The Right to Bargain - 201
A couple days ago, I posted a primer on the right to bargain in Canada.
I've also gotten involved in some Twitter discourse about it among other lawyers (mainly from outside the labour discipline), and quite frankly Twitter's not well-suited to the discussion. So this is a bit more advanced.
But the thing I want to be absolutely clear about is this: None of this is necessary to find Bill 28 objectionable. This analysis is abstract, talking about the extent of the obligations and limits that s.2(d) puts on government legislation. However, once you accept that s.2(d) has ANY collective bargaining content at all, it's a very short step from there to "government can't legislatively impose collective agreements without infringing 2(d)".
The outer limits deal more with government regulation of private sector bargaining relationships - whether or not the Charter imposes a positive obligation on governments to require employers to bargain in good faith, or to reinstate employees once a strike ends.
The Wagner Model
In 1935, the US government passed the "National Labor Relations Act", better known as the "Wagner Act". This established rights of workers to form unions, negotiate collectively with their employers, and to strike. However, it also imposed limits on their ability to strike.
By contrast to the earlier labour era, where unions could be ad hoc (and often illegal) associations that sought better working conditions and could effectively organize a strike at any time, the Wagner Act was a compromise, protecting and regulating organized labour, that has become foundational to labour relations in both the US and Canada.
The key features of Wagner model legislation include things like:
- Explicit recognition of the exclusive representational rights of labour unions, based on majority support
- Statutory obligations on both sides to bargain in good faith
- Effective means of resolving a bargaining impasse - usually through strike or lockout
- Grievance mechanisms for disputes during a collective agreement
- Prohibitions against strikes or lockouts during the term of a collective agreement
In other words, employees are entitled to join labour unions, and employers are obligated to take those unions seriously - recognize their authority and try to reach a deal - but employers know who they have to deal with, and once they've struck a deal they can count on labour peace while that agreement is in force. They won't get workers making additional or different demands during the term, or organized labour disruptions based on those demands.
The Wagner model isn't the only model of labour relations. European countries tend to take very different approaches. But it's the one we've mostly (but, critically, not entirely) used in Canada for many decades.
So, having established that there's some collective bargaining content to s.2(d), what is the extent of that content? Is the government obligated to maintain a statutory Wagner-esque legislative scheme?
Positive Charter Obligations
Under most circumstances, Charter rights are about freedom from government interference. Freedom of religion is infringed if the government closes your church, but it doesn't require the government to build you a church. Freedom of speech is infringed if the government censors your book, but it doesn't require government to force Indigo to carry and sell it.
So the 'right to bargain' has always sounded a little odd in that respect, because bargaining is, by definition, an activity engaged in by two or more parties. Presumably, it would prevent government from prohibiting or interfering with collective bargaining, but a constitutional obligation for government to proactively require employers to participate in collective bargaining seems, at a glance, to be something more than usual constitutional interpretation rules would permit.
In my earlier post, I discussed Dunmore. This was a fairly groundbreaking case, finding that omission of a group from a statutory regime could violate s.2(d) based on its creation of a chilling effect.
I didn't discuss Fraser, a 2011 sequel where the majority of the court concluded that the obligations considered in Dunmore - that employees had a right to make and have heard representations - implicitly (and by virtue of BC Health Services) extended to a duty to bargain in good faith.
Justice Rothstein argued that this amounted to (inappropriately) constitutionalizing the Wagner model. Successive majorities of the SCC, however, have been expressly clear that that's not what they're doing.
So, to be clear on the scope of this particular debate, the question isn't whether or not s.2(d) should be understood as constitutionally entrenching Wagner; it's whether you need to understand it in that way to rationalize the conclusions reached by the majority in cases like Dunmore, BC Health Services, Fraser, and Saskatchewan Federation of Labour. In other words: Regardless of what the SCC says it's doing, has the case law nonetheless constitutionalized Wagner?
Wagner and the Charter
I maintain - and I think you'd be hard-pressed to point to authority to the contrary - that if a government chose to scrap labour relations regulation in its entirety (which used to be unthinkable, but with some of the politicians coming to power, it's hard to rule anything out), or chose to abandon Wagner altogether in favour of a functional Europe-inspired labour relations model, this would not necessarily infringe s.2(d).
But as long as they maintain the Wagner model as the guiding principle of their regulation of labour relations, derogations from Wagner principles will attract a lot of scrutiny under s.2(d). Paul-Erik Veel points out that there are no other domains where modification of rights attracts the same sort of scrutiny. He's right about that, but what he misses is that there are no other domains where fundamental freedoms are subject to such onerous and comprehensive sets of regulatory restrictions as in labour relations.
This whole discussion arose from an abstract question posed by Andrew Bernstein about how, absent jurisprudence, we would interpret s.2(d) on the face of its language. To further explore the abstraction, let's set aside the historical context of the Wagner model and the prior discussions of freedom of association I explored yesterday.
Most people responded to Bernstein's question agreeing that the right to *join* a union would likely be protected, but that collective bargaining rights are trickier. Some rejected it out-of-hand because it requires third-party engagement; others in a more nuanced way acknowledged that government interference would be problematic. Some rejected the right to strike on a similar basis, but when challenged acknowledged that their problem is really with a right to resume employment after the strike - and not the right to collectively withdraw services, in and of itself.
The audience here largely has decent enough credentials in the area of constitutional law, but relatively little in labour law specifically. So it's not surprising that the answers took this form: The first instinct, well-grounded in most constitutional jurisprudence, is that Charter rights are almost always negative - a freedom from government interference, not a right for positive intervention by the government to bolster them. The originalists among the responders (and there are some) probably don't know the context of how the phrase was used in Canadian labour relations discourse in and prior to 1982. They're intelligent people who largely know what the Wagner model is, broadly speaking, but I'd wager that most of them can't tell you much about how and why it works.
One of the most elegant answers is from labour lawyer Karen Ensslen, that the associational aspects of bargaining and striking are protected, but recognizing that it's not a right to the specific statutory forms.
Paul-Erik Veel's initial response was similar in form: It's a constitutional protection of the associational elements of forming a union, bargaining, and striking, but is NOT a constitutional protection against getting fired for doing any of those.
Let's stay on that, for a second, and ask about the impact of the Wagner model as against that kind of First Principles 'negative rights' interpretation.
The Wagner Model Infringes the Freedom of Association in Countless Ways
1. Exclusive and majoritarian representation rights
What if I were to tell you that, in a large number of cases, Wagner model legislation actually prohibits groups of employees from getting together and negotiating with the employer?
When that's the case, it's usually because you're already unionized. You may not support THAT union, and you may not agree with their bargaining goals, but that doesn't matter. That union represents your interests in bargaining despite your objections, and the agreement they enter into (again, despite your objections) is going to be binding on your employment. If you try to negotiate with the employer outside of the union - whether in association with a different group of employees or otherwise - the employer is legally prohibited from bargaining with you. If you try to get rid of that union, or bring in a different union, you can only do so in legally-designated 'open periods' - outside of those open periods, there is no way of even attempting to associate for bargaining purposes other than through the existing and recognized union.
You can even, in some cases, be required to join that union you dislike, or at least to pay dues to the union.
2. Rules surrounding bargaining
There are a range of substantive and procedural rules, applicable to both sides, like a duty to bargain in good faith, and processes surrounding notice to bargain, required disclosures, form of bargaining, statutory conciliation processes, etc.
While the rules against hard bargaining are most often interpreted on the employer side, they exist on the union side, too. If a union bargained to impasse demanding 52 weeks of paid vacation per year for all employees, they'd likely get told that they've violated their obligations.
Collective bargaining is a heavily regulated process; these are restrictions that all, prima facie, infringe 2(d).
3. Strike prohibitions
When we talk about a right to strike, in the context of Wagner models, we're usually talking about 'after the collective agreement expires and bargaining reaches an impasse'.
But all labour relations statutes in Canada actually have restrictions on when workers can strike - not just "If you strike, you can be fired", but "If you engage in a concerted work refusal, that's illegal and you can be punished or ordered back to work by the government".
These strike prohibitions during the life of the collective agreement are so commonplace, so ubiquitous, so fundamental to the compromise struck by Wagner, that we never even really talk about it...but yes, "We're unhappy with something the employer did, so we're all calling in sick tomorrow" is actually prohibited by law in pretty much every unionized workplace in Canada.
4. Deemed provisions in collective agreements
Most labour relations statutes have some provisions deeming certain terms to form part of the collective agreement - often, with no right to bargain around it - or an obligation upon a party to accept a given term if it is asked for.
This, too, is a clear and affirmative government interference with the collective bargaining process.
The list goes on, but the point is clear: While our first instinct may be to look at the Wagner model in terms of what it gives to workers, like statutory protection in forming a union and the right to a meaningful collective bargaining process, it's actually much more than that: A legislative rulebook as to how, when, and with whom that bargaining can occur.
As I told Paul-Erik Veel, once you accept that s.2(d) protects collective bargaining activities, the really extraordinary thing about Wagner model legislation is that it's constitutional at all. We'd never accept that kind of regulatory regime for speech rights, religious rights, or assembly rights. That's why there are no analogous approaches in other disciplines.
These are countless deep cuts to freedom of association, and so - on an abstract, first principles, devoid of context approach - they all need to be justified by s.1 of the Charter.
But there's no real discussion about that, and there's a simple reason for that: The Wagner model is a historical compromise, and on the whole it is a reasonable compromise. Nobody challenges the Wagner model's strike prohibitions, because we all understand that the protection against disruption of the collective agreement is a necessary trade-off for a protected process for reaching the collective agreement in the first place. Nobody challenges the closed periods for union certification applications or the bar on secondary associations participating in bargaining, because we understand it's a corollary of the protection and recognition rights afforded to unions in the first place.
The Wagner model is one big 'reasonable limit'. Adopted wholesale, there's an easy case to make that its inherent balancing act is justifiable under s.1. But once you start pulling out elements, we need to start seriously asking whether what's left preserves the meaningful right to collective bargaining that lies at the heart of the compromise. Because if that isn't sought or achieved by the scheme, then the whole scheme is constitutionally suspect.
Permissible Derogations from the Wagner Model
Perhaps the most important observation to make against an argument that we've effectively constitutionalized the Wagner model is that...we don't always follow the Wagner model.
Parts of the Canadian market, such as the Ontario construction industry, have adopted sectoral bargaining approaches that are really quite different from Wagner.
If you unionize your construction industry workplace in Ontario, odds are that a bargaining process won't immediately follow. Most of the time, that's going to mean that certain types of work are immediately and automatically subject to a Province-wide collective agreement, and other types of work are likely to be immediately and automatically subject to regional collective agreements, which were previously negotiated and agreed to on behalf of employer and union representatives.
When bargaining does happen - usually every three years - it's not actually between individual locals and individual employers, but between a team for the union's umbrella organization, and representatives of an employer association.
It's a complex hybrid model, not entirely without its reasons, and not entirely without its drawbacks. Employers are often surprised that they have these collective agreements foisted upon them. Many lawyers will be surprised that there's no ratification vote for collective agreements, so the 'open period' for changing your union might actually be quite narrow even if your bargaining unit has lost faith in the union.
Similarly, a few years ago Ontario studied the possibility of adopting a broader-based bargaining approach to provide access to collective bargaining for industries that have typically been shut out of it, like food services. While that never ended up coming to pass, there was no serious contention that it would have been paradigmatically unconstitutional, nor do I believe that the existing body of case law provides a basis for such an argument.
Conclusion on the Constitutionalization of the Wagner Model
So the argument that we've effectively constitutionally mandated the Wagner model is full of holes: The better analysis of what's going on is that we've recognized a constitutional exception that permits the Wagner model, but reliance on that exception generally requires adherence to its principles, or at least to its objectives.
The Right to Strike
More briefly, I just want to address Gerard's approach to the right to strike, because this is a question that never really arises outside the question of 'government interference'. In other words, to the best of my knowledge we've never litigated a case where the government legally permitted a strike but failed to affirmatively protect a right to return to work. (Fun fact: That regime does exist. In Ontario, strike prohibitions are applicable only to unionized workplaces, which means that 'out-of-scope' workers [i.e. workers who have not unionized in accordance with the LRA framework] may be (subject only to common law contract and tort rules) able to engage activities that might be considered a strike - without any protection against getting fired for it. If we've constitutionalized the right to reinstatement after a strike in the way suggested by Gerard, then that framework is problematic. Alberta's framework, by contrast, doesn't obviously limit its strike prohibition to unionized workers - meaning that a concerted refusal of, say, optional overtime by non-union workers would be illegal. That might well have constitutional problems to it.)
So this is unlike 'right to good faith bargaining' questions, which sometimes arise in contexts of direct government interference, and sometimes only in their exclusion from established bargaining regimes.
I think, when it comes right down to it, most of the participants in that Twitter exchange would agree that the right to collectively withdraw services is a fair extension of s.2(d), and that making strikes illegal is an infringement.
The right to resume employment is secondary - again an extrapolation of the compromise that we find in labour relations regulation. It doesn't arise in the case law for complex reasons: Even without an express prohibition, it would be usually be tough for an employer to fire lawfully striking workers.
But I think that what the skeptics need to remember here is this: That, at most times and under most circumstances, unionized employees are, in fact, legally prohibited from striking. That's what the Wagner model does. It prohibits strikes to provide a useful and protected mechanism of trying to reach a deal, first. And when it does authorize strikes, that is still with the goal of furthering those useful and protected mechanisms of trying to reach a deal.
The right to go back to work at the end of the strike can't be seen as a free-standing operation of s.2(d). But it can be seen as a necessary extension of a complex and invasive statutory regime that carefully regulates striking and bargaining activities to promote deal-making and labour peace - a necessary part of the compromise that justifies the rest of the infringements.
The Remedial Question
Veel posed a question about the appropriate remedy: Even accepting my analysis, that freedom of association extends to positive assertions of Wagner model rights because they're necessary to justify the various limitations that Wagner puts on the freedom of association, wouldn't the fix be to get rid of those Wagner model limitations on rights (i.e. prohibit the now-unjustified breach), as opposed to requiring government to extend additional rights?
So, first off...most of the cases we look at here don't actually deal with positive assertions of rights. SFOL dealt with a strike prohibition. BC Health dealt with legislation that overrode collective agreements. When we talk about Bill 28, that's a legislatively-imposed collective agreement.
Even in Fraser, the remedial question doesn't arise: The majority came to the conclusion that the statute, as written, could (and therefore should) be interpreted as conferring the necessary rights.
Basically, we're talking about Dunmore - where the court basically concluded that when the legislature extended union recognition rights to everyone except agricultural workers, that amounted to an interference with agricultural workers' rights.
And yes, when we say "the problem is the exclusion", in a very conventional sense of constitutional interpretation - in that the constitution limits what governments can do, instead of mandating affirmative acts - we're actually saying that the problem is the recognition they are extending to everyone else. So, remedially, the most obvious fix is to stop them from extending that recognition to everyone else.
But whether or not it's the only available remedy for the court is arguable, and at minimum it's clearly an absurd remedy: Deleting the entire labour relations framework because it was underinclusive? Ridiculous, disruptive, and outrageous.
In Vriend v. Alberta, there was some discussion about the court's ability to 'read in' terms to bring a statute into compliance, where there was clearly a deliberate legislative intent otherwise. That was a case about Alberta having - expressly and deliberately - excluded sexual orientation as a prohibited ground under its human rights legislation. The majority of the court concluded that the appropriate resolution under the circumstances was to read in. They relied on language from a decision by the Ontario Court of Appeal in a similar case dealing with the Federal Act, calling it "inconceivable...that Parliament would have preferred no human rights Act over one that included sexual orientation as a prohibited ground of discrimination. To believe otherwise would be a gratuitous insult to Parliament."
I tend to agree with the logic. Realistically, there were two options: Effectively (or actually) get rid of the whole human rights statute, or fix it. This wasn't a case where the legislature had a wide range of choices in terms of how to fix the deficiency. And frankly, if they disagreed with the court's chosen remedy, they could legislate a different one.
The same applies in Dunmore: It's simply an absurdity to think that the Ontario legislature would have fixed the underinclusion problem by repealing the Labour Relations Act altogether.
*****
This post does not contain legal advice, but only general legal information. It does not create a solicitor-client relationship with any readers. If you have a legal issue or potential issue, please consult a lawyer.
Comments
Post a Comment