Each Thursday we present a significant excerpt, usually from a recently published book or journal article. In every case the proper permissions have been obtained. If you are a publisher who would like to participate in this feature, please let us know via the site’s contact form.
Brian Langille, Professor of Law, University of Toronto and Pnina Alon-Shenker, Associate Professor and Director of Ryerson Law & Business Clinic, Ted Rogers School of Management at Ryerson University
Reprinted with permission. (2015) 4:2 Canadian Journal of Human Rights 211
Excerpt: Introductory abstract and Parts III and IV
[Footnotes omitted. They can be found in the original via the link above]
In this article, we argue for purposive interpretation of statutory labour laws when issues of their “scope” or “range of application” arise. While this purposive approach has been rhetorically dominant, it often fails to fulfill its promise in our case law. Drawing on Tussman and tenBroek’s work, this article calls attention to the structure of thought involved in legislative “classifications”, which is not a new idea but has been absent from current discussions. We stress that determining appropriate coverage of labour laws requires rational and pragmatic reasons for treating people differently which go beyond legislative classifications to the purposes of the specific law. This article critically reviews the Supreme Court of Canada’s recent decision on the application of human rights laws to law firm partners in McCormick v Fasken Martineau DuMoulin LLP in an effort to show how the purposive approach is invoked, how it is then either ignored or applied incorrectly, and how the purposive approach ought to have been deployed if we had remained faithful to its structure and demands.
III. Some Canadian Illustrations
A. When Things Go Right
We refer to Canadian cases from several labour law contexts as a starting point and to cement our basic idea of “purpose” in understanding the proper scope of application of various labour laws. We begin by noting that when determining coverage of labour laws (and thus possible and proper segmentation), Canadian courts, at least in some circumstances, simply think in terms of specific statutory purpose (and have not been tripped up along the way by labour law’s commonly understood overall purpose). In these cases, the concepts with which we have been dealing, such as “employer” and “employee”, even though present, provide no barrier to the protection of those who should be protected, i.e. to the achievement of specific statutory purposes. However, the central and potentially radical import of these cases has gone, as far as we can tell, largely unnoticed.
Our first case example involves a truck driver who was crushed between his own truck and that of a fellow independent contractor. The Ontario Court of Appeal held that under Ontario’s Occupational Health and Safety Act, the broker for this group of independent truck drivers had to establish a health and safety committee. Such a committee was required “at a workplace at which 20 or more workers are regularly employed.” The question raised in this appeal was whether the truck drivers, who were from one legal point of view independent truck owner-operators (and thus not employees), were “regularly employed” within the meaning of section 9(2)(a) of the Act. The answer is yes. The goal of creating safe workplaces is not to be hobbled or thwarted by passing this goal through another substantive filter of legal categories having no rational relationship to that objective. The idea of “employer” and “employee” to be deployed in this case is completely filled up by this way of thinking. All that is required is the statute, an understanding of its purposes and a rational line of thought from purpose to “coverage”.
The same logic has been applied to human rights statutes. Our second example involves a cook employed by a catering company who was assigned to a railroad gang in rural Saskatchewan. Soon after his assignment the railroad gang discovered that the cook was HIV positive. Whilst the group had a variety of responses, the roadmaster was seriously concerned and refused to eat food prepared by the cook. The Human Rights Tribunal found that whilst the cook had not explicitly been fired, the inhospitable environment constituted constructive dismissal. The issue on appeal was whether the railroad gang or the roadmaster had “employed” the cook. Canadian Pacific argued that if the cook was not employed by the railroad gang, he could not be constructively dismissed by them. The court disagreed. The terms “employer”, “employ” and “employment” within human rights statutes are to be interpreted to advance the purposes of those statutes. Narrow definitions of these terms defeat those purposes. In the cook’s case, the railroad gang’s “utilization” of the cook was sufficient to find an employment relationship for the purposes of human rights statutes.
These are the sort of cases we wish to explore. The basic ideas we take from these cases are the following. First, that the key idea in these cases is simply that of purposive statutory interpretation. Second, that statutory purpose can take us just about everywhere we need to go in thinking about most labour laws. Third, this means that we will have as much segmentation as we need and should have if we simply apply labour law to whom it should be applied. Fourth, that this does not require the creation of new categories of workers, or new labels – in fact we can use any term we wish as long as we use it sensibly, i.e. purposively. Fifth, and more controversially, the urge to define new categories is a legacy of our starting point, particularly the traditional employer/employee relationship and the idea that labour law emerged as a need to redress imbalances therein. The trick, in short, lies in simply sticking to the statutory purposes at issue in any case. This would help liberate our thinking and enable us to reconceive our account of labour law’s overall purpose in a way which fits with rational legal decision making, and also with much more profound normative thinking than the simple tale of inequality of bargaining power permits.
These cases set an example for, and issue an invitation to, all of our labour laws. In a case such as that of the Ontario health and safety committee, there are statutory “definitions” but they do no “work”. They are on their face “empty”. The real work has to be done by thinking about what the statute is trying to do. Our courts and tribunals have generally done so with no difficulty, but unfortunately not always.
B. When Things Go Wrong
Faskens is a striking example of how things can go wrong if we ignore these basic points. The case is one of overt discrimination on the basis of age. The facts are as follows. McCormick was a partner in the Vancouver office of a large (by Canadian standards), law firm which we will refer to simply as “Faskens”. About 650 lawyers worked for the firm and of these, 260 were “equity partners” who had an ownership interest in the firm. McCormick was an equity partner. The Partnership Agreement at Faskens contained a mandatory retirement policy at age 65. McCormick complained, not unreasonably, that this violated the Code, which prohibits discrimination in employment on the basis of, among other grounds, age. There is no denying that this was age discrimination. It was not sophisticated, covert or “adverse effect” based. There was no “facially neutral rule” which had an “impact” upon a protected group here. Faskens is simply a case of overt, intentional discrimination against an individual which cost him his job. It was as if Faskens had simply said you have to go because you are a Catholic or a Jew or black or a woman.
Given this obvious point, Faskens’ legal escape route was to argue that the Code, in its entirety, did not apply to the relationship between law firms and their partners. Their argument, unattractive as it may be, was that law firms can legally discriminate against partners of the firm. This means that partners of law firms can be forced out of the firm not simply on the basis of age, but also on the basis of sex, or religion, or race. That is quite a striking claim. The idea that one group of Canadians, no matter whether rich lawyers or poor agricultural workers, were somehow carved out of these sorts of basic human rights protections is, on its face, quite an extraordinary one. Yet this is the claim which Faskens made. Faskens asserted that the legal basis for this most unattractive legal conclusion was the idea that McCormick was not an “employee” of Faskens. Because McCormick was a partner, he could not be an employee under the Code. Since the Code protected only employees, the protections were not applicable to him.
This was the argument which the British Columbia Court of Appeal accepted. Human rights codes do not apply to law firms in their relations with their partners. The law firms are immune from human rights law. That is quite a “knock out” result. By contrast, the Supreme Court rejected the notion that “a partner in a firm can never be an employee under the Code”. But it also admitted that “the structure and the protections normally associated with equity partnerships mean they will rarely be employment relationships for purposes of human rights legislation”. That is, while the Court of Appeal focused on partnership as a legal concept which cannot be reconciled with an employment relationship, the Supreme Court examined “the substance of the actual relationship” and reached very similar results. How did the Supreme Court get there?
First, the Court expressed its view of the issue before it as follows:
The issue before this Court, therefore, is how to characterize Mr. McCormick’s relationship with his firm in order to determine if it comes within the jurisdiction of the Code over employment. That requires us to examine the essential character of the relationship and the extent to which it is a dependent one.
No rationale or explanation was offered for this requirement which turns out to be critical to the Court’s conclusion.
Second, and in our view surprisingly, the Court began not with the Code, but with a long explanation of how law partnerships, in this case at Faskens, actually work. At Faskens, the 260 equity partners run the firm through a Partnership Board of 13 partners from across the firm’s various offices. The Board appoints the Managing Partner. There were about 60 partners in Vancouver and the British Columbia region sent three partners to the Board. Board members are elected by the partners and at one point, McCormick himself had served on the Board (under its previous form as the Executive Committee). The “constitutional” basis for all of this was the Partnership Agreement (which also contained the mandatory retirement policy of which McCormick complained). The Partnership Agreement could be amended only by vote of the partnership as a whole. (Thus, McCormick was in fact “fired” under the mandatory retirement policy by a majority vote, or perhaps nonvote to change the policy, of his partners). Partners such as McCormick have an “ownership” interest in the firm and a share of the equity. The mandatory retirement policy took the form of requiring that all equity partners divest their share of ownership at 65. It was theoretically possible under the policy to make an arrangement to continue working for the firm after having ceased to be an equity partner, but the Partnership Agreement stated this was to be “the exception rather than the rule”. Basically, once you cease to be an equity partner, you are out. This is what happened to McCormick.
Third, and eventually, the Court turned its attention to the Code. It began its treatment with the almost ritual invocation of the ideas that the Code is “quasi-constitutional” legislation and that it “attracts a generous interpretation to permit the achievement of its broad public purposes”. But after their incantation, these ideas exit the judgment.
Fourth, the Court then expressed its view about those purposes:
Those purposes include the prevention of arbitrary disadvantage or exclusion based on enumerated grounds, so that individuals deemed to be vulnerable by virtue of a group characteristic can be protected from discrimination.
That sounds reasonable enough. In fact exactly right. And is this not precisely what happened to McCormick?
Fifth, the Court made the very important assertion:
The Code achieves those purposes by prohibiting discrimination in specific contexts. One of those contexts is “employment”. The definition of employment must be approached consistently with the generous, aspirational purposes set out in s. 3 of the Code and understood in light of the protective nature of human rights legislation, which is “often the final refuge of the disadvantaged and the disenfranchised” and of “the most vulnerable members of society”. This is the philosophical framework for ascertaining whether a particular workplace relationship represents the kind of vulnerability the Code intended to bring under its protective scope.
Again, that sounds right – and would give McCormick heart.
Sixth, the Court then turned to the language of the Code – it does apply to “employment”. That was the basis of McCormick’s claim. He claimed that he was being forced out of employment at his law firm because he was 65. So the legal issue is just our issue – what does “employment” mean for the purposes of the Code? The Code itself is singularly unhelpful – as are most labour or employment law statutes – in this regard.
For years, we both have been confident of the view that the Code does not concern itself with the niceties, or inadequacies, of the familiar labour law distinction between “employees” and “independent contractors”. Whatever the distinction’s usefulness, say in regards to issues of vicarious liability, it is intellectually and legally useless here because no one cares, and the Code does not care, whether you refused to employ a person as an employee or as an independent contractor if the reason you did so was because of sex, race, religion or age. The Court helpfully confirms this “expansive” approach. Again, so far so good for McCormick.
Seventh, unfortunately, the Court then took the following, unexplained tack. Rather than developing the idea of the purposes of the Code, it embarked upon a very doctrinal, arid and terribly familiar discussion of “employment” in the abstract – all unrelated to the discussion thus far about the allegedly broad purposes of the Code. The Court simply asserted that
[d]eciding who is in an employment relationship for purposes of the Code means in essence, examining how two synergetic aspects function in an employment relationship: control exercised by an employer over working conditions and remuneration, and corresponding dependency on the part of the worker.
The Court very briefly explained that “the emphasis on control and dependency … is consistent with approaches taken to the definition of employment in the context of protective legislation both in Canada and internationally”. The Court accepted that “while significant underlying similarities may exist across different statutory schemes dealing with employment, it must always be assessed in the context of the particular scheme being scrutinized.” But control and dependency seems to flow from a large scale notion of labour law’s standard and tired story (of protecting vulnerable workers), which gets in the way of, and defeats everything just said about, the purposes of our human rights laws (of preventing victimization based upon a prohibited ground) and McCormick’s claim.
The rest of the decision constitutes a meditation upon the ideas of “control” and “dependency”, along with the development of the notion that being a partner is inconsistent with these two ideas.
Along the way to this conclusion, the Court added to our problems in a variety of ways. For example, it seemed at ease with the idea that it is appropriate to look at the jurisprudence under collective bargaining statutes to ascertain the meaning of employment for the purpose of human rights law. Note that the Court cited Guy Davidov’s work to explain its shift to control and dependency tests when examining employment for the purpose of the Code. The Court asserted that this move was “consistent with the approaches taken to the definition of employment in the context of protective legislation” and again referred to Davidov’s work. But even labour law scholars who have defended the traditional idea of labour law’s mission must find this move puzzling. Davidov, for example, acknowledged that some laws (such as human rights and health and safety regulations) have broader application. When constructing his test for who is an “employee”, which is built on criteria such as dependency and control, Davidov stressed that his focus was on “the appropriate scope of employment standards and collective bargaining laws”, rather than human rights and health and safety regulations which are quite different.
The Court also suggested that partnership law and the very idea of partnership have relevance to the coverage of our human rights codes. This seems implausible – partnership surely is as irrelevant to the coverage of the Code as the idea of “independent contractor”, flowing from vicarious liability law, was to the health and safety committee case discussed above. The Court also stressed the fact that partners have the chance to participate in decision making processes which affect them, but the relevance of this fact remains unexplained. (One might, perhaps, entertain the idea that this would be a relevant consideration if the issue were collective bargaining coverage).
The bottom line is that law firm partners, such as McCormick, have no human rights protection because they do not fit the standard model of an “employee” according to the test of “control” and “dependency”. Rather, because they have some degree of control they look more like the standard narrative’s picture of an “employer”. The most radical statements are the Court’s conclusions that “[i] n most cases, … partners are not employees of the firm, they are, collectively, the employer”, and that as a partner, McCormick is seen “more as someone in control of, rather than subject to, decisions about workplace conditions.”
Note that no explanation at all was offered for the implicit conclusion that someone cannot be an employer for some purposes and also an employee for another purpose. Nor was there a robust engagement with the obvious point – that, while as a partner McCormick exercised more control and suffered less dependency than many other employees, he was still ousted from the partnership because of his age simply because a majority of his colleagues thought that was a good idea. His lack of control on the very point at issue was the reason for the litigation. The Court seems to be of the view that because he could object and vote against the mandatory retirement policy, he was the author of his own misfortune. So too the Court was of the view, it seems, that because partners have a sort of “tenure” (i.e. are much more difficult to fire), this was a reason to take their removal from the partnership less seriously, rather than more seriously.
On the test the Court adopted, one is left wondering if there are not a lot of other employees who are now no longer covered by human rights law. So, to take an example close to home, tenured university professors exercise a lot of control and suffer much less dependency than most workers. Are they (we) now excluded?
The main point for our purposes is this. The Court elided the idea of the rationale for extending human rights protection to those groups identified by the prohibited grounds (which, as we have just seen and as the Court itself noted, is the idea of their vulnerability in virtue of the group characteristic) with the idea of the various contexts in which the Code seeks to protect those so identified as vulnerable, e.g. housing, service provision and employment. Here the Court slid into not only a moralized (vulnerability) account of why we have human rights protections for certain groups (a sound move), but also to a moralized (vulnerability) account of the contexts in which these protections are to be operable (a very unsound move).
Here, in the form of a question, is a simple way of putting this point: what if a wealthy black person, say an equity partner at Faskens, were denied accommodation on the basis of race? (Or refused service in a restaurant?) We would not “moralize” the housing market (or provision of restaurant services). We would not say that only those members of protected groups (who are statutorily deemed to be vulnerable) who are also weak and desperate for housing (or other services such as food) get human rights protection. That is, you don’t have to be “doubly vulnerable”, once in virtue of membership of a protected group and again in the particular sphere of market activity (services, accommodation and employment) which is at issue. That is what the Court constructed here – a double vulnerability.
A second critical point is this – the idea, and any discussion, of the specific statutory purposes at stake were completely dropped out. As a result, we were never told why these ideas of “control” or “dependency” bear at all on our reflections about “coverage” (i.e. what is an “employment” for the purposes) of our human rights law.
What should have been done? Our suggestion is that we should not moralize, in this case via a familiar account of the general purpose of labour law as a whole, a straightforward legal problem of determining statutory purposes and applying the statute rationally in light thereof. We need to advance a sensible, purposive, rational, pragmatic, non-moralised question about allocation of responsibility, incentives and relationships which are quite oblivious to any standard notion of “employment”. If this view is right, we are able to end up with just the right amount of segmentation, or lack thereof, which our various laws demand.
IV. Getting It Right Next Time
There are three parts to a plan to avoid the problem we have identified and which is so well represented by Faskens. First, we need to revisit the idea of statutory purposes and explore the structure of reasoning involved in such an endeavor. Second, we need to carefully articulate our statutory purposes and examine concrete cases such as Faskens through this lens. Third, which we will defer for another day, we have to offer a new and better account of labour law and its basic purpose, which fits with the specific purposive approach, does not “get in our way”, and thus avoids the unnecessary problems on display in Faskens.
A. Purposes and Classification
Hugh Collins has alluded to the possibility discussed here – that the problem of “coverage” of various labour laws could be solved by thinking about purposes. But he dismissed this possibility, writing:
It is always tempting to urge the courts to adopt a purposive approach, and indeed this was attempted for a brief period in the USA. But without additional guidance this seems highly indeterminate and vulnerable to judicial misconceptions of purpose, and furthermore it seems unlikely that this would overcome the problem of choice described above. If the courts are to engage in the imposition of forms of government over economic relations, they require a firmer set of criteria which both make sense in the world and establish intelligible boundaries to the reach of employment protection legislation.
There are three objections listed here – indeterminacy, misperceptions of purpose and “the problem of choice”. None of these seem, to our mind, to be particularly powerful. The first two can be met by the observation that nothing could be more indeterminate nor misconceived than the type of reasoning we see on display in Faskens, which is visited upon us by our current thinking. Collins’ view is that we need “firmer criteria”. Our view is that we have too many criteria which are too firm and too familiar. By the “problem of choice”, Collins refers to his discussion of the courts deferring to the intention of the parties as to the form of the engagement of labour. It is hard to see how this is not the very point in issue – that is, that the issue is one of public policy, not private choice. This is precisely the focus that the idea of attending to purpose is meant to achieve.
As usual, however, Collins has accurately uncovered a real problem. It was one which Tussman and tenBroek laid bare, and provided the basics of a solution to, decades before. In their famous essay “The Equal Protection of the Laws”, the authors provided the intellectual building blocks for rational legal thought about legislative “classifications”. The basics of the approach are as follows. All laws will not apply to all people. Some laws will be general and apply to all, and some will be “special” and apply to some. In the latter case, the law will by necessity “classify” those to whom it does and does not apply. The problem of classification – of the legitimacy of classification57 – has a structure. To classify is to define a “class” to whom the law applies. To do this involves designation of a “quality or characteristic or trait or relation, or any combination of these, the possession of which, by an individual, determines his membership in or inclusion within the class” (e.g. aliens, all over age 25, foreign corporations). Further, a reasonable classification is one which includes all who are similarly situated and none who are not. But determining who is similarly situated presents a problem. We need some independent point of reference to do this job; it cannot be enough that all in the class possess the trait set out to define the class. That would be tautological and any classification would be therefore reasonable. As Tussman and tenBroek observed:
The inescapable answer is that we must look beyond the classification to the purpose of the law. A reasonable classification is one which includes all persons who are similarly situated with respect to the purpose of the law.
The purpose of a law may be either the elimination of a public “mischief” or the achievement of some positive public good. … We shall speak of the defining character or characteristics of the legislative classification as the trait. We can thus speak of the relation of the classification to the purpose of the law as the relation of the Trait to the Mischief.
Then our authors came to the problem which Collins confronted decades later when addressing the problem of classification in labour law, and they hit the same nail on the head. Tussman and tenBroek wrote that “[a] problem arises at all because the classification in a law usually does not have as its defining Trait the possession of or involvement with the Mischief at which the law aims.”
This seems, at first blush, to be the case. Most statutes do not have a definition section which reads “this act applies to all those tainted by the mischief that this statute aims to cure”. It would make life a lot easier if more statutes did so. This is because if a statute defined the class to whom it applied directly in terms of those who suffer from the mischief to be ameliorated, legal life would be much simpler because “[t]he class, being defined directly in terms of the Mischief, automatically includes all who are similarly situated with respect to the purpose of the law.”
The drawbacks of such an approach which Collins identified above, it will be recalled, are just problems with adjudicative “discretion” – i.e. indeterminacy and misconception of purpose and so on. Tussman and tenBroek observed along the same lines that “[t]his procedure requires, however, delegation of considerable discretion to administrators. … Legislators, reluctant to confer such discretion, tend to classify by Traits which limit the range of administrative freedom.”
So, it seems we are left with the problem of two classes: the mischief (M) and the identification of a trait (T) which is a sort of proxy for the “ideal” solution of simply one class – the mischief itself. Thus, the law often generates for itself the following familiar problem. The group identified by the Trait, i.e. the nonpurposively described class (T) to which the law in its attempt to capture the class defined by the statutory purpose (M), may or may not in fact overlap with the class (M) we seek to help. As Tussman and tenBroek outlined for us:
In other words, we are really dealing with the relation of two classes to each other. The first class consists of all individuals possessing the defining Trait; the second class consists of all individual possessing, or rather, tainted by, the Mischief at which the law aims. The former is the legislative classification; the latter is the class of those similarly situated with respect to the purpose of the law. We shall refer to these two classes as T and M respectively.
Now, since the reasonableness of any class T depends entirely upon its relation to a class M, it is obvious that it is impossible to pass judgment on the reasonableness of a classification without taking into consideration, or identifying, the purpose of the law.
There are five possible relationships between the class defined by the Trait and the class defined by the Mischief. These relationships can be indicated by the following … :
(1) All T’s are M’s and all M’s are T’s
(2) No T’s are M’s
(3) All T’s are M’s but some M’s are not T’s
(4) All M’s are T’s but some T’s are not M’s
(5) Some T’s are M’s; some T’s are not M’s; and some M’s are not T’s
One of these five relationships holds in fact in any case of legislative classification.
We can pause here to say that in our view, the Supreme Court in Faskens produced an under-inclusive result (3, probably). Their understanding of T (employee) did not reach all in M (those in need of and entitled to human rights law protection). This was because their T proxy for M was really, if anything, a proxy for something else. (And the best account of “control and dependency” is that it is the Trait for identifying those involved in the Mischief of harms caused in circumstances we see as calling for vicarious liability).
Now we face the real problems – and have at least some material to help us do so. Is Collins right in saying that it is preferable to not define T as M, so that T = M? Is it true that we need to separate T and M? Are we worried more about discretion than irrational definitions (T’s)? And is it what our statutes are actually doing?
We can begin by noting that Tussman and tenBroek are, unlike Collins, agnostic on this point. They simply noted that as a matter of fact, legislatures are reluctant to “confer such discretion” on adjudicators and tend to “classify by trait”. They passed no explicit judgment on whether this is a good legislative judgment or sound legal procedure. Yet implicitly, their whole article is testimony to the problems the law gets itself into when we refuse a grant of rationally governed discretion and insist on using both T and M in an attempt to curb discretion. Their article would not have been necessary without this legislative tendency.
Recall that Tussman and tenBroek were making a point about the constitutional demand of equality. We are not. We are not making a constitutional argument that, for example, Faskens is wrong (although such an argument is clearly available). But it turns out that the idea of equality – treating like cases alike and how we know what counts as being alike – is basic to legal thinking. At the root is the idea of having rational reasons for treating people differently, and this is just the key to deciding to whom a statute should apply.
As we have noted, however, most of the time we do not really have a very helpful definition of T in our labour law statutes. Most of the time we have, as in Faskens, simply the invocation of the word “employee” and perhaps “employer” as one who employs employees. As a result, many of our statutes define, as we have noted, T in terms of “striking circumlocution”.
Occasionally we find very awkward attempts to define T (such as “dependent contractor”). However, these usually turn out to be unhelpful responses to poor adjudication, under a definition of the “striking circumlocution” variety, in which there has been a failure to equate T with M.
The point being made here is that in cases where there is no real statutory definition of T, there is an obvious tendency to fill this void, by visiting upon this non-defined T, the standard and tired labour law narrative. Our point is that this is a large mistake, as Faskens shows so well. This is what the Supreme Court of Canada did in Faskens and the result was, as we have seen, that the Court did not attend at all to the relevant specific purposes of the Code. If Faskens shows anything, it is that we are always doing purposive interpretation, whether we know it or not. It is better to do it overtly and correctly.
B. What the Supreme Court of Canada Should Have Done in Faskens
Let us now go back to the decision in Faskens and see whether this better understanding of classification could have led to a different analysis and perhaps a different result. The central issue in Faskens was an issue of classification – to whom the Code applied, and whether it applied to McCormick or not. The Code prohibits discrimination by a “person” against another “person” (the “complainant”) on the basis of various prohibited grounds, such as race, sex and age in various social contexts, such as provision of services, housing and employment.
The statutory definitions, however, are not very helpful to McCormick nor to Faskens. The definition of a “complainant” is completely empty, while the definition of a “person” is circumlocutory. There is no definition of an “employee” or an “employer” (which is listed under the definition of a “person”), and the definition of “employment” tells us absolutely nothing.
An attempt to extract a definition of the class to whom the Code applies by various characteristics (T) would be completely tautological: it covers those who are discriminated against on the basis of prohibited grounds and those who discriminate against them in various social contexts, such as employment.
An independent point of reference is therefore needed. We need to articulate the mischief (M) which the Code aims to eliminate, in order to identify all the persons who are similarly situated with respect to the purposes of the Code, and whether McCormick is one of them. As noted above, the Code does not provide a definition of the class to whom it applied (T), and, all the more so, a definition which relates to the mischief to be ameliorated by the Code (M).
Zooming in on the context of employment, we have to think about the purposes of the Code in order to have a better understanding of the terms “employee”, “employer” and “employment” – those persons who are similarly situated for the purposes of the Code. It should be acknowledged that the meaning of “employee”, “employer” and “employment” may be different for the purposes of different acts and even different statutory provisions within a single act.
Probing the purposes of the Code, the Court could have started its analysis in Faskens with section 3 of the Code, which was mentioned only in passing. While the Court correctly acknowledged that “employment” should be interpreted in a way that was consistent “with the generous, aspirational purposes set out in s. 3”, the Court does not go about this analysis. Here is the full text of section 3:
The purposes of this Code are as follows:
(a) to foster a society in British Columbia in which there are no impediments to full and free participation in the economic, social, political and cultural life of British Columbia;
(b) to promote a climate of understanding and mutual respect where all are equal in dignity and rights;
(c) to prevent discrimination prohibited by this Code;
(d) to identify and eliminate persistent patterns of inequality associated with discrimination prohibited by this Code;
(e) to provide a means of redress for those persons who are discriminated against contrary to this Code.
Section 3 is only somewhat helpful. It talks about promoting public purposes such as “equality”, “mutual respect” and “dignity” and preventing a public mischief – that is, “discrimination” – without explaining these very big terms. Here our case law and scholarly literature can be auxiliary. Discrimination is often understood as
a distinction, whether intentional or not but based on grounds relating to the personal characteristics of the individual or group, which has the effect of imposing burdens, obligations, or disadvantages on such individuals or groups not imposed on others, or which withholds or limits access to opportunities, benefits, and advantages available to other members of society. Distinctions based on personal characteristics attributed to an individual solely on the basis of association with a group will rarely escape the charge of discrimination, while those based on an individual’s merits and capacities will rarely be so classed.
Three additional points should be noted. First, discrimination is more than just a distinction between individuals. Generally, we are allowed to make a distinction in favour of or against another person on the basis of one’s favourite rock band or the colour of one’s shirt. Distinction amounts to discrimination when the unequal treatment is based on an enumerated ground, i.e. a personal characteristic, such as colour of skin, race or sex, which is immutable (or constructively immutable), and should therefore not be the basis for the assessment or treatment of individuals. The list of prohibited grounds under the Code corresponds to historically disadvantaged groups whose stigmatization and marginalization the Code aims at redressing.
Second, treating like people alike may still result in discrimination. The purpose of the Code is therefore to promote “substantive equality” rather than identical treatment. Promoting (substantive) equality entails “the promotion of a society in which all are secure in the knowledge that they are recognized at law as human beings equally deserving of concern, respect and consideration.”
Many legal and philosophical scholars have attempted to provide substantive content to the concept of equality by identifying several principles that equality aims to protect. While some advanced a single principle or purpose – such as ending oppression, ensuring sufficiency, or protecting the interest of belonging and the benefits of full membership in social, economic and political lives – others took a more pluralistic approach. Sophia Moreau, for example, spoke about the three wrongs of discrimination: unequal treatment which is associated with stereotyping and prejudice, oppression and denial of basic goods. Denise Réaume specifically discussed the harms that are associated with discrimination in the private sector as implicit or explicit conduct motivated by stereotypes or prejudice, or by a denial of a fair opportunity to participate in important activities and social institutions. Finally, T.M. Scanlon articulated five diverse reasons for eliminating inequality, including alleviation of suffering, prevention of unacceptable forms of power or domination and elimination of stigmatizing differences in status.
Third, discrimination is prohibited only in various social spheres such as housing, provision of services and employment. Generally, people are autonomous moral agents. They are obliged only to respect others’ rights and liberties (including the right to equality) by not restricting or interfering with other persons’ freedom to enjoy their rights and liberties. However, in some social spheres, there is a case for imposing broader obligations on private actors. When they are in a position to distribute and redistribute benefits or resources comparable to the State and have the ability to create social change, these actors are subject to a duty to treat people with equal concern and respect.
Accordingly, the Code applies to persons who are treated unequally on the basis of personal immutable characteristics (rather than their merits) corresponding to the enumerated grounds (such as black, gay or older people) and are denied certain goods or full participation in economic, social, political and cultural lives. They are those who suffer from the mischief that the Code is designed to cure (M).
In the context of employment, this includes employees, independent contractors, temporary, contract, seasonal and casual workers, volunteers, interns, co-op students and any person who engages in economic or productive work in the labour market. They are all similarly situated with respect to the purpose of the Code because any type of work is “one of the most fundamental aspects in a person’s life, providing the individual with a means of financial support and, as importantly, a contributory role in society. A person’s employment is an essential component of his or her sense of identity, self-worth and emotional well-being.”
The Code also applies to those who are in a position to create or exploit this mischief – those who have the power to bring about social change and redistribute benefits or goods in a way that will repair injustice and social exclusion. In the context of employment, this could include employers, temporary employment agencies, contractors and many more.
Considering that McCormick engaged in productive work (providing legal services to clients on behalf of the partnership) and that Faskens could and had, in fact, affected his ability to have full and meaningful participation in economic and social life by imposing a mandatory retirement policy, the Code applies to his relationship with Faskens and generally to the relationship between a law firm partner and a law firm partnership.
Although Faskens’ mandatory retirement policy clearly established a prima facie case of age discrimination, the Supreme Court of Canada could have then examined whether it was justified for other reasons – either some sort of estoppel given that McCormick had benefited for years from the arrangement of forcing other partners to retire, or statutory exceptions such as a bona fide occupational requirement.
Here an analysis of the meaning of age discrimination and the purposes of anti-age discrimination law becomes relevant. In short, while age-based distinctions are considered “a common and necessary way of ordering our society” – and less harmful than other forms of discrimination – age discrimination might be associated with significant wrongs especially when it comes to older people. As Alon-Shenker has written,
[p]eople of advanced age do represent a historically disadvantaged group, particularly in the workplace. Although seniors may not be a typical “minority group,” and may even include privileged individuals, they have some central characteristics of minority groups such as identifiable physical characteristics and shared social and institutional expectations (including the expectation of retirement). They are often subject to negative stereotypes and they face discrimination in many spheres, including employment, health services and housing.
Accordingly, anti-age discrimination in employment law aims at promoting various purposes to remedy these wrongs, including preventing opportunistic behaviour by employers, who are tempted to dismiss older workers when their labour costs are higher than their marginal productivity; promoting displaced older workers who experience major challenges to reemployment; protecting against ageist stereotypes and ageism; and alleviating social isolation, oppression and economic deprivation among older workers.
Although no categorical answer can be given here, it seems as though Faskens could have provided evidence to establish a claim that its mandatory retirement policy did not involve any of the aforementioned wrongs. For example, the law firm could have argued that the policy was not designed on the basis of a stereotypical belief that older partners were not productive, but rather was designed to benefit all partners by ensuring the regenerative turnover of partnership shares. The result would have been that while the Code applied to McCormick (and to law firm partners in general), and the mandatory retirement policy was discriminatory on its face, the policy was justifiable under the Code. This would have been perhaps a disappointing result for McCormick, but a promising and sound result for many other people who engage in productive work and are vulnerable, by virtue of a personal characteristic, to discrimination.