Canadian Appeals Monitor Information and Commentary on Upcoming and Recent Appeal Court Decisions

Follow My Lead: The SCC Rules on the Role of Its Own Precedents

Posted in Case Comments, Tax
Brendan Owen Brammall

Are lower courts obliged to follow precedents of the Supreme Court of Canada, even if those precedents are decades-old and have been the subject of extensive criticism? And when should the Court overrule one of its own precedents? The Court recently addressed both of these questions in Canada v. Craig, a tax case involving the deductibility of losses from a lawyer’s horse racing business. Notably, in this case the Court overruled a precedent from 1977, on the basis that there were compelling reasons indicating that the precedent was wrongly decided. At the same time, the Court held that the lower courts should have applied this precedent and were not at liberty to overrule it themselves.


In 2000 and 2001, the taxation years at issue, Mr. Craig was a partner at a downtown Toronto law firm and derived most of his income from his legal practice. He also had significant capital gains on equity investments. In addition, he had been actively involved for many years in the standardbred horse racing industry and operated a horse racing business on a year-round basis, with as many as 20 horses in 2000 and 14 horses in 2001. He would devote more than 600 hours annually to this business, which incurred losses of $222,642 in 2000 and $205,655 in 2001. This case arose when Mr. Craig sought to deduct these losses from his other income.

The deductibility of these losses turned on the interpretation of section 31(1) of the Income Tax Act (the “Act”), which applies “[w]here a taxpayer’s chief source of income for a taxation year is neither farming nor a combination of farming and some other source of income …” (emphasis added) Where section 31(1) applies, the maximum allowable deduction for a farming loss is $8,750. Section 248(1) of the Act provides that “farming” includes “maintaining of horses for racing.” The Minister of National Revenue took the position that section 31(1) applied to Mr. Craig and therefore allowed him only the maximum deduction of $8,750, which meant that most of his horse racing losses for these taxation years were not deductible.

The correctness of the Minister’s position depended on whether the Supreme Court of Canada’s 1977 decision in Moldowan v. The Queen remained good law. In that case, Dickson J. (as he then was) interpreted section 13(1) (now section 31(1)) to apply to “the taxpayer who does not look to farming, or to farming and some subordinate source of income, for his livelihood but carries on farming as a sideline business.” (p. 487; emphasis added) This interpretation effectively read in the word “subordinate” before “other source of income,” such that a taxpayer would be excepted from the deduction limitation of $8,750 only if farming constituted the taxpayer’s chief source of income. Moldowan was subsequently criticized for having thus made the second exception in section 31(1) (“a combination of farming and some other source of income”) redundant, as any taxpayer whose chief source of income was farming would be excepted from the deduction limitation by virtue of the first exception (“farming”).

This criticism culminated in the Federal Court of Appeal’s 2006 decision in Gunn v. Canada, which adopted a “more generous” approach to the question of whether a taxpayer’s chief source of income is a “combination of farming and some other source of income”:

In my view, the combination question should be interpreted to require only an examination of the cumulative effect of the aggregate of the capital invested in farming and a second source of income, the aggregate of the income derived from farming and a second source of income, and the aggregate of the time spent on farming and on the second source of income, considered in the light of the taxpayer’s ordinary mode of living, farming history, and future intentions and expectations. This would avoid the judge-made test that requires farming to be the predominant element in the combination of farming with the second source of income, which in my view is a test that cannot stand with subsequent jurisprudence. (emphasis added; para. 83)

Decisions Below

The Tax Court of Canada

As the Minister conceded at trial that Mr. Craig’s horse racing operation was a “source of income” (and therefore not a mere personal endeavour), the case focussed on the applicability of the deduction limitation in section 31(1). Hershfield J. held that this limitation did not apply to Mr. Craig, on the basis of the more generous interpretation in Gunn.

In assessing the authority of Gunn in light of Moldowan, Hershfield J. noted that Dickson J. (in Moldowan) also considered the meaning of the term “chief source” of income. In particular, Dickson J. held:

The distinguishing features of “chief source” are the taxpayer’s reasonable expectation of income from his various revenue sources and his ordinary mode and habit of work. These may be tested by considering, inter alia in relation to a source of income, the time spent, the capital committed, the profitability both actual and potential. (Molowan, p. 486)

Given that these “distinguishing features” (reasonable expectations, mode of work, time, capital, and profitability) are similar to the factors that Gunn applies to the section 31(1) combination question (as set out in the above excerpt), Hershfield J. sought to reconcile Gunn and Moldowan:

Indeed, it is my view that Gunn simply puts the Moldowan analysis back on track as a workable construction of section 31 – one that does not render it sterile while paying heed to the language of the section. The Gunn analysis uses the personal and commercial commitment factors embraced in Moldowan in a manner that fits into the combination formula stipulated in section 31. (para. 61; emphasis added)

Accordingly, Hershfield J. held that Gunn did not violate the underlying principles in Moldowan, and he formulated the applicable legal test as follows:

I am suggesting then that the test is whether the taxpayer’s mode of operation has sufficient commitment and commerciality and profit potential to be recognized as a chief source applying the Moldowan commitment and profitability criteria. Looking at time spent, capital invested, and a meaningful profit potential arising from a dedication to profitability, the question of whether the taxpayer is recognizable as a committed, viable commercial player in a genuine economic sector of the economy should be readily answered. (para. 72; emphasis in original)

Hershfield J. concluded that Mr. Craig satisfied this test and that the deductibility of his horse racing losses was therefore not limited by the $8,750 ceiling in section 31(1).

Federal Court of Appeal

Dismissing the Minister’s appeal, Evans J.A., on behalf of a unanimous Federal Court of Appeal, held that there was no basis on which to depart from Gunn. Evans J.A. noted, firstly, that the Federal Court of Appeal normally follows its own precedent unless that precedent was made per incuriam, without regard to a decision that it ought to have followed. As Gunn considered Moldowan at length, Evans J.A. wrote that “Gunn was thus anything but a per incuriamdecision.” (para. 14)

Evans J.A. provided other reasons for following Gunn, including:

Since Moldowan was decided before Gunn, we are not at liberty to depart from Gunn on the ground that it has been overruled by the Supreme Court of Canada.

… a decision by a panel of this Court on the precedential effect of a prior decision by the Supreme Court of Canada [i.e., Moldowan] deserves as much respect from a subsequent panel of this Court as a decision by a previous panel on any other question of law. …

… in view of subsequent decisions from the Supreme Court of Canada on the interpretation of taxation statutes, the aspect of Moldowan in question here might be decided differently today.

judge-made rules relating to precedent are not like other legal rules, in the sense that the Supreme Court of Canada does not reverse the decision of an intermediate appellate court on the ground that it failed to follow the principle of stare decisis. Rather, when the Supreme Court grants leave to appeal, the question before the Court will be whether the lower court’s decision is consistent with substantive law, including extant decisions of the Supreme Court, or whether the Supreme Court should modify its own jurisprudence on the point. (paras. 16-18, 20; emphasis added)

Having concluded that Hershfield J. did not err in law in relying on Gunn, Evans J.A. then held that Hershfield J. did not misapply Gunn on the facts of this case:

… [Hershfield J.] concluded on the basis of the factors and analytical framework set out in Gunn that farming constituted a significant part of Mr Craig’s income in 2000 and 2001, and was more than a “sideline business”. Hence, section 31 did not apply.

The facts of this case may be fairly close to the line, and I might not have made the same decision as the Judge. However, I cannot say that his decision is based on a palpable and overriding error warranting the intervention of this Court. (paras. 25-26)

The Supreme Court of Canada’s Decision

Writing on behalf of a unanimous Supreme Court of Canada, Rothstein J. dismissed the Minister’s appeal. There were, in essence, three aspects to Rothstein J.’s decision.

Firstly, Rothstein J. stated that the courts below should have followed Moldowan. He characterized Moldowan and Gunn as “inconsistent precedents” (para. 19). And though he acknowledged that Moldowan had been the subject of “extensive criticism” (para. 20), Rothstein J. wrote that the courts below were not at liberty to depart from it:

But regardless of the explanation, what the court in this case ought to have done was to have written reasons as to why Moldowan was problematic, in the way that the reasons in Gunn did, rather than purporting to overrule it.

… the question of whether the Federal Court of Appeal should have followed Gunn simply did not arise, in view of the Moldowan Supreme Court precedent. (paras. 21-22)

Secondly, and notwithstanding the fact that the courts below were required to follow Moldowan, Rothstein J. held that it was appropriate for the Supreme Court to overrule that decision. He noted that the Supreme Court does not lightly overrule its own precedents, and that, in doing so:

… the Supreme Court engages in a balancing exercise between the two important values of correctness and certainty. The Court must ask whether it is preferable to adhere to an incorrect precedent to maintain certainty, or to correct the error. (para. 27)

In this case, Rothstein J. concluded that correctness weighed more heavily in the balance, and that Moldowan was incorrect as it “essentially read the combination test out of s. 31(1)” (para. 28). This was contrary to the words used by Parliament, which had created both a “farming” exception and a “combination” exception to the application of that provision.

Thirdly, Rothstein J. held that the “combination” exception should be interpreted as proposed in Gunn, such that farming itself need not be the taxpayer’s predominant source of income:

The factors identified [in Gunn], namely, the capital invested in farming and the second source of income, the income from each of the two sources of income, the time spent on the two sources of income, and the taxpayer’s ordinary mode of living, farming history, and future intentions and expectations, are all factors involved in running a farming business together with another source of income. If these factors tend to show that the taxpayer places significant emphasis on both his farming and non-farming sources of income, there is no reason that such a combination should not constitute a chief source of income, avoiding the application of the loss deduction limitation of s. 31(1). The determination is a factual one for the trial judge. (para. 42; emphasis added)

Having adopted this interpretation of section 31(1), Rothstein J. held that there was no basis on which to disturb Hershfield J.’s factual conclusion that Mr. Craig’s horse-racing operation was more than a sideline business and that the loss deduction limitation was therefore not applicable.

Potential Significance

The Supreme Court’s decision is perhaps most significant because it confirms that lower courts are required to follow its precedents, even if the precedent in question is relatively old and has been the subject of widespread judicial and academic criticism. The decision is also significant because it sets out when the Supreme Court will overrule one of its own precedents, and because it settles the interpretation of a frequently-litigated provision of the Act.

Case Information

Canada v. Craig, 2012 SCC 43

Date Decided: August 1, 2012

SCC Docket: 34144