If there is one aspect of Canadian administrative law that is relatively understudied, it is the constitutionality (or, less ambitiously, the desirability) of the delegation of legislative power from Parliament to the Cabinet and administrative actors. The recent Saskatchewan Carbon Tax Reference puts into stark relief the underdeveloped nature of the law in this area, and the stakes underlying the issue of delegated legislative power. Despite a patina of agreement that there is no non-delegation doctrine in Canada, there are some underlying disagreements that the Saskatchewan case presents as to when delegation could go too far in a system of parliamentary sovereignty. Specifically, and in my view, there is no delegation issue in this case, as the majority held; no matter the scope of delegated power, Parliament is free to delegate. But there could be delegation issues in future cases, based not on the scope of delegated power, but based on who can receive the power.
First things first. While the Saskatchewan case was largely built up as a division of powers case, lurking underneath the surface were arguments about the delegation of legislative power. This is because of the way in which the “fuel charge” (which the challengers argued was a “tax”) was to be implemented under the statute left a great deal of discretion to the executive. Indeed, the Court said [36]:
Part 1 implements a fuel charge. It applies to 22 GHG-producing fuels listed in Schedule 2 of the Act. These include such things as gasoline, kerosene, fuel oil, propane, coke oven gas, and methanol. The Governor in Council may add to, or delete from, this list (s 3, definition of “fuel”; s 17(1)).
Of course, controversially, the Cabinet also has discretion to delineate the circumstances under which the impugned statute would apply, based on whether provinces adopted similarly stringent carbon pricing mechanisms [48].
The two parts of the statute at issue here present different putative delegation problems. First is the narrow issue of taxation and delegation, and the special rules around which that issue arises in our constitutional system. Second is the broader context of delegation, including the legislation’s use of a so-called “Henry VIII” clause to permit the Cabinet to amend the statute without going through the legislative proves.
On the first issue, the reason there has never been a serious question raised as to the constitutionality of delegation in Canada is because of the heritage of parliamentary sovereignty. The seminal cases on the matter all insinuate that so long as Parliament does not “abdicate” its legislative power, there is no constitutional problem. No attempt has ever been offered to determine what abdication might mean (though, I am currently working on a paper attempting to answer this question).
So, take the seminal case of Gray, which Professor Daly discusses here in the context of Henry VIII clauses. Gray took place during the context of World War I, where Parliament delegated broad power to the Cabinet to
[d]o and authorize such acts and things, and to make from time to time such orders and regulations, as he may be reason of the existence of real or apprehended war, invasion or insurrection, deem necessary or advisable for the security, defence, peace, order and welfare of Canada…
Clearly, a broad delegated power. But what’s more, the delegation also contained a Henry VIII clause. To the Court, this was no matter. So long as Parliament did not “abdicate its functions” and did not disturb “the existing balance of constitutional authority by aggrandizing the prerogative at the expense of the legislature” [157], there was no constitutional problem. This basic holding has been repeated many times: Chemicals Reference, Furtney, Pan-Canadian Securities Reference.
Fast forward to the Saskatchewan case, where this background arose in the context of a tax. The power to tax in Canada cannot be delegated, because of s.53 of the Constitution Act, 1867—which says that all taxes must originate in the House of Commons. In Re Eurig Estate, the Court rooted this textual command in the age-old idea of “no taxation without representation.” This issue does not arise, though, if the tax is in reality just a regulatory charge. The Court drew this conclusion in the Saskatchewan case. But it went on to note, in partial reliance on Supreme Court precedent requiring that taxes do not arise “incidentally in delegated legislation” (Eurig Estate, at para 32), that the power of the Governor-General to decide when the exaction applied (para 105) and the breadth of the power itself (para 106) was limited.
As the Court noted, this analysis was not strictly necessary—and in my view, neither are the Supreme Court precedents requiring a special clear statement rule to delegate “limited” power to fill in details of a tax (see Eurig, at para 30). Once a court has determined that the exaction is not a tax, but a regulatory charge, in a system of parliamentary sovereignty (where no division of powers or Charter issues are raised with respect to the delegation itself), there is no warrant for courts to attach limits on the scope of the delegation as such. Accordingly, the majority’s analysis here was not at all necessary.
The more interesting issue in the case—one which clearly divided the majority and the dissent—was the issue of Henry VIII clauses. The authority of Gray establishes Henry VIII clauses as constitutionally permissible. But other cases show that they are, as Professor Daly notes, constitutionally “exceptionable”: see Ontario Public School Boards Association v Ontario (Attorney General) at para 51. The dissent in the case jumped on this intuition, objecting to the Henry VIII clause in the legislation as repugnant to the principle of parliamentary sovereignty, which the Court most recently endorsed as relevant and powerful in the Pan-Canadian Securities Reference to uphold the cooperative federalism scheme to regulate securities (see para 58). Piggybacking on the principle of parliamentary sovereignty, the dissent noted that (at para 372):
In fact, even if Part 1 were seen to be a regulatory levy, we would question the constitutionality of the delegation of law-making power under it…because [the provision] sets forth a wholly-unfettered grant of broad discretion to amend Part 1 of the Act that…may trump any exercise of legislative power by Parliament…This would seem unconstitutional, and contrary to the Westminster model, regardless of where the so-delegated law-making power lies under the Constitution Act, 1867.
With respect, this seems backwards. It is the case, of course, that Henry VIII clauses put the legislative onus in the executive rather than Parliament, and for that reason, may be suspect. But this is just a special case of the general rule that, in a system of parliamentary sovereignty, Parliament can “make or unmake any law whatever,” subject to express constitutional limits (as Dicey noted). This is the basic holding of Gray and the other cases endorsing broad powers to delegate as an analogue of parliamentary sovereignty. So, to put it differently, parliamentary sovereignty does not cut in the direction of limiting delegation, even Henry VIII clauses. Rather, it cuts in the direction of permitting broad delegated power, so long as it does not constitute impermissible interdelegation (see the NS Interdelegation Case) or runs afoul of Charter rights. This is true no matter one’s view of the merits in the Saskatchewan case.
It should be noted that the dissent’s analysis arguably missed the point of the Pan-Canadian Securities Reference, which it cited to support its view of parliamentary sovereignty. In fact, the Pan-Canadian Securities Reference presents the opposite conclusion that the dissent makes in the Saskatchewan case. At issue in that case, as noted above, was a cooperative federalism scheme to regulate securities. The scheme included model provincial statutes and a federal statute, and it established a new body called the Council of Ministers along with a regulator. The Council of Ministers is composed of responsible ministers in the provinces and the federal government. The Council has power to recommend statutory amendments to the model statutes and the federal statutes, but ultimately those amendments must be approved by the responsible legislatures. And the regulator had the delegated power to make regulations, but those regulations had to be approved by the Council of Ministers.
The Attorneys General of Quebec and Alberta challenged the scheme as a legislative body unknown to the constitutional structure—put differently, an unconstitutional delegation. But the Court rebuffed this argument, based on the idea of parliamentary sovereignty: so long as all of the legislative acts were subject to parliamentary control, there was no concern (see paras 77-78). The Court also held that the regulation-making power was not problematic from a perspective of parliamentary sovereignty: “…this form of delegation of administrative power is entirely consistent with the principle of parliamentary sovereignty, since the delegated authority can always be revoked by the sovereign legislature” (para 123).
The dissent may have tried to make it’s point stronger by saying that the case presented a rare case of “abdication,” adverted to in Gray. But when we are dealing with a delegation from Parliament to the executive, this is a hard argument to make. This is because Gray says that Parliament “could not abandon any of its own legislative jurisdiction (at 170, my emphasis). Of course, this makes sense. If, as the Pan-Canadian Securities Reference says, parliamentary sovereignty includes an element of control over the executive, then Parliament always has control over the Cabinet through the principle of responsible government. Put differently, it is hard to imagine a situation in which Parliament could ever withdraw control in a responsible government sense.
Indeed, I can’t see a non-abdication principle working in the context of the Saskatchewan Reference because here the executive is accountable to Parliament and Parliament exercises a natural mechanism of control. Instead, I think the non-abdication principle has much more force in the context of independent agencies. In that context, Ministers are merely answerable rather than accountable for the agencies, and there is a degree of independence built into the constating statutes or garnered through conventions of independence. In such a case, responsible government is not as powerful, because the Minister can merely point to the agency as “independent” (the McClintic case is a good example of this). Taken too far, Parliament might try to create a new Parliament, separated from parliamentary control. This is probably explicitly prohibited by s.17 of the Constitution Act, 1867 which says there shall be “one” Parliament. It also runs counter to cases, like the Re Initiative and Referendum Act case, where the Manitoba legislature’s effort to create a citizen initiative process–which would bypass legislative reading and Royal Assent–was rebuffed by the Judicial Committee of the Privy Council.
Indeed, for the Court in the Pan-Canadian Securities Reference, the only reason the Council of Ministers and the regulator were permissible was because Parliament exercised a sufficient degree of control over them. While one can quibble with this holding in the context of the specifics–as I do because, to my mind, there is insufficient parliamentary control over the process of regulation-making by the regulator and the Council of Ministers–there is still the basic idea. Parliament cannot abdicate its power by withdrawing control, and this is most likely to happen in the context of independent agencies.
But this is a bit of a tangent, because in the Saskatchewan Reference, none of these interpretive difficulties arise. The only real issue is the scope of the delegation, not the recipient; in my view, “non-abdication” as a constitutional rule only applies to the latter situation. That said, and all told, the Saskatchewan Reference raises important questions. What does it mean for Parliament to abdicate its legislative power? Could Henry VIII clauses ever constitute “abdication”? As I noted above, I am going to try to tackle these questions in an upcoming paper.