On February 6, 2015 the Supreme Court of Canada struck down two provisions of the Criminal Code making it an offence for physicians to help certain persons end their lives. Or, at least, it appeared to. The Court declared that ss. 14 and 241(b) are of no force and effect, but only

to the extent that they prohibit physician-assisted death for a competent adult person who (1) clearly consents to the termination of life and (2) has a grievous and irremediable medical condition (including an illness, disease or disability) that causes enduring suffering that is intolerable to the individual in the circumstances of his or her condition. [The ”˜Carter conditions’.]

At the same time, the Court expressly said that people who satisfy the ”˜Carter conditions’ are not eligible for a ”œconstitutional exemption” – that judges may not grant relief to individuals on a case-by-case basis. In this post, we want to think about that claim. In doing so, we focus on the remedy in Carter, and set aside the Court’s additional order to suspend the remedy for one year. After all, there is now a very real possibility that the suspension will lapse before the federal government can or will implement a regime of its own. We may need to live with Carter for much longer than the Court anticipated.


Our first point is this: viewed on its own terms, the remedy ordered in Carter seems to cash out as a constitutional exemption in all but name. It is certainly not self-executing. At some point, either a physician or the Crown will want to know whether charges may be brought under s. 241(b) for helping a given person end his or her life. (We will leave aside whether and how the issue might reach the courts before an actual prosecution is undertaken.) In answering that question, a judge will presumably need to decide whether that individual falls within the scope of the protected class articulated in Carter. If she could not do so, on the basis that she is precluded from issuing constitutional exemptions, we would be left with two unhappy possibilities: either sections 14 and 241(b) apply to no one, even though the Court in Carter implied that the sections are valid insofar as they apply to individuals not satisfying the Carter conditions; or they apply to everyone, even though the Court found that the sections are invalid to the extent they apply to people who do satisfy those conditions. Neither possibility strikes us as plausible or acceptable.

So why did the Court in Carter refuse to expressly recognize a role for constitutional exemptions, given that the practical effect of its declaration of invalidity was to create a de facto regime of constitutional exemptions? Why might it decide to make this sort of rhetorical move?


In refusing to permit constitutional exemptions, the Court purported to follow a 2008 decision: Ferguson. In that case, the Court considered whether exemptions are available for mandatory minimum sentences that violate section 12 of the Charter. A mandatory minimum sentence that is appropriate in the vast majority of cases may nonetheless be grossly disproportionate for a relatively small class of offenders. In such circumstances, one could argue that it does not make sense to strike down the offence altogether. It might be seen as more respectful of Parliament to simply permit exemptions in those rare instances where the prescribed minimum sentence is unconstitutional.

The Court in Ferguson unanimously rejected that possibility, going so far as to say that constitutional exemptions are disrespectful of the role of Parliament. When the judiciary uses constitutional exemptions, the Court reasoned, it effectively rewrites the statutory provision: it takes a law of general application, and turns it into something requiring the exercise of judicial discretion. That is especially clear in a case like Ferguson. After all, the purpose of mandatory minimum sentences is to constrain judicial discretion – to prevent judges from imposing sentences that fall below the minimum. Allowing a judge to exempt a particular offender from a mandatory minimum does exactly the sort of thing Parliament wished to avoid by enacting the minimum in the first place. It creates uncertainty, in a context where Parliament intended to impose certainty.

Arguably, there are similar factors at work in a case like Carter. Parliament (it would appear) proceeded on the basis that anything less than a blanket prohibition on physician-assisted suicide would fail to protect vulnerable Canadians. (We acknowledge that we are here attributing the arguments made by the Crown to Parliament.) A legal regime relying upon highly nuanced, case-by-case analysis, Parliament implicitly supposed, is bound to produce ”˜false positives’ – people who do not satisfy the Carter conditions, but whose deaths would be authorized (whether before or after the fact). There is no reason to believe that the Parliament that enacted section 241(b) would have been satisfied with a case-by-case assessment of individuals’ ”˜eligibility’ for physician-assisted suicide.

And all of this leaves aside a further cause for concern: the problem of ”˜false negatives’. A judge could wrongly conclude that a patient fails to satisfy the Carter conditions. In that case, a patient with a constitutional right to relief would remain in a state of suffering.

So it is possible to see Carter, like Ferguson, as a case about the perils of judicial discretion. Those perils might explain the Court’s reluctance to permit constitutional exemptions. But if the Court was worried about discretion, there is something funny about the remedy it did provide. Although the Court in Carter might seem to have replaced one bright-line rule (”œno physician-assisted suicides”) with another (”œno physician-assisted suicides except where the Carter conditions hold”), in fact the replacement rule is quite murky. How ”œcompetent” must a person be to consent to his or her own death? Is it the same degree of ”œcompetence” required to refuse life-saving medical treatment? Must ”œconsent” be express and contemporaneous in order to be ”œclear”? If not – if, say, patients can have legally effective advance directives – how temporally remote can consent be before we conclude that it is not informed, or does not reflect the priorities and values of the patient in the here and now? ”œConsent” means different things in different legal contexts. What, then, might it mean in cases of physician-assisted suicide? And what about ”œirremediable”? Must a patient show that medical science will be unable to treat her condition in the foreseeable future, or only in the immediate present? Does it matter that the condition is treatable in principle if its cost is not covered by provincial health care plans?

The language used in the Carter conditions – however neutral it might appear – glosses over a range of important policy questions. Before the Carter conditions can become useful as a test, those questions must be resolved. If, as we have assumed, it will be the responsibility of the courts to give content to the Carter conditions, it follows that they will be doing much more than deciding baseline questions of constitutional validity. They will be engaging in the quasi-legislative task of deciding when, as a matter of social policy, we should regard consent to death (at the hands of third parties) as legally effective.

The Court in Carter avoids judicial discretion in fact-finding but effectively embraces judicial discretion in value-finding. Is the latter kind of discretion less problematic? We’re not sure the Court in Ferguson would have thought so. Constitutional exemptions to mandatory minimums leave it to the individual judge to decide, not just what facts he or she accepts, but what significance those facts ought to have when crafting a proportionate sentence. Mandatory minimums reflect Parliament’s view that those who commit certain offences necessarily deserve at least X amount of punishment. Exemptions permit individual judges to make the kind of value judgment that Parliament purported to pre-empt – to seize a core aspect of social policy-making over which Parliament has claimed dominion. It was partly on this basis that the Court in Ferguson held that it is preferable, in cases where a mandatory minimum is found unconstitutional, to simply issue a declaration of invalidity.

It is worth noting that the Court in Ferguson found that other forms of textual tinkering – in particular, severance and reading in – were problematic for much the same reason. It observed:

When a court opts for severance or reading in as an alternative to striking down a provision, it does so on the assumption that had Parliament been aware of the provision’s constitutional defect, it would likely have passed it with the alterations now being made by the court by means of severance or reading in. …. If it is not clear that Parliament would have passed the scheme with the modifications being considered by the court " or if it is probable that Parliament would not have passed the scheme with these modifications " then for the court to make these modifications would represent an inappropriate intrusion into the legislative sphere. In such cases, the least intrusive remedy is to strike down the constitutionally defective legislation under s. 52. It is then left up to Parliament to decide what legislative response, if any, is appropriate.

The Court took the view that since severance and reading in risk usurping the policy-making function of legislatures, these judicial remedies should be used with extreme caution. A declaration of invalidity is the safer path.

But assuming that usurpation is a genuine problem, does the declaration of invalidity issued in Carter really avoid it? We have seen that it will require at least as much judicial policy-making as reading in. It looks less intrusive, because striking down appears to be an all-or-nothing remedy that demands only compliance with the baseline requirements of the Constitution. It does not expressly leave it to the courts to make nuanced, value-laden decisions about who should be deemed competent to consent to death, etc. But it cashes out much the same way.

Now, the Court has done something very like what it has done in Carter before. Consider the ruling in Sharpe. Sharpe involved a challenge to the offences of creating and possessing child pornography. The provisions were found to be overbroad in that they presupposed that all visual and written materials meeting the Criminal Code’s definition of ”˜child pornography’ promote the sexual abuse of children. The Court didn’t agree, observing that some instances of so-called ”œchild pornography” depict activities that do not pose such a risk. For example, a 16-year-old girl engaged in a sexual relationship with her 16-year-old boyfriend who sends him an intimate photograph of herself, does nothing wrong by creating the photo, and the boyfriend does nothing wrong by keeping it so long as he does not distribute it to others. To cure the overbreadth problem, the Court carved out two exceptions to the offences in issue.

The Sharpe exceptions required something very like the judicial unpacking required by the Carter conditions. Precisely because there was some uncertainty about the parameters of the private use exception, the Court returned to the issue this year. In Barabash, the Court essentially gave additional content to the exception it had crafted in Sharpe (though it seemed to proceed as though it was engaging in just another exercise of statutory interpretation).

Leave aside whether the remedy in Sharpe is appropriate. It is enough to note that the Court explicitly acknowledged that it was reading in the exceptions it created. The Court in Carter didn’t deploy similar language. But the remedy it decided upon is, in terms of judicial craft, equally intrusive. Merely describing the remedy as declaratory cannot change its fundamental character.

Photo by Derek Tsang / CC BY 2.0 / modified from original