rev.Trib.). Paragraph 14(a) provides:

It is not a discriminatory practice if (a) any refusal ... limitation ... or preference to any employment is established by an employer to be based on a bona fide occupational requirement.

95The Review Tribunal in Carson approved an American version of the Etobicoke test. The main authority for this American test was Smallwood v. United Airline Inc., 661 F.2d 303 (4th Cir. 1981), cert. denied 102 S.Ct. 2299 (1982). This case examined a fact situation very similar to that in Carson. In finding that an employer had failed to establish the "bona fide occupational qualification" defence afforded by the Age Discrimination in Employment Act, 29 U.S.C. ss. 621–34 (the "ADEA"), the United States Court of Appeals relied upon the following statement of the test for a "bona fide occupational qualification":

To justify a refusal to hire under the BFOQ exception contained in the ADEA, the burden is on the employer to meet a two-prong test:

(1) that the BFOQ which it invokes is reasonably necessary to the essence of its business ... and

(2) that the employer has reasonable cause, ie., a factual basis for believing that all or substantially all persons within the class ... would be unable to perform safely and efficiently the duties of the job involved, or that it is impossible or impractical to deal wan persons over the age limit on an individualized basis (307)

Mahoney, J. approved this analysis in the Federal Court of Appeal decision in Carson by writing:

Thus, in asking what is reasonably necessary to assure the safe performance by pilots of their duties as they age, it seems entirely reasonable to enquire if it is not possible or practical to deal with those pilots on an individual basis rather than preventing their initial employment by a blanket refusal to hire. (5)

96In a separate, concurring opinion in the Carson case MacGuigan J. endorsed this American version of the Etobicoke test. He stated:

According to the American test the first prong in the employer's burden of proof is to show that the BFOR it invokes is reasonably necessary to the essence of its business; this is the risk-safety element and could be satisfied by proving that the maximum hiring age requirement is reasonably necessary for public safety, which is admittedly the essence of the carrier's business. The second prong is for the employer to show that it has reasonable cause for believing that all or substantially all persons within the class would be unable to perform the duties of the position safely and effectively, or that it would be impossible or impractical to safeguard public safety through individualized testing; this is the availability-of-alternatives factor and could be satisfied by proving that, although safety was not imperilled by the results of aging in more than a few cases, it could not be effectually safeguarded by individualized testing, on the basis of our present scientific capabilities (17).

97The Smallwood test has been used in leading American decisions on the issue of sex discrimination in the context of employee pension plans. In the case of Weeks v. Southern Bell Telephone and Telegraph Co., 408 F.2d 228 (5th Circ. 1969), the Court of Appeals considered an allegation of denial of employment on the basis of discrimination under section 703(a)(1) of the Civil Rights Act of 1964, 42 U.S.C.A. section 2000 - 2(a)(1). Section 703(a)(1) provides for a BFOQ exception:

on the basis of religion, sex, or national origin in those certain instances where religion, sex, or national origin is a bona fide occupation qualification reasonably necessary to the normal operation of that particular business or enterprise ...

The plaintiff in Weeks had been denied the job of "switchman" because she was a woman and the job required strenuous lifting. The court held that the employer had simply not shown a factual basis for believing that all or substantially all women would be unable to perform the job safely and efficiently. Instead, the employer had asked the Court to adopt a stereotype of women and assume that the plaintiff was incapable of lifting more than 30 pounds.

98Counsel for the Complainant in this case also discussed two other American cases of City of Los Angeles Dept. of Water and Power v. Manhart, 98 S.Ct. 1370, (1978) and Arizona Governing Committee for Tax Deferred Annuity and Deferred Compensation Plans v. Norris, 103 S.Ct. 3492 (1983). Both are decisions on sex discrimination in the administration of employee pension plans and both are decisions made under section 703(a)(1) of the Civil Rights Act of 1964 (Title VII), the same prohibition which was considered in Weeks. In Manhart, the plaintiff alleged discrimination in employment on the basis of sex, contrary to section 703(a)(1), because the employer required its female employees to make larger contributions to its pension fund than its male employees. The employer had imposed this requirement on females because actuarial statistics demonstrated that women, as a class, live longer than men, and both male and female employees were entitled to equal monthly pension benefits upon retirement.

99The United States Supreme Court held that the employer had unlawfully discriminated against its female employees. The Court had little difficulty in dismissing the employer's actuarial statistics as insufficient justification for the disparate treatment:

The Department treated its women employees differently from its men employees because the two classes are in fact different. It is equally true, however, that all individuals in the respective classes do not share the characteristic that differentiates the average class representatives. Many women do not live as long as the average man ... The question, therefore, is whether the existence or non-existence of "discrimination; is to be determined by comparison of class characteristics or individual characteristics.

The statute's focus on the individual is unambiguous. It precludes treatment of individuals as simply components of a racial, religious, sexual, or national class ... Even a true generalization about the class is insufficient reason for disqualifying an individual to whom the generalization does not apply.

It is true, of course, that while contributions are being collected from the employees, the Department cannot know which individuals will predecease the average woman. Therefore unless women as a class are assessed an extra charge, they will be subsidized, to some extent, by the class of male employees.

But the question of fairness to various classes affected by the statute is essentially a matter of policy for the legislation to address. Congress had decided that classification based on sex, like those based on national origin or race, are unlawful. Actuarial studies could unquestionably identify differences in life expectancy based on race or national origin, as well as sex ...

Finally, there is no reason to believe that Congress intended a special definition of discrimination in the context of employee group insurance coverage. It is true that insurance is concerned with events that are individually unpredictable, but that is characteristic of many employment decisions. Individual risks, like individual performance, may not be predicted by resort to classifications prescribed by Title VII. Indeed, the fact that this case involves a group insurance program highlights a basic flaw in the Department's fairness argument. For when insurance risks are grouped, the better risks always subsidize the poorer risks. Healthy persons subsidize medical benefits for the less healthy ... Treating different classes of risks as though they were the same for the purposes of group insurance is a common practice that has never been considered inherently unfair. (emphasis added) (1375–1376)

The Norris case is a reaffirmation of Manhart dealing with a similar allegation of sex discrimination in the context of the administration of a pension plan. It concerned the payment of lower monthly retirement benefits to women than to men, despite equal contributions by both sexes to the pension plans.

100While some of the statements quoted from Manhart support the Complainant in this inquiry, they must be distinguished in that the employer had no statutory exception provision similar to section 21 of the Code to rely on in defence of differential treatment based on a true class generalization. Manhart remains relevant, however, in relation to its treatment of the use of statistics. As the Court held, a true generalization about a class does not, in itself, justify class-based treatment as there will always be individuals who do not share the characteristics that differentiate the average class-members. While, as stated above, it is unreasonable to expect automobile insurance companies to do individual testing, they should be scrutinized in terms of reasonable alternatives which may indicate whether the present scheme of rate classification is "reasonably necessary". Without requiring that insurance companies assess individual characteristics for everyone, it is reasonable to inquire if it is possible or practical to deal with applicants for insurance under-25 in the context of a classification system that relies on characteristics other then age, sex or marital status. In other words, are there any characteristics which can be used which are non-discriminatory in the sense that they are not prohibited by sections 1 and 3 of the Code? We can ask whether Zurich has demonstrated that it would be impractical or impossible to operate the automobile insurance system with the use of non-discriminatory criteria for rate classification. We may also ask whether Zurich has demonstrated a factual basis for believing that all or substantially all persons within the class must pay substantially higher premiums than the majority of drivers in Ontario.

101As for the degree of reasonableness that may be required in the present circumstances, the Board is mindful of the general rule that exceptions to discrimination prohibitions in human rights legislation should be narrowly construed. As stated by Manhony J. in Carson:

As is evidenced by section 2 of the Canadian Human Rights Act [similar in preamble to the Code], Parliament has made a fundamental decision to give preference to individual opportunity over competing social values. The preference is not absolute. Indeed it is limited in the present context by an employer's right to establish a BFOR. But, the Courts must be zealous to ensure that Parliament's primary intention that people should for the most part be judged on their merits rather than on group characteristics is not eroded by overly generous exceptions. This necessitates a narrow interpretation of the exceptions. (22)

102This approach has been also articulated in American case law. In Diaz v. Pan American Airways, 442 F.2d 385 (5th Circ. 1971), a refusal to consider males for flight attendant positions was found to constitute employment discrimination on the basis of sex contrary to the Civil Rights Act 1964, (Title VII) provision considered in Weeks. Relying on the Weeks decision, the Court held that the BFOQ provision must be narrowly and strictly construed to avoid the danger of having the exception ”˜swallow up' the rule of nondiscrimination (387). Similarly, in Orzel v. City of Wauwatosa Fire Department, 697 F 2d 743 (7th Circ. 1983), the United States Court of Appeals held that the BFOQ provision of the ADEA must be narrowly and strictly construed as an exception to a remedial statute. To do otherwise would run the risk of overuse of the exception and thereby introduce the very stereotyping that the ADEA was designed to prevent (748).

103The role of economics and efficiency in defining the parameters of reasonableness has been the subject of some controversy. In Etobicoke, the Supreme Court held that where economic concern does not raise public safety considerations, it may be difficult, if not impossible, to establish a BFOQ. However, the Review Tribunal in Carson held that, where the appropriate evidence is advanced by an employer, economic factors may well give rise to a BFOQ defence. The Tribunal pointed out that this possibility was indeed contemplated in the Etobicoke decision with the Court's reference to questions of "economy" and "economic performance". Nevertheless, the Tribunal made it clear that the availability of a BFOQ defence on economic grounds alone would be quite limited:

... the Tribunal is also of the opinion that...non-discrimination entails an economic cost. This would open the door too widely for many forms of discrimination ... It is not possible to believe that Parliament intended the Canadian Human Rights Act to apply only if there was a cost involved to employers ...

Parliament must have realized this when it enacted section 7 of the Code, which specifically prohibits discrimination in hiring because of age. If the increased cost of hiring resulting from the age of the employee can readily give rise to a BFOQ defence, Parliament substantially took away by section 14(a) the right to freedom from age discrimination in hiring which it specifically granted in section 7. It would take clear and unmistakable language to convince this tribunal that Parliament had such contradictory intentions.

It is our view that there is an inherent cost benefit assessment in Parliament's espousal of individual freedom set forth in section 2 of the Act, and in extending protection to all individuals in Canada. In any situation, the added possible direct economic cost to an employer, and possible inherent costs to society ... should be weighed against the obvious benefits to society through the protection and enhancement of the freedoms recognized in human rights legislation. They are so fundamental to the fabric of Canadian society that Parliament has stated clearly that the benefit of protection extended to an individual is to the ultimate benefit of all Canadians. (D1881–1883)

104American case law suggests that to establish a BFOQ defence, employers must demonstrate that the discrimination relates to business "necessity" and not just business convenience. In the Diaz case, the Court of Appeals stated:

We begin with the proposition that the use of the word "necessary" in s. 703(e) requires that we apply a business necessity test, not a business convenience test. That is to say, discrimination based on sex is valid only when the essence of the business operation would be undermined by not hiring members of one sex exclusively. (388)

105The legislature could well have had economic considerations in mind when it created the section 21 exception for insurance contracts. Still, the section calls for a cost-benefit analysis in terms of a balancing of harm: Is the harm to individual integrity under the Code sufficiently counter-balanced by the harm that would result to the industry if the present system was discontinued? In narrower terms, has Zurich established that the very essence of its business would be undermined if it could no longer rely on discriminatory group characteristics for its rate classification system?

106Applying the second part of the two-prong Etobicoke test of what is "reasonably necessary", I must determine whether the present insurance system is efficient? That is, is there reasonable cause to assume that the present risk and rate categories are accurate? Counsel for Zurich maintained that their statistics show that those groupings under age 25 which pay higher premiums – unmarried people, males and the general category of under-25's – are, in fact, a higher risk, as their claim frequencies and average claim cost are significantly higher than other insured drivers. Zurich goes further and asserts that the statistics show that young, single males even pay less than they should.

107While these statistics may be correct, one must be extremely careful in giving too much weight to statistical evidence. In the view of the Review Tribunal in Carson:

The fallacy in the approach of Dr. St. Pierre and Dr. Busby is that it tends to assume that a correlation between age and impairment is not only a necessary, but also a sufficient, basis to support age as a BFOR. The basic premise of human rights legislation is that the merits of the individual should be assessed. Otherwise, bona fide occupational requirements might be established simply on the basis of statistical averages of group characteristics. This would merely be stereotyping in a new format which is, if anything more invidious than traditional prejudices because it has an apparently scientific base. Even if the correlation between a discriminatory characteristic and a legitimate disqualification is shown to include a causal relationship, such as, in this case, the undisputed relationship between exposure to the life environment and the development of impairing conditions in some pilots, consideration must still be given to the rights of individuals for whom the correlation does not hold. (at D1876)

This statement is particularly applicable in this case, as there are many problems in accepting the Respondent's statistics at face value. There has simply been no evidence, in my opinion, offered to support the assertion that it has been scientifically proven that there is a direct, causal relationship between the discriminatory group factors used – age, sex and marital status and high risk. Age, sex and marital status are used for rate classification purposes only for those drivers under 25. Where did this distinction between those over and under the age of 25 originate? According to the evidence, there is no statistical or actuarial basis for choosing the age of 25 as the "magic age". This distinction appears to be based on little more than historical accident. Apparently, the age of maturity for insurance purposes could just as easily have been established at 23, 27, or 30. (Evidence, Volume II, p. 108–109 and Volume VII, p. 105) Any of these ages would likely reveal a similar differential in terms of claims experience, to that revealed currently between the under-25 and over-25 group.

108There has also been no evidence offered by Zurich to show that unmarried males under the age of 25 constitute a high risk. There was evidence which was offered in support of the contrary view that the factors of age, sex and marital status are, very likely, mere proxy factors which are not, in fact, causally related to the differentials in the claims experience demonstrated for the various groupings. (Evidence, Volume VII, p. 101–107) Thus, as age, sex and marital status have never been controlled or isolated in the statistics to ascertain if they are high risk factors, it has never been shown that there is reasonable cause to assume that all or substantially all unmarried males under 25 present a high risk. Many anomalies and non-statistical factors have raised doubts as to whether the present system is, in fact, scientifically or statistically based and, thus, whether statistics should be used to justify discrimination. There is no statistical basis to support a surcharge for convictions, multi-vehicle discounts, driver-training discounts or good-student discounts. Similarly, there is no statistical basis for the view that administrative costs are higher for the under-25 group. Furthermore, there are certain anomalies, such as the inconsistency between companies vis-a-vis their forgiveness policies, and the fact that a principal driver over 25 in combination with an occasional driver under 25 will pay more than a principal driver under 25 with an occasional driver over 25. These anomalies reinforce my opinion that the automobile premium rate-structure in Ontario is as much if not more based upon economic and business decisions than upon the Statistical Plan.

109Is the present anomalous system nevertheless reasonably necessary to assure the accomplishment of the objectives of the industry? Certainly, if the central objective is the ease of establishing insurance rates then this aspect of the Etobicoke test would be satisfied. But with the statutory duty of providing insurance rates that differentiate only on certain limited, reasonable and bona fide grounds, the ease of establishing rates is not an acceptable objective. It is clear that the Respondent must show that there are no other group factors which are non-discriminatory that can be used to determine the premiums for insured drivers.

110It is also quite clear that they have also failed to meet this test. First since 1977, there have been many meetings between the Superintendent's Office and the industry with the express purpose of amending the industry's classification system. The thrust of the Superintendent's position, in accordance with the comments of the Select Committee of the Legislature on Company Law, has been to de-emphasize or to eliminate entirely the factors of sex, marital status and age from the classification system for the under-25 group. Thus, the Superintendent feels that there are various alternatives which are possible and desirable.

111Counsel for Zurich maintained that the Superintendent's Office supported the discriminatory rate classification system as it did not legislate against it. While it is true that, aside from the Code, there has been no specific legislative reform, this absence is not absolute proof that the Superintendent's Office supports the present classification system. Mr. Newton testified that, despite the Superintendent's views agains[t] the classification system, legislation has not been imposed due to the working philosophy of mutual cooperation rather than confrontation with the industry. Mr. Newton also testified that it is the Superintendent's hope that the industry will be able to change the system on its own, rather than being forced by the Government. Zurich further argued that as the industry has relied on discriminatory groupings for many years (almost fifty), it can only rely on statistics based on the present groupings for the establishment of its rate classification system. MacGuigan J. in Carson addressed this issue quite well in relation to the status quo of a low maximum hiring age for airline pilots:

Finally, while the general practice in the industry undoubtedly favours a low maximum hiring age, even the status quo cannot alone sufficiently establish a BFOR, in the absence of other proof. (20).

112Zurich's position clearly does not meet the "reasonably necessary" standard. Simply asserting that a system has been in place for fifty years is an inadequate response to the argument that there are alternatives to the system. The insurance industry cannot rely on its inaction and tradition to support a discriminatory rate classification system.

113There is, in fact, an alternative to discriminatory rate classification factors which is being used today by the automobile insurance industry. For those insured persons who are 25 or older, the industry relies solely on non-discriminatory, causally connected factors such as driving record, driving experience and vehicle use to establish rate differentials. The evidence indicates that this system has worked well. There has been no evidence offered to suggest that these factors cannot be used for insured persons who are under 25.

114As stated above, the test to be used when the defence of discriminatory groupings is based on economic factors, alone, is one of high degree. Zurich failed to establish that the very essence of its business would be undermined if it could no longer rely on discriminatory group characteristics for its rate classification system. On the contrary, it is my decision that Zurich has presented no evidence to convince this Tribunal that non-discriminatory driver classifications would affect its economic efficiency.

115As I indicated at the outset of this decision, this hearing presents an interesting, yet difficult matter. The role of the Tribunal chairman is, in my opinion, to apply the new legislation in a careful and cautious fashion. I am mindful of the fact that the Ontario legislature has decided to apply the Code to contracts of automobile insurance and has provided the industry with a limited defence in section 21. As this decision indicates, it is my decision, after a careful review of both the evidence and the recent Canadian and American jurisprudence, that the respondent, Zurich, has not brought itself within the parameters of section 21 of the Code. As I have said earlier, I have no doubt that Zurich and the insurance industry have acted in good faith and that the current classification of automobile drivers developed for legitimate economic and business reasons. However, it is my conclusion that the current automobile driver classifications of unmarried drivers under 25 contravene the Code and do not fall within the provisions of section 21.

116I am also aware of the fact that this decision deals only with the question of liability and I, therefore, order that this Hearing resume on September 26th, 1985 at 10:00 A.M. to hear evidence and argument as to the appropriate remedy to be granted in this matter.

Footnotes

1. 1The Facility Association came into existence in Ontario in 1979 by virtue of Section 7(1) of the Compulsory Automobile Insurance Act. The Facility Association was introduced as a means of guaranteeing a readily accessible insurance market for all licensed owners and owners and operators of motor vehicles in Ontario.