Vriend v. Alberta: Making the Private Public

Vriend v. Alberta: Making the Private Public

Timothy Macklei”

Constitutions bind governments, not private per-
sons. Or so it has always been assumed. Yet if constitu-
tions bind not only what a government does but also
what it does not do, then they bind private persons as
much as governments; for in that case a government’s
failure to secure constitutional values in the dealings
between private persons would be unconstitutional. If
that is so, then the courts, not the legislatures, would
have the exclusive power to decide whether the private
should be made public and subjected to the public re-
quirements of liberty and equality.

In Vriend v. Alberta the Supreme Court of Canada
has confirmed that the Charter does not apply to pri-
vate persons and so does not apply to government
omissions generally. Yet the Court has also held that the
Charter does apply to the discriminatory treatment of
one private person by another on the basis of sexual
orientation, in that it does apply to a government’s fail-
ure to prohibit such treatment. Are these coherent con-
clusions? More importantly, are they the right conclu-
sions for Canadians generally or for the gays and lesbi-
aas to whose benefit they were principally directed?

This comment explores the implications of hold-
ing governments liable for their failure to act in fulfill-
ment of Charter values. Should constitutions bind pri-
vate persons in this way? Should the courts take upon
themselves ultimate authority over the values contained
in the Charter? There are reasons to think not. Are
there better ways, then, of using the Constitution to se-
cure justice for gays and lesbians?

Les constitutions lient les gouvemements, non pas
les personnes privies. Du moins c’est ce que l’on a
toujours prdsum& Cependant, si une constitution lie
non seulement ce qu’un gouvemement fait, mais aussi
ce qu’il ne fait pas, alors il lie 6galement les personnes
privdes, parce que dans un tel cas le manquement du
gouvemement A son devoir d’assurer les valeurs cons-
titutionnelles dans le cadre de transactions entre per-
sonnes privdes serait inconstitutionnel. Si tel dtait le
cas, alors les tribunaux, et non les 1dgislatures, auraient
le pouvoir exclusif de d6cider si le privd devrait devenir
public et tre subjugu6 aux exigences publiques de li-
bert6 et d’dgalitd.

Dans Vriend c. Alberta la Cour supreme du Ca-
nada a confirm6 que la Charte ne s’appliquait pas aux
personnes prives et qu’elle ne s’appliquait done pas
g6ndralement aux omissions gouvernementales. Toute-
fois, ]a cour a 6galement dacid6 que la Charte
s’appliquait au traitement discriminatoire par une per-
sone priv~e envers une autre en raison de son orienta-
tion sexuelle, dans le sens oit Ia Charte s’appliquait au
manquement du gouvemement d’avoir prohib6 une
telle discrimination. Peot-on concilier ces deux conclu-
sions? De plus, sont-elles les bonnes conclusions pour
les canadiens en gdndral ou pour les personnes homo-
sexuelles qu’elles ont cherch6 hk bdn6ficier principale-
ment?

La chronique en question explore la port~e d’une
d6cision de tenir les gouvemements responsables pour
leur manquement d’agir dans l’esprit des valeurs con-
tenues dans la Charte. Les constitutions devraient-elles
lier les personnes privies d’une telle fagon? Les tribu-
naux devraient-ils assumer une autorit6 ultime sur
l’applidation de ces valeurs? ll existe de nombreuses
raisons pour penser que non. Y a-t-il alors de meilleu-
res fagons d’utiliser Ia Constitution afin d’assurer Ia
justice pour les personnes homosexuelles?

” Lecturer in law, King’s College, London. I would like to thank John Gardner, Peter Oliver and
David Lillico for their help and comments on this piece.

McGill Law Journal 1999

Revue de droit de McGill 1999
To be cited as: (1999) 44 McGill L.J. 197
Mode de r6fdrence: (1999) 44 R.D. McGill 197

MCGILL LAW JOURNAL / REVUE DE DROITDE MCGILL

[Vol. 44

Introduction

I. The Facts

II. Red Herrings

III. Actions and Omissions

IV. Constitutions as Constraints on Power

V. The Supreme Court’s Reasons

A. Liability for Omissions
B. The Application of the Charter to Private Bodies
C. Mirroring

VI. The Impact on Vulnerable Minorities

A. Legal Implications
B. Political Implications
C. The Importance of Reasons

VII. Alternative Solutions

1999]

T MACKLEM – VRIEND V. ALBERTA

Introduction

There is good reason to be grateful for the Supreme Court of Canada’s decision in
Vriend v. Alberta,’ and yet as good or greater reason, unfortunately, to regret it. This
may seem an overly harsh, even churlish response to what is clearly-in terms of its
immediate consequences at least-an enlightened ruling. After all, Vriend corrected
what was acknowledged by all but a few militant heterosexists to be a grave injustice
against gays and lesbians in Alberta, namely, the continuing refusal of their govern-
ment to extend to them the full protection of anti-discrimination legislation, a protec-
tion that gays and lesbians had previously been accorded in virtually every other ju-
risdiction in Canada.

What is more, in achieving this end Vriend illustrated once again an attractive and
yet unforeseen consequence of the entrenchment of the Canadian Charter of Rights
and Freedoms;2 it showed that in certain circumstances at least, the Charter enables a
decision to be made that most agree should be made but that the legislature is, for one
reason or another, politically incapable of making. Prior to Vriend, the government of
Alberta had stoutly refused to accede to repeated demands that it add “sexual orienta-
tion” to the list of prohibited grounds of discrimination in the Individual’s Rights
Protection Act,’ despite the fact that those demands- were widely recognized to be le-
gitimate. Yet when the Supreme Court of Canada invoked the Charter to do precisely
what the government of Alberta had refused to do, the government of Alberta ac-
knowledged that the Court’s decision to act in its stead had been right and proper. In
effect, the Charter permitted the Supreme Court to break a political deadlock, a
deadlock in which a government had been held hostage to a political position that it
did not ultimately approve of. Had it not been for the Charter, no amendment to the
Individual’s Rights Protection Act could have been made, at least not for some time
and not without significant changes in the political landscape. Yet once made, the
amendment was broadly acceptable to the Alberta government and to the Alberta
public. Clearly, then, not only gays and lesbians, but all Albertans who accept the jus-
tice of their cause have reason to be grateful to the Charter and to the Supreme Court.
Nor is Vriend an entirely isolated case in this respect The sequence of events that
took place in Vriend is simply the reverse of that which occurred in R. v. Morgen-
taler,” where the Supreme Court removed from the Criminal Code’ provisions gov-
erning the obtaining of abortions. Subsequent events in Parliament made it clear that
no majority in Parliament supported those provisions, but no majority in Parliament

‘ [199811 S.C.R. 493, 156 D.L.R. (4th) 385 [hereinafter Vriend cited to S.C.R.].
2 Part I of the Constitution Act, 1982, being Schedule B to the Canada Act 1982 (U.K.), 1982, c. 11

[hereinafter Charter].
3 R.S.A. 1980, c. 1-2. The Act is now known as the Human Rights, Citizenship and Multiculturalism

Act, R.S.A. 1980, c. H-11.7.

4 [1988) 1 S.C.R. 30,44 D.L.R. (4th) 385.
5 R.S.C. 1985, c. C-46.

200

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

[Vol. 44

could have removed or modified them either. Cases like these suggest intriguing and
appealing prospects for the Charter. All are familiar with the idea that the Charter
was entrenched so as to protect minorities from the will of the majority, at least in
certain realms. What the Charter has done in each of these settings, however, is to
protect minorities from the consequences of an absence of will on the part of the ma-
jority. In all those cases where democratic inertia, supported by the voice of a malign
but politically significant minority, would otherwise prevent a change in the status quo
and so ensure that an unjust state of affairs is maintained, the Charter supplies the
necessary dynamic for change. This is undoubtedly a good thing. The Supreme
Court’s decision in Vriend, then, confirms that there is yet another reason to be grate-
ful for the presence of the Charter in our political life.

So why is there as much reason to complain as to cheer? Surely in cases like this it
is the result that matters, and the result in Vriend is not simply that a small measure of
justice has been secured to a historically disadvantaged group in Canadian society, but
that the capacity of the Charter to help others who find themselves in the same posi-
tion as that group, namely, the position of being disadvantaged in their relations with
their fellow citizens and unable to obtain the protection and support of their govern-
ment, has been confirmed and extended.

Of course it is the result that matters here, but results matter for reasons and it is.
the startling weakness of the reasons in Vriend that brings its results into doubt. An-
other way to put this is to say that the reasons for judgment in Vriend are part of its
result, and the weakness of those reasons is likely to cause as much harm to gays and
lesbians as did the omission of sexual orientation from the Individual’s Rights Protec-
tion Act, harmful as that was. It simply cannot be the case that a government is con-
stitutionally liable for any failure on its part to enforce Charter values in dealings
between private citizens, for otherwise the Charter would cease to be a constitutional
document, limiting the power of the government over the people-as it states itself to
be-and would instead become what all deny that it is or should be, namely, a politi-
cal mandate or agenda, its content determined by the courts, limiting the power of the
people over their government. If that assertion is true, as is the thesis of this comment,
and if a government cannot be held liable for every failure to legislate, then one of two
things follows. Either there is no constitutional liability for a government’s failure to
legislate (except in those rare cases where there is an explicit constitutional duty to
legislate, as all were agreed there was not in Vriend), or a principled basis must be
found that will enable the courts and the public to distinguish those omissions for
which a government can be held constitutionally liable from those for which it cannot.
Vriend, unfortunately, offers no such basis, and its failure to do so damages not only
the reputation and standing of the Supreme Court, but the cause of gays, lesbians, and
other disadvantaged groups in Canadian society.

I. The Facts

At the time that the events giving rise to his complaint took place, Delwin Vriend
was employed as a coordinator in a laboratory at King’s College in Edmonton. He
was also a homosexual, a fact that he disclosed to College authorities, at their request,

1999]

T MACKLEM – VRIEND V. ALBERTA

in 1990. In early 1991, he was dismissed from his position at the College, purely on
the ground of his homosexuality. At no point was it ever suggested by the College
that, his homosexuality aside, Vriend had in any way failed to do his job. Nor was it
ever suggested that the fact of his homosexuality had any functional bearing upon his
capacity to discharge his duties as a laboratory coordinator, as it might conceivably
have been suggested in certain unusual circumstances. On the facts, Vriend was dis-
missed solely because he was gay and without any pretence that not being gay con-
stituted a bonafide occupational requirement in his particular field of employment.

In most parts of Canada, such treatment would have constituted unlawful discrimi-
nation. Not so in Alberta. At the time that the dismissal took place, that is to say, be-
fore Vriend’s successful challenge to the law, Alberta’s human rights legislation, the
Individual’s Rights Protection Act, offered no protection against discrimination on the
basis of sexual orientation. What is more, the Charter, which did prohibit such dis-
crimination,’ had been held in RWDSU v. Dolphin Delivery Ltd.’ to bind only gov-
emments, and so to have no application to private institutions such as King’s College.
The result of this gap in the law was that Vriend had no legal recourse for the wrong
that had plainly been done to him. As the law then stood, private employers in Alberta
were free to discriminate on the basis of sexual orientation, notwithstanding the fact
that every employer in most other provinces and every public employer in Alberta
would have been forbidden to do so.

Understandably but also bravely, Vriend refused to accept this state of affairs. Six
months after his dismissal, he attempted to file a complaint with the Alberta Human
Rights Commission alleging a breach of the Individual’s Rights Protection Act. When
the Commission rejected the complaint as inadmissible, as it was bound to do, he
sought a declaration in the Alberta courts that the failure of the Alberta government to
prohibit discrimination on the basis of sexual orientation constituted a violation of
section 15 of the Charter. The Alberta Court of Queen’s Bench granted Vriend the
relief he sought, and ordered that a prohibition against discrimination on the basis of
sexual orientation be read into the Act. The Alberta Court of Appeal reversed this de-
cision, but the Supreme Court of Canada reversed again, allowed Vriend’s appeal, and
restored the order made by the trial judge.

II. Red Herrings

There is a real difficulty in analyzing Vriend and in addressing the very important
issues that it raises, and that is that the case has become highly charged politically, and
so has attracted to itself a number of arguments-whether about the content of gay
rights or about the wisdom of judicial deference-that have no real bearing on the
relatively straightforward question that Delwin Vriend sought to bring before the
courts. What is more, these peripheral political arguments have not simply obscured,

6 Egan v. Canada, [1995] 2 S.C.R. 513, 124 D.L.R. (4th) 609 [hereinafter Egan cited to S.C.R.].
[1986] 2 S.C.R. 573,33 D.L.R. (4th) 174 [hereinafter Dolphin Delivery cited to S.C.R.].

MCGILL LAW JOURNAL / REVUE DE DROITDE MCGILL

[Vol. 44

but have actually distorted the resolution of Vriend’s straightforward question. It
seems fair to say that in many respects his case has been hijacked by those who wish
to use it as a vehicle for issues that it does not genuinely raise. Had this tendency to
mistake and obscure the real issues in Vriend been confined to journalists and aca-
demics, there might have been little harm in it, for confusion is harmful only where
clear guidance is both reasonably expected and reasonably depended upon, and few
would dispute that journalistic and academic audiences today are anything other than
highly skeptical of what they read. What is distressing in Vriend, then, is that quite
unusually and quite seriously the tendency to mistake the issues in the case has actu-
ally entered into the reasons for judgment of the Alberta Court of Appeal and the Su-
preme Court of Canada, bodies from whom clear guidance is indeed both reasonably
expected and depended upon. For that reason, it is important now to make clear what
Vriend was not about, despite what the courts may have said.

First and most obviously, Vriend contributes little or nothing to the debate over the
content of gay and lesbian rights. The once much contested question of whether sex-
ual orientation should be added to the grounds of discrimination set out in Canada’s
human rights codes, and further, to the grounds of distinction prohibited by section 15
of the Charter, has long since been resolved in favour of gays and lesbians. By 1991,
when Vriend launched his complaint, and certainly by 1997, when that complaint
reached the Supreme Court of Canada, few if any people seriously pretended that
gays were not both discriminated against in Canadian society and greatly disadvan-
taged as a result; or that such discrimination could in any way be said to be justified;
or that a legal prohibition was not essential to removing that discrimination; or, what
amounts to the same thing, that failure to enact a legal prohibition did not compound
that discrimination. In short-and as its subsequent behaviour in response to the Su-
preme Court’s decision tacitly recognized-the Alberta government certainly should
have amended the Individual’s Rights Protection Act so as to prohibit discrimination
on the basis of sexual orientation, whether or not it was compelled to do so by the
terms of the Charter. Vriend’s case turned, therefore, not on the question of whether a
change in the law was morally required, but on the question of whether the courts
were the proper body to effect that change. Or to put the point in the language of the
Charter, Vriend’s case did not turn on the question of whether there had been a
breach of section 15, or the question of whether such a breach could possibly be justi-
fied under section 1, but on the question of whether the Charter had any application to
conduct that if constitutionally reviewable was indisputably unequal, discriminatory
and unjustifiable. If the Charter applied, then the courts were empowered, indeed
obliged, to remedy an obvious injustice. If the Charter did not apply, then the respon-
sibility for remedying that injustice would have to remain with the legislature.

Second, and despite all that has been said on the subject by the Alberta Court of
Appeal and the Supreme Court of Canada, not forgetting the Toronto Globe and Mail,
Vriend contributes little or nothing to the debate over the proper degree of judicial
deference to the legislature, whether generally or in the course of fashioning a remedy
for the breach of a constitutional right or freedom. Once it had been decided that a
failure to legislate was subject to Charter review and, what necessarily followed, that
a failure to legislate against discrimination on the basis of sexual orientation

1999]

T MACKLEM – VRIEND V. ALBERTA

amounted to an unjustifiable breach of section 15 of the Charter, then the only possi-
ble remedy was to read the missing prohibition into the legislation. Just as with the is-
sue of the moral status of sexual orientation discrimination, in 1992, when Vriend
brought his application, and certainly by 1998, when the Supreme Court of Canada
finally disposed of that application, few if any seriously pretended that the device of
reading a provision into legislation was not, in certain circumstances at least, a legiti-
mate remedy for the breach of a Charter right or freedom: the Supreme Court had de-
cided as much in Schachter v. Canada and both academic and public opinion had
rightly endorsed that decision.

It is true, of course, that in any particular case it remains a question, as it must,
whether the circumstances of that case warrant such a remedy, and it is also true that
in Vriend Major J. held that they did not. Yet on the facts of Vriend this is a highly
implausible conclusion, and one that Major J. had obvious difficulty in finding argu-
ments to support. In principle and in law the device of reading a provision into legis-
lation in order to bring it into conformity with the terms of the Charter is unaccept-
able only where there is more than one way of remedying a constitutional defect, so
that by reading the provision into the legislation the Court does not merely bring the
legislation into line with the terms of the Charter-as it is constitutionally bound to
do-but goes further, so as to make a political decision as to which of several equally
constitutional options it prefers, a decision that is not governed by the terms of the
Charter and that the Constitution gives it no power to make. This was not the case in
Vriend, however, where the only possible response to the finding that a failure to pro-
hibit discrimination on the basis of sexual orientation was itself discriminatory was to
read that prohibition into the legislation.” Once the Supreme Court had decided that a
failure to prohibit discrimination on the basis of sexual orientation was a breach of the
Charter, what other options were available to the Alberta legislature that ought to
have been left to it by the Court? In the circumstances, the legislature could only have
done what the Supreme Court had held it must do, in the terms laid down by the
Court, unless it chose to invoke section 33, the notwithstanding clause, so as to oust
the Charter obligation that the Court had identified, something that it remained as free
to do in response to the Court’s decision to read in as it would have been had the
Court decided to strike down the entire statute or to stay its remedy.'” Vriend offered

‘ [1992] 2 S.C.R. 679, 93 D.L.R. (4th) 1, curing a discriminatory act.
9 Subject to the comments in infra note 11.
‘0 It must be remembered that a decision to strike down a legislative provision or to read a provision
into legislation does not actually do what it appears to do, for a court has no power to enact or to
amend legislation. Provisions that are struck down remain in the statute book unless and until they are
removed by the legislature; provisions that are read into legislation do not appear in the statute book
unless and until they are enacted by the legislature. Constitutional declarations are just that: declara-
tions as to the treatment that is to be given to legislation that is brought before the courts, so as to en-
sure that the content of that legislation, as enforced by the courts, is in accordance with the terms of
the constitution. It follows that whichever course a court adopts, be it to strike down or to read in, the
legislature will have to legislate if it wants its statute book to reflect the reading that the courts have
decided to impose upon it, and that is something that the legislature is free to do or not to do. Ac-

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

[Vol. 44

no possibility of different ways of acting constitutionally, the choice among which
should have been left to the Alberta legislature. It follows that the Supreme Court was
bound by its own jurisprudence to read in the missing provision; the Alberta legisla-
ture was then free to rephrase the Court’s requirements, or to repeal the Individual’s
Rights Protection Act altogether, subject to future constitutional challenge or an invo-
cation of section 33.”

III. Actions and Omissions

A further difficulty in analyzing and responding to Vriend arises from the extent to
which the question of the Alberta government’s responsibility for the treatment visited
upon Vriend by King’s College has been conflated with the question of whether and
in what ways the exercise of that responsibility was discriminatory. This is not to sug-
gest that there was ever any doubt that the treatment that Vriend experienced
amounted to discrimination on the basis of sexual orientation. What was open to
doubt-and what can be usefully clarified-is how and in what ways the Alberta gov-
ernment’s behaviour exposed it to Charter review and ultimately to a finding of dis-
crimination.

The essence of Vriend’s claim was that he was discriminated against not only by
King’s College, but by the Alberta government as well. If that claim was a valid one it
was because the Alberta government was responsible to Vriend for the manner in
which he was treated by his employer and, furthermore, discriminated against him in
the discharge of that responsibility. It will be clear, then, that the question of the scope
of the Alberta government’s responsibility and the question of its discrimination are
quite separate issues, issues that are unfortunately sometimes run together in the Su-
preme Court’s reasons for judgment despite the Court’s evident desire to distinguish
them. Yet it is vital to keep these issues separate for, as has already been indicated

cordingly, the legislature remains free to amend the requirements laid down by the court and, if it so
wishes, to invoke s. 33 to insulate itself against those requirements.

” It is true that to show that the device of reading in is a legitimate remedy is not to show that it is
the proper remedy, or that it is to be preferred to striking down the offending provision. According to
the Supreme Court, the Alberta govemment was constitutionally entitled to repeal the entire Individ-
ual’s Rights Protection Act if it so wished, rather than add “sexual orientation” to the Act’s list of pro-
hibited grounds of discrimination. The Court was confident, however, that the Alberta legislature
would rather add the disputed ground than repeal the entire Act, and so held that reading in was the
proper remedy in the circumstances. In fact, I find it difficult to see how repealing the entire Act
would have been constitutional, for two reasons, which I will sketch here and expand upon in the bal-
ance of this comment. First, if the Court is correct in holding that underinclusiveness is constitution-
ally reviewable, then I find it difficult to see that permitting one form of discrimination is constitution-
ally unacceptable but permitting all forms of discrimination is constitutionally acceptable. After all,
both permissions have an adverse impact on disadvantaged groups protected by the Charter, and
surely the greater impact must be the more unjustifiable. Second, in my view it is not the legislature’s
act-in the form of the Individual’s Rights Protection Act-but its failure to act that is discriminatory
here, so that striking down the non-discriminatory Act as a remedy for the legislature’s discriminatory
failure to act would be incoherent.

1999]

1T MACKLEM – VRIEND V. ALBERTA

205

above, only the first of them is open to real debate. The best way to untangle them is
to be clear about the exact nature of the discrimination that Vriend suffered at the
hands of the Alberta government.

Vriend’s challenge was brought against several sections of the Individual’s Rights
Protection Act, on the ground that those sections were discriminatory and so breached
section 15 of the Charter because of their failure to list sexual orientation as a prohib-
ited ground of discrimination. Exactly how were those provisions discriminatory? A
number of possibilities present themselves, depending on whether the Alberta gov-
ernment’s role in relation to Vriend’s employment amounted to a government action
or a government omission and, further, on whether the discrimination alleged to flow
from that action or omission was direct or indirect.2

Assume first what is in fact the more implausible possibility, that the Alberta leg-
islature acted discriminatorily by enacting the Individual’s Rights Protection Act in
the form that it did, a form that although protecting Vriend and other homosexuals
from a number of different types of discrimination, offered them no protection against
the type of discrimination likely to concern them most, namely, discrimination on the
basis of their sexual orientation.’3 In other words, the question here is whether the Al-
berta legislature acted discriminatorily in prohibiting discrimination on the grounds of
race, religious belief, colour, gender, physical disability, mental disability, age, ances-
try and place of origin.

Clearly, in so formulating the Individual’s Rights Protection Act, the legislature did
not act so as to discriminate directly against gays and lesbians for the confinement of the
law’s protection to nine listed grounds, and the consequent failure to refer either to sex-
ual orientation or to any other potential ground of discrimination, did not involve the
drawing of an express distinction on the basis of sexual orientation. In short, there was
nothing directly discriminatory in what the Alberta legislature did. To offer one body of
people protection against discrimination is not to discriminate against any other body of
people, unless the protected body of people is described in discriminatory terms. In this
respect, the Individual’s Rights Protection Act is to be contrasted to the provision of the
Ontario Human Rights Code, 1981″1 that was successfully challenged in Re Blainey v.

22 By direct discrimination I mean intentional discrimination, or what is sometimes called a viola-
tion of formal equality; by indirect discrimination I mean adverse effect discrimination, or what is
sometimes called a violation of substantive equality. See R. Wintemute, Sexual Orientation and Hu-
man Rights (Oxford: Clarendon Press, 1995) at 10 for a fuller description of the different terminology
used in Canada, the United States and the United Kingdom to describe the two basic types of dis-
crimination.

” I am assuming here what is open to doubt, that all homosexuals see themselves primarily in terms
of their sexual orientation, and that all suffer discrimination primarily on that basis rather than some
other. It seems more likely in fact that many homosexuals take their race, or their religion, or their dis-
ability to be a more important element of their identity than their sexual orientation, and similarly that
many homosexuals suffer more for their race, religion or disability than for their sexual orientation.
However, nothing in what follows turns on the truth of what I have assumed.

24S.O. 1981, c. 53. s. 19(2) [hereinafter HRC].

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

[Vol. 44

Ontario Hockey Association,” which expressly excluded sporting activities from the
HRC’s protection against sex discrimination, thereby invoking a distinction that was a
prohibited ground of discrimination in the Charter, and thus describing its protection
against sex discrimination in sexually discriminatory terms.”

Nor was it claimed, and understandably so, that the Individual’s Rights Protection
Act indirectly discriminated against gays and lesbians in prohibiting certain listed
grounds of discrimination that did not include sexual orientation, for discrimination of
that kind could only be established if the groups protected by the prohibitions con-
tained in the Act could be shown to be disproportionately heterosexual. It seems clear
that no such proof would have been possible, given that the Act protected both dimen-
sions of the binary distinctions that it prohibited, and so protected men as well as
women, whites as well as blacks, the able-bodied as well as the disabled; in short, it
protected people of both genders, all races, all colours, all ages, all abilities and all
faiths. Even if one were to take the view, quite reasonably, that the benefit of the Act’s
prohibitions was intended to be and was in fact enjoyed by women, racial and relig-
ious minorities, the young, the old and the disabled, rather than their privileged coun-
terparts, it seems implausible that those beneficiaries of the Act could ever have been
shown to be disproportionately heterosexual.

The other, more plausible possibility then is that it was the Alberta legislature’s.
failure to act, or omission, the failure to include sexual orientation as a prohibited
ground of discrimination in the Act, the omission of sexual orientation from the list of
prohibited grounds, that constituted unlawful discrimination.” Clearly, this failure or
omission could not be said to constitute direct discrimination, for it did not depend,
explicitly or implicitly, upon a distinction on the basis of sexual orientation. This is
not, as it might seem to be, simply a matter of form. In confining the grounds of dis-
crimination protected by the Act to the nine grounds set out above, the legislature
failed to prohibit a large number of grounds of discrimination, some of which are
prohibited by the human rights codes of other provinces, some of which are prohib-
ited by section 15 of the Charter,” and many of which have yet to find recognition in

,i(1986), 54 O.R. (2d) 513,26 D.L.R. (4th) 728 (C.A.) [hereinafter Blainey].
16 Blainey was a case of an indirectly discriminatory act. It has been argued by Dianne Pothier in
“The Sounds of Silence: Charter Applications when the Legislature Declines to Speak” (1996) 7
Constitutional Forum 113 at 114, 116 that the difference between Blainey and Vriend is one of form
only, and that form should not be allowed to prevail over substance. I agree that the difference is one
of form but it seems to me that form matters here, as Pothier’s own argument confirms. Pothier takes
the view that the government is constitutionally responsible for deliberate omissions from a benefit
that it has decided to confer, and the form of a law may make clear, as in Blainey, that an omission is
deliberate.

‘” On this view it is not, strictly speaking, the legislation that is constitutionally objectionable, but

the absence of legislation.

” I have in mind here the analogous grounds of discrimination of which sexual orientation is but
one; the grounds fisted in s. 15 are reproduced in the Individual’s Rights Protection Act, albeit not in
precisely the same terms.

1999]

T MACKLEM – VRIEND V. ALBERTA

207

any law, or indeed in the experience of any others than those who have been made to
suffer on account of them. It is not as if, as sometimes seems to have been assumed,”
sexual orientation was the only ground of discrimination not included in the Individ-
ual’s Rights Protection Act.

However, and just as clearly, the failure to include sexual orientation as a prohib-
ited ground of discrimination in the Act did indeed constitute indirect discrimination
on the basis of sexual orientation, for the impact of that failure was bome almost ex-
clusively by homosexuals. No proof is needed to establish that it is homosexuals and
not heterosexuals who bear the burden of discrimination on the basis of sexual orien-
tation in our society, as indeed in every other society. Nor can there be any serious ar-
gument that the failure to prohibit sexual orientation discrimination did not constitute
a disadvantage to those who suffered from that discrimination, so as to amount to a
breach of section 15 of the Charter, or that the failure could be justified as reasonable
within the terms of section 1. As has already been indicated above, if it is true that a
legislature can be held constitutionally liable for a failure to legislate, then there was
no possible answer to Vriend’s claim of discrimination. Liability for omissions, or
perhaps more accurately for a failure to act, was therefore the real issue in his case.

The point here is that the issue in Vriend was not what the legislature did but what
it did not do. The significance of that issue is this. Given that the combination of what
the legislature does and does not do embraces all imaginable human endeavour, to
hold the legislature constitutionally liable for all that it has not done would be to ren-
der the legislature constitutionally liable for all acts and omissions within its jurisdic-
tion or, to use the language of the Charter, all matters within its authority. This is, to
put it mildly, a somewhat surprising claim as to the scope of the Constitution, one that
has been denied by the courts and by, most academic commentators. That does not
mean, of course, that it is incorrect. In order to assess its correctness, however, it is
necessary to look at least briefly at what constitutions are and should be about, so as
to determine whether they properly embrace all legislative omissions or some only,
and if they properly embrace some only, to determine which ones they embrace and
which they do not.

IV. Constitutions as Constraints on Power

Constitutions describe the limits of political authority. Those limits protect the
authority of other governments and of individuals. A federal constitution, for exam-
ple, assigns different realms of authority to different governments. Each assignment
determines what each level of government can and cannot do. If, as is the case in
Canada, authority over international and inter-provincial trade is assigned to the fed-
eral government, and authority over trade within a province is correspondingly as-
signed to provincial governments, then each level of government is entitled to legis-

‘9 See Pothier, supra note 16 at 119: ‘”here the legislature enters the fray, but holds back in one

particular aspect?”

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

[Vol. 44

late within its own realm of authority-if it so decides and in whatever manner it de-
cides-and is correspondingly forbidden to legislate within the realm of authority as-
signed to the other level of government. The assignment of a subject matter to one
government limits the authority of that government and correspondingly protects the
authority of another.

In the case of the relationship between individuals and the State, the limitation of
authority is more explicit. Since constitutions assign plenary authority to the govern-
ments that they describe (a combination of governments in a federal State), they set no
limits to that authority and, correspondingly, do nothing to secure the authority of in-
dividuals over the conduct of their own lives, unless and to the extent that those con-
stitutions also contain a guarantee of individual rights and freedoms. A constitution
that contains such a guarantee limits the powers otherwise assigned to any level of
government and so protects the authority of the individual over the design and pursuit
of his or her own plan of life.

This is an extremely brief sketch of the nature of a constitution, one which in no
sense pretends to be a complete account of all that constitutions do. It does, however,
reveal an important aspect of what constitutions do not do-at least as they are pres-
ently understood-and indeed should not be expected to do, for reasons that will be
developed in a moment. What constitutions do not do is require a legislature to enact
any laws, let alone laws of a particular kind. On the contrary, they assume a govern-
ment’s desire to exercise power and take for granted its will to legislate. Another way
of putting this is to say that constitutions confer powers coupled with negative duties.
Certain exceptions aside, constitutions do not require governments to govern, and
that, it must be said, is a good thing, not because government is undesirable, but be-
cause democracy depends upon the capacity of democratically elected governments to
decide for themselves whether and when to act. It is up to governments to decide how
far to govern, and if they decide wrongly, as they often do, the people’s remedy is at
the ballot box, not the bar. Constitutions may diminish the capacity to govern by
transferring some portion of it to other governments or to the people themselves, but
they should not diminish the capacity to govern by transferring it to the courts. In
other words, and as will be argued more fully below, it is one thing for the courts to
police the relationship between two levels of government, or between governments
and the people, and quite another thing for the courts to assume the task of govern-
ment themselves.

So when a constitution assigns the authority to make criminal law to the federal
government, as does the Constitution Act, 1867, it does not tell that government what
conduct to make criminal. Nor does the Constitution tell the federal government how
to prosecute the crimes that it chooses to create, other than to set limits to prosecuto-
rial behaviour through its guarantee of certain rights and freedoms to individuals, such
as the right to a fair trial. The federal government may fail to criminalize conduct that

” (U.K.), 30 & 31 Vict., c. 3, s. 91(27), reprinted in R.S.C. 1985, App. II, No. 5.

1999]

T MACKLEM – VRIEND V. ALBERTA

209

should be made criminal; it may criminalize conduct that should not be made crimi-
nal; it may establish methods of prosecution that are morally unacceptable without
violating an accused’s right to a fair trial (although that might be difficult to do given
the breadth of the constitutional right to a fair trial). If the government does any of
these things it will face moral censure from the press and from the public. It may even
jeopardize its chances of re-election. However, it will not thereby face censure from
the courts, for in assigning authority over criminal law to the federal government the
Constitution assigns to that government complete responsibility for the proper, i.e.,
the moral, development of that branch of the law, and leaves to the Canadian people
the task of holding the federal government to account for what it does and does not do
in the discharge of its responsibility.

The Charter is no exception to this constitutional pattern, as the language of sec-
tion 32 suggests and the Supreme Court’s decision in Dolphin Delivery confirms.
Freedom of expression does not oblige people to speak or to write, does not require
that newspapers be published or that movies be made; nor does it require the govern-
ment to foster or to regulate any of these activities. Similarly, the right to privacy does
not oblige people to be private or require the government to foster privacy under the
scrutiny of the courts. On the contrary, freedom of expression sets limits to whatever
regulatory ambitions a government may have with respect to whatever expressive ac-
tivity its population may choose to engage in, just as the right to privacy sets limits to
whatever desire a government may have to intrude in whatever privacy its population
may seek.

To say that this is the way things are is not, of course, to say that this is the way
they should be. Could matters be otherwise then? Should the Constitution tell the
government how to govern? After all, as is well known, there are exceptions to the
general rule that constitutions give rise to negative rather than positive duties. For in-
stance, the Canadian Constitution requires the federal government to maintain certain
transportation links to British Columbia2′ and to Prince Edward Island,’ and it re-
quires provincial governments to provide certain educational services to members of
minority religions’ and to minority language groups.’4 Would it be desirable for these
exceptions to be made the norm? This would require a revolutionary seizure of power
by the courts, of course, but it might be argued that such a seizure would simply be a
somewhat more dramatic version of what took place in the United States as the result
of the Supreme Court’s decision in Marbury v. Madison.’

Even were such a shift in power possible, however, the answer to the question of
whether it would be desirable must be no. One way of approaching that answer is to
ask what the government would look like if the provisions of the Constitution were to

2′ British Columbia Terms of Union, R.S.C. 1985, App. II, No. 10.

Constitutional Amendment, 1993 (Prince Edward Island), C. Gaz. 1993.11.2022.
Constitution Act, 1867, s. 93.

24 Charter, s. 23.

5 U.S. (1 Cranch) 137 (1803), online: WL (SCr-OLD).

210

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

[Vol. 44

be treated as positive duties. It would then be for the courts to establish, for example,
the government’s role in relation to health care, which would mean deciding whether
health care should be public or private or something in between, deciding the level of
funding that health care should receive from the government, and deciding how that
funding should be distributed. It would be for the courts to decide what kind of armed
forces Canada should have, what their numbers and equipment should be, and what
their policies and goals should be. It would be for the courts to set the direction for the
economy, to establish the curriculum for the schools, to determine environmental
policy-in short, to govern. Clearly this would be undesirable, for it would have the
effect of transferring virtually all democratic authority from the people’s present rep-
resentatives to the authors of the Constitution and to the courts whose duty it is to in-
terpret and enforce that Constitution.

In offering these examples from the division of powers, I do not mean to mock the
proposal that the Constitution could and should be read in obligatory terms. The ex-
amples suggest the price of such an obligation in terms of democracy but they do not
suggest that the price is never worth paying. While the examples make clear that the
creation of such an obligation would be objectionable if framed in general terms, it
remains entirely possible that the obligation might yet be attractive if it were confined
to a limited number of subjects, such as the Charter perhaps, or certain sections of the
Charter, particularly when one remembers that a number of aspects of the division of
powers and of the Charter are already read in obligatory terms without giving rise to
any concerns about the state of democracy in Canada. Indeed, the very existence of
the Charter could be said to constitute a limitation on democracy and yet its presence
in our Constitution has been accepted as compatible with both the practice and theory
of democracy. Would it be desirable, then, if either the Charter as a whole or its guar-
antee of equality rights were to be read in obligatory terms, so as to impose upon the
governments of Canada an obligation to secure the liberty and equality of all Canadi-
ans? In that case, the governments of Canada would have a positive duty to secure the
rights and freedoms of Canadians, including the right to equality, a duty that would be
interpreted and enforced by the courts.

Once again, however, the answer to that question mtist be no. Were section 15 of
the Charter, for example, to be read in obligatory terms, it would be for the courts to
determine the direction and the detail of human rights policy in Canada. It would be
for the courts to establish the content of human rights codes, which would mean de-
termining what grounds of discrimination are prohibited by the Charter and which of
them must be prohibited in human rights codes in order to ensure that both public and
private activity are governed by the principle of equality; deciding which realms of
conduct, such as the provision of employment, services and accommodation, are to be
addressed by human rights codes; deciding what justifications for discrimination
should be accepted, thereby anticipating what would otherwise be decided in a section
1 analysis; deciding how human rights codes should be enforced and by what kinds of
bodies; and so on. In other words, it would not simply be a matter of ensuring that the
grounds of discrimination prohibited by section 15 were mirrored in the human rights
codes of all Canadian jurisdictions, although that would clearly be required; the courts
would have to ensure that the legislatures implemented whatever social and economic

1999]

T MACKLEM – VRIEND V. ALBERTA

policies were necessary to ensure the equality of all who are protected by the grounds
of discrimination set out in the Charter.

This would be undesirable, not simply because it would take the courts well out-
side their recognized areas of expertise-for it is possible that the courts could be per-
suaded to acquire new areas of expertise (although it is a mystery why a society
should want to transfer power from representative bodies to the courts only to ask the
courts to function as if they were representative bodies)-and not simply because it
would be undemocratic-for the existence of positive duties of limited scope is
clearly compatible with democracy. Rather, it is undesirable because the issues of
equality and non-discrimination are issues of such vital importance to a society of
immigrants such as Canada that the moral responsibility for the manner in which we
as a country come to terms with them must for the most part remain with individual
Canadians and with their elected representatives. In other words, and as this comment
has already indicated, it is one thing to limit the scope of democratic authority by in-
sisting that the government not discriminate against Canadians unless it has powerful
reasons for doing so-reasons that are capable of satisfying the courts-and quite an-
other thing to eliminate democratic authority by insisting that the government have no
voice in the enterprise of securing the equality of all Canadians, an enterprise that
would then become the exclusive property of the courts.” It is one thing to transfer
power to the people, subject to the scrutiny of the courts, and quite another thing to
transfer power to the courts.”

26 In what follows, I will sometimes use the term “equality” as shorthand for the principle of equal-
ity and non-discrimination that is protected by s. 15 of the Charter, where “equality” and “non-
discrimination” are treated as correlatives.

‘ It is commonly believed that the Charter transfers power from Parliament and the legislatures to
the courts. In my view, however, the Charter transfers power from Parliament and the legislatures to
the people. Freedom of expression, for example, transfers the power to govern expression to the peo-
ple, not the courts. The power that is given to the courts to enforce that freedom is a new power, but it
no more enables the courts to govern expression than their power to enforce the division of powers
between the federal and provincial governments enables them to govern generally. Correspondingly,
the power that is taken from Parliament and the legislatures is for the most part transferred to the peo-
ple, acting individually and in combination; otherwise it falls into abeyance. Therefore, when the
courts decide that certain acts of expression are constitutionally protected and that government regu-
lation of those acts is constitutionally prohibited, they transfer the power to govern those acts from the
people’s representatives to the people themselves. This is why bills of rights tend to be popular, to the
dismay of those academics who think that the genuine interests of the people are always to be found
in the expression of the democratic will.

Two further observations might be made here. First, while the moral right to freedom of expres-
sion (or any other freedom) may require government intervention if it is to be fully realized (an inter-
vention that recognizes in the right a dimension that is sometimes known as positive liberty), that in-
tervention is not and should not be part of what is guaranteed by the Constitution for the reasons given
in the text. It follows that in this setting, the content of legal rights does not and should not correspond
to the content of moral rights. Second, the reason why bills of rights normally do not contain social
and economic rights is that it is in the nature of those rights to be understood as positive duties, the en-

212

MCGILL LAW JOURNAL /REVUE DE DROITDE MCGILL

[Vol. 44

The suggestion here, then, is that the ultimate responsibility for securing equality
must rest with the Canadian people and with their political representatives. Let me go
beyond that bald assertion by looking first at the reasons in support of preserving the
responsibility of Parliament and the legislatures for securing equality. Canadian gov-
ernments cannot govern successfully unless and until they develop a proper apprecia-
tion of the distinctive character of the particular community for whose well-being they
are responsible and to whose judgment they are accountable. This is not simply a
matter of a government’s ability to appreciate the right way in which to design human
rights legislation or employment equity programs in its particular jurisdiction. Rather,
it is a matter of a government’s ability to appreciate the right way in which to design
the entire legislative program for a community the members of which have diverse
backgrounds and commitments, a community that the government must understand
properly in order to govern justly. In other words, a government that does not under-
stand equality and what it requires cannot govern well, at least not in a pluralistic so-
ciety such as Canada. Yet a government’s capacity to understand equality would be
seriously diminished if the mandate for equality were to be transferred from Parlia-
ment and the legislatures to the courts.

This is not, however, the only reason for preserving the responsibility of Parliament
and the legislatures for securing equality. As far as individual Canadians are con-
cemed, the formation of full relationships among members of different cultural, social
and sexual communities once again requires a level of understanding and appreciation
that is diminished when those relationships are specified in law. We would all be the
poorer if every aspect of the relationship between homosexual and heterosexual were
to be laid down in human rights codes and other legislation. It -is of course essential
that we treat one another with proper understanding and respect, but the fact is that we
can only do so if we retain a significant measure of personal responsibility for our re-
lationships with one another. This is but one reason why liberal societies like Can-
ada’s take individual freedom seriously, indeed constitutionally guarantee it.

Of course, to convert a legislature’s power to secure equality (coupled with a duty
not to violate equality) into a positive duty to secure equality is not necessarily to di-
minish the realm of private judgment. It is possible thaf the courts would be capable
of setting out a positive duty that left as much responsibility to individuals as indi-
viduals now enjoy, perhaps more. This, though, is unlikely. The reason for thinking
this is that there would be absolutely no point in converting a power into a positive
duty if the consequence were to entrench the very sense of obligation that Canadian
legislatures presently recognize, or something less demanding. That being the case,
entrenchment of a positive duty to secure equality would in practice be bound to di-
minish the realm of private judgment in the construction of personal relationships
among Canadians.

forcement of which would require-as the enforcement of civil and political rights does not–the
transfer of governmental authority from the legislatures to the courts.

1999]

T MACKLEM – VRIEND V. ALBERTA

How essential is it, though, to secure a realm of private judgment in these matters?
It may seem to some that I am seeking here to vindicate the importance of the distinc-
tion between the private and the public realms, a distinction that has been regularly
disparaged by feminists among other social critics. If so, I would contend that this
does not in fact put me at odds with feminists and other champions of equality, or at
least with the substance of what they have argued. What feminists have done is to
point out that the distinction between private and public in any particular culture is a
moral distinction as well as a social fact, and to argue that many activities that are, at
least in cultures that they are concerned with, private as a matter of social fact should,
for sound moral reasons, be a matter of public concern. I do not disagree with any of
this. What feminists have not done is to argue that there should no longer be a private
realm, for they too are committed to personal freedom, provided that its exercise does
not promote gender inequality or otherwise cause harm to others. Feminists simply
want to ensure that, where women are concerned, free also means equal, as it has not
done in the past. What is more, feminists have not argued that the distinction between
private and public, whatever form it might take, should be a matter for the courts and
the courts alone to determine. Some aspects of that distinction are for the courts to
determine but the balance must remain with the legislature, or a feminist would have
no reason to seek public office.

Third and last, it seems that the entrenchment of an obligation to secure equality
would be almost certain to diminish the sensitivity of the law’s response to the re-
quirements of local circumstances and conditions, for the courts have no capacity and
no obligation to represent local people and local circumstances. On the contrary, one
of the functions of the courts is to be blind to such questions. By contrast, as this
comment has already indicated, it is a fundamental part of the responsibility of any
legislature to understand its constituency and to design a set of social obligations that
is appropriate to that constituency.S

Two things follow from all this. First, section 15 of the Charter does not and
should not impose upon Parliament and the legislatures of Canada a duty to ensure the
equality of all those within their respective jurisdictions. For that reason, Canadian
governments cannot be held constitutionally liable merely for their failure to secure
equality, either generally or in some particular respect. If they do not do what they
should do for disadvantaged Canadians, they will to that extent be bad governments,
but the responsibility for that fact will lie with them and the accountability for that fact
will rest in the hands of their electors. The courts and the Constitution are not now-
perhaps could not be, and certainly should not be-the only vehicles for the
vindication of the moral rights of the Canadian people, in this case the moral right to
equality. Nor should they be the only authority for the limitation of liberty that the
vindication of those rights may require of other Canadians.

‘ This does not mean that I endorse the reasons of McClung J.A. in Vriend v. Alberta (1996), 184

A.R. 351,141 D.L.R. (4th) 44 (C.A.).

MCGILL LAW JOURNAL / REVUE DE DROITDE MCGILL

[Vol. 44

Second, however, it may well be the case that a government can be held constitu-
tionally liable for a failure to act in certain circumstances if, in all those circum-
stances, its failure to act can be said to amount to an act. Context can and often does
reveal that what is from one perspective a failure to act is from another perspective an
act. Such acts may attract constitutional liability in the same way as any other gov-
ernment act. In other words, while there cannot be liability for all omissions, even in
the realm of equality, there may well be liability for certain omissions that can be
shown in context not to be omissions but acts. In Vriend, then, it may well be that the
failure of the Alberta government to legislate against discrimination on the basis of
sexual orientation amounted to an act of discrimination against gays and lesbians in
that province. While the omission of one or more grounds of discrimination from hu-
man rights legislation does not, in and of itself, constitute an infringement of section
15, the omission of a ground as prominent as sexual orientation might do so if other
conditions were sufficient to make that omission an act. This possibility will be ex-
plored in Part VII, below.

What is critical in a case like Vriend, then, and what must be determined in what
follows, is what it is that makes a failure to act into an act, what conditions must be
met in order to establish constitutional liability for a legislative omission. Reasons
must be found for departing from the general rule and for holding a legislature to be
liable for its failure to act, in this case, reasons to hold that homosexuals in Alberta
were actively discriminated against when their government refused to extend to them
the benefit of its protection against discrimination. It may be that the Supreme Court
has succeeded in identifying those reasons in Vriend, although it should be clear by
now that I do not think so.

V. The Supreme Court’s Reasons

Virtually all of what has been said above was assumed to be correct by the Su-
preme Court of Canada, although obviously not in quite the same terms. The Court
assumed that not all government omissions are subject to Charter review. It assumed
that Dolphin Delivery had been correctly decided. It suggested, however, that the
terms of the Constitution did not forbid a constitutional challenge to certain govern-
ment omissions, of which it took Vriend to be one. Unfortunately, at no point in its
reasons did the Court explain what it was about Vriend that made the Alberta gov-
ernment’s failure to prohibit discrimination on the basis of sexual orientation consti-
tutionally reviewable. Instead, it by and large contented itself with rebutting the Al-
berta government’s arguments that theirs was not a reviewable omission, as if to rebut
the argument for one side was to make the argument for the other side, something that
would only be true in a zero sum game, which Vriend clearly was not given that either
the Alberta government or Vriend might have been either right or wrong for reasons
other than those they offered to the Court.

A. Liability for Omissions

The Supreme Court, speaking through Cory J., dealt with the question of the appli-
cation of the Charter to a failure to legislate in sixteen short paragraphs, six of which

1999]

T MACKLEM – VRIEND V. ALBERTA

215

were taken up with a digression on judicial deference, a subject that Cory J. rightly
regarded as not being at issue in the case but nevertheless felt it necessary to address
given its treatment in the courts below. Disregarding some further prefatory com-
ments, Cory J. then devoted five paragraphs to the issue of the Charter’s application
to government omissions, two more paragraphs to the issue of the Charter’s applica-
tion to private activity, and one paragraph to the problem of mirroring. Even a Laco-
nian would have had trouble addressing questions of such difficulty in so brief -a com-
pass, and since I am no Laconian, I will address them under this and the next two
headings.

Cory J. began by observing that there is nothing in the language of section 32 of
the Charter (which states that the Charter applies to the legislature and government of
each province in respect of all matters within the authority of the provincial legisla-
ture), that would prevent the courts from treating government omissions as subject to
the Charter.” This must surely be right, for otherwise a provincial government would
not be subject to Charter review for a failure to fulfil its constitutional duties, duties
that are set out in the Charter itself. To take the most obvious example, if section 32
ruled out any constitutional challenge to a provincial government’s failure to act, it
would be impossible to enforce the Charter’s minority language educational rights, or
indeed its voting rights. To read section 32 in that way would make the Charter a self-
defeating document.

Cory J. further recognized, at least implicitly, that not all government omissions are
subject to Charter review. Again this must surely be right, not simply because it
amounts to conventional wisdom in the realm of constitutional law, and not simply
because that conventional wisdom has been confirmed by the Supreme Court of Can-
ada in Dolphin Delivery, but for the moral reasons that have been developed at some
length in Part IV, above.

What Cory J. did not do, however, was offer any argument that might have ex-
plained how and why the particular omission before him differed from other omis-
sions and so was subject to Charter review, an argument that would necessarily have
involved describing the principles that govern liability for omissions in sufficient de-
tail to make their application to Vriend clear. Yet such an argument was essential to
the Court’s decision, for unless the Court was willing to endorse the view that all gov-
ernment omissions are subject to constitutional challenge-a view that it in fact re-
jected-it was bound to show, through adequate reasons for its decision, how and why
the omission before it was subject to challenge, unless it was prepared to act arbitrar-
ily, as it should not have been here.

The subject of reasons and their importance will be revisited later in this comment.
At this point, it is necessary to address Cory J.’s repeated description of the Alberta
legislation as “underinclusive”,3 a description that was invoked at many points in his

29Vriend, supra note 1 at 532-33.
‘ Ibid. at 541,545, 553.

216

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

[Vol. 44

reasons for judgment and that seems to be intended to carry much of the weight of his
argument. The idea of underinclusiveness was in effect used by Cory J. to finesse the
resolution of two separate questions: the question of the Alberta government’s respon-
sibility for its failure to act, and the further and consequent question of whether and in
what ways the Alberta government’s discharge of that responsibility was discrimina-
tory.

Underinclusiveness is an idea borrowed from United States constitutional law, spe-
cifically from that body of doctrine known as “rationality review”, a body of doctrine
that has been developed by the United States Supreme Court in the course of inter-
preting and applying the fourteenth amendment to the United States Constitution, the
equal protection clause. Quite apart from any general concerns one might have about
borrowing from American constitutional law in order to solve problems of Canadian
constitutional law,” two things in particular need to be remembered about the four-
teenth amendment and its relevance to the interpretation of the Charter.

First, the form of equality that the fourteenth amendment secures embraces both a
prohibition against certain forms of discrimination and a prohibition against the use of
unreasonable classifications in the design and implementation of legislation. Ration-
ality review and its attendant idea of underinclusiveness are features of the fourteenth
amendment’s prohibition against the use of unreasonable classifications, a prohibition
that insists upon the presence of a rational connection between a legislature’s goals
and the means chosen to achieve those goals. Second, in Andrews v. Law Society of
British Columbia” the Supreme Court of Canada rejected an earlier line of Canadian
authority that had interpreted the form of equality that is guaranteed by section 15 of
the Charter as a prohibition against the use of unreasonable classifications in the de-
sign and implementation of legislation. The Supreme Court held in Andrews that the
Canadian and American constitutions were different in this respect, and that the
equality guaranteed by section 15 is to be understood exclusively in terms of non-
discrimination.

In the wake of Andrews, one might understandably object to judicial reliance upon
the idea of underinclusiveness on the basis that underinclusiveness is part of rational-
ity review, a form of review that was rejected in Andrews. In fact, however, it would
be a mistake to object to the reasons of Cory J. on that basis, for it seems at least pos-
sible that the idea of underinclusiveness could be usefully employed in some setting
other than rationality review. What makes the reasons of Cory J. objectionable is that
they import the idea of underinclusiveness into the setting of a debate on discrimina-
tion, a setting in which to rely upon underinclusiveness is blatantly to assume one’s
conclusions. Let me explain.

See R. v. Rahey, [1987] 1 S.C.R. 588 at 639, 39 D.L.R. (4th) 481, La Forest J.: “Canadian legal
thought has at many points in the past deferred to that of the British; the Charter will be no sign of
our national maturity if it simply becomes an excuse for adopting another intellectual mentor”

” [1989] 1 S.C.R. 143, 56 D.L.R. (4th) 1 [hereinafter Andrews cited to S.C.R.].

1999]

T MACKLEM – VRIEND V. ALBERTA

217

Underinclusiveness describes a particular kind of failure of fit between a legisla-
ture’s goals and the classification that the legislature has used to achieve those goals.
If the classification used by the legislature fails to embrace all those people who
would be rationally understood as falling within the legislature’s goals, then the classi-
fication is, from a rational point of view, underinclusive. This means that the idea of
underinclusiveness expresses the outcome of a comparison between a classification
and the purpose that the classification is deemed to serve. Laws that are found to be
underinclusive as a result of that comparison are, under the fourteenth amendment,
unequal.

This interpretation of underinclusiveness is not some special feature of American
constitutional law. Rather, it is part of the very concept of underinclusiveness, for a
classification can only be described as underinclusive by measuring it against some
standard. When so measured, the classification may be found to be either underinclu-
sive or overinclusive, depending on the standard that it is deemed to serve. In the set-
ting of rationality review, legislative classifications are deemed to serve the legisla-
ture’s purpose, and so are determined to be underinclusive or overinclusive by com-
paring them to that purpose-a purpose that is not itself subject to constitutional re-
view. Whether or not this is an attractive interpretation of the concept of equality, it at
least makes sense.

From the perspective of rationality review, if it was the purpose of the Alberta leg-
islature to prohibit all forms of discrimination, then the classification that it used to
achieve that purpose was clearly underinclusive and so irrational. From that perspec-
tive, Alberta’s legislation would be unequal because it was underinclusive. In the set-
ting of rationality review, underinclusiveness is the basis for a finding of inequality.

In the wake of Andrews, though, Alberta’s rationality was not the issue before the
Supreme Court in Vriend. The issue before the Court was whether Alberta’s failure to
extend the benefit of its human rights legislation to homosexuals was discriminatory.
From that perspective the legislation could be said to be underinclusive only if it was
discriminatory. In other words, in the setting of an inquiry into discrimination, the de-
scription “underinclusive” expresses the conclusion of a finding of discrimination’ In
discrimination analysis, legislation is underinclusive if-it is unequal, that is, if and
only if it is discriminatory.

On that analysis, if the Alberta government acted discriminatorily in failing to

protect homosexuals-something that it was the Supreme Court’s task to determine-
then the list of grounds of discrimination in its legislation was clearly underinclusive;

” In discrimination analysis, the standard against which legislation is found to be underinclusive is
established not by the legislature’s purpose but by the conclusion that the legislature has discrimi-
nated. To begin one’s reasons for judgment by describing legislation as underinclusive, as Cory J. did,
is to import a standard into one’s reasons without acknowledging that one is doing so, the very stan-
dard that it is one’s task to determine, the standard by which, in Vriend, sexual orientation should have
been included in the Individual’s Rights Protection Act and against which the Alberta government
acted discriminatorily.

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

[Vol. 44

if it did not act discriminatorily, then that list of grounds was not underinclusive. It
follows that for Cory J. to have begun his reasons for judgment by describing the
grounds of discrimination set out in the Individual’s Rights Protection Act as underin-
clusive was for him to assume the conclusions to the very questions that he had been
asked to decide, namely, whether the Alberta government had behaved discriminato-
rily and, more fundamentally and more troublingly, whether the Charter rendered the
Alberta government liable for its failure to act. Those assumptions may have been
correct, but it would take an argument to know that, an argument that Cory J. did not
offer and that the use of the label “underinclusive” blatantly short-circuits.

To complain about the Court’s reasons in this way is not simply to engage in an
academic cavil, to exhibit an unwarranted attachment to intellectual tidiness, as some
might think. In certain settings it may well be that weak or absent or untidy reasons do
not much matter, but they matter greatly here, as will be explained below. First, how-
ever, it is important to take a brief look at the remaining considerations to which Cory
J. referred in the course of his judgment.

B. The Application of the Charter to Private Bodies
In the second part of his discussion of the application of the Charter, Cory J. ac-
knowledged, quite rightly, that the Charter does not apply to the private actions of
private actors. Indeed, that much at least seems to have been accepted by all parties.
Constitutions bind governments, not citizens, as section 32 of the Charter makes clear
and Dolphin Delivery has confirmed. Yet if a government’s failure to bind the private
actions of private actors to Charter values is ipsofacto to be treated as a breach of the
Charter, then the Charter effectively applies to those private actions, albeit in two
stages, for it compels the government to compel private actors to abide by Charter
values in their private conduct. On that reading of section 32, a government’s failure
to end discrimination, or to end any other infringement of a Charter value by a private
actor, would in and of itself amount to a breach of the Charter. It follows that to apply
the Charter to a government’s failure to act (in the absence of a constitutional duty to
act) is indeed to extend the Charter to the private actions of private actors, and thus to
silently overrule Dolphin Delivery.

Unfortunately, once again this is an issue that Cory J. failed to address adequately,
contenting himself instead with the truism that it would be unacceptable if any legis-
lation that regulated private activity were for that reason immune from Charter scru-
tiny. This is a truism because it is surely indisputable. Were it not the case, then the
governments of Canada would be virtually immune from Charter scrutiny, since vir-
tually all legislation regulates private activity. If Alberta’s case had truly depended on
the contention that the Individual’s Rights Protection Act should be immune from
constitutional scrutiny because it regulated private activity, then Vriend’s appeal
should have been allowed from the bench. However, there is no evidence that the Al-
berta government ever pretended anything so foolish, so that Cory J. has here an-
swered an argument that was not made, and in doing so has failed to confront the very
real implications of his decision.

1999]

T MACKLEM – VRIEND V. ALBERTA

219

One of the sources of the error here-and indeed one of the reasons why Cory J.
earlier allowed himself to call the Individual’s Rights Protection Act underinclu-
sive4
is the all too common mistake of equating Charter values and Charter rights.”
Freedom of speech is a Charter value, as is liberty, as is equality. Freedom of speech
is only a Charter right, however, to the extent that section 32 renders some govern-
ment liable for its infringement. In other words, the difference between a Charter
value and a Charter right is the difference between a moral right and a legal right, the
difference that in Canadian law is defined by the scope of section 32. Charter values
bind all Canadians, but are not enforceable in law. Charter rights, on the other hand,
bind only the actions of Canadian governments, but are fully enforceable in law. It
follows that to say that something is a Charter value is not to say that it is a Charter
right, at least in the absence of additional argument as to the scope of section 32.
Furthermore, and what is the point here, to treat a Charter value as if it were a Char-
terright is, once again, to assume that section 32 embraces all government omissions
as well as all government acts, and so to assume that the Charter embraces all private
activity or, at a minimum, to assume that section 32 embraces the government omis-
sion that is at issue in the case at hand, and so to assume that the Charter embraces
the private activity that is affected by that omission.

C. Mirroring

The Court’s final, very brief consideration of the application of the Charter to gov-
ernment omissions took place in the context of its section 15 analysis’ Counsel for

” See Vriend, supra note 1 at 533: “The [Individual’s Rights Protection Act] is being challenged as
unconstitutional because of its failure to protect Charter rights, that is to say its underinclusiveness”
If what the Act had failed to protect was a Charter right, Vriend could have invoked the Charter di-
rectly. In fact what the Act failed to protect was a Charter value, so that Vriend was forced to contend
that the failure to protect a Charter value infringed a Charter right.

” Cory J. may have been led to believe that his decision in Vriend would not extend the Charter’s
application to private activity by a belief that Alberta’s failure to legislate, as reflected in the terms of
the Individual’s Rights Protection Act, affected Charter rights rather than Charter values. Such a be-
lief could have led him to think that the Individual’s Rights Protection Act differed from other legisla-
tion in affecting Charter rights, so that to apply the Charter to the failure to legislate that was implicit
in its terms would not be to apply the Charter to the failure to legislate that is implicit in the terms of
all legislation, and so would not be to apply the Charter to all that activity that is private simply be-
cause all governments have failed to regulate it, notwithstanding the impact of their failure upon
Charter values.

3 In fact, s. 15 was clearly not the proper place for consideration of this issue; s. 15 only comes into
play once it has been decided that the Charter is applicable. If the Charter is applicable, then the
question of discrimination falls to be determined according to the law laid down in Andrews and sub-
sequent cases. It follows that in Vriend, s. 15 only came into play once the threshold question of
whether the Charter applied to government omissions had been resolved. If the Charter applied to
government omissions, then Alberta’s omission was discriminatory and so contrary to s. 15, assuming
that the omission had an adverse impact upon homosexuals, as it clearly did. The mirroring argument,
however, is not a s. 15 argument about the presence or absence of discrimination, but a s. 32 argument

220

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

[Vol. 44

the Alberta government had argued that if the Charter compels Canadian govern-
ments to compel private actors to abide by Charter values in their private conduct,
then the anti-discrimination provisions of human rights statutes must necessarily mir-
,ror the requirements of section 15. As has been indicated above, this simply follows
from acceptance of the contention that the Charter applies to government omissions.
Yet one might legitimately object to such a conclusion on the ground that endorse-
ment of it would mean that the Charter applies to the private actions of private actors,
contrary to section 32 and Dolphin Delivery.

Cory J., however, denied that the application of the Charter to government omis-
sions meant that human rights codes must mirror the requirements of section 15. In
his view, whether the omission of a ground of discrimination from a human rights
code infringed the Charter would depend on the nature of the exclusion, the type of
legislation, and the context in which it was enacted.” But how could this be so? In the
course of his reasons in Vriend, Cory J. did not rely on any of these factors to find that
Alberta’s failure to mirror the requirements of section 15 was discriminatory, with one
exception. He suggested that the comprehensive nature of the Individual’s Rights
Protection Act made its omission of sexual orientation as a ground of discrimination
discriminatory. It would be different, Cory J. suggested, if the legislature had chosen
to deal with only one type of discrimination: “In such a case it might be permissible to
target only that specific type of discrimination and not another.”

It seems clear that Cory J. has run together here the resolution of two separate
questions, the question of the application of the Charter and the question of discrimi-
nation. He has tried to suggest that there is no need to be concerned at the prospect of
a government being held liable for all omissions from its human rights legislation, and
so no need to resist the application of the Charter to those omissions, for an omission
will only give rise to liability for discrimination if the human rights legislation that it
affects is comprehensive in its scope. An argument about the meaning of discrimina-
tion is thus used to address a concern about the scope of Charter liability. If the two
questions are separated, as they must be, then there are two possible ways to
read Cory J.’s judgment. On one reading, Cory J. intended to say that the
comprehensiveness of a human rights statute may render a government constitution-
ally liable for any omissions from that statute. It would follow that there is no need to
be concerned about the problem of mirroring, for a government omission will only

in support of the position that governments should not be held liable for their omissions, constitutional
duties aside. Consideration of the mirroring argument implied that the omissions question had not
been resolved, in which case s. 15 could not have come into play.

” Vriend, supra note 1 at 552-53. Cory J. was less than clear as to whether a failure to mirror the
Charter would not necessarily be discriminatory and so contrary to s. 15, or whether it would be dis-
criminatory but not necessarily unreasonable according to the terms of s. 1. The better view would
appear to be that Cory J. believed that a failure to mirror the Charter’s requirements in legislation
would not necessarily be discriminatory, since if discriminatory, it is hard to imagine how it might be
justified as reasonable.

” Ibid. at 548 [emphasis added].

1999]

T MACKLEM – VRIEND V. ALBERTA

give rise to constitutional liability if the legislation that it affects is comprehensive in
its scope.

Part VII of this commentwill show that there is some merit in this type of argu-
ment. However, it cannot be accepted as a good reading of what Cory J. intended to
say, partly because it seems unlikely that he would have made what amounts to a sec-
tion 32 argument in the context of an analysis of section 15, and partly because if he
did in fact intend to make a section 32 argument in that setting, he did not begin to
explain why the comprehensiveness of a human rights code should make omissions
from that code constitutionally reviewable, as he surely would have been required to
do.

The better reading of Cory J.’s judgment is that he intended to say that the
comprehensiveness of a human rights statute may make omissions from that statute
discriminatory. It would follow that there is still no need to be concerned about the
problem of mirroring, for while governments are liable under section 32 for all their
omissions, those omissions will only give rise to a finding of discrimination if the
legislation that they affect is comprehensive in its scope. In other words, it is only if a
human rights statute is comprehensive in its scope-so as to omit one ground of dis-
crimination while covering all others-that the statute is likely to be found to be dis-
criminatory. If, on the contrary, a human rights statute is limited in its focus so as to
omit many grounds of discrimination while covering only one, then the statute is un-
likely to be found discriminatory. To the extent that human rights legislation is limited
rather than comprehensive in scope, then it will not have to mirror the terms of the
Charter.

While this may be what Cory J. intended to say, it is not only an unsatisfactory
analysis of discrimination, but offers small comfort to those who are concerned over
the potential implications of constitutional liability for government omissions. As an
analysis of discrimination, the position taken by Cory J. is surely implausible. Would
it not be both natural and in line with the case law on section 15 to think that a human
rights code that prohibited only one ground of discrimination while omitting ten oth-
ers was more disadvantageous to vulnerable groups in Canadian society and so more
discriminatory than a code that prohibited ten grounds of discrimination while omit-
ting one? Would it not be natural to think that the absence of any human rights code
would be even more disadvantageous to such groups and so even more discrimina-
tory? Cory J. suggests otherwise. Yet if a government can be held liable under section
32 of the Charter for any and all of its omissions, it is difficult to see how and why
comprehensive omissions on its part should be regarded as less discriminatory than
limited omissions. What could make the omission of one ground of discrimination,
rather than several, discriminatory and so contrary to section 15?”‘ Did Cory J. mean
to suggest that the Individual’s Rights Protection Act would have been less discrimi-
natory if it had omitted more grounds of discrimination, say discrimination on the ba-

39 In fact, the Act omitted more than one ground of discrimination, as I have already indicated.

222

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

[Vol. 44

sis of physical or mental handicap as well as discrimination on the basis of sexual ori-
entation?

Cory J. pointed out, probably correctly, that exclusion from a human rights stat-
ute’s grounds of discrimination not only denies homosexuals the protection of the law,
but may actually send a message to society that discrimination on the basis of sexual
orientation is acceptable.’ Even assuming that this is true, however, it would be just as
true of any failure to mirror the requirements of section 15. It cannot explain why the
Alberta government’s failure to mirror the Charter was discriminatory but a more
comprehensive failure to mirror the Charter might not have been, as Cory J. sug-
gested.”

As a reassurance to those who are concerned over the potential implications of
constitutional liability for government omissions, Cory J.’s reference to the idea of
comprehensiveness offers small comfort. First and most obviously, since the basis for
reassurance here is that while all government omissions are constitutionally review-
able not all are discriminatory, the argument implicitly accepts that any failure to mir-
ror the terms of the Charter is constitutionally reviewable and simply seeks to contain
the implications of that holding by its analysis of discrimination. Even were that
analysis of discrimination correct, then, it would do nothing to relieve governments of
liability for their failure to fulfil Charter values other than equality. Second, since Ca-
nadian governments, unlike governments in some other parts of the world, have con-
sistently chosen to address the problem of discrimination through comprehensive hu-
man rights statutes, if omissions from comprehensive statutes are likely to be held to
be discriminatory, then Canadian human rights statutes will indeed have to mirror the
terms of the Charter.

Perhaps the best explanation for Cory J.’s reliance upon the idea of comprehensive-
ness is that it simply followed from his earlier reference to the idea of underinclusive-
ness. In the course of reaching his decision, Cory J. characterized the Individual’s
Rights Protection Act as “legislation that purports to provide comprehensive protec-
tion from discrimination for all individuals in Alberta.. 2 If that was indeed the pur-
pose of the Act (as in fact it clearly was not), then the list of grounds of discrimination
that the Alberta legislature chose to achieve that purpose was plainly underinclusive.
Yet as was argued above, sensible as the idea of underinclusiveness may be in the
context of rationality review, to rely upon it in the setting of an analysis of discrimina-

40 Vriend, supra note I at 550.
4, To fail to prohibit a certain ground of discrimination is not necessarily to approve, let alone to
authorize, discrimination on that ground. A legislature may legitimately take the view that it is up to
each one of us, in the exercise of our individual moral responsibility, to determine what constitutes
discriminatory behaviour on this ground and how best to prevent it. A failure to prohibit a certain
ground of discrimination only becomes discriminatory when the circumstances show that the creation
of a non-discriminatory environment depends upon government regulation, as in fact was probably
the case in Alberta.

4 Vriend, supra note 1 at 548.

1999]

T MACKLEM – VRIEND V. ALBERTA

tion is to assume the conclusion that it is one’s task to demonstrate. To characterize
legislation as “comprehensive” is simply another way of characterizing it as “underin-
clusive”, and suggests that despite its protestations to the contrary, the Court feels a
continuing attachment to the idea of equality as rationality that it rightly rejected in
Andrews.

VI. The Impact on Vulnerable Minorities

The gist of what was decided in Vriend seems to be this. The Supreme Court has
denied that section 15 of the Charter imposes a duty upon the governments of Canada
to secure equality and end discrimination in their respective jurisdictions, such that
any failure to fulfil that duty would constitute a breach of the Charter.’ The Charter
does not make Canadian governments liable for the discriminatory behaviour of pri-
vate actors. That is a conclusion that one might well have expected, for to have held
otherwise would have been to overrule Dolphin Delivery, something the Court has
shown no inclination to do.

On the other hand, the Supreme Court has insisted that a government’s failure to
secure equality can indeed give rise to constitutional liability, not because that failure
violates a constitutional duty to secure equality among private actors-for the Court
has held that no such duty exists-and apparently not because in light of all the cir-
cumstances a government’s failure to secure equality can sometimes amount to a dis-
criminatory act.” The Charter simply makes Canadian governments liable for their
failure to prohibit discrimination by private actors on any ground of discrimination
prohibited by the Charter, unless that failure is so complete that human rights legisla-
tion cannot be characterized as comprehensive, in which case it might not be dis-
criminatory. That is plainly a conclusion that one could not have expected, for it flies
in the face both of what was decided in Dolphin Delivery and one’s best understand-
ing of what constitutions are and should be about. Expectations aside, it is difficult to
distinguish that conclusion from the conclusion that the Court rejects, namely, that the
Charter does not make Canadian governments liable for the discriminatory behaviour
of private actors. One is left at a loss as to how to reconcile the two positions taken by
the Court, positions that more or less contradict one another. More generally, one is at
a loss to understand what it is that makes a government constitutionally liable for its
failure to act in fulfilment of Charter values.

Does this matter to anyone but legal scholars? If sexual orientation should clearly
have been added to the grounds of discrimination prohibited by the Individual’s
Rights Protection Act, and if the Alberta government was unlikely to do any such
thing in the foreseeable future, should the Supreme Court not be applauded for com-

” Unless, of course, the failure to fulfil the duty could be justified as reasonable within the terms of

s. 1.

” See Vriend, supra note I at 533: “[]t is not necessary to rely on this position in order to find that

the Charter is applicable ‘

224

MCGILL LAW JOURNAL / REVUE DE DROITDE MCGILL

[Vol. 44

ing to the aid of Alberta’s homosexuals, and indeed to the aid of all those who support
the quest for a more inclusive society-a society that genuinely reflects the values set
out in the Charter? Surely the quality of the reasons that the Court happened to offer
for its conclusion was neither here nor there as far as Alberta’s gays and lesbians were
concerned. That being the case, is it not far-fetched to attribute to those reasons any-
thing like the significance that attaches to the decision that the reasons purport to ac-
count for? How much do reasons matter, here or anywhere else?

The answer is that they matter greatly, not because there is some sort of aesthetic
objection to untidy reasoning, but because the reasons in Vriend are likely to injure
the cause not only of homosexuals, but of other disadvantaged groups as well as un-
dermining the status of the Court. Let me try to explain what I mean.

A. Legal Implications

The first, although not necessarily the most important, reason to be concerned
about the decision in Vriend is the potential scope of the constitutional obligation that
it gives rise to. Since the Court offers no basis upon which to distinguish those omis-
sions for which a government can be held liable from those omissions for which it
cannot be held liable, there is a real possibility, one that is bome out by the tenor of
the judgments in the case, that a government will be held liable for any failure to fulfil
Charter values. I have already indicated that I agree with counsel for the Alberta gov-
ernment that if the Charter compels Canadian governments to compel private actors
to abide by Charter values in their private conduct, then the anti-discrimination provi-
sions of human rights statutes must necessarily mirror the grounds of discrimination
set out in section 15. Indeed, it seems that Cory J. did not seriously question this, for
he accepted that it followed from his decision that any failure to mirror the Charter’s
grounds of discrimination would be vulnerable to constitutional challenge, and merely
queried whether the concept of mirroring was not too simplistic, given that the Court
would remain free to find that the omission of a ground of discrimination was either
not discriminatory under section 15 or justifiable under section 1.’ In light of the fact
that in Vriend the Court treated the bare fact of exclusion as conclusive proof of dis-
criminatory effect and, further, found itself unable to imagine any pressing govern-
mental objective that could justify that exclusion, mirroring seems a fair description of
what the Court has required.

The implications of mirroring do not stop there, however. If a failure to prohibit
discrimination on certain grounds is itself discriminatory and contrary to the Charter,
then a failure to prohibit discrimination in certain settings must also be discriminatory.
Canadian human rights laws typically prohibit discrimination in employment, ac-
commodation and the provision of services. By implication, however, they permit dis-
crimination in all other settings. That permission must now be constitutionally vulner-
able, as Cory J. appears to have recognized when he asserted that the consequence of

5Ibid. at 552-53.

1999]

T MACKLEM – VRIEND V. ALBERTA

225

his decision was that the constitutionality of a government’s failure to prohibit certain
acts of discrimination would have to be assessed “by taking into account the nature of
the exclusion, the type of legislation, and the context in which it was enacted”” If that
is right, then Vriend has given to Canadian courts the task of determining the appro-
priate anti-discrimination strategies for the various Canadian jurisdictions.

Nor do the implications of mirroring stop even there. While it is perfectly clear that
the Court has not made the governments of Canada liable for their failure to fulfil the
values set out in the Constitution as a whole, there seems no reason to distinguish the
liability that the Court has recognized with respect to section 15 of the Charter from
liability for a failure to fulfil those values that underlie other sections of the Charter,
such as its guarantees of freedom of speech, privacy, and other personal liberties. If
the Charter compels Canadian governments to compel private actors to abide by
Charter values in their private conduct, then it must compel private actors to abide by
all Charter values, not just the value of equality. If that is the case, then a government
is likely to be found liable for a failure to enact legislation preventing people from
limiting one another’s liberties.

My concerns at these kinds of outcomes have already been explained in the course
of the discussion above on the nature of constitutions, so that there is no need to dwell
on them here.” For the reasons given above, such outcomes are damaging to the long-
term interests not only of gays and lesbians, but of all disadvantaged Canadians, in-
deed, Canadians of every kind.

B. Political Implications

Suppose, however, that I am wrong in all that I have said under the previous head-
ing. Assume that I have created a bogey by my description of the possible legal impli-
cations of the Supreme Court’s decision in Vriend. After all, the Court expressly de-
nied that the Charter rendered Canadian governments constitutionally liable for all
their omissions, even if it failed to offer any basis upon which to distinguish the con-
stitutional liability that it insisted upon from the constitutional liability that it denied.
Assume, then, that Vriend merely established what no one can dispute that it estab-
lished, namely, that the governments of Canada will, in certain circumstances, be
found to be constitutionally liable for their failure to fulfil Charter values, and that the
grounds on which they will be found liable, while not specified in Vriend itself, will
be developed by the Court as and when it is asked to determine the constitutional va-
lidity of particular government omissions. There is just as much, perhaps more, reason
to be concerned at this possibility.

The Supreme Court has suggested that incomplete human rights legislation-leg-
islation that prohibits some but not all grounds of discrimination-is likely to be
found to be discriminatory. It has also suggested that the absence of any human rights

46Ibid.
41 See Part IV, above, particularly the text following supra note 27.

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

[Vol. 44

legislation is unlikely to be found to be discriminatory, however baffling that conclu-
sion may seem in light of all else that the Court decided in Vriend. In doing so it has
made human rights policy an all or nothing matter. From now on a government that is
contemplating measures for the protection and benefit of those who have been disad-
vantaged on the basis of a particular ground of discrimination-such as race or sex-
will have to decide whether it is prepared to extend those measures to all grounds of
discrimination protected by section 15 of the Charter, or whether it would rather do
nothing.

This prospect is likely to have two adverse effects on disadvantaged groups. First,
the presence of constitutional liability is bound to multiply the resistance that policy
makers all too often exhibit toward anti-discrimination initiatives, for it will increase
the burden of those initiatives by foreclosing the possibility of their being approached
in a piecemeal manner. Policy makers and the governments for whom they act are
likely to ask why they should do anything new for women or racial minorities if it
means that they will have to do it for every other disadvantaged group. Second, the
requirement that all grounds of discrimination be addressed together makes it difficult
if not impossible, even for a well-meaning policy maker, to direct or tailor legislation
to the distinctive needs of particular groups. Human rights legislation is thus encour-
aged by the Court to be generic rather than context-sensitive, to treat disadvantage as
a shared circumstance rather than as an affliction that takes different forms and re-
quires different remedies for different groups of people.

What is even more damaging to the interests of disadvantaged groups is that
Vriend pits those groups against one another in a contest for government recognition,
so that they become rivals and not, as they should be, allies in the quest for justice at
the hands of their privileged counterparts. At trial, in the Alberta Court of Appeal, and
in the Supreme Court of Canada, homosexuals in Alberta were said to be distin-
guished from other disadvantaged groups, and so discriminated against by their in-
ability to gain access to the protection offered by the Individual’s Rights Protection
Act.” In other words, the Act’s denial to homosexuals of a benefit that it granted to
other disadvantaged groups was said to constitute a discriminatory distinction. The
courts’ error here is the common one of interpreting section 15 as if it prohibited dis-
advantage to certain people rather than disadvantage on certain grounds. It overlooks
the fact that the impugned distinction was not made on the ground of sexual orienta-
tion. Homosexuals are distinguished from heterosexuals by their sexual orientation.
They cannot be distinguished from women or racial minorities on that basis, however,

Vriend, supra note I at 541:

It is clear that the Individual’s Rights Protection Act, by reason of its underinclusive-
ness, does create a distinction. The distinction is simultaneously drawn along two dif-
ferent lines. The first is the distinction between homosexuals, on one hand, and other
disadvantaged groups which are protected under the Act, on the other. Gays and lesbi-
ans do not even have formal equality with reference to other protected groups, since
those other groups are explicitly included and they are not.

1999]

T MACKLEM – VRIEND V. ALBERTA

for women and racial minorities are homosexual as well as heterosexual and in the
same proportions as all other Canadians, or at least presumably so.

Even assuming that the error of finding discrimination in the distinctions among
disadvantaged groups is unlikely to be repeated in future decisions, it is an inescap-
able consequence of the decision in Vriend that disadvantaged groups in search of
protection from their governments are bound to look to and challenge the condition of
other, more fully recognized disadvantaged groups, for it is only by doing so that they
can acquire constitutional support for their cause. It follows, to pick some likely rival-
ries, that racial minorities will be prompted to attack the exclusive character of the
benefits that women have won for themselves; that gays, lesbians and the handi-
capped will be prompted to attack the exclusive character of the benefits that have
been won by racial minorities as well as women; that all disadvantaged groups will
find themselves focusing upon the limited access to justice that has been won by their
neighbours in disadvantage rather than upon the injustice of their own position in so-
ciety and the extent to which that injustice is the product of a prohibited ground of
discrimination. Given that the consequence of such rivalries is as likely to be no bene-
fits for any of the disadvantaged as benefits for all, it is possible that their existence
may suit the privileged. From the point of view of the disadvantaged, however, they
are distracting at best, destructive at worst.

All this follows not from some exaggerated reading of what Vriend decided, as
might be said of the concerns expressed under the previous heading. Rather, it follows
from the bare fact that in rendering Canadian governments liable for their failure to
extend the benefit of a particular human rights statute to all those who are discrimi-
nated against on a basis prohibited by section 15 of the Charter, Vriend indisputably
encourages governments and citizens to assimilate the condition of each disadvan-
taged group to the condition of every other disadvantaged group, so as to assume that
women, racial and religious minorities, the handicapped, the elderly and homosexuals
all suffer similarly, all need similar support, and so should all be responded to together
and in like manner.

C. The Importance of Reasons

It is customary for courts to offer reasons for judgment and to set out those reasons
in sufficient detail, at sufficient length, and with sufficient clarity to account for the
decision that they have rendered. It is not always necessary to do this of course. Cer-
tain appeals to the Supreme Court of Canada, for example, are dismissed with merely
nominal reasons for judgment.” In a case such as Vriend however, where what is at
stake is the well entrenched prejudice of a majority toward a vulnerable minority, it
seems essential to offer clear and coherent reasons for judgment, and conversely,
highly damaging to offer incomplete or contradictory reasons. There are two grounds
for thinking this.

9See e.g. an appeal as of right such as R. v. Bekoe, [1998] 1 S.C.R. 90, online: QL (SCJ).

228

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

[Vol. 44

First, people who are being asked to change their view of the world, people such as
those who opposed the prohibition in Alberta of discrimination on the basis of sexual
orientation, need to be offered reasons whose force they can acknowledge. If, instead,
those people recognize in the language of the Court’s judgment the rhetoric of their
ideological opponents, they will treat the Court as no more than one of those oppo-
nents, and will invoke against it whatever weapons they have in their political arsenal,
weapons that are bound to diminish the authority and standing of the Court, including
the weapon that the majority has reserved to itself for occasions such as this, the not-
withstanding or override clause contained in section 33 of the Charter. The Court in
Vriend expressly recognized that the Alberta legislature was entitled to invoke section
33,’ despite the fact that using the override clause in circumstances such as these
would nullify the Charter when it most mattered.”

Second (and more important in my view), reasons are the only weapon (short of
terrorism, which is likely to be met by counter-terrorism) that a disadvantaged minor-
ity such as Alberta’s homosexuals can hope to deploy against the majority. It is the
claim to justice that reason and reason alone can establish that offers the only real
possibility of overcoming the status, power and democratic authority of the majority
so as to secure the well-being of the disadvantaged and excluded. Indeed, without
good reason on their side, the disadvantaged will remain disadvantaged for the source
of their condition is the weakness of their cause in all realms but that of reason.
Therefore, they must insist upon reason, even if in the short term it costs them to do
so, as might have been the case in Vriend had better reasons been insisted upon. The
final issue to address, then, is whether good reasons can be found in support of a deci-
sion to read a prohibition against discrimination on the basis of sexual orientation into
the Individual’s Rights Protection Act.

VII. Alternative Solutions

As was already indicated, it seems at least possible to see the behaviour of the Al-
berta government in refusing to include sexual orientation as a ground of discrimina-
tion prohibited by the Individual’s Rights Protection Act.not as an omission, but as an
act. While I have emphasized that a government cannot be held constitutionally liable
for its omissions (a few positive duties aside), in certain marginal cases government
behaviour that takes the form of an omission may in fact amount to an act, in this case
an act of discrimination on the ground of sexual orientation, and so may be subject to
constitutional challenge on the same basis as any other government act.

See Vriend, supra note I at 565, 578, per lacobucci J.

SI I am assuming here that the omission of sexual orientation from the grounds of discrimination set
out in the Individual’s Rights Protection Act could be shown to be discriminatory by developing an
argument along the lines that I have tried to sketch under Part VII, below, namely, an argument that
would show that in all the circumstances that omission amounted to an act, and more particularly, to
an act of discrimination.

1999]

T MACKLEM – VRIEND V. ALBERTA

In the course of our everyday experience, we commonly recognize that certain
failures to act are acts in their own right. If I run into a colleague on the street and fail
to offer my hand or say hello I do not simply fail to act, as I would have if I had failed
to telephone that colleague. On the contrary, my failure to engage in the socially ex-
pected greeting amounts to an act of rudeness, as my failure to telephone would not,
unless of course a call had been promised. Similarly, we recognize that a failure to
applaud a performance-amounts to an act of criticism or censure; that a failure to defer
to those in authority amounts to an act of disrespect, perhaps rebellion; that a failure
to speak up for a friend amounts to an act of disloyalty. What these examples reveal is
that social context affects the application of the distinction between acts and omis-
sions. Context sometimes shows that what is an omission from one perspective is an
act from another perspective. The same behaviour can function as both act and omis-
sion. This does not mean, of course, that the distinction between act and omission is
indeterminate. Rather, it means that the distinction is always applied in some setting,
that a given piece of behaviour may operate in several settings at once, and that the
same piece of behaviour may therefore be at once an omission and an act.

Would it be accurate, then, to describe the Alberta government’s failure to prohibit
discrimination on the basis of sexual orientation as an act of discrimination in its own
right? It seems to me that it might well be. The omission of a ground of discrimina-
tion as notorious as sexual orientation (with all due respect to the reasons of Sopinka
J. in Egan’), from legislation that has been held by the Supreme Court to possess
quasi-constitutional status, 3 has powerful symbolic overtones. On the one hand, the
implication of those overtones might be that the omission in itself amounted to a dis-
criminatory act, in the same way that a failure to include someone in one’s guest list
might amount to an act of exclusion if the party was important enough and the person
in question was an expected guest there. On the other hand, the implication of those
overtones might be that the omission went beyond a legitimate recognition by the
government of the existence of a private realm of individual moral responsibility, a
realm in which each of us is expected to determine what constitutes discriminatory
behaviour and how best to avoid it, so as to amount to an authorization from the Al-
berta government to the Alberta people to discriminate on the basis of sexual orienta-
tion.’

‘2 Supra note 6 at 576.
51 See e.g. Canadian National Railway Co. v. Canada (Canadian Human Rights Commission),
[1987] 1 S.C.R. 1114 at 1136,40 D.L.R. (4th) 193.

, Notice that these arguments would be no less compelling if the legislation had prohibited only
two grounds of discrimination rather than nine. They might even hold in the absence of any legisla-
tion. The reason is that the comprehensiveness of a law does not, in and of itself, make omissions
from that law acts. Notice too that the deliberateness of an omission is not enough to make that omis-
sion an act. One must assume that in enacting legislation in a particular form, a responsible legislature
has considered and rejected the various alternatives to that form, alternatives that are normally set out
in cabinet submissions and/or are raised by the opposition in debate. It follows that every omission

MCGILL LAW JOURNAL / REVUE DE DROIT DE MCGILL

[Vol. 44

I am not claiming that an argument along these lines would or should necessarily
succeed. It may be that the best description of the Alberta government’s behaviour is
that it simply failed to act in circumstances where it was morally but not legally bound
to act. If that is true, then the responsibility for that failure must lie with the Alberta
government, and more to the point, the enforcement of that responsibility must be a
matter for the people of Alberta, not the courts. As was argued above, if the courts
possessed or were to be granted the power to right every wrong committed by gov-
ernment-even in the realm of Charter rights and equality-then government would
to that extent be deprived of moral responsibility for its conduct, and democracy
would be deprived of moral significance. As I have tried to make clear, such a conse-
quence would be very much to the disadvantage of the disadvantaged. Making the
private public by deciding, for example, that the elimination of discrimination requires
a direct legal prohibition rather than a gradual, consensual evolution, of social atti-
tudes, is what government is very largely about. The burden of government does not
and should not belong to the courts alone, but to the people of Canada and their repre-
sentatives in Parliament and in the legislatures. Where a legislature does wrong there
must indeed be a remedy, but not necessarily a Charter remedy.

from legislation is deliberate, so that to render Parliament and the legislatures liable for all their delib-
erate omissions would be to render them liable for all their omissions.