In two cases decided in the 1990s, the Supreme Court of Canada rejected the doctrine that courts cannot consider the international legal context of a provision they are interpreting unless there is some ambiguity on the face of the statute or regulation that permits resort to international law. The Supreme Court has since embraced international law as a contextual consideration in statutory interpretation, leaving the ambiguity doctrine behind. But the Federal Court of Appeal recently cited one of its own precedents on ambiguity (National Corn Growers) despite that decision having been overturned by the top court on this very point. The FCA seems divided on the issue, with some members following SCC precedent and others trying to bring ambiguity back.
NOTE: This post was originally published on 16 May 2019. I updated it on 18 June 2019 to reflect a correction issued in the Tapambwa decision, discussed below.
The story of ambiguity and international law in Canadian jurisprudence is inseparable from the interpretive presumption that Parliament does not intend to legislate contrary to international law. The concept of ambiguity is meant to answer the question, Should the court apply the presumption of conformity?
The old cases said Yes
In older cases, the answer to that question tended to be Yes. The prevailing judicial approach was to apply the presumption of conformity to avoid internationally unlawful interpretations of domestic laws unless the legislature unambiguously expressed its intention to act contrary to it. An illustration is Lord Escher MR in Colquhoun v Brooks (1888) 21 QBD 52 at 57-8:
…the English parliament cannot be supposed merely by reason of its having used general words to be intending to do that which is against the comity of nations. It is true that if we came to the conclusion that this had been intentionally done we must carry out the law and leave to the government of the country the task of answering objections, but unless that is perfectly clear we ought to limit the words so as to make them reasonable and proper…
(My emphasis.) Similarly, a leading English statutory interpretation text in the late nineteenth century, Maxwell’s Interpretation of Statutes (1875), insisted that if Parliament intends to violate international law or comity “it must express its intention with irresistible clearness, to induce a Court to believe that it entertained it; for if any other construction is possible, it would be adopted, in order to avoid imputing such an intention to the Legislature” (my emphasis).
This is not to deny that there are old English authorities insisting on Parliament’s power to violate international law and the courts’ duty to give effect to such exercises of its sovereignty. There are plenty. An example is R v Keyn (1876) 2 Ex Div 63 where Cockburn CJ affirmed that the international legal rule that domestic laws do not apply extraterritorially
must be taken subject to this qualification—namely, that if the Legislature of a particular country should think fit by express enactment to render foreigners subject to its law with reference to offences committed beyond the limits of its territory, it would be incumbent on the Courts of such country to give effect to such enactment, leaving it to the State to settle the question of International Law with the Governments of other nations.
Note, however, that even this sentiment is qualified by the necessity for an “express enactment”.
The regularity with which the presumption of conformity was applied in English law up to 1939 was commented upon by Prof. Lauterpacht in his well-known paper to the Grotius Society, “Is International Law a Part of the Law of England?” (1939) Transactions of the Grotius Society 51. He observed (at pp. 58-9) that the “practical effects” of the presumption of conformity “have been considerable” and that:
The absolute superiority of Acts of Parliament, even when they conflict with International Law, has indeed been repeatedly and emphatically affirmed. But this has been a theoretical affirmation having the probably not unintended effect of stressing the duty of Judges to do their utmost to interpret statutes so as not to impute to the Legislature the intention of disregarding International Law….
The presumption that Parliament did not intend to commit a breach of the law of nations has been a powerful weapon wielded with a determination which on occasions has come near to a denial of the supremacy of Parliament.
Under the Maxwell/Colquhoun approach, the court will presume legislative provisions were intended to conform to international law, and interpret them accordingly, unless the legislature unambiguously expresses its intention to breach the State’s international obligations. While these authorities did not speak in terms of the international context in which laws are made (as our Supreme Court has done since the 1990s), the starting-point is nevertheless an awareness of the international legal order and a commitment to reaching internationally-compliant results unless Parliament makes “perfectly clear” that it means to exercise its undoubted power to legislate contrary to international law. If there was rule about ambiguity in the old presumption cases, it was this: apply the presumption of conformity unless the legislature has unambiguously signalled its intent to breach international law.
The ambiguity flip
Sometime in the mid 20th century, the role of ambiguity in internationally-informed statutory interpretation was reversed. The English courts went from requiring unambiguous words to rebut the presumption of conformity to requiring ambiguous words to permit it.
It is easy to see how this flip could happen. The claim, “A statute that is unambiguously contrary to international law must be followed”, resembles the claim, “A statute that is unambiguous must be given effect even if it is contrary to international law”. The first claim can be confused with the second, and vice versa. But the two claims are in fact different.
The first claim recognizes the constitutional orthodoxy, in Westminster model states such as the UK and Canada, that legislatures are sovereign to violate international law, and implies a judicial “no ambiguity” rule: we (judges) won’t interpret domestic laws as contrary to international law unless you (Parliament) make yourself perfectly clear. Under this approach, the international law context of the provision at issue is a relevant interpretive consideration from the outset.
The second claim, by contrast, sets up ambiguity as a prerequisite to considering the international law context of Parliament’s enactment at all. The judicial interpretive strategy, under the second claim, begins by asking: Is this provision unambiguous? If the answer is Yes, the court must give effect to it whatever the international legal context or consequence. The presumption of conformity with international law is ousted by the provision’s lack of ambiguity. Indeed, the presumption can only apply if there is some ambiguity in Parliament’s expression that permits the courts to resort to international law as an interpretive consideration.
This new approach—ambiguity as a prerequisite to giving interpretive weight to the international legal context of the provision under scrutiny—flips the older “no ambiguity” rule on its head. The old “no ambiguity” rule is a judicial technique promoting the State’s compliance with international law. This new ambiguity requirement, by contrast, is a judicial avoidance technique, an excuse not to comply with international law. International law is a mandatory contextual consideration in the one, and an extrinsic interpretive aid in the other.
In more extreme expressions of the ambiguity requirement, domestic courts cannot even look at a relevant international norm (usually found in a treaty) if there is not some ambiguity on the face of the enactment to permit resort to it. In its milder form, courts do not forbid regard for international legal considerations entirely but use the supposed lack of ambiguity on a provision’s face to brush aside the presumption of conformity. As Murray Hunt describes this development in Using Human Rights Law in English Courts (Oxford, 1998) at p. 19, under the new ambiguity requirement, “The terms of the treaty only became relevant to the judicial interpretive task once the court had first decided that there was an ambiguity or lack of clarity in the statutory language.”
The Supreme Court flips back
The new rule that legislative provisions had to be ambiguous on their face before courts could interpret them in their international legal context, and apply the presumption of conformity, had this big problem: facial ambiguity is obviously in the eye of the judicial beholder. If the judge wants to consider international law, she can usually find some ambiguity to permit it. And if she doesn’t, she won’t.
The new rule had a deeper problem, too. It was methodologically out of step with prevailing ideas about the construction of laws. A passage from Professor Côté’s The Interpretation of Legislation in Canada, 4th ed (2011) at 268–269 rejects “the idea that interpretation is legitimate or appropriate only when the text is obscure”, saying:
This idea is based on the view, incorrect, that the meaning of a legal rule is identical to its literal legislative wording. The role of the interpreter is to establish the meaning of rules, not texts, with textual meaning at most the starting point of a process which necessarily takes into account extra-textual elements. The prima facie meaning of a text must be construed in the light of the other indicia relevant to interpretation. A competent interpreter asks whether the rule so construed can be reconciled with the other rules and principles of the legal system: Is this meaning consistent with the history of the text? Do the consequences of construing the rule solely in terms of the literal rule justify revisiting the interpretation? and so on.
The ambiguity question came up for decision by the Supreme Court of Canada in National Corn Growers Assn v Canada (Import Tribunal)  2 SCR 1324. In the Federal Court of Appeal (1989 CanLII 5205), Iacobucci CJ (as he then was) had declined to resort to the international agreements at issue, applying the dictum of Estey J in Schavernoch v. Foreign Claims Commission  1 SCR 1092 at 1098:
If one could assert an ambiguity, either patent or latent, in the regulations it might be that a court could find support for making reference to matters external to the regulations in order to interpret its terms. Because, however there is in my view no ambiguity arising from the above-quoted excerpt from these regulations, there is no authority and none was drawn to our attention in argument entitling a court to take recourse either to an underlying international agreement or to textbooks on international law with reference to the negotiation of agreements or to take recourse to reports made to the Government of Canada by persons engaged in the negotiation referred to in the regulations.
Here, Iacobucci CJ held that there was no ambiguity.
In dissent, MacGuigan JA quoted approvingly Lord Denning MR’s statement in Salomon v. Commissioners of Customs & Excise  3 All E.R. 871 at 874 that “we ought always to interpret our statutes so as to be in conformity with international law. Our statute does not in terms incorporate the convention, nor refer to it; but that does not matter”.
At the Supreme Court of Canada, Iacobucci CJ was reversed. Gonthier J for the majority of the court explained Schavernoch, and re-cast the ambiguity requirement, as follows (at 1371-2):
The first comment I wish to make is that I share the appellants’ view that in circumstances where the domestic legislation is unclear it is reasonable to examine any underlying international agreement. In interpreting legislation which has been enacted with a view towards implementing international obligations, as is the case here, it is reasonable for a tribunal to examine the domestic law in the context of the relevant agreement to clarify any uncertainty. Indeed where the text of the domestic law lends itself to it, one should also strive to expound an interpretation which is consonant with the relevant international obligations.
Second, and more specifically, it is reasonable to make reference to an international agreement at the very outset of the inquiry to determine if there is any ambiguity, even latent, in the domestic legislation. The Court of Appeal’s suggestion that recourse to an international treaty is only available where the provision of the domestic legislation is ambiguous on its face is to be rejected. As I. Brownlie has stated at p. 51 of Principles of Public International Law (3rd ed. 1979):
If the convention may be used on the correct principle that the statute is intended to implement the convention then, it follows, the latter becomes a proper aid to interpretation, and, more especially, may reveal a latent ambiguity in the text of the statute even if this was ‘clear in itself’. Moreover, the principle or presumption that the Crown does not intend to break an international treaty must have the corollary that the text of the international instrument is a primary source of meaning or ‘interpretation’. The courts have lately accepted the need to refer to the relevant treaty even in the absence of ambiguity in the legislative text when taken in isolation.
In Schavernoch v. Foreign Claims Commission,  1 S.C.R. 1092, this Court had occasion to comment upon the circumstances in which it is proper for the courts to consult an underlying international agreement. Though the language used by Estey J. is perhaps not explicit, I do not understand his remarks to mean that consultation of the treaty is proper only where it appears that the text to be interpreted is ambiguous on its face….
The suggestion that recourse can be had to an underlying international agreement where a latent ambiguity can be asserted implies that there is no need to find a patent ambiguity before consultation of the agreement is possible. As a latent ambiguity must arise out of matters external to the text to be interpreted, such an international agreement may be used, as I have just suggested, at the preliminary stage of determining if an ambiguity exists.
Having found that the rules of statutory interpretation allow consideration of an underlying agreement at the preliminary stage of determining if the domestic legislation contains an ambiguity, I do not hesitate to conclude in this case that the Tribunal did not act unreasonably in consulting the GATT. The question that remains, therefore, is whether the Tribunal’s interpretation of s. 42 in this case, given the relevant SIMA and GATT provisions, is unreasonable having regard to all the circumstances.
In short, Gonthier J rejected the notion that consultation of a treaty is proper only where it appears that the text to be interpreted is ambiguous on its face, and affirmed that there is no need to find a patent ambiguity before such consultation is possible. Rather, an international agreement may be used at the preliminary stage of determining if an ambiguity exists. Ambiguity is not jettisoned here entirely, but it loses its gatekeeper function. The court cannot safely say whether a provision is ambiguous or not until comparing its plain meaning with its international context, such as an underlying treaty. The court reviews the legislative provisions, reviews the relevant international provisions, and decides upon consideration of both whether there is ambiguity in the legislation to be resolved according the the presumption of conformity with international law.
Five years later, the Supreme Court of Canada dropped ambiguity as a precursor to consideration of international law entirely in Crown Forest Industries Ltd v Canada  2 SCR 802 at para 44. Notably, it was Justice Iacobucci, now a member of the Supreme Court of Canada, who affirmed that “a court may refer to extrinsic materials [in this case, international agreements] which form part of the legal context…without the need first to find an ambiguity before turning to such materials”.
Where Gonthier J. had seemed to say that ambiguity was still a consideration, but could be found by comparing the domestic provision with its international counterpart, Iacobucci J abandoned it. Like Lord Escher in Colquhoun and Maxwell in 1875, the court in Crown Forest seems, in retrospect, to be espousing the application of the presumption of conformity in all cases save those where Parliament unambiguously exercises is sovereignty to put Canada in breach of its international obligations. The role of ambiguity has shifted back from “A statute that is unambiguous must be given effect even if it is contrary to international law” to “A statute that is unambiguously contrary to international law must be followed”.
Context not ambiguity
From Crown Forest forward, the Supreme Court’s approach to international law considerations in statutory interpretation has been governed not by ambiguity but by context. International law and the State’s obligations under it are contextual considerations in the interpretive exercise, and ambiguity is not part of the inquiry.
Justice Linden of the Federal Court of Appeal recognized this change in the law in Canada v Seaboard Lumber Sales Co  3 FC 113 at 120, saying:
It is now established that courts will look to relevant international documents to aid interpretation of implementing legislation from the outset of the investigation, and even absent ambiguity on the face of that legislation. Ambiguity may arise out of the consideration of any manner or variety of contextual factors; it should not be taken as a necessary precondition to looking to those factors.
The Supreme Court has kept to this path ever since. In Baker v Canada  2 SCR 817, L’Heureux-Dubé J for the majority (at para 70) approved Prof. Sullivan’s statement that “the legislature is presumed to respect the values and principles enshrined in international law, both customary and conventional” as these “constitute a part of the legal context in which legislation is enacted and read”. Eight years later, LeBel J in R v Hape 2007 SCC 26 at para. 53 approved this passage from Sullivan and Driedger on the Construction of Statutes again, observing that the presumption of conformity has two aspects:
First, the legislature is presumed to act in compliance with Canada’s obligations as a signatory of international treaties and as a member of the international community. In deciding between possible interpretations, courts will avoid a construction that would place Canada in breach of those obligations. The second aspect is that the legislature is presumed to comply with the values and principles of customary and conventional international law. Those values and principles form part of the context in which statutes are enacted, and courts will therefore prefer a construction that reflects them.
(My emphasis.) Eight years later again, McLachlin CJ in B010 v. Canada (Citizenship and Immigration) 2015 SCC 58, in a discussion headed “International Law as Context: General Relevance”, repeated (at para 47) Baker and Hape’s affirmation that “the values and principles of customary and conventional international law form part of the context in which Canadian laws are enacted” and added that this proposition “follows from the fact that to interpret a Canadian law in a way that conflicts with Canada’s international obligations risks incursion by the courts in the executive’s conduct of foreign affairs and censure under international law”. The chief justice added (at para 49) that courts are directed to “relevant international instruments at the context stage of statutory interpretation”, then proceeded to do just that (at paras 51-66).
These (and other) recent Supreme Court precedents establish beyond argument that consideration of relevant international legal considerations are part of the modern approach to statutory interpretation. They enter into the analysis at the contextual stage—without depending on a finding of ambiguity at the textual/ordinary meaning stage.
But Parliament is still sovereign
The Supreme Court’s abandonment of the ambiguity requirement in favour of international-law-as-context does not undermine legislative sovereignty to enact laws inconsistent with international law. The Supreme Court of Canada has often affirmed that sovereignty in the same breath as it has affirmed the presumption of conformity. In doing so, however, it has again recast the role of ambiguity. In the place of the ambiguity-as-precursor doctrine of some twentieth century English cases, the court has given support for the no ambiguity/”perfectly clear” approach I have associated with Lord Escher and Maxwell. Legislatures can enact laws contrary to the State’s obligations, but they must express themselves unambiguously before the courts should give effect to such interpretations. Hape at para 53 is again relevant here (my underlining):
…The presumption of conformity is based on the rule of judicial policy that, as a matter of law, courts will strive to avoid constructions of domestic law pursuant to which the state would be in violation of its international obligations, unless the wording of the statute clearly compels that result. … The presumption is rebuttable, however. Parliamentary sovereignty requires courts to give effect to a statute that demonstrates an unequivocal legislative intent to default on an international obligation….
This is the Supreme Court’s strongest statement of how clearly a legislature must express itself to rebut the presumption of conformity. Other Supreme Court authorities are less emphatic, or at least less clear, on what is needed to rebut the presumption. In Kazemi Estate v Iran 2014 SCC 62 at para 60, LeBel J noted that “International law cannot be used to support an interpretation that is not permitted by the words of the statute”, adding that “clear words” are needed to rebut the presumption. Is this “clear words” standard hard or easy to satisfy? Similarly, in Németh v. Canada (Justice) 2010 SCC 56 at para 35, Cromwell J noted that the presumption of conformity is rebuttable, saying, “If the provisions are unambiguous, they must be given effect”. Again, is this a high bar or a low one? Ironically, the word “unambiguous” in this quotation can bear two meanings: either Parliament must unambiguously express its intent to breach international law (a high standard consistent with Hape, Lord Escher in Colquhoun, and B010‘s concerns about international censure and incursion in the conduct of foreign affairs), or the provision in question must simply admit of no ambiguity (a low standard that recalls the ambiguity-as-prerequisite approach).
Recent Federal Court of Appeal decisions
I come now to recent Federal Court of Appeal case law on ambiguity. That court is alive to the rule in National Corn Growers and the approach to ambiguity it rejected. In Najafi v Canada (Public Safety and Emergency Preparedness) 2014 FCA 262 at para 61, Justice Gauthier expressed her view that “relevant international law, like other relevant elements of the legal context, should ideally be taken into account before concluding whether or not a text is clear or ambiguous” but noted that Sullivan’s Construction of Statutes observes that “many courts still consider ambiguity a prerequisite”. Despite this, Nadon JA for the Federal Court of Appeal followed National Corn Growers (SCC) in Pembina County Water Resource District v Manitoba (Government) 2017 FCA 92. Unusually, the learned judge noted that the Supreme Court in that case “specified that recourse can be had to international treaties even where the legislative provision is not ambiguous (overturning this Court on this point)” (at para 46).
One might have thought that a 2017 Federal Court of Appeal decision from a leading jurist on international law questions, noting that the Supreme Court of Canada had settled the ambiguity question in 1990—overturning the Federal Court of Appeal to do it—would be the last word on the matter.
Yet in Tapambwa v Minister of Citizenship and Immigration 2019 FCA 34, Rennie JA seems to breathe new life into the ambiguity doctrine. The appellants argued that s. 112(3)(c) of the Immigration and Refugee Protection Act had to be interpreted to permit reconsideration of a prior exclusion finding by a PRRA officer. They contended that the international law obligation of non-refoulement required such an interpretation. Rennie JA for the court rejected this argument in the following terms (paras. 42, 44):
There is a well-established presumption that, where possible, Canada’s domestic legislation should be interpreted to conform to international law (R. v. Hape, 2007 SCC 26 at para. 53,  2 S.C.R. 292 (Hape)). Absent contrary indication, legislative provisions are also presumed to observe
“the values and principles of customary and conventional international law” (Hape at para. 53; B010 at para. 47; Ruth Sullivan, Sullivan on the Construction of Statutes, 6th ed., (Markham: LexisNexis Canada Inc., 2014) at §18.6; see also de Guzman v. Canada (Minister of Citizenship and Immigration), 2005 FCA 436 at paras. 82–87,  3 F.C.R. 655).
There is, however, an important counter-weight to these principles – the doctrine of Parliamentary supremacy. An unambiguous provision must be given effect even if it is contrary to Canada’s international obligations or international law (Németh at para. 35; Schreiber v. Canada (Attorney General), 2002 SCC 62 at para. 50; National Corn Growers Assn. v. Canada (Canadian Import Tribunal),  2 F.C. 517 at p. 530, 58 D.L.R. (4th) 642 at p. 651 (F.C.A.); Gitxaala Nation v. Canada, 2015 FCA 73 at para. 16; Hape at para. 54).
(My underlining.) As we have seen, the mere statement that an unambiguous provision must be given effect even if contrary to international law is not inconsistent with National Corn Growers (SCC), Crown Forest or later context-based authorities. Parliamentary sovereignty means that our legislatures are competent to enact internationally unlawful laws. But Rennie JA’s citation of National Corn Growers (FCA) muddies things. The cited part of that decision (p. 530) reads,
Applying the foregoing guidelines to the case before us, the question arises whether in the language of section 42 there are any ambiguities or obscurities which would warrant resort to the related international agreements for clarification. In my view section 42 is clear and unambiguous…
This is the very reasoning on ambiguity Gonthier J rejected in the Supreme Court.
Tapambwa has already been cited by the Federal Court of Appeal on this point in two cases: Nova Tube Inc./Nova Steel Inc. v. Conares Metal Supply Ltd. 2019 FCA 52 at paras 58-60 and Hillier v Canada (Attorney General) 2019 FCA 44 at para 38. Laskin JA’s phrase in Nova Tube that “…I see no ambiguity that would justify resort to Canada’s international obligations” (para 60, my emphasis) is especially remarkable for its inconsistency with National Corn Growers (SCC), Crown Forest and the whole tenor of Supreme Court international-law-as-context jurisprudence in the last 30 years.
(In making these points I express no opinion on the correctness of the results in Tapambwa, Nova Tube or Hillier.)
Update (18 June 2019): On 31 May 2019 (fifteen days after I originally published this), the Federal Court of Appeal issued a correction to Tapambwa. The citation to National Corn Growers (FCA) at para 44 was replaced by a citation to National Corn Growers (SCC) at p 1371. The reasoning, however, is unchanged. So, too, is Laskin JA’s reasoning in Nova Tube at para 60.)
This leaves the law in a rather awkward place. Of all Canadian appellate courts, the Federal Court of Appeal is likely the one that most frequently (and most fluently) considers international legal submissions. Many of these come to that court via the Federal Court, of course. Confronted with both Nadon JA in Pembina County and Rennie JA in Tapambwa, what is a Federal Court trial judge to do? What, for that matter, is the Federal Court of Appeal itself to do the next time round?
The answer is easy: follow the Supreme Court of Canada. That court does not appear to have changed course on the interpretation of statutes in light of the State’s international obligations. It is hard to see how the Federal Court of Appeal can do so.