I’ve been meaning to blog about an interesting Charter case that went to the Supreme Court a few years ago, Dunmore v. Ontario, which challenged the exclusion of agricultural workers from the Ontario Labour Relations Act. Until the NDP came to power in the early 1990s, agricultural workers had essentially been excluded from the scope of the LRA, which is the framework governing unionized labour in Ontario. The result of this of course was that agricultural workers couldn’t effectively unionize, much less strike. The NDP passed legislation to include agricultural workers in the LRA, which was promptly reversed by the Tories when they came to power in the mid-1990s.

The union (or one of the unions) representing agricultural workers mounted a Charter challenge to this exclusion. The challenge got nowhere at trial and in the Court of Appeal, but it ultimately succeeded at the Supreme Court of Canada. The union set out two arguments: (1) that the exclusion violated freedom of association (s. 2(d)) and (2) that the exclusion of agricultural workers from a regime that included most other workers violates the equality provision in s. 15(1) of the Charter.

The union won at the SCC under s. 2(d). Unlike the lower courts, the SCC was willing to extend s. 2(d) to include some requirement that the government act positively (i.e., pass legislation) to protect freedom of association rights. This however creates a problem: once you say that the Charter creates a positive obligation for the government to act, the question immediately arises, what in particular does the Charter require that the government do? And this becomes a difficult question because in general the courts are hesitant to constitutionally entrench any particular legislative scheme–the courts are much more comfortable telling governments they’re not allowed to do something (because it violates a right) than that they must as a matter of constitutional law actively legislate in a particular way. This problem is especially acute in the context of labour law since in general the courts have been unwilling to use the Charter to interfere with legislated balances of power between union and management. In the event, the Court set out a certain minimal baseline, which protects agricultural worker organization but doesn’t give them many rights beyond that. These minimal requirements were duly enacted and have not been changed since.

Another approach the Court could have taken (rejected by all but one of the nine justices) would have been to say that the exclusion of agricultural workers violated s. 15 by denying agricultural workers “equal protection and equal benefit of the law” (quoting s. 15). On this analysis the question becomes: If, in general, employees enjoy the right to unionize, why shouldn’t agricultural workers? The Court rejected this approach because, since early in the Charter’s history, they have rejected a general “equal protection” approach in favour of one based on specific grounds of discrimination. The government doesn’t have to justify unequal treatment, as long as unequal treatment is not based on specific prohibited grounds, e.g., race, sex, sexual orientation, etc. The Court wasn’t willing to find that occupation or occupational status was one of these grounds, and so the equality argument failed.

The basic idea behind a “prohibited grounds” approach to s. 15 is understandable: you don’t want the government to be dragged into court to justify itself every time it makes a distinction between two groups of people. And in general, you don’t want occupational status to be one of these prohibited grounds. Usually, you’d want the government to be able to distinguish between (say) legal powers and obligations of doctors vs. nurses, lawyers vs. para-legals, or for that matter doctors vs. lawyers, without being liable to a court challenge for every distinction it makes between these two groups of people. But on the other hand you would think the disadvantaged status usually associated with some occupations (e.g., as here, agricultural workers) would be the kind of thing that ought to engage an equality rights analysis. Given that agricultural workers are arguably among those workers most in need of the benefits of unionization, you might expect the government to have to justify why these workers in particular are altogether deprived of these benefits. Another advantage of the equality analysis is that on the equality analysis the courts wouldn’t be telling the government to legislate specific benefits for agricultural workers: it would just be telling it to extend whatever benefits it gives to other classes of workers to this class of workers, or alternatively justify its decision not to do so.
It seems to me that Dunmore ought to be as much an equality case as it is a general freedom of association case. But explaining how and why this can be the case is not an easy matter.