R v Comeau: Who decides history?

By now, the Supreme Court of Canada’s boozy federalism decision in R v Comeau is old news. And – no doubt – many important things can be said about Comeau and cooperative federalism, originalism, and precedent. My interest, however (and not surprisingly), is in the expert evidence issues it contains. Most notably, Comeau raises important issues about the factual determination of history in courtrooms and the roles of judges and expert witnesses in that task. In Comeau, I think these issues could have been handled a lot better.

Background: The Facts and Dispute (see paras 1-21 of the SCC decision)

Mr. Comeau, a New Brunswicker, got caught in a police sting. He drove across the border into Quebec and bought a great deal of beer and hard liquor. On his way back, the RCMP stopped him, found the putative contraband, and fined him under subsection 134(b) of the New Brunswick Liquor Control Act (the “Act”). This provision, along with others in the Act, prevent New Brunswick residents from importing alcohol from other provinces.

Unlike many people that get fined for such behaviour, Comeau got a bit creative – he hired an expert (a historian, Andrew D. Smith). Smith provided evidence that ultimately convinced a trial judge that section 134(b) is unconstitutional. More specifically, Comeau successfully argued that 134(b) infringed section 121 of the Constitution Act, which reads: “All Articles of the Growth, Produce, or Manufacture of any one of the Provinces shall, from and after the Union, be admitted free into each of the other Provinces.”

In demonstrating that section 134(b) was unconstitutional, Comeau faced an uphill battle. This was because previous Supreme Court and Privy Council case law read 121 restrictively, holding that it only banned tariffs and tariff-like burdens. Dr. Smith disagreed, drawing on evidence of the political climate and speeches from the Canadian founding fathers. Smith interpreted this evidence to mean that the drafters of 121 intended that section as a broad free trade clause. Importantly, not only did Smith provide this framework evidence, he also provided an interpretation of 121 in light of that background (Comeau SCC at para 40).

Dr. Smith, by all accounts, succeeded. The judge accepted his evidence and ruled subsection 134(b) unconstitutional. The Crown appealed the decision up to the Supreme Court.

The Supreme Court Decision

A full panel of the Supreme Court issued a unanimous decision, holding that the trial judge erred for two reasons. First, the evidence did not justify departing from existing Supreme Court precedent. Deviating from such precedent requires that the evidence “fundamentally shift the parameters of the debate” (Comeau SCC at para 31, citing Bedford). This could take the form of the “evolution in the foundational legislative and social facts” relevant to the dispute. A new interpretation of history doesn’t meet this threshold (Comeau SCC at para 37).

Second, on proper interpretation of the Constitution and the Act, the liquor law was constitutional. The Court gave particular consideration to the history of the section 121. It found ample evidence that the founders were concerned with tariffs, but only “limited support” that they contemplated abolishment of all barriers to trade (the expert’s position, see Comeau SCC at para 67). This finding, among others, narrowed the ambit of 121 to prohibiting laws that impacted interprovincial movement of goods (e.g., tariffs) and which did so as their primary purpose (para 114). The impugned law did restrict trade but not as its primary purpose, which was to regulate alcohol in New Brunswick with an eye towards supervising its production, use, and sale (Comeau SCC at paras 122-124). The law was constitutional.

The Expert Evidence

The defence’s witness was Dr. Andrew D. Smith [bio], who appears to be a senior lecturer at the University of Liverpool (with a PhD in history from Western; CV). The trial judge described Smith as “one of the world’s most renowned experts on the constitutional moment” with “unimpeachable” credentials. Moreover, his testimony was “beyond reproach” (Comeau Trial at para 52). The trial judge relied on Dr. Smith’s testimony at several points during his analysis of section 121. I will survey just a few.

As to the text of section 121, Smith contrasted its final wording with that of an original draft that individually listed the provinces and used the term “admitted free into all Ports in Canada” (Comeau Trial at para 57). To Smith and the trial judge, removing the individual province names and seemingly broadening the scope from ports to generally admitting trade into the provinces suggested the founders had expansion on their minds. And this, in turn, suggested expansion in terms of unfettered free trade (this seemed like a bit of a leap to me). Beyond the changes in the drafts, Smith noted that “admitted free from duty” was a principle the founders were aware of and thus was significant that they did not use that term. Contextually, Smith also opined on the importance of section 121’s position among other economic provisions (Comeau Trial at para 66). He suggested this reinforced its meaning as a broad free trade protection. The trial judge did not find this final argument as convincing (Comeau Trial at para 68).

Dr. Smith, naturally, also provided a great deal of evidence about the historical context of section 121, most of it supporting a broad interpretation (Comeau Trial at paras 70-101). The first arm of this evidence was that the drafting of the Constitution came on the heels of a trade treaty with the United States and during a time when the British were moving towards free trade. As a result, Dr. Smith believed – and the trial judge accepted – that the drafters were influenced by British practices and sought intra-Canadian free trade as a means to offset lost trade with the United States. The second arm – a selection of speeches from around the time of the drafting of the Constitution – generally corroborated the first. More precisely, some founding fathers expressed that free trade in any forthcoming union would promote a stronger economy in that union.

The Crown tendered Dr. Tom Bateman, an associate professor in political science at St. Thomas University (at paras 150-165). The description of Bateman’s testimony fell under the heading “The Nature of the Canadian Constitution”. Generally, this seemed to include a discussion of how section 121 has been interpreted from a separation of powers standpoint and the public policy issues related to that. For instance, he noted that the provinces do many things that may impact trade (e.g., issue professional licenses, develop product standards). Accordingly, the restrictive reading of 121 that had prevailed until then was a “workable way to reconcile” free trade with provincial powers under section 92 of the Constitution Act.

As I said above, the trial appeared to find Dr. Smith’s evidence quite convincing and ultimately found for Comeau. The relevant provisions of Act were unconstitutional because section 121 protected free intra-Canadian trade, broadly construed.

What is going on here?

Comeau is fascinating from a fact-finding perspective. In particular, it raises some serious questions about how courts should make findings of historical facts and incorporate those facts into their decisions. And it what role should experts play in that process? Although the scope for judicial fact-finding for “legislative facts” (i.e., facts that go beyond the dispute itself) is wider than that for other facts, it’s not clear to me that judges (trial judges especially) are well-equipped to do historical research. On the other hand, experts bring to court with them a host of biases. For instance, they may have theories they have grown quite attached to. And they also labour under “adversarial bias”, the natural inclination to do good things for the person paying for your services. In Comeau, I think that the trial judge was overly deferential to Dr. Smith and that this may have introduced error into the decision.

First, it certainly seems like the rules of evidence were not vigorously argued or enforced here. That’s fine – there may have been strategic and economic reasons for this – but it is important to note. Both experts were admitted on consent. And moreover, judges tend to be more relaxed about the rules of evidence in bench trials. Justice Lax provided one of the most thorough justifications for this practice (Chan v Erin Mills Town Centre Corp at para 31):

…(T)he metaphor of the judge as gatekeeper loses much of its symbolic force when it is the judge who is the trier of fact. This is not to say that a trial judge is excused from scrutinizing evidence as improperly admitted evidence can surely have an impact on a trial, but the likelihood of a judge being overwhelmed by the “mystic infallibility” of the evidence and misusing the evidence to distort the fact-finding process, is far more remote. The dangers that the principles are designed to avoid begin to fall away.

Similarly, other contextual aspects may have militated towards less scrutiny of the experts. For instance, some have argued (see Edmond & Roach) that trial judges should be most vigilant towards prejudicial evidence when it is brought against the accused in criminal trials. D. Michael Risinger has said – speaking of the reliability of expert evidence, but I think it applies more generally – something similar:

Elsewhere, I have argued that there ought to be varying levels of foundational reliability, with that required for prosecution-proffered expertise in criminal cases being very high, especially when it goes to issues of “brute fact” guilt or innocence, such as the identity of the defendant as the perpetrator.

Here, the stakes seemed relatively low and both the Crown and Comeau had (ostensibly) well-qualified experts. Neither party was at the mercy of an expert. And this was certainly not a case in which an accused could not afford to find an expert to rebut a Crown’s forensic (or otherwise) evidence (see, for instance, the Motherisk miscarriages of justice). I might quibble more with Justice Lax’s comment about judges being less prone to the fall under the sway of a mystically infallible expert. Here, with the trial judge’s description of the expert being world-renowned and of unimpeachable credibility, it certainly seems like he may not have given enough scrutiny to Dr. Smith’s evidence (Scott Dallen and I discussed this at pages 545-548 of our 2016 paper).

Had the trial judge considered the admissibility of the experts’ evidence, he might have found that it should have at least been scoped to not so clearly infringe the “necessity” requirement in R v Mohan. As the Supreme Court (in Comeau) found: “…it is difficult if not impossible to contemplate a situation where evidence on domestic law (e.g. interpreting a Canadian statute) would ever be admissible as expert opinion evidence under R. v. Mohan, … application of contextual factors, including drafters’ intent, to the interpretation of a statutory provision is not something that is ‘outside the experience and knowledge of a judge’”

Moreover, opining on domestic law has also been conceived as an exclusionary rule (Andersen v St. Jude Medical Inc. at paras 5-7). In that case (also a decision of the late Justice Lax), the defendant called Professor Michael Trebilcock (a law professor at the University of Toronto) to explain the problems with waiver of tort (i.e., a doctrine saying damages are not required for a negligence claim in certain circumstances) from an economics and law perspective. Here, you might draw a parallel between Trebilcock’s opinion and that of Dr. Bateman. Both explained the public policy issues raised by a potential change in the law. And they are both distinguishable from the task of legally interpreting section the relevant provisions, which apparently Dr Smith did and the trial judge accepted.

The justification for excluding expertise on domestic law (either as an exclusionary rule or as being unnecessary) is that it is the trier of law’s job to interpret the law and determine if it applies. He or she would then explain that to the jury, if it was a jury trial. Expertise is a superfluity. That said, I think there might also be an issue of reliability here. In particular, just as I wouldn’t want a forensic scientist trained in DNA analysis to also opine on whether two fingerprints match, I don’t know if Dr. Smith – a historian – was competent to interpret legal provisions. He may well have been competent at unearthing and interpreting historical documents and placing those documents in their historical context. But does he have any training or education in interpreting the law? Helena Likwornik and I have suggested that in such cases, the task of establishing the scope of an expert’s testimony can help in determining if they are competent to give that testimony – it may promote analytic clarity (See Chin & Likwornik, 2017). There are also many cases of experts giving evidence that seems related to their field of expertise, but is only bears only a “specious nexus” to it (See Chin et al, 2017).

One of the reasons we worry about the reliability of expert evidence is that experts are often biased. This is not intentional; human cognition is by its nature motivated. It’s impacted by irrelevant details (i.e., contextual bias) and by the decision that we want to come to (i.e., confirmation bias). By my reading of the trial decisions and Dr. Smith’s CV, it seems that the hypothesis that the framers wanted free trade is important to him. In these conditions, people tend to emphasize the data that supports their view and deemphasize the data that does not support it. Again, I do not think that Dr. Smith was being duplicitous, just that he is human. In fact, recent empirical research finds that many scientists present their results in a falsely attractive way and that these practices have been generally accepted for a long time (Munafò et al). Further, New Brunswick’s rules do not require an oath that the expert has presented both evidence confirming and disconfirming their view (New Brunswick Rules of Court).

Amos Tversky was one of the world’s leading researchers on cognitive bias. Daniel Kahneman, his long time research partner, won the Nobel Prize for their research in this area (you can’t win a Nobel posthumously). I say this because Tversky once said the following to a group of historians (Lewis, 2016):

All too often, we find ourselves unable to predict what will happen; yet after the fact we explain what did happen with a great deal of confidence. This ‘ability’ to explain that which we cannot predict, even in the absence of any additional information, represents an important, though subtle, flaw in our reasoning. It leads us to believe that there is a less uncertain world than there actually is...

While cognitive bias among forensic science practitioners has received a lot of attention (e.g., NAS Report), Tversky’s worry here suggests the same applies to historians. It is easy to read a few historical documents that confirm your view and assume it is as certain as that. Dr. Smith could very well have been fooling himself into thinking the historical picture was as clear cut as he wanted it to be. 

Who decides history?

While I have been critical of the Comeau trial judge’s uncritical acceptance of expert evidence, I certainly do not mean to say that his perspective was not valuable. And I also don’t think that the trial judge was better placed to research the history of section 121 and interpret that history. And while judicial factfinding is considered more permissible with legislative facts, that practice carries its own serious dangers (Larsen). And the same goes for simply following other courts for their findings of fact (i.e., “factual precedent”; Larsen).

Instead, I think Comeau makes salient the importance of two commons sense safeguards. First, even in bench trials, trial judges should seriously consider the admissibility and scope of expert evidence. Had the trial judge gone through this exercise, he might have realized that a significant portion was plainly inadmissible evidence. The expert provided a domestic legal opinion that he was apparently not proficient to give. Second, it suggests radical transparency ought to be demanded of experts. Smith should have been required to disclose not just the history that supported his opinion, but the history that did not support it (or attest that he had provided all of the relevant history).


  • Allison Orr Larsen, “Factual Precedents” (2013) 162 University of Pennsylvania Law Review 59.

  • Andersen v St. Jude Medical Inc., 2011 ONSC 2178.

  • Canada (Attorney General) v Bedford, 2013 SCC 72.

  • Chan v Erin Mills Town Centre Corp., 2005 CarswellOnt 6741

  • D Michael Risinger, “Defining the ‘Task at Hand’: Non-Science Forensic Science After Kumho Tire Co. v. Carmichael” (2000) 57 Washington & Lee Law Review 767.

  • Gary Edmond & Kent Roach, “A Contextual Approach to the Admissibility of the State’s Forensic Science and Medical Evidence” (2011) 61:3 UTLJ 343.

  • Jason M Chin & Helena Likwornik, “R v Bingley and the Importance of Scientifically Guided Legal Analysis” (2017) 43:1 Queen’s LJ 33.

  • Jason M Chin, Jan Tomiska & Chen Li, “Drawing the Line Between Lay and Expert Opinion Evidence” 63:1 McGill LJ (forthcoming, 2017).

  • Jason M Chin & Scott Dallen, “R v Awer and the Dangers of Science in Sheep’s Clothing” (2016) 63 Crim LQ 527.

  • Marcus R Munafò et al, “A Manifesto for Reproducible Science” (2017) 1 Nature Human Behaviour 0021.

  • Michael Lewis, The Undoing Project: A Friendship That Changed Our Minds (New York, WW Norton & Company: 2016).

  • Ontario, Harmful Impacts: The Reliance on Hair Testing in Child Protection Report of the Motherisk Commission (Ontario: Ministry of the Attorney General, 2018) (The Honourable Judith C Beaman)

  • R v Comeau, 2016 NPPC 3.

  • R v Comeau, 2018 SCC 15.

  • R v Mohan, 114 DLR (4th) 419.

  • William Twining, "Taking Facts Seriously - Again" (2005) 55 Journal of Legal Education 360.