Dickason v. University of Alberta, 1992 CanLII 30 (SCC), [1992] 2 SCR 1103

Olive Patricia Dickason Appellant

v.

The Governors of the University of Alberta Respondent

and

The Alberta Human Rights Commission Respondent

Indexed as: Dickason v. University of Alberta

File No.: 22700.

1992: May 5; 1992: September 24.

Present: La Forest, L'Heureux‑Dubé, Sopinka, Gonthier, Cory, McLachlin and Iacobucci JJ.

on appeal from the court of appeal for alberta

Civil rights ‑‑ Equality rights ‑‑ Mandatory retirement ‑‑ Provincial legislation prohibiting discrimination on basis of age ‑‑ Employer may show that alleged contravention "reasonable and justifiable in the circumstances" ‑‑ Whether university's policy of mandatory retirement at age 65 justified ‑‑ Whether criteria for limiting Charter rights under s. 1 apply ‑‑ Individual's Rights Protection Act, R.S.A. 1980, c. I‑2, ss. 7, 11.1.

Courts ‑‑ Appellate review ‑‑ Findings of fact ‑‑ Curial deference ‑‑ Whether appellate court should disturb findings of fact made by board of inquiry and Court of Queen's Bench.

Appellant, a tenured full professor at the University of Alberta, was forced to retire at the age of 65 pursuant to a mandatory retirement clause in the collective agreement between the university and its academic staff. She filed a complaint with the Alberta Human Rights Commission alleging that her forced retirement contravened s. 7 of the Individual's Rights Protection Act by discriminating against her on the basis of her age. Section 11.1 of the Act provides that discrimination on a prohibited ground will be permitted if the employer shows that the breach was "reasonable and justifiable in the circumstances". The board of inquiry appointed to hear appellant's complaint decided in her favour and ordered that she be reinstated. The Court of Queen's Bench upheld that decision, but it was overturned by the Court of Appeal.

Held (L'Heureux‑Dubé, Sopinka and McLachlin JJ. dissenting): The appeal should be dismissed. The university has shown that the impugned practice of mandatory retirement is reasonable and justifiable within the meaning of s. 11.1 of the Individual's Rights Protection Act.

Per La Forest, Gonthier, Cory and Iacobucci JJ.: In the construction of human rights legislation, the rights enunciated must be given their full recognition and effect, while defences to the exercise of those rights should be interpreted narrowly. In applying the test set out in Oakes for determining whether legislation can be justified under s. 1 of the Canadian Charter of Rights and Freedoms, the Court has adopted a flexible standard of proof which responds to the varying contexts in which the state seeks to invoke justification for the impugned legislation. Since a challenge brought under the Charter to legislation enacted by the state obviously affects a state interest, due deference must be given to the actions of the state manifested by the legislation under attack. The policy rationale for this varying standard cannot be automatically transferred to the consideration of human rights legislation, where the challenge will be to the actions of a private party. Charter cases may thus assist in the development of the test for determining whether a defence under s. 11.1 of the Individual's Rights Protection Act has been established, but the Oakes model is only appropriate if it is applied without any trace of deference to a private defendant, and only with a large measure of flexibility and due regard to the context.

The common law principle of curial deference to findings of fact made by a court of first instance has been to a large extent adopted in reviewing the decisions of administrative tribunals, although the standard of review will always be governed by the tribunal's empowering legislation. Curial deference must be given to a tribunal's findings of fact where there is a privative clause or where the findings were made within the tribunal's field of specialized knowledge. In this case, however, the Court should not be constrained by the conclusions of the board of inquiry. The Act has clearly indicated that a very broad standard of review would be appropriate. On a plain reading of the Act, it is clear that the legislature specifically intended that appellate courts should examine the evidence anew and, if deemed appropriate, make their own findings of fact. The judge of the Court of Queen's Bench did not hear any viva voce testimony, but rather reassessed the evidence on the basis of the transcript of the board hearing. The Court of Appeal and this Court thus stand in the exactly the same position as he did with precisely the same record to consider, and the policy reasons in favour of deferring to the findings of fact of a trial court do not apply.

While this Court's decision in McKinney can provide some guidance, it does not determine the outcome of this case. When weighing the arguments on the issue of minimal impairment as part of the analysis under s. 1 of the Charter, the majority in McKinney considered whether the government had a reasonable basis for concluding that it impaired the relevant right as little as possible. To frame the question in this way imposes a significantly lower burden of proof on the defendant than s. 11.1, which requires a defendant to prove that the discriminatory policy, viewed objectively, constitutes only a minimal impairment of the right. While no deference should be given to the policy choice of the defendant, other factors may well be relevant. The courts have respected the unique role of universities in our society as self‑governing centres of learning, research and teaching safeguarded by academic freedom, and over the years have been very cautious in intervening in university affairs. The nurturing of academic freedom and the ensuring of faculty renewal are most delicate matters that do not lend themselves to a single clear-cut answer as to the proportionality between the burden of the discrimination complained of and the objectives sought. The evidence and assessment of competent and experienced university administrators will be of particular significance.

In this case the collective bargaining agreement authorizing compulsory retirement can also properly be taken into account. Parties may not generally contract out of a human rights statute. This rule resulted from the concern that there may be a great discrepancy in bargaining power between the person contracting out of human rights legislation and the party receiving the benefit of that term. Labour codes are specifically designed to overcome or compensate any imbalance in bargaining power, however, and with these statutory safeguards in place collective agreements take on a new and important significance. A collective agreement may very well provide evidence of the reasonableness of a practice which appears on its face to be discriminatory. It should be shown, however, that the agreement was freely negotiated by parties with relatively equal bargaining positions and that it did not discriminate unfairly against minorities. Here the term of the collective agreement relating to compulsory retirement will apply to every member of the faculty association. Moreover, the union did not negotiate the term in a vacuum, but rather in the context of a system of tenure which protects all members of faculty from dismissal without just cause, and provides a pension scheme assuring the financial security of all retiring members of faculty.

The objectives of mandatory retirement were stated to be the preservation of tenure, the promotion of academic renewal, the facilitation of planning and resource management and the protection of "retirement with dignity" for faculty members. Like the objectives put forward in McKinney, in which they are subsumed, they are of sufficient significance to justify the limitation of a constitutional right to equality. The impugned retirement practice is rationally connected to the objectives cited. The retirement of faculty members at the age of 65 ensures that the university may readily predict the rate at which employees will leave the institution and that positions are opened for new faculty. Mandatory retirement also allows the university to renew its faculty by introducing younger members who may bring new perspectives to their disciplines. It provides a means of remedying the twofold problem of limited funding and a "bulge" in the age distribution of professors. As well, the policy supports the existence of a tenure system which creates barriers to the dismissal of faculty members thereby enhancing academic independence. In the university setting, mandatory retirement also withstands the minimal impairment test. No obvious alternative policy exists which would achieve the same results without restricting the individual rights of faculty members. Finally, the effects of the prima facie discrimination are proportional to the legitimate objectives served.

Per L'Heureux‑Dubé and McLachlin JJ. (dissenting): Curial deference to the board of inquiry's findings of fact is consistent with both principle and precedent, and recognizes the "signal advantage" enjoyed by courts of first instance and administrative tribunals, which see and hear witnesses. A board's findings on social fact evidence should also be accorded some deference where these findings fall within the board's specific and primary mandate. A board whose determinations are not protected by a privative clause may be afforded less deference than a board which has the protection of such a clause, but this is only a question of degree.

In this case, it is clear that the trial judge appreciated the advantages enjoyed by the board, and that he relied on the board's findings of fact in making his own. This Court should accord the board's findings a similar measure of respect.

It has been established that human rights legislation is to receive a broad and purposive interpretation. Provisions which provide defences for discriminatory conduct must therefore be read narrowly. The analysis of s. 11.1 of the Act may be approached in a manner consistent with the model established in R. v. Oakes. The test for justifying discrimination under s. 11.1 is a strict one. This test requires the employer, in the face of a discriminatory practice, to prove the absence of a practical alternative to the discriminatory rule, satisfying a civil burden of proof. The flexible standard, whereby the defendant need not show that it adopted the least intrusive and offensive means of implementing policy, does not apply here. The rationale for the flexible standard is judicial deference to legislative choice, based on the idea that, with respect to resources and training, law makers are in a better position than courts to make policy choices between competing interests. This deference to legislative choice is completely unwarranted where, as here, the defendant is not a legislative body.

The university's policy of mandatory retirement at age 65 is not reasonable and justifiable under s. 11.1 of the Act. First, since parties generally may not contract out of human rights legislation, a collective agreement is not evidence of the reasonableness of a discriminatory practice. The prohibition against waiver of human rights provisions arises not only from a concern about inequality in bargaining power, but also because the rights guaranteed by human rights codes are seen as inherent to the dignity of every individual within our society. While the existence of a collective agreement whereby employees agree to limit their own rights may exceptionally be a factor in considering the justifiability of an employer's discriminatory policy, any such agreement must be scrutinized to ensure that it does not discriminate unfairly against a minority of the union membership, and that it was freely negotiated. The particular context of the bargain, including relevant legislation in place at the time of its conclusion, may greatly mitigate its evidentiary weight, as is the case here. In the circumstances, the collective agreement between appellant and the university is not evidence of the reasonableness of its mandatory retirement policy.

In light of this Court's conclusions in McKinney, the university's stated objectives are pressing and substantial. The university's fears about how tenure might be affected should mandatory retirement be eliminated are insufficient, however, to prove a rational connection between its objective of preserving the tenure system and its discriminatory policy. Peer evaluation is a fair and equitable way of assessing professors in good faith, on the basis of their teaching, research and publication records, rather than on their age. Unless abused, it poses no threat to academic freedom, and in fact enhances the value of tenure by ensuring that incompetent professors, young or old, are dismissed. The university has also failed to prove a rational link between the goal of faculty renewal and its policy of mandatory retirement. The argument that mandatory retirement at a fixed age allows the university to open positions to younger academics, thereby at once allowing a fresh infusion of ideas into the institution and remedying the problem of underfunding, does not stand up to scrutiny. It is based on the false premise that older workers are uniformly less productive and original than their younger colleagues. Further, the elimination of mandatory retirement would have only a limited effect on the number of jobs for young academics owing, in part, to the small number of academics who actually wish to keep working beyond the normal retirement age. The institutional planning argument, according to which mandatory retirement is necessary because it allows the administration to plan ahead, should also be dismissed. Other variables, such as resignations, deaths, and early retirement, are predicted with relative certainty by means of statistical forecasts. The slight inconvenience which inevitably remains cannot alone be used as a justification for the denial of equality on the basis of age. The retirement with dignity argument depends entirely on the idea that professors who reach the age of 65 must necessarily fear assessment on the basis of their performance, because that performance has necessarily declined rapidly with age. Given that the evidence conclusively refutes the myth of universal decline, and that peer evaluation is an effective tool for identifying incompetence, this proposition clearly fails at the outset. The policy of mandatory retirement does not in any event meet the minimal impairment test. Peer evaluations, based on objective assessment, offer a far more dignified approach to academic work and are thus infinitely preferable. Encouragement of early retirement is another non‑discriminatory alternative that would achieve the university's goals. Finally, the devastating effects that forced retirement has on a worker's finances, health, and self‑esteem are grossly disproportionate to any advantages the university gains by its discriminatory practice.

Per Sopinka J. (dissenting): The conclusion and for the most part the reasons of L'Heureux‑Dubé J. were agreed with. In McKinney this Court decided that mandatory retirement at a specified age was not constitutionally impermissible. Parliament or a provincial legislature by appropriate legislation could prohibit or permit it. The province of Alberta, in s. 11.1 of its Individual's Rights Protection Act, has left the decision to employers and employees provided that where mandatory retirement is resorted to, the employer must satisfy a board of inquiry that this discriminatory practice is reasonable and justifiable.

The jurisprudence developed with respect to s. 1 of the Charter is a useful guide in applying s. 11.1. In determining whether the proportionality factor has been met, the Court should not defer to the decision of an employer in the same way as in the case of a governmental actor. The tests developed for the application of s. 1 of the Charter and the defence to discrimination under human rights legislation are similar. While the former as expounded in Oakes is more elaborate, they both require that the impugned measure bear a rational relationship to a legitimate objective. The test is whether the requirement is "reasonably necessary" to assure the performance of the job. This is a question of fact to be determined by a board of inquiry subject to appeal proceedings. All the circumstances must be taken into account, including any agreement or collective agreement between the employer and employees. This is a factor, but in this case it has little or no weight.

The appeal should be allowed because the board of inquiry found on the evidence that the connection between the university's objective and its mandatory retirement policy was weak. Moreover, the board found that there were other more reasonable means for the university to achieve its objectives. No valid reason for disturbing these findings has been made out.

Cases Cited

By Cory J.

Considered: McKinney v. University of Guelph, 1990 60 (SCC), [1990] 3 S.C.R. 229; referred to: Canadian National Railway Co. v. Canada (Canadian Human Rights Commission), 1987 109 (SCC), [1987] 1 S.C.R. 1114; Ontario Human Rights Commission v. Borough of Etobicoke, 1982 15 (SCC), [1982] 1 S.C.R. 202; Brossard (Town) v. Quebec (Commission des droits de la personne), 1988 7 (SCC), [1988] 2 S.C.R. 279; R. v. Oakes, 1986 46 (SCC), [1986] 1 S.C.R. 103; Irwin Toy Ltd. v. Quebec (Attorney General), 1989 87 (SCC), [1989] 1 S.C.R. 927; Stein v. The Ship "Kathy K", 1975 146 (SCC), [1976] 2 S.C.R. 802; N.V. Bocimar S.A. v. Century Insurance Co. of Canada, 1987 68 (SCC), [1987] 1 S.C.R. 1247; Klimashewski v. Klimashewski Estate, 1987 13 (SCC), [1987] 2 S.C.R. 754; National Corn Growers Assn. v. Canada (Import Tribunal), 1990 49 (SCC), [1990] 2 S.C.R. 1324; Lester (W.W.) (1978) Ltd. v. United Association of Journeymen and Apprentices of the Plumbing and Pipefitting Industry, Local 740, 1990 22 (SCC), [1990] 3 S.C.R. 644; Gendron v. Supply and Services Union of the Public Service Alliance of Canada, Local 50057, 1990 110 (SCC), [1990] 1 S.C.R. 1298; Bell Canada v. Canada (Canadian Radio‑television and Telecommunications Commission), 1989 67 (SCC), [1989] 1 S.C.R. 1722; Winnipeg School Division No. 1 v. Craton, 1985 48 (SCC), [1985] 2 S.C.R. 150.

By L'Heureux‑Dubé J. (dissenting)

Harrison v. University of British Columbia (1986), 1986 1297 (BC SC), 30 D.L.R. (4th) 206; R. v. Oakes, 1986 46 (SCC), [1986] 1 S.C.R. 103; McKinney v. University of Guelph, 1990 60 (SCC), [1990] 3 S.C.R. 229; Ontario Human Rights Commission v. Borough of Etobicoke, 1982 15 (SCC), [1982] 1 S.C.R. 202; Saskatchewan (Human Rights Commission) v. Saskatoon (City), 1989 18 (SCC), [1989] 2 S.C.R. 1297; Clarke v. Edinburgh and District Tramways Co., [1919] S.C. (H.L.) 35; Laurentide Motels Ltd. v. Beauport (City), 1989 81 (SCC), [1989] 1 S.C.R. 705; Stein v. The Ship "Kathy K", 1975 146 (SCC), [1976] 2 S.C.R. 802; Klimashewski v. Klimashewski Estate, 1987 13 (SCC), [1987] 2 S.C.R. 754; N.V. Bocimar S.A. v. Century Insurance Co. of Canada, 1987 68 (SCC), [1987] 1 S.C.R. 1247; Lapointe v. Hôpital Le Gardeur, 1992 119 (SCC), [1992] 1 S.C.R. 351; Bell Canada v. Canada (Canadian Radio‑television and Telecommunications Commission), 1989 67 (SCC), [1989] 1 S.C.R. 1722; National Corn Growers Assn. v. Canada (Import Tribunal), 1990 49 (SCC), [1990] 2 S.C.R. 1324; Lester (W.W.) (1978) Ltd. v. United Association of Journeymen and Apprentices of the Plumbing and Pipefitting Industry, Local 740, 1990 22 (SCC), [1990] 3 S.C.R. 644; Zurich Insurance Co. v. Ontario (Human Rights Commission), 1992 67 (SCC), [1992] 2 S.C.R. 321; Re Ontario Human Rights Commission and City of North Bay (1977), 1977 1253 (ON CA), 17 O.R. (2d) 712; Insurance Corp. of British Columbia v. Heerspink, 1982 27 (SCC), [1982] 2 S.C.R. 145; Ontario Human Rights Commission v. Simpsons‑Sears Ltd., 1985 18 (SCC), [1985] 2 S.C.R. 536; Canadian National Railway Co. v. Canada (Canadian Human Rights Commission), 1987 109 (SCC), [1987] 1 S.C.R. 1114; Brossard (Town) v. Quebec (Commission des droits de la personne), 1988 7 (SCC), [1988] 2 S.C.R. 279; Central Alberta Dairy Pool v. Alberta (Human Rights Commission), 1990 76 (SCC), [1990] 2 S.C.R. 489; R. v. Edwards Books and Art Ltd., 1986 12 (SCC), [1986] 2 S.C.R. 713; Irwin Toy Ltd. v. Quebec (Attorney General), 1989 87 (SCC), [1989] 1 S.C.R. 927; Reference Re Public Service Employee Relations Act (Alta.), 1987 88 (SCC), [1987] 1 S.C.R. 313; PSAC v. Canada, 1987 89 (SCC), [1987] 1 S.C.R. 424; Andrews v. Law Society of British Columbia, 1989 2 (SCC), [1989] 1 S.C.R. 143; Black v. Law Society of Alberta, 1989 132 (SCC), [1989] 1 S.C.R. 591; Stoffman v. Vancouver General Hospital, 1990 62 (SCC), [1990] 3 S.C.R. 483; Tétreault‑Gadoury v. Canada (Employment and Immigration Commission), 1991 12 (SCC), [1991] 2 S.C.R. 22; Winnipeg School Division No. 1 v. Craton, 1985 48 (SCC), [1985] 2 S.C.R. 150.

By Sopinka J. (dissenting)

McKinney v. University of Guelph, 1990 60 (SCC), [1990] 3 S.C.R. 229; Irwin Toy Ltd. v. Quebec (Attorney General), 1989 87 (SCC), [1989] 1 S.C.R. 927; R. v. Keegstra, 1990 24 (SCC), [1990] 3 S.C.R. 697; R. v. Butler, 1992 124 (SCC), [1992] 1 S.C.R. 452; R. v. Oakes, 1986 46 (SCC), [1986] 1 S.C.R. 103; Ontario Human Rights Commission v. Borough of Etobicoke, 1982 15 (SCC), [1982] 1 S.C.R. 202; Brossard (Town) v. Quebec (Commission des droits de la personne), 1988 7 (SCC), [1988] 2 S.C.R. 279; Stein v. The Ship "Kathy" K, 1975 146 (SCC), [1976] 2 S.C.R. 802; Ontario (Attorney General) v. Bear Island Foundation, 1991 75 (SCC), [1991] 2 S.C.R. 570; Lapointe v. Hôpital Le Gardeur, 1992 119 (SCC), [1992] 1 S.C.R. 351.

Statutes and Regulations Cited

Alberta Rules of Court, Alta. Reg. 390/68, r. 518.

Canadian Charter of Rights and Freedoms, ss. 1, 7, 15.

Charter Omnibus Act, S.A. 1985, c. 15, s. 14.

Individual's Rights Protection Act, R.S.A. 1980, c. I‑2, ss. 7 [am. 1990, c. 23, s. 4], 11.1 [ad. 1985, c. 33, s. 5], 31(1), 33, 38(a) [rep. & sub. 1985, c. 15, s. 14].

Individual's Rights Protection Act, S.A. 1972, c. 2, s. 6.

Individual's Rights Protection Amendment Act, 1985, S.A. 1985, c. 33, s. 5.

Human Rights Code, R.S.O. 1990, c. H.19, s. 42(3).

Universities Academic Pension Act, R.S.A. 1980, c. U‑6.

Universities Academic Pension Act, S.A. 1978, c. 36, s. 9.

Universities Academic Pension Plan Act, S.A. 1985, c. U‑6.1.

Authors Cited

Alberta Hansard, May 21, 1985, p. 1085.

Blackburn, Robert T. and Janet H. Lawrence. "Aging and the Quality of Faculty Job Performance" (1986), 23 Review of Educational Research 265.

Cole, Stephen. "Age and Scientific Performance" (1979), 84 Am. J. Soc. 958.

Ford, Robert C. and Myron D. Fottler. "Flexible Retirement: Slowing Early Retirement of Productive Older Employees", [1985] Human Resource Planning 147.

Krashinsky, Michael. "The Case for Eliminating Mandatory Retirement: Why Economics and Human Rights Need Not Conflict" (1988), XIV Canadian Public Policy 40.

Necheles‑Jansyn, Ruth F. "Retirement in Academe: Special Case or Social Model?", [1983] Aging and Work 175.

Ontario. Commission on the Future Development of the Universities of Ontario. Ontario Universities: Options and Futures. Toronto: The Commission, 1984.

Reid, Frank. Economic Aspects of Mandatory Retirement. Toronto: Centre for Industrial Relations and Department of Economics, University of Toronto, 1987.

Roadburg, Alan. Aging: Retirement, Leisure and Work in Canada. Toronto: Methuen, 1984.

Walker, James W. and Harriet L. Lazer. The End of Mandatory Retirement: Implications for Management. New York: John Wiley & Sons, 1978.

APPEAL from a judgment of the Alberta Court of Appeal (1991), 1991 ABCA 208, 81 Alta. L.R. (2d) 393, 117 A.R. 11, 2 W.A.C. 11, 83 D.L.R. (4th) 1, [1991] 6 W.W.R. 377, reversing a judgment of Murray J. (1988), 1988 3530 (AB KB), 62 Alta. L.R. (2d) 209, 91 A.R. 350, 9 C.H.R.R. D/5403, affirming a decision of a Board of Inquiry (1987), 1987 8520 (AB HRC), 9 C.H.R.R. D/4468, upholding appellant's complaint of unlawful discrimination. Appeal dismissed, L'Heureux‑Dubé, Sopinka and McLachlin JJ. dissenting.

Sheila J. Greckol and Jo‑Ann R. Kolmes, for the appellant.

Peter M. Owen, Q.C., and Greg A. Harding, for the respondent Governors of the University of Alberta.

J. Leslie Wallace and Sarah FitzGerald, for the respondent Alberta Human Rights Commission.

Cory J.

The judgment of La Forest, Gonthier, Cory and Iacobucci JJ. was delivered by

Cory J. -- At issue on this appeal is whether the mandatory retirement policy for faculty of the University of Alberta contravenes the Individual's Rights Protection Act, R.S.A. 1980, c. I-2, of that province and if it does whether the policy can be justified under s. 11.1 of that Act.

Factual Background

On June 24, 1975 the University of Alberta hired the appellant Dr. Olive Patricia Dickason. On June 22, 1976 Dr. Dickason signed a contract in which she specifically agreed to be bound by the Academic Staff Agreement between the University and the Association of Academic Staff of the University of Alberta (AASUA). Article 18.01 of the agreement provided a normal retirement date for staff members at the age of 65 years. The appellant was appointed an assistant professor on July 1, 1976. On July 1, 1978, the Academic Staff Agreement was re-named as the Faculty Agreement. Article 18.01 was amended to refer to new pension legislation; 65 remained the age of retirement. On November 2, 1979 the University awarded her tenure, pursuant to the provisions of the Faculty Agreement. On June 1, 1985, Dr. Dickason was promoted to the position of full professor. On June 30, 1985, the University retired Dr. Dickason, pursuant to the terms of art. 18.01 of the Faculty Agreement. She was appointed professor emeritus, an indication of the esteem in which she was held. Indeed, her competence as a professor was never questioned.

Dr. Dickason filed a complaint alleging that her forced retirement was a violation of the Individual's Rights Protection Act (the "IRPA"). A Board of Inquiry was appointed and ruled in her favour: (1987), 1987 8520 (AB HRC), 9 C.H.R.R. D/4468. The University appealed this result to the Court of Queen's Bench without success: (1988), 1988 3530 (AB KB), 62 Alta. L.R. (2d) 209, 91 A.R. 350, 9 C.H.R.R. D/5403. A further appeal was taken by the University to the Court of Appeal and that Court overturned the Board's decision: (1991), 1991 ABCA 208, 81 Alta. L.R. (2d) 393, 117 A.R. 11, 2 W.A.C. 11, 83 D.L.R. (4th) 1, [1991] 6 W.W.R. 377. It is from that decision that this appeal is brought.

Pertinent Legislation

A brief history of the pertinent human rights and pension legislation must be given as these Acts were amended over the years that Dr. Dickason was a member of the Faculty of the University. On January 1, 1973, the Individual's Rights Protection Act, S.A. 1972, c. 2, came into force. It prohibited employment discrimination on the basis of age. However the protected age group was limited to those between 45 and 65 years of age. On July 1, 1978 the province proclaimed the Universities Academic Pension Act, S.A. 1978, c. 36. This provided that every member of the pension plan shall retire on "the normal retirement date" which was stated to be the age of 65 (s. 9). In 1985 the IRPA was amended by the introduction of a provision allowing exceptions to the bar against discrimination where it was "reasonable and justifiable" (Individual's Rights Protection Amendment Act, 1985, S.A. 1985, c. 33, s. 5). On April 18, 1985 the province introduced the Universities Academic Pension Plan Act, S.A. 1985, c. U-6.1. This Act replaced the previous pension legislation and came into force November 1, 1985. The new legislation provided for "a normal pensionable date" of 65 years and allowed the participation of all employees in the plan, including those over that age. On June 5, 1985 the Charter Omnibus Act, S.A. 1985, c. 15, was passed. This removed the cap of 65 years on the age definition contained in the IRPA.

Legislation

Individual's Rights Protection Act

7(1)No employer or person acting on behalf of an employer shall

(a) refuse to employ or refuse to continue to employ any person, or

(b) discriminate against any person with regard to employment or any term or condition of employment,

because of the race, religious beliefs, colour, gender, physical disability, mental disability, marital status, age, ancestry or place of origin of that person or any other person.

11.1 A contravention of this Act shall be deemed not to have occurred if the person who is alleged to have contravened the Act shows that the alleged contravention was reasonable and justifiable in the circumstances.

33(1) A party to a proceeding before a board of inquiry may appeal the order of the board to the Court of Queen's Bench by originating notice filed with the clerk of the Court of the judicial district in which the inquiry was held.

(2) If the grounds of appeal consist wholly or partly of a question of fact or mixed fact and law, no appeal lies without the leave of a judge of the Court of Queen's Bench.

(4) The judge may, after taking into consideration any matters that he considers appropriate, grant leave to appeal under subsection (2).

(6) The Court may confirm, reverse or vary the order of the board and may make any order that the board can make under section 31.

(The wording in s. 33 has since been altered, S.A. 1990, c. 23, s. 14.)

38 In this Act,

(a) "age" means 18 years of age or older;

Proceedings in the Courts Below

Board of Inquiry

The Chairman of the Board of Inquiry found that the University policy of mandatory retirement at age 65 contravened the law. He observed that s. 7 of the IRPA prohibits discrimination on the basis of age. Once that infringement was established the onus then shifted to the employer to show that the discrimination was reasonable and justifiable pursuant to s. 11.1 of the IRPA. The Chairman observed that human rights legislation is fundamental law. As a result, in his view, s. 11.1 should be interpreted restrictively so as to ensure that a right was limited only in "compelling circumstances".

He found that the similarity in wording and legislative purpose between s. 11.1 of the IRPA and s. 1 of the Canadian Charter of Rights and Freedoms was such that the Oakes test should be used as a guide to the application of s. 11.1 of the IRPA. Using that test he found that the University's objectives, namely retirement with dignity, planning, academic renewal and preservation of the tenure system, were of sufficient importance to warrant overriding s. 7. He found as well that mandatory retirement was connected to all four objectives of the University. However, he concluded that the policy did not survive the minimal impairment test.

The Chairman specifically rejected the University's submission that the collective agreement justified the forced retirement of Professor Dickason. Rather, it was his opinion that the collective agreement in providing for mandatory retirement of faculty members represented an impermissible attempt to contract out of the IRPA.

In the result he ordered that Dr. Dickason be reinstated with compensation for her loss of income and benefits.

Court of Queen's Bench

The judge of first instance agreed that the rights guaranteed by provincial human rights statutes should be protected to the same degree as comparable ones guaranteed in the Charter. He, too, found that the criteria for limiting Charter rights in accordance with s. 1 should be used as a guide to applying s. 11.1 of the IRPA. He concluded that three of the objectives proposed by the University were sufficiently pressing and substantial to justify overriding the s. 7 right although he expressed some doubts as to whether the "retirement with dignity" objective could be sufficient to override s. 7. He concluded however that the University had not satisfied the onus resting upon it of demonstrating that mandatory retirement was rationally connected to those objectives. If it were required he would also have found that the minimal impairment requirement had not been satisfied by the University.

Alberta Court of Appeal

The Court of Appeal rejected the submission that it should not interfere in findings of fact made by the chambers judge. In doing so it observed that the chambers judge had relied on precisely the same record that was made available on the appeal. The court considered the case in light of the judgment of this Court in McKinney v. University of Guelph, 1990 60 (SCC), [1990] 3 S.C.R. 229. It noted that La Forest J. had held that universities were not government actors within the scope of the Charter but had then gone on to analyze the mandatory retirement requirement on the hypothetical basis that universities were in fact government actors. The Court of Appeal agreed with the chambers judge that an employer should not receive the "benefit of the doubt legislators may be extended". That is to say, the court was of the view that the deference extended by the courts to legislators should not apply to private employers. The Court of Appeal went on to assume that La Forest J. did not adopt a position of deference when considering the application of s. 1 to universities. It concluded on that basis that the issue in the case at bar was rendered moot and that the university must succeed.

Issues

Two issues are presented in this appeal.

1. Did the Court of Appeal err in concluding that the retirement policy of the University was justified pursuant to s. 11.1 of the IRPA?

2. Should this Court or the Court of Appeal disturb the findings of fact which were made by the Board of Inquiry and the Court of Queen's Bench?

Analysis

The Statutory Scheme Which Applied to Dr. Dickason

Before the merits of this appeal can be explored an explanation must be given as to the statutes which will be considered. At the date of Dr. Dickason's retirement in June 1985 the Universities Academic Pension Act contained a provision which mandated the retirement of faculty members at the age of 65. The new legislation which removed the mandatory retirement age had not yet been proclaimed. Thus, in June 1985 the University of Alberta was required by legislation to enforce the retirement of Dr. Dickason. The University could have raised this statutory requirement to support its defence that the retirement practice was reasonable. It chose not to do so. At the hearing, appellant's counsel explained that the parties had agreed to proceed as if the amendments to the pension legislation had been proclaimed before Dr. Dickason retired. Accordingly, it is not necessary to consider the appellant's submission, referred to in oral argument, that the IRPA guarantee against age discrimination without an age cap had primacy over the pension legislation. Rather, the appeal can be considered in the context of the relevant legislation in its present form.

Age Discrimination

Section 7 of the IRPA prohibits discrimination in employment on the basis of race, religious beliefs, colour, gender, physical disability, mental disability, marital status, age, ancestry or place of origin. Dr. Dickason's competence has never been questioned by the University. In these circumstances it is not unexpected that the parties agree that the impugned policy of compulsory retirement contravenes s. 7 of the IRPA. The sole question is whether the practice is justifiable pursuant to the provisions of s. 11.1 of the IRPA.

Section 11.1 of the IRPA

Section 11.1 of the IRPA provides that discrimination on a prohibited ground will be permitted if the defendant (University) establishes that the breach was reasonable and justifiable in the circumstances. The appellant contends that the University failed to meet the onus of proof it bears as an employer to satisfy the test set out in this section. The resolution of this appeal will turn upon the interpretation of the justification defence provided by the section.

What is the Nature of the Test Set by s. 11.1?

Although this Court has not yet considered the application of this section, a number of cases have dealt generally with the interpretation of human rights statutes. Dickson C.J. set out the guiding principles in Canadian National Railway Co. v. Canada (Canadian Human Rights Commission), 1987 109 (SCC), [1987] 1 S.C.R. 1114, at p. 1134:

Human rights legislation is intended to give rise, amongst other things, to individual rights of vital importance, rights capable of enforcement, in the final analysis, in a court of law. I recognize that in the construction of such legislation the words of the Act must be given their plain meaning, but it is equally important that the rights enunciated be given their full recognition and effect. We should not search for ways and means to minimize these rights and to enfeeble their proper impact. Although it may seem commonplace, it may be wise to remind ourselves of the statutory guidance given by the federal Interpretation Act which asserts that statutes are deemed to be remedial and are thus to be given such fair, large and liberal interpretation as will best ensure that their objects are attained.

The right against discrimination provided in human rights statutes will be subject to any defence provided by those same statutes to those who discriminate. However, as McIntyre J. recognized in Ontario Human Rights Commission v. Borough of Etobicoke, 1982 15 (SCC), [1982] 1 S.C.R. 202, a defence which allows discrimination to continue stands as an exception to the rule of non-discrimination. In Brossard (Town) v. Quebec (Commission des droits de la personne), 1988 7 (SCC), [1988] 2 S.C.R. 279, this Court held that while rights should receive a broad interpretation, defences to the exercise of those rights should be interpreted narrowly. Current human rights enactments seek to broaden the impact of individual rights, yet they strive to provide a balancing mechanism so that the many competing interests of society can be accommodated. For example, the Charter provides such a means of balancing by way of s. 1. Indeed, the parties to this appeal stressed the similarities that exist between s. 1 of the Charter and s. 11.1 of the IRPA. Both sections require that those who are prima facie in breach of the provisions of the Act establish that the breach is "reasonable and justifiable". I think it can be taken that the similarity exists between the Charter and the IRPA because they are designed to serve many of the same functions.

The resemblance between the two sections is apparent on their face. Indeed, the legislature of Alberta did rely upon s. 1 of the Charter as a model when the amendments of the IRPA were enacted. See Alberta Hansard, May 21, 1985, p. 1085. The fact that the two sections are remarkably similar and fulfil comparable roles should be taken into account when interpreting them. Yet it must be remembered there is a crucial difference between human rights legislation and constitutional rights. Human rights legislation is aimed at regulating the actions of private individuals. The Charter's goal is to regulate and, on occasion, to constrain actions of the state. This essential difference must be borne in mind when the defences provided by s. 1 of the Charter and s. 11.1 of the IRPA are considered. The cases which have dealt with s. 1 can only provide a useful source for identifying the requirements of s. 11.1 if this basic distinction is kept in mind.

The first and fundamental test for determining whether legislation can be justified under s. 1 of the Charter is set out in R. v. Oakes, 1986 46 (SCC), [1986] 1 S.C.R. 103. That familiar test directs the party raising a s. 1 defence to demonstrate: (i) that the restriction of a right is undertaken in the pursuit of a pressing and substantial objective and (ii) that the impugned restrictive measure is proportional to the enacted measure as evidenced by the fact that it is (a) rationally connected, (b) constitutes a minimal impairment to the right and (c) is proportional in its effects. In its application, the Court has adopted a flexible standard of proof which responds to the varying contexts in which the state seeks to invoke s. 1 justification for the impugned legislation.

In Irwin Toy Ltd. v. Quebec (Attorney General), 1989 87 (SCC), [1989] 1 S.C.R. 927, Dickson C.J. and Lamer and Wilson JJ. explained that in performing a s. 1 analysis the Court should measure state action against a varying standard which would be dependent upon the context in which it was invoked. There it was stated at pp. 993-94:

When striking a balance between the claims of competing groups, the choice of means, like the choice of ends, frequently will require an assessment of conflicting scientific evidence and differing justified demands on scarce resources. Democratic institutions are meant to let us all share in the responsibility for these difficult choices. Thus, as courts review the results of the legislature's deliberations, particularly with respect to the protection of vulnerable groups, they must be mindful of the legislature's representative function. . . .

In other cases, however, rather than mediating between different groups, the government is best characterized as the singular antagonist of the individual whose right has been infringed. . . . [W]henever the government's purpose relates to maintaining the authority and impartiality of the judicial system, the courts can assess with some certainty whether the "least drastic means" for achieving the purpose have been chosen, especially given their accumulated experience in dealing with such question . . . .

The policy rationale set out in Irwin Toy cannot be automatically transferred to the consideration of human rights legislation which seeks to regulate the behaviour of private actors. A challenge brought under the Charter to legislation enacted by the state obviously affects a state interest. In those circumstances, due deference must be given to the actions of the state manifested by the legislation which is under attack. Consideration must be given to the mischief or evil the legislation was designed to control or eradicate.

On the other hand, in the realm of human rights legislation, the challenge will be to the actions of a private party such as an employer or landlord who, it is alleged, has adopted a discriminatory practice. The actions of private parties will often relate to private aims such as increasing the profit or efficiency of a business. Praiseworthy as those aims may be in themselves, no deference should be accorded to them once a complainant has demonstrated a prima facie case of discrimination. To adopt a deferential attitude to such private aims would undermine the professed goal of human rights legislation to guarantee the rights of minority groups, women and individuals against arbitrary and abusive treatment. Legislation aimed at abolishing or reducing discrimination should be given a liberal and generous reading. It follows that any legislated defence to acts of discrimination should be construed narrowly as was done in Brossard, supra.

Nonetheless, the similarity in wording and in function that exists between s. 1 of the Charter and s. 11.1 of the IRPA does indicate that Charter cases may assist in the development of the test for determining whether a defence under s. 11.1 of the IRPA is reasonable and justified. All the parties to this appeal take the position that the established structure of the s. 1 Charter analysis should be utilized when a s. 11.1 defence is being considered. I would agree but with some words of caution and restraint. First the Oakes model is only appropriate if it is applied without any trace of deference to a private defendant such as the employer or landlord. Secondly, only with a large measure of flexibility and due regard to the context should it be applied to the regulation of private relationships. The inquiry into what is reasonable and justifiable within the meaning of s. 11.1 should not be rigidly constrained by the formal categories set out in the Oakes test. The courts below, in my view, applied the correct legal standard in considering the defence put forward pursuant to s. 11.1 of the IRPA. Before applying the Oakes test to the case at bar, some other arguments must be considered.

What Deference Should Be Paid to Findings of Fact of the Commission Pursuant to the IRPA?

The appellant contends that neither the Court of Appeal nor this Court should interfere with the findings of fact made by the Board of Inquiry or the Court of Queen's Bench. In support of this view, reliance is placed on the common law principle of curial deference to findings of fact made by a court of first instance. In Stein v. The Ship "Kathy K", 1975 146 (SCC), [1976] 2 S.C.R. 802, Ritchie J. wrote at p. 808:

These authorities are not to be taken as meaning that the findings of fact made at trial are immutable, but rather that they are not to be reversed unless it can be established that the learned trial judge made some palpable and overriding error which affected his assessment of the facts.

See also N.V. Bocimar S.A. v. Century Insurance Co. of Canada, 1987 68 (SCC), [1987] 1 S.C.R. 1247 and Klimashewski v. Klimashewski Estate, 1987 13 (SCC), [1987] 2 S.C.R. 754.

The common law position set out above is echoed in r. 518 of the Alberta Rules of Court, Alta. Reg. 390/68, which restricts the Court of Appeal's jurisdiction to review trial courts on questions of law or inferences drawn from the facts. Following this rule, the appellate court may not interfere with findings of fact except in cases in which the trier of fact has made a palpable and overriding error.

The principle of deference to findings of fact made at first instance has been to a large extent adopted in reviewing the decisions of administrative tribunals, although the standard of review of decisions made by administrative bodies will always be governed by their empowering legislation. Where the legislature has enacted a privative clause restricting review it has been held by this Court that appellate courts must defer to a tribunal's finding of fact. See: National Corn Growers Assn. v. Canada (Import Tribunal), 1990 49 (SCC), [1990] 2 S.C.R. 1324; Lester (W.W.) (1978) Ltd. v. United Association of Journeymen and Apprentices of the Plumbing and Pipefitting Industry, Local 740, 1990 22 (SCC), [1990] 3 S.C.R. 644, and Gendron v. Supply and Services Union of the Public Service Alliance of Canada, Local 50057, 1990 110 (SCC), [1990] 1 S.C.R. 1298.

In Bell Canada v. Canada (Canadian Radio-television and Telecommunications Commission), 1989 67 (SCC), [1989] 1 S.C.R. 1722, at p. 1746, this Court further recognized that curial deference must be given to findings made within the field of specialized knowledge of an administrative decision-maker.

However, the situation is different where there is neither specialized skill and knowledge exercised by an administrative decision-maker nor a statutory restriction imposed upon the Court's review of those decisions. Here the IRPA has clearly indicated that a very broad standard of review would be appropriate to decisions rendered by a Board of Inquiry under the Act. Section 33(2) provides that a party may appeal a decision of the Board on questions of fact or mixed fact and law with the leave of a judge of the Court of Queen's Bench. In the case at bar, Murray J. granted leave for an appeal on the facts. On a plain reading of the IRPA, it is clear that the legislature specifically intended that appellate courts should examine the evidence anew and, if deemed appropriate, make their own findings of fact. Under this Act, no particular deference is owed by the Court of Appeal to the findings of the initial trier of fact. This Court possesses the same statutory jurisdiction as the Court of Appeal.

In support of this position I would note that the provision for appeal in the Ontario Human Rights Code (formerly R.S.O. 1970, c. 318) which this Court considered in Etobicoke, supra. The statutory basis for an appeal from an Ontario Board of Inquiry is found in s. 42(3) of the Code, R.S.O. 1990, c. H.19. It provides for an appeal on any question of law or fact and states that the Court may substitute its opinion for that of the Board. In Etobicoke, McIntyre J. held that this section (then s. 14d(4)) granted an appellate court broader powers to review findings of the trier of fact than exist at common law. The wording of s. 42(3) of the Ontario Code is more explicit than that found in s. 33(2) of the IRPA. However, the import of the two sections must be the same, as the right to an appeal on questions of fact would be meaningless if the appellate court were not empowered to substitute its own opinion for that of the Board. Nor is this a situation in which the administrative decision-maker possesses a specialized expertise which would merit curial deference. It can be seen that the IRPA grants the Court of Appeal and thus this Court the jurisdiction to make findings of fact based on a review of the evidence on the record, without deferring to the conclusions drawn by the Board of Inquiry.

The question remains whether this Court may disturb findings of fact made by a judge of the Court of Queen's Bench. It must be remembered that Murray J. did not hear any viva voce testimony. He did not, nor was it necessary for him to, make any assessment as to credibility. Indeed, almost all of the evidence in this case is so-called "social fact" evidence. Rather, the judge reassessed the evidence on the basis of the transcript of the Board of Inquiry hearing. As a result the Court of Appeal of Alberta and this Court stand in exactly the same position as the judge of the Court of Queen's Bench with precisely the same record to consider. In the circumstances the policy reasons in favour of deferring to the findings of fact of a trial court simply do not apply. There is no basis or justification for the position that this Court should be constrained by the conclusions reached by either the Chairman of the Board of Inquiry or the Judge of the Court of Queen's Bench on s. 11.1 evidence. One further preliminary matter must be dealt with.

The Relevance of McKinney to the Case at Bar

The Alberta Court of Appeal, correctly in my view, concluded that a private rights violator should not receive the "benefit of the doubt" extended to legislatures in Charter cases. However, that court went on to hold that the decision in McKinney determined the result of this case. With respect, I cannot agree.

Undoubtedly, this appeal raises a question that is closely related to that which was determined in McKinney. Namely, do the objectives of promoting tenure, academic renewal, planning and resource management and retirement with dignity justify the placing of limits by way of age on the substantive rights to equal treatment? In McKinney La Forest J. writing for the majority determined that the defendant universities were not government for the purposes of the Charter. Nevertheless, he proceeded to examine the institutions' retirement policies on the assumption that they were government actors for the purposes of considering those policies in light of s. 15 and s. 1. Although this section of the judgment is in obiter it clearly disposes of the constitutionality of such policies in government settings. The issue remains however, does McKinney fully decide whether a mandatory retirement policy in a private employment setting can be justified pursuant to the provisions of s. 11.1 of the IRPA?

There can be no doubt that the evidence presented in McKinney was in part identical and in part remarkably similar to that presented in this case. As a result, the treatment of the evidence by La Forest J. in the McKinney case can provide some guidance to the resolution of this one. Yet the decision cannot simply be transferred to this case as suggested by the University of Alberta. La Forest J. invoked, at least in part, the deferential standard of s. 1 analysis when he considered the arguments pertaining to minimal impairment. That same deference cannot be applicable in the case at bar, which is concerned with the employment arrangements of private parties. When weighing the arguments on the issue of minimal impairment La Forest J. in the McKinney case framed the question as: "whether the government had a reasonable basis for concluding that it impaired the relevant right as little as possible . . ." (at p. 286). To frame the question in this way imposes a significantly lower burden of proof on the defendant than that required by s. 11.1. That section requires a defendant to prove that the discriminatory policy, viewed objectively, constitutes only a minimal impairment of the right. As a result, the evidence adduced in this case must be considered in light of the test appropriate to the implementation of s. 11.1 of the IRPA even though the evidence presented is so very similar to that put forward in McKinney.

Application of s. 11.1 to the Evidence

A party seeking to establish a defence pursuant to s. 11.1 must demonstrate that the discriminatory practice furthers a substantial objective and is proportional to that objective. In considering the defence, no deference should be given to the policy choice of the defendant as would be the case in the s. 1 analysis of a social policy. However, other factors may well be relevant when the s. 11.1 test is applied in this regulatory context. The role of universities in our society as self-governing centres of learning, research and teaching safeguarded by academic freedom is unique. The courts have respected this and over the years have been very cautious in intervening in university affairs. The determination of the matters at stake involves the weighing of a number of factors not susceptible of exact measurement. The nurturing of academic freedom and the ensuring of faculty renewal are most delicate matters as they bear on the governance of persons and must take into account both group and individual responses. The subject matter does not lend itself to a single clear-cut answer as to the proportionality between the burden of the discrimination complained of and the objectives which are sought. The evidence and assessment of competent and experienced university administrators who are entrusted with the responsibility for fulfilling the role of the university in society will certainly be of particular significance. Furthermore, in this case the practice which they uphold and is now attacked was initially imposed by the terms of a collective agreement. The employees through their faculty association agreed with the employer to incorporate mandatory retirement at age 65 into the collective agreement. This term was the result of freely conducted negotiations. It is significant that the parties originally agreed to this term before the University's pension legislation was amended to require retirement at age 65.

The Relevance of a Collective Bargaining Agreement to a s. 11.1 Defence

Generally speaking, parties may not contract out of a human rights statute. This Court has enforced that general rule in the context of age discrimination; see Etobicoke, supra, and Winnipeg School Division No. 1 v. Craton, 1985 48 (SCC), [1985] 2 S.C.R. 150. Yet this Court has accepted the existence of a collective bargaining agreement as a factor to be taken into consideration in a s. 1 analysis. See McKinney, per La Forest J. at p. 287 and Cory J. at pp. 447-48. The Craton rule resulted from the concern that there may be a great discrepancy in bargaining power between the person contracting out of human rights legislation and the party receiving the benefit of that term. Unequal bargaining power may be utilized to force a party to agree to a contract which undermines the very purpose of human rights legislation. That type of imbalance may typically arise in an employment relationship.

Yet labour codes, providing for the formation of unions and collective bargaining, are specifically designed to overcome or compensate any imbalance in bargaining power. With these statutory safeguards in place collective agreements take on a new and important significance. They can represent a carefully constructed, fairly negotiated bargain between employer and employees. It follows that a collective agreement cannot be readily dismissed from consideration.

There remains a danger with collective bargaining agreements that must be considered. Unions must act in the best interests of all their members. Yet, a union could choose to negotiate terms which are solely for the benefit of the majority of the membership at the expense of the interests of the minority. Courts must be alert to this possibility. For example, any collective agreement which includes provisions adversely affecting a minority of individuals on the basis of their race or religion should be closely scrutinized. Similarly, a collective agreement term which discriminates against women in the workplace in situations where they do not constitute a clear voting majority of the bargaining unit should not be readily accepted as evidence of the reasonableness of a discriminatory practice.

Nonetheless, a collective agreement may very well provide evidence of the reasonableness of a practice which appears on its face to be discriminatory. However, certain precautions must be taken. For example, it should be demonstrated that the collective agreement was freely negotiated by parties with relatively equal bargaining positions. It should be shown that the agreement did not discriminate unfairly against minorities. In essence, there must be substantive fairness demonstrated by the agreement itself. As well, when considering the weight to be attached to the terms of a collective bargaining agreement, a court must remain flexible and be aware that it is always difficult to obtain complete equality between all the groups forming part of a bargaining unit.

Generally, in assessing the weight to be attached to a collective bargaining agreement as a factor in demonstrating that the discrimination was reasonable and justified, my reasons in McKinney, supra, pp. 447-48, may be apposite:

. . . in the course of negotiating a collective bargaining agreement, it may become apparent that the union membership is overwhelmingly in favour of an agreement that embraces compulsory retirement as part of the consideration for obtaining higher wages at an earlier age--an age when houses must be bought and children raised and educated. That is to say, at a time when the need for family funds is at the highest.

It is often the case that, before a collective bargaining agreement is ratified, the union members will have received very careful advice concerning its terms and their significance not only from union officials, but also from skilled economists and lawyers. The collective agreement represents a total package balancing many factors and interests. It represents the considered opinion of its members that it would be in their best interests to accept the proposed contract. Bargains struck whereby higher wages are paid at an earlier age in exchange for mandatory retirement at a fixed and certain age, may well confer a very real benefit upon the worker and not in any way affect his or her basic dignity or sense of worth. If such contracts should be found to be invalid, it would attack the very foundations of collective bargaining and might well put in jeopardy some of the hard won rights of labour.

The collective agreement reflects the decision of intelligent adults, based upon sound advice, that it is in the best interest of themselves and their families to accept a higher wage settlement for the present and near future in exchange for agreeing to a fixed and certain date for retirement. In those circumstances, it would be unseemly and unfortunate for a court to say to a union worker that, although this carefully made decision is in the best interest of you and your family, you are not going to be permitted to enter into this contract. It is a position that I would find unacceptable.

In this case the collective bargaining agreement authorizing compulsory retirement can properly be taken into account as a factor in considering whether the discriminatory policy can be found to be reasonably justified. Here, the term of the collective agreement relating to compulsory retirement will apply to every member of the faculty association. The term was reasonable and appropriate in the circumstances. First, age differs from other grounds of discrimination since everyone of no matter what religion, colour, social origin, nationality or gender becomes older with the passage of time. Eventual retirement with pension security is a matter of concern for all employers in every field. As a consequence, the mandatory retirement term will have an impact on all those members of the faculty association who remain employees of the University until the end of their career. Second, the union did not negotiate the term in a vacuum, but rather, in the context of a system of tenure which protects all members of faculty from dismissal without just cause, and provides a pension scheme assuring the financial security of all retiring members of faculty. It is safe to assume that the terms of the collective agreement pertaining to compulsory retirement were not the manifestation of an abuse of its power by the employer University. Rather, they represent a carefully considered agreement that was negotiated with the best interests of all members of the faculty association in mind.

In determining whether the practice of mandatory retirement constitutes reasonable justification for age discrimination the fact that this option was chosen by way of an agreement between an employer and the faculty association should be recognized. Members of a faculty association may, even more than some rank and file members of other unions, appreciate the significance and consequences of a mandatory retirement policy for themselves, for the faculty association and for the university. Such an agreement should not be lightly disregarded. In this case the University may properly rely on the collective agreement as evidence which tends to confirm the reasonableness of the mandatory retirement policy. It is a factor which must be kept in mind when applying the Oakes model test to the facts of this case.

Are the Objectives of the Policy Pressing and Substantial?

In the proceedings below, the objectives of mandatory retirement were stated to be: the preservation of tenure, the promotion of academic renewal, the facilitation of planning and resource management and the protection of "retirement with dignity" for faculty members. They are subsumed in the objectives put forward to justify mandatory retirement by the defendant universities in McKinney which the University of Alberta adopted in argument before this Court. These were (at p. 281):

(1) To enhance and maintain a university's capacity to seek and maintain excellence by permitting flexibility in resource allocation and faculty renewal, and

(2) To preserve a university's academic freedom and the collegial form of association by minimizing distinctive modes of performance evaluation.

In McKinney this Court concluded that these objectives were of sufficient significance to justify the limitation of a constitutional right to equality. This conclusion must also apply to these same objectives put forward by the University of Alberta in the case at bar.

Is the Policy Proportional to the Objective?

According to the Oakes model, the defendant must demonstrate that the policy is rationally connected to the objective, minimally impairs the right and is proportional in its effects.

(a) Rational Connection

The University has submitted that the objectives outlined above as the rationale for negotiating a compulsory retirement term in the collective agreement all relate to the specific labour requirements in the academic market. As discussed in McKinney, the traditional career pattern of university faculty involves a deferred compensation scheme with a comparatively low entry level salary which increases with seniority (see Michael Krashinsky, "The Case for Eliminating Mandatory Retirement: Why Economics and Human Rights Need Not Conflict" (1988), XIV Canadian Public Policy 40, at p. 47). Within the first few years of appointment, junior faculty members have the opportunity to apply for tenured status. If accepted, they will be protected from dismissal except on grounds of incompetence or misconduct. Typically, as in the case of the University of Alberta, the power to dismiss professors on this basis is rarely exercised. The University dismissed only 10 professors over the years 1974-1984.

In the 1960s and early '70s, Canadian universities experienced dramatic growth which necessitated the hiring of a large number of new professors. Since then, financial constraints have prevented the expansion of our academic institutions. As a result, there exists a "bulge" in the age distribution of professors. A disproportionate number are concentrated among those in the latter part of their careers.

These circumstances support a finding that the impugned retirement practice is rationally connected to the objectives cited. The retirement of faculty members at the age of 65 ensures that the University may readily predict the rate at which employees will leave the institution. By requiring the retirement of all professors on reaching that age, positions are opened for new faculty. In accepting this argument, I do not wish to appear to have fallen victim to the "lump of labour" fallacy, which is the assumption that the total number of jobs in an economy is fixed. (See Frank Reid, Economic Aspects of Mandatory Retirement (1987), at p. 9.) In the economy as a whole, retirements are not essential to open new positions for young workers as the quantity of jobs is not fixed. In the current academic market, however, funding pressures have prevented the expansion of the total number of teaching and research positions. The presence of the "bulge" has concentrated a large proportion of jobs in one age group. In this particular context, mandatory retirement is an essential tool to provide a degree of inter-generational equity.

Mandatory retirement also allows the University to renew its faculty by introducing younger members who may bring new perspectives to their disciplines. The Report of the Bovey Commission on the Future Development of the Universities of Ontario (1984) identified academic renewal as a crucial goal for universities. This report recognizes that it is only through new appointments that universities can improve the ratio of women faculty members from the current numbers, which are extremely low (at pp. 21-22). The same Report estimates that without mandatory retirement, the numbers of new faculty hired in Ontario would have declined by approximately 300 from an estimate of 750 by the end of the 1980s (at p. 22). Mandatory retirement provides a means of remedying the twofold problem of limited funding and a "bulge" in the age distribution of professors.

As well, the policy of mandatory retirement supports the existence of a tenure system which creates barriers to the dismissal of faculty members thereby enhancing academic independence. Certainly, the University of Alberta has in place a system designed to remove demonstrably incompetent professors, but the administration has taken such steps most infrequently. It might be suggested that any necessary trimming of faculty could be more fairly accomplished by an active exercise of the power to dismiss for incompetence. Yet this argument fails to recognize the difficulties inherent in assessing either the quality of a professor's work or her productivity (see Robert T. Blackburn and Janet H. Lawrence, "Aging and the Quality of Faculty Job Performance" (1986), 23 Review of Educational Research 265, at p. 271).

Typically, teaching ability is only monitored by student evaluation. Research can only be measured by the work published. A greater risk of dismissal could well discourage individual professors from choosing to concentrate on teaching skills. There are many obstacles that must be overcome in order to make a fair objective assessment of the quality of work produced. Although these difficulties are faced by universities when tenure is first considered, a policy which increased the frequency with which these difficult assessments were made would be cumbersome, threatening to faculty and disruptive to the university. It would undermine the sense of security which is so essential for the free exercise of academic inquiry.

On these facts, the University of Alberta has demonstrated that the practice of compulsory retirement is rationally connected to its stated objectives.

(b) Minimal Impairment

There are only a very limited number of alternatives to mandatory retirement which would satisfy the University's legitimate objectives. The University could increase the rigour with which it enforces its procedures to dismiss professors for cause. While this might not necessarily undermine the tenure system, which is designed to protect professors from political interference and not from their own incompetence, it could make the security of tenure easier to circumvent. Also, as noted earlier, there are serious difficulties inherent in any attempt to measure the competence and productivity of professors.

Another possible alternative would be to encourage early retirement. This option would not contravene any of the University's objectives. Yet an optional programme would not necessarily achieve the numbers or certainty of vacancies among faculty that are required for renewal. Thus, this method would be an inadequate substitute for mandatory retirement.

In the University setting, the policy of mandatory retirement withstands the minimal impairment test. No obvious alternative policy exists which would achieve the same results without restricting the individual rights of faculty members. The fact that the practice is the result of a fair and freely negotiated collective agreement supports the conclusion that the practice is reasonable and justifiable within the meaning of s. 11.1.

(c) Proportionality of Effects

Lastly, the University of Alberta must demonstrate that the effects of the impugned practice are proportional to the objectives which it is intended to serve. Without question, mandatory retirement constitutes prima facie discrimination. Those faculty members who, like Professor Dickason, would choose to continue working past the age of 65 will suffer from the denial of this opportunity. Yet these individuals have the benefit of secure and reasonable pensions. The policy forms part of a collective agreement which benefits, as a whole, the academic staff of the University. In this context, the effects of the prima facie discrimination are proportional to the legitimate objectives served.

Disposition

The University has shown that the impugned practice of mandatory retirement is reasonable and justifiable within the meaning of s. 11.1 of the IRPA. In the result, the appeal is dismissed. As this proceeding is in the nature of a test case, there shall be no order as to costs.

L'Heureux-Dubé J.

The reasons of L'Heureux-Dubé and McLachlin JJ. were delivered by

L'Heureux‑Dubé J. (dissenting) -- This appeal once again raises the issue of mandatory retirement at age 65, this time, however, in the context of the policy of a private employer, the respondent University of Alberta (the "University"). That policy is being challenged by the appellant professor, Dr. Olive Dickason, under the Alberta Individual's Rights Protection Act, R.S.A. 1980, c. I‑2 (the "IRPA"). Specifically, the issue is whether the University, through a collective agreement with its academic staff, can retire Dr. Dickason at age 65, despite the provisions of s. 7 of the IRPA, which reads:

7(1) No employer or person acting on behalf of an employer shall

(a) refuse to employ or refuse to continue to employ any person, or

(b) discriminate against any person with regard to employment or any term or condition of employment,

because of the race, religious beliefs, colour, gender, physical disability, mental disability, marital status, age, ancestry or place of origin of that person or any other person. [Emphasis added.]

The University relies upon s. 11.1 of the IRPA, which provides a general defence for discriminatory acts:

11.1 A contravention of this Act shall be deemed not to have occurred if the person who is alleged to have contravened the Act shows that the alleged contravention was reasonable and justifiable in the circumstances.

I have considered the opinion of my colleague Justice Cory and, with respect, I cannot agree for the following reasons.

Facts

Given their particular relevance to the issue, I will briefly review the facts surrounding this appeal, with an emphasis on the chronology of events leading up to Dr. Dickason's retirement.

Dr. Dickason joined the academic staff at the University of Alberta on June 24, 1975. At the time, while s. 6 of the Individual's Rights Protection Act, S.A. 1972, c. 2, then in force prohibited discrimination in employment on the basis of various factors, including age (subject to a defence of bona fide occupational qualification), age was defined as between 45 and 65 years of age, for the purposes of the Act.

A little over a year later, Dr. Dickason became an assistant professor and signed a contract agreeing to be bound by the Academic Staff Agreement ("the Faculty Agreement") between the University and the Association of Academic Staff of the University of Alberta ("the Faculty Association"). Article 18.01 of the Agreement, which came into effect on July 1, 1976, established a "normal retirement date" for academic staff at age 65:

18.01 The normal retirement date for a staff member will be 30 June following the attainment of age sixty‑five (65) or 31 August for those staff members who were employed in anticipation of an end of year date, being 31 August.

In 1978, the Universities Academic Pension Act, S.A. 1978, c. 36 (the "UAPA") came into effect. Section 9 provided that academic staff would retire on the "normal retirement date", which was defined as age 65. The Faculty Agreement which bound Dr. Dickason was amended that same year to reflect its compliance with this legislation, although this did not involve any material change in the University's policy.

On November 2, 1979, Dr. Dickason was awarded tenure under the terms of the Faculty Agreement between the University and the Faculty Association.

In 1985, various amendments to the IRPA were enacted to make it compatible with s. 15 of the Canadian Charter of Rights and Freedoms ("the Charter"). On June 5, 1985, four days after Dr. Dickason became a full professor at the University, these amendments came into force. Section 11.1 was added to the IRPA, and s. 38(a) of the IRPA was replaced with the present provision, which defines age as "18 years of age or older".

On June 30, 1985, Dr. Dickason, having reached her 65th birthday earlier in the year, was retired pursuant to art. 18.01 of the Faculty Agreement. Upon her retirement, she automatically acquired the status of professor emeritus. On July 3, 1985, Dr. Dickason filed a complaint with the Alberta Human Rights Commission, alleging that the University had contravened s. 7 of the IRPA by discriminating against her on the basis of her age. A Board of Inquiry was appointed on April 1, 1986.

It should be noted at this point that, at the time of Dr. Dickason's forced retirement, the UAPA of 1978 was still in force. It was not until November 1, 1985, that it was superseded by the Universities Academic Pension Plan Act, S.A. 1985, c. U‑6.1 (the "UAPPA"), in which references to "normal retirement age" were replaced with references to "normal pensionable age". Notwithstanding this chronology, the parties to this appeal agreed at the outset that they would argue the case as though the UAPPA had already come into force on the date of Dr. Dickason's retirement, and the case proceeded before the Board of Inquiry, as well as before the courts, including ours, on this understanding. The argument that the University could have perhaps made under the former UAPA was therefore abandoned.

The Board of Inquiry appointed to hear Dr. Dickason's complaint decided in her favour and ordered that she be re‑instated at the University: (1987), 1987 8520 (AB HRC), 9 C.H.R.R. D/4468. On appeal to the Court of Queen's Bench, the decision was upheld ((1988), 1988 3530 (AB KB), 62 Alta. L.R. (2d) 209, 91 A.R. 350, 9 C.H.R.R. D/5403), but a further appeal to the Court of Appeal was allowed: (1991), 1991 ABCA 208, 81 Alta. L.R. (2d) 393, 117 A.R. 11, 2 W.A.C. 11, 83 D.L.R. (4th) 1, [1991] 6 W.W.R. 377.

Board of Inquiry (1987), 1987 8520 (AB HRC), 9 C.H.R.R. D/4468

In his decision, the Chairperson, K. J. Alyluia, first emphasised the special role of human rights legislation as "fundamental law", and noted the similarity between the equality provisions at s. 7 of the IRPA and at s. 15(1) of the Canadian Charter of Rights and Freedoms. He concluded that the provisions of the collective agreement could not be used to circumvent human rights legislation.

The Chairperson then embarked on an extensive review of the evidence, canvassing the four points argued by the University in support of mandatory retirement at age 65, and rejecting each of them in turn. With respect to the argument that professors should be allowed to "retire with dignity", he found that the University had not presented evidence showing a decline in ability "which would render the vast majority of academic staff unable to function effectively as they approach or reach 65" (p. D/4472). Accordingly, it would be "unwise, unfair and now illegal" to penalize all professors who reach age 65, particularly when incompetent professors of any age could be dismissed for cause through procedures already in place at the University.

On the need for institutional planning, the Chairperson concluded that there were so many uncertainties in the planning process that the removal of the "constant" of mandatory retirement would not have a great effect.

He next turned to the University's "academic renewal" argument, to the effect that it was necessary to replace older academics with young professors who had new ideas and heightened creativity. While conceding that intellectual renewal was necessary for a healthy academic atmosphere, the Chairperson concluded that, in the face of competing evidence, the proposition of originality as a by-product of youth had not been proved. As well, in his view, the increasing popularity of flexible retirement made mandatory retirement less crucial to academic renewal. Based on evidence collected from institutions which had abolished mandatory retirement, the Chairperson held that, after a five to ten year transitional period, the flow of replacement returns to its initial level. He therefore disagreed with the contrary conclusion by the trial judge in Harrison v. University of British Columbia (1986), 1986 1297 (BC SC), 30 D.L.R. (4th) 206 (B.C.S.C.), in these terms (at p. D/4475):

The Court concedes that mandatory retirement poses disadvantages to the fit and willing person who must now retire at a fixed age but balances these factors with youth unemployment and decides for the latter, one of the facts being the economic advantage enjoyed by older workers with pension benefits versus youth with no similar financial resources. . . . Again, with respect I cannot concur with these conclusions based on the evidence presented. Admitting that there may be a reduction in the number of young academics entering into employment if mandatory retirement were abolished, the evidence indicates that the reductions would not be large.

Finally, again drawing on data from universities where mandatory retirement had been abolished, the Chairperson rejected the University's argument that the abolition of mandatory retirement threatened the tenure system.

Turning to the applicable law, the Chairperson found that the onus of proof on the University, in the face of its prima facie violation of human rights, was akin to that established by this Court in R. v. Oakes, 1986 46 (SCC), [1986] 1 S.C.R. 103, under s. 1 of the Charter. While he agreed that there was a rational connection between mandatory retirement and the University's aims, he concluded that "none of the links is vital to their continuance" (p. D/4478); all the aims could be met without forced retirement. In his view, mandatory retirement impairs rights more than minimally, and he concluded that "the effect of the policy ‑‑ namely loss of tenured professor status with all the rights and privileges flowing therefrom, including of course the right to continue working ‑‑ is disproportionate to its objectives" (pp. D/4478‑79).

The Board of Inquiry accordingly allowed Dr. Dickason's complaint and ordered the University to reinstate her.

Court of Queen's Bench (1988), 1988 3530 (AB KB), 62 Alta L.R. (2d) 209

On appeal to the Court of Queen's Bench, Murray J. began by considering the onus of proof under s. 11.1 of the IRPA and the principles set out in Charter cases. He observed that the IRPA enjoys a special status, "not quite constitutional in character", but as a "fundamental law" which expresses public policy (at p. 222). While Supreme Court decisions on the onus of proof under s. 1 of the Charter set out appropriate guidelines, the chambers judge concluded that the Board erred in over‑emphasising the link between s. 15(1) of the Charter and s. 7 of the IRPA. However, in his opinion, this error was not fatal to the Board's decision.

Agreeing that there was a prima facie violation of s. 7 of the IRPA, the chambers judge then turned to the Oakes test to determine whether the respondent University's policy regarding mandatory retirement could be saved (at p. 230):

We are here concerned with human rights legislation enacted at the provincial constitutional level. What better way to ensure proper protection of those rights than to use as a guide the criteria formulated by the highest court in the land to scrutinize actions which violate a comparable protected right under the Charter. One of the purposes in each case is to determine if on balance the limiting action is reasonable and can be justified having regard to the objectives it seeks to achieve. I am of the opinion that the criteria which must be satisfied to save a government law by virtue of s. 1 of the Charter can and should be used as a guide in examining an otherwise prohibited action under s. 11.1 of the I.R.P.A., at least to the extent that those considerations are applicable.

After a review of the case law in other jurisdictions as well as the evidence, he applied the civil standard of "proof by a preponderance of probability" from Oakes and held that, although "certain of the objectives are valid . . . [and] of a pressing and substantial nature", mandatory retirement had no rational connection to the University's aims (at p. 247). The appeal was accordingly dismissed. The issue of the constitutionality of s. 11.1 of the IRPA raised by the appellant was not discussed. Dr. Dickason's appeal on the question of damages was also dismissed.

Court of Appeal (1991), 1991 ABCA 208, 81 Alta. L.R. (2d) 393

The Court of Appeal (Lieberman, Belzil and Foisy JJ.A.) allowed the University's appeal on the basis of this Court's decision in McKinney v. University of Guelph, 1990 60 (SCC), [1990] 3 S.C.R. 229. Although the Court of Appeal acknowledged that this case did not involve an application of the Canadian Charter, it held that McKinney was binding insofar as, in that decision, La Forest J. determined that mandatory retirement in a university context was justifiable under s. 1.

In the court's view, the appeal raised the question of the stringency of the test applied under s. 11.1 of the IRPA, that is, the burden which the University had to discharge in order to have its mandatory retirement policy upheld. The Court of Appeal agreed with the chambers judge that "the university, as a private employer, would not be entitled to the deference shown by the courts to legislators in a s. 1 of the Charter analysis" (p. 398). However, in the court's opinion, La Forest J.'s conclusion on the justifiability of mandatory retirement in Canadian universities was binding and rendered moot any question on the evidentiary burden under s. 11.1 (at pp. 399‑400):

It is now apparent from the decision of the Supreme Court of Canada in McKinney that the learned chambers judge was in error in his assessment of the evidence and in his conclusions. Either he applied too stringent a test or he misapprehended the evidence.

These basic social and economic conclusions of the Supreme Court of Canada are judicially approved criteria and policy guidelines binding on this court, whether in the context of a Charter analysis under s. 1 or an analysis under s. 11.1 of the Individual's Rights Protection Act.

The court further held that the issues raised in the appeal were identical to those in McKinney and should not be decided on a case‑by‑case basis "unless more than minor factual differences exist" (p. 400).

Issues

Since the respondent University has conceded that its policy of mandatory retirement at age 65 for its academic staff is discriminatory under s. 7 of the IRPA, the central issue in this case is whether that policy is "reasonable and justifiable" under s. 11.1. Before attempting to answer that question, however, several legal issues must be addressed:

1.The role of the Board of Inquiry;

2.Section 11.1 of the IRPA;

3.The impact of the collective agreement.

1. The Role of the Board of Inquiry

After extensive review of the evidence, which included viva voce testimony from twenty witnesses (see the appendix at the end of the reasons for judgment of Murray J., not included in the reported version), the Board concluded that, though there was some connection between the policy and the University's goals, mandatory retirement was not vital to achieving the stated objectives. The effect of the policy was disproportionate to its objectives. The differing conclusions by this Court in Ontario Human Rights Commission v. Borough of Etobicoke, 1982 15 (SCC), [1982] 1 S.C.R. 202, and Saskatchewan (Human Rights Commission) v. Saskatoon (City), 1989 18 (SCC), [1989] 2 S.C.R. 1297, two cases arising in the context of mandatory retirement for firefighters and involving nearly identical provisions in human rights legislation, show that discrimination cases such as this are fact‑driven, and that outcome will vary with the evidence submitted. The recent decision by this Court in McKinney, supra, has cast a long shadow over these proceedings. Hence, it is crucial that we establish at the outset why, aside from the fact that it was based on "government" action challenged under the Charter and this involves private action challenged under the IRPA, the outcome in McKinney is not necessarily determinative here.

The basis for review of the Board of Inquiry's decision in this case is s. 33 of the IRPA, the relevant portions of which read as follows:

33(1) A party to a proceeding before a board of inquiry may appeal the order of the board to the Court of Queen's Bench by originating notice filed with the clerk of the Court of the judicial district in which the inquiry was held.

(2) If the grounds of appeal consist wholly or partly of a question of fact or mixed fact and law, no appeal lies without the leave of a judge of the Court of Queen's Bench.

(4) The judge may, after taking into consideration any matters that he considers appropriate, grant leave to appeal under subsection (2).

(6) The Court may confirm, reverse or vary the order of the board and may make any order that the board can make under section 31.

My colleague Cory J. concludes, after a brief review of the relevant principles, that, under the IRPA, "no particular deference" is owed by this Court either to the findings of the Board or to the findings of the Court of Queen's Bench. On the first point, he relies on the "plain meaning" of s. 33(2) of the IRPA, which empowers the Court of Queen's Bench to substitute its opinion for that of the Board where the appeal is granted on mixed issues of law and fact (as was the case here). On the second point, he notes that Murray J. did not hear any viva voce witnesses and, in fact, relied on the same written transcript and evidence as is before the Court now. Hence, in his opinion, "[t]here is no basis or justification for the position that this Court should be constrained by the conclusions reached by either the Chairman of the Board of Inquiry or the Judge of the Court of Queen's Bench on s. 11.1 evidence" (p. 000).

With respect to the findings of the Board of Inquiry, I must of course agree that there is a wide scope of review afforded to the Court of Queen's Bench under the IRPA, given that the Board's findings are not protected by a privative clause and that human rights tribunals are not as "specialized" as some other administrative bodies such as labour boards. In my opinion, however, a sweeping dismissal of the role of the Board overlooks the very rationale for the rule of curial deference to findings of fact made by courts of first instance and administrative tribunals.

The principle of curial deference is not simply based upon what appellate courts may do, acting within their jurisdiction. It is based on the recognition of the "signal advantage" enjoyed by lower courts and quasi‑judicial bodies which see and hear witnesses. The classic statement on this subject is found in Clarke v. Edinburgh and District Tramways Co., [1919] S.C. (H.L.) 35, where Lord Shaw of Dunfermline states (at p. 36):

When a Judge hears and sees witnesses and makes a conclusion or inference with regard to what is the weight on balance of their evidence, that judgment is entitled to great respect, and that quite irrespective of whether the Judge makes any observation with regard to credibility or not. . . . In Courts of justice in the ordinary case things are much more evenly divided; witnesses without any conscious bias towards a conclusion may have in their demeanour, in their manner, in their hesitation, in the nuance of their expressions, in even the turns of the eyelid, left an impression upon the man who saw and heard them which can never be reproduced in the printed page. What in such circumstances, thus psychologically put, is the duty of an appellate Court? In my opinion, the duty of an appellate Court in those circumstances is for each Judge of it to put to himself, as I now do in this case, the question, Am I ‑‑ who sit here without those advantages, sometimes broad and sometimes subtle, which are the privilege of the Judge who heard and tried the case ‑‑ in a position, not having those privileges, to come to a clear conclusion that the Judge who had them was plainly wrong? If I cannot be satisfied in my own mind that the Judge with those privileges was plainly wrong, then it appears to me to be my duty to defer to his judgment.

This rationale has been adopted consistently by this Court in such cases as Laurentide Motels Ltd. v. Beauport (City), 1989 81 (SCC), [1989] 1 S.C.R. 705, where a trial judge's evaluation of expert testimony was re‑instated in favour of the assessment substituted by the Court of Appeal (at p. 799):

In these circumstances, an appellate court which has neither seen nor heard the witnesses and as such is unable to assess their movements, glances, hesitations, trembling, blushing, surprise or bravado, is not in a position to substitute its opinion for that of the trial judge, who has the difficult task of separating the wheat from the chaff and looking into hearts and minds of witnesses in an attempt to discover the truth.

See also Stein v. The Ship "Kathy K", 1975 146 (SCC), [1976] 2 S.C.R. 802, at pp. 806 and 808, and Klimashewski v. Klimashewski Estate, 1987 13 (SCC), [1987] 2 S.C.R. 754, at p. 754.

Furthermore, in N.V. Bocimar S.A. v. Century Insurance Co. of Canada, 1987 68 (SCC), [1987] 1 S.C.R. 1247, this Court confirmed that the advantage which the trier of fact enjoys in evaluating the testimony of witnesses extends to experts. Le Dain J. wrote at pp. 1249‑50:

The Court of Appeal took the position that because of the nature of the evidence in this case, which consisted of expert testimony and documentary evidence, the Court, to use its own words, was "almost in the position of conducting the trial de novo and making our own assessment of the evidence." I cannot agree. The limits to the scope of appellate review of the findings of fact by a trial court, which were affirmed by this Court in Stein v. The Ship "Kathy K", 1975 146 (SCC), [1976] 2 S.C.R. 802, and other decisions, also apply in my opinion to the review of the findings of a trial court based on expert testimony, as indicated in Joseph Brant Memorial Hospital v. Koziol, 1977 6 (SCC), [1978] 1 S.C.R. 491; Schreiber Brothers Ltd. v. Currie Products Ltd., 1980 11 (SCC), [1980] 2 S.C.R. 78; and Joyce v. Yeomans, [1981] 2 All E.R. 21 (C.A.).

The Court of Appeal appears to have taken the position that it could assess the weight of the evidence in support of the facts on which the expert witnesses expressed an opinion because the evidence of those facts was before the trial court in a documentary form. The findings of the trial judge were similar to the assumed facts and the opinions in the affidavits of the expert witnesses. In coming to a different conclusion, on a balance of probabilities, from that of the trial judge with respect to some of those facts, the Court of Appeal in effect rejected the expert testimony which was based in part on those facts. It did so without having heard the expert witnesses and without being in a position to determine what their testimony would have been had the factual basis for their testimony been qualified to the extent considered necessary by the Court of Appeal. In doing so, the Court of Appeal, in my respectful opinion, erred. [Emphasis added.]

In addition to the cases cited in the foregoing passage, see also Lapointe v. Hôpital Le Gardeur, 1992 119 (SCC), [1992] 1 S.C.R. 351, at p. 358.

A general discounting of findings of fact of the Board of Inquiry also ignores the other advantages it enjoys as the preliminary trier of fact. The Board is in a privileged position with respect to the assessment of evidence, in that it sees and hears witnesses. The mandate of members of the Human Rights Commission, who preside over hearings, is always the interpretation and application of the IRPA, which necessarily involves the adjudication of policy issues and the consideration of social fact evidence. The findings of the Board on social evidence must be accorded some deference to the extent that they fall within the realm of its specific and primary mandate under the IRPA. These findings, not being protected by a privative clause under the IRPA, may be afforded less deference than findings of an administrative tribunal which has the protection of a privative clause. This, however, is only an issue of degree. See Bell Canada v. Canada (Canadian Radio-television and Telecommunications Commission), 1989 67 (SCC), [1989] 1 S.C.R. 1722, at pp. 1745‑46; National Corn Growers Assn. v. Canada (Import Tribunal), 1990 49 (SCC), [1990] 2 S.C.R. 1324, at p. 1336; Lester (W.W.) (1978) Ltd. v. United Association of Journeymen and Apprentices of the Plumbing and Pipefitting Industry, Local 740, 1990 22 (SCC), [1990] 3 S.C.R. 644, at p. 650; and the recent decision in Zurich Insurance Co. v. Ontario (Human Rights Commission), 1992 67 (SCC), [1992] 2 S.C.R. 321, at p. 338 of Sopinka J.'s reasons.

This approach has been adopted by this Court in other appeals from decisions of human rights tribunals very similar to this one. As Cory J. points out, the scope of review in this case is no greater than that in Etobicoke, supra. There, McIntyre J., at p. 211, noted that because of the broader scope of review it would not always be an error for the appellate court to substitute its findings for those of the Board. He nonetheless emphasized the advantageous position of administrative tribunals in assessing viva voce evidence, relying on this passage from the Ontario Divisional Court's decision in Re Ontario Human Rights Commission and City of North Bay (1977), 1977 1253 (ON CA), 17 O.R. (2d) 712, at p. 716:

. . . the learned Inquiry Officer's finding that the mandatory retirement provision was a bona fide occupational qualification and requirement for Cosgrove's employment was a finding of fact. . . . It was he who heard the evidence and under the circumstances was the only authority capable of judging the good faith of the assertions made before him. No appellate Court, in my respectful view, should overlook this signal advantage.

In Saskatchewan (Human Rights Commission) v. Saskatoon (City), supra, Sopinka J. took a similar view (at pp. 1314‑15):

It was the role of the Board, admittedly a very difficult one given the complexity and divergence of medical views on the subject, to determine if individualized testing was feasible. The Board concluded that this was not the case. In my opinion, the Board correctly applied the law, found as a fact that there was no practical alternative available to the appellant, and concluded that in adopting the policy with respect to retirement at age sixty, the employer was acting reasonably. No error of law having been committed by the Board, the Court of Appeal was wrong to reverse its decision. [Emphasis added.]

In my view, therefore, curial deference to the findings of fact of the Board of Inquiry is consistent both with principle and precedent. I note in particular that, while most of the witnesses who testified were qualified as experts, the Board also heard ordinary witnesses. In any event, as I have already mentioned, the assessment of expert evidence is also the province of the trier of fact.

This leads us to the second issue, the deference to be accorded to the findings of the Court of Queen's Bench. According to Cory J., it is open to an appellate court to substitute its own determinations of fact when it relies on the same written record as the trial court which first reviewed a tribunal's decision. While this might be true in some circumstances (and I make no comment on its general application, noting that there seems to be no authority on this proposition), this power of substitution of findings based on a written record is only valid to the extent that the trial court made its own findings of fact, and did not rely on the conclusions of the Board of Inquiry. Otherwise, an appellate court would once again risk failing to respect the privileged position of the tribunal which heard and saw witnesses.

In the case at bar, it is clear that Murray J. appreciated the advantage enjoyed by the Board of Inquiry with respect to viva voce evidence. After reproducing the passage from Clarke v. Edinburgh and District Tramways Co. cited above, he wrote (at p. 221):

This court has not had the privilege of hearing or observing the witnesses testify and to the extent that that may be important in reviewing the evidence which was before the board, Lord Shaw's comments must be remembered and taken into consideration.

Murray J. ultimately concluded that the respondent University had failed to discharge its burden of proof under s. 11.1 of the IRPA and wrote (at p. 247):

Having reviewed the evidence and bearing in mind the comments of the board with respect to certain of the evidence which it had the privilege of hearing firsthand, I am satisfied that the university has not discharged the onus which rests upon it to establish that art. 18.01 rationally relates to those objectives which the university claims it assists in accomplishing. The evidence does not show that it was carefully designed to meet and achieve the objectives put forward. [Emphasis added.]

In arriving at this conclusion, it is true that Murray J. re‑examined the evidence presented before the Board. However, his conclusions were completely consistent with those of the Chairperson, as he himself noted at various points in his judgment (see, for example, pp. 234, 238 and 248).

From these remarks made by the trial judge, indicating his acknowledgement of the Board's role as primary trier of fact, and from the similarity of the conclusions of the Board and the Court of Queen's Bench, I must conclude that Murray J. relied on the Board's findings of fact in making his own. In other words, the trial judge properly understood the rationale underlying the principle of curial deference and accorded the Board's findings of fact the measure of respect appropriate to the circumstances. In my opinion, this Court should do no less.

Even had I not been of the opinion that deference should have been given to the findings of the tribunal, I would have come to the same conclusions for the reasons that follow.

2. Section 11.1 of the IRPA

While I agree generally with my colleague Cory J.'s remarks on this issue, I believe it important to emphasize the principles of interpretation of human rights legislation, this Court's decisions in previous age discrimination cases, and the strict standard applicable in the absence of legislative deference, an element which again distinguishes this case from McKinney.

In order to further the goal of achieving as fair and tolerant a society as possible, this Court has long recognized that human rights legislation should be interpreted both broadly and purposively. Once in place, laws which seek to protect individuals from discrimination acquire a quasi‑constitutional status, which gives them preeminence over ordinary legislation. As Lamer J. (as he then was) wrote in Insurance Corp. of British Columbia v. Heerspink, 1982 27 (SCC), [1982] 2 S.C.R. 145, at pp. 157‑58:

When the subject matter of a law is said to be the comprehensive statement of the "human rights" of the people living in that jurisdiction, then there is no doubt in my mind that the people of that jurisdiction have through their legislature clearly indicated that they consider that law, and the values it endeavours to buttress and protect, are, save their constitutional laws, more important than all others. Therefore, short of that legislature speaking to the contrary in express and unequivocal language in the Code or in some other enactment, it is intended that the Code supersede all other laws when conflict arises.

As a result, the legal proposition generalia specialibus non derogant cannot be applied to such a code. Indeed the Human Rights Code, when in conflict with "particular and specific legislation", is not to be treated as another ordinary law of general application. It should be recognized for what it is, a fundamental law.

This Court again stressed the fundamental nature of human rights legislation in Ontario Human Rights Commission v. Simpsons‑Sears Ltd., 1985 18 (SCC), [1985] 2 S.C.R. 536, when McIntyre J. wrote at p. 547:

The accepted rules of construction are flexible enough to enable the Court to recognize in the construction of a human rights code the special nature and purpose of the enactment . . . and give to it an interpretation which will advance its broad purposes. Legislation of this type is of a special nature, not quite constitutional but certainly more than the ordinary ‑‑ and it is for the courts to seek out its purpose and give it effect.

A short time later, Dickson C.J. gave eloquent expression to the principle in a well‑known passage in Canadian National Railway Co. v. Canada (Canadian Human Rights Commission), 1987 109 (SCC), [1987] 1 S.C.R. 1114 (the Action Travail case), at p. 1134:

Human rights legislation is intended to give rise, amongst other things, to individual rights of vital importance, rights capable of enforcement, in the final analysis, in a court of law. I recognize that in the construction of such legislation the words of the Act must be given their plain meaning, but it is equally important that the rights enunciated be given their full recognition and effect. We should not search for ways and means to minimize those rights and to enfeeble their proper impact.

As a concomitant principle, provisions in human rights codes which provide defences for discriminatory conduct must be read narrowly if they are not to overwhelm and defeat the protection extended by this legislation. In Zurich Insurance Co. v. Ontario (Human Rights Commission), supra, this point was emphasized at pp. 358-59 of my dissenting reasons:

The starting point for any analysis of human rights legislation is the recognition that the purpose of such legislation is the protection of fundamental individual rights. These rights are violated if stereotypical group characteristics are ascribed to individuals. Since the protection of human rights is vitally important in our society, legislation prohibiting discrimination must be construed broadly using a purposive approach. . . .

A logical corollary to this purposive approach is the narrow interpretation of exceptions within human rights legislation. Defences to discrimination under the Code must be narrowly construed so that the larger objects of the Code are not frustrated.

Sopinka J., writing for the majority, agreed at p. 339 of his reasons:

In approaching the interpretation of a human rights statute, certain special principles must be respected. Human rights legislation is amongst the most pre‑eminent category of legislation. It has been described as having a "special nature, not quite constitutional but certainly more than ordinary". . . . One of the reasons such legislation has been so described is that it is often the final refuge of the disadvantaged and the disenfranchised. As the last protection of the most vulnerable members of society, exceptions to such legislation should be narrowly construed. . . .

Accordingly, in general, s. 11.1 of the IRPA must be read narrowly, so as to give the fullest possible breadth to the protection against discrimination afforded in s. 7, as both the Board of Inquiry and the Court of Queen's Bench found in this case.

With respect to the specific test to be applied under s. 11.1, it has been argued that the established structure of the model set out in R. v. Oakes, supra, provides the most appropriate test. In considering the applicability of the Oakes test to s. 11.1 of the IRPA, it is helpful to examine commonalities between this test and the two‑part test developed for those seeking to uphold discriminatory action under other exceptional provisions in human rights codes. I would note that where legislation provides tribunals with a specific test for discriminatory justifications, the tribunals should apply that test. The current "BFOQ" test, comprised of a subjective and an objective branch, was first set out by this Court in Ontario Human Rights Commission v. Borough of Etobicoke, supra, by McIntyre J., who wrote at p. 208:

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 which could defeat the purpose of the Code. 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.

In Brossard (Town) v. Quebec (Commission des droits de la personne), 1988 7 (SCC), [1988] 2 S.C.R. 279, a case involving an infringement of the Quebec Charter of Human Rights and Freedoms, Beetz J., writing for the majority, discussed the components of the objective branch (at pp. 311‑12):

The respondent must also demonstrate that the aptitude or qualification is related in an objective sense to the performance of the employment concerned. McIntyre J. suggested in Etobicoke that the purpose of the objective test is to determine whether the employment requirement is "reasonably necessary" to assure the performance of the job. In the case at bar, I believe that this "reasonable necessity" can be examined on the basis of the following two questions:

(1) Is the aptitude or qualification rationally connected to the employment concerned? This allows us to determine whether the employer's purpose in establishing the requirement is appropriate in an objective sense to the job in question. In Etobicoke, for example, physical strength evaluated as a function of age was rationally connected to the work of being a fireman.

(2) Is the rule properly designed to ensure that the aptitude or qualification is met without placing an undue burden on those to whom the rule applies? This allows us to inquire as to the reasonableness of the means the employer chooses to test for the presence of the requirement for the employment in question. The sixty‑year mandatory retirement age in Etobicoke was disproportionately stringent, for example, in respect of its objective which was to ensure that all firemen have the necessary physical strength for the job.

In other decisions, the Court has stressed the need for employers seeking to justify discriminatory practices to prove the absence of alternative means of testing employees. In Central Alberta Dairy Pool v. Alberta (Human Rights Commission), 1990 76 (SCC), [1990] 2 S.C.R. 489, Wilson J. explained the rationale for this approach (at p. 513):

The essence of direct discrimination in employment is the making of a rule that generalizes about a person's ability to perform a job based on membership in a group sharing a common personal attribute such as age, sex, religion, etc. The ideal of human rights legislation is that each person be accorded equal treatment as an individual taking into account those attributes. Thus, justification of a rule manifesting a group stereotype depends on the validity of the generalization and/or the impossibility of making individualized assessments.

In Saskatchewan (Human Rights Commission) v. Saskatoon (City), supra, Sopinka J. clarified the rule at pp. 1313‑14:

While it is not an absolute requirement that employees be individually tested, the employer may not satisfy the burden of proof of establishing the reasonableness of the requirement if he fails to deal satisfactorily with the question as to why it was not possible to deal with employees on an individual basis by, inter alia, individual testing. If there is a practical alternative to the adoption of a discriminatory rule, this may lead to a determination that the employer did not act reasonably in not adopting it.

See also Zurich Insurance Co. v. Ontario (Human Rights Commission), supra, at pp. 340-41 of Sopinka J.'s reasons for judgment, and my reasons at pp. 363-65.

In justifying a discriminatory policy, an employer must satisfy the civil burden of proof, that is, proof on the preponderance of probabilities. In the context of s. 1 of the Charter, the standard is rigorously applied. As Dickson C.J. first wrote in R. v. Oakes, supra, at p. 138:

Having regard to the fact that s. 1 is being invoked for the purpose of justifying a violation of the constitutional rights and freedoms the Charter was designed to protect, a very high degree of probability will be, in the words of Lord Denning, "commensurate with the occasion". Where evidence is required in order to prove the constituent elements of a s. 1 inquiry, and this will generally be the case, it should be cogent and persuasive and make clear to the Court the consequences of imposing or not imposing the limit.

Given the terminology, legislative history and purpose of human rights legislation, those seeking to justify discrimination attract a similarly heavy onus.

The review above illustrates some of the commonalities between the "BFOQ" test currently applied in human rights cases, and the Oakes test used in Charter cases. The "BFOQ" test, like the Oakes test, involves a consideration of objectives, of connections between discriminatory policies and objectives, and of disproportional impacts of policies. There is considerable interplay between the Charter and provincial human rights legislation, due to the similarity of their goals and the specific guarantees they provide. The scope for interplay is even more apparent in considering s. 11.1 of the IRPA. Both the Charter and the IRPA guarantee equality and prohibit discrimination on the basis of age. The language of s. 11.1 of the IRPA and s. 1 of the Charter is strikingly similar: both require that limitations on rights be both "reasonable" and "justifiable", and both clearly put the burden of proof on the party seeking to uphold discriminatory conduct. Given these similarities, it is appropriate for the court to approach the analysis of s. 11.1 in a manner consistent with that set out in Oakes.

The respondent University does not argue that analogy to s. 1 is inappropriate, but submits that the Court should apply the "flexible" standard first used in R. v. Edwards Books and Art Ltd., 1986 12 (SCC), [1986] 2 S.C.R. 713 (at pp. 781-83), and further developed in Irwin Toy Ltd. v. Quebec (Attorney General), 1989 87 (SCC), [1989] 1 S.C.R. 927 (at pp. 990 and 993‑94). (See also: Reference Re Public Service Employee Relations Act (Alta.), 1987 88 (SCC), [1987] 1 S.C.R. 313, at p. 391; PSAC v. Canada, 1987 89 (SCC), [1987] 1 S.C.R. 424, at p. 442; Andrews v. Law Society of British Columbia, 1989 2 (SCC), [1989] 1 S.C.R. 143, at pp. 185‑86, and Black v. Law Society of Alberta, 1989 132 (SCC), [1989] 1 S.C.R. 591, at pp. 627‑28.) My colleague La Forest J. concluded in McKinney that a less stringent standard of judicial scrutiny was mandated, because he found that the University of Guelph, in that case, had to make choices in the allocation of its scarce resources between academic staff over the age of 65 and younger professors. He wrote at pp. 285‑86:

In assessing proportionality and particularly the issue whether there has been a minimal impairment to a constitutionally guaranteed right, it must be remembered that we are concerned here with measures that attempt to strike a balance between the claims of legitimate but competing social values. In the case of broadly based social measures like these, where government seeks to mediate between competing groups, it is by no means easy to determine with precision where the balance is to be struck. . . .

The approach taken to these cases has been marked by considerable flexibility having regard to the difficulty of the choices, their impact on different sectors of society and the inherent advantages in a democratic society of the legislature in assessing these matters. Implicit in earlier cases, this was expressly adopted in Irwin Toy Ltd. v. Quebec (Attorney General).

. . . the question is whether the government had a reasonable basis for concluding that it impaired the relevant right as little as possible given the government's pressing and substantial objectives. [Emphasis in original.]

Thus, the respondent University argues, the flexible standard, whereby the defendant need not show that it adopted the least intrusive and offensive means of implementing policy, would apply here, and would militate strongly for an outcome identical to that in McKinney.

This argument simply cannot succeed and in that I agree totally with my colleague Cory J. when he states (at p. 000):

The policy rationale set out in Irwin Toy cannot be automatically transferred to the consideration of human rights legislation which seeks to regulate the behaviour of private actors . . . .

. . . in the realm of human rights legislation, the challenge will be to the actions of a private party such as an employer or landlord who, it is alleged, has adopted a discriminatory practice. The actions of private parties will often relate to private aims such as increasing the profit or efficiency of a business. Praiseworthy as those aims may be in themselves, no deference should be accorded to them once a complainant has demonstrated a prima facie case of discrimination. To adopt a deferential attitude to such private aims would undermine the professed goal of human rights legislation to guarantee the rights of minority groups, women and individuals against arbitrary and abusive treatment.

The rationale for the flexible standard is judicial deference to legislative choice, based on the idea that, with respect to resources and training, law makers are in a better position than courts to make policy choices between competing interests. Legislative representatives have better access to information, and they are accountable to the electorate. Because legislative bodies are democratic, their decisions are representative of the will of the majority in our society. This important latter consideration was spelled out by Dickson C.J. and Lamer and Wilson JJ. in Irwin Toy, supra, at p. 993:

When striking a balance between the claims of competing groups, the choice of means, like the choice of ends, frequently will require an assessment of conflicting scientific evidence and differing justified demands on scarce resources. Democratic institutions are meant to let us all share in the responsibility for these difficult choices. Thus, as courts review the results of the legislature's deliberations, particularly with respect to the protection of vulnerable groups, they must be mindful of the legislature's representative function. [Emphasis added.]

In McKinney, the majority of the Court conducted its s. 1 analysis from the premise that the policies under scrutiny were laws promulgated by a legislature. The reasons of La Forest J. are rife with references to "the Legislature" and to legislative action, as the following passage from p. 318 amply illustrates:

The Charter, we saw earlier, was expressly framed so as not to apply to private conduct. It left the task of regulating and advancing the cause of human rights in the private sector to the legislative branch. This invites a measure of deference for legislative choice. . . . Not, I repeat, that the courts should stand idly by in the face of a breach of human rights in the Code itself, as occurred in Blainey. But generally, the courts should not lightly use the Charter to second‑guess legislative judgment as to just how quickly it should proceed in moving forward towards the ideal of equality. The courts should adopt a stance that encourages legislative advances in the protection of human rights. [Emphasis added.]

(See also, for example, pp. 285‑86, 304‑5, and 314‑15 of La Forest J.'s reasons.)

This deference to legislative choice, however, is completely unwarranted where, as here, the defendant is not a legislative body. Leaving aside, for the moment, the question whether the allocation of scarce resources is even the central issue in this matter, this Court owes no deference to the policy of a private employer. On this point, I can do no better than to refer to the argument presented by the Alberta Human Rights Commission, which submitted at paragraph 33 of its factum:

This Court's reasons in Irwin Toy and McKinney closely tie entitlement to the Court's deference to the democratic, representative and neutral nature of the decision‑maker. With very few exceptions, employers are not representative of the general public. They are not subject to democratic checks and balances. They are not in the business of making difficult decisions of social policy and mediating between competing claims. They cannot be expected to be free from private interest in deciding whether to adopt a policy that infringes upon protected rights.

In formulating their policy towards their employees, bodies such as the respondent University must be expected to protect their own interests. They may even have an adversarial relationship with their employees with respect to such issues as salaries and the allocation of budget and other resources. Private employers are unrepresentative and accountable to no one, except to the extent that their conduct is regulated through law. As well, their ability to exploit available information in evaluating and determining policy is not necessarily any better than, and may indeed be distinctly inferior to, the ability of boards of inquiry and courts. As I wrote in Stoffman v. Vancouver General Hospital, 1990 62 (SCC), [1990] 3 S.C.R. 483, at p. 562, even if the respondent University were a lawmaking body for the purposes of the Charter, it "does not have all the requisite characteristics of a legislative body considering resource allocation, to warrant these special considerations being applied".

In enacting s. 11.1, the Alberta legislature delegated to private employers the task of setting up policies which were "reasonable" and "justifiable" under the IRPA. It chose to structure its human rights legislation in such a way that employers would have the freedom to choose the course of action most suitable to them in their particular circumstances. However, the legislature recognized that this scope of action could not be unchecked since employers, operating primarily in their own interests, might establish policies which were unreasonably and unjustifiably discriminatory. Accordingly, in s. 31(1) of the IRPA, it gave a broad supervisory and remedial role to the Alberta Human Rights Commission.

If the Alberta legislature had made its own social policy decision with respect to mandatory retirement of university employees, the Court might have reason to show some deference to this policy, since we would expect the legislature to examine its own proposed laws critically, in order to ensure its own compliance with its human rights code. Instead, the legislature delegated this critical and supervisory task to the Human Rights Commission. To the extent that this Court owes judicial deference to legislative choice, we should respect the law makers' decision to invest the Commission with the responsibility for determining which discriminatory employment policies are nevertheless valid under s. 11.1 of the IRPA. This reinforces the need for some deference by appellate courts, including this one, to the findings of fact of the Board of Inquiry, and makes the use of the flexible standard yet more inappropriate to this case than it would otherwise be.

Finally, on this point, I remain unconvinced, as I was in both McKinney and Stoffman, that this case falls into the exceptional category which attracts a relaxation of the strict standard in Oakes. The crux of the matter here is not the allocation of scarce resources, but rather the denial of the appellant's right to equality. My remarks made in dissent in Stoffman also apply here, mutatis mutandis (at p. 562):

My colleague La Forest J. also maintains that the hospital as an institution is forced to choose between "competing social groups" and, as a result, special considerations apply in considering the "minimal impairment" aspect of the Oakes formula. I agree that the Trustees do make a choice between groups in this situation. However, I cannot agree that in the present case the allocation of resources is a fundamental issue. That is not the kind of choice being made pursuant to this policy. The choice as I see it is made between competent medical practitioners who happen to be over 65, and competent doctors under 65, usually entering the medical practice. I do not see these circumstances warranting special considerations.

See also Tétreault-Gadoury v. Canada (Employment and Immigration Commission), 1991 12 (SCC), [1991] 2 S.C.R. 22, at pp. 43-44.

For these reasons, with respect to the burden of proof on the University under s. 11.1 of the IRPA, the Oakes strict standard is appropriate, rather than the more relaxed test adopted by this Court in Edwards Books, Irwin Toy and, latterly, McKinney. Simply put, this Court owes no deference to the policy decisions of a private employer, and the use of the flexible standard in these circumstances is completely unjustified. To paraphrase the remarks of Wilson J. in Andrews v. Law Society of British Columbia, supra, at p. 154, given that s. 7 of the IRPA "is designed to protect those groups who suffer social, political and legal disadvantage in our society, the burden resting on government [or, in this case, the respondent University] to justify the type of discrimination against such groups is appropriately an onerous one." This Court has recognized in other age discrimination cases that mandatory retirement is particularly difficult to justify. McIntyre J. noted in Ontario Human Rights Commission v. Borough of Etobicoke, supra, at p. 209:

We all age chronologically at the same rate, but aging in what has been termed the functional sense proceeds at widely varying rates and is largely unpredictable. In cases where concern for the employee's capacity is largely economic, that is where the employer's concern is one of productivity, and the circumstances of employment require no special skills that may diminish significantly with aging, or involve any unusual dangers to employees or the public that may be compounded by aging, it may be difficult, if not impossible, to demonstrate that a mandatory retirement at a fixed age, without regard to individual capacity, may be validly imposed under the Code. In such employment, as capacity fails, and as such failure become evident, individuals may be discharged or retired for cause.

There is a presumption that policies of mandatory retirement at a specific age are illegitimate, save in the clearest of cases. This presumption is not consistent with a posture of deference to employers, whether governmental or private, who seek to brand as incompetent all employees aged 65 and over.

In assessing the reasonableness and justifiability of the University of Alberta's mandatory retirement policy under s. 11.1 we must first examine the University's purposes in establishing the policy and ask whether these goals are legitimate or, in Charter terms, "pressing and substantial". Next, we must assess whether the policy is rationally connected to the goals it purports to further. If the University's goals are legitimate, and if the policy used to further those goals is rationally connected to them, we must then ask whether there is an alternative to the discriminatory policy and, specifically, whether the employer has shown why individual testing is not feasible in the circumstances. Throughout, we must bear in mind that the standard of proof upon the respondent is strict, akin to the s. 1 standard set out in Oakes. The employer must not merely show that its discriminatory policy is a reasonable alternative, but that it is the least burdensome option available, the one which least infringes on employees' rights while furthering the employer's legitimate goals.

Before applying this test to the facts in this case, however, we must pause briefly to consider the principles governing private contracts which purport to waive individuals' guaranteed rights, and to assess the impact of the particular collective agreement between the University and its academic staff in this case.

3. The Impact of the Collective Agreement

As a general rule, parties may not contract out of human rights legislation. This Court has consistently held that agreements which discriminate contrary to human rights codes are invalid. The first decision to address this issue explicitly was, like the case at bar, an age discrimination case under a provincial human rights statute, Ontario Human Rights Commission v. Borough of Etobicoke, supra. In response to the city's argument that mandatory retirement at age 65 was justified because the parties had agreed to it in a collective agreement, McIntyre J. wrote at pp. 213‑14:

While this submission is that the condition, being in a collective agreement, should be considered a bona fide occupational qualification and requirement, in my opinion to give it effect would be to permit the parties to contract out of the provisions of The Ontario Human Rights Code.

Although the Code contains no explicit restriction on such contracting out, it is nevertheless a public statute and it constitutes public policy in Ontario as appears from a reading of the Statute itself and as declared in the preamble. It is clear from the authorities, both in Canada and in England, that parties are not competent to contract themselves out of the provisions of such enactments and that contracts having such effect are void, as contrary to public policy. . . .

The Ontario Human Rights Code has been enacted by the Legislature of the Province of Ontario for the benefit of the community at large and of its individual members and clearly falls within that category of enactment which may not be waived or varied by private contract; therefore this argument cannot receive effect. [Emphasis added.]

This principle was again stressed in Insurance Corp. of British Columbia v. Heerspink, supra, where Lamer J. (as he then was) wrote at p. 158:

Furthermore, as it [the Human Rights Code] is a public and fundamental law, no one, unless clearly authorized by law to do so, may contractually agree to suspend its operation and thereby put oneself beyond the reach of its protection.

In Winnipeg School Division No. 1 v. Craton, 1985 48 (SCC), [1985] 2 S.C.R. 150, the Court was again faced with a collective agreement which an employer attempted to use as proof of the reasonability of its discriminatory policy of mandatory retirement. McIntyre J. stated flatly at p. 154:

There is no merit in the argument raised below, but not pressed in this Court, that the parties by agreeing to article 14 in the Collective Bargaining Agreement have contracted themselves out of the provisions of s. 6(1). The Human Rights Act is legislation declaring public policy and may not be avoided by private contract.

While my colleague Cory J. acknowledges that, generally speaking, parties may not contract out of human rights legislation, he concludes that "a collective agreement cannot be readily dismissed from consideration" (p. 000), because labour codes may overcome the traditional inequality of bargaining power between employer and employee which underlies the rule. In his opinion, a collective agreement which provides for a discriminatory practice may be evidence of the practice's reasonableness if it can be shown that the agreement was freely negotiated and that it does not discriminate "unfairly" against minorities.

This analysis represents a substantial departure from this Court's previous statements on the legality of contracting out of human rights legislation, and we should, in my opinion, be cautious in embracing it. Using an agreement as evidence of the reasonableness of a discriminatory practice must still be improper, in my opinion, where the agreement clearly violates human rights legislation to the point of being indefensible. The use of such a contract as the basis for a defence is only valid insofar as human rights legislation is ambiguous and imposes only broad guidelines which are open to interpretation, so that the parties to the bargain may honestly claim that they believed they were operating within the limits of the law.

But leaving aside these very real concerns about "bootstrapping", the prohibition against waiver of human rights provisions arises not only from a concern about inequality in bargaining power, but also because the rights guaranteed by human rights codes are seen as inherent to the dignity of every individual within our society. As a matter of public policy, such rights are not the common currency of contracts, but values which, by their very nature, cannot be bartered. For example, if employees, through their union, voted to sign a collective agreement with their employer whereby they agreed that female employees would be laid off first in case of an economic slowdown, that contract would be invalid. Even if a female employee sincerely believed that it made economic sense that male employees have greater job security, because in her view they are the primary bread‑winners in most families, courts would not condone a contract which, on its face, defined and devalued the position of an individual on the sole basis of her gender.

Constitutionally entrenched rights cannot be changed by a simple vote of Parliament. Human rights legislation is analogous in many respects to constitutional law. Indeed, s. 7 of the IRPA protects the same equality right as s. 15 of the Canadian Charter. Once extended, rights provided by provincial legislation cannot easily be withdrawn or circumscribed; that is surely what is implied by the statement that "short of [the] legislature speaking to the contrary in express and unequivocal language . . . it is intended that the Code supersede all other laws when conflict arises" (Insurance Corp. of British Columbia v. Heerspink, supra, at p. 158). It is equally unacceptable that, by a simple majority vote, a group of private citizens would be permitted to waive fundamental rights, barring truly exceptional circumstances.

Accordingly, in my view, while the existence of a collective agreement whereby employees agree to limit their own rights may exceptionally be a factor in considering the justifiability of an employer's discriminatory policy, any such agreement must be scrutinized to ensure that it does not discriminate unfairly against a minority of the union membership, and that it was freely negotiated. The particular context of the bargain, including relevant legislation in place at the time of its conclusion, may greatly mitigate its evidentiary weight. This is, in fact, just such a case. In the circumstances, the collective agreement between Dr. Dickason and the University of Alberta is not evidence of the reasonableness of its mandatory retirement policy.

The agreement which bound Dr. Dickason was not necessarily the result of freely conducted negotiations. Given the law which came in force in 1978, the UAPA, the Faculty Association and the University had to provide for mandatory retirement at age 65 for academic staff. The Faculty Agreement governing the contract which Dr. Dickason signed in 1979 made explicit reference to this legislation:

18.01 In accordance with the provisions of The Universities Academic Pension Act, a staff member shall retire from the service of the University (a) on June 30 following the attainment of age sixty‑five, or (b) on August 31 following the attainment of age sixty‑five for a staff member whose appointment date was prior to July 1, 1966. [Emphasis added.]

Since the parties merely abided by the law at the time Dr. Dickason was awarded tenure, it cannot be said that the Faculty Agreement resulted from free negotiations on that point. Nor is it evidence of Dr. Dickason's or the Faculty Association's consent to the policy, or of the reasonableness of the University's conduct. If art. 18.01 had not been part of the Agreement, mandatory retirement at age 65 would still have been in effect at the University through the operation of law.

It is true that, when Dr. Dickason first joined the academic staff of the University, there was no provincial legislation governing the retirement of academic staff. At the time, however, she signed only a fixed‑term probationary contract. Although the contract did specify that Dr. Dickason would be bound by the Faculty Agreement, the retirement provisions of the Agreement were not relevant to Dr. Dickason, since, by the end of her fixed term on June 30, 1978, she would still be well below normal retirement age according to the Agreement then in force (see art. 18.01 of the 1976 agreement).

Even in the absence of a law like the UAPA, I would have serious doubts about drawing any conclusion about the reasonableness of the University's policy from the provisions of the Faculty Agreement. At the time the relevant mandatory retirement provisions were first drawn up, members of the Faculty Association in all probability simply assumed (as did the legislature) that it was "normal" to retire at age 65, and that such issues as flexible retirement age were simply not on the bargaining table. Until amendments were made in 1985 to make it compatible with Charter provisions, "age" was defined in the IRPA as between 45 and 65 years, for the purposes of claims of discrimination under the provincial human rights law.

Moreover, Dr. Dickason herself clearly did not perceive the age of retirement as a negotiable issue when she became a full professor at the University. She testified before the Board to this effect:

Q.:And, ma'am, at about the same time that you got tenure did you also know that it was part of the agreement or arrangement between you and the university that you would leave when you became 65?

A.:I knew that that was the law of the land at that particular time, that that was the legislated law, but I also knew that there were often arrangements made to carry on employment afterward, right, correct.

Q.:And did you know then, ma'am, that you had also agreed to leave at age 65 in the sense that it was part of the deal between the university and yourself?

A.:It was a sine ‑‑ in order to get the job I would be required to sign that agreement, yes. [Emphasis added.]

Dr. Dickason's responses show the serious limitations in using a collective agreement as evidence of the consent of any particular member of a union. The mandatory retirement provisions of the Faculty Agreement had long been in place by the time she was offered a permanent position. She had to agree to the discriminatory policy in order to get the job. Where, as in this case, an employee had no part in negotiating offensive provisions, or any power to change policy through direct negotiations with her employer, the relevance of a contract to the reasonableness of a discriminatory policy is highly questionable.

My colleague has argued that discrimination on the basis of age is more subject to rational negotiation than discrimination under other heads, such as gender and race, because "everyone . . . becomes older with the passage of time" (p. 000). In fact, in my opinion, the opposite may be true. Because, in our society, old age tends to be less associated with wisdom and tranquillity and more with infirmity and dependence, we fear it. We may be more likely to discriminate against elderly people, in a futile attempt to distance ourselves from what will inevitably occur to each one of us.

This is borne out by studies which have shown that young people are notoriously unwilling to accept that they will grow old or plan for their old age, either financially or emotionally. The prospect of retirement is simply not a consideration most young workers take to heart in negotiating employment contracts. Even assuming that employees have access to full and correct information about their future financial situation, we cannot assume that they act upon this, rather than upon custom and stereotypes about aging, in assessing the cost of mandatory retirement clauses. As to how collective negotiation affects this state of affairs, Michael Krashinsky writes in "The Case for Eliminating Mandatory Retirement: Why Ecomomics and Human Rights Need Not Conflict" (1988), XIV Canadian Public Policy 40, at p. 44:

Of course workers are often represented by union leaders whose job it is to negotiate contracts in the interests of their members. But if workers systematically underestimate the cost of mandatory retirement provisions, there is little reason to believe that union leaders are better informed. And an enlightened union leader might well find little popular support for assigning the "correct" weight to mandatory retirement in contract negotiations.

Furthermore, employers themselves may also have bad information about older workers. If employers as a group act on stereotypes and underestimate the productivity of older workers, then they may put too much weight on the mandatory retirement clause in the contract.

The point of this section is not to assert that mandatory retirement is necessarily inefficient. Rather it is to raise serious questions about the contract theory approach that asserts that because mandatory retirement occurs in contracts affecting about half the work force, it is necessarily efficient. [Emphasis in original.]

While it is true that, as my colleague Cory J. states, "[e]ventual retirement with pension security is a matter of concern for all" (p. 000), different individuals have different expectations and different levels of concern regarding retirement. Most individuals do plan to stop working at age 65, as statistics from workplaces with flexible retirement plans bear out; others, given the choice, will retire even earlier. A small but significant minority of individuals see no reason to stop working at age 65. I fail to see why the expectations of the workers in the first two groups, who might find nothing objectionable about the University's mandatory retirement policy, should dictate the plans of those like Dr. Dickason who wish to prolong their careers past the "normal" pensionable age.

In other words, the fact that a majority of professors at the University would have elected to retain the policy of mandatory retirement at age 65 does not show, in my view, the reasonableness of that policy with respect to the appellant. This conclusion is strongly warranted not only by the principle of the inviolability of human rights, but by the extraordinarily negative effects that forced retirement can have on the small percentage of the population who wish to continue working past the age of 65. As I wrote in McKinney at p. 431:

Forced removal from the work force strictly on account of age can be extraordinarily debilitating for those entering their senior years. Aging is not a reversible process. Those yearning to carry on with their livelihood, career, and ambitions cannot have this aspiration stultified or decimated by some arbitrary scheme. The fact that we all experience the aging process is not a safeguard which prevents discriminatory acts by the majority. The prospect that current decision‑makers may some day be 65 and older is no guarantee against their acting in a discriminatory fashion against older individuals today, or against their acting on the basis of negative stereotypes. [Emphasis added.]

In the circumstances, the fact that the majority has purported to waive a minority's human rights cannot be used as proof that the waiver is valid.

In any event, the evidence shows that there was considerable support for a flexible retirement policy among the University's academic staff when Dr. Dickason retired. An ad hoc committee on the issue of retirement was struck within the Faculty Association in 1985; it tabled a report strongly favouring a flexible retirement plan. Drawing on this report, the Faculty Association discussed the abolition of mandatory retirement at age 65 with the University, although no amendment to the collective agreement was made by the time this case came before the Board. Before this Court, counsel for Dr. Dickason gave this explanation for the failure to amend the agreement (at pp. 50‑51 of the Court transcript):

In fact, the Faculty Association had tabled in the pre‑collective bargaining scenario proposals to amend the collective agreement to put flexible retirement in place instead of mandatory retirement. But those discussions came to nothing, and in the formal negotiations an amendment to the collective agreement did not occur.

It seems to me . . . that that consideration ought to be also viewed in the context of the fact that these proceedings were under way at that time. It is reasonable to conclude, I submit, that the parties were relying on the outcome of these proceedings to determine the situation at the university.

The record also shows that the Faculty Association, the AASUA, supports the concept of flexible retirement and, at the lower stages of the proceedings, was supportive of Dr. Dickason's initiative in challenging the collective agreement.

I submit that nothing can be drawn in the overall context from the fact that, by the time we got to Tribunal, the collective agreement had not been amended.

It appears that, once no longer required by provincial statute, the University's policy of mandatory retirement at age 65 remained in place not because the faculty supported it, but because all parties involved were awaiting the final outcome of this case. This is consistent with the actions of the Canadian Association of University Teachers, which has been lobbying for flexible retirement policies for its 27,000 members since 1979 and, which, together with the Ontario Confederation of University Faculty Associations, has invested significant funds in court challenges to mandatory retirement policies.

Having analyzed the legal issues underlying this case, I now turn to their application to the evidence in this case. As I noted earlier, I would defer to the findings of fact of the Board of Inquiry. My discussion of the evidence here is not for the purpose of reviewing or reconsidering the findings of the Board, but is rather for the purpose of responding to the arguments of the respondents, and to the analysis of my colleague Cory J.

In light of the conclusions of this Court in McKinney, the appellant concedes that the University's stated objectives ‑‑ the preservation of the tenure system, academic renewal, institutional planning and retirement with dignity ‑‑ are "pressing and substantial". Similarly, the good faith of the University is not contested. Therefore, we may proceed directly to the question of whether the respondent has succeeded in proving that the policy is reasonable and justifiable under s. 11.1 of the IRPA. Each of the University's stated objectives must be examined to ascertain whether it is rationally connected to the policy. If the policy of mandatory retirement at age 65 is rationally connected to any or all of the objectives, we must then ask whether the policy nonetheless places an undue burden on employees or goes beyond "minimal impairment", bearing in mind that the onus on the University is stringent and that, in the circumstances, no deference to its policy is warranted if there is a non‑discriminatory alternative.

Rational connection

(a) Tenure

The respondent University urged, as did its analogue in McKinney, that the elimination of mandatory retirement will threaten the tenure system. In the absence of forced retirement at age 65, the University speculates that it would be forced to conduct more faculty evaluations, and this in turn would inhibit scholarship. My colleague Cory J. accepts this argument, concluding that flexible retirement policy "would undermine the sense of security which is so essential for the free exercise of academic inquiry" (p. 000).

In my view, this argument misapprehends the purpose of tenure and underrates peer evaluation. The function of tenure is not to shield faculty members from their own incompetence, but rather to prevent spurious challenges actually based on the unpopularity of their academic views. The role of academics within universities and within our larger society would be greatly devalued if we accepted the idea that, once tenured, professors can become unproductive and unthinking. The most ardent proponent of academic freedom would admit that there must be some mechanism to remove professors who are taking advantage of their protected position in order to contribute little or nothing to teaching or scholarship. Otherwise, tenure would become a euphemism for tolerated incompetence.

But is mandatory retirement at age 65 an acceptable way of determining who is incompetent? To ask the question is to answer it, since it is obvious that the blanket policy of firing people when they become old has nothing to do with assessing their continuing contribution or their overall competence, but rather relies entirely on a stereotype of declining abilities with age. To the extent that the policy of mandatory retirement is unrelated to this stereotype, the University appears to be arguing that there is no other way to get rid of incompetent professors but to wait until they are old enough to retire. This argument is, of course, unacceptable. Besides, alternative mechanisms of evaluating professors exist and are actually in place.

Peer evaluation is one such mechanism already used. It is a fair and equitable way of assessing professors in good faith, on the basis of their teaching, research and publication records, rather than on their age. Unless abused, it poses no threat to academic freedom, and in fact enhances the value of tenure by ensuring that incompetent professors, young or old, are dismissed. These evaluations are difficult, as they should be, but no more difficult than the assessments that are made before an academic is first hired for a term, offered tenure, promoted, and awarded merit increases. In effect, evaluations of a professor's competence, as opposed to the popularity of their specific views, are made throughout his or her career.

Dismissal for cause at Canadian institutions has been infrequent in the past decade. I suspect, however, that this speaks not to the inadequacies of peer evaluation, but to the instinct of most professors to resign or transfer if they receive consistent signals, from merit assessments and from other colleagues, that their performance is substandard. This suspicion is borne out by studies showing that unsatisfactory faculty are more likely to retire early; see Ruth F. Necheles‑Jansyn, "Retirement in Academe: Special Case or Social Model?", in [1983] Aging and Work 175, at p. 179.

Most importantly, peer evaluation represents an assessment of merit. Because it is a somewhat unwieldy mechanism, it will never be used frequently, but its existence safeguards against the abuse of the tenure system in a way that mandatory retirement at a fixed age never could. As I remarked in McKinney at pp. 426-27:

The fear that aging professors will rest on their laurels and wallow in a perpetual and interminable quagmire of unproductivity and stagnation may be a real one. Yet it applies with equal force to younger tenured faculty as well. Peer review, so long as it is predicated on the premise of unbiased good faith, provides a healthy injection of critical evaluation and will serve to promote the scholastic standards indispensable to a flourishing university.

Given that a fair way of assessing merit already exists, and that the link between mandatory retirement and academic freedom and excellence is tenuous at best, there is no basis to conclude that the elimination of the University's policy would threaten tenure.

The best proof that the University's fears are groundless comes from universities which have abolished mandatory retirement without any noticeable effect on tenure. Drawing on both expert and ordinary testimony it had heard, the Board of Inquiry found (at p. D/4476):

No evidence was presented to satisfy me that the elimination of mandatory retirement would result in the demise of or even serious injury to the tenure system at the University of Alberta....

No evidence was led in relation to the University of Manitoba or the universities in the province of Quebec where mandatory retirement has been eliminated to indicate jeopardy to the tenure system at those universities. On the contrary, Prof. Ian J. Kerr of the University of Manitoba, a past-president of that university's faculty association testified unequivocally that, "It [elimination of mandatory retirement in 1981] has had no effect on tenure at the University of Manitoba" (Trans. 1165). Aside from the removal of the pertinent clause, Dr. Kerr also confirmed that there had been no changes in the collective agreement at the university related to the abolition of mandatory retirement; nor have there been changes in the university's professor evaluation procedures (Trans. 1165).

Similarly, tenure has not been abandoned in universities in the United States, 15 per cent of which have no mandatory retirement for tenured faculty, as was noted by both majority and minority in McKinney; see pp. 427‑28 and 283. This lack of change is attributable, in part, to the fact that very few employees choose to continue working past age 65, a point made by Frank Reid, an expert on gerontology who testified before the Board (Economic Aspects of Mandatory Retirement (1987), at pp. 11‑12):

. . . most employers would not find it worthwhile to make substantial changes to their personnel practices as a result of a small number of employees who would want to work past the normal retirement age. The need for increased performance reviews is also reduced by the fact that the few employees who elect to work beyond normal retirement age will do so for only a relatively short period of time. Evidence from the Conference Board survey (Dunlop 1980 p. 40) ". . . indicates that those few employees presently aged 65 who elect to continue working will, in general, leave their employment prior to age 70 and will, on average, probably work for a period of three years beyond age 65".

Evidence from Quebec (where mandatory retirement was abolished in 1982) suggests that the period of work may be even shorter. After examining the evidence on those who continued to work past age 65 Labour Canada (1985, Appendix p. 2) concluded ". . . this prolongation of employment is not apparently very long. After 6 to 8 months many decide to retire."

In short, employers would not have to change their system of evaluation to avoid a "bulge" of older workers.

Accordingly, the University's fears about how tenure might be affected should mandatory retirement be eliminated are insufficient to prove a rational link between its objective and its discriminatory policy. In light of the evidence given, it was not unreasonable for the Board and the Court of Queen's Bench to conclude that the link had not been proven. As Murray J. said (at p. 247):

The weight of evidence falls on the side of establishing that a policy of mandatory retirement is not necessary to ensure that tenure is preserved. There is a fear of this happening but it is speculative. There are equally well qualified people who hold the opposite opinion. The empirical evidence which is available to this time does not support this fear. [Emphasis added.]

(b) Academic Renewal

My colleague Cory J. accepts the argument that mandatory retirement at a fixed age allows the University to open positions to younger academics, thereby at once allowing a fresh infusion of ideas into the institution and remedying the problem of underfunding. Although a superficially attractive proposition, in my view, this reasoning does not stand the close scrutiny mandated under the IRPA. Aside from being based on the false premise that older workers are uniformly less productive and original than their younger colleagues, a myth refuted decisively by the evidence before us in this case, it ignores uncontradicted expert evidence about trends in retirement.

The very purpose of human rights legislation is the elimination of stereotypes associated with inherent characteristics. Yet one of the most persistent prejudices in our society is that old people are uniformly senile and physically incapacitated. In fact, the effects of age vary tremendously from person to person. As life expectancy has increased, the definition of old age itself has had to be revised. In creative fields, such as literature and the performing arts, certain older individuals continue to contribute long after "normal" retirement age; examples include Will Durant, who published the last volume of his Story of Civilization when he was in his 80s, great artists such as Pablo Picasso, and such musical notables as Arturo Toscanini, Leonard Bernstein and Herbert von Karajan. As well, societies around the world have long recognized that certain qualities, such as wisdom and leadership, are most likely to be acquired with age. Many judges, including judges of this Court, continue to work after they would have been forced to retire from other jobs. A pantheon of celebrated political leaders from this century, including Winston Churchill, Charles de Gaulle, and Konrad Adenauer, made some of their greatest contributions at a mature age.

In the academic context, as the Board found, "the proposition of originality as a by‑product of `youth' . . . has not been proved" (p. D/4473). Decline in complex cognitive processes does not occur until the late 70s or early 80s; see Frank Reid, supra, at p. 3, and Necheles‑Jansyn, supra, at p. 178. Academics, as a group, tend to suffer still less decline because they are in a relatively high social strata, enjoy good health and work in jobs which are not very physically demanding. As well, the intellectual stimulation they enjoy from their work and contact with colleagues inhibits decline. The maturity, wisdom and insight older faculty have acquired even give them distinct advantages over younger colleagues.

The stereotype of older professors clinging desperately to their posts despite declining abilities simply is not warranted on the evidence. Generally speaking, those who start by being highly productive and creative remain so as they get older, and age seems to have very little influence on the quality and quantity of work produced; see Stephen Cole, "Age and Scientific Performance" (1979), 84 Am. J. Soc. 958, at pp. 968‑69. Truly exceptional performance is neither the province of the young nor of the old. Of course there are unproductive older staff; there are also unproductive younger staff. The only difference between the two is that, once someone reaches a certain age, society tends to assume that any substandard performance is a direct consequence of age, whereas when they are younger, we make no such assumption.

At the University of Alberta specifically, the record of merit increases indicates that the greatest decline in performance occurs in the age groups 40‑44 and 45‑49. According to a statistical expert who testified before the Board, there were no statistically reliable differences in performance between the three oldest age groups, 50 to 54, 55 to 59, and 60 to 65, even though the expectation for senior professors was higher than that for younger staff, making it more difficult for them to achieve a satisfactory rating. This led the Board to conclude (at pp. D/4472-73):

On the evidence presented, I cannot find that there is a decline in ability which would render the vast majority of academic staff unable to function effectively as they approach or reach 65. There will obviously be some professors who are giving substandard performance at 65, but similarly such deficiencies could exist for staff at any age. It would be unwise, unfair and now illegal to stereotype and penalize the high or even adequate performance because he, or she, has reached the age of 65.

I defer to the Board's finding of fact in this regard, and would be ready to dismiss the University's argument on this basis alone.

The assumption that the elimination of mandatory retirement will significantly reduce the number of jobs for young academics must also be examined critically. Experts before the Board testified that the "flow of replacement" would return to its initial level in five years or less, a short term effect which certainly cannot justify the wholesale dismissal of all competent academics over age 65. The limited effect of flexible retirement on the rate of hiring of young academics is due, in part, to the small number of academics who actually desire to keep working beyond the "normal" retirement age. According to a 1980 Conference Board of Canada report, only a tiny fraction of the labour force, about one fifth of one per cent, is likely to be forced to retire at age 65; the rest retire voluntarily at age 65 or earlier (see F. Reid, supra, at p. 2, and Krashinsky, supra, at pp. 45 and 48).

The average age of retirement amongst workers in general may decline still further with the increasing popularity of early retirement. In the United States, for example, although the age of retirement has been increased from 65 to 70, the average age of retirement has declined from 63 to 62. Labour shortages are projected for the 1990s based on the upswing in early retirement combined with current population trends (see Robert C. Ford and Myron D. Fottler, "Flexible Retirement: Slowing Early Retirement of Productive Older Employees", [1985] Human Resource Planning, at p. 147, and James W. Walker and Harriet L. Lazer in The End of Mandatory Retirement: Implications for Management (1978), at p. 7). The Board commented at p. D/4474:

With the increasing popularity of flexible retirement options, compulsory retirement wanes in importance as a function in achieving renewal. Indeed, it was the evidence of Dr. Gideon Rosenbluth, professor of Economics (macro economics) at the University of British Columbia that, "The early retirements balance out the ones that postpone their retirements". I concur with this statement, based on the evidence before me. [References omitted.]

The University of Alberta's own records show that of the 80 retirements which took place in 1985 and 1986, 47 were early and 33 were regular. Even if the age cap were removed, the vast majority of faculty would continue to retire at age 65 or earlier.

While the rate of hiring in Canadian universities has declined in the last decade, the cause is not recalcitrant older academics refusing to retire, but a government policy of fiscal restraint and the population bulge caused by the baby boom. As Michael Krashinsky, an expert before the Board, wrote at p. 48 of "The Case for Eliminating Mandatory Retirement: Why Economics and Human Rights Need Not Conflict", supra:

. . . the issue of faculty renewal is something of a red herring. There has been precious little renewal in recent years, despite mandatory retirement, due largely to government policy that has cut real allocations to the universities, and exacerbated by the skewed age distribution of the faculty. Furthermore, the Bovey Commission itself suggested that short‑term government funding be provided to reduce the transitional impact of eliminating mandatory retirement, financing new appointments until expected retirements materialized. Faculty renewal, therefore, has little to do with mandatory retirement and everything to do with public policy, and blaming older faculty members who do not wish to retire for the problem is unreasonable. [Emphasis added.]

The Court of Queen's Bench too identified the "real villain" in this situation as the demographic bulge of academics currently in their forties, and noted that increased attention paid to the performance of professors could well alleviate the problem of academic renewal (see p. 248).

Finally, it is deeply offensive to the notion of human rights that a private employer should offer up a disadvantaged group as a sacrificial lamb to another group in the absence of a legislative mandate, or compelling evidence that this is either desirable or necessary. Despite my colleague Cory J.'s oblique reference to affirmative action, there is no proof in this case that the policy of mandatory retirement of professors over the age of 65 is linked in any way to the promotion of the hiring of women or other traditionally underrepresented groups within Canadian university faculties. While his call for "inter‑generational equity" may be laudable, the "young" have not traditionally and are not today regarded as a disadvantaged group.

Even if there were competition between two truly disadvantaged groups in this case, surely we would not even consider permitting private employers to make the choice between them if the group that was being penalized was anyone other than the elderly. Imagine, for example, the reaction if a university announced that it would arbitrarily fire a certain number of physically handicapped professors in order to make room on faculty for women and racial minorities. Yet, under the IRPA, discrimination on the basis of age is proscribed in exactly the same terms as discrimination on the basis of physical handicap, gender and race. I fail to see why this Court should make a distinction where the legislature has not in its treatment of groups disadvantaged by stereotype and prejudice.

Hence, in my opinion, the respondent University has failed to prove a rational link between the goal of faculty renewal and its policy of mandatory retirement.

(c) Institutional Planning

The institutional planning argument is, in many ways, an adjunct to the academic renewal argument, except that it appeals more explicitly to administrative efficiency and costs. According to the University, mandatory retirement is necessary because it allows the administration to plan ahead, secure in the knowledge that every professor will retire at age 65.

In dismissing this argument, I refer simply to the findings of fact of the Board of Inquiry, which wrote at p. D/4473:

With so many uncertainties in the planning process, I cannot accept mandatory retirement being vital to the process. Early retirement . . . for example has become increasingly popular among the academic staff, essentially being available anywhere between ages 55‑65. Illness and death of academic staff cannot be predicted. The level of government funding varies annually; student enrolment and course selection vary from year to year. In total the number of uncertain factors overwhelm the one constant of mandatory retirement.

As the Court of Queen's Bench pointed out, other variables, such as resignations, deaths, and early retirement, are predicted with relative certainty by means of statistical forecasts. The slight inconvenience which inevitably remains cannot be used, alone, as a justification for the denial of equality on the basis of age.

(d) Retirement with Dignity

Of all the arguments put forth by the University, this is the one which stems most clearly from stereotype. It depends entirely on the idea that professors who reach the age of 65 must necessarily fear assessment on the basis of their performance, because that performance has necessarily declined rapidly with age. Given that the evidence conclusively refutes the myth of universal decline, and that peer evaluation is an effective tool for identifying incompetence, in my view this proposition clearly fails at the outset.

I find it curious, however, that this argument is associated with dignity. Faculty members such as Dr. Dickason, who have contributed significantly until age 65, must find particularly demeaning the insinuation that they have been allowed to "coast" to retirement. As I commented in McKinney at p. 430:

The threat that an evaluation scheme will "constitute a demeaning affront to individual dignity" (at p. 284) is difficult to accept. Are objective standards of job performance a demeaning affront to individual dignity? Certainly not when measured against the prospect of getting "turfed‑out" automatically at a prescribed age, and witnessing your younger ex‑colleagues persevere in condoned relative incompetence on the strength of a "dignifying" tenure system.

Surely those who have worked hard all their lives would prefer to be treated as the equals of their younger colleagues, rather than being suffered on the assumption that age makes them redundant. In the words of the Board, "such practice does little to maintain the dignity of either the individual professor concerned or the University as an institution of scholarship" (p. D/4478).

Having concluded that the University has failed to prove a rational connection between any of its objectives and its impugned policy of mandatory retirement, it is not strictly necessary for me to discuss the proportionality of the policy to the goals or the issue of non‑discriminatory alternatives. However, since my colleague has dealt with these questions, I will address them briefly as well.

Minimal Impairment

According to my colleague Cory J., the policy of mandatory retirement withstands the minimal impairment test because "[n]o obvious alternative policy exists which would achieve the same results without restricting the individual rights of faculty members" (p. 000). In my view, this conclusion is untenable given the strict standard of scrutiny mandated under the Oakes test and applied by this Court in employment discrimination cases. The University, upon which the burden falls to show the reasonability and justifiability of its discriminatory policy, has not presented convincing evidence as to why alternative non‑discriminatory practices would not achieve the same results as or better results than mandatory retirement at age 65.

Assessments are already used to review the performance of individual professors throughout their careers, without posing a threat to academic freedom. Peer evaluation can and does distinguish between incompetence and unpopular academic views. Its use may increase the value of tenure, since it ensures that substandard performance is not rewarded. Based on objective assessment, peer evaluations offer a far more dignified approach to academic work than the assumption that all those over a certain age should be allowed to coast to retirement. Hence, such evaluations are infinitely preferable to the blanket policy of dismissal without cause on the basis of a stereotype of declining ability with age.

Encouragement of early retirement is another non‑discriminatory alternative that would achieve the University's goals. My colleague Cory J. has given this option short shrift, claiming that it "would not necessarily achieve the numbers or certainty of vacancies among faculty that are required for renewal" (p. 000). In fact, the evidence before the Board and before this Court indicates that statistical projections could be used to predict the rate of vacancies without mandatory retirement, and that the flow of retirement would return to its initial level within five years.

Of course, the search for alternatives is premised on the idea that the elimination of mandatory retirement will have a significant impact at the University of Alberta. However, the University has not presented any evidence to this effect from other academic institutions where mandatory retirement has been eliminated. There is no doubt that the University genuinely fears that dire consequences will follow the elimination of mandatory retirement, but these apprehensions are insufficient alone to satisfy its evidentiary burden. In fact, the evidence clearly suggests that the effect of eliminating mandatory retirement at the University of Alberta will be minimal, recalling the small number of professors who will wish to remain after age 65. The negligible impact of the removal of the policy can readily be gauged by the experience of academic institutions in Quebec, Manitoba, and the United States.

I accordingly agree with both the Board and the Court of Queen's Bench that the University's policy of mandatory retirement does not meet the minimal impairment test.

Proportionality

Again, my colleague Cory J. disposes of the issue of proportionality very briefly, writing (at p. 000):

Without question, mandatory retirement constitutes prima facie discrimination. Those faculty members who, like Professor Dickason, would choose to continue working past the age of 65 will suffer from the denial of this opportunity. Yet these individuals have the benefit of secure and reasonable pensions. The policy forms part of a collective agreement which benefits, as a whole, the academic staff of the University. In this context, the effects of the prima facie discrimination are proportional to the legitimate objectives served.

First of all, any conclusions regarding pensions are unfounded in this case, the University having failed to adduce any evidence whatsoever on economic consequences of the elimination of its policy. With respect to the acquiescence to mandatory retirement by the faculty in the collective agreement, as I concluded earlier, this cannot be used as an indication of the reasonability of the policy, since a law in force at the time Dr. Dickason was awarded tenure required that this provision be included. Nor can the administrative and financial burden of setting up alternative policies be considered a strong argument on this point, since the peer evaluation system is already in place and the cost of encouraging early retirement would be negligible.

Thus, benefits to the University as a result of its discriminatory policy are, to say the least, unproved. Yet forced retirement at age 65 can have a devastating effect on individuals like Dr. Dickason. Some individuals in her position may be financially ill‑prepared for retirement. Although the elderly make up only 10 per cent of Canada's population, they represent 42 per cent of the poor; see Alan Roadburg, Aging: Retirement, Leisure and Work in Canada (1984), at p. 22 et seq. Women are penalized, in particular, because they tend to have lower paying jobs which are less likely to offer pension coverage, and they often interrupt their careers to raise families. (These socio‑economic patterns, combined with private and government pension plans which are calculated on years of participation in the work force, in some ways make mandatory retirement at age 65 as much an issue of gender as of age discrimination.) According to one study, only people in the top 25 per cent of earning brackets can provide for themselves in old age, since both savings and proceeds from indexed pensions are ravaged by inflation (see Roadburg, supra, at p. 24).

For an academic, the consequences of retirement include not only the loss of full income, but loss of access to computers, books, students, laboratories and other resources. As the Board found, the awarding of the status of professor emeritus is a poor substitute, since it confers almost no access to University facilities. Mandatory retirement for a professor hence means not only the loss of a job, but the denial of resources essential to any work at all.

Moreover, the psychological effects of forced retirement may be as harmful as financial consequences. This Court remarked in Reference Re Public Service Employee Relations Act (Alta.), supra, at p. 368:

Work is one of the most fundamental aspects in a person's life, providing the individual with a means of financial support and, as importantly, a contributory role in society. A person's employment is an essential component of his or her sense of identity, self‑worth and emotional well‑being. [Emphasis added.]

Given the central importance that our society accords to career as a way of defining an individual's status and self‑worth, it is hardly surprising that being dismissed without cause on account of one's age is extremely traumatic. According to Roadburg, supra, at pp. 12‑13:

. . . the effects of retiring voluntarily, as opposed to forced retirement owing to company policy or health reasons, may be significant. For example, researchers have found that voluntary retirees are more likely to have a positive attitude and a higher level of satisfaction compared with non voluntary retirees. . . . This result may be expected especially in a nation which professes to follow democratic principles of equality and freedom.

The American Medical Association's assessment is even more dramatic, as reported by Walker and Lazer, supra, at pp. 11‑12:

. . . the American Medical Association, in opposing mandatory retirement at any age, observed that the sudden shock of compulsory retirement and loss of earning power and productive work often leads to physical and emotional deterioration and premature death. [Emphasis added.]

The devastating effects that forced retirement has on a worker's finances, health, and self‑esteem are grossly disproportionate to any advantages accruing to the University by its discriminatory practice of firing professors on the basis of their age, and this aspect of the applicable test under s. 11.1 of the IRPA has not been met.

Conclusion

For all these reasons, the University's policy of mandatory retirement at age 65 is not reasonable and justifiable under s. 11.1 of the IRPA. To borrow the words of Michael Krashinsky, supra, at p. 49, quoting J. R. London, "Universities Should Prepare for Abolition of Mandatory Retirement", University of Toronto Bulletin, June 6, 1983, p. 10:

Historically, all advances in human rights have been opposed by those with vested interests, who inevitably have predicted all manner of resulting evil and destruction. My own view is that the benefits of eliminating "ageism" from our employment systems far out‑weigh the costs. Stereotypical victimization cannot be tolerated. Administrative inconvenience can.

The appeal should be allowed, the decision of the Court of Appeal reversed and the judgment of the Queen's Bench reinstated, the whole with costs.

Sopinka J.

The following are the reasons delivered by

Sopinka J. (dissenting) -- I have had the privilege of reading the reasons of my colleagues L'Heureux-Dubé and Cory JJ. While I agree with the conclusion and for the most part with the reasons of L'Heureux-Dubé J., I prefer to state my own reasons for arriving at a conclusion that may seem at variance with that of the majority in McKinney v. University of Guelph, 1990 60 (SCC), [1990] 3 S.C.R. 229, of which I was a member.

In McKinney we decided that mandatory retirement at a specified age was not constitutionally impermissible. Parliament or a provincial legislature by appropriate legislation could prohibit or permit it. The decision could be changed from time to time. If not prohibited by human rights legislation, the issue would be decided as between employer and employee. In the case of unionized employees, the issue would be dealt with in the collective bargaining process. This result was more democratic and thus more consistent with values of the Canadian Charter of Rights and Freedoms. This approach accorded some deference to the legislature, which was an important premise of the decision. La Forest J. stated (at pp. 304-5):

In undertaking this task, it is important again to remember that the ramifications of mandatory retirement on the organization of the workplace and its impact on society generally are not matters capable of precise measurement, and the effect of its removal by judicial fiat is even less certain. Decisions on such matters must inevitably be the product of a mix of conjecture, fragmentary knowledge, general experience and knowledge of the needs, aspirations and resources of society, and other components. They are decisions of a kind where those engaged in the political and legislative activities of Canadian democracy have evident advantages over members of the judicial branch, as Irwin Toy [1989 87 (SCC), [1989] 1 S.C.R. 927], at pp. 993-94, has reminded us.

I stated in my concurring reasons (at p. 446):

The current state of affairs in the country, absent a ruling from this court that mandatory retirement is constitutionally impermissible, is the following. The federal government and several provinces have legislated against it. Others have declined to do so. These decisions have been made by means of the customary democratic process and no doubt this process will continue unless arrested by a decision of this Court. Furthermore, employers and employees through the collective bargaining process can determine for themselves whether there should be a mandatory retirement age and what it should be. They have done so in the past, and the position taken by organized labour on this issue indicates that they wish this process to continue. A ruling that mandatory retirement is constitutionally invalid would impose on the whole country a regime not forged through the democratic process but by the heavy hand of the law. Ironically, the Charter would be used to restrict the freedom of many in order to promote the interests of the few.

The Province of Alberta, in its Individual's Rights Protection Act, R.S.A. 1980, c. I-2, as amended (the "IRPA"), has not legislated to prohibit mandatory retirement. Rather, by virtue of s. 11.1 of the IRPA, it has left that decision to employers and employees provided that where mandatory retirement is resorted to, the employer must satisfy a Board of Inquiry that this discriminatory practice is reasonable and justifiable.

The first question in this appeal is the meaning to be ascribed to the phrase "reasonable and justifiable". I agree with the Board of Inquiry, the courts below and my colleagues that the jurisprudence developed with respect to s. 1 of the Charter is a useful guide in applying s. 11.1 of the IRPA. There is also substantial agreement that in determining whether the proportionality factor has been met, the Court should not defer to the decision of an employer in the same way as in the case of a governmental actor. The rationale for this approach in the latter case is that in dealing with governmental actors, it is often difficult, if not impossible, to prove in the ordinary way whether a particular measure will in fact achieve its objective. Accordingly, if Parliament, a legislature or other governmental body had a reasonable basis for concluding that the measure would achieve its objective, that is ordinarily a basis for concluding that there is a rational connection between the measure and the governmental objective. Accordingly, although the government could not prove that advertising toys on television had a manipulative effect on children, nor that hate propaganda actually promoted hatred against an identifiable group, nor that pornography caused harm to women, the fact that there was sufficient evidence to provide a reasonable basis for the legislature to adopt the impugned legislation in aid of its objective was sufficient to save it. See Irwin Toy Ltd. v. Quebec (Attorney General), 1989 87 (SCC), [1989] 1 S.C.R. 927, R. v. Keegstra, 1990 24 (SCC), [1990] 3 S.C.R. 697, and R. v. Butler, 1992 124 (SCC), [1992] 1 S.C.R. 452.

As my colleagues point out, we cannot use this same model when applying s. 11.1 to employers. My colleague L'Heureux-Dubé J. would apply the strict version of the test in R. v. Oakes, 1986 46 (SCC), [1986] 1 S.C.R. 103, while my colleague Cory J. would apply a more flexible test, with the Oakes test serving as a guideline.

In my view, it must be stressed that the tests developed for the application of s. 1 of the Charter and the defence to discrimination under human rights legislation, commonly referred to as BFORQ (bona fide occupational requirement or qualification), are similar. While the former as expounded in Oakes is more elaborate, they both require that the impugned measure bear a rational relationship to a legitimate objective. In the case of an employer, the objective is usually the efficient and economical operation of the employer's business or activity. Section 11.1, therefore, is essentially a generalized BFORQ which, while worded more generally than the provision with which we dealt in Ontario Human Rights Commission v. Borough of Etobicoke, 1982 15 (SCC), [1982] 1 S.C.R. 202, serves the same purpose. Accordingly, while it is appropriate to apply the jurisprudence developed under s. 1 as a guideline, when that guideline is wanting we should interpret s. 11.1 by reference to the jurisprudence relating to the BFORQ. In this regard I therefore see no reason to depart from the test which was enunciated by McIntyre J. in Etobicoke and adopted by Beetz J. in Brossard (Town) v. Quebec (Commission des droits de la personne), 1988 7 (SCC), [1988] 2 S.C.R. 279. In the latter case, Beetz J. stated, at pp. 311-12:

The respondent must also demonstrate that the aptitude or qualification is related in an objective sense to the performance of the employment concerned. McIntyre J. suggested in Etobicoke that the purpose of the objective test is to determine whether the employment requirement is "reasonably necessary" to assure the performance of the job.

What is reasonably necessary is a question of fact to be determined by a Board of Inquiry subject to appeal proceedings as authorized by the human rights code. All the circumstances must be taken into account. I would include any agreement or collective agreement between the employer and employees. This is a factor, but I agree with L'Heureux-Dubé J. that in this case, for the reasons that she gives, it has little or no weight.

I would allow the appeal because the Board of Inquiry found on the evidence that the connection between the respondent university's objective and its mandatory retirement policy was weak. Moreover, the Board found that there were other more reasonable means for the respondent university to achieve its objectives. No valid reason for disturbing these findings has been made out. My colleague Cory J. correctly points out that by virtue of s. 33(2) of the IRPA, an appellate court is empowered to substitute its opinion for that of the Board of Inquiry. That does not, however, justify a conclusion that no attention need be paid to findings of fact by the Board. As the trier of fact, the Board of Inquiry is entitled to the benefit of the principle that its findings should not be set aside unless they were induced by an error of law or for some other valid reason are shown to be clearly wrong. The assessment of the facts by the trier of fact should not be disturbed by an appellate court in the absence of a palpable and overriding error. See Stein v. The Ship "Kathy" K, 1975 146 (SCC), [1976] 2 S.C.R. 802; Ontario (Attorney General) v. Bear Island Foundation, 1991 75 (SCC), [1991] 2 S.C.R. 570, and Lapointe v. Hôpital Le Gardeur, 1992 119 (SCC), [1992] 1 S.C.R. 351. This principle applies to all triers of fact whether or not they are accorded any special curial deference. There is a natural temptation in this case to apply the findings made by the majority in McKinney to the facts of this case. This is essentially what occurred in the Court of Appeal. This is not what was intended by McKinney. As I stated above, McKinney left a legislature free to decide on mandatory retirement or to delegate that decision to employers and employees. The legislature of Alberta has done that. It provides for a hearing before a Board of Inquiry. A hearing has been held in this matter. To reverse that decision because it fails to conform to McKinney would make a sham out of the hearing process and negate a basic premise on the basis of which McKinney was decided by this Court.

I would dispose of this appeal as proposed by L'Heureux-Dubé J.

Appeal dismissed without costs, L'Heureux‑Dubé, Sopinka and McLachlin JJ. dissenting.

Solicitors for the appellant: Chivers, Greckol & Kanee, Edmonton.

Solicitors for the respondent Governors of the University of Alberta: Field & Field Perraton Masuch, Edmonton.

Solicitor for the respondent Alberta Human Rights Commission: The Alberta Human Rights Commission, Edmonton.