Balancing is the main method used by a number of constitutional courts around the world to resolve conflicts of fundamental rights. The European Court of Human Rights routinely balances human rights against each other and against conflicting public interests; it has elevated proportionality to the status of a basic principle of interpretation of the European Convention on Human Rights (ECHR). This paper examines the debate on balancing in the context of American constitutional law and the convention and discusses theories that claim some form of balancing is inherent in human rights adjudication. It argues that proportionality constitutes a misguided quest for precision and objectivity in the resolution of human rights disputes, and it suggests that courts should focus, instead, on the real moral issues underlying such disputes.
Balancing is in vogue outside of the United States. In Europe, Canada, India, South Africa, and elsewhere courts invoke balancing as the proper method of human rights adjudication. The European Court of Human Rights (ECtHR), by its own admission, routinely balances human rights against each other and against conflicting public interests and, in many countries, proportionality has been elevated, implicitly, to a basic constitutional principle. Not only have the nations beyond the United States embarked upon the age of balancing, but there is a marked complacency about it.1 There are theories that claim not only that balancing is the proper way of resolving human rights issues but also that it is the only way; this, according to such theories, is because the very concept of human rights implies balancing and is inseparable from it.2
Are we actually facing, here, a novel and radical method of human rights adjudication, or are we, on the contrary, reinventing old theories and concerns? I will argue that balancing, in the form of proportionality, is nothing but a manifestation of the perennial quest to invest adjudication with precision and objectivity. And, as such, it is vulnerable to some well-known arguments and concerns that the American debate on balancing has revealed.
The American debate on balancing
What is so appealing about balancing? It is a powerful metaphor that claims to capture, as a whole, the best method of decision making.3 According to this metaphor, rational people place, on one side of the scale, considerations in favor of a course of action and, on the other side, considerations against it, they weigh them, and then come up with a decision that follows from the outcome of this balancing process. The metaphor is sufficiently vague so as to include a great variety of reasons and human actions. Should I go to the movies tonight or not? In order to make up my mind and act accordingly, I will probably have to do some kind of reasoning. One way to describe this reasoning is to say that I balance the pros and the cons of going to the movies, and if the former outweigh the latter I will go, if not I will stay home.
In a sense, balancing appears to be a basic way of reasoning and, certainly, a basic way of practical reasoning. This seems plausible only on the assumption that every thought we have or choice we make is (or may be represented as being) in conflict with its opposite. The idea of everything in constant conflict with its opposite (something analogous to the Hegelian notion that every thesis has to be confronted with an antithesis) has the great appeal of simplicity and all-inclusiveness. Every course of action can be represented as the outcome of a choice between itself and other options. And every choice we make can be depicted as the resolution of a divergence among countless alternatives: going to the movies is in tension with a myriad other activities I could pursue. These tensions, especially those that concern courses of action, demand some kind of resolution.
Simplicity and all-inclusiveness are not the only appealing features of balancing. The metaphor also suggests precision. We weigh things and our decisions have the precision that weighing—as calculation—produces. The scale as a symbol of justice expresses the ancient and well-known quest of investing judicial judgments with the precision of natural sciences.
Why has American adjudication resisted the appeal of balancing? The reason is that it was tested in the context of a powerful right—freedom of speech—and was found problematic as a method of adjudication. In fact, the issue was openly debated in the setting of the First Amendment during the famous dispute between Justice Hugo Black, on the one hand, and Justices Felix Frankfurter and John Marshall Harlan II, on the other, over the meaning of freedom of speech. The dispute was often presented as one between absolutists against balancers. The stubborn insistence of Justice Black on the absolute character of the First Amendment was an easy target for the balancers who, confident that there was no such thing as an absolute right, were quick to reach the conclusion that balancing was unavoidable.
Justice Black, as well as many scholars, fiercely criticized the balancing approach and, in the process, revealed in many ways the ambiguity of the metaphor. Black and others variously pointed out that it is not clear what is weighed (interests, principles, rights, considerations); how it is weighted (with what metric); and who is doing—or should do—the balancing (judges or legislators).
To the question of what is weighted, Justice Frankfurter replied that it is interests that are weighted: “The demands of free speech in a democratic society as well as the interests in national security are better served by candid and informed weighing of competing interests, within the confines of the judicial process, than by announcing dogmas too inflexible for the non-Euclidean problems to be solved.”4 The same answer was given by Justice Harlan: “Where First Amendment rights are asserted to bar governmental interrogation[,] resolution of the issue always involves a balancing by the courts of the competing private and public interests at stake in the particular circumstances.”5
The view that constitutional rights are nothing but private interests whose protection depends, on each occasion, on being balanced with competing public interests, in fact, renders the Constitution futile. Indeed, if constitutional rights protect the same kind of interests as those of the government, and if the protection depends on considerations of some kind of relative “weight” given to the conflicting interests, it follows that the protection accorded by the Constitution can never be stable but is always conditional on various circumstances and depends on the outcome of balancing. On this view, not only is it doubtful whether the Constitution is the kind of law that includes stable and knowable propositions, but it also renders the very idea of such a constitution futile. Laurent Frantz has made both of these claims when he was provocatively asking if the First Amendment was law at all6 and, consequently, was asserting that “[t]he balancing test assures us little, if any, more freedom of speech than we should have had if the first amendment had never been adopted.”7
It should be noted that in the most simplistic version of balancing, there cannot be any concept of fundamental rights having priority over other considerations. Interests protected by rights find themselves in the scale on a par with any of the other interests that individuals or the government have. On this account, the interests of the majority tend to outweigh the interests of individuals and minorities. It is not surprising that, under the balancing approach, the outcomes of most free speech cases that involved communist speech during the Cold War were decided against freedom of speech.8
The critics of balancing never accepted the either/or framing of the issue, that is, rights either are absolute or balancing is unavoidable. They insisted that, without some kind of definition (categorization), the whole idea of rights is without any meaning, and they stressed that the ambiguity of balancing reaches every single element in the process: what the metric should be and who ought to do it.
The most effective critique of balancing concerns the assumption of a common metric in the weighing process. The metaphor says nothing about how various interests are to be weighted, and this silence tends to conceal the impossibility of measuring incommensurable values by introducing the image of a mechanistic, quantitative common metric.9 The only way to attempt introducing a common metric is to subscribe to some form of utilitarianism, namely, to a moral theory that assumes all interests are ultimately reducible to some shared metric (money or happiness or pleasure), and that, once translated into this common standard, they can be measured against each other. But, this would be a high-risk strategy. For one thing, it seems to make our theory of constitutional adjudication stand or fall on having the correct answer to an extremely vexing and controversial question in moral philosophy. Besides, if we decide to go down this road, we strip the balancing approach of much of its theoretical motivation. Arguably, balancing makes sense only against the backdrop of various conflicting values. If all values are reducible to a common metric, the problem that gave rise to the need for a balancing method dissolves.
Finally, a third point of criticism contests the legitimacy of judicial balancing. If we assume that human rights protection is the result of balancing interests, one wonders whether judges should perform it instead of legislators. What is the aim of judicial review? Is it to replicate or supervise the balancing of the legislators? Or is it constrained by an overarching requirement of judicial deference, as Justice Frankfurter, a keen balancer himself, was arguing?
And even more clearly:
Free speech cases are not an exception to the principle that we are not legislators, that direct policy-making is not our province. How best to reconcile competing interests is the business of legislatures and the balance they strike is a judgment not to be displaced by ours, but to be respected unless outside the pale of fair judgment.10
Primary responsibility for adjusting the interests which compete in the situation before us of necessity belongs to the Congress.11
If we are not going to take metaphors very seriously,12 then we must start by altogether rejecting the myth of mathematical precision. Whatever else it may be, it is quite certain that judicial reasoning has nothing to do with going to the grocer’s. Very few, if any, genuine and important values are amenable to any meaningful form of quantification. And, even if they were, balancing them would require, in addition, coming up with a way to compare their respective “weights,” which hardly anyone but the most hard-nosed utilitarian would think is more than a chimera. In this sense, Justice Scalia is merely scoring an easy point, when he says that we cannot compare the length of a line with the heaviness of a rock.13
It is important to circumvent a possible misunderstanding, at this point. Scalia's statement might be taken to suggest that values are incommensurable in the sense that we could never adjudicate between them rationally or, put differently, that we could never have rational grounds for preferring one over the other. This is not the position I will be defending. Rather, I want to side with Jeremy Waldron who argues that belief in such a “strong” incommensurability would lead to a total agnosticism about morality, which goes against our most strongly held intuitions regarding morality and the point of moral reasoning. Instead, Waldron has argued in favor of a “weak” incommensurability, which, while it acknowledges the lack of a common metric for “balancing,” nevertheless permits us to bring values into relation with each other.14 He suggests that we do this intuitively when, for example, we say things like: “Any reasonable person can see that saving an innocent child from a painful death is to have priority over the preservation of the statue that has fallen on top of her.”15 He also suggests that we do it by reasoning, that is, when we relate values, when we propose ways of ordering them and of putting them into a system. That is exactly, says Waldron, what John Rawls, Ronald Dworkin, or even Robert Nozick do, when they insist on the lexical priority of basic liberties, on rights as trumps, or as side constraints.16
The idea of putting values in an order and assigning them priorities is also a way of reasoning about more ordinary courses of action. I do not go to the movies when I have a class, and there is no balancing taking place in this case. Going to the movies is ruled out simply because having a class enjoys a higher priority; put another way, it trumps the consideration of going to the movies. Now, one could describe the reasoning involved in relating values or having priorities as a kind of balancing. In fact, Waldron maintains that “often when people talk about weighing or balancing one value, principle, or consideration against another, what they mean is not necessarily Benthamite quantification but any form of reasoning or argumentation about the values in question.”17 He goes on to say that for “most ordinary people” elaborate moral arguments, like those of Dworkin or Rawls, seem like balancing. And because our moral reasoning certainly includes considerations in favor or against an argument his conclusion is that “the reasoned articulation of our moral principles and priorities inescapably involves what ordinary people might regard as weighing and balancing.”18
The question is whether “weighing” or “balancing,” even taken in a broad sense, are of any value in characterizing human rights adjudication. In theory—and keeping in mind that these are just metaphors—there is no reason to exclude them as shortcuts for describing the judicial process. However, in practice, the term “balancing” has become tantamount to the principle of proportionality, that is, a specific judicial test that pretends to balance values while avoiding any moral reasoning. In fact, the principle of proportionality, bypassing any discourse on priorities, pretends to resolve value conflicts by assessing the degree of their relative coexistence. Human rights values, for example, do not enjoy any priority vis-à-vis other public interests. They compete on a par with them.19 Such rights may prevail when the public interest can be attained with a less restrictive measure, but they may be curtailed when the measure seems proportional to the objective.
More precisely, the principle of proportionality consists in a three-prong test that assesses (a) whether a measure that interferes with a right is suitable for achieving its objective, (b) whether it is necessary for that purpose, and (c) whether it burdens the individual excessively compared with the benefits it aims to secure.20 Since it is only rarely that measures are completely irrational, and it is always possible to argue that they are suitable and necessary to accomplish a legitimate aim, a measure fails only very occasionally on the first two counts. Essentially, the proportionality test is reduced, more often than not, to measuring the relative intensity of the interference with the importance of the aim sought. The principle of proportionality assumes that conflicts of values can be reduced to issues of intensity or degree and, more importantly, it assumes further that intensity and degree can be measured with a common metric (something like a natural force), and that this process will reveal the solution to the conflict. Thus it pretends to be objective, neutral, and totally extraneous to any moral reasoning.
Now, while one might accept the idea of the commensurability of values within the ambit of a moral discourse, there is no way to accept the notion that values are commensurable without a moral argument, that is, an argument that relates them and justifies degrees of priority. If the moral discourse is lacking, there is no way to demonstrate that values, indeed, are commensurable, and it makes no sense, therefore, to pretend that the principle of proportionality allows us to do it.
That is the reason, it seems to me (and setting aside the “strong” incommensurability of values), why we should not accept balancing—even in the loose sense Waldron proposes—as the method of human rights adjudication. Moreover, this method is, by its very nature, figurative. The imagery of balancing unavoidably carries with it connotations of mathematical precision or, at any event, seems to allude to some kind of quantification,21 Benthamite or other, and thus tends to neglect any moral reasoning. The result is, as the principle of proportionality proves, to impress on us an illusion of some kind of mechanical weighing of values similar to that of weighing apples and oranges.
The adoption of a balancing test, according to the principle of proportionality, risks neglecting the complexity of moral evaluation and, especially, the complexity of rights. More specifically, it tends to overlook, or at least not appreciate adequately, the fact that our moral universe includes ideas not amenable to quantification, with the result that these ideas are not given due consideration in our reasoning. Most importantly, as I am going to argue, among the moral concepts that this kind of balancing is likely to distort are fundamental individual rights.
These are strong claims, and the only way to defend them is to turn our attention to real cases to see how balancing, according to the principle of proportionality, unfolds. For this purpose, there is no better starting point than the jurisprudence of the ECtHR, which is engaging in precisely this kind of balancing.
By definition, any treaty for the protection of human rights gives priority to rights. Its goal is to protect certain individual fundamental interests not only from arbitrary state power but also from collective interests. So, although accurate, it sounds somewhat strange to say, as did the former president of the ECtHR Rolv Ryssdall, that “[t]he theme that runs through the Convention and its case law is the need to strike a balance between the general interest of the community and the protection of the individual's fundamental rights.”22 The former president was simply repeating, almost verbatim, the dictum of the Court that “inherent in the whole of the Convention is a search for a fair balance between the demands of the general interest of the community and the requirements of the protection of the individual's fundamental rights.”23 There is no doubt that the ECtHR engages in a balancing approach both as a method of interpretation and as a method of adjudication. This balancing approach, under the rubric of the principle of proportionality, “has acquired the status of general principle in the Convention system.”24
Now, one should expect that the rights convention itself reflects such a balance, the outcome of which must be that human rights are to be protected before other interests are even taken into consideration. If that is so, what does it mean to say that the issue is to strike a further balance between the general interest of the community and individual rights? The obvious answer is that the vast limitations contained in articles 8 to 11 (the rights to respect of private and family life, home, and correspondence; the right of freedom of thought, religion, and conscience; the right of speech; and the right of association and assembly), namely, the restrictions necessary in a democratic society for the protection of public security, safety, protection of public order, health or morals, and the rights and freedoms of others,25 give rise to new considerations and balancing. The concept of the restrictions necessary in a democratic society is supposed to lead to the principle of proportionality, that is, a balancing approach that requires the intensity of the restriction not to be excessive in relation to the legitimate needs and interests, which the specific restriction aims to redress. “The scale the Court utilizes seems to imply that the more far-reaching the infringement or more essential the aspect of the right that has been interfered with, the more substantial or compelling the legitimate aims pursued must be.”26
There are, at least, two controversial assumptions underlying this approach: first, that public interests, as a matter of principle, can always be weighted against human rights; and, second, that measures aimed at promoting a public interest may prevail unless they impose an excessive restriction compared to the benefit they secure (the violation seems to depend rather on the intensity of the restriction than on its incompatibility with the right in case).
Let us see what the impact of these assumptions was in two cases, Otto-Preminger-Institut v. Austria27 and İ.A. v. Turkey,28 both of which concerned freedom for blasphemous speech. I find these cases paradigmatic because, although there was disagreement between majority and minority on what should be put on the scale (rights or interests), there was, nevertheless, agreement that a scale should be used and the principle of proportionality should resolve the cases.
In the Otto-Preminger-Institut case a private nonprofit art cinema complained about a violation of article 10 of the convention because the Austrian authorities, at the request of the Innsbruck diocese of the Roman Catholic Church, had seized and confiscated a film that was scheduled to be shown to the public. The film Das Liebeskonzil (Council in Heaven) was based on a play written by Oskar Panizza in 1894 that portrayed God, Christ, and the Virgin Mary plotting with the devil how to punish mankind and deciding to infect human beings with syphilis. The devil's daughter assumes the task of spreading it to the worldly and powerful, to the court of the pope, to the bishops, to convents and monasteries, and, finally, to the common people. Panizza was found guilty of “crimes against religion” and was sentenced to a term of imprisonment in 1895 in Germany. However, recent productions of the play were performed and the film actually reproduced one performance that took place in Rome with the addition of some small parts in the beginning and the end of the film, along with comments about the trial of Panizza. The film depicted God, Christ, and the Virgin Mary in a demeaning way and also contained some erotic scenes and innuendos. The seizure and confiscation in Austria were based on article 188 of the Austrian Penal Law that punishes the “disparage[ment] of a dogma, a lawful custom or a lawful institution of [a] church or religious community.”
The court (with a majority of six out of nine) held that there was no violation of freedom of speech. At the outset, it examined whether the seizure and confiscation of the film constituted interference in pursuit of a “legitimate aim.” The court found that these measures were meant “to protect the right of citizens not to be insulted in their religious feelings by the public expression of views of other persons,”29 and, thus, it came to the conclusion that the impugned measures pursued a legitimate aim under article 10 (2) of the convention, namely, “the protection of the rights of others.” The court then proceeded to examine whether the measures were “necessary in a democratic society.” It referred to its case law on freedom of speech and its finding that this freedom includes not only “information” or “ideas” that “are favourably received or regarded as inoffensive or as a matter of indifference, but also … those that shock, offend or disturb the State or any sector of population” (Handyside v. United Kingdom judgment of December 7, 1976). However, the court went on to stress that those who exercise their freedom of speech also undertake duties and responsibilities and “among them—in the context of religious opinions and beliefs—may legitimately be included an obligation to avoid as far as possible expressions that are gratuitously offensive to others and thus an infringement of their rights, and which therefore do not contribute to any form of public debate capable of furthering progress in human affairs.”30 Finally, having established that states may sanction improper attacks on objects of religious veneration, the court proceeded to a final exercise in balancing by examining whether the seizure and the confiscation of the film were restrictions proportionate to the legitimate aim pursued.
On this final balancing the court was not unanimous. The majority dismissed various arguments to the effect that many precautions were taken to prevent offending the feelings of the believers. The film was to be screened in a cinema and was addressed to a specific audience interested in avant-garde culture; the public was to pay for a ticket to see the film; persons under seventeen were not admitted; and there was an information bulletin helpfully describing the theme of the film in detail,31 so there was no danger of anyone's being exposed against his will to material he would find offensive. The majority reasoned that because the film was advertised and precisely because there was adequate public knowledge of its content, the expression had been made “sufficiently” “public to cause offence.” Without elaborating, the majority accepted the judgment of the Austrian courts that the film lacked any artistic merit that could outweigh offense to the public. The majority went on to point out that the Roman Catholic religion is the religion of the overwhelming majority of Tyroleans (87 percent), and it thought that the authorities had not overstepped their margin of appreciation by seizing the film and in wanting to ensure religious peace and to secure some people from feeling offended. On the other side, a minority of three judges out of nine was of the view that the seizure and confiscation of the film, far from being the less restrictive solution, amounted to a complete prevention of freedom of expression, which could be accepted only if the speech was so abusive as to come close to a denial of the freedom of religion for others. Arguing that “there was little likelihood […] of anyone being confronted with objectionable material unwittingly,”32 the minority found that “on balance […] the seizure and forfeiture of the film in question were not proportionate to the legitimate aim pursued.”33
In what follows I wish to focus on two problematic aspects of this decision. The first has to do with the specification of the items that the court put in the balance and the second aspect with the way the “weight” of those items was compared in the balancing exercise. Let us take each of these aspects in turn.
What is to be compared?
Before we discuss the balancing stage, we should consider how the court has structured the case up to that point. There is no need to dwell on the question of whether there was interference in the first place. No one could deny that there was an obvious (and I would say brutal) interference with the applicant's speech rights. What is more interesting is to examine how the court next inquired whether the purpose of this interference was formally included in the vast categories of restrictions listed by article 10 (2). The court seems to have treated this as little more than a kind of formal inquiry, as mere taxonomic exercise. Thus, for the majority, the interference fell under the “protection of the rights of others” restriction. The minority, by contrast, pointed out that “[t]he Convention does not, in terms, guarantee a right to protection of religious feelings. More particularly, such a right cannot be derived from the right to freedom of religion, which in effect includes a right to express views critical of the religious opinion of others.”34 However—and this is very important—although the minority rejects the idea of a right to have one's religious feelings protected, it does not have any difficulty accepting the proposition that such protection is “legitimate” since “the democratic character of a society will be affected if violent and abusive attacks on the reputation of a religious group are allowed.”35 Thus, regardless of whether limitations of the kind in question are premised on a right or not, both sides agree that “it is necessary in a democratic society to set limits to the public expression of such criticism or abuse.”36
Is it so trivial to affirm or deny the existence of a right? Does it make so little practical difference whether we decide to ground a limitation of speech rights on a public interest or on a competing right? Perhaps for the balancers it does, since the methodology they recommend will be the same, whichever way we go. But the truth is that at the level of moral theory, at least, we do attach great importance to rights claims, and we do want to distinguish such claims from claims based on mere public interest; therefore, before we go along with the balancers’ suggestion, we should pause to think.
Let us see, then, how someone could come to the conclusion that there is a right of protection for religious feelings. The majority inferred it from the right to freedom of religion; however, since it did not elaborate on its reasons for thinking so, we have to reconstruct them ourselves for the sake of argument. Here is how the claim might go. One might say that since I am free to believe in some religion and since religious beliefs typically arouse strong feelings, I should be protected, in some fashion, from verbal attacks against my religion. If I am not protected, such attacks will hurt my feelings and, hence, impede my religious life. Of course, against this line of argument we can assert, with the minority, that freedom of religion includes the right of others to advocate their own religion and thus express views critical of my own religious beliefs. The mere fact that there are people who do not share my religious beliefs may hurt my feelings, but, obviously, I cannot seek any protection against this sort of discomfort without denying others their freedom of religion.
Considerations of the same kind apply to other feelings we may have. We may, for example, have strong feelings about certain political ideas, and contrary opinions may deeply hurt our feelings. I may be deeply distressed (even terrorized) by advocacy of the dictatorship of the proletariat, but I cannot lay claim to be protected against this kind of distress if I am willing to accept a right to free speech. I may have strong feelings about a person. I may be in love with Jennifer Lopez, for instance. Still, the strength of my feelings for her does not entitle me to any special protection. For instance, it does not give me the right to demand the press stop gossiping about her or publishing provocative photographs.
Of course, one can imagine cases where verbal attacks against one's religious feelings may constitute a genuine burden on the exercise of freedom of religion. Imagine the following situation: a group of nonbelievers parades every day outside a church shouting hostile slogans against the religion of the believers inside. But, as so often in law and morality, context is everything. In the example just given, our moral reaction stems not from the mere fact that someone holds views that can hurt the religious feelings of other people but, rather, from the circumstances in which these views are actually expressed, with the apparent purpose of intimidating the believers.
This is no more than a rough outline of a much more complex argument challenging the claim that religious feelings in themselves give rise to a right, on the part of those who hold them, to be protected from the expression of views that may hurt them. My aim was to show that the existence of such a right must be premised on certain assumptions—themselves contestable and in need of argumentative support—about what is worthy of being included in the ambit of a right. When I say that such assumptions stand in need of justification, I mean that they must draw on broader conceptions of the nature of rights and of how an alleged right must fit with other rights recognized in the convention and with the more general moral principles we happen to hold. These assumptions may prove mistaken (as I think they are in the case of an alleged right to have one's religious feelings protected), with the result that the case for the existence of a certain right must fail.
Now, this form of reasoning lies in stark contrast to the majority's rather cavalier approach toward the meaning of freedom of religion. Regardless of how we choose to characterize it, however, the majority's approach is in line with one of the basic methodological principles of the balancing approach, which we may call the “principle of definitional generosity.”37 According to this principle the interpreter assumes a broad definition of what can conceivably count as an instance of the exercise of a certain right. He asks: What can count as expression? What can count as religion? Value judgments about the importance of a right or the salience of one form of its exercise may inform this stage, though not necessarily in any particularly demanding way. The interpreter's purpose, here, is merely to assess whether a given act or behavior will be prima facie included within the ambit of a provision safeguarding, say, freedom of expression or freedom of religion. Since the threshold is not demanding, the normative implications that such a specification of a right carries with it are correspondingly limited. The interpreter can be generous at the stage of specification, safe in the knowledge that all the crucial normative issues may be deferred to the balancing stage.
But is he really safe? After all, if there is no such thing as a right to have one's religious feelings protected, then it makes no sense to speak of balancing in the first place, since we would be lacking that against which we are supposed to balance freedom of speech. This I take to be an embarrassing implication of the balancing method. In response, the balancer can always point to the strictures of the balancing stage as his safety net, but if the balancing stage is supposed to remedy a confusion that the balancer's approach itself has engendered, you might start asking whether it would be better to scrap the approach altogether.
At any rate, as we have said, the issue of whether or not the protection of religious feelings was a matter of right did not seem, in any case, to make much difference in Otto-Preminger-Institut, since the minority considered that such protection constituted a public interest worth balancing against the right to freedom of speech. So, let us now examine whether the principle of definitional generosity is more at home in the specification of the concept of public interest.
While we are familiar with the idea that there are different theories about rights, we sometimes pay little attention to the fact that there are also different theories about the concept of public interest.38 The reason for this is that we assume the public interest is the interest of the majority and, hence, we can tell whether something is in the public interest just by looking at that for which the elected representatives of the people have voted. The court seems to favor this understanding when it assumes that the interest of 87 percent of the Tyrolese not to be offended in their religious feelings constitutes a public interest, stressing that it “cannot disregard the fact that the Roman Catholic religion is the religion of the overwhelming majority of Tyrolese.”39
But suppose, for a moment, that 87 percent of the Tyrolese hated the Eskimos. Suppose that, when Eskimo plays are staged or Eskimo films screened, the “overwhelming majority” of Tyrolese feel stirred by violent feelings of moral indignation and uncontrollable fear. Would we be willing to include protection of these feelings within the ambit of public interest? If not, it is probably because we have to be more discriminating in our specification of what counts as public interest. We may, for example, not want external preferences to count, that is, preferences people have not regarding themselves but preferences they have about how other people should be treated.40 Unsurprisingly, the set of assumptions we need to bring to bear in this exercise are very similar to the assumptions driving our specification of rights. Our conception of the public interest also must incorporate or flow from normative ideas about the relationship between the individual and society, the importance of rights in structuring this relationship, and so forth.
Balancing in the strict sense
Earlier, I said that the balancing stage is the balancer's last ditch. But considering its importance within the balancing methodology, it is rather surprising to observe the dearth of argument that supports the court's balancing exercise in Otto-Preminger-Institut. Admittedly, once you jettison the idea that values are quantifiable and concede that the talk of “weight” is no more than a metaphor, it is hard to imagine what shape arguments at the balancing stage must take or, put otherwise, how we should tailor arguments to fit the balancing methodology.
One of the putative advantages of the balancing approach is its rigor. But rigor is one thing, and elegant formal structures are quite another. Otto-Preminger-Institut amply demonstrates that the balancing approach fails, spectacularly, to deliver what it promises. At the very least, we would expect that the balancing approach would throw some light on the “black box” of comparisons among weakly incommensurable values. What we find, instead, is a characteristically impressionistic assessment of the relative weights of competing considerations, which does not lend itself to a rational reconstruction of the argumentative path that has led to a particular decision. The reasoning is terse and fails to identify the contribution that different considerations make to the outcome.
The preceding analysis suggests one possible explanation for this opacity. According to the principle of definitional generosity, it is perfectly conceivable that certain items will make their way into the balancing process that are not genuine. Go back to the Eskimo example. I claimed in my analysis of that example that the preferences of the Eskimo-haters should not be taken into account at all. But a balancer, arguably, would let them play out in the balance. How, then, would he assign a value to such preferences? Presumably, he would assign them a very low value that would make them easy to override by way of competing considerations. But that sounds hopelessly ad hoc. It is not that the preferences of the Eskimo-haters should count or only count for little. It is that they do not count at all.
In fact, even in those cases where the court does attempt to specify, with more precision, the distinct contribution of different considerations, the result it reaches is far from self-evident. Take the following example. The minority in Otto-Preminger-Institut held that, while some restrictions might be thought necessary in order to further the stated public interest (protection of religious feelings), nevertheless the measures in question (seizure and confiscation of the film) restricted the applicants’ freedom of speech in a manner disproportionate to the benefit thereby achieved. The minority, therefore, meant to suggest that although, in principle, restrictions on freedom of speech for the protection of religious feeling are legitimate, they ought not to go too far. If a less severe restriction can achieve the same goal, it must be preferred.41
At this point, however, one might wonder what a less severe restriction would look like. Here's one suggestion. The minority seemed to favor taking precautions with regard to the time and manner of expression as preferable to seizure and confiscation. But if we take the offense to religious feelings to stem from the mere knowledge that some people are engaging in this kind of speech, then no precautions concerning the time and manner of expression can cure it. The idea that some people may depict my God in a demeaning way can hurt my feelings, whether they do it in private or in public. The only way to be protected from such an offense, it seems, is to restrict that kind of speech altogether. This is not, of course, to suggest that the majority approach is preferable. Rather, it serves to point out that the problem lies less with the severity of the restriction and more with the justifiability of imposing a restriction, in the first place, on the grounds that it offends someone's religious feelings. By deferring all the crucial judgments to the final stage, the balancing approach clouds the real problem and provides crude resources to resolve it.
One might want to recall the famous Cohen v. California42 case, where the Supreme Court reversed a conviction for disturbing the peace by “offensive conduct.” The offensive conduct of the appellant consisted of being in a courthouse corridor wearing a jacket bearing the words “Fuck the Draft.” Justice Harlan, who wrote the majority opinion, a keen balancer himself, dismissed the possible annoyance of people confronting the four-letter word. “Those in the Los Angeles courthouse,” he said, “could effectively avoid further bombardment of their sensibilities simply by averting their eyes.”43 Gerald Gunther has praised Harlan for his balancing approach, 44 but, whatever his balancing technique was, it certainly did not take into account any claim of people aiming to avoid disturbance at the sight of the controversial slogan. Advising these people to avert their eyes meant, simply, that they did not have any claim whatsoever that could carry weight on the balancing scale. In fact, if their claim enters the scale, it is not at all obvious that the appellant's right of free speech will prevail.
The İ.A. v. Turkey case: Proportionality unraveled
In our analysis of Otto-Preminger-Institut, we pointed out the failure on the part of the court to articulate, carefully, the competing considerations and their normative import for the determination of the outcome. I believe that this failure is far from restricted to that particular case. Rather, it permeates the ECtHR's methodological approach. To illustrate this, I want briefly to consider a more recent case, İ.A. v. Turkey,45 where the restriction on freedom of speech for the protection of religious feelings was, again, at issue. The case is interesting not because it brings out novel aspects of the issue but, rather, because it shows how the balancing approach can unravel and produce decisions that are hardly recognizable as the adjudication of human rights questions.
The applicant, a publisher, had published a novel (entitled The Forbidden Phrases) with a print run of two thousand copies. The content of the book contained various critical remarks about religion, in general, and the Muslim religion, in particular. The most provocative passage was the following:
Some of these words, moreover, were inspired in a surge of exultation, in Aisha's arms… . God's messenger broke his fast through sexual intercourse, after dinner and before prayer. Muhammad did not forbid sexual relations with a dead person or a live animal.46
The applicant was indicted on the basis of blasphemy (according to a Turkish law that punishes blasphemy “against God, one of the religions, one of the prophets, one of the sects or one of the holy books”) and was convicted and sentenced to two years’ imprisonment and a fine. The Turkish courts commuted the prison sentence to a fine, so that the applicant was obliged, ultimately, to pay a total fine equivalent to sixteen U.S. dollars.
A majority of the court (four out of seven), relying on its previous rulings in Otto-Preminger-Institut, constructed the case as a clash between two fundamental freedoms, “namely the right of the applicant to impart to the public his views on religious doctrine, on the one hand, and the right of others to respect for their freedom of thought, conscience and religion, on the other hand.”47 The majority “therefore” explicitly engaged in balancing. It held that there was no violation of freedom of speech because the law of blasphemy was a measure intended to provide protection against offensive attacks on matters regarded sacred by Muslims and, thus, was a reasonable measure meeting a “pressing social need.”48
While there is nothing new in the reasoning (except maybe it shows how far down the slippery slope can descend), it does include some striking thoughts about the fine imposed. Thus, the majority said:
As to the proportionality of the impugned measure, the Court is mindful of the fact that the domestic courts did not decide to seize the book, and accordingly considers that the insignificant fine imposed was proportionate to the aims pursued.49
I am not saying that the severity of the penalty should never be a consideration in moral and legal reasoning. In fact, in some cases it makes all the difference in the world.50 However, when we are preoccupied with this form of exercise, we risk losing sight of the battles of principle with which human rights law is so intimately intertwined. We also risk losing sight of the characteristic moral attitude that recognition of a right is supposed to display, and the message it is supposed to convey. Imagine, by way of contrast, what attitude toward individuals the following statement displays: “Why do you make so much fuss over sixteen dollars?”
It may, of course, be objected that we can always discard this piece of the court's jurisprudence as a grotesque mistake. But even if reference to the fine imposed in İ.A. were thought an unfortunate mistake, it is still true that balancing according to proportionality must feel more at home with considerations that seem amenable, at least prima facie, to some sort of scaling, such as the penalty's degree of severity. As a result of such thinking, adherents of this approach will tend to privilege such considerations or, at any rate, assign them a role in the reasoning process that they would otherwise lack.
Moreover, we can find traces of this type of problematic reasoning in other cases as well. Thus, in F. v. Switzerland,51 the court had to evaluate a law imposing a temporary prohibition on remarriage. The Swiss government defended the law on the basis that it gave people time to reflect before committing themselves to a new marriage. The court was not convinced that this argument could apply to “a person of full age in possession of his mental faculties.”52 This reasoning seems to suggest that the court rejected out of hand the paternalism underpinning the measure. However, the court subsequently muddied the waters when it concluded that “the disputed measure, which affected the very essence of the right to marry, was disproportionate to the legitimate aim pursued.”53 Should we infer from this that, in the court's view, the temporary prohibition of remarriage served a legitimate aim, and that its sole concern was the disproportionate impact on the applicant's rights? Would the measure be allowed to stand, if, while still paternalistic in spirit, it prescribed a shorter temporary prohibition?
A similar ambivalence affects the court's treatment of a number of related issues, notably the scope of the margin of appreciation. This can be seen, for example, in Hirst v. UK,54 which concerned interference with the ECHR right to vote of criminal offenders (article 3 of protocol 1 of the convention). The court largely bypassed the question whether negating the right of criminal offenders to vote for the purpose of “enhancing civic responsibility and respect for the rule of law” and “[conferring] an additional punishment” were aims compatible with the recognition of the right.55 Instead, it based its judgment—that there had been a violation of the right to vote—on the fact that the measure impugned, being a “blunt instrument,” fell way outside the margin of appreciation accorded the contracting states. The court continued: “It [the measure] strips of their Convention right to vote a significant category of persons and it does so in a way which is indiscriminate. The provision imposes a blanket restriction on all convicted prisoners in prison. It applies automatically to such prisoners, irrespective of the length of their sentence and irrespective of the nature or gravity of their offence and their individual circumstances.”56 Arguing in this manner meant that the court had withdrawn from a battle regarding the general principle without a fight and had focused, instead, on the more quantifiable aspects of the case such as “the general, automatic, and indiscriminate” nature of the restriction.57
Similarly, in Hatton v. UK,58 a case where the applicants complained that the increase in noise levels from night flights at Heathrow airport deprived them of sleep and thus was a violation of article 8 (respect for home, private and family life), the court was satisfied with the assessment of the balance struck by the national authorities. This balance weighed the conflicting interests of the individual and the economic welfare of community, as a whole, and came out in favor of the latter. By framing the discussion in terms of interests and relative weights the court was able to dodge a number of the difficult questions the case raised: Do the applicants have a actual right to sleep? If so, should it be given priority over the interests of the community? If not, is it the court's business to address the applicant's claim in the first place?
Putting human rights back in focus
The balancing approach, in the form of the principle of proportionality, appears to pervert rather than elucidate human rights adjudication. With the balancing approach, we no longer ask what is right or wrong in a human rights case but, instead, try to investigate whether something is appropriate, adequate, intensive, or far-reaching.59
This is true not only of the jurisprudence of the ECtHR. It is characteristic of the thinking of advocates of proportionality around the world. It is exemplified, too, in the work of David Beatty. In his recent book, he has launched a comprehensive and wide-ranging defense of the principle of proportionality as the cornerstone of constitutional adjudication. He has gone so far as to suggest that the best way to interpret the issue at stake in the landmark desegregation case of Brown v. Board of Education is to see it in terms of a conflict between the harm inflicted on black children from segregation and the harm inflicted on white children from integration. He writes: “Telling black children they cannot be educated in the same schools as white students is brutally offensive to their dignity and self-worth in a way that forcing whites to share their class-rooms is not. Segregationists may be deeply offended by having to mix with people with whom they want no association, but their stature and status in the community is not diminished by their forced integration.”60 If we take Beatty's words at face value, the reason why desegregation was required by the U.S. Constitution is that the harm to black children outweighed the harm to whites. It would seem to follow from this that if the loss of the whites’ sense of superiority (or self-esteem) was greater than the blacks’ loss of self-worth, the outcome would be different.61 But this would be an absurd and extreme conclusion that goes against our basic intuitions concerning the point of human rights. It erodes these rights’ distinctive meaning by transforming them into something seemingly quantifiable.
This is due to the very methodology the balancer employs. First, the balancing method does not pay sufficient attention to the specification of the items it purports to place in balance. It rests content with a prima facie specification of the ambit of a human right or of the public interest that is set against it. I remarked, earlier, that this strategy is grounded in the principle of definitional generosity. The motivation behind this principle is that by keeping an open mind about what is to go in the balance, you avoid excluding some claims from the outset, and, hence, you do not unduly restrict the range of claims you undertake to consider. But, in proceeding this way, the balancing approach trades inclusiveness for superficiality. The proper specification of the content of a human right is a specification guided by an understanding of its importance, which is the point in awarding it this unique status; such a specification needs to be sensitive to the important evaluative questions that recognition of a right entails. This involves coming to terms with what we value about that right and firmly placing the right in the constellation of our other political and moral values. In short, it involves a good deal of moral reasoning. This reasoning is likely to be lost when our analysis, at the first stage, is insufficiently fine grained.
One particularly striking way in which the principle of definitional generosity fails to capture the importance of the items it places in the scales is by not weeding out, at the first stage, interests and preferences powered by what I would characterize as illicit justifications. There are some types of justification that are not just less weighty than the right with which they conflict. Rather, their invocation is incompatible with the recognition of that right. An illicit justification goes against the very core of what it is that we value in the right. Freedom of speech, which was at issue in Otto-Preminger-Institut, provides a useful illustration of this point. Before we decide to balance the protection of religious sentiment against freedom of speech, we have to examine whether this goal can ever be a ground for prohibiting freedom of speech. My reason for believing that such a prohibition can never be accepted is not because, in such cases, religious sentiment loses out in its comparison with freedom of artistic expression; rather, a part of what we hold dear about freedom of expression is fatally compromised whenever the state prohibits one view in order to support another. Religious sentiment and freedom of expression can never be put on the scale, whatever we take that scale to be. The balancing approach, by contrast, reduces conflicts between rights or between rights and the common good to comparisons of relative weights and thus overlooks the justification-blocking function of rights.
Now, of course, this is a controversial claim. Reasonable people (like the majority in Otto-Preminger-Institut) would reject the view that religious feelings cannot ever be protected against irreverent speech. Thus, they would argue that the balancing approach has the advantage of bypassing this disagreement, without denying any claim, however frivolous, its day in court. Weak claims, they would go on, are adequately dealt with at the balancing stage, since they will not carry much weight and thus be easily overridden.
In response, the following may be said: Even if there is room for reasonable disagreement in the case of protecting religious feelings, there are other cases that self-evidently fit into the category of illicit justification. My example was the feelings of the Eskimo-haters. Does it make sense to say that feelings like those be allowed to play out at the balancing stage? To say that it does is to miss the distinctive moral status that a rights claim presupposes and affirms. We could say, following Dworkin, that this is the status of being entitled to equal concern and respect, or, following Thomas Nagel, that it is the status of inviolability.62 However we decide to characterize such a claim, we have an intuitive understanding of its implications in political argument; it removes some issues from the table, or it trumps competing considerations.63 The balancing approach, by contrast, is committed to a view whereby everything, even those aspects of our life most closely associated with our status as free and equal, is, in principle, up for grabs. This is echoed in Robert Alexy's famous distinction between rules and principles. Alexy writes that “[r]ules are norms that, given the satisfaction of specific conditions, definitively command, forbid, permit, or empower. Thus they can be characterized as ‘definitive commands.’ …” “Principles … are commands to optimalize… . They are norms commanding that something must be realized to the highest degree that is actually and legally possible.”64 Principles are optimization requirements; they can be satisfied to varying degrees, which depends on the legal and factual possibilities, while rules are always either fulfilled or not.65
In a sense, the İ.A. decision can be characterized as an optimization enterprise. The “light” interference (an insignificant fine) still leaves freedom of speech somehow in place while, at the same time, it serves public interest; in other words, it optimizes the competing values. This idea assumes that human rights guarantee degrees of liberty; the more liberty they guarantee the more the right is affirmed. On the other hand, less severe interferences are not negating the right altogether but, rather, accommodate public interest.
This view leads to a complete erosion of the notion of human rights. It overlooks the idea that human rights are not merely quantities of freedom but protect some basic status of people as moral agents. As Dworkin puts it: “If we have a right to basic liberties not because they are cases in which the commodity of liberty is somehow especially at stake, but because an assault on basic liberties injures us or demeans us in some way that goes beyond its impact on liberty, then what we have a right to is not liberty at all, but to the values or interests or standing that this particular constraint defeats.”66 One might think that the sixteen-dollar fine is not a big constraint on the freedom of speech for the Turkish publisher; he may continue to publish controversial books, and on each occasion pay an insignificant fine. Still, there is a way to see this sanction in a much deeper sense: as assault on him, as a moral agent, who has a right not to be sanctioned because of his ideas.
The balancer may reply to this that it is wrong to view the balancing approach as anything more than a handy heuristic device. Its purpose is not to articulate any deep moral truths or to be faithful, in all its details, to our most considered judgments about individuals and their relationship to society. If it helps us find the right answer, it achieves everything expected of it. In fact—the balancer will go on—it has an additional advantage over its rivals: it provides a simple, structured, and manageable method to adjudicate human rights issues that does not embroil judges in deep moral questions with all their complexity and contestability, not to mention the legitimacy problems they raise, when they are decided by judges. In this vein, some have pointed out that it is not feasible to ask of judges that they engage every time in a full-scale moral discourse calling upon all our basic moral values before they reach a decision. “To expect judges to develop their own unifying theory … is simply unrealistic—a task for Hercules perhaps, but not ordinary judges.”67 Judge Frank Coffin has made the same claim even more emphatically: “When we try to see what would be substituted for all balancing in the areas covered by the first ten and fourteenth amendments, we are told only to ‘give up feigned mathematical precision and objective constitutional science for serious theoretical investigations on the meaning of constitutional language and structure.’ We are urged to ‘begin again a lively discussion about the fundamental principles that we believe undergird our political system.’ Finally we are entreated ‘to begin to search for new liberating metaphors’. As a practicing judge with a backlog of opinions to write and cases to decide, I hope for forgiveness if, pending the result of theoretical investigations of the meaning of the language and structure, I continue to resort to balancing.”68
One obvious objection to this line of thought is that a methodology is unlikely to yield correct outcomes unless it does reflect, as far as possible, the true nature of our moral concepts. However, even if we set this problem aside, we must note that the balancing methodology is no less taxing on the intellectual powers of judges than the full-scale moral argument they want to avoid. It is noteworthy that, after having disparaged the alternatives as inappropriate for “a practicing judge with a backlog of opinions to write and cases to decide,” Judge Coffin goes on to explain how balancing must be properly conducted. He suggests two prerequisite qualities (openness and carefulness) and then six stages of balance! The whole process does not sound much easier than the Herculean task of the Dworkinian judge. It becomes easier only if we skip all these stages and rush to compare apples and oranges.
It seems to me that it would be a genuine assault on the very concept of human rights if adjudication were based merely on the principle of proportionality, and we were content with arguing about human rights issues in terms of adequacy or intensity instead of right or wrong. Fortunately, although our judges pay lip service to balancing and proportionality, it is more than obvious that, most of the time, their judgment relies, in fact, on moral considerations. In the F. v. Switzerland case, for example, they made reference to the essence of the right in order to indicate that the measure offended the very core of what we value in a right, such as the right to marry.
More revealingly, even when judges cloak their reasoning in the terminology of proportionality, they often discount certain justifications as incompatible with a right before they even reach the balancing stage. A characteristic example, in this respect, is the court's decision in Lustig-Prean and Beckett v. UK, where the issue was the compatibility with the right to privacy (article 8) of a policy to discharge homosexuals from the armed forces in the name of national security and operational effectiveness. The court ruled that insofar as the purported “threat to the fighting power and operational effectiveness of the armed forces [was] founded solely upon the negative attitudes of heterosexual personnel toward those of homosexual orientation”69 and “a predisposed bias on the part of a heterosexual majority against a homosexual minority,”70 these conditions could not furnish sufficient justification for interference with the applicants’ privacy. Obviously, the court could not exclude offhandedly the fact that the existence of such negative attitudes might adversely affect the operational effectiveness of the armed forces. Its insistence on discounting those attitudes stemmed, rather, from the fact that a state committed to the protection of human rights cannot condone or protect such attitudes or make them the basis of its policies.
The fact that courts often use the language of balancing and proportionality while, in reality, they engage in substantive moral reasoning has also been noted by authors who otherwise take a favorable stance toward balancing. Julian Rivers, for instance, suggests that there are two conceptions of proportionality: one common in Continental Europe, where rights and public interests are formally indistinguishable (a situation he favors), and one predominant in the English common law where public interest reasons are treated as limitations on rights and the role of the court is to police those limitations.71 His explanation of the role of courts in human rights adjudication, within the context of common law proportionality, is that “all the court does is maintain an efficiency-based oversight to ensure that there are no unnecessary costs to rights, that sledgehammers are not used to crack nuts, or rather, that sledgehammers are only used when nutcrackers prove impotent.”72 But sometimes, Rivers concedes, the jargon of proportionality is used by courts to denote the existence of an inviolable core of the right at issue: “Finally—and this is less frequently observed—the state-limiting conception of proportionality sometimes assumes that there is an absolute minimum to each right, a core content, which may not be violated on any account. This is supposedly defined without any reference to any public interest and is, once again, the preserve of the courts.”73
It is this last point that poses difficulties for balancers. The position that there is an inviolable core content of the right implies a substantive moral assessment about what is right and wrong. Once we have accepted that this core content cannot be compromised under any circumstances we have left behind the idea that the right at stake can be weighed against competing public interests. Put simply, there is no balance to talk about in the first place. This explains why definitional generosity, no matter how broadly we understand it, can never accommodate certain rights-restricting reasons. To go back to my earlier example of Eskimo-haters, we reject the view that their preferences should be taken into account when deciding whether to prohibit the screening of an Eskimo film—not because that would have been an excessive interference with free speech, but because such preferences are altogether excluded from the range of permissible reasons the state could invoke to prohibit the screening. A court that assumes that there is an absolute minimum to each right is no longer concerned with issues of intensity and degree, and, thus, proportionality. Or, to use the imagery suggested by Rivers, sledgehammers and nutcrackers are irrelevant; the court's concern is to keep the nut intact.
The problem with the rhetoric of balancing in the context of proportionality is that it obscures the moral considerations that are at the heart of human rights issues, and it thus deprives society of a moral discourse that is indispensable. It may be that our judges are worried about moral disagreements and that is why they try to bypass the moral arguments by masking their reasoning in neutral language. However, the best way to resolve our disagreements is to spell them out and openly debate them.
This paper was written during my residency at NYU Law School in 2007–2008 and presented in the Jean Monnet Working Papers series. I am grateful to Sarah Barringe Gordon, Mattias Kumm, Michel Rosenfeld, and Joseph Weiler for valuable comments. I am also grateful to Nicholas Hatzis and Dimitris Kyritsis for comments and discussions on earlier drafts.