John Stuart Mill versus the Supreme Court of Canada

One of the great intellectual dustups among the intelligentsia in late Victorian English society was the controversy about the nature of the state’s legitimate power of compulsion.

In his classic On Liberty, John Stuart Mill wrote, “The only purpose for which power can be rightfully exercised over another member of a civilized community, against his will, is to prevent harm to others. His own good, either physical or moral, is not a sufficient warrant.” Among the most prominent opponents of this “neo-utilitarian” formulation was Sir James Fitz-James Stephen, who, appealing to old-school Benthamite utilitarianism, disagreed with Mill’s proposition that moral harm is not properly included in the category of harm to others. Stephen, whose view of the purpose of criminal law is aptly captured in his aphorism “the criminal law stands to the passion for revenge in much the same relation as marriage to the sexual appetite,” argued in his counterblast to On Liberty that “immoral behaviour” can properly be criminalized.

Although the debate was lively at the time, the long-term influence of the two antagonists and their views has been very different.

Given his Benthamite view that the state needs to give fair notice of the rules it intends to compel people to obey, Stephen soon turned his attention to the task of codifying various aspects of English common law, including the criminal law. Although a bill to enact his draft criminal code was introduced in Parliament, the measure was never passed, and Stephen’s influence on English legal and constitutional theory went into rapid decline. Were that the end of the story, Stephen would be remembered today, if at all, primarily as Virginia Woolf’s uncle. In the event, however, while Stephen’s project fizzled out in England, its impact was quite profound in Her Majesty’s overseas dominions, notably in Canada, where it served as the basis for the Criminal Code of Canada, which, despite extensive amendment, still defines criminal law in Canada today.

Mill’s theory, by contrast, had little tangible effect on the articulation of criminal law, either in its common law English form or its statutory Canadian version for the next hundred years. It did, however, continue to be an important influence on liberal democratic legal and constitutional thought, as well as a touchstone for the evolving doctrine of human rights (a concept to which Bentham famously referred as “nonsense upon stilts”). In 1982, the Canadian Charter of Rights and Freedoms was entrenched in Canada’s constitution, a development that not a few commentators saw as a realization of Mill’s principles.

Fast-forward to the early 21st century and what could quite plausibly be seen as a rematch by proxy between Mill and Stephen.

Mill’s Harm-Based Principle Takes on the Supreme Court of Canada

L.W. Sumner’s The Hateful and the Obscene: Studies in the Limits of Free Expression is self-consciously in the line of John Stuart Mill in its project to explore the limits of free expression. The task Sumner sets himself is twofold. He aims first to articulate a theory, based not on personal preferences but on “firm principles,” that is “capable of addressing the most important conceptual, moral, and political questions about expressive freedom and its limits.” He also aims to apply that framework to the specific challenges to free speech rights raised by two of the least popular and ostensibly least valuable examples of expression, namely hate speech and pornography.

The firm principles Sumner refers to constitute what he calls Mill’s Harm-Based Framework. This framework consists of two components. First is Mill’s harm principle itself, which rejects paternalism (prohibiting harmful acts consented to by willing adults) and moralism (prohibiting acts causing only moral harm or moral distress), insisting rather on clearly proven and tangible harm to the rights and interests of others as the sole justification for state intervention. The second component of the framework is the consequentialism principle, which dictates that before liberty is interfered with, even to prevent harm to others, it is necessary to conduct a cost/benefit analysis to prove that restricting liberty will yield a better balance of benefits over costs than unrestricted liberty.

The justification for applying this framework from Mill to the Canadian case law on freedom of expression under the Charter is that Sumner sees Mill as the “godfather” of the so-called “Oakes test,” the analysis the Canadian courts have adopted to assess infringements on Charter rights.

The reason that the Oakes test (and its presumed intellectual paternity) is important is that section 1 of the Charter guarantees those rights and freedoms “subject only to such reasonable limits prescribed by law as are demonstrably justifiable in a free and democratic society.” Just what the Charter guarantees, therefore, depends crucially on how courts approach the issue of what constitutes a reasonable limit. This is especially true with a right such as freedom of expression since the Supreme Court of Canada has defined “expression” in the broadest possible way to include anything intended to convey meaning.

The Oakes test, named after the case in which it was first articulated by Chief Justice Brian Dickson, says that any attempt to infringe on freedom of expression must satisfy two criteria. First, it must show that the legislation in question responds to a “pressing and substantial legislative objective” of a sort capable of justifying infringing a Charter right. In other words, this better be important. Second, it must show that there is a “proportionality” among the legislative objective, the means chosen to achieve it, the extent of the encroachment on the Charter right and the impact of that encroachment. In other words, balance is everything.

Sumner sees Mill as the godfather of this test because he sees its two steps as reflecting Mill’s Harm-Based Framework. He identifies the first step with the harm principle itself, presumably on the view that only legislative objectives aimed at responding to or preventing harm to others are potentially important enough to justify infringing on Charter rights. As for the second step, he views proportionality as involving a cost/benefit analysis of the infringement, echoing Mill’s consequentialism principle.

It is hard to overestimate the importance to Sumner’s project of his claim about the Oakes test. If his claim is correct, then the exercise of applying the Harm-Based Framework to the Supreme Court of Canada’s jurisprudence on hate propaganda and obscenity becomes more than just a Philosophy 300 essay question (“What would Mill have thought of the Supreme Court’s decisions on freedom of expression?”). It becomes an objective evaluation of how accurately the Supreme Court of Canada has applied the philosophical and analytical theory implicit in section 1 of the Charter to those cases. It leads, in other words, to apparently reliable conclusions as to whether the Supreme Court has gotten it right in deciding freedom of expression cases since 1982.

For Sumner, the upshot of the evaluation is that, with one exception (R. v. Zundel, where the Criminal Code offence of “spreading false news” was struck down), all of that jurisprudence is wrong. In his view, none of the infringements on expressive freedom in those cases should legitimately have survived a section 1 analysis if the Harm-Based Framework had been properly applied. Indeed, as Sumner’s catalogue of judicial errors unfolds, it turns out that even the father of the Oakes test, Chief Justice Dickson, got it consistently wrong by applying the test in a manner at odds with Mill’s Harm-Based Framework.

Specifically, Sumner argues that in almost all of the decisions concerning pornography, the prohibitions should not have survived a proper application of the “legislative objective” portion of the Oakes test because, contrary to the harm principle, they are aimed at either moral harm or moral distress or at (alleged) harm to consenting users.

As for the remaining decisions, mainly involving laws dealing with hate propaganda, Sumner concedes that there may be at least “potential harm” to others but, he argues, those laws should not have passed an objective analysis weighing the “real cost” of regulation against the “potential cost” to vulnerable groups of leaving the expression unregulated. In the final analysis, he concludes that the only form of expression whose suppression could be justified on both parts of the Harm-Based Framework (and, therefore, implicitly, by the Oakes test) is the explicit depiction of actual sexual activity involving children. And even here the Supreme Court was wrong to allow a much broader range of expressive materials to be prohibited under the category of child pornography.

By the end of his work, however, Sumner nevertheless sounds a pessimistic note about the prospects for his “blueprint for a society that strikes an appropriate balance” in regulating expression. Seeing no plausible political or legal means to reverse the Supreme Court’s (erroneous) pronouncements, he concludes: “Being neither a politician nor a lawyer, I have no way forward to propose. As a philosopher I must limit myself to mapping the promised land, leaving to others the task of getting us there.”

The Supreme Court of Canada Takes on Mill’s Harm Principle

One can only speculate how much more pessimistic Sumner might have been as a result of the Supreme Court’s decision in R. v. Malmo-Levine, released just weeks after the date of his preface to The Hateful and the Obscene. This case is not about political expression or pornography; rather, it deals w ith that other 1960s culture–related issue, the criminalization of marijuana (or, as it is quaintly spelled in the Narcotic Control Act, “marihuana”). (1) But the legal theory that Malmo-Levine relies on goes to the heart of Sumner’s hypothesis about the relationship between Mill’s Harm-Based Framework and the section 1 guarantees of the Charter.

David Malmo-Levine argued that a conviction for possession of marijuana violated the right guaranteed by section 7 of the Charter to “life, liberty and security of the person and the right not to be deprived thereof except in accordance with the principles of fundamental justice.” His claim was that one of the principles of fundamental justice is precisely Mill’s harm principle. Since criminalizing marijuana could only be justified on the basis of moralism or paternalism, it must fail Mill’s test of harm to others and should be struck down.

In other words, Malmo-Levine put before the court precisely the thesis of Sumner’s The Hateful and the Obscene, namely that the harm principle, and specifically its exclusion of moralism and paternalism, does or should constitute a normative legal principle in Canadian law.

And it is here that the godfather of the Criminal Code of Canada re-enters the fray.

True to Sir James Fitz-James Stephen’s view, Canadian criminal law is full of examples of offences based on moralism (for example, cruelty to animals, cannibalism, bestiality) and paternalism (such as duels between consenting adults). The Charter, on the other hand, is said to embody Mill’s contrary vision through section 7 (according to Malmo-Levine) and through section 1 (according to Sumner). If Sumner and Malmo-Levine are correct, then it should be irrelevant that many offences in Canadian criminal law—often dating back to Victorian days and Stephen’s codification—do accept moral harm as a basis for criminal sanctions. The Charter, as part of the Canadian constitution, is meant to be the supreme law of Canada and to prevail over any laws inconsistent with it. If Sumner and Malmo-Levine are correct in placing the harm principle at the heart of the Charter, then the number of laws that would have to be invalidated is no more relevant than the number of Supreme Court decisions on freedom of expression that had previously gotten it wrong. The harm principle ought to have trumped them all.

But are Sumner and Malmo-Levine right about the harm principle?

Significantly, the Supreme Court faces the issue head on. It does not simply side with Stephen because he got to the criminal law before Mill got to the Charter. Instead, the Court examines the harm principle on its own terms, using observations by modern-day followers of Mill, notably H.L.A. Hart and J. Feinberg (to whom, ironically, The Hateful and the Obscene is dedicated) to point out the limitations that the Court believes make it unsuitable to serve as a normative principle for Charter analysis.

The trouble with the harm principle is that, once you look at it carefully, it does not really have much explanatory power. If, as Sumner proposes, one gives harm the limited meaning proposed by Mill, then the definition is too narrow to be useful since, in Hart’s words, “the grounds for interfering with human liberty are more various than the single criterion of ‘harm to others’ suggests … there are multiple criteria, not a single criterion determining when human liberty may be restricted.” On the other hand, if one gives the concept of harm the broader meaning suggested by Hart, then, in the words of the legal theorist B.E. Harcourt, “the Harm Principle …effectively collaps[es] under the weight of its own success. Claims of harm … become so pervasive that the Harm Principle … become[s] meaningless: the Harm Principle no longer serves the function of a critical principle because non-trivial harm arguments permeate the debate. Today, the issue is no longer whether a moral offence causes harm, but rather what type and what amounts of harms the challenge conduct causes, and how the harms compare. On those issues, the Harm Principle is silent.”

In the final analysis, the Supreme Court does not reject harm as a constitutional prerequisite for valid criminal law. Even Stephen accepted that criminal law must be confined to acts and omissions that “inflict definite evils.” Based on Hart and others, the Court sides with Stephen’s formulation that such evils do not consist solely of harm to others but rather may be inflicted “either on specific persons or on the community at large.”

The Hateful and the Obscene in Light of Malmo-Levine

So where does all this leave The Hateful and the Obscene and its conclusions about freedom of expression?

The book uses Sumner’s Harm-Based Frame-work to reject the jurisprudence of the Supreme Court on expressive freedom. In Malmo-Levine, the Court rejects Mill’s harm principle, which lies at the heart of the Harm-Based Framework, as a fundamental concept underlying the Charter. Stripped of this justification, the Harm-Based Framework will therefore have to stand or fall on its own merits.

In considering those merits, it is interesting to reflect on an additional passage from Harcourt set out in Malmo-Levine that notes that it is the “hidden normative dimensions … [that] do the work in the Harm Principle, not the abstract, simple notion of harm.” In other words, the Harm-Based Framework is only as strong as these hidden “normative dimensions”—that is, its (often untested) value assumptions.

If it is true that what really does the work in the Harm-Based Framework are these value assumptions, then in The Hateful and the Obscene most of the heavy lifting is done offstage. This is especially true in Sumner’s application of the consequentialism principle. The book’s objective analysis of the respective costs of regulating speech and leaving it unregulated does consist of various factors on each side of the ledger, but its conclusions about the respective weight to be assigned to them are stated rather than demonstrated. This should hardly be surprising, since it is no more possible today to reduce such factors to a common currency for purposes of comparison than it was when Bentham proposed a “felicific calculus” to measure pleasure and pain. Ultimately, where the balance falls will be a factor of the very “hidden normative dimensions” to which Harcourt refers. Or, to put it another way, notwithstanding Sumner’s stated purpose, it is almost unavoidable that elements of “personal preference” find their way into the “firm principles” of his theory.

In assessing Sumner’s topography for his promised land, the reader is therefore ultimately thrown back on his or her own assumptions about what values the law can and should protect, how they stack up against one another, and—to use a concept of which Sumner would undoubtedly disapprove—how much of an incursion on these values societies can and should tolerate (Sumner rejects the concept that tolerance is a meaningful principle to apply to rights).

The normative assumptions that underlie Sumner’s costs/benefit analysis are not always explicit. Among the assumptions at or near the surface of this analysis are that the default position ought to be that the right to see and decide for oneself trumps any potential consequences resulting from that right, that exceptions to this default position must be justified on the basis of conclusive and scientific evidence on all counts, and that social values are irrelevant in such an assessment.

The trouble with these assumptions is that they are precisely that, assumptions. Accept them and Sumner’s description of a promised land in which interference with expression is almost never allowed will not only logically follow, but will likely seem a very desirable destination.

But because these are all assumptions, there remains plenty of room for scepticism and alternate assumptions. Some readers, for instance, may—contrary to Sumner’s views—believe that the preservation and protection of the social values at the core of a free and democratic society can justify restrictions and that there can be clear instances, perhaps reinforced by a social consensus, where line drawing is not only possible but desirable. For such readers, Sumner’s promised land—where the presumed costs to personal autonomy and to “robust debate” of restricting the interests of producers and consumers of violent pornography and racist incitement axiomatically outweigh the costs inflicted by the free circulation of these forms of expression—may seem somewhat less enticing. Those readers might, accordingly, find considerable solace in the fact that the Supreme Court of Canada appears to have said “you can’t get there from here.”

Notes