Bradley Miller: Some Problems in Constitutional Architecture

B-Miller-webAs my house approaches its centenary, it becomes harder to deny that some architectural designs age better than others.   No doubt it all looked great in 1930, but some of the roof lines have proved better than others at keeping the melting snow away from the foundation. Not to mention that some of the house’s later additions – though addressing real needs – employed questionable craftsmanship and don’t really cohere with the rest of the design.

Similarly, some parts of the Canadian constitution are holding up better than others.

Grégoire Webber and I have each argued that the Charter of Rights and Freedoms – a 1982 addition to the Canadian constitution – suffers from an overarching design flaw: its limitation clause. Like all limitations clauses (whether generalized, as in the Canadian, New Zealand, and South African examples, or contained in individual rights provisions such as in the European Convention on Human Rights) it has provided an environment congenial to several interpretive pathologies: chief among them a proportionality doctrine and a two-stage adjudicative structure that first asks whether a person’s right has been “violated” before inviting government to marshal reasons to justify the “violation”.

A feature of this analytical division is that the first instance the focus is to be entirely on the rights holder and no one else. (Set aside, for present purposes, whether it is actually possible to assess whether a person’s right to equality has been “violated” without considering the situation of others.) Thereafter, the onus shifts on the government to establish that the violation (better, “limitation”) of the right is justified. Thus we have the absurdity that legislation can be conceptualized as being at once fully justified and a violation of someone’s rights. Webber and I have both drawn criticism for thinking that this is a problem, and Webber has recently provided a characteristically careful and elegant response in “On the Loss of Rights”, a chapter in Proportionality and the Rule of Law, Rights, Justification, Reasoning(CUP 2014), a new collection of essays edited by Grant Huscroft, Bradley Miller, and Grégoire Webber.

I won’t summarize Webber’s arguments here, but will instead illustrate one way in which the problem that we insist is real is manifesting itself in Canadian constitutional law.

In practice, the two-stage division between limitation and justification has proven to be highly unstable. Some sections of the Charter, like the right to freedom of expression, have simply become speed bumps on the way to s.1’s justification phase. Almost no analytical work is done at the first stage. Other sections, like s.7’s guarantee that any deprivation of life, liberty, and security of the person accord with principles of fundamental justice, have gone the other way, with all of the work taking place at the first stage and s.1 becoming functionally irrelevant.

Given the criticisms that Webber and I have brought against the two-stage process, you might think I would welcome its collapse as a step in the right direction. But like my back deck, this collapse has a curious asymmetry about it, one that further deforms rights adjudication.

Consider the history of s.7 adjudication. Straight out of the gate, the Supreme Court of Canada announced that it would not ask itself what constitutional settlement the drafters or ratifiers of s.7 meant to enact through the words ‘principles of fundamental justice’. That settlement, had the Court been interested, was accessible enough: the principles of fundamental justice were entirely a matter of procedural fairness. Having rejected that proposition, the Court set a course for deep water. The idea that s.7 was intended to authorize judges to strike down laws on the basis of their non-conformity to some substantive theory of justice has some obvious drawbacks. Which theory? Who decides? Sensibly enough, the Court searched for some limiting principles, and rejected the ‘harm principle’ and other candidates as too controversial and insufficiently legal. The principles of fundamental justice had to be found within the fundamental tenets of the legal system, it held, as opposed to free-standing, controversial political philosophies.

So where to ‘find’ these legal, fundamental principles of justice without having to delve into contested political moralities? The Court’s strategy has been to adopt, incrementally (and quietly), principles adapted from the proportionality test used to give effect to the limitation clause. (Leave to one side my argument in ‘Proportionality’s Blind Spot’ in Huscroft, Miller, and Webber (eds) that the proportionality analysis that the court uncrated from the European Court of Human Rights came pre-loaded with the commitments of Rawlsian liberalism.) So the requirement that legislative means be rationally connected to its ends yields the principle that law must not be arbitrary. The requirement that legislation must be minimally impairing of rights yields the principle that law not be overbroad. And the requirement of proportionality stricto sensu yields the principle that law not be grossly disproportionate to its ends.

So now we have, in effect, the principles of proportionality doing double-time as the principles of fundamental justice. I address some of the drawbacks of the use of proportionality analysis in limitations clauses in Huscroft, Miller, and Webber (eds). These criticisms apply equally to their deployment in the context of s.7. But what is of more immediate concern is the redundancy of s.1. After all, could it make any sense for a court, having determined that a statute limits a person’s liberty in a way that is grossly disproportionate to its ends, then proceeding to ask whether that restriction on rights is nevertheless proportionate? So it is not at all surprising that where the Court has concluded that a s.7 right has been limited, it has never found that limit to be justified under s.1.

Recently addressing this issue in Canada (AG) v. Bedford(2013), the Court defended the two-stage analysis, in terms that fail conceptually:

[125]   Section 7 and s. 1 ask different questions. The question under s. 7 is whether the law’s negative effect on life, liberty, or security of the person is in accordance with the principles of fundamental justice. With respect to the principles of arbitrariness, overbreadth, and gross disproportionality, the specific questions are whether the law’s purpose, taken at face value, is connected to its effects and whether the negative effect is grossly disproportionate to the law’s purpose.  Under s. 1, the question is different — whether the negative impact of a law on the rights of individuals is proportionate to the pressing and substantial goal of the law in furthering the public interest.

What the Court is suggesting is that there is a difference between a law’s purpose (in s.7 analysis) and its public interest goal (in s.1). If this isn’t nonsense, then the distinction that the Court must have in mind – and it only hints at it – is quite troubling: that s. 1 justification, unlike the more principled s.7 analysis, allows for ‘quantitative’ (read ‘utilitarian’) justifications for the ‘greater public good’. Webber and I have each argued at length as to how such a reading of a limitation clause rests on a poor conception of the common good, one which is contrary to the dominant stream of the Court’s s.1 adjudication as exemplified in Newfoundland (Treasury Board) v. N.A.P.E. (2004).

In any event, the telling fact is that 30 years out, the returns of cases where the Court has found a s.7 limitation to have been justified under s.1 are still a null set.

Why isn’t this effective collapse of the two-stage analytical structure happy news to Webber and me? Perhaps it would be better to call it a partial collapse. After all, the Court remains committed to the proposition that it is in fact engaged in a two-stage enquiry. What is entailed by this commitment is that in what is deemed to be the first stage – but is in effect the only stage – the Court remains focussed entirely on the rights-holder. There is to be no consideration, at this stage, to the needs of other persons living in community with the rights-holder. Justice and justification are to be considered from one side only. All other considerations are to be postponed to the second stage that never comes.

To be sure, the Court throws out three paragraphs in Bedford under the requisite heading of s.1, in which it repeats some of its s.7 analysis, but there is no real work being done. Bedford demonstrates how the Charter’s architecture has resulted in s.7 and s.1 doctrines that make it profoundly difficult for the AG Canada to find any analytical space in which to articulate the reasoning behind much criminal legislation. Unless the s.7 doctrine is reformed, the persons and interests that protective legislation is meant to serve will not be present – or at best present dimly – in the minds of judges.

Bradley W Miller is an Associate Professor at the Faculty of Law, the University of Western Ontario.

(Suggested citation: B. Miller, ‘Some Problems in Constitutional Architecture’ U.K. Const. L. Blog (30th April 2014) (available at: