The Supreme Court of Canada yesterday gave judgment in Canada (Attorney General) v. Collins Family Trust, 2022 SCC 26, including what seems to me a rather unfortunate treatment of the UKSC decision in Pitt v Holt. The context is that lower courts in Canada had in recent years taken a very wide view of rectification, allowing it to be used to conduct what the Court calls ‘retroactive tax planning’: that is, where there were unforeseen tax consequences. This was stopped in a pair of 2016 decisions, from Ontario and Quebec (thus covering both of Canada’s principal legal systems).

It is of course striking to see a common law trust named as a litigant. The only explanation for this is that trusts (and some other non-persons) are deemed to be persons for the purposes of the federal Income Tax Act, which deeming seems inevitably to ‘leak’ outside of the Act and into civil procedure (at least).

In Collins Family Trust, the attempt was to use rescission to undo a transaction that had unforeseen taxation consequences. The claim was arguably strengthened, at least in one dimension, by the consideration that the federal tax authorities had adopted and shared the same understanding of the Income Tax Act on which the transaction was based, but which was later rejected by a court. In earlier litigation in British Columbia, the CA had applied Pitt v Holt to allow rescission in a similar case to Collins Family Trust, and so in this case, which also came from BC, the courts below had followed that.

The Supreme Court of Canada by an 8-1 majority allowed the appeal. Some of the holdings seem very sound: that a party cannot set aside a transaction simply because it had unforeseen taxation consequences (eg [16]). Thus, this should no more be possible using rescission than using rectification. This, however, is quite consistent with Pitt v Holt I think.

What is missing from the majority a general theory of what rescission is, and when it is rightly available. I hear my doctoral supervisor spinning in his grave when I read ([11]): “Generally speaking, a court of equity may grant relief where it would be unconscionable or unfair to allow the common law to operate in favour of the party seeking enforcement of the transaction.”

What seems to have happened is a desire to say, “you should no more be able to change your mind using rescission than using rectification” (see eg [19]-[23]); but surely this calls for an explanation of when rescission is available, so as to explain its differences from rectification. We should have a test for when it is available.

The treatment of Pitt v Holt at [24]-[27] is cursory. The majority suggests [25] that our law is different from UK law inasmuch as in Canada, the Minister is obliged to apply the Income Tax Act. There seems to have been no understanding that Pitt v Holt, rather like Collins Family Trust, was part of a series of developments aiming to articulate when exactly rescission is available, and to restrict its availability in a principled way.

In my personal view, the judgment in Pitt v Holt is not well written and may be (being a single judgment of an extraordinary seven-judge panel) what is sometimes called a ‘committee job’. It is strangely structured, and it is not easy to identify the crucial points in the reasoning. But I think it is usually understood as having overruled a line of cases that were too generous about rescission. Part of it holds that rescission is only available in relation to the exercise of a fiduciary power when there was a breach of ‘duty’ by the fiduciary—‘breach of duty’ in the sense of an absence of the elements which are necessary for the exercise of the power to be unimpeachable (and not ‘breach of duty’ in the sense of violation of an existing right). I think the discussion in Pitt v Holt of ‘improper purpose’ as being distinct from ‘inadequate deliberation’ merely identifies two possible ways in which a power can be exercised without complying with all the requirements for its unimpeachable exercise (see Hart on power-conferring rules). Part of Pitt v Holt is about rescission for mistake and insists upon a causative mistake of sufficient seriousness. This seems to be the part that the Supreme Court of Canada has deprecated, although there is no discussion of the fiduciary part of the UKSC holding.

What is arguably unfortunate about this for Canadian law is that the Supreme Court of Canada has sideswiped Pitt v Holt—at least the part about mistake—seemingly without having understood its import. The UKSC case did not simply say that unfairness was needed, but that what was needed was [122] “a causative mistake of sufficient gravity; and, as additional guidance to judges in finding and evaluating the facts of any particular case, that the test will normally be satisfied only when there is a mistake either as to the legal character or nature of a transaction, or as to some matter of fact or law which is basic to the transaction.” This can be seen as an attempt to provide a test that is at least a little clearer than “where it would be unconscionable or unfair”. Pitt’s [122] is quoted in Collins Family Trust at [24] but it is said to be inapplicable in Canadian law: “This divergence is unsurprising, given that English law lacks the prohibition against retroactive tax planning stated in Fairmont Hotels and Jean Coutu, and operates under a different legislative framework.” This is puzzling, as the whole point of Pitt was to rein in retroactive tax planning and to say that rescission is available to trustees only where it would be available on general (not statutory) principles, and to try to clarify what those general principles require.

For me personally, the failure to even mention the fiduciary part of Pitt v Holt while summarily dismissing what it says about mistake is unfortunate, as we are left with little indication of what is the Canadian law for the rescission of fiduciary powers where, for example, the power-holder has considered irrelevant matters in breach of duty. The powers exercised in Collins Family Trust were not fiduciary powers. Is that part of Pitt still arguably relevant in Canada? Moreover, what does the new holding mean for rescission that does not involve tax consequences? Is Pitt also rejected for that?

Côté J dissents and says that an unavailability of rectification does not entail an unavailability of rescission; that rescission must be based on mistake and not just misprediction; that rescission should require a mistake of sufficient seriousness; and that these tests were met in this case.
For those who like such things, the relevant transactions were entered into in 2008 and the court ruling that showed that everyone, including the tax authorities, had the law wrong was issued in 2012. Thus the relevant mistake for Côté J was one that affected the [65] “general understanding in the tax community” before that ruling.

To modify a maxim: do tax cases make bad non-tax law?

Lionel