Lavoie v. Canada, 2002 SCC 23, [2002] 1 SCR 769

Lavoie v. Canada, [2002] 1 S.C.R. 769, 2002 SCC 23

Elisabeth Lavoie and Jeanne To‑Thanh‑Hien                                               Appellants

v.

Her Majesty The Queen in Right of Canada and

the Public Service Commission                                                                   Respondents

and between

Janine Bailey                                                                                                    Appellant

v.

Her Majesty The Queen in Right of Canada and

the Public Service Commission                                                                   Respondents

and

Center for Research‑Action on Race Relations                                           Intervener

Indexed as:  Lavoie v. Canada

Neutral citation:  2002 SCC 23.

File No.:  27427.

2001:  June 12; 2002:  March 8.

Present:  McLachlin C.J. and L’Heureux‑Dubé, Gonthier, Iacobucci, Major, Bastarache, Binnie, Arbour and LeBel JJ.

on appeal from the federal court of appeal

Constitutional law — Charter of Rights  — Equality rights — Citizenship — Preference given to Canadian citizens for employment in federal Public Service under Public Service Employment Act — Whether preference on basis of citizenship infringing equality guarantee — If so, whether preference justified — Canadian Charter of Rights and Freedoms, ss. 1 , 15(1)  — Public Service Employment Act, R.S.C. 1985, c. P‑33, s. 16(4) (c).

Canadian citizens receive preferential treatment in federal Public Service employment by virtue of s. 16(4)(c) of the Public Service Employment Act (“PSEA”).  The appointment of qualified persons to the Public Service is the exclusive responsibility of the Public Service Commission, as is the exercise of discretion to prefer Canadian citizens under s. 16(4)(c).  Staffing takes place by either open or closed competition, the difference being that closed competitions are restricted to existing employees of the Public Service.  Open competitions generally involve three stages:  the inventory stage, in which persons submit applications to the Commission for general consideration; the referral stage, in which the Commission responds to departmental staffing requests by referring qualified applicants to the requesting department; and the selection stage, in which the requesting department prepares an eligibility list from the list of qualified referrals and chooses from the eligibility list.  The citizenship preference at issue in this appeal occurs at the referral stage of open competitions.  The appellants, foreign nationals who sought employment in the Public Service without having obtained Canadian citizenship, were, in one way or another, disadvantaged by the application of s. 16(4)(c), and challenge this provision as a violation of their equality rights under s. 15(1)  of the Canadian Charter of Rights and Freedoms .  The Federal Court, Trial Division, allowed the s. 15(1)  claim, but held that the legislation could be justified under s. 1  of the Charter .  The Federal Court of Appeal, in a majority judgment, dismissed the appellants’ appeal.

Held (McLachlin C.J. and L’Heureux‑Dubé and Binnie JJ. dissenting):  The appeal should be dismissed.  Section 16(4)(c) of the PSEA is constitutional.

Per Gonthier, Iacobucci, Major, and Bastarache JJ.:  Section 16(4)(c) of the PSEA infringes s. 15(1)  of the Charter .  The impugned provision conflicts with the purpose of s. 15(1) , which is to prevent the violation of essential human dignity and freedom through the imposition of disadvantage, stereotyping, or political or social prejudice, and to promote a society in which all persons enjoy equal recognition at law as human beings or as members of Canadian society, equally capable and equally deserving of concern, respect and consideration.

The integrated approach to s. 15(1)  as set out in Law v. Canada (Minister of Employment and Immigration) involves three broad inquiries.  As to the first and second inquiries, the impugned law draws a clear distinction between citizens and non‑citizens, and citizenship constitutes an analogous ground of discrimination under s. 15(1) .  The third inquiry, which determines whether the distinction is discriminatory, assesses the subjective experience of the claimant against an objective standard, having regard to four contextual factors.  Of these four, the second factor explores the extent to which differential treatment may in fact be acceptable under s. 15(1) :  where there is a genuine relationship between the ground upon which the claim is based and the nature of the differential treatment, it may be acceptable to make certain legislative distinctions.  In the context of laws whose very raison d’être is the definition of citizenship (as in this case), the assertion that citizens and non‑citizens are so differently situated that they do not merit equal treatment, and that citizenship is a relevant (and indeed necessary) category on which unequal treatment is based, goes beyond what is contemplated in Law.  The law or government action must take into account the particular situation of those affected, including any relative advantage or disadvantage.  In this case, to the extent non‑citizens are “differently situated” than citizens, it is only because the legislature has accorded them a unique legal status.  The distinction is not made on the basis of any actual personal differences between individuals.  If anything, the distinction places an additional burden on an already disadvantaged group.  Such a distinction is impossible to square with this Court’s finding in Andrews v. Law Society of British Columbia.  All three remaining contextual factors further militate in favour of a s. 15(1)  violation.  First, while the claimants in this case are all relatively well‑educated, it is settled law that non‑citizens suffer from political marginalization, stereotyping and historical disadvantage.  Second, s. 16(4)(c) of the PSEA does not aim to ameliorate the predicament of a group more disadvantaged than non‑citizens; rather, the comparator class in this case enjoys greater status on the whole than the claimant class.  Finally, the nature of the interest in this case — namely employment — is one that warrants constitutional protection.

The Law factors should not however be applied too mechanically.  Whether the law perpetuates the view that non‑citizens are less capable or less worthy of recognition or value as human beings or as members of Canadian society is the overarching question.  The Law methodology requires a contextualized look at how a non‑citizen legitimately feels when confronted by a particular enactment.  That subjective inquiry into human dignity requires the claimant to provide a rational foundation for her experience of discrimination in the sense that a reasonable person similarly situated would share that experience.  In this case, the claimants felt legitimately burdened by the idea that, having made their home in Canada, their professional development was stifled on the basis of their citizenship status.  Freedom of choice in work and employment are fundamental aspects of this society and, perhaps unlike voting and other political activities, should be, in the eyes of immigrants, as equally accessible to them as to Canadian citizens.

The government has demonstrated that, on a balance of probabilities, s. 16(4)(c) is a reasonable limit on equality that can be demonstrably justified in a free and democratic society under s. 1  of the Charter .  The objectives behind s. 16(4) (c) are sufficiently important to justify limiting the appellants’ equality rights.  Canada’s citizenship policy embodies two distinct objectives:  to enhance the meaning of citizenship as a unifying bond for Canadians, and to encourage and facilitate naturalization by permanent residents.  The signal effect of the impugned provisions is not to discourage immigration but to underscore the value of citizenship.  In an era of increased movement across borders, citizenship provides immigrants with a basic sense of identity and belonging.  Parliament has attempted to achieve the goal of enhancing Canadian citizenship in a manner that respects cultural diversity.

With respect to rational connection, Parliament is entitled to some deference as to whether one privilege or another advances a compelling state interest. As to the first objective, Parliament’s view is supported by common sense and widespread international practice, both of which are relevant indicators of a rational connection in this case.  With regard to the second objective, there is a very close relationship between immigration and naturalization rates in Canada, meaning that a high proportion of immigrants choose to naturalize upon meeting the three‑year residency requirement.  The government’s efforts to enhance the value of citizenship can reasonably be assumed to play a role.

The minimum impairment test has been met.  The test asks whether there are less intrusive ways of enhancing the value of citizenship among public servants. Certain features of s. 16(4)(c) render it less intrusive than it might be: it is a preference only and not an absolute bar; it does not apply to closed competition, the most common means of staffing Public Service positions; it only applies to the referral stage of open competition; and dual citizenship is permitted in Canada, such that Canadian law does not burden non‑citizens with a choice between renouncing their foreign citizenship and entering the Public Service.  While certain individuals undoubtedly fall through the cracks of s. 16(4)(c) of the PSEA, it is uncertain whether a reasonable alternative is available that would fill these cracks in a fair, consistent and principled manner.  Parliament has conscientiously considered alternatives to s. 16(4)(c) and has chosen not to pursue them.  The role of this Court is not to order that Parliament should have decided otherwise.

Finally, the infringing effects of s. 16(4)(c) do not outweigh the importance of the objective sought.  The disadvantage to non‑citizens relative to citizens does not appear significant: it is almost as difficult for citizens to enter the Public Service as non‑citizens; promotion via open competition is a distinct possibility for non‑citizens despite their disadvantage relative to their colleagues; and non‑citizens who are members of the Public Service have unfettered access to closed competitions, which are by far the more conventional avenue of Public Service promotion.  Absent greater evidence of the impact on the claimants’ career prospects, the inconvenience suffered is not too high a price to pay for the government’s right to define the rights and privileges of its citizens.

Per Arbour J.:  Section 16(4)(c) of the PSEA does not infringe s. 15(1)  of the Charter .  The appellants have failed to establish that their claim satisfies the third branch of the Law test for assessing equality claims.  The reasonable person in circumstances similar to those of the claimants would, upon consideration of the various contextual factors set out in Law, conclude that s. 16(4)(c) of the PSEA does not offend the essential human dignity of the claimants and therefore does not discriminate.

At the heart of the third Law inquiry is the recognition that not all distinctions resulting in differential treatment at law can properly be said to violate equality rights under s. 15(1)  of the Charter .  An investigation into whether a legal distinction made on enumerated or analogous grounds is discriminatory is vital to that determination.  The appropriate perspective from which to analyse a claim of discrimination has both a subjective and an objective component.  To read out the requirement of an objective component would be to allow a claimant simply to assert without more that his or her dignity has been adversely affected by a law in order to ground a s. 15(1)  claim and, in so doing, would irrevocably damage the Law methodology.  While there may be certain legislative distinctions, such as those made on the basis of race, that can be labelled infringements of s. 15(1)  without the need for a detailed investigation into whether or not they are discriminatory, this is the exception that proves the rule.  In an understandable eagerness to extend equality rights as widely as possible, stripping those rights of any meaningful content must be avoided.  Otherwise, the result will be the creation of an equality guarantee that is far‑reaching but wafer‑thin, leaving equality rights at the mercy of a diluted justificatory analysis under s. 1  in almost every case.  When the subjective‑objective perspective is properly applied as a necessary condition for making a finding of discrimination, it becomes more difficult to establish that one’s equality rights have been infringed.  It also becomes more difficult, having made a finding of discrimination, to establish that the resulting s. 15(1)  violation can be justified.  Freed of the need to guard the integrity of the legislative process against too easy findings of s. 15(1)  infringements, the justificatory analysis under s. 1  will then be conducted with the uncompromising rigour that it was intended to have.  While this approach to s. 15(1)  may blur the distinction between the kinds of considerations that are appropriate under that section and the kinds of considerations that are appropriate under s. 1 , the overlap is to some extent merely a function of the fact that s. 15(1)  contains its own internal limitation:  specifically, its differentiation between legislative distinctions and discrimination.

Virtually all liberal democracies impose citizenship‑based restrictions on access to their public services.  These restrictions indicate widespread international agreement that such restrictions do not implicate the essential human dignity of non‑citizens and that the partial and temporary difference of treatment imposed by these restrictions is not discriminatory.  An analysis of the non‑exhaustive list of contextual factors suggested in Law further militates against a finding that s. 16(4)(c) of the PSEA violates the essential human dignity of reasonable non‑citizens.  First, while in many aspects of their lives, non‑citizens in general suffer from the sort of pre‑existing disadvantage, stereotyping, prejudice, and vulnerability that s. 15(1)  of the Charter  is directed at remedying, there is doubt as to whether these specific claimants suffer from pre‑existing disadvantage.  On the contrary, this is in some ways a case about the maintaining of pre‑existing advantage by the claimants, who want to retain all of the valuable benefits legally accruing to them as members of the European Union and citizens of other countries while claiming similar privileges and benefits afforded to Canadian citizens under an analogous legislative arrangement.  Second, where the ground upon which the claim is made actually corresponds to personal differences that are relevant to the legislative purpose, the claimant will have difficulty in proving a violation of essential human dignity, even if differential treatment on the basis of that ground is unjustifiable in the vast majority of cases.  Citizenship is relevant to the public distribution of benefits to the extent that it tracks the class of people who have taken on correlative or reciprocal duties in exchange for the receipt of the benefits in question, such that the withholding of those benefits from non‑citizens cannot constitute an affront to human dignity.  Use in this case of the analogous ground of citizenship as a basis for legislating differential treatment between individuals is both: (a) unavoidable, inasmuch as legislating over matters of citizenship itself entails differential treatment between citizens and non‑citizens; and (b) appropriate, inasmuch as the ground of citizenship corresponds to real personal differences between the various individuals who would claim benefits from the state.  Finally, the nature and scope of the interests affected by s. 16(4)(c) of the PSEA are not sufficiently vital and large, nor the effects of that provision sufficiently severe and localized, to allow the claimants to successfully make out a violation of their essential human dignity.  The interest at stake here falls considerably short of being an interest in work per se.  Unlike Andrews v. Law Society of British Columbia, this is not a case in which the claimants are simply refused entry into their chosen profession because of their status as non‑citizens.  At most, what s. 16(4) (c) deprives these claimants of is a chance to enter into open competition with others for positions in the federal Public Service.

Per LeBel J.:  Section 16(4)(c) of the PSEA does not violate s. 15  of the Charter .  The appellants’ claim does not meet the third branch of the test designed in Law as the citizenship preference does not affect the essential dignity of non‑citizens.  Whether s. 1 could justify a breach of s. 15  in this case need not be addressed.  However, the approach to the Oakes test must reflect jurisprudential developments  which acknowledge that the minimal impairment branch of the test may leave a significant margin of appreciation as to the selection of the appropriate remedies to Parliament and legislatures, provided they fall within a range of reasonable alternatives.

Per McLachlin C.J. and L’Heureux‑Dubé and Binnie JJ. (dissenting):  Section 16(4)(c) of the PSEA infringes s. 15(1)  of the Charter  in a way that marginalizes immigrants from the fabric of Canadian life.  A law which bars an entire class of persons from certain forms of employment, solely on the grounds of a lack of citizenship status and without consideration of the qualifications or merits of individuals in the group, violates human dignity.  It is Parliament’s task to draft laws in relation to citizenship that comply with s. 15(1) .  Defining Canadian citizenship does not require that Parliament be allowed to discriminate against non‑citizens.  That some of the appellants in this case could have become citizens, but chose not to, does not militate against a finding of discrimination.  That a person could avoid discrimination by modifying his or her behaviour does not negate the discriminatory effect.  The very act of forcing some people to make such a choice violates human dignity, and is therefore inherently discriminatory.

The infringement in this case is not justified under s. 1  of the Charter .  Assuming that enhancing citizenship and encouraging a small class of civil servants to become Canadian citizens are pressing and substantial objectives, the discrimination complained of is not rationally connected to either of these objectives.  First, the impugned provision confers an advantage upon citizens by discriminating against non‑citizens.  Far from being rationally connected to the goal of enhancing citizenship, the impugned provision undermines this goal, by presenting Canadian citizenship as benefiting from discrimination against non‑citizens, a group which this Court has long recognized as a “discrete and insular minority” deserving of protection.  Secondly, the assessment that the citizenship preference seems generally to have worked as an incentive to naturalize is not persuasive.  There is no evidence to suggest that high rates of naturalization were in any way attributable to the citizenship preference.  That the citizenship preference confers only a minimal advantage upon citizens, because it is almost as difficult for citizens to enter the Public Service as non‑citizens, militates against finding a rational connection.  Finally, that citizenship requirements for civil service are a widespread international practice is neither relevant nor indicative of a rational connection.  There is no evidence that other countries with citizenship‑based restrictions on access to Public Service employment share the same objectives as Parliament in this case.

Cases Cited

By Bastarache J.

Applied:  Law v. Canada (Minister of Employment and Immigration), [1999] 1 S.C.R. 497; referred to: Andrews v. Law Society of British Columbia, [1989] 1 S.C.R. 143; Winner v. S.M.T. (Eastern) Ltd., [1951] S.C.R. 887; Chiarelli v. Canada (Minister of Employment and Immigration), [1992] 1 S.C.R. 711; Corbiere v. Canada (Minister of Indian and Northern Affairs), [1999] 2 S.C.R. 203; Eaton v. Brant County Board of Education, [1997] 1 S.C.R. 241; Eldridge v. British Columbia (Attorney General), [1997] 3 S.C.R. 624; Weatherall v. Canada (Attorney General), [1993] 2 S.C.R. 872; Brooks v. Canada Safeway Ltd., [1989] 1 S.C.R. 1219; Reference Re Public Service Employee Relations Act (Alta.), [1987] 1 S.C.R. 313; McKinney v. University of Guelph, [1990] 3 S.C.R. 229; Harrison v. University of British Columbia, [1990] 3 S.C.R. 451; Stoffman v. Vancouver General Hospital, [1990] 3 S.C.R. 483; Egan v. Canada, [1995] 2 S.C.R. 513; Irwin Toy Ltd. v. Quebec (Attorney General), [1989] 1 S.C.R. 927; Thomson Newspapers Co. v. Canada (Attorney General), [1998] 1 S.C.R. 877; R. v. Oakes, [1986] 1 S.C.R. 103; Mathews v. Diaz, 426 U.S. 67 (1976); Sugarman v. Dougall, 413 U.S. 634 (1973); RJR‑MacDonald Inc. v. Canada (Attorney General), [1995] 3 S.C.R. 199; Mow Sun Wong v. Hampton, 435 F.Supp. 37 (1977); Dagenais v. Canadian Broadcasting Corp., [1994] 3 S.C.R. 835.

By Arbour J.

Applied:  Law v. Canada (Minister of Employment and Immigration), [1999] 1 S.C.R. 497; referred to:  Andrews v. Law Society of British Columbia, [1989] 1 S.C.R. 143; Smith, Kline & French Laboratories Ltd. v. Canada (Attorney General), [1987] 2 F.C. 359; Re B.C. Motor Vehicle Act, [1985] 2 S.C.R. 486; Weatherall v. Canada (Attorney General), [1993] 2 S.C.R. 872; Egan v. Canada, [1995] 2 S.C.R. 513.

By LeBel J.

Applied:  Law v. Canada (Minister of Employment and Immigration), [1999] 1 S.C.R. 497.

By McLachlin C.J. and L’Heureux‑Dubé J. (dissenting)

Andrews v. Law Society of British Columbia, [1989] 1 S.C.R. 143; Corbiere v. Canada (Minister of Indian and Northern Affairs), [1999] 2 S.C.R. 203; Little Sisters Book and Art Emporium v. Canada (Minister of Justice), [2000] 2 S.C.R. 1120, 2000 SCC 69; Law v. Canada (Minister of Employment and Immigration), [1999] 1 S.C.R. 497; Adler v. Ontario, [1996] 3 S.C.R. 609; R. v. Oakes, [1986] 1 S.C.R. 103; Egan v. Canada, [1995] 2 S.C.R. 513; M. v. H., [1999] 2 S.C.R. 3; Irwin Toy Ltd. v. Quebec (Attorney General), [1989] 1 S.C.R. 927; McKinley v. BC Tel, [2001] 2 S.C.R. 161, 2001 SCC 38; Reference Re Public Service Employee Relations Act (Alta.), [1987] 1 S.C.R. 313; Vriend v. Alberta, [1998] 1 S.C.R. 493.

Statutes and Regulations Cited

Canadian Charter of Rights and Freedoms, ss. 1 , 2 (b) 7, 8, 15(1).

Canadian Citizenship Act, S.C. 1946, c. 15.

Civil Service Act, R.S.C. 1906, c. 16.

Civil Service Act, S.C. 1960‑61, c. 57, s. 40(1)(a), (b), (c).

Civil Service Act, 1918, S.C. 1918, c. 12, ss. 38, 41(1).

Civil Service Amendment Act, 1908, S.C. 1908, c. 15, s. 14.

International Covenant on Civil and Political Rights, 999 U.N.T.S. 171 (1966), Art. 25(c).

Public Service Employment Act, R.S.C. 1985, c. P‑33, s. 16(4) (c).

Public Service Employment Act, S.C. 1966‑67, c. 71.

Public Service Staff Relations Act, R.S.C. 1985, c. P‑35 , Sch. I, Part I.

Universal Declaration of Human Rights, G.A. Res. 217 A (III), U.N. Doc. A/810, at 71 (1948), Art. 21(2).

Authors Cited

Berezowski, Nan M., and Benjamin J. Trister.  Citizenship 1996.  Scarborough, Ont.: Carswell, 1996.

Canada.  Department of Justice.  Equality Issues in Federal Law:  A Discussion Paper.  Ottawa:  Department of Justice, 1985.

Canada.  House of Commons.  House of Commons Debates, vol. II, 1st Sess., 20th Parl., October 22, 1945, pp. 1335 et seq.

Canada.  House of Commons.  Report of the Standing Committee on Citizenship and Immigration.  Canadian Citizenship:  A Sense of Belonging, June 1994.

Canada.  House of Commons.  Sub‑Committee on Equality Rights.  Equality for All:  Report of the Parliamentary Committee on Equality Rights, 1985.

Canada.  House of Commons.  Sub-Committee on Equality Rights.    Toward Equality:  The Response to the Report of the Parliamentary Committee on Equality Rights, 1986.

Hogg, Peter W.  Constitutional Law of Canada, vol. 2, loose‑leaf ed.  Scarborough, Ont.:  Carswell, 1992 (updated 2000, release 1).

Kymlicka, Will.  Multicultural Citizenship:  A Liberal Theory of Minority Rights.  Oxford:  Clarendon Press, 1995.

Schuck, Peter H.  “The Re‑Evaluation of American Citizenship” (1997), 12 Geo. Immigr. L.J. 1.

Sharpe, Robert J.  “Citizenship, the Constitution Act, 1867 , and the Charter ”.  In William Kaplan, ed., Belonging:  The Meaning and Future of Canadian Citizenship.  Montreal & Kingston:  McGill‑Queen’s University Press, 1993, 221.

APPEAL from a judgment of the Federal Court of Appeal, [2000] 1 F.C. 3, 174 D.L.R. (4th) 588, 242 N.R. 278, 64 C.R.R. (2d) 189, [1999] F.C.J. No. 754 (QL), affirming a judgment of the Trial Division, [1995] 2 F.C. 623, 95 F.T.R. 1, 125 D.L.R. (4th) 80, 31 C.R.R. (2d) 109, 95 C.L.L.C. ¶210‑023, [1995] F.C.J. No. 608 (QL).  Appeal dismissed, McLachlin C.J. and L’Heureux‑Dubé and Binnie JJ. dissenting.

David J. Jewitt, for the appellants Elisabeth Lavoie and Jeanne To‑Thanh‑Hien.

Andrew Raven and David Yazbeck, for the appellant Janine Bailey.

Graham R. Garton, Q.C., and Yvonne Milosevic, for the respondents.

Joanne St. Lewis and Milton James Fernandes, for the intervener.

The reasons of McLachlin C.J. and L’Heureux-Dubé and Binnie JJ. were delivered by

1                                   The Chief Justice and L’Heureux-Dubé J. (dissenting) — We agree with Bastarache J. that s. 16(4) (c) of the Public Service Employment Act, R.S.C. 1985, c. P-33  (“PSEA ”), infringes s. 15(1)  of the Canadian Charter of Rights and Freedoms  in a way that marginalizes immigrants from the fabric of Canadian life, and endorse his reasons on this point.  In our view, this conclusion is mandated by Andrews v. Law Society of British Columbia, [1989] 1 S.C.R. 143, which we find to be indistinguishable on the question of discrimination.  We respectfully disagree, however, that s. 1  justifies the infringement as a “reasonable limit on equality” (para. 21).

I.              Question 1: Does Section 16(4)(c) of the PSEA Infringe Section 15(1) of the Charter ?

2                                   Violation of s. 15(1) depends on finding a discriminatory distinction, based on an enumerated or analogous ground.  On both counts, this case is similar to Andrews.  First, the distinction at issue is made on the basis of citizenship, the very ground held to be analogous in Andrews.  Once identified, an analogous ground stands as “a constant marker of potential legislative discrimination” and need not be established again in subsequent cases: Corbiere v. Canada (Minister of Indian and Northern Affairs), [1999] 2 S.C.R. 203, at paras. 7-10; see also Little Sisters Book and Art Emporium v. Canada (Minister of Justice), [2000] 2 S.C.R. 1120, 2000 SCC 69, at para. 119, per Binnie J.  The distinction here at issue, denial of employment opportunity, is the same distinction recognized in Andrews.  A discriminatory distinction is one that violates human dignity: Law v. Canada (Minister of Employment and Immigration), [1999] 1 S.C.R. 497.  Law affirms Andrews, and must therefore be taken as finding that a law which bars an entire class of persons from certain forms of employment, solely on the grounds of a lack of citizenship status and without consideration of the qualifications or merits of individuals in the group, violates human dignity.  See Andrews, supra, per McIntyre J., at p. 183.

3                                   It is argued that Andrews is distinguishable as having turned on provincial legislation, whereas this case centers upon federal legislation under the citizenship power.  Parliament, it is said, must be granted significant leeway in determining the rights and privileges attached to citizenship if this power is not to be trivialized.  This argument, it seems to us, sets up a false dichotomy between Parliament’s right to make laws regarding citizenship and Parliament’s duty to ensure that its laws conform to s. 15(1).  Parliament need not choose between legislating with respect to citizenship and discrimination.  Rather, it is Parliament’s task to draft laws in relation to citizenship that comply with s. 15(1).  This leaves ample scope for the exercise of the citizenship power, so long as Parliament does not make distinctions that unjustifiably violate human dignity: Law, supra.  We cannot agree that defining Canadian citizenship requires that Parliament be allowed to discriminate against non-citizens.

4                                   It is also argued that Andrews involved an outright ban on a form of employment by non-citizens, whereas this case is closer to a lost chance of employment.  Again, the distinction eludes us.  In both cases, non-citizens were denied employment opportunities, solely because of their citizenship status and for no other reason.

5                                   Finally, much has been made of the fact that some of the appellants in this case could have become citizens, but chose not to.  In our view, this consideration does not militate against a finding of discrimination.  First, such a choice can be attributed to only two of the appellants.  Second, in any event the benefit is denied during the period that is required before a permanent resident can obtain citizenship.  Third, the fact that a person could avoid discrimination by modifying his or her behaviour does not negate the discriminatory effect.  If it were otherwise, an employer who denied women employment in his factory on the ground that he did not wish to establish female changing facilities could contend that the real cause of the discriminatory effect is the woman’s “choice” not to use men’s changing facilities.  The very act of forcing some people to make such a choice violates human dignity, and is therefore inherently discriminatory.  The law of discrimination thus far has not required applicants to demonstrate that they could not have avoided the discriminatory effect in order to establish a denial of equality under s. 15(1).  The Court in Andrews was not deterred by such considerations.  On the contrary, La Forest J. specifically noted that acquiring Canadian citizenship could in some cases entail the “serious hardship” of losing an existing citizenship.  He left no doubt that this hardship was a cost to be considered in favour of the individual affected by the discrimination: Andrews, supra, at p. 201.

II.      Question 2: Is the Breach of Section 15(1) Justified Under Section 1 of the Charter ?

6                                   This brings us to s. 1  of the Charter  and the question of whether the discrimination this law effects is justified in a free and democratic society.  In conducting the s. 1  analysis, “it must be remembered that it is the right to substantive equality and the accompanying violation of human dignity that has been infringed when a violation of s. 15(1)  has been found” (Corbiere, supra, per L’Heureux-Dubé J., at para. 98 (emphasis deleted)).  Indeed, “cases will be rare where it is found reasonable in a free and democratic society to discriminate” (see Adler v. Ontario, [1996] 3 S.C.R. 609, per L’Heureux-Dubé J., at para. 95 (citing Andrews, supra, per Wilson J., at p. 154)).  Discrimination on the basis of non-citizenship will attract close scrutiny.  To quote La Forest J. in Andrews, supra, at p. 201:

If we allow people to come to live in Canada, [we] cannot see why they should be treated differently from anyone else.  Section 15  speaks of every individual.  There will be exceptions no doubt, but these require the rigorous justification provided by s. 1 .

The majority of this Court in Andrews held that the burden of justification in cases such as this is “onerous”.

7                                   This Court has held that in order to invoke the protection of s. 1 , the government must demonstrate that an infringement of the Charter  “is ‘reasonable’ and ‘demonstrably justified in a free and democratic society’” (R. v. Oakes, [1986] 1 S.C.R. 103, at p. 135).  The test that this Court has fashioned to make such a determination requires that (1) the objective of the legislation be pressing and substantial; (2) the rights violation be rationally connected to the aim of the legislation; (3) the impugned provision minimally impair the Charter  guarantee; and (4) the effect of the measure be proportional to its objective so that the attainment of the legislative goal is not outweighed by the abridgment of the right (see Egan v. Canada, [1995] 2 S.C.R. 513, at para. 182, per Iacobucci J. (citing Oakes, supra, at pp. 138-39)).

8                                   We agree with the majority that two objectives can be attributed to the impugned legislation: encouraging non-citizens to naturalize, and enhancing citizenship.  We note in passing that the majority reasons appear to restate or modify  the objectives as the s. 1  analysis progresses.  However, since our disagreement turns on the rational connection component of the s. 1  analysis, this point need not detain us here.  In our view, when two objectives are accepted as pressing and substantial, the s. 1  analysis must be applied to each of them separately, so that an individual is not left guessing as to the state objective purported to justify the infringement of his or her Charter  rights.  We will consider these objectives in turn.

9                                   Assuming that “enhancing citizenship” and encouraging a small class of civil servants to become Canadian citizens are pressing and substantial objectives, thereby satisfying the first requirement of the Oakes test, we are not satisfied that the discrimination complained of is rationally connected to either of these objectives.  In order to satisfy this portion of the s. 1  test, the government must show that the impugned law is “carefully designed to achieve the objective in question”; it must not be “arbitrary, unfair or based on irrational considerations” (Oakes, supra, at p. 139).

10                               It is argued that a law giving citizens an advantage in connection with Public Service employment is rationally connected to the legislative objective of enhancing citizenship.  With respect, we think this characterization misses the crucial point, which is that the impugned provision confers an advantage upon citizens by discriminating against non-citizens.  Far from being rationally connected to the goal of enhancing citizenship, the impugned provision undermines this goal, by presenting Canadian citizenship as benefiting from, as nourished by, discrimination against non-citizens, a group which this Court has long recognized as a “discrete and insular minority” deserving of protection (Andrews, supra, at p. 152).  It seems to us that such reasoning is incompatible with the view of Canadian citizenship as defined by “tolerance”, “a belief in equality” and “respect for all individuals” (Citizenship and Immigration Canada, available at <http://www.cic.gc.ca/english/about/faq/ask-23e.html> and <http://www.cic.gc.ca/english/newcomer/welcome/wel-03e.html>).  As the majority points out at para. 52, “[i]mmigrants come to Canada expecting to enjoy the same basic opportunities as citizens”.  Accordingly, the majority argues that work and employment, which are “fundamental aspects” of Canadian society, “should be . . . as equally accessible to them as to Canadian citizens” and that “[d]iscrimination in these areas has the potential to marginalize immigrants from the fabric of Canadian life and exacerbate their existing disadvantage in the Canadian labour market”.

11                               To put it another way, we fail to see how the value of Canadian citizenship can in any way be enhanced by a law that the majority concedes discriminates against non-citizens, particularly given La Forest J.’s recognition in Andrews, supra, at p. 197, that “[o]ur nation has [historically] drawn strength from the flow of people to our shores”.  In this regard, we also find Linden J.A.’s evolutionary view of Canadian citizenship compelling:  “The broader, inclusive, Canadian view of citizenship which has emerged brings with it important legal ramifications . . . . It is a tool of equality, not exclusion” (Lavoie v. Canada, [2000] 1 F.C. 3 (C.A.), at para. 121).  A law that favours the relatively advantaged group of Canadian citizens over the relatively disadvantaged group of non-citizens serves to undermine, not further, the value of Canadian citizenship, based as it is on principles of inclusion and acceptance.  The anomaly of this reasoning is accentuated by the majority’s contention that the citizenship preference only minimally advantages citizens.  The notion that a trivial advantage, secured at the cost of violating s. 15(1)’s equality guarantee, could enhance citizenship, is difficult for us to fathom.

12                               Moreover, the government presented no evidence that excluding non-citizens in fact furthers the objective of enhancing citizenship.  The majority addresses this difficulty by arguing, at para. 59, that “Parliament is entitled to some deference as to whether one privilege or another advances a compelling state interest”.  But judicial deference alone cannot establish a rational connection.  In M. v. H., [1999] 2 S.C.R. 3, at paras. 78-79, Iacobucci J., writing for the majority, stated:

As Cory J. stated in Vriend, supra, at para. 54: “The notion of judicial deference to legislative choices should not . . . be used to completely immunize certain kinds of legislative decisions from Charter  scrutiny.”

Under s. 1, the burden is on the legislature to prove that the infringement of a right is justified.  In attempting to discharge this burden, the legislature will have to provide the court with evidence and arguments to support its general claim of justification.

In that case, this Court concluded the impugned legislation was not saved by s. 1 after finding “no evidence” of a rational connection (M. v. H., supra, at paras. 109-15).

13                               In previous decisions, this Court has duly granted a greater degree of deference to legislation with a valid objective related to social justice, for example, legislation that promotes the protection of a socially vulnerable group (see Irwin Toy Ltd. v. Quebec (Attorney General), [1989] 1 S.C.R. 927).  It follows from this principle that “[a] less deferential stance should be taken and a greater onus remain on the state to justify its encroachment on the Charter  right”, where, as here, “the nature of the infringement lies at the core of the rights protected in the Charter  and the social objective is meant to serve the interest of the majority” (see Adler, supra, per L’Heureux-Dubé J., at para. 95).  As our colleague Bastarache J. concedes, at para. 53,  “[s. 16(4) (c)] does not promote the interests of a vulnerable group, is not premised on particularly complex social science evidence, and interferes with an activity (namely employment) whose social value is relatively high”.  Indeed, this Court has recognized that employment is a fundamental aspect of an individual’s life and an essential component of identity, personal dignity, self-worth and emotional well-being (see McKinley v. BC Tel, [2001] 2 S.C.R. 161, 2001 SCC 38, at para. 53 (citing Reference Re Public Service Employee Relations Act (Alta.), [1987] 1 S.C.R. 313, per Dickson C.J., at p. 368)).  Given the onus on the state in this case, we are of the view that it is incumbent on the government to offer at least some evidence that the impugned law furthers the objective of promoting the value of Canadian citizenship before the s. 15(1) violation can be justified.

14                               We conclude that in this case, as in M. v. H., supra, and Vriend v. Alberta, [1998] 1 S.C.R. 493, the infringing measure was antithetical to the objective sought to be achieved.  There is no rational connection between the discrimination effected by s. 16(4) (c) of the PSEA  and the objective of enhancing citizenship.

15                               We now turn to the second objective.  It is argued that a law giving citizens an advantage in connection with Public Service employment is rationally connected to the legislative objective of encouraging naturalization.  At first blush, this may seem plausible: since non-citizens could avail themselves of this same benefit by naturalizing, they have an incentive to do so.  However, it seems to us this benign characterization fails to capture the significance of the government’s position.  In essence the government’s argument is this: we discriminate against people lawfully in Canada so that they will value citizenship and be motivated to become citizens, at which point we will cease to discriminate against them.  As noted above, the discrimination in question is at odds with the values of tolerance, equality and respect that the government acknowledges lie at the heart of Canadian citizenship.

16                               Moreover, as with the first objective asserted, the government introduced no evidence capable of supporting the contention that the discrimination complained of actually works as an incentive to naturalize.  We are not persuaded by the majority’s assessment, at para. 60, that the citizenship preference “seems generally to have worked”.  That the impugned provision was in effect at a time when the naturalization rate was high does not prove that the impugned provision caused the high naturalization rate.  There is no evidence to suggest that high rates of naturalization were in any way attributable to the citizenship preference.

17                               Indeed, the majority’s assertion that the citizenship preference confers a minimal advantage upon citizens militates against finding a rational connection.  That it is “almost as difficult for citizens to enter the Public Service as non-citizens” (emphasis deleted) and “the latter’s disadvantage relative to the former does not appear significant”, as the majority believes (para. 71), works against the notion that the citizenship preference causally contributed to high rates of naturalization.  Still less can it be argued that the citizenship preference was “carefully designed” to achieve the objective of encouraging non-citizens to naturalize (Oakes, supra, at p. 139).

18                               For these reasons, we conclude there is no rational connection between s. 16(4) (c) of the PSEA  and the objective of encouraging non-citizens to naturalize.

19                               Finally, we would add this.  The fact that citizenship requirements for civil service are a “widespread international practice” (para. 59) is neither relevant nor indicative of a rational connection in this case.  There is no evidence that other countries with citizenship-based restrictions on access to Public Service employment share the same objectives as Parliament in this case.  In fact, the government itself argued at trial that most democratic countries with citizenship requirements on Public Service employment have different immigration policies and realities and therefore base their citizenship requirements on different legislative objectives.  For instance, the respondents’ record suggests that Germany’s citizenship requirement is tied to concerns about loyalty and commitment: “The fundamental duty of the civil servant is derived from the concept of a civil servant’s position as one of service and loyalty”. Moreover, the government argued that New Zealand’s citizenship restriction on national security postings “is tailored to a narrower legislative purpose which would not serve the broader citizenship objectives of the Canadian Parliament” (Lavoie v. Canada, [1995] 2 F.C. 623 (T.D.), at pp. 670-71).  While we take no position on why other countries impose citizenship-based restrictions, we do not believe the practice of these countries can form the basis of our decision without at least some evidence that they share similar objectives as Parliament.  In arriving at this conclusion, we place no restrictions on Parliament’s ability to impose citizenship-based restrictions on certain Public Service jobs (such as positions that relate to a political function or national security) as legitimate qualifications of employment.

20                               Since both of the stated objectives fail the rational connection component of the test set out in Oakes, the infringement of s. 15(1)  of the Charter  cannot be justified under s. 1 .  We would allow the appeal with costs throughout and declare s. 16(4) (c) of the PSEA  to be of no force and effect.

The judgment of Gonthier, Iacobucci, Major and Bastarache JJ. was delivered by

21                               Bastarache J. — Canadian citizens receive preferential treatment in federal Public Service employment by virtue of s. 16(4) (c) of the Public Service Employment Act, R.S.C. 1985, c. P-33  (“PSEA ”).  The appellants, foreign nationals who sought employment in the Public Service without having obtained Canadian citizenship, challenge this provision as a violation of their equality rights under s. 15(1)  of the Canadian Charter of Rights and Freedoms .  They claim the preference withholds a benefit from them that is enjoyed by Canadian citizens and, in so doing, undermines their essential human dignity.  In addition, they claim such treatment cannot be justified as a reasonable limit on equality under s. 1  of the Charter .  For the reasons that follow, I conclude that s. 16(4) (c) of the PSEA  violates s. 15(1)  of the Charter  but can be justified as a reasonable limit on equality under s. 1 .

I.        Factual Background

22                            Canadian citizens have enjoyed privileged access to the federal Public Service ever since the enactment of amendments to the Civil Service Act, R.S.C. 1906, c. 16; see The Civil Service Amendment Act, 1908, S.C. 1908, c. 15.  The amended Act replaced a process of patronage appointment with the principle of open competition to the Civil Service; it originally applied only to the Ottawa region, but was extended country-wide in 1918 (The Civil Service Act, 1918, S.C. 1918, c. 12, s. 38).  The original version of the present citizenship preference was inscribed in s. 14 of the 1908 Act, which provided:  “No person shall be admitted to [the Civil Service] unless he is a natural-born or naturalized British subject, and has been a resident of Canada for at least three years”.  This provision made citizenship a prerequisite for Civil Service employment, not a preference; under the 1918 Act, it could only be waived by Order-in-Council (s. 41(1)).  In 1961, the requirement of citizenship was changed to a preference for Canadian citizens in consideration for open competition; see Civil Service Act, S.C. 1960-61, c. 57, s. 40(1)(c).  The 1961 Act also accorded preference to veterans and widows of veterans (s. 40(1)(a) and (b)).  All three preferences continue to this day, despite a complete revision to the legislative scheme in 1967, at which time the Civil Service Act was repealed and the Public Service Employment Act, S.C. 1966-67, c. 71, was enacted in its place.  The 1967 Act also established the Public Service Commission (“PSC” or “Commission”), which took on a modified role of its predecessor, the Civil Service Commission.  No relevant amendments have taken place since that time.

23                            Today, the appointment of qualified persons to the Public Service is the exclusive responsibility of the PSC.  By authority of the PSEA , the Commission makes appointments in all government departments and agencies that do not have separate staffing authority under specific legislation; see Public Service Staff Relations Act, R.S.C. 1985, c. P-35 , Part I of Schedule I.  Staffing takes place either by open or closed competition, the difference being that closed competitions are restricted to existing employees of the Public Service.  Open competitions generally involve three stages:  the inventory stage, in which persons submit applications to the Commission for general consideration; the referral stage, in which the Commission responds to departmental staffing requests by referring qualified applicants to the requesting department; and the selection stage, in which the requesting department prepares an eligibility list from the list of qualified referrals and, assuming the competition is not cancelled, chooses from the eligibility list.

24                            The citizenship preference at issue in this appeal occurs at the referral stage of open competitions.  This means that non-citizens are eligible (and indeed encouraged) to submit their resumes to the Commission for consideration, and that, for the sake of this appeal, non-citizens who are referred by the Commission face no disadvantage compared to citizens.  It also means that non-citizens enjoy the same privileges as citizens with respect to closed competitions; these are the principal means by which the Public Service fills its staffing needs.  Finally, the citizenship preference is just that:  a preference.  Non-citizens are routinely referred to open competition where, in the opinion of the Regional PSC Director, there are insufficient qualified Canadians to fill the particular position; see PSC, Personnel Management Manual (1985), ss. 5.1(2) and 5.4(3).  Given the demand for Public Service employment among the citizenry, such referrals are rare in proportion to the overall number.  While no statistics are kept on the number of non-citizens referred to open competition, the Regional Director of the National Capital Region of the PSC, Mr. Peter Stewart, recalled approximately “a dozen” such referrals in a one-year period.  This represented less than 2 percent of the total number of referrals that year, although it is not known what percentage of the referred candidates were successful.

25                            The appellants are all foreign nationals who, in one way or another, were disadvantaged by the application of s. 16(4) (c).  Janine Bailey is a Dutch citizen and a citizen of the European Union (“EU”); she moved to Canada with her husband in 1986 and was admitted as a permanent resident.  Although eligible to apply for Canadian citizen in 1989, Bailey chose not to do so because it would have meant relinquishing her Dutch citizenship.  She testified that she had emotional ties to the Netherlands and may one day have to return to take care of family members.  Despite her foreign citizenship, Bailey was appointed by open competition to a three-month term position as a shift clerk with the Canada Employment and Immigration Commission.  This employment was extended through a succession of term appointments, during which time she repeatedly sought promotion through open and closed competition.  In the closed competitions, she was screened out at least three times for lack of relevant experience and/or knowledge; a fourth time she successfully obtained a position as an Immigration Examining Officer (PM-01).  She also had mixed success in open competition.  In one case, she was screened out of a PM-02 competition to which 144 Canadians were referred; in another, she was screened out of a PM-03 position to which 40 Canadians were referred.  Both times she registered a complaint with the PSC, which ruled its discretion had been properly exercised under s. 16(4) (c).  Eventually, Bailey was referred by open competition to a PM-03 competition but failed to meet the department’s rated requirements.  As of trial, she remained at the PM-01 classification.

26                            Elisabeth Lavoie is an Austrian citizen and a citizen of the EU; like Bailey, she moved to Canada with her husband and was admitted as a permanent resident.  Lavoie also declined to apply for Canadian citizenship because she feared it would jeopardize her Austrian and EU citizenship.  She testified she would become “a foreigner in [her] own country” and would have limited work opportunities.  Unlike Bailey, Lavoie never obtained Public Service employment by open competition; rather, she obtained a short-term contract with the Department of Supply and Services (“DSS”) through a personnel agency, Harrington Temporary Services.  This contract lasted for 22 weeks, during which time Lavoie applied to fill the position permanently.  DSS even submitted a “named referral request” to the PSC on her behalf, formally requesting she be referred as a candidate for the position.  To everyone’s “shock”, the named referral request was refused.  The PSC used its s. 16(4) (c) discretion to refer a Canadian citizen, who was appointed on a term basis the following month.  Lavoie’s contract was thereby terminated, and she sought employment in the provincial Public Service and the private sector.  The following year, the term appointment was not renewed and the position was declared redundant.

27                            Jeanne To-Thanh-Hien is a French citizen born in Vietnam; she moved to Ottawa in 1987 at the suggestion of her sister, a translator for the federal government.  Unlike Bailey and Lavoie, To-Thanh-Hien obtained Canadian citizenship in 1991 and did not have to relinquish her foreign citizenship to do so.  Before arriving in Canada, To-Thanh-Hien applied for employment as a French-language editor and was referred to the PSC’s Employment Services for Visible Minorities Program.  The Program informed her that all possible efforts would be made to find her a job, but they did not mention the existence of a citizenship preference.  After applying to several government agencies, To-Thanh-Hien was informed that the Program could not do anything for her until she became a Canadian citizen.  She did, however, manage to obtain temporary work with several government departments, notably Agriculture Canada, both on her own and through a personnel agency.  In the spring of 1988, To-Thanh-Hien applied unsuccessfully for two open competitions; in one of them, she felt encouraged to apply but was eventually told her citizenship precluded referral.  Eventually, she was appointed in an open competition to a secretarial position, something for which she felt overqualified.  She was again successful in an open competition in 1993, by which time she had obtained Canadian citizenship, and held a term position with the Department of Human Resources until 1994.  As of trial, To-Thanh-Hien was a project coordinator for the Employment Equity Branch of Human Resources Development Canada.

28                            In all three cases, the appellants sought declaratory relief and damages on constitutional grounds due to the application of s. 16(4) (c) of the PSEA , arguing it breached s. 15(1)  of the Charter .  In the Federal Court, Trial Division, Wetston J. allowed the s. 15(1)  claim, but held that the legislation could be justified under s. 1  of the Charter .  He rejected less intrusive alternatives on the basis of administrative inconvenience and held, further, that Parliament was entitled to a margin of deference in balancing the state interest in enhancing citizenship against the non-citizen’s interest in pursuing Public Service employment.  Wetston J. did not have the benefit of Law v. Canada (Minister of Employment and Immigration), [1999] 1 S.C.R. 497, when rendering his s. 15(1)  judgment.  On appeal, Marceau J.A. dismissed the s. 15(1)  claim on the grounds that citizenship was not a suspect marker of discrimination in the context of federal laws defining the rights and privileges of citizenship.  Desjardins J.A. concurred in the result, but for essentially the same reasons as the trial judge:  that the law violated s. 15(1)  but was justified under s. 1  of the Charter .  Both Marceau and Desjardins JJ.A. applied Law, supra, to the s. 15(1)  analysis.  Linden J.A. dissented, holding that the law violated s. 15(1)  and failed both the minimum impairment and final balancing stages of s. 1 .  The appellants were granted leave to appeal to this Court on May 25, 2000, [2000] 1 S.C.R. xiv.

II.    Relevant Statutory Provisions

29                               Public Service Employment Act, R.S.C. 1985, c. P-33 

16.  . . .

(4)  Where, in the case of an open competition, the Commission is of the opinion that there are sufficient qualified applicants who are

. . .

(c) persons who are Canadian citizens who do not come within paragraph (a) or (b),

to enable the Commission to establish an eligibility list in accordance with this Act, the Commission may confine its selection of qualified candidates under subsection (1) to the applicants who come within paragraph (a), paragraphs (a) and (b) or paragraphs (a), (b), and (c).

Canadian Charter of Rights and Freedoms 

1.  The Canadian Charter of Rights and Freedoms  guarantees the rights and freedoms set out in it subject only to such reasonable limits prescribed by law as can be demonstrably justified in a free and democratic society.

                                                                   . . .

15.  (1) Every individual is equal before and under the law and has the right to the equal protection and equal benefit of the law without discrimination and, in particular, without discrimination based on race, national or ethnic origin, colour, religion, sex, age or mental or physical disability.

III.    Judgments Below

A.           Federal Court, Trial Division, [1995] 2 F.C. 623

30                               As noted above, Wetston J. did not have the benefit of Law, supra, when rendering his s. 15(1)  judgment.  At the time, the controlling authority on s. 15(1)  was Andrews v. Law Society of British Columbia, [1989] 1 S.C.R. 143, which set forth three criteria for a violation of equality rights: (1) the existence of a legislative distinction based on a personal characteristic; (2) the distinction being based on an enumerated or analogous ground; and (3) the existence of discrimination in the substantive sense through the imposition of a burden or denial of a benefit.  The trial judge noted that s. 16(4) (c) created a distinction on its face, and that the distinction was based on the prohibited ground of “citizenship”:  see Andrews, supra, at p. 183.  He then considered the respondents’ argument that while citizenship could not be used as a proxy for merit in provincial laws governing the professions, it constituted an acceptable ground of discrimination in the context of federal laws defining the rights and privileges of citizenship.  After canvassing American authority on this subject (which generally allows citizenship preference in the context of the federal Public Service, but scrutinizes it more strictly at the state level), he concluded that Canadian equality law would not countenance such a result.  Rather, the relevant question in assessing substantive discrimination was whether the law imposes a burden or denies a benefit based on a prohibited ground; in his view, s. 16(4) (c) did so.

31                               At the s. 1  stage, Wetston J. reviewed the legislative history of the provision and concluded it stemmed from two objectives:  to enhance the meaning, value and importance of citizenship in the context of federal Public Service employment and, as part of this, to provide an incentive for non-citizens to naturalize.  He went on to hold that the objectives were pressing and substantial in so far as “[a] nation-state clearly has the right, as part of its domestic law, to determine who is a citizen and what rights and obligations may flow from that status” (p. 658; citing Winner v. S.M.T. (Eastern) Ltd., [1951] S.C.R. 887, at pp. 918-19).  Moreover, Wetston J. relied on the evidence of Peter H. Schuck, an American citizenship expert, for the proposition that citizenship serves important political, emotional and motivational purposes.  He rejected the appellants’ argument that concrete evidence of increased naturalization rates was required in order to establish the naturalization objective as pressing and substantial.

32                               With respect to proportionality, the trial judge first held, at p. 664, that “in light of international practice alone, Parliament had a reasonable basis for assuming that the means chosen would achieve the desired ends”.  This brought him to the minimum impairment test, in which he canvassed four alternatives to the current citizenship preference.  The first, an all-out ban on the preference, was rejected on the grounds that Parliament need not adopt “the absolutely least intrusive means of achieving its objective” (p. 667).  The others — a preference for jobs serving a “political function” only (the American state and European models), an exception for permanent residents who seek naturalization as soon as possible (the Australian model), and a preference for jobs affecting “national security” only (the New Zealand model) — were all rejected on the grounds that, in cases requiring a balance of competing interests, it was not the court’s role to second-guess Parliament’s conclusions.  Finally, the trial judge held that any burden inflicted on non-citizens by the preference was not of a serious enough nature to outweigh the salutary effects of the legislation (p. 677).

B.        Federal Court of Appeal, [2000] 1 F.C. 3

33                               On appeal, Marceau J.A. agreed with the respondents that, in contrast to provincial laws such as those impugned in Andrews, qualifications imposed by Parliament with respect to the status of landed immigrants could not be criticized under s. 15(1) .  The reason for this is that in the latter context, there is not sufficient similarity between citizens and non-citizens to give rise to a discrimination claim; the latter have such a special status that “[t]o try to apply equality rights between citizens and non-citizens . . . [would] negate or abolish the concept of citizenship altogether” (para. 11).  In the alternative, Marceau J.A. held that a preference for citizens in Public Service employment was “relevant” to the aims of the PSEA  and, on that basis, acceptable under s. 15(1) .  In his view, the relevancy test helped answer the question whether the impugned provision prejudicially affected the human dignity of the s. 15(1)  claimant:  see Law, supra.  In any case, he concluded, at paras. 25-26, that “[t]he intent to enhance the value of citizenship does not denigrate the landed immigrant in a manner based upon a personal characteristic”, and that the law “cannot . . . be seen objectively as demeaning in any way the human dignity of the appellants or non-citizens generally”.

34                               Both Desjardins J.A. (concurring) and Linden J.A. (dissenting) disagreed with this finding and found a s. 15(1)  violation.  The former agreed with Marceau J.A. that, unlike the legislation considered in Andrews, the citizenship preference in the PSEA  was “another feature of the rights and privileges of Canadian citizens” (para. 58); however, she focused on the history of discrimination against aliens in Canada and concluded, at para. 64, that “the impugned legislation puts, in a serious disadvantageous position, members of a discrete and insular minority and affects them in their search for employment”.  The latter provided a lengthy summary of Law, and had no difficulty finding that the law differentiated between two groups on the basis of an analogous ground.  Focussing mainly on the third branch of Law, Linden J.A. based his finding of discrimination on four points:  (1) the citizenship preference further discriminates against an already disadvantaged group; (2) denying people the chance to work is far more serious than refusing them some monetary benefit or procedural right; (3) derogating from the rights of non-citizens does little to enhance the rights of citizens, and (4) the provision makes no reference to the needs and capacities of the targeted group.  At one point in this discussion, Linden J.A. stated, at para. 167, that “[b]eing told that your ‘kind’ is not permitted to apply for a job, seriously demeans the human dignity of the applicant”.

35                               At the s. 1  stage, Desjardins and Linden JJ.A. parted ways.  The former essentially upheld the findings of the trial judge, although she paid less attention to the final balancing under s. 1 .  In her view, drawing the line with respect to citizenship preference in the Public Service was a “political consideration” (para. 99) which she was not prepared to disturb.  Linden J.A., in dissent, held that the trial judge made a palpable and overriding error in not finding that the citizenship preference was also enacted to address concerns of commitment and loyalty which arise when non-citizens are hired to serve the Canadian public.  In his view, this objective was apparent from a 1908 speech in the House of Commons and a 1985 discussion paper circulated by the Minister of Justice.  Although he found such an objective was not pressing and substantial, he conceded that the two objectives identified by the trial judge “may warrant some compromise of equality rights” (para. 194).  Linden J.A. proceeded to decide the case on minimum impairment grounds, finding no less than five ways in which Parliament could have achieved its objectives less intrusively.  While conceding the legislation was “not as bad as it could be” (para. 208), Linden J.A. observed that Parliament has never turned its mind to the above alternatives other than an all-out ban on the preference and, for that reason, has not passed legislation which is “carefully tailored to minimize impairment of the Charter  right” (para. 208).  He would have reached this conclusion no matter what standard of deference was applied to the legislation.  Finally, with respect to the final balancing test, Linden J.A. provided five reasons why the deleterious effects of the legislation outweighed its benefits:  (1) the legislation excludes up to 600,000 people from nearly 250,000 jobs and, on an individual level, “halts individual growth and opportunity at self-attainment in order to ensure that cooks, deckhands and curators, and for that matter interpreters, prison guards and secretaries, are Canadian citizens” (para. 215); (2) the legislation undermines people’s legitimate reasons for maintaining a connection with their homeland, something Canada already recognizes by allowing dual citizenship; (3) any benefits in terms of commitment and loyalty to Canada are remnants of a bygone era; (4) there is no evidence that, as compared to the burdens of obtaining citizenship, the benefits of enhanced citizenship and increased naturalization actually accrue; and (5) one of the effects of the legislation is to undermine the merit principle underlying the statute as a whole.

IV.    Issues

36                            The following two constitutional questions were stated by the Chief Justice on October 31, 2000:

1.    Does paragraph 16(4)(c) of the Public Service Employment Act, R.S.C. 1985, c. P‑33 , on its own or in its effect, discriminate against persons on the basis of citizenship by providing a preference to Canadian citizens over non‑citizens in open competitions in the federal public service, contrary to section 15(1)  of the Canadian Charter of Rights and Freedoms ?

2.    If the answer to question one is yes, is the discrimination a reasonable limit prescribed by law which can be demonstrably justified in a free and democratic society under section 1  of the Canadian Charter of Rights and Freedoms ?

V.     Analysis

A.        Section 15(1)

37                                  This Court has twice considered the relationship between citizenship and s. 15(1)  of the Charter .  The first time was Andrews, supra, which concerned a provincial law barring non-citizens from access to the legal profession; the law was struck down as a violation of s. 15(1)  and was not saved under s. 1 .  The second time was Chiarelli v. Canada (Minister of Employment and Immigration), [1992] 1 S.C.R. 711, which, by contrast, involved a federal law authorizing the deportation of permanent residents convicted of serious criminal offences; as s. 6  of the Charter  specifically authorized differential treatment of non-citizens for immigration purposes, the law was held not to be discriminatory (p. 736).  This case has much in common with both Andrews and Chiarelli.  Like Andrews, it involves differential treatment in employment that is not explicitly authorized by the Charter ; like Chiarelli, it involves a federal law that is part of a recognized package of privileges conferred on Canadian citizens.  This combination of factors makes it difficult to decide whether, at the end of the day, the law conflicts with the purpose of s. 15(1)  of the Charter .  Based on this Court’s recent s. 15(1)  jurisprudence, I conclude that it does.

38                                  The integrated approach to s. 15(1)  is set forth in Law, supra.  In that case, Iacobucci J. summarized, at para. 88, the proper approach to s. 15(1)  as follows:

. . . a court that is called upon to determine a discrimination claim under s. 15(1)  should make the following three broad inquiries:

(A)     Does the impugned law (a) draw a formal distinction between the claimant and others on the basis of one or more personal characteristics, or (b) fail to take into account the claimant’s already disadvantaged position within Canadian society resulting in substantively differential treatment between the claimant and others on the basis of one or more personal characteristics?

(B)     Is the claimant subject to differential treatment based on one or more enumerated and analogous grounds?

and

(C)     Does the differential treatment discriminate, by imposing a burden upon or withholding a benefit from the claimant in a manner which reflects the stereotypical application of presumed group or personal characteristics, or which otherwise has the effect of perpetuating or promoting the view that the individual is less capable or worthy of recognition or value as a human being or as a member of Canadian society, equally deserving of concern, respect, and consideration?

The third of these inquiries is perhaps the most challenging; it is to be assessed from the perspective of the claimant, having regard to several “contextual factors”.  The factors suggested in Law, while not exhaustive, are (1) pre-existing disadvantage, stereotyping, prejudice or vulnerability, (2) correspondence between the ground claimed and the actual needs, capacity or circumstances of the claimant or others, (3) any ameliorative purpose or effects of the impugned law upon a more disadvantaged person or group, and (4) the nature and scope of the interest affected by the impugned law.  Essential to any s. 15(1)  claim is a conflict between the effect of the impugned legislation and the purpose of s. 15(1) ; the latter is defined as “to prevent the violation of essential human dignity and freedom through the imposition of disadvantage, stereotyping, or political or social prejudice, and to promote a society in which all persons enjoy equal recognition at law as human beings or as members of Canadian society, equally capable and equally deserving of concern, respect and consideration” (Law, at para. 51).

39                                  At first blush, the first two broad inquiries raise little controversy in this appeal:  the impugned law draws a clear distinction between citizens and non-citizens, and the latter constitutes an analogous ground of discrimination under s. 15(1) :  see Andrews, supra, at p. 183.  Nevertheless, the respondents argue that the whole point of federal citizenship legislation is to treat citizens and non-citizens differently, and therefore that the two groups cannot validly be compared for s. 15(1)  purposes.  As they put it, “[b]y universal definition and by constitutional fiat, . . . citizens and non-citizens are unequal in status.  To treat them equally would be to negate or abolish the concept of citizenship”.  This argument is animated by the following passage from Law, supra, at paras. 56-57:

Locating the appropriate comparator is necessary in identifying differential treatment and the grounds of the distinction.  Identifying the appropriate comparator will be relevant when considering many of the contextual factors in the discrimination analysis.

To locate the appropriate comparator, we must consider a variety of factors, including the subject-matter of the legislation.  The object of a s. 15(1)  analysis is not to determine equality in the abstract; it is to determine whether the impugned legislation creates differential treatment between the claimant and others on the basis of enumerated or analogous grounds, which results in discrimination.  Both the purpose and the effect of the legislation must be considered in determining the appropriate comparison group or groups.  Other contextual factors may also be relevant.  The biological, historical, and sociological similarities or dissimilarities may be relevant in establishing the relevant comparator in particular, and whether the legislation effects discrimination in a substantive sense more generally:  see Weatherall, supra, at pp. 877-78.  [Emphasis added.]

On the basis of this passage, the respondents concede that citizens and non-citizens may, in certain contexts, appropriately be compared for equality purposes.  In their view, however, such a comparison is not appropriate in the case of “a citizenship defining law that draws a constitutionally permitted distinction between citizens and non-citizens”.  In such a case, the s. 15(1)  analysis would undermine the fundamental difference between citizens and non-citizens and invade Parliament’s exclusive jurisdiction over naturalization and aliens.

40                                  Whether citizens are an appropriate comparator in this case is, in my view, better dealt with as a contextual factor under the third branch of the Law analysis than as a bar to recognizing a legislative distinction.  Although Iacobucci J. stressed the importance of identifying an appropriate comparator group, there is nothing in Law to indicate that the first inquiry is anything but a threshold test.  On the contrary, the precise inquiry at the first stage is whether the law draws a formal distinction “between the claimant and others” (para. 88 (emphasis added)).  Not only is it normally the claimant’s prerogative to choose the appropriate comparator group, but the court is only to step in where “the differential treatment is not between the groups identified by the claimant, but rather between other groups” (para. 58 (emphasis added)).  By contrast, the type of scrutiny proposed by the respondents — namely, to choose comparator groups based on jurisdictional considerations — finds no support either in Law or in any other s. 15(1)  case.  On the contrary, the very essence of an entrenched bill of rights such as the Charter  is to analyse differential treatment as an issue of equality rights, not of federal versus provincial jurisdiction.  Professor Hogg makes this point as follows (P. W. Hogg, Constitutional Law of Canada (loose-leaf ed.), vol. 2, at pp. 52-2 and 52-3):

. . . the position before April 17, 1985, when s. 15  of the Charter of Rights  came into force, was dictated by the doctrine of parliamentary sovereignty:  generally speaking, the Parliament or a Legislature could discriminate as it pleased in enacting otherwise competent legislation. . . .  Before the coming into force of s. 15 , discrimination against aliens and naturalized subjects, and against Indians, was undoubtedly competent to the federal Parliament.

In my view, the respondents’ argument promises a return to the days when federalism, not Charter  principles, governed the constitutionality of citizenship laws:  see R. J. Sharpe, “Citizenship, the Constitution Act, 1867 , and the Charter ”, in W. Kaplan, ed., Belonging:  The Meaning and Future of Canadian Citizenship (1993), 221, at pp. 221-44.  The modern approach is to scrutinize differential treatment according to entrenched rights and freedoms and, in the s. 15(1)  context, the concept of essential human dignity and freedom.  I am confident that such an approach would not “abolish the concept of citizenship” as stated by the respondents.  Although the contextual factors weigh in favour of the appellants in this case, many federal alienage laws could, depending on the context, survive the third branch of the Law analysis.  This is, of course, essentially what happened in Law itself.

41                                  The respondents’ argument is similarly problematic at the second stage, which asks only whether the claimant is “subject to differential treatment based on one or more enumerated and analogous grounds”:  see Law, supra, at para. 88.  As citizenship was recognized as an analogous ground in Andrews, I can find no authority for qualifying this finding according to the context of a given case.  The point of the analogous grounds, according to Law and subsequent cases, is that they are “suspect markers” of discrimination:  the groups occupying them are vulnerable to having their interests overlooked no matter what the legislative context.  Further, as the third inquiry in Law functions to constrain s. 15(1)  claims to cases of genuine discrimination, such analysis should not be pre-empted at the second stage.  This is especially so given this Court’s recent finding that once a ground is found to be analogous, it is permanently enrolled as analogous for other cases:  see Corbiere v. Canada (Minister of Indian and Northern Affairs), [1999] 2 S.C.R. 203, at para. 8.

42                                  At the third stage of Law, the precise issue is whether the impugned law perpetuates the view that the claimants are less capable or less worthy of recognition or value as human beings or as members of Canadian society:  see Law, supra, at para. 99.  Under this rubric, the respondents’ distinction between citizenship preference in the context of federal laws defining the rights and duties of Canadian citizens, as opposed to provincial laws using citizenship as a proxy for merit, becomes relevant.  In the court below, Marceau J.A. articulated this point in two ways: (1) that s. 15(1)  permits of differential treatment to the extent individuals are differently situated (paras. 10-12); and (2) that s. 15(1)  permits distinctions that are relevant to the underlying legislative objective (paras. 22-26).  In his view, either of these principles could provide a basis for guaranteeing equal protection to non-citizens in the context of laws having nothing to do with citizenship per se (as in Andrews), but not in the context of laws whose very raison d’être is the definition of citizenship (as in this case).  In the latter case, it may be argued, first, that citizens and non-citizens are so differently situated that they do not merit equal treatment and, second, that citizenship is a relevant (and indeed necessary) category on which unequal treatment is based.  These arguments find apparent support in Law, supra, wherein Iacobucci J. stated the following, at paras. 70-71:

. . . it will be easier to establish discrimination to the extent that impugned legislation fails to take into account a claimant’s actual situation, and more difficult to establish discrimination to the extent that legislation properly accommodates the claimant’s needs, capacities, and circumstances.

Examples are prevalent in the jurisprudence of this Court of legislation or other state action which either failed to take into account the actual situation of a claimant, or alternatively quite properly treated a claimant differently on the basis of actual personal differences between individuals.  [Emphasis added.]

This dictum is the only direct support I can find in Law for Marceau J.A.’s position.  It appears as the second of four “contextual factors” in the Law analysis, labelled “Relationship Between Grounds and the Claimant’s Characteristics or Circumstances”.  At its broadest, this contextual factor explores the extent to which differential treatment may in fact be acceptable under s. 15(1) :  where there is a genuine “relationship between the ground upon which the claim is based and the nature of the differential treatment” (Law, at para. 69), it may be acceptable to make certain legislative distinctions.  This principle has traditionally functioned to uphold special treatment for groups distinguished by disability (Eaton v. Brant County Board of Education, [1997] 1 S.C.R. 241, and Eldridge v. British Columbia (Attorney General), [1997] 3 S.C.R. 624), as well as gender (Weatherall v. Canada (Attorney General), [1993] 2 S.C.R. 872, and Brooks v. Canada Safeway Ltd., [1989] 1 S.C.R. 1219).  The respondents imply that it should also function to permit differential treatment on the basis of citizenship.  In their words:

. . . it is the essence of the concept of citizenship that it confers certain rights and entitlements on citizens that are necessarily denied to non-citizens.  Both constitutional and federal law prescribe these rights and entitlements.  Consonant with historical and international practice, preferential access to Public Service employment is one of them.  What the appellants inveigh as simply another entitlement-denying law operating against non-citizens is, in reality, an original and fundamental citizenship-defining provision that establishes a basic and universal attribute of the status of citizen.

43                                  Although s. 15(1) permits some differential treatment, the respondents’ citizenship argument goes beyond what is contemplated by the second contextual factor in Law.  In the past, this factor has meant that a disadvantaged class might deserve special accommodation on account of being differently situated; such as, for example, the right to sign-language interpreters for the hearing impaired in public hospitals (Eldridge, supra) or, conversely, that groups who are “more advantaged in a relative sense” may be denied benefits that correspond with the “different circumstances experienced by the more disadvantaged group being targeted by the legislation” (Law, supra, at para. 103).  What matters, in my view, is that the law or government action take into account the particular situation of those affected, including any relative advantage or disadvantage.  This is the requirement of the contextual and substantive approach to s. 15(1) that was unanimously endorsed by this Court in Law, as opposed to a formal approach that, at least in some cases, functions to exacerbate historical disadvantage.

44                                  In this case, to the extent non-citizens are differently situated than citizens, it is only because the legislature has accorded them a unique legal status.  In all relevant respects — sociological, economic, moral, intellectual — non-citizens are equally vital members of Canadian society and deserve tantamount concern and respect.  The only recognized exception to this rule is where the Constitution itself withholds a benefit from non-citizens, as was the case in Chiarelli, supra.  In such a case, it may be said that the Charter  itself authorizes differential treatment, and that finding a s. 15(1)  violation would amount to finding the Charter  in violation of itself.  Such is not the case in the present appeal.  On the contrary, the distinction in this case finds no authorization in the Charter  and, more broadly, is not made on the basis of any “actual personal differences between individuals”:  see Law, supra, at para. 71.  If anything, the distinction places an additional burden on an already disadvantaged group.  Such a distinction is impossible to square with this Court’s finding in Andrews, supra, at p. 183, which held that “[a] rule which bars an entire class of persons from certain forms of employment, solely on the grounds of a lack of citizenship status and without consideration of educational and professional qualifications or the other attributes or merits of individuals in the group, would . . . infringe s. 15  equality rights”.

45                                  Turning to the remaining contextual factors in Law, the questions to be asked are whether (1) the claimants in this case suffer from pre-existing disadvantage, stereotyping, prejudice or vulnerability; (2) the law aims or operates to ameliorate the predicament of a more disadvantaged person or group; and (3) the nature and scope of the interest affected by the impugned law is such that it merits constitutional protection.  In my view, all three of these factors militate in favour of a s. 15(1)  violation.  First, while the claimants in this case are all relatively well-educated, it is settled law that non-citizens suffer from political marginalization, stereotyping and historical disadvantage.  Indeed, the claimant in Andrews, who was himself a trained member of the legal profession, was held to be part of a class “lacking in political power and as such vulnerable to having their interests overlooked and their rights to equal concern and respect violated”:  see Andrews, supra, per Wilson J., at p. 152.  In my view, this dictum applies no matter what the nature of the impugned law.  Second, s. 16(4)(c) of the PSEA  does not aim to ameliorate the predicament of a group more disadvantaged than non-citizens; rather, the comparator class in this case (unlike in Law, perhaps) enjoys greater status on the whole than the claimant class.  Finally, the nature of the interest in this case — namely, employment — is most definitely one that enjoys constitutional protection.  As repeatedly held by this Court, work is a fundamental aspect of a person’s life, implicating his livelihood, self-worth and human dignity:  see Reference Re Public Service Employee Relations Act (Alta.), [1987] 1 S.C.R. 313, per Dickson C.J., at p. 368, and subsequent cases.  Although the scope of the affected interest in this case is fairly narrow owing to the fact that s. 16(4) (c) is limited to public sector employment and does not impose a complete bar on non-citizens, in my view the nature and scope of the affected interest still warrants constitutional protection.  As stated above, work is a fundamental aspect of a person’s life, and a law which operates to limit the range of employment options for non-citizens is still likely to implicate the individual’s livelihood, self-worth and human dignity.  Indeed, much of the discussion in this case was centered on the appellants’ argument that Parliament’s intention was to distinguish between citizens and non-citizens on the basis of their relative loyalty and commitment to Canada.  In this context, a cursory look at the four Law factors suggests that s. 16(4) (c) of the PSEA  violates s. 15(1)  of the Charter .

46                                  Needless to say, the Law factors should not be applied too mechanically.  One must never lose sight of the overarching question, which is whether the law perpetuates the view that non-citizens are less capable or less worthy of recognition or value as human beings or as members of Canadian society:  see Law, supra, at para. 99.  It may be, in light of the above discussion, that a law defining the core rights and privileges of citizens is incapable of perpetuating such a view; indeed, such a law finds support in numerous international treaties and is accepted by almost every country in the world.  In my view, however, this misses the point of the Law methodology; what is required is a contextualized look at how a non-citizen legitimately feels when confronted by a particular enactment.  Even if the non-citizen knows the preference has nothing to do with her capabilities — as most reasonable people would — she may still feel “less . . . worthy of recognition . . . as a member of Canadian society”:  see Law, supra, at para. 88.  This subjective view must be examined in context, that is, with a view to determining whether a rational foundation exists for the subjective belief.

47                                  In measuring the appellants’ subjective experience of discrimination against an objective standard, it is crucial not to elide the distinction between the claimant’s onus to establish a prima facie s. 15(1)  violation and the state’s onus to justify such a violation under s. 1 .  Section 15(1)  requires the claimant to show that her human dignity and/or freedom is adversely affected.  The concepts of dignity and freedom are not amorphous and, in my view, do not invite the kind of balancing of individual against state interest that is required under s. 1  of the Charter .  On the contrary, the subjective inquiry into human dignity requires the claimant to provide a rational foundation for her experience of discrimination in the sense that a reasonable person similarly situated would share that experience.  In this case, the claimants submit that a reasonable person similarly situated would believe that the reduced opportunity of working in the federal Public Service fails to account for their individual capacities and, moreover, implies they are less loyal and worthy of trust.  The existence of a s. 15(1)  violation depends on the validity of this submission.

48                                  By contrast, the government’s burden under s. 1  is to justify a breach of human dignity, not to explain it or deny its existence.  This justification may be established by the practical, moral, economic, or social underpinnings of the legislation in question, or by the need to protect other rights and values embodied in the Charter .  It may further be established based on the requirements of proportionality, that is, whether the interest pursued by the legislation outweighs its impact on human dignity and freedom.  However, the exigencies of public policy do not undermine the prima facie legitimacy of an equality claim.  A law is not “non-discriminatory” simply because it pursues a pressing objective or impairs equality rights as little as possible.  Much less is it “non-discriminatory” because it reflects an international consensus as to the appropriate limits on equality rights.  While these are highly relevant considerations at the s. 1  stage, the suggestion that governments should be encouraged if not required to counter the claimant’s s. 15(1)  argument with public policy arguments is highly misplaced.  Section 15(1)  requires us to define the scope of the individual right to equality, not to balance that right against societal values and interests or other Charter  rights.

49                                  It is not, as my colleague Arbour J. suggests at para. 86, an “eagerness to extend equality rights as widely as possible” that informs the distinction between s. 15(1)  and s. 1 .  It is the very structure of the Charter  that mandates this distinction, as well as the methodology adopted by this Court since Andrews.  Nor do I accept my colleague’s suggestion that s. 15(1)  rights are all but absolute, in that their violation should only be justifiable in rare circumstances.  This Court has often applied s. 1  to breaches of s. 15(1) , in so doing recognizing that s. 15(1)  merits a liberal and purposive construction: see McKinney v. University of Guelph, [1990] 3 S.C.R. 229; Harrison v. University of British Columbia, [1990] 3 S.C.R. 451; Stoffman v. Vancouver General Hospital, [1990] 3 S.C.R. 483; Weatherall, supra; Egan v. Canada, [1995] 2 S.C.R. 513.  In each of these cases, it was incumbent on the Court to consider the extent and impact of the s. 15(1)  breach and, in appropriate contexts, to find that a particular breach of s. 15(1)  was minor.  This is not indicative of undue deference to the legislature, but of the need for a flexible approach to s. 1  justification and, more broadly, the recognition that any balancing between individual rights and societal needs occurs in s. 1 , not s. 15(1) .  Indeed, conducting this balancing at the s. 15(1)  stage would accord far greater deference to the legislature than I suspect my colleague Arbour J. intends.

50                                  The balancing conducted at the s. 15(1)  stage would transform that subsection into a variant of s. 7, whereby violations are difficult to establish and, in turn, difficult to justify under s. 1 .  Only the “most important” objectives would be sufficiently pressing to violate s. 15(1) , and the proportionality test would in turn be conducted with “uncompromising rigour”: see Arbour J., at para. 91.  My central concern with this approach is not only that, in my view, it departs from previous s. 15(1)  jurisprudence, but that it substitutes a rigid and categorical approach to s. 1  justification for a contextual one: see Irwin Toy Ltd. v. Quebec (Attorney General), [1989] 1 S.C.R. 927, and Thomson Newspapers Co. v. Canada (Attorney General), [1998] 1 S.C.R. 877.  Even if such an approach could be defended, however, it is not supported by the different wording of s. 15(1)  and s. 7 .  Section 7 contains an explicit internal limitation on the right to life, liberty and security of the person, in that such rights may be infringed in accordance with the “principles of fundamental justice”.  Section 15(1) , by contrast, contains no such limitation.  What my colleague Arbour J., at para. 92, refers to as a “differentiation between legislative distinctions and discrimination” is not an internal limitation on s. 15(1)  in the s. 7 sense, but a judicial interpretation of the normative parameters of the term “discrimination”.  Such parameters are no different than those used to define the term “expression” in s. 2 (b) or the term “unreasonable” in s. 8.  They function to define the right or freedom at issue, not to place an internal limitation on it.

51                                  With respect, I must disagree with my colleague Arbour J. that the difficulties posed by blurring the distinction between s. 15(1)  and s. 1  are not insurmountable.  At the very least, such an approach creates significant uncertainty for lower courts in terms of the kinds of considerations they are permitted to adduce in adjudicating a s. 15(1)  claim.  Yet on a deeper level, the approach has the potential to create a hierarchy of rights within s. 15(1)  itself, whereby public policy considerations may defeat a s. 15(1)  claim in certain cases (e.g. citizenship), but not others (e.g. race).  I cannot find support for such a hierarchy in the Charter .  Not only does my colleague stop short of providing any criteria for ranking the analogous grounds, but she does not persuade me that the kinds of considerations she uses in this case — for example, the fact that two of the claimants chose to forego Canadian citizenship — could not, even unwittingly, be applied invidiously in future cases.  While such an approach would certainly filter out vexatious s. 15(1)  claims, I think it would do so at great cost to our Court’s liberal interpretation of equality rights — an interpretation which, it must be said, goes beyond s. 15(1)  of the Charter  and affects Canada’s human rights jurisprudence generally.

52                                  Turning to the subjective-objective evaluation in this case, I think the claimants in this case felt legitimately burdened by the idea that, having made their home in Canada (and, in To-Thanh-Hien’s case, begun to seek citizenship), their professional development was stifled on the basis of their citizenship status.  Their subjective reaction to the citizenship preference no doubt differed from their reaction to not being able to vote, sit in the Senate, serve on a jury, or remain in Canada unconditionally.  An obvious difference in this context is that employment is vital to one’s livelihood and self-worth; another is that there is no apparent link between one’s citizenship and one’s ability to perform a particular job; finally, the distinction can reasonably be associated with stereotypical assumptions about loyalty and commitment to the country, even if that is not Parliament’s intention.  There is certainly no shortage of evidence in this case to support these views.  As the respondents’ own expert, Peter H. Schuck, recognized in a 1997 article (“The Re-Evaluation of American Citizenship” (1997), 12 Geo. Immigr. L.J. 1, at p. 14):

The . . . policy of barring aliens from federal employment, which is similar to the practice of most nations, is likely to be a greater concern to aliens than the bar to jury service for most aliens.  Few if any legal permanent residents (“LPRs”) are likely to seek high elective or appointive offices prior to naturalization.  Many LPRs, however, might want to pursue employment in the federal, state, and local civil service systems.

This point is borne out on the record.  All three appellants in this case sought Public Service employment immediately upon arriving in Canada and, for the first three years, could not have changed their citizenship even had they wanted to.  During that time, one of them was told the PSC would sooner lower the qualifications for a job than hire a non-citizen; another was barred from a position she had been competently performing on contract; and a third was rejected despite her obvious desire to become a Canadian citizen.  The impact of placing obstacles in the way of the appellants’ professional development does not vary according to whether the PSEA  imposes a “preference” or a “ban”.  Immigrants come to Canada expecting to enjoy the same basic opportunities as citizens and to participate fully and freely in Canadian society.  Freedom of choice in work and employment are fundamental aspects of this society and, perhaps unlike voting and other political activities, should be, in the eyes of immigrants, as equally accessible to them as to Canadian citizens.  Discrimination in these areas has the potential to marginalize immigrants from the fabric of Canadian life and exacerbate their existing disadvantage in the Canadian labour market.  This is true whether or not the discrimination operates on the basis of stereotyping; if it makes immigrants feel less deserving of concern, respect and consideration, it runs afoul of s. 15(1) :  see Law, supra, at para. 88.  For these reasons, I conclude that s. 16(4) (c) of the PSEA  violates s. 15(1)  of the Charter  and requires justification under s. 1 .

B.      Section 1 

53                                  At the s. 1  stage, it is for the government to demonstrate that, on a balance of probabilities, s. 16(4) (c) is a “reasonable limit” on equality that can be “demonstrably justified in a free and democratic society”:  see R. v. Oakes, [1986] 1 S.C.R. 103, at pp. 136-37.  To qualify as such, the provision must (1) pursue an objective that is sufficiently important to justify limiting a Charter  right, (2) be rationally connected to that objective, (3) impair the right no more than is reasonably necessary to accomplish the objective, and (4) not have a disproportionately severe effect on the persons to whom it applies:  see Oakes, supra, at pp. 138-39.  These criteria will be applied with varying levels of rigour depending on the context of the appeal:  see Thomson Newspapers, supra.  In this case, we are presented with a law that attempts to promote the value of Canadian citizenship by detracting from the rights of non-citizens; as this inevitably requires Parliament to balance the interests of competing groups, some degree of deference is required in the application of Oakes, supra.  That being said, the law does not promote the interests of a vulnerable group, is not premised on particularly complex social science evidence, and interferes with an activity (namely employment) whose social value is relatively high:  see Thomson Newspapers, supra, and Irwin Toy, supra, at pp. 993-94.

(1) Sufficiently Important Objective

(a) What Is the Legislative Objective?

54                                  At trial, Wetston J. was presented with radically different views of the objective behind the citizenship preference.  The appellants claimed the objective was to ensure a loyal and committed Public Service; on this view, s. 16(4) (c) stemmed from a dubious legacy of according citizens greater privileges on account of their supposed merit.  The respondents insisted the preference had nothing to do with merit; they claimed it was meant to further Canada’s citizenship policy by granting citizens certain privileges not enjoyed by immigrants — the right to vote, for example.  In turn, the respondents identified a twofold objective behind Canada’s citizenship policy:  first, to enhance the meaning of citizenship as a unifying symbol for Canadians; and second, to encourage permanent residents to naturalize.  In my view, the respondents’ view must prevail.  Even if concerns about commitment and loyalty informed the enactment of the amendments to the Civil Service Act in 1908, on which I make no comment, there is no denying that the citizenship preference is also intended to further Canada’s citizenship policy.  This was affirmed by all four judges in the courts below, including Linden J.A., and it is reflected in the legislative record:  see Department of Justice, Equality Issues in Federal Law:  A Discussion Paper (1985), at pp. 49-50.  Whether this privilege is pressing and substantial is, as we shall see, a matter of some controversy.

55                                  In oral argument, the appellants urged this Court to consider the testimony of one of the respondents’ witnesses, Mr. John J. Carson, as conclusive evidence of legislative objective.  Mr. Carson served as chairman of the PSC in the 1960s; he testified that he objected to a repeal of the citizenship preference at the time and, moreover, that concerns about “commitment and loyalty” motivated his views.  In particular, he stated that “if you’re undertaking an application for employment you want to give evidence of good faith and your willingness to partner in the venture that you’re going into” and, further, that “someone who has shown evidence of commitment and a desire to be fully involved is usually an indication of their motivation”.  The appellants argue that Wetston J. erred in not considering this testimony, especially having ruled that Mr. Carson’s opinion was admissible.  In my view, this would unduly interfere with the trial judge’s discretion.  The trial judge was not bound to accept Mr. Carson’s answer on the legislative objective, particularly where he had restricted Mr. Carson’s testimony to whether he felt it was “necessary in that responsibility to make any recommendations to repeal that provision”.  Much less was the trial judge bound to ignore the other legislative objectives which the Crown proffered, which were subsequently upheld by all three judges at the Court of Appeal.  In short, I think the appellants are hard-pressed to modify the legislative objective at this stage based on their interpretation of the events at trial.  Not only were the trial judge’s actions perfectly legitimate, but the objectives he identified were never questioned by the reviewing court.

(b) Are the Objectives Sufficiently Important?

56                                  According to Oakes, supra, Parliament’s objectives must be sufficiently important to justify overriding a Charter  right.  The respondents note that virtually all liberal democracies impose citizenship-based restrictions on access to the national Public Service; these restrictions vary from virtual bans on federal Public Service employment (as in Switzerland and the United States) to policies allowing permanent residents to work in the Public Service on a probationary basis (as in Australia).  The respondents further argue that international conventions support citizenship-based restrictions by guaranteeing the right of all citizens to work in the Public Service:  see Article 21(2) of the Universal Declaration of Human Rights, G.A. Res. 217 A (III), U.N. Doc. A/810, at 71 (1948) (“Everyone has the right of equal access to public service in his country”); Article 25(c) of the International Covenant on Civil and Political Rights, 999 U.N.T.S. 171 (1966) (“Every citizen shall have the right and the opportunity . . . [t]o have access, on general terms of equality, to public service in his country”).  For their part, the appellants claim that widespread acceptance of the citizenship-based restrictions do not justify them, and that Canada should be held to a higher standard than countries which discourage immigration.  The appellants also note that citizenship-based restrictions imposed by state legislatures have routinely been struck down by the United States Supreme Court as violations of equality rights:  see Mathews v. Diaz, 426 U.S. 67 (1976), and Sugarman v. Dougall, 413 U.S. 634 (1973).

57                                  In my view, a cursory examination of Canada’s citizenship policy provides a normative foundation for the impugned law.  This policy dates to the enactment of The Canadian Citizenship Act in 1946 (S.C. 1946, c.15); it sought to clarify confusion over the use of the terms “citizen” and “national” in federal legislation and create a unifying symbol for Canadians:  see House of Commons Debates, vol. II, 1st Sess., 20th Parl., October 22, 1945 at pp. 1335 et seq. (the Hon. Paul Martin).  Since then, Canada’s citizenship policy has embodied two distinct objectives:  to enhance the meaning of citizenship as a unifying bond for Canadians, and to encourage and facilitate naturalization by permanent residents.  In my view, these objectives are non-controversial.  In any liberal democracy, the concept of citizenship serves important political, emotional and motivational purposes; if nothing else, it fosters a sense of unity and shared civic purpose among a diverse population:  see W. Kymlicka, Multicultural Citizenship: A Liberal Theory of Minority Rights (1995), at pp. 173-76.  This was recognized by this Court in Winner, supra, in which Rand J. defined citizenship, at p. 918, simply as “membership in a state”.  Rand J. went on to affirm the very basis of Canada’s citizenship policy:  “in the citizen”, he held, “inhere those rights and duties, the correlatives of allegiance and protection, which are basic to that status”.  The signal effect of the impugned provisions is not to discourage immigration but to underscore the value of citizenship as a unifying bond for Canada.

58                                  The appellants question the very premise of Canada’s citizenship policy, arguing that one does not enhance the meaning of citizenship by detracting from the rights of non-citizens.  In their view, this is a “perverse” approach to social unity and undermines the spirit of inclusion represented by the Charter  and our liberal immigration laws.  In my view, this argument is unrealistic.  It only makes sense for a country as open and diverse as Canada to enact a policy that integrates its population; in an era of increased movement across borders, citizenship still provides immigrants with a basic sense of identity and belonging.  The question that challenges multicultural polities like Canada is not whether to enact a citizenship policy, but how to do so in a way that is respectful of cultural and linguistic differences.  At trial, the appellants’ chief expert, Joseph Carens, essentially admitted this view; he testified that “there may be certain points on which we may draw legal distinctions between citizens and non-citizens but other points in which everyone is regarded as a member of the community”.  Canada has sought to strike this balance not only by limiting the number of privileges accorded to Canadian citizens, but by allowing dual citizenship, relaxing naturalization requirements and, in the appellant To-Thanh-Hien’s case, making special efforts to find employment for qualified visible minorities:  see Canadian Citizenship: A Sense of Belonging (1994), Report of the Standing Committee on Citizenship and Immigration, at pp. 5-7, 11 and 15.  By taking measures such as these, Parliament attempts to reconcile the goals of enhancing Canadian citizenship and respecting cultural diversity.  I am thus comfortable concluding that the objectives behind s. 16(4) (c) are sufficiently important to justify limiting the appellants’ equality rights.

(2) Rational Connection

59                                  With respect to rational connection, the appellants suggest it is irrational to pursue Canada’s citizenship policy by making Public Service employment a privilege of citizenship.  In their view, there is no end to the amount of discrimination Parliament could inflict on non-citizens if such an objective is accepted.  Moreover, they argue that s. 16(4) (c) actually undermines Parliament’s objective by making Canada a less desirable country in which to live.  In my view, this opinion is unrealistic; furthermore, this is something for Parliament to decide.  While there is a point at which granting privileges to citizens may be unjustifiable under s. 1  — banning immigrants from social housing, perhaps — that point is not the same as the point at which this Court finds a s. 15(1)  violation.  Rather, as contemplated by s. 1  of the Charter , Parliament is entitled to some deference as to whether one privilege or another advances a compelling state interest.  In this case, Parliament’s view is supported by common sense and widespread international practice, both of which are relevant indicators of a rational connection.  Short of rejecting Canada’s entire citizenship policy, it seems rather speculative to suggest that this privilege is so arbitrary and unreasonable that it detracts from the value of Canadian citizenship.  If this logic were accepted, even the less intrusive alternatives proposed by Linden J.A. would have to be rejected as failing the rational connection test.

60                                  With respect to the second objective, encouraging naturalization, the appellants question whether granting employment privileges to non-citizens actually persuades permanent residents to naturalize.  In their view, it would be no less surprising to find otherwise given the limited reach of the preference and the fact that many other factors inform the decision to naturalize.  From a statistical perspective, however, Canada’s citizenship policy seems generally to have worked.  There is a very close relationship between immigration and naturalization rates in Canada, meaning that a high proportion of immigrants choose to naturalize upon meeting the three-year residency requirement.  While this may be due to several factors — an immigrant’s personal circumstances, the fact that citizenship is so easy to acquire in Canada, or the mere fact that Canada is a desirable country in which to live — the government’s efforts to enhance the value of citizenship can reasonably be assumed to play a role.  This is apparent from the personal rewards that accrue from being able to vote, remain in Canada unconditionally, serve appointed political office or join the Public Service.  This common sense view is shared by almost every country in the world, including those that make citizenship more difficult to obtain.  In this context, it would not be appropriate to hold Parliament to an exacting standard of proof:  see Hogg, supra, at p. 35-8, citing Oakes, supra, at p. 138.  The real issue, in my view, is whether the law is tailored in such a way that it does not unduly burden non-citizens in its laudable efforts to promote Canadian citizenship.

(3) Minimum Impairment

61                                  This brings me to the minimum impairment test, which asks whether there are less intrusive ways of enhancing the value of citizenship among public servants.  Before examining the alternatives in any detail, it is important to note the features of s. 16(4) (c) which render it less intrusive than it might be.  Among these features are (1) the fact that it is a preference only and not an absolute bar, (2) the fact that it does not apply to closed competition, which is the most common means of staffing Public Service positions, (3) the fact that it only applies to the referral stage of open competition, not the inventory or eligibility stage, and (4) the fact that dual citizenship is permitted in Canada, such that Canadian law does not burden non-citizens with a choice between renouncing their foreign citizenship and entering the Public Service.  These factors were all recognized by the Court of Appeal, which also noted that the preference is ultimately discretionary.  In my view, the factors all go to whether s. 16(4) (c) falls within the “range of reasonable alternatives” permitted by s. 1  of the Charter ; see RJR-MacDonald Inc. v. Canada (Attorney General), [1995] 3 S.C.R. 199, at para. 160.

62                                  The hallmark of s. 16(4) (c) is that it is merely a preference for Canadian citizens, as opposed to an absolute bar on non-citizens.  Non-citizens are frequently referred to open competition, either along with qualified Canadian citizens or after the pool of qualified citizens is exhausted.  While the former occurs rarely, the PSC director who testified for the respondents, Peter Stewart, recalled several such referrals in the year leading up to trial.  Mr. Stewart recalled even more referrals in the latter category:  co-op assignments for librarians where no qualified Canadians applied, computer systems positions where no qualified Canadians were willing to take short-term positions, term research positions with Agriculture Canada where the requirements were highly stringent, and a position at the Department of Justice requiring, among other things, experience and knowledge of war crimes trials and German fluency.  Indeed, the appellants Bailey and To-Thanh-Hien were themselves referred to open competition; the former was referred the year she arrived to Canada, and the latter was referred both before and after obtaining Canadian citizenship.

63                                  A second feature of s. 16(4) (c) is that it does not apply to closed competition.  This restriction cannot be underestimated; it means the vast majority of Public Service positions are equally extended to citizens and non-citizens.  While this provides little comfort to those who lack the experience generally required of closed competitions — and who therefore must seek promotion through open competition — such a lack of experience is not unique to non-citizens.  Many citizens who are members of the Public Service lack extensive experience and thus, like the appellant Bailey, seek promotion through open competition rather than closed.  The desire to do this stems from a strategic decision to seek a job that does not require experience rather than to seek one that does.  Thus, while I accept that Bailey’s opportunities for promotion are less than her colleagues’, I think this is only partly because of the citizenship preference and also because of her lack of experience.  More compelling, in my view, is the fact that she is as eligible as any Canadian citizen to compete in the vast majority of Public Service competitions.

64                                  Third, s. 16(4)(c) only applies to the referral stage of open competition.  This means that non-citizens are equally entitled to submit applications to the PSC inventory and, for the purposes of this appeal, equally eligible for Public Service employment once referred to the requesting department.  This substantially increases their chances of Public Service employment:  if the preference applied to non-citizens at the inventory stage, it would amount to an all-out ban given the demand for Public Service jobs among citizens.

65                                  Finally, Parliament has substantially reduced the burden on non-citizens by permitting them to hold dual citizenship upon naturalizing in Canada.  This spares many immigrants the choice between becoming a Canadian citizen (and assuming all of the privileges and responsibilities thereof) and maintaining citizenship in their country of origin.  While it does not assist individuals like Bailey and Lavoie, whose countries of origin do not permit dual citizenship, this can hardly be considered the responsibility of the Canadian government.  On the contrary, the burden faced by such individuals is a combination of their own countries’ legislation and their personal decision to maintain citizenship abroad.  Parliament cannot be expected to abandon its citizenship preference in order to lessen this burden, much less to establish a regime that furthers its objective in an entirely different way.

66                                  Despite these features of s. 16(4)(c), parliamentary committees have twice recommended the repeal of the citizenship preference, and other jurisdictions have enacted arguably less impairing restrictions.  Linden J.A. summarized these alternatives, and the jurisdictions that have adopted them, as follows (at para. 206):

First, the preference could be legislated to apply only after a functional analysis of the open position revealed it to be one which was appropriate for non-citizens.  This is not unlike the citizenship preference employed in the 50 states of the United States.  Similarly, evidence was adduced in this case to explain that New Zealand imposes a citizenship requirement for positions which can be classified as “security positions.”  Second, the preference could be legislated to apply only after people were eligible for citizenship and chose not to apply for it.  This is similar to the citizenship preference as it exists in Australia. . . .  Third, the citizenship preference could be eliminated in the case of permanent residents, but maintained for non-landed visa holders. . . .  Fourth, the citizenship preference could apply as a true affirmative action program —  if all other considerations were equal, citizens would be preferred over non-citizens.  Finally, the preference could be struck entirely, following which the Commission could rely on subsection 12(3)  of the PSEA , which would permit a position to be limited on the basis of a bona fide occupational requirement, e.g., a residence requirement to ensure familiarity with the country, and perhaps commitment and loyalty with regards to those positions which require it.  [Emphasis added.]

67                                  In assessing these alternatives, it is crucial not to lose sight of the objective underlying the legislation; as I stressed in Thomson Newspapers, supra, the point is not just to look for anything less intrusive, but something that would fulfill the objective less intrusively.  In this regard, I am sceptical whether a “political function” test would accomplish the objectives articulated by the respondents in this case.  While such a rule would no doubt impair s. 15(1)  less than the current rule, in my view it would decrease the incentive to naturalize and erode the value of Canadian citizenship.  The notion that certain employment functions enhance Canadian citizenship more than others is not only counterintuitive, but it undermines the ideal of an open and egalitarian Public Service.  This was affirmed at trial by Mr. Carson, who noted that such distinctions would fragment the Public Service and subject employees to different rules and regulations.  Even assuming this were administratively possible, it would certainly require arbitrary distinctions between different classes of employment.  At what point does a position become so “political” that it enhances the value of Canadian citizenship?  Perhaps such a point exists; however, I am more inclined to the view that all jobs are worthy of equal respect, and that drawing distinctions based on political function would, in light of Parliament’s stated objectives, implicitly denigrate certain types of work.

68                                  Of the remaining alternatives, I am most compelled by the Australian model of referring permanent residents to open competition pending the outcome of their citizenship applications.  To the extent permanent residents are committed to Canadian citizenship but nevertheless burdened by citizenship preferences, s. 16(4) (c) of the PSEA  might be considered overbroad.  Indeed, it may be argued that the Australian model increases the incentive to naturalize so that permanent residents can remain in the Public Service after they become eligible for citizenship.  That being said, the Australian model presents some obvious administrative difficulties.  As noted in the American case of Mow Sun Wong v. Hampton, 435 F.Supp. 37 (1977), at pp. 45-46, such a scheme “would be excessively disruptive to the service, in that significant numbers of alien employees would automatically be terminated upon their failure, for one reason or another, to become naturalized”.  These difficulties would be especially acute in Canada, where an unsuccessful applicant would technically be entitled to keep her job if there were not sufficiently qualified Canadians; thus, the Commission would presumably have to conduct an open competition every time a probationary employee failed to naturalize.  It was for such reasons that the trial judge rejected this alternative, reiterating the importance of deferring to Parliament’s discretion.  Moreover, it is not even clear that the Australian model is less impairing than the Canadian one:  not only does Australia prohibit dual citizenship, but it creates an all-out restriction on non-citizens who are not seeking or who fail to obtain naturalization.  This seems entirely unwelcoming to those permanent residents who do not apply for citizenship, as opposed to the across-the-board preference in effect in Canada.  Indeed, the Australian model would have been of no assistance to the appellants Bailey and Lavoie.

69                                  In the final analysis, there is little doubt that certain individuals fall through the cracks of s. 16(4) (c) of the PSEA :  those who are committed to naturalization and awaiting final determination of their citizenship application; those who are committed to naturalization but have legitimate reasons for maintaining permanent resident status; and, perhaps, those who are otherwise qualified for the Public Service and whose full-time employment would in no way undercut Parliament’s objectives.  What is less certain, however, is whether a reasonable alternative is available that would fill these cracks in a fair, consistent and principled manner.  In this regard, I am struck by the fact that Parliament has repeatedly considered less intrusive ways of furthering its citizenship policy and in some cases has lessened the burden on non-citizens.  The most obvious example is in 1961 when Parliament changed the restriction on non-citizens to a preference and thereby departed from the path taken by numerous other countries.  This amendment was followed by numerous reviews of the citizenship preference between 1961 and 1985:  a 1967 overhaul of the legislative scheme which did not question the value of the preference; a 1974 parliamentary committee which re-examined the preferences in the PSEA  and recommended the retention of the citizenship preference; a 1979 report by the D’Avignon Committee recommending the extension of the preference to permanent residents; and a 1985 parliamentary committee which recommended, to no avail, that the citizenship preference be eliminated:  see Equality for All: Report of the Parliamentary Committee on Equality Rights (1985); Toward Equality:  The Response to the Report of the Parliamentary Committee on Equality Rights (1986).  In my view, the fact that Parliament did not adopt the position of the D’Avignon and Equality Rights Committees is not a reason to fail the minimum impairment test; on the contrary, it is evidence that Parliament has conscientiously considered alternatives to s. 16(4) (c) and chosen not to pursue them.  The role of this Court is not to order that Parliament should have decided otherwise.  This is precisely the type of policy review that is beyond our reach, particularly given the delicate balancing that is required in this area of the law.

(4) Final Balancing

70                                  Having passed the minimum impairment stage, the final stage asks whether the infringing effects of s. 16(4) (c) outweigh the importance of the objective sought.  This final stage should not, as I pointed out in Thomson Newspapers, supra, be conflated with the first three stages.  If the first three relate to reasonableness of the legislation itself, the fourth examines the nature of the infringement and asks whether its costs outweigh its benefits.  The implication of finding a violation at the fourth stage is that even a minimum level of impairment is too much:  the costs to the claimant so outweigh the benefits that no solace can be found in the fact that the legislation violates the Charter  “as little as reasonably possible”.  Moreover, if the costs of the legislation are significant enough, and the legislation only partially achieves its objectives, greater evidence of its benefits may be necessary in order to survive s. 1 :  see Dagenais v. Canadian Broadcasting Corp., [1994] 3 S.C.R. 835, at p. 889.  In Thomson Newspapers, I summarized the proper approach as follows, at para. 125:

The third stage of the proportionality analysis provides an opportunity to assess, in light of the practical and contextual details which are elucidated in the first and second stages, whether the benefits which accrue from the limitation are proportional to its deleterious effects as measured by the values underlying the Charter .

On this point, the appellants argue that a preference in open competitions effectively prevents non-citizens from breaking into the Public Service.  The reason for this is most positions within the Public Service are filled internally (up to 75-80 percent in a given year), such that giving citizens preferential treatment in open competition effectively denies non-citizens their only opportunity to enter the Public Service.  In addition, the citizenship preference is said to preclude non-citizens from valuable promotions once they become employees of the Public Service.  The parties acknowledge, of course, that any burden imposed by s. 16(4) (c) is temporally limited for those non-citizens who successfully undergo the naturalization process.

71                                  In my view, the appellants’ argument assumes that lifting the citizenship preference in open competition would substantially increase the employment prospects of non-citizens.  However, the fact that most positions are filled internally shows that it is almost as difficult for citizens to enter the Public Service as non-citizens; thus, the latter’s disadvantage relative to the former does not appear significant.  This is not to deny that some non-citizens would have jobs in the Public Service but for the citizenship preference; however, given the scarcity of Public Service openings and the resultant competitiveness of the positions, I do not think these exceptional cases ground a constitutional violation.  With respect to promotions within the Public Service, I acknowledge that non-citizens are at a disadvantage relative to their colleagues when it comes to promotion by open competition.  This is essentially the complaint of the appellant Bailey, who was already a member of the Public Service when she applied for various open competitions but was excluded because of her citizenship.  However, the record shows that promotion via open competition is a distinct possibility for non-citizens, and indeed that Bailey herself was successful in some cases.  Moreover, non-citizens who are members of the Public Service have unfettered access to closed competitions, which are by far the more conventional avenue of Public Service promotion.  For these reasons, I have difficulty characterizing the effect of s. 16(4) (c) as a disproportionate and unjustified breach of the Charter .  Absent greater evidence of the prevalence of this problem, or of the impact on the claimants’ career prospects, I do not think the inconvenience they suffered is too high a price to pay for the government’s right to define the rights and privileges of its citizens.

VI.    Conclusion

72                                  Based on the foregoing, I conclude that s. 16(4) (c) is a breach of s. 15(1)  of the Charter  that can be demonstrably justified in a free and democratic society.  I acknowledge that the legislation creates differential treatment which, in some cases, functions to impair the dignity and freedom of non-citizens.  However, I note that the Charter  permits certain forms of discrimination where they pursue an important objective in a proportionate manner.  I would therefore dismiss this appeal with costs in this Court, substantially for the same reasons as Wetston J. and Desjardins J.A. I would answer the constitutional questions as follows:

1.                  Does paragraph 16(4)(c) of the Public Service Employment Act, R.S.C. 1985, c. P‑33 , on its own or in its effect, discriminate against persons on the basis of citizenship by providing a preference to Canadian citizens over non‑citizens in open competitions in the federal public service, contrary to section 15(1)  of the Canadian Charter of Rights and Freedoms ?

Yes.

2.                  If the answer to question one is yes, is the discrimination a reasonable limit prescribed by law which can be demonstrably justified in a free and democratic society under section 1  of the Canadian Charter of Rights and Freedoms ?

Yes.

The following are the reasons delivered by

73                                  Arbour J. — I have read Justice Bastarache’s thorough reasons and, although I would also dismiss the appeal, I would do so for different reasons.  In my view, s. 16(4) (c) of the Public Service Employment Act, R.S.C. 1985, c. P-33  (“PSEA ”), does not infringe s. 15(1)  of the Canadian Charter of Rights and Freedoms .  On this record, I cannot conclude that the third branch of the test in Law v. Canada (Minister of Employment and Immigration), [1999] 1 S.C.R. 497, is met and that the law discriminates.

74                                  Had I found a breach of s. 15(1)  I would have been unable to save it under s. 1 , if for no other reason than that I cannot be persuaded that the federal objective of promoting the acquisition of citizenship is sufficiently pressing to be pursued by discriminatory means.

I.           Section 15(1) 

75                                  As my colleague Bastarache J. has pointed out, the proper approach to conducting a s. 15(1)  analysis was set out by this Court in Law, supra.  A summary of that approach is already provided in Bastarache J.’s reasons.  Nevertheless, it bears repeating (Law, at para. 88):

. . . a court that is called upon to determine a discrimination claim under s. 15(1)  should make the following three broad inquiries:

(A)     Does the impugned law (a) draw a formal distinction between the claimant and others on the basis of one or more personal characteristics, or (b) fail to take into account the claimant’s already disadvantaged position within Canadian society resulting in substantively differential treatment between the claimant and others on the basis of one or more personal characteristics?

(B)     Is the claimant subject to differential treatment based on one or more enumerated and analogous grounds?

and

(C)     Does the differential treatment discriminate, by imposing a burden upon or withholding a benefit from the claimant in a manner which reflects the stereotypical application of presumed group or personal characteristics, or which otherwise has the effect of perpetuating or promoting the view that the individual is less capable or worthy of recognition or value as a human being or as a member of Canadian society, equally deserving of concern, respect, and consideration?

Of these three inquiries, the third is undoubtedly, as Bastarache J. suggests, at para. 38, “the most challenging”.  It is also this third inquiry that has traditionally received the least amount of attention from the courts, and upon which this Court in particular has only recently begun to provide guidance.  We would do well, then, to remind ourselves of the exact purpose and function of this third branch of the Law test.

76                                  At the heart of the third Law inquiry is the recognition that not all distinctions resulting in differential treatment at law can properly be said to violate equality rights under s. 15(1)  of the Charter .  This proposition finds support in a number of judgments of this Court going back at least as far as Andrews v. Law Society of British Columbia, [1989] 1 S.C.R. 143, at pp. 168-69:        

It is not every distinction or differentiation in treatment at law which will transgress the equality guarantees of s. 15  of the Charter .  It is, of course, obvious that legislatures may – and to govern effectively – must treat different individuals and groups in different ways. . . . The classifying of individuals and groups, the making of different provisions respecting such groups, the application of different rules, regulations, requirements and qualifications to different persons is necessary for the governance of modern society.

These reflections, as McIntyre J. noted in that case, immediately give rise to the following question:  “What kinds of distinctions will be acceptable under s. 15(1)  and what kinds will violate its provisions?” (p. 169).

77                                  This Court has consistently answered that question in the following manner:  those and only those distinctions that are (a) based on enumerated or analogous grounds, and (b) discriminatory, will violate the equality guarantee in s. 15(1)  of the Charter .  Hence the three broad inquiries that were set out in Law.

78                                  It cannot be overemphasized that the third Law inquiry, requiring an investigation into whether a legal distinction made on enumerated or analogous grounds is discriminatory, is as vital to determining the presence of a s. 15(1)  violation as are the other two.  Thus, it is important to be clear about precisely what is entailed by such an investigation.

79                                  In Law, this Court stated in unequivocal terms that the appropriate perspective from which to analyse a claim of discrimination has both a subjective and an objective component (at para. 59):

As applied in practice in several of this Court’s equality decisions, . . . the focus of the discrimination inquiry is both subjective and objective: subjective in so far as the right to equal treatment is an individual right, asserted by a specific claimant with particular traits and circumstances; and objective in so far as it is possible to determine whether the individual claimant’s equality rights have been infringed only by considering the larger context of the legislation in question, and society’s past and present treatment of the claimant and of other persons or groups with similar characteristics or circumstances.  The objective component means that it is not sufficient, in order to ground a s. 15(1)  claim, for a claimant simply to assert, without more, that his or her dignity has been adversely affected by a law.

Iacobucci J. went on to say, “the relevant point of view is that of the reasonable person, dispassionate and fully apprised of the circumstances, possessed of similar attributes to, and under similar circumstances as, the claimant” (para. 60).  For clarity, he added that “the appropriate perspective is not solely that of a ‘reasonable person’ — a perspective which could, through misapplication, serve as a vehicle for the imposition of community prejudices.  The appropriate perspective is subjective-objective” (para. 61).

80                                  I do not see how these authoritative statements can be squared with Bastarache J.’s suggestion, in the case at bar, that the discrimination inquiry “is to be assessed from the perspective of the claimant” (para. 38) and that “[e]ven if the non-citizen knows the preference has nothing to do with her capabilities — as most reasonable people would — she may still feel ‘less . . . worthy of recognition . . . as a member of Canadian society’” (para. 46).  It is this aspect of human dignity which is most relevant to this appeal.  In my view, the latter comments have the effect of reading out the requirement of an objective component in the analysis of claims of discrimination.  To do so would be to allow, contrary to the dictum in Law, that it is after all sufficient, in order to ground a s. 15(1)  claim, for a claimant simply to assert without more that his or her dignity has been adversely affected by a law.

81                                  There are strong reasons for resisting this result.  To begin with, we cannot accede to it without doing irrevocable damage to the Law methodology for assessing equality claims under the Charter .  The third inquiry set out in Law would be rendered vacuous were we to resort to a purely subjective perspective in analysing claims of discrimination.  Indeed, if the claimant’s own subjective experience of discrimination were all that mattered, we might legitimately take the fact that he or she had launched a s. 15(1)  Charter  challenge, by itself, as sufficient evidence that the claimant felt his or her dignity had been adversely affected by a law.  The discrimination inquiry would thus be trivially satisfied in every case before the courts, shifting the entire analytical burden in assessing equality claims under s. 15(1)  to the first two inquiries set out in Law.

82                                  This in turn would be tantamount to adopting an approach to equality jurisprudence that was expressly repudiated by this Court in Andrews.  If nothing else, Andrews stands for the proposition that a straight line should not be drawn from the finding of a distinction — even one made on enumerated or analogous grounds — to a determination of its validity under s. 1  of the Charter .  Referring approvingly to comments made by McLachlin J.A. (as she then was) in the court below, McIntyre J. noted in that case that “the labelling of every legislative distinction as an infringement of s. 15(1)  trivializes the fundamental rights guaranteed by the Charter  . . .” (p. 181).  In my view, the same holds true even when the distinction in question is made on enumerated or analogous grounds.

83                                  In saying this, I do not mean to deny that there may in fact be certain legislative distinctions, such as those made on the basis of race, that can be labelled infringements of s. 15(1)  without the need for a detailed investigation into whether or not they are discriminatory.  Even allowing that there are such distinctions, we must not conclude that the discrimination inquiry is unnecessary and that it is sufficient, in order to establish a s. 15(1)  violation, to demonstrate that a distinction has been made on enumerated or analogous grounds.  Rather the labelling of such distinctions as s. 15(1)  violations without the need for conducting a detailed discrimination inquiry is, as it were, the exception that proves the rule.  There are some distinctions made on certain enumerated or analogous grounds — I refer again to those made on the basis of race as an obvious example — which a reasonable person could not but view as presumptively, if not unavoidably, discriminatory.  The discrimination inquiry may get short-circuited where these kinds of distinctions are at issue, not because it is unnecessary or unimportant but because its outcome will seem all too readily apparent.

84                                  In most cases, however, the mere presence of a distinction made on enumerated or analogous grounds should not, in the absence of a detailed discrimination inquiry, determine the existence of an infringement under s. 15(1) .  An approach to equality jurisprudence that gives insufficient attention to the discrimination inquiry, as McIntyre J. observed in Andrews, “virtually denies any role for s. 15(1) ” (p. 181).  The following remarks, quoted by McIntyre J. from the judgment of Hugessen J.A. in Smith, Kline & French Laboratories Ltd. v. Canada (Attorney General), [1987] 2 F.C. 359 (C.A.), at pp. 367-68, illustrate what he meant by this (Andrews, supra, at p. 180):

The rights which it [s. 15 ] guarantees are not based on any concept of strict, numerical equality amongst all human beings.  If they were, virtually all legislation, whose function it is, after all, to define, distinguish and make categories, would be in prima facie breach of section 15  and would require justification under section 1 .  This would be to turn the exception into the rule.  Since courts would be obliged to look for and find section 1  justification for most legislation, the alternative being anarchy, there is a real risk of paradox: the broader the reach given to section 15  the more likely it is that it will be deprived of any real content.

The reasons of my colleague Bastarache J. in the case at bar provide a striking example of this paradox.

85                                  Having moved quickly from a finding that s. 16(4) (c) of the PSEA  makes a distinction on an enumerated or analogous ground to the conclusion that the claimants’ s. 15(1)  rights were violated on the basis they felt subjectively discriminated against, Bastarache J. proceeds to find that the violation is justified under s. 1 .  For myself, I cannot accept that the violation of so sacrosanct a right as the guarantee of equality is justified where the government is pursuing an objective as abstract and general as the promotion of naturalization.  To find that this objective is sufficiently pressing and substantial to be pursued by discriminatory means would, I believe, leave scarcely any legitimate state objective seriously constrained by the constitutional fetter of equality.  Nor can I be persuaded that a law that supposedly undermines the essential human dignity of the claimants, and is therefore considered sufficiently egregious to fail s. 15(1)  scrutiny, is also properly characterized for the purposes of a s. 1  analysis as nothing more than an “inconvenience”, the price the claimants must “pay for the government’s right to define the rights and privileges of its citizens” (para. 71).

86                                  We must be careful, in our understandable eagerness to extend equality rights as widely as possible, to avoid stripping those rights of any meaningful content.  Lack of care can only result in the creation of an equality guarantee that is far-reaching but wafer-thin, an expansive but insubstantial shield with which to fend off state incursions on our dignity and freedom.  This of course is precisely the paradox that so exercised this Court in Andrews.  It is a paradox that will prove inescapable if we are too quick to find s. 15(1)  violations on the basis of a discrimination inquiry devoid of real content.  For we shall then be forced in almost every case to turn to a justificatory analysis under s. 1  which, although suitably rigorous in other contexts, will inevitably become diluted in the s. 15(1)  context.  The Oakes test was not designed to bear the considerable strain of salvaging under s. 1  a plethora of laws that would otherwise offend a s. 15(1)  analysis essentially lacking consideration for the existence of objectively discernible discrimination.  Yet this is exactly what s. 1  is asked to do, on pain of unravelling the legislative process, when s. 15(1)  infringements are too easily found.  In response, courts are forced to engage in a s. 1  analysis that pays an undue amount of deference to the legislatures, both in the objectives they choose to pursue and in the means they adopt in pursuing them.  For it is only by continually loosening the strictures imposed under the  test that s. 1  can discharge the onerous burden that it has been placed under.  The problem is that in thus discharging its burden s. 1  effectively denudes the equality rights guaranteed under s. 15(1) of their meaning and content while paying lip service to a broad and generous concept of equality.

87                                  It would in my opinion be preferable, from the perspectives of analytical integrity, justificatory force and fidelity to this Court’s prior equality jurisprudence, to avoid this paradox altogether.  This can only be accomplished by allowing the third branch of the Law test — the discrimination inquiry — to do the kind of sorting that it was intended to do.  Again, not all distinctions made on enumerated or analogous grounds constitute infringements of s. 15(1)  of the Charter .  We cannot do justice to this basic fact without recognizing that the proper perspective from which to analyse a claim of discrimination is not the claimant’s perspective alone.  Rather, as was stated by this Court in Law, “the relevant point of view is that of the reasonable person, dispassionate and fully apprised of the circumstances, possessed of similar attributes to, and under similar circumstances as, the claimant” (para. 60).

88                                  An appropriate emphasis on the objective component in the discrimination analysis, in addition to the subjective component, makes sense of the concept of equality rights in a way that an exclusive focus on the subjective component in the analysis cannot.  Where conducting the discrimination analysis from the perspective of the claimant alone allows the fair terms of interaction between the individual and the state — the boundaries of individual rights — to be unilaterally determined by the claimant, attention to the objective component in the analysis recognizes the essentially bilateral character of rights.  In the end a rights claim is nothing other than a legally binding demand for recognition of, and respect for, one’s interests on the part of others.  As a result it cannot avoid engaging the interests of those others.  For if others are to be duty-bound to respect one’s rights, fairness requires that they be given some say, that their own interests be taken account of, in determining those rights.

89                                  It is of course trite to point out that one’s rights end where those of others begin.  Nevertheless it is a truth that we should endeavour to keep constantly before our minds.  The objective component in the discrimination analysis gives voice to this truth by allowing equality rights to be determined inter-subjectively, with proper regard for the interests of both the individual claimant and the state, rather than subjectively, paying attention only to the interests and feelings of the claimant.  No doubt this is part of what Iacobucci J. had in mind when he stated in Law, supra, at para. 59, that the discrimination analysis is “objective in so far as it is possible to determine whether the individual claimant’s equality rights have been infringed only by considering the larger context of the legislation in question . . .” and that “[t]he objective component [in the analysis] means that it is not sufficient, in order to ground a s. 15(1)  claim, for a claimant simply to assert, without more, that his or her dignity has been adversely affected by a law.”

90                                  We should be explicit about the impact that a general implementation of this reading of s. 15(1)  of the Charter  will have on the future course of equality jurisprudence.  Admittedly, this understanding of the Law test has the effect of narrowing the range of successful Charter  challenges that could be made under s. 15(1) .  Once the subjective-objective perspective is properly applied as a necessary condition for making a finding of discrimination, it becomes more difficult to establish that one’s equality rights have been infringed.  Yet I think that it also becomes more difficult, having made a finding of discrimination, to establish that the resulting s. 15(1)  violation can be justified.

91                                  Under this approach equality rights, once found, will not be at the mercy of a s. 1  analysis that would otherwise, of necessity, be too deferential to the legislative process and hence too heedless of the importance of s. 15(1)  rights.  Freed of the need to guard the integrity of the legislative process against too-easy findings of s. 15(1)  infringements, the justificatory analysis under s. 1  will then be conducted with the uncompromising rigour that I believe it was intended to have.  No longer will keeping the legislatures functional necessitate tolerating violations of Charter  rights, the embodiments of our freedom and of this society’s most cherished values, in favour of less valued state objectives such as the one at issue in this case.  Without wanting to decide the question in advance, the class of state objectives which might then qualify as sufficiently pressing and substantial to limit equality rights under s. 1  may become restricted to only those that are most important: perhaps the need to protect the Charter  rights of others; or more generally, the need for laws that advance the values underlying the Charter , conceived of as a coherent document expressing our highest values and the supreme law of this country.  Ultimately, then, this reading of s. 15(1)  entails an ideological preference for spreading equality rights somewhat less broadly but with much greater substance.  I believe that this is what is required in order to properly situate the debate on the limits of constitutional guarantees.

92                                  It may seem that this approach to s. 15(1)  blurs the distinction between the kinds of considerations that are appropriate under that section and the kinds of considerations that are appropriate under s. 1 .  I confess that there appears to be considerable overlap between the two, but a number of points should be made in this regard.  To begin with, the overlap is to some extent merely a function of the fact that we are dealing here with a section that contains its own internal limitation (as opposed to the external limitation imposed under s. 1 ):  specifically, its differentiation between legislative distinctions and discrimination.  Similar problems in defining the precise contours of the relationship between a rights-granting provision of the Charter  and s. 1  have been faced by this Court in the past when dealing with the internal limitation in s. 7 , to take one example among many.

93                                  I do not think that these problems are insurmountable.  Nor should we assume that their resolution will be identical in the case of all internally qualified rights-granting provisions.  Thus, while in Re B.C. Motor Vehicle Act, [1985] 2 S.C.R. 486, at p. 518, Lamer J., as he then was, indicated that a limit to a right under s. 7 effected through a violation of the principles of fundamental justice could be sustained under s. 1  “only in cases arising out of exceptional conditions, such as natural disasters, the outbreak of war, epidemics, and the like”, thereby leaving virtually no role for s. 1  in the context of defining rights under s. 7, it does not follow that s. 1  would have to be accorded a similarly negligible role in the face of a s. 15(1)  violation.  The exact nature of the interaction between the two constitutional provisions should in my view be left to an incremental development of the case law.

94                                  In any event, I do not believe that analytical convenience should lead us to make perfunctory findings of s. 15(1)  violations as a mere prelude to the justificatory analysis under s. 1 .  Apart from concerns relating to the burden of proof, which rests on the claimant under s. 15(1)  but on the state under s. 1 , there is little practical difference from the perspective of the claimant between a finding of no discrimination and one of justified discrimination.  But there is considerable difference between the two from the perspective of jurisprudential integrity, for reasons that I have already set out.

95                                  As regards the burden of proof, although I do not think it necessary to resolve all of the concerns it may raise at this point, I believe that accommodations could easily be made.  For instance, I see no reason why these concerns could not be dealt with simply by recognizing that in some cases it will be reasonable for the court to infer discrimination on the basis of the circumstances themselves as well as evidence put forth by the claimant in respect of his or her own subjective experience of discrimination.  Should the circumstances warrant the drawing of such an inference, this will accrue to the claimant’s benefit.  Whether that will be sufficient for the claimant to succeed in his or her claim will then depend on the evidence relating to the existence of objective discrimination tendered by the state in order to negate this inference.  If the state produces sufficient evidence to suggest that there is no objectively discernable discrimination, this will tend to neutralize any inference drawn on the basis of the circumstances and the claimant’s own subjective testimony.  If the state chooses not to lead such evidence, it will be taking the chance that a s. 15(1)  violation will be made out on the strength of a validly drawn inference.  In other words, if need be, we could effect a partial shift in the evidential burden (as opposed to the legal burden) under s. 15(1)  to the state in order to address concerns over the burden of proof, recognizing that each party is differently situated for the purposes of leading evidence that is relevant to the different components of the discrimination analysis.

96                                  These matters aside, I turn now to an application of the foregoing analysis to the specific facts in the case at bar.

II.    Application to the Case at Bar

97                                  This Court has considered the relationship between citizenship and s. 15(1)  of the Charter  in the context of employment opportunities once before, in the case of Andrews, supra.  A superficial reading of that case might lead one to conclude that the discrimination inquiry in the instant case can be quickly disposed of in favour of the claimants.  Such a reading ostensibly finds support in the following passage (at p. 183):

A rule which bars an entire class of persons from certain forms of employment, solely on the grounds of a lack of citizenship status and without consideration of educational and professional qualifications or the other attributes or merits of individuals in the group, would, in my view, infringe s. 15  equality rights.

There are, in my view, a number of reasons for exercising caution in applying this general statement of law to the particular facts of the case at bar.

98                                  The first and most obvious of these reasons is that the statement purports to be speaking only about laws that effectively bar non-citizens from certain forms of employment.  Section 16(4) (c) of the PSEA  does not, on its face, impose such a bar; it merely creates a preference at the referral stage of open competitions for employment in the federal Public Service.  Nor can it be seriously maintained that the s. 16(4) (c) preference has the effect in practice of creating such a bar.  As my colleague Bastarache J. notes in his reasons (at para. 24):

[N]on-citizens are eligible (and indeed encouraged) to submit their resumes to the Commission for consideration . . . and . . . non-citizens who are referred by the Commission face no disadvantage compared to citizens . . .  [Moreover,] non-citizens enjoy the same privileges as citizens with respect to closed competitions; [which] are the principal means by which the Public Service fills its staffing needs.  Finally, the citizenship preference is just that: a preference.  Non-citizens are routinely referred to open competition where, in the opinion of the Regional PSC Director, there are insufficient qualified Canadians to fill the particular position; . . .

These factual findings, even by themselves, go a significant distance towards distinguishing this case from Andrews.  Indeed it is largely on the strength of these findings that Bastarache J. has decided that, unlike in Andrews, the supposed s. 15(1)  violation in this case is a relatively trivial one that can be justified under s. 1 .  I have already indicated my discomfort with the idea that any s. 15(1)  violation could be seen as a matter of mere inconvenience.  From my perspective the significance of these facts is not that they render the alleged s. 15(1)  violation any less serious, but rather that they interfere somewhat with the conclusion that s. 15(1)  has been infringed in the first place.

99                                  Seen in this light, the factual findings noted by this Court are suggestive of the need to engage in a more careful and thorough discrimination analysis before making a determination under s. 15(1)  of the Charter .  For the purposes of conducting such an analysis, this Court’s decision in Andrews is of limited assistance.  It hardly warrants mentioning that Andrews was decided without the benefit of the detailed analytical framework for assessing equality claims that was set out by this Court in Law.  This is not to suggest that Andrews would have been decided differently under the Law framework: I have little doubt that the outcome of that case would have been the same.  Still, it does suggest that we must be slow to decide the s. 15(1)  question in this case on the basis of the general proposition of law set out in Andrews and quoted above.  It remains necessary, in assessing this equality claim just as in assessing any other, to pay careful attention to the Law methodology for determining the scope of the claimants’ s. 15(1)  rights.

100                              Here as always s. 15(1)  rights extend only as far as is necessary to preserve the claimants’ immunity from laws that are discriminatory.  In Law this Court cast the question to be determined by the discrimination analysis in the following terms: “. . . does the differential treatment discriminate in a substantive sense, bringing into play the purpose of s. 15(1)  of the Charter  . . .?” (para. 39 (emphasis in original)).  Iacobucci J. went on to describe the purpose of s. 15(1)  as being “to prevent the violation of essential human dignity and freedom . . .” (para. 51).  Human dignity is thus at the centre of the discrimination inquiry.  A law will only be discriminatory for the purposes of s. 15(1)  if it can be said to violate the claimant’s essential human dignity or freedom.  Moreover, the proper perspective from which to make this assessment, as I have already emphasized, is not simply the claimant’s own subjective perspective but that of “the reasonable person, dispassionate and fully apprised of the  circumstances, possessed of similar attributes to, and under similar circumstances as, the claimant” (Law, at para. 60).  The proper perspective is thus subjective-objective: “objective in so far as it is possible to determine whether the individual claimant’s equality rights have been infringed only by considering the larger context of the legislation in question” (Law, at para. 59).

101                              In this regard, useful reference can be made to the international context within which the impugned legislation in this case is situated.  As my colleague Bastarache J. observes (at para. 56):

[V]irtually all liberal democracies impose citizenship-based restrictions on access to the national Public Service; these restrictions vary from virtual bans on federal Public Service employment (as in Switzerland and the United States) to policies allowing permanent residents to work in the Public Service on a probationary basis (as in Australia) . . . . [Moreover,] international conventions support [such] citizenship-based restrictions by guaranteeing the right of all citizens to work in the Public Service; . . .

The value of these observations, in my view, is not that they help to justify what would otherwise be discriminatory restrictions on access to the federal Public Service but that they indicate widespread international agreement that such restrictions do not implicate the essential human dignity of non-citizens to begin with.  To my mind there could scarcely be better evidence of what the reasonable non-citizen would conclude in respect of any claim of discrimination that might be made against these restrictions: in short, he or she would conclude that the partial and temporary difference of treatment imposed by these restrictions is not discriminatory.

102                              The reasonableness of this conclusion is confirmed on the particular facts of this case.  A non-exhaustive list of contextual factors is suggested in Law as relevant to the determination of whether or not, from the perspective of a reasonable person in circumstances similar to those of the claimant, the claimant’s essential human dignity is violated by an impugned law.  These include:  (a) whether those in circumstances similar to the claimant have been subjected to pre-existing disadvantage, stereotyping, prejudice, or vulnerability; (b) whether there is a relationship between the ground upon which the claim is based and the nature of the differential treatment; and (c) what the nature and scope of the interest affected by the impugned law is.  An analysis of these various factors militates against making a finding in the present case that s. 16(4) (c) of the PSEA  is discriminatory in the sense that it violates the essential human dignity of reasonable non-citizens.

A.        Pre-Existing Disadvantage

103                              In many aspects of their lives, non-citizens in general suffer from the sort of pre-existing disadvantage, stereotyping, prejudice, and vulnerability that s. 15(1)  of the Charter  is directed at remedying.  This was the basis for the holding in Andrews that non-citizenship is an analogous ground for the purposes of s. 15(1)  and that non-citizens in general are “a good example of a ‘discrete and insular minority’ who come within the protection of s. 15 ” (Andrews, supra, at p. 183).  At first blush, the present factor would therefore appear to be an aggravating one in determining whether s. 16(4) (c) of the PSEA  offends human dignity.

104                              Kept at that level of generality, however, the truths set out in the preceding paragraph are useful merely for the purpose of finding an analogous ground under s. 15(1)  and in my view tell only half the story that is relevant to this particular appeal.  My colleague, Bastarache J., has nevertheless chosen to focus almost exclusively on this half of the story (para. 45).  Ironically, while my problem with his reasons as a whole can be traced to his having adopted an insufficiently objective perspective for the purposes of conducting the discrimination inquiry, what this partial account leaves out constitutes a deficiency that arises from adopting a perspective that is also insufficiently subjective.  As Iacobucci J. stated in Law, at para. 59, the inquiry into whether an impugned law discriminates is “subjective in so far as the right to equal treatment is an individual right, asserted by a specific claimant with particular traits and circumstances” (emphasis added).  In light of this required focus on the particularity of the claimant, I do not believe that the question of pre-existing disadvantage can be settled in this case simply by adverting, without further ado, to the non-citizenship status of the claimants.

105                              Indeed, on closer inspection there is considerable room for doubt as to whether the particular traits and circumstances of these specific claimants are such that the claimants can properly be said to suffer from pre-existing disadvantage.  Especially telling in this regard is the fact that, at least in the case of two of the claimants, their continuing status as non-citizens is a matter of personal choice.  As noted by Marceau J.A. in the court below (Lavoie v. Canada, [2000] 1 F.C. 3 (C.A.), at para. 3):

All three plaintiffs are citizens of European countries and, at least two of them, who are citizens of Austria and the Netherlands — countries which do not permit the holding of dual citizenship — in their testimony conceded that they refused to naturalize to Canada in part because of the benefits concomitant with their original citizenships, one of which was preferential employment within the European Union countries, and within the public service of their native countries.

In the circumstances, I have difficulty with the proposition that these claimants suffer from pre-existing disadvantage as a result of their status as non-citizens.  On the contrary, this is in some ways a case about the maintaining of pre-existing advantage by the claimants, who want to retain all of the valuable benefits legally accruing to them as members of the European Union and citizens of other countries while claiming similar privileges and benefits afforded to Canadian citizens under an analogous legislative arrangement.  If there is any disadvantage here it arises principally from the fact that their countries of citizenship do not permit these claimants to hold dual citizenship (a disadvantage not suffered by the more fortunate third claimant, who was able to obtain her Canadian citizenship in 1991 without having to relinquish her French citizenship).  Yet this disadvantage is not suffered as a result of their status as non-citizens of Canada — whose laws, I note in passing, do permit the holding of dual citizenship — but as a result of their status as citizens of other countries.  In any event, a reasonable observer might conclude that this is something of an enviable problem to have, as is evidenced by the fact that the claimants are themselves unwilling to remedy their claimed disadvantage by naturalizing to Canada.  Let me add, in connection with this last point, that in my view the acquisition of Canadian citizenship for the purpose of being granted equal access to the federal Public Service cannot be said to come at an unacceptable personal cost.  If there are any costs involved at all — and there are none in the case of non-citizens who are citizens of countries that permit dual citizenship — those costs are relatively minor and are, as I have already suggested, properly attributable not to the acquisition of Canadian citizenship per se but to the fact that other countries do not always permit the holding of dual citizenship.

106                              I recognize that not all those affected by s. 16(4) (c) of the PSEA  are likely to be as advantaged in the ways just canvassed as are the claimants in this appeal.  Nor are all non-citizens likely to be as well educated as these claimants.  Given this, and consonant with the holding in Andrews that non-citizens are in general a disadvantaged group, I will not go so far as to conclude that the preceding analysis favours a finding that s. 16(4) (c) does not offend human dignity.  Still, it is difficult to find any pre-existing disadvantage in the particular case at bar and to that extent difficult to locate a violation of human dignity.  I therefore conclude that this contextual factor is indeterminate.

B.     Relationship Between Grounds and the Nature of the Differential Treatment

107                           A more damaging contextual factor for the claimants in this case is the second factor set out in Law: “the relationship between the ground upon which the claim is based and the nature of the differential treatment” (para. 69).  In Law this Court recognized that it will in general be “more difficult to establish discrimination to the extent that legislation properly accommodates the claimant’s needs, capacities and circumstances” (para. 70).  It further recognized that “[s]ome of the enumerated and analogous grounds have the potential to correspond with need, capacity, or circumstances” (para. 69).  Iacobucci J., at para. 71, cited Weatherall v. Canada (Attorney General), [1993] 2 S.C.R. 872, as an example of a case in which legislation “quite properly treated a claimant differently on the basis of actual personal differences between individuals” and where “it was stated that the decision to permit cross-gender prison searches of male prisoners but not of female prisoners likely did not violate s. 15(1) , because such a difference in treatment was appropriate in light of the historical, biological and sociological differences between men and women”.

108                           The combined effect of these statements is that not all cases of legislated differential treatment on the basis of an enumerated or analogous ground will give rise to a valid claim of discrimination.  In particular, where, as in Weatherall, the ground upon which the claim is made (in that case, the enumerated ground of sex) actually corresponds to personal differences that are relevant to the legislative purpose, the claimant will have difficulty in proving a violation of essential human dignity, even if differential treatment on the basis of that ground is unjustifiable in the vast majority of cases (as it is in most cases where legislated differences in treatment are based on the sex of those affected by the impugned law).

109                           The instant case provides a further example of this.  Even if the use of the analogous ground of citizenship as a basis for withholding advantages from some individuals while extending them to others is discriminatory in the context of a provincial law using citizenship as a proxy for loyalty or trust-worthiness, as it was found to be in Andrews, it does not follow that it is discriminatory in the context of a federal law that forms part of a package of incentives to naturalize while at the same time defining the rights and duties of Canadian citizens.  The reason for this is clear.  In the second case, but not in the first, there is actual correspondence between the ground of citizenship itself and the nature of the differential treatment.

110                           It is the essence of the concept of citizenship that it distinguishes between citizens and non-citizens and treats them differently.  As the respondent’s expert, Professor Schuck, explained in evidence given by way of affidavit:

[The] political, emotional, and motivational purposes of citizenship cannot be fully achieved unless there is a difference in legal status between citizens and non-citizens, a difference that can help motivate non-citizens to invest the time, energy, and resources necessary to acquire [citizenship] . . . . Were the differences in rights and status between citizens and non-citizens completely eliminated so that all rights available to citizens were also immediately and equally available to non-citizens, the notion of citizenship would become meaningless.  [Emphasis in original.]

At issue in this case is a federal law that is validly enacted in the exercise of exclusive federal jurisdiction over matters of citizenship for the dual purposes of defining one of the historical and internationally recognized entitlements of citizenship and providing an incentive to naturalize.  As such, s. 16(4) (c) of the PSEA  cannot help but give rise to differential legal treatment: for that is precisely what is entailed in the act of legislating over matters of citizenship.

111                           By way of contrast, there was no such correspondence between the differential treatment in Andrews and the ground upon which that treatment was based.  In that case, the law in issue was a provincial law establishing qualifications for admission to the practice of law in British Columbia.  The law had nothing to do with citizenship per se, or defining the entitlements of citizenship.  In fact, it could not have purported to be directed at matters of citizenship since such matters are not within the competence of provincial legislatures.  Instead, the law was merely a profession-regulating law within provincial jurisdiction which drew a distinction between citizens and non-citizens for the purposes of setting an entry requirement to the legal profession.  It was in this context that the differential treatment in that case was found to be based on an irrelevant ground of distinction and thus discriminatory.

112                           While my colleague, Bastarache J., seems to acknowledge this distinction between the present case and Andrews — and, more pointedly, accepts that the law in this case is “meant to further Canada’s citizenship policy by granting citizens certain privileges not enjoyed by immigrants” (para. 54) — he nevertheless concludes that this “citizenship argument goes beyond what is contemplated by the second contextual factor in Law” (para. 43).

113                           This conclusion, which effectively limits the operation of the second contextual factor set out in Law, reflects an inappropriately exclusive focus on the subjective perspective of the claimant.  Only on the supposition that it is the claimant’s interests alone that are implicated by the discrimination inquiry does one arrive at the conclusion that the third branch of the Law test cannot have the result of curtailing the scope of claimants’ rights.  Yet as I have emphasized throughout these reasons, this supposition mistakes the proper perspective from which the discrimination inquiry is to be conducted.  The proper perspective is not the purely subjective perspective of the claimant but a perspective that is “both subjective and objective:  . . . objective in so far as it is possible to determine whether the individual claimant’s equality rights have been infringed only by considering the larger context of the legislation in question” (Law, supra, at para. 59).  This mandates a consideration of the larger context in which the claimant’s interests are not the only interests that figure in the discrimination inquiry.  What is required is that the claimant’s interests be defined and constrained by reference to those other interests that are revealed through a contextual analysis.

114                           Nowhere is this requirement more evident than in the context of laws that purport to govern matters of citizenship by defining the incidents thereof and providing incentives to naturalize.  Citizenship law is about defining not just the rights of citizens but also their correlative duties towards the state.  These include: “voting in elections; obeying the laws of Canada; respecting the rights and freedom of others; working to help others in the community; eliminating discrimination and injustice; and caring for Canada’s heritage” (N. M. Berezowski and B. J. Trister, Citizenship 1996 (1996), at pp. 5-6).  I might also hasten to add the “right” to sit on jury, which is more commonly referred to as the right to serve as a juror, or more simply as “jury duty”.  In connection with this I note with bemusement my colleague Bastarache J.’s observation that the claimants’ “subjective reaction to the citizenship preference no doubt differed from their reaction to not being able to vote, sit in the Senate, serve on a jury, or remain in Canada unconditionally” (para. 52).  This is quite the point: these other incidents of citizenship are at best likely to be perceived as a matter of indifference to non-citizens and at worst likely to be perceived more as a burden than as a benefit.  The latter is especially true in the case of serving on juries, which many prospective jurors see as a major imposition, and in the case of voting, which some may view merely as a civic duty to be performed perfunctorily if at all.

115                           The challenge faced by the federal government, in the light of these observations, is to establish a package of incentives — rights and privileges of citizenship — that will provide sufficient motivation for non-citizens to naturalize and in the process take on these more burdensome incidents, or duties, of citizenship.  It can only do this by distributing rights and benefits unequally between citizens and non-citizens.  This differential treatment should not, however, be viewed entirely from the one-sided subjective perspective of the claimant.  Indeed, legislating over matters of citizenship can only be understood as an exercise directed at achieving mutual respect and recognition, or reciprocal concern, between the citizen and the state.  It is only insofar as the individual submits selflessly to the demands and duties imposed by membership in the state that the state reciprocally submits fully to the individual’s needs by according him or her the entire complex of advantages that are the concomitants of state membership.  Citizenship is thus relevant to the public distribution of benefits to the extent that it tracks the class of people who have taken on correlative or reciprocal duties in exchange for the receipt of the benefits in question.

116                           There are naturally limits to the extent to which this relevance will obtain.  Some benefits — the provision of basic health and policing services, for example — may in fact, though I need not decide this question here, be owed to all persons as of right just by virtue of their humanity.  In the case of such benefits, the right to an equal share will not be contingent upon an act of reciprocity by the recipient.  It follows that these benefits, assuming them to exist, cannot be tied to the rights of citizenship.  But it is not suggested here that immediate access to employment in the federal Public Service falls within this category of benefits, such that we can dispense with any notion of reciprocity.  As is noted in Citizenship 1996, supra, at p. 6:

Canada has no military draft nor are its citizens required to partake in obligatory military service.  The Canadian scheme adheres to an individualist definition of citizenship with few economic obligations toward the state and a wide degree of liberty, thus leaving it to the individual conscience in determining his or her duties as a citizen.

In such circumstances, we might reasonably accord the state a similarly wide latitude in determining some of the special rights of citizenship, including the one at issue in this case.

117                           In conclusion, I find that this contextual factor militates strongly against finding that s. 16(4) (c) of the PSEA  is discriminatory in the sense that it violates human dignity.  Once an appropriately subjective-objective perspective is adopted it becomes evident that there is a valid state interest in tying the receipt of certain benefits to citizenship such that the withholding of those benefits from non-citizens cannot constitute an affront to human dignity.  Use in this case of the analogous ground of citizenship as a basis for legislating differential treatment between individuals is both: (a) unavoidable, inasmuch as legislating over matters of citizenship itself entails differential treatment between citizens and non-citizens; and (b) appropriate, inasmuch as the ground of citizenship corresponds to real personal differences between the various individuals who would claim benefits from the state.

C.        Nature of the Interest Affected

118                              The fourth and final contextual factor in the non-exhaustive list mentioned in Law as potentially relevant to the discrimination inquiry (I have skipped over the third factor—whether the legislation is ameliorative in purpose or effect—since it clearly has no application to this case) requires a determination of the nature and scope of the interest affected by the impugned legislation.  As was stated by L’Heureux-Dubé J. in Egan v. Canada, [1995] 2 S.C.R. 513, at para. 63, and affirmed in Law, supra, at para. 74:

[T]he more severe and localized the . . . consequences [of the legislation] on the affected group, the more likely that the distinction responsible for these consequences is discriminatory within the meaning of s. 15  of the Charter .

Iacobucci J. went on in Law, at para. 74, to further elaborate on L’Heureux-Dubé J.’s comments in Egan by indicating that:

. . . the discriminatory calibre of differential treatment cannot be fully appreciated without evaluating not only the economic but also the constitutional and societal significance attributed to the interest or interests adversely affected by the legislation in question.

In my view the nature and scope of the interests affected by s. 16(4) (c) of the PSEA  are not sufficiently vital and large, nor the effects of that provision sufficiently severe and localized, to allow the claimants to successfully make out a violation of their essential human dignity.

119                              I cannot accept my colleague Bastarache J.’s overly broad characterization of the interest at issue in this case as an interest in “employment” or “work” itself.  It is of course true, as this Court has repeatedly held, that “work is a fundamental aspect of a person’s life, [implicating his] livelihood, self-worth and human dignity” (para. 45).  There are however a number of ways in which the interest at stake in the present case falls considerably short of being an interest in work per se.

120                              To begin with, one should not overlook the various features of s. 16(4) (c) of the PSEA  that serve to limit the scope of the interest it affects.  Bastarache J. identifies these features as follows (at para. 61):

(1) the fact that it is a preference only and not an absolute bar[;] (2) the fact that it does not apply to closed competition, which is the most common means of staffing Public Service positions [; and] (3) the fact that it only applies to the referral stage of open competition, not the inventory or eligibility stage, . . .

When one adds to this list of features the fact that s. 16(4) (c) only regulates access to the federal Public Service, leaving access to provincial Public Service entirely unrestricted to non-citizens, it becomes difficult to see how the interest it implicates can be appropriately characterized as an interest in work itself.  Unlike Andrews, this is not a case in which the claimants are simply refused entry into their chosen profession because of their status as non-citizens.  These various features of s. 16(4) (c)  ensure that the claimants here need neither leave their province of residence in order to find work in their chosen field nor even settle for employment in the provincial Public Service if the federal Public Service is what they would prefer.

121                              Indeed, upon more careful scrutiny it becomes apparent that the only interest that is really at stake here on the side of the claimants is something more akin to a lost chance than to an interest in employment itself.  In this respect, the instant case is once again distinguishable from Andrews.  The claimant in Andrews had met all the requirements necessary to the practice of law in British Columbia and was prevented from doing so only because of a law that excluded him from the profession on the basis of his status as a non-citizen.  His interest in securing employment as a lawyer was thus sufficiently crystallized—sufficiently proximate in the sense that it was entirely within his control to do so but for the legal impediment at issue—that to deny him access to the profession on the irrelevant (in that case) ground of citizenship necessarily touched his essential human dignity.  There is no parallel to this situation in the present case.  Even if the citizenship preference in s. 16(4) (c) of the PSEA  were to be struck down there is no sense in which the claimants here would be ensured, as Mr. Andrews would have been, of being able to pursue employment in their chosen field.  It would still be necessary for them to go on to succeed in open competition with others in order to secure any given position.  Thus the nature of the interest here is not proximate and crystallized as it was in Andrews but remote and tenuous.  At most, what s. 16(4) (c) deprives these claimants of is a chance to enter into open competition with others for positions in the federal Public Service.  In fact it is not even clear that it deprives them of this much given that, as my colleague, Bastarache J. points out, “[n]on-citizens are routinely referred to open competition where . . . there are insufficient qualified Canadians to fill the particular position” (para. 24) and “the appellants Bailey and To-Thanh-Hien were themselves referred to open competition” (para. 62).

122                              Thus, in terms of both its nature, which is tenuous and remote, and its scope, which is considerably narrowed by the limiting features of s. 16(4) (c), the interest at stake in this case does not approach being an interest in employment, or work, per se.  Properly understood, in fact, this interest has little connection to the essential human dignity of the claimants.  The foregoing analysis has already suggested why the effects of s. 16(4) (c) on the claimants cannot be considered particularly severe.  I would conclude by adding that they are not very localized either, as my colleague Bastarache J.’s own reasoning seems to imply (at para. 71): “it is almost as difficult for citizens to enter the Public Service as non-citizens; thus, the latter’s disadvantage relative to the former does not appear significant” (emphasis in original).

III.    Conclusion

123                              In light of all of this, I find that the appellants have failed to establish that their claim satisfies the third branch of the Law test for assessing equality claims under s. 15(1)  of the Charter .  The reasonable person in circumstances similar to those of the claimants would, upon consideration of the various contextual factors set out in Law, conclude that s. 16(4) (c) of the PSEA  does not offend the essential human dignity of the claimants and therefore does not discriminate.  I would accordingly dismiss this appeal.

The following are the reasons delivered by

124                              LeBel J. — With respect for other views forcefully held in this case, I share Justice Arbour’s opinion that s. 16(4) (c) of the Public Service Employment Act, R.S.C. 1985, c. P-33 , does not violate s. 15  of the Canadian Charter of Rights and Freedoms .  The appellants’ claim does not meet the third branch of the test designed in Law v. Canada (Minister of Employment and Immigration), [1999] 1 S.C.R. 497.  The citizenship preference does not affect the essential dignity of non-citizens.  It is but a stage in an open process of integration in a fully shared citizenship.  During this period, the future citizen is not viewed as an inferior member of Canadian society, but as a person who will be entitled to the full rights of citizenship and will have to bear its burdens and obligations in the near future.  This person is fully valued in the eyes of others as someone who is engaged in the process of becoming a citizen.  If a person chooses to remain outside this process, by reason of the application of foreign legislation and not of Canadian law, this has little to do with a claim of discrimination.  If this is so, it is largely self-inflicted and does not flow from state action in Canada.

125                              Given this conclusion, I do not need to discuss whether s. 1  could justify a breach of s. 15  in this case.  I will thus refrain from expressing views which would be just so much obiter literature.  I feel it necessary, though, to express my disagreement with my colleague Arbour J.’s approach to the Oakes test.  In my view, it fails to reflect jurisprudential developments since Oakes which acknowledge that the minimal impairment branch of the test may leave a significant margin of appreciation as to the selection of the appropriate remedies to Parliament and legislatures, provided they fall within a range of reasonable alternatives as Bastarache J. points out in his opinion.  For these reasons, I agree that the appeal should be dismissed.

Appeal dismissed, McLachlin C.J. and L’Heureux‑Dubé and Binnie JJ. dissenting.

Solicitors for the appellants Elisabeth Lavoie and Jeanne To‑Thanh‑Hien:  Jewitt & Associates, Ottawa.

Solicitors for the appellant Janine Bailey:  Raven, Allen, Cameron & Ballantyne, Ottawa.  

Solicitor for the respondents:  The Deputy Attorney General of Canada, Ottawa.

Solicitors for the intervener:  Joanne St. Lewis and Milton James Fernandes, Montréal.