Introduction
The present paper attempts to provide an abstract and theoretical account of conflicts of rights in international human rights law. It does not focus on the jurisprudence of international courts, it takes a distance from practice and reflects on the options that international human rights law (as an academic discipline) has for resolving conflicts of rights, keeping in mind the contributions of neighbouring fields such as constitutional law, ethics and political philosophy. It aims to develop a taxonomy of the methods available for resolving conflicts of rights in international human rights law and to determine which of these options is better.
The study is structured around the notion of a meta-rule. It opposes the notion of resolving conflicts according to one’s own wisdom to that of resolving conflicts following a rule. A meta-rule is a rule about the application of other rules. A specific sort of meta-rule – a conflict rule or collision norm – could determine, inter alia, which right to prefer in case of conflict.Footnote 1 Nevertheless, international human rights law does not provide us with ready-made, effective meta-rules for resolving conflicts of rights. Therefore the question is: can judicial practice come up with useful and defensible meta-rules for this purpose?
Conflicts of rights are a neglected topic in international human rights law. A rapid overview of the main textbooks on the subject reflects an almost complete lack of coverage of the issue.Footnote 2 This is perplexing considering that a neighbouring topic, the limitation of rights due to concerns of public policy, is well covered. One can hypothesize that given the fact that most international human rights litigation usually involves only one right-claiming party versus the State, the rights of third parties fade from view.Footnote 3 This does not mean that the rights of third parties not present in the litigation are not relevant. If the defending State alleges that it is pursuing an interest related to rights of third parties, it would be improper to judge the legality of state conduct through the restrictive lens of the limitations doctrine. Limitations of rights are meant to be exceptional and the justificatory burden on the State invoking a limitation is very high.Footnote 4 If there is a right of a third party on the other side of a case, that right is just as worthy as the right of the petitioner, and there is no reason to stack the odds against it.
The title to the present study presupposes that human rights really conflict. This can be seen as problematic, some scholars have argued that proper human rights do not conflict, or that conflicts between rights are only illusory because once human rights receive their proper interpretation the conflict fades away. Whether this is true or not cannot be dealt with now.Footnote 5 At this point it is enough to assert that in actual practice one can find situations of real or apparent conflict quite often. For instance consider the following streamlined examples, abstracted from actual cases:
Celebrity: a paparazzo asserts his right to publish photographs of the private life of a celebrity under freedom of speech, and the celebrity asserts her right to prevent such publication under the right to privacy.Footnote 6
Kidnapping: A person is suspected of having kidnapped a child, whose life may be in danger. The police consider torture as a means to extract information that may save the life of the victim. The right to life of the child is pitted against (inter alia) the right of the suspected kidnapper of being free from torture.Footnote 7
In general, this paper proceeds in five steps. First it addresses what precisely a conflict of rights is. Second, it explains the relationship between conflicts of rights and the rights’ catalogue. Third, it explains the exhaustion of traditional rules of international law with regard to conflicts of rights and systematically describes the judicially crafted meta-rules that can be developed for resolving these conflicts. Fourth, it takes a position with regard to the question of whether conflicts of rights should be considered real or illusory. Finally, it takes a position as to which method of resolving conflicts is better.
While the topic of this paper has been extensively studied in a domestic context, this contribution does not aim to be a mere application of domestic knowledge to the international environment. Rather it attempts to move the discussion forward, by presenting a more refined analytical description of the methods used to resolve conflicts of rights and a systematic analysis of their added value.
What is a conflict of rights?
Rights prescribe duties. Every right prescribes at least one duty, and most prescribe more than one duty.Footnote 8 What is meant by ‘conflict of rights’ is that the duties prescribed by two or more rights are not co-possible at a given time and place, that they cannot be jointly realized, because the performance of the actions required to comply with a duty arising under one right renders impossible the performance of the actions required to comply with a duty arising from another right.Footnote 9 These duties may be negative duties (duties of inaction) or positive duties (duties of action). Conflict may occur when: (1) one right prescribes something and another right prohibits it and (2) when one right prescribes something and another right prescribes something else, and both prescriptions cannot be done jointly. This can be illustrated with examples as follows:
(1) Positive duty vs negative duty: The state has the duty to protect its citizens from discrimination, but also the negative duty to respect the freedom of speech of people harshly criticizing a particular minority.
(2) Positive duty vs positive duty: Rights require the state to increase the number of schools and hospitals in a province, but there is not enough money to do both.
Additionally one could imagine the following situation: (3) a right prohibits something and prohibits something else, and the two prohibitions should not be complied with jointly. Consider the hypothetical case of Jim devised by Bernard Williams:
(3) Negative duty vs negative duty: A foreigner sees that a rebel group is about to execute a group of natives. The rebel leader tells the foreigner that if he shoots the chief, he will let the rest of the group members live. The foreigner is put into a situation of choosing which right to violate, the right to life of the chief, or the right to life of the whole tribe, and both courses of action are prohibited.Footnote 10
But, do these prohibition vs prohibition cases really reflect a conflict between negative duties? The agent put in such a dilemma may always honour both negative duties by simply refusing to do anything, even if this leads to a terrible consequence.Footnote 11 On the other hand, if he does this, he will be missing an opportunity to reduce the amount of rights violations in the world, irrespective of who the perpetrator is. But this smuggles in a sort of positive duty to reduce violations. Consequently it seems that negative duties cannot conflict. Conflict in this case requires a third rule, establishing a positive duty of some sort.Footnote 12 This may be a vague principle, like stating that one should take opportunities to reduce the amount of rights violations in the world, or something more concrete, like the foreigner having a positive duty to protect the life of the tribe given his specific situation. But in any case, (3) can be reduced to an indirect version of (1).
Rights do not only encapsulate duties, but also permissions. Permissions should be understood as realms of free choice, where a person that is permitted to do something may freely choose to do it or not.Footnote 13 Therefore another case can be added to our list: (4) a permission for person α to do φ clashes with both (A) the existence of a duty for person α not to do φ, and (B) a duty for person α to do φ, as in both cases free choice is eliminated. Thus:
(4) Permission (free choice) vs negative duty or positive duty: (A) The State-endorsed permission of citizens to defend their life with deadly force when threatened conflicts with the right to life of the assailant. (B) The State has recognized freedom to choose one’s work; on the other hand, there is a national health crisis and a shortage of medical personnel. A conflict would arise if the State made the practice of medicine for accredited doctors compulsory in an attempt to further the right to health.
It must be emphasized that actual rights are more complex than the positive duties, negative duties and permissions discussed above. It is better to see each human right as a broad, sometimes unspecified normative regime which may include a range of negative duties, positive duties and permissions, and it is important to identify what is really conflicting in each case. For instance it would be wrong to identify a ‘liberty right’ like freedom of speech only with a permission to speak. It also includes negative duties for the state not to censor speech, and even some positive duties for the state to provide protection for the speaker.Footnote 14 If the state cannot provide protection for a speaker because of scarce resources – because there are no available policemen at the time – it is not the ‘permission’ aspect that is entering into play, but the positive duty.Footnote 15
Conflict has been defined as a situation where it is impossible to carry out the conducts prescribed or allowed by the rights jointly, but this notion of impossibility is vague and can only be assessed with a view to certain assumed constraints on action.Footnote 16 Certainly ‘impossible’ includes the physically impossible, but it may refer also to institutional impossibilities. For instance, with regard to example (2) above, it may be possible to build both schools and hospitals if the state is willing to confiscate private property, but such a move is institutionally prohibited and therefore unavailable. Institutional constraints may also be understood as unstated rules.Footnote 17
The catalogue of rights and the centrality of conflicts
Taking the notion of co-possibility seriously, it is clear that a catalogue of rights should not include conceptually conflicting rights such as freedom from slavery and the right to own slaves. This is trivial. But the more serious question is: can the problem of conflict of rights be avoided by reducing the catalogue of rights to only rights that are co-possible in all circumstances? This is attractive because the solutions to the problem of conflicting rights are not wholly satisfactory, and the phenomenon of conflicts puts into question the idea that rights can function as self-sufficient apolitical trumps.
Keeping in view the undesirable side effects of admitting conflicts of rights, some scholars have developed an extremely revisionist view of the human rights catalogue, asserting that it is necessary to accept that some rights therein are mistakenly included, or at least that they do not fulfil the requirements of an ideal rights catalogue.Footnote 18 The revisionist proposals tend to suggest that one can elaborate a co-possible bill of rights by setting forth only negative rights in independent domains.Footnote 19 Negative rights, they argue, do not conflict with one another because these demand only inaction and nobody’s inaction can conflict with anybody else’s inaction.Footnote 20
These proposals are unsatisfactory at two levels. First, negative rights, to be meaningful legal rights, require enforcement and some form of compensation in case of violation, and both entail positive duties open to conflict in a context of scarce resources.Footnote 21 Second, such restricted catalogues of rights might be undesirable. They are disconnected from actual human needs and thus uproot human rights from their justifying soil. If one thinks that what is dislikeable from conflict of rights is that they allow too much discretion on the part of the judge, the restricted catalogues may eliminate such a problem, but at the price of offering a list of rights that is unconvincing politically.Footnote 22
Another enterprise on this direction is that of reducing conflicts, not eliminating them. Rawls argues that human rights should have a co-possible central range of application.Footnote 23 They may conflict at the margins, but as long as such a conflict is peripheral, the catalogue of rights is admissible. This is a commendable approach aimed, not at making the problem of conflict disappear at any price, but at reducing it to a marginal size. It goes hand in hand with the sensible preoccupation of preventing human rights dilution; as more and more human rights are created there is a risk of making everything that is desirable a human right and of thereby undermining the urgency of the subject.Footnote 24 Nevertheless, the Rawlsian approach is not the end of the problem. Lawyers are usually called on to litigate the ‘hard cases’, which, infrequent as they may be, often represent intractable conflict of rights in need for resolution. For instance freedom of speech and privacy do conflict, and few would suggest taking them out of the catalogue of rights.
In any case, one must keep in mind that international lawyers work with the real catalogue of rights, and not the ideal, minimal and coherent catalogues of the philosophers.Footnote 25 Human rights catalogues nowadays are – if anything – expansive.Footnote 26
Methods for resolving conflict: the need for judicially created meta-rules
The problem of conflicting human rights is different from (and more intractable than) the problem of conflicting ordinary rules due to two reasons. First, there are no legally specified meta-rules that can be used to solve a conflict of rights, and second, there are no (legally established) higher values to which one can appeal.
When an ordinary legal norm enters into conflict with another legal norm, one can usually resolve the conflict through application of pre-established rules of interpretation. At the international level, these sorts of rules can be found in the Vienna Convention on the Law of Treaties (1969). Nevertheless, some of these rules like hierarchy, lex posterior, lex specialis seem to be inapplicable for human rights, which are meant to stand jointly at the apex of the legal hierarchy, and are not meant to be derogated by other rights, but rather are expected to coexist harmoniously, and do not seem to stand in relations of more general to more specific to each other as they are all required as a minimum for human dignity.Footnote 27 Although these features are more obvious in constitutional bills of rights, they are also part of human rights treaties, specially through the theory of ‘objective obligations’, according to which human rights treaties are meant to constitute limits to the freedom of action of states and operate beyond reciprocity defining a realm of jointly policed obligations,Footnote 28 and the ‘living instrument approach’ according to which human rights treaties develop a life of their own, beyond the limited intentions of the drafters.Footnote 29 This has led to the idea of the constitutionalization of human rights treaties.Footnote 30
Other rules such as effet utile, intent of the drafters and teleological interpretation seem to be applicable, but are not by themselves very illuminating. Effet utile can suggest the importance of coherence, that in principle all the rights in a conflict should be honoured, but it does not tell us how to achieve that result. Looking at the drafting history of the treaty may suggest the preferred solutions for some conflicts, but this is unlikely. Drafting history may devote some space to the content of rights, but it will rarely foresee their conflicts, much less resolve them. Finally, teleology asks us to interpret rights with a view to their object and purpose, but if there are no legally recognized values beyond the rights themselves, it seems difficult to find a rich, non-polemical source of value, against the background of which conflicts may be defused.
Furthermore, when ordinary legal rules conflict, one can use the higher values of the legal system to guide the application of the meta-rules. In this sense, it is true that the meta-rules will not always point in the same direction, but taken together with the values of the constitution, it is possible that a legal decision-maker will find sufficient guidance in the law to resolve the conflict satisfactorily. Nevertheless, as already mentioned, recourse to higher values is arguably not available for conflicts of human rights as human rights are already the highest point in the system.
The doctrine of limitations of rights – as developed in the jurisprudence of the European Court of Human Rights – and the problems posed by conflicts of rights are also different. For limitations of rights there is a relatively fixed decision procedure already in place in international law, whereupon limitation is only valid if (1) the right in question allows limitations in the first place, (2) the limitations are prescribed by law, (3) the limitations pursue a legitimate aim, and (4) the limitations are proportional and necessary in a democratic society.Footnote 31 This decision procedure is often strengthened by a presumption that limitations are meant to be exceptional.Footnote 32 No such detailed, generally-agreed-to conceptual apparatus exists for conflicts of rights.
In the absence of applicable, effective collision norms, conflicts of rights may be handled by a primordial form of balancing where the judge simply decides which of the rights in question is more important, or as the metaphor goes, which right has more weight.Footnote 33 But doing this seems to enable unfettered discretion, to the point that the resolution to the conflict of rights can hardly be seen as a legal decision. It is not possible to say that a judge would be ‘following rules’, even grossly indeterminate rules, in coming up with the solution.
To constrain this, doctrine has proposed special methods for resolving conflicts of rights. These methods can be explained as meta-rules that allow conflicts of rights to be resolved through rule application. Given the fact that these methods do not appear expressly in the law, they would have to be judicially adopted meta-rules, but as we will see, they might be able to derive their justification from an interpretation of the law. Borrowing from American constitutional doctrine, one could say that these judicially adopted collision norms may be understood as meta-rules drawing their inspiration and authority from international law, but not required by international law, which are used to resolve conflicts of rights, or at least constrain judicial discretion when conflicts occur.Footnote 34
As we will see, all the meta-rule approaches have strong evaluative elements, they do not eliminate discretion, but they do constrain it. Furthermore, since the approach is rule based, it tends towards stability. It creates expectations that the same meta-rules will be applied on future cases in the same manner. Therefore, through the application of a meta-rules approach, every legal dilemma becomes slightly less dilemmatic with the passage of time, as a value of legal certainty is added to one side of the conflict, with increasing weight.Footnote 35
Hierarchy
A first solution is to create a very simple, very determinate meta-rule for solving conflicts of rights, a rule of hierarchy. Earlier we argued that all human rights are formally on the same level, but maybe this is not true if one looks at things more substantively. It seems intuitively clear that life is more important than freedom of speech, and that the right to food is more important than the right to paid vacations, so one is tempted to consider the possibility of putting all human rights in a hierarchy and resolving conflicts lexically, so that the higher rights always trump the lower ones.Footnote 36 This meta-rule would have the benefit of clarity; after the location of the conflicting rights in the hierarchy is determined, the solution to the conflict becomes almost mechanical. Nevertheless, this advantage seems to be outweighed by the disadvantages.
First, in many cases one cannot say without controversy that one right has precedence over another in a general fashion, because the importance of a right will depend fundamentally on context. Going back to Celebrity, in most accounts the right to freedom of speech is of higher status than the right to privacy. Free speech is central to the conservation of a democratic society, and it is also the tool through which other rights are protected.Footnote 37 Nevertheless, in the particular case, it seems intuitive to wish to give more value to privacy. Even extremely important rights such as the right to life might need to give way to less essential rights in specific contexts; for instance,
Euthanasia: A terminally ill person challenges the non-existence of laws permitting or facilitating euthanasia in his country on the basis of the right to privacy, arguing that it should be a personal decision for him to choose when and how to end his life. The state argues that his challenge cannot proceed, because he has a non-waivable right to life, which the state can not infringe, and which the state must protect against third parties.Footnote 38
It is not unreasonable to consider that in Euthanasia, the right to privacy outweighs the right to life (admittedly, depending on the nature of the illness, some elements of the right to be free from torture can also be brought to bear against the right to life in this case).
Second, it is dangerous to try to identify more important rights. Such an enterprise can easily devolve into an ideological exercise, where ‘Western’ states push for a higher rank for civil and political rights, while other states push for recognition of ‘development’, economic, social and cultural rights and rights of communities (or peoples) above ‘bourgeoisie’ freedoms. The difficulty of selecting more important rights seems to introduce a strong discretional element that the hierarchy meta-rule meant to remove.
One way to restrain this is to try to find a basis for hierarchal order in international law itself. Despite the UN’s refrain of interdependency, interrelatedness and equal value of all human rights,Footnote 39 there may be indications in international law that some rights are more important than others.Footnote 40 A few rights, such as freedom from torture, freedom of conscience, the prohibition of slavery and the right to personality appear to be absolute, and maybe they can even be considered to be ius cogens.Footnote 41 Other human rights, such as the right to life and fair trial are defined, within their respective treaties, as non-derogable in times of emergency, other rights have no special qualification, and other rights are expressly limitable for matters of the public good. All these qualifications can be seen as signals by the drafters to the treaty interpreters of the relative importance of the rights.Footnote 42 Nevertheless, not all of these ‘signals’ are univocal. For instance, the fact that certain rights that cannot be derogated at times of emergency may simply reflect that there is no reason to think it is useful to derogate the right in such a context, and what the treaty drafter is trying to do is to avoid an opportunistic derogation, not signal the importance of the right.Footnote 43
Third, further complications appear if one takes into account that rights are actually bundles of obligations and permissions, and the importance of a right is not necessarily shared by all its duties and permissions. So for instance, while most would agree that the right to life is at the top or near the top of any reasonable hierarchy, some could object that what is at the top is the duty of the state not to kill arbitrarily. Other duties, like criminal prosecution of murderers, are more nuanced, and should be balanced against other concerns (consider in this light the peace vs justice debate with respect to transitional societies). This understanding undermines the attempt to derive a hierarchy from signals in international law. The prohibition of torture is clearly ius cogens, but the duty to punish the torturers severely – usually seen as a duty arising under the right to be free from torture – might not share this property.
Finally, hierarchy is an inherently limited meta-rule; it does not solve all possible conflicts of rights. Unless one ranks every right (which seems impossible) there will always be the possibility of conflicts between rights in the same rung of the hierarchy.Footnote 44 And even if one ranks every right, hierarchy does not tell us how to resolve conflicts between the same right of two persons, or between different duties contained in the same right.Footnote 45 Consider:
Conjoined twins: A and B are born as conjoined twins. Tragically, if they are not surgically separated, they will both die, but if they are separated, only one of them will die. Consequently, the rights to life of A and B are pitted against each other.Footnote 46
These objections suggest that meta-rule of hierarchy is too stiff and limited to solve the problem of conflicts of rights, and the benefits it brings at the level of determinacy are outweighed by its inflexibility or undermined by the difficulty of constructing the hierarchy. Nevertheless, it might be reasonable to defend the notion of hierarchy having a lesser role. After all that has been stated, it is still reasonable to say that generally speaking, the rights to life, food, and freedom from torture should come first, and that novelty rights (possibly not real human rights) such as the rights to culture and paid holidays should come last.Footnote 47 Maybe what is necessary is to have a defeasible hierarchy, where in specific cases the abstract value of certain rights may be defeated by contextual circumstances. This moves us to the next method, balancing.
Balancing (as a meta-rule)
If a simple hierarchy meta-rule does not always work because it is too stiff, it may be beneficial to substitute it for a more nuanced meta-rule that allows for greater sensitivity to context. One option then is to use a balancing meta-rule. This is different from the primordial balancing described in the part of this section on methods for resolving conflicts. Here the idea is that while the judge will evaluate something as being more important than something else, this act will not be wholly free, but constrained as to which factors to take into account by the balancing formula.Footnote 48 Such meta-rule based balancing strategies can be divided in two categories: those that conceive balancing as an act of displacement and those that conceive balancing as an act of optimization.
Displacement balancing. Displacement balancing departs from the idea that one can either comply or not comply with a right, without there being any middle ground.Footnote 49 Consequently, when rights conflict, one must choose which right to comply with, and in choosing one of the rights, the other right is displaced completely. Unlike primordial, rule-free balancing, the judge is given some parameters in order to decide which right to prefer. For instance, the judge may be asked to consider only:
(1a) The abstract weight of the rights.
(2a) Concrete weight of the right in a specific context of application.Footnote 50
This coheres nicely with intuitions of some rights being more important than others – as discussed in the section on hierarchy – but while the hierarchy is inflexible, balancing allows other considerations to outweigh the abstract importance of the right on occasions. A right may have a very high abstract weight, but may have a much lower weight in the particular case or vice versa. In this way, one is able to retain the good elements of hierarchy, while allowing for greater sensitivity to context. So cases like Celebrity and Euthanasia become easier to resolve in a way that satisfies our intuitions. Freedom of speech and the right to life do have a higher abstract weight than privacy, but in both of these cases, the concrete weight of privacy in the context of application is high enough to displace the abstract weight of the countervailing rights. Conversely, the prohibition of torture in a case like Kidnapping becomes more uncertain.
Nevertheless, balancing as a meta-rule introduces its own difficulties. For one, multiple formulas for balancing can be brought up, and there is no clear methodology to decide which one should be adopted. For instance, to the three elements set forth above, one could add a hypothetical third principle:
(3a) Expectations of society that a certain right will prevail over another.
Or a fourth:
(4a) The amount of aggregate societal utility expected from the decision.
And so forth.
Also, the example above (1a–2a) assumes that each abstract consideration merits the same degree of attention. One could deviate from it by establishing that the abstract weight of the rights is the predominant consideration, and that any other consideration can only overcome its abstract weight if it is significantly higher (as if the value of the first consideration were multiplied by 5 and the second by 1).
A key element of most attempts at devising balancing formulas is a reference to weight, as in (1a–2a). But what is this ‘weight’ that is referred to here? One could speak of importance, but it would still be unclear, importance according to what? The notion of weight seems to presuppose that there is a common currency underneath the rights, which can serve as basis for the act of balancing, and that this currency can in fact be measured. This leads into the objections of incomparability, incommensurability and arbitrariness.Footnote 51
Incomparability here refers to the fact that because most of the rights are extremely important, it is not possible to rationally consider trade-offs between them. It is nonsensical to ask how many lives lost justify one act of torture, because both represent ultimate values that cannot be traded against the background of a common currency, nor judged against the background of an even higher value.Footnote 52 Alternatively, rights may be only incommensurable, so that in a conflict of rights a person may be asked to pick one of the rights in conflict, no matter how terrible the choice is, but he will not be able to say by how much, or according to which objective metric one right is better than the other. In this case it is important to consider on what grounds other than mere preference is the person making this choice, leading to the problem of arbitrariness. One may consider that rights are comparable and commensurable because there is a common currency of all human rights after all, a key good that dominates all rights like freedom or human dignity. Even if this were true, it does not avoid the charge of arbitrariness, because lacking a reliable mechanism to measure freedom or dignity, all judgments assigning concrete or abstract value to rights would be highly subjective.Footnote 53
Optimization balancing.
Unlike displacement balancing, optimization balancing departs from the notion that rights are not ‘rules’ that can only be complied with or not, but that they are ‘principles’, which allow different degrees of compliance.Footnote 54 Then the logic of principles suggests that one should find a point of optimization, in which all competing principles are complied with in their highest degree.Footnote 55
For optimization balancing one can think of a formula that is almost identical to the one used for displacement balancing, only that instead of asking the judge to identify the most important right, it asks the judge to identify the point at which a limitation of one right to provide for another right is optimal with a view to the maximization of a core value, such as human dignity across both rights. So the judge will be directed to consider the following factors:
(1b) The abstract weight of the rights in question.
(2b) The relative weights of the rights, jointly considered, so that ‘the greater the degree of non-satisfaction of, or detriment to, one right or principle, the greater must be the importance of satisfying the other’.
(3b) The degree of certainty one has about the justifiability of interfering with one right in order to provide for another so that ‘the more heavily an interference with a constitutional right weighs, the greater must be the certainty of its underlying premises’.Footnote 56
The main difference to the scheme of 1a–2a is made in the second criterion (in the context of this essay, we can ignore the third criterion), which presupposing that rights can be complied with in degrees points towards optimization, the situation in which both principles are complied with to the greatest extent possible, taking into account their weight. This notion of weight seems to make reference to a value that has to be maximized across rights, and the usual candidate is human dignity.Footnote 57 An early Rawlsian approach also was oriented towards optimization of rights on the background of a common currency, but attempted to maximize freedom (the most extensive liberty) instead.Footnote 58
A benefit of optimization balancing versus displacement balancing is that the former explicitly entertains the possibility of finding the optimal degree of rights’ protection in a situation of partial compliance with various rights.Footnote 59 Such flexibility might be beneficial, so for example in a case of speech that offends the rights of others, criminal liability may be considered too great a restriction of freedom of speech, but not civil liability; consequently civil liability may be thought to be the optimal point of balance between the principles embedded in the rights. It may also be worrisome. In the actual case that is behind Kidnapping, an important issue of contention was whether the reduction of penalties for a torturer with good intentions was a justifiable compromise.Footnote 60 Although such compromises are strongly thematically connected to optimization balancing, it is dubious that they should be unavailable to displacement balancers.Footnote 61 To achieve the compromise result of the example sketched above, all that the displacement balancer has to do is to decide that when a civil liability claim is at issue, privacy displaces freedom of speech, but when a criminal liability is at issue, freedom of speech displaces privacy. The only thing that would be missing in this account is the notion of civil liability being an optimum balance point between the two rights, which is an interesting ‘view from above’, but it is unclear if it really adds something new.
In general, optimization balancing suffers from the same problems as displacement balancing. First, there is no agreement on the components of the balancing formula. Second, more than displacement balancing, these attempts to optimize rights depart from the basis that the rights are comparable, that there is a common currency or metric beneath them, which must be maximized across rights. This means that balancing is open to the objections of incomparability, incommensurability and arbitrariness that were discussed in the section on displacement balancing.Footnote 62
Another issue is that rights do not always seem to behave as principles. In certain cases it seems difficult or impossible to find a middle ground between fully complying and not complying with a right. Sometimes the only way in which rights can be optimized is by displacing one of them.Footnote 63
A technique that may ease the difficulty of balancing is the identification of the core element of rights. According to this theory, rights have both a periphery and a core, and while restrictions to the periphery are permissible, the core must always be maintained. This suggests that before balancing, one can discern whether one core aspect of a right is pitted against the periphery of another right. If such is the case, the core should trump and one avoids balancing altogether. Balancing proceeds if the conflict is between two cores, or between two peripheries. This procedure combines some of the flexibility of balancing with hierarchy’s lexicality. It also inherits some of hierarchy’s weaknesses. Just as it is difficult to identify the hierarchy of a right, it is questionable if a judge can impartially identify a ‘core’ in a non-evaluative fashion in the absence of a meta-rule for identification of cores, or consensus on what the core elements of a right are.Footnote 64
One final point must be made about the defensibility of balancing. Displacement balancing and optimization balancing may be both justified through reference to the idea of teleology. It may be said that human rights serve a function, to promote a value of human dignity, and balancing, in both of its forms, is the procedure to be used to ensure as much respect as possible for the aims of rights. Optimization balancing can be further defended through the notion of effet utile, as it endeavours to give each right a proper degree of concern.
Harmonizing meta-rules
A third sort of meta-rule is a harmonizing meta-rule. A rule that tries to bridge the apparent conflict between the rights, and through refining their scope, ensures co-possibility. Unlike the preceding meta-rules, which apply to rights in general, harmonizing meta-rules need to be tailored for a specific class of conflicts. Consider:
Hate speech: Freedom of speech, naively understood, clashes with the right to equality, when speech is used by a certain majority group to trash the status of a minority group in a way that constitutes hate speech.Footnote 65
This conflict can be resolved if one adds a mediating meta-rule stating that freedom of speech does not protect hate speech, hate speech is an exception to freedom of speech, or is outside the scope of this right. Once this sort of meta-rule is recognized, the only thing that needs to be done is to subsume the facts of the case to the meta-rule: is the particular thing being said hate speech, or just harsh words? Although these sort of harmonizing meta-rules do not exist for all rights, one can imagine that they can be developed, and a case like Euthanasia could be dealt with them. The meta-rule in question would simply state that the right to life does not extend to matters of suicide.
In a way, these sorts of meta-rules seem to be the most determinate method for resolving conflicts of rights. Furthermore, they do not suffer from the rigidity of hierarchy. On the negative side, they may create too much complexity, as every potentially conflictive rights relation would need its own meta-rule. Some have argued that this complexity denaturalizes the simple nature of bills of rights.Footnote 66 Still, the major problem with this strategy is, how does one come up with such harmonizing meta-rules? All these meta-rules are essentially contestable. For instance, although most would agree to set the limits of freedom of expression in hate speech, others would prefer to make freedom of speech stronger, and only refuse protection to direct encouragements of violence.
As with optimization balancing, this sort of strategy would be supported by the effet utile canon of interpretation, as it attempts to ensure that both rights in conflict receive their due. Teleology seems also to be honoured, as identifying an appropriate harmonizing meta-rule requires reflecting on the object and purpose of the rights.
As possible ways to come up with defensible harmonizing meta-rules, we can consider the following possibilities.
Harmonizing rules as the product of optimization balancing.
First, one can consider harmonization to be the outcome of optimization balancing. So at first glance, the meta-rule in operation is a harmonizing meta-rule that defines (or redefines) the scope of application of rights in order to make them co-possible. But to find out which right is to be limited in which fashion, the judge makes recourse to another technique, a formula for optimization balancing. This formula constrains the judge to find a meta-rule that maximizes a value over the conflicting rights, so for example, the judge could ask himself, which of these two rules better optimizes human dignity: the rule according to which freedom of speech stops at hate speech, or the rule according to which freedom of speech stops only at cases of direct encouragement to violence? Is a third rule superior?
On one view, this simply restates the virtues and vices of optimization balancing on a more abstract level, but this is not entirely so. Following up optimization balancing with the creation of a very clear-cut meta-rule (freedom of speech does not include hate speech) makes optimization balancing more stable, more predictable, as in future cases judges would only have to apply the harmonizing meta-rule, and only if this rule appears to be wholly inadequate would they re-do the balancing process to find a better rule.
Harmonizing rules as inference from an ideal.Footnote 67
Another option is to derive harmonizing rules from an ideal that informs us about the actual extent of the rights.Footnote 68 This must be differentiated from a purely discretionary act, where the judge simply chooses which harmonizing meta-rule he prefers and from making the harmonizing meta-rule the outcome of optimization balancing. Instead of asking himself, ‘what distribution of the scopes of freedom of speech and protection from discrimination is best?’ or ‘what distribution of the scopes of freedom of speech and protection from discrimination maximizes human dignity?’, the judge will ask himself, ‘what is the true extent of freedom of speech as informed by the ideal of a democratic society?’,Footnote 69 or for some other ideal such as ‘free and equal citizens, capable of having a sense of justice and a conception of good’,Footnote 70 or for ‘rational, social animals that need to exercise certain capabilities to flourish’.Footnote 71
These ideals must be differentiated from principles understood according to Dworkin’s terminology.Footnote 72 In our view, the main difference is that ideals are images, that they are complex.Footnote 73 While liberty is just one interest, or one value, the ideal of a democratic society, of the free and equal citizen, or of the good life for a rational, social animal reflects or at least intimates a just distribution of various principles and values. Correspondingly, inference from ideals can be used in two ways, depending on how great is the contribution expected from them.
The more ambitious use makes ideals analogous to the virtuous person in virtue ethics.Footnote 74 The image of the virtuous person already incorporates a balance of the competing values in such a way that persons who are trying to be virtuous could be instructed to act like a virtuous person would act, and not asked to balance competing values. So, for instance, the good citizen enjoys freedom, equality, privacy, freedom of speech, freedom of association, etc in some defined, proper measure.Footnote 75
Less ambitiously, ideals can be seen as starting points that can guide one towards a proper identification of the right balance of freedom and rights. In this sense, they might not be seen necessarily as an image having all the information that is needed, but as the basis for a process of construction that aims to solve a particular problem, that commands some allegiance, and that is not arbitrary. In this regard, one can consider Rawls’ idea of citizens as free and equal underpinning his construction of Justice as Fairness in Political Liberalism; had he begun with a view of citizens as self-interested maximizers, his concussions would have been necessarily different.Footnote 76 In both uses, ideals are historically contingent, as far as we can see.Footnote 77
In the best situation, the ideals that are invoked form part of the social fabric, but they can also be derived from a notion of reasonableness.Footnote 78 What is necessary is that ideals are justifiable (not justified) to both parties in litigation and society at large.
As vague as these ideals are, they can provide interpretive guidance. For instance, once the ideal of a democratic society is accepted, one would be at pains to deny that certain ‘problematic’ forms of speech must be always protected, for instance, speech criticizing the government’s policy on minorities. Interestingly, once some harmonizing meta-rules are devised for resolving conflicts, others may be inferred from them, with the initial meta-rules operating as a fixed point allowing for analogy. So if speech that asks to overthrow the government must be tolerated as long as it does not present an immediate danger of violence, speech that is highly satirical of the government must also be tolerated. And maybe if a factual direct link can be made between hate speech and violence, the aforementioned ‘fixed point’ might support a clear decision to ban hate speech.Footnote 79 Over the long run, these sorts of arguments may constitute a tradition, with a stronger presence over the minds of the judiciary – and of citizens – than a collection of balancing decisions.
Reliance on ideals is not devoid of problems. For one, ideals are vague and require interpretation. Additionally, there may be multiple ideas to choose from and the choice of ideals may be highly contested. Nevertheless, it might be productive to discuss the ideals rather than to discuss the specific conflict of rights. A discussion on ideals will probably force people to argue in terms that are closer to ‘public reason’ than defence of particular interests in the sense that ideals tend to constrain those that invoke them on pain of being declared hypocrites, and, because ideals are complex and overshoot the specific case, the constraint is rather far-reaching.Footnote 80 So if a person argues that abortion should be prohibited on grounds of the sanctity of life, he will find it difficult to advance a claim in favour of the death penalty.
Finally, it is difficult to imagine that all ideals will provide guidance to all conflict of rights questions. Some ideals such as that of ‘a democratic society’ are inherently limited, and do not say anything of conflicts which have little relevance for democratic politics, like those that might take place in a medical setting, although such an ideal might suggest that the judge must abstain from resolving such conflict, that he should defer to the political branches. Other ideals seem to be thicker, but the thicker the ideal, the more difficult it is to achieve consensus as regards it.
Conflicts: real or illusory?
Whether conflicts are real or merely illusions to be dispelled through intellectual effort depends more on the spirit in which the topic is approached than on the strategy that is used. This is because all the methods to resolve conflicts of rights can be consistently interpreted as methods that deal with the rights themselves or with our conception of the rights.
As a practical matter, the methods are not fully neutral to the question of whether conflict is real or illusory. The discourse of hierarchy and balancing tends to assume that conflicts are real, while the discourse of specific conflict rules tends to assume that conflicts are illusory. But there is nothing necessary about these identifications. Hierarchy can be seen as a mere ‘epistemological aid’ for putting order into our conception of rights, and the strategy of inference from an ideal as a means developing specific conflict rules can be seen as a mere starting point that should lead us in the right path in practical reasoning in face of a real conflict.
That said there are two substantive reasons to consider that conflicts are real. The first is one of economy. What sense does it make to postulate that there is a level where rights are in harmony, if such a level operates beyond what we can reach? At the practical level, there is no gap between rights and our conceptions of rights, the only rights that we can worry about are the ones that figure in our conception of the world. The second is one derived from the fact of tragedy. Certain choices are so unsatisfactory that one would not be willing to say ‘this person didn’t really have a right in the first place, it was just an illusion’ as is well illustrated by Conjoined Twins.
Evaluating the approaches
It is very difficult to evaluate the approaches discussed so far without getting entangled in larger controversies which include, but are not limited to, the controversy about the status of normative judgments, the extent to which legal interpretation is indeterminate or the role of ‘the political’ in law and ethics. It would be impossible to resolve all this issues here, or even to discuss them satisfactorily. The next best thing seems to be, to evaluate the methods for solving conflict of rights while making explicit the assumptions that are made with respect to these intractable questions. For the most part, an attempt has been made to cast the net widely: the assumptions made should be acceptable from multiple theoretical viewpoints.
In the field of constitutionalism, it is very common to compare the value of judicial review with that of democratic decision-making. Yet comparing judicial resolution of conflicts of rights with democratic resolution of such conflicts is something that cannot be achieved in this contribution. What this study can establish is the relative value of different methods of judicial resolution of conflicts of rights, which method exemplifies the practice in its best light, and this can be thought of as a first step towards comparing them to ideal and non-ideal democratic deliberation in the future.
Assumptions
To evaluate the approaches it is necessary to first identify what are the desiderata of a decision procedure for conflicts of rights. It is submitted that a decision procedure will have two key objectives: it will bind judges discretion and it will steer them towards the right answer. These two goals cover two overarching commitments of the law, that of being ethically good and that of being politically viable. Recognizably, there is a tension between the goal of finding the right answer and limiting the judge’s discretion. The more free the judge is, the more capable he is of considering all the facts and all the reasons that may be pertinent to reach the right decision, but likewise, the more possible that he abuses his discretion and suits his own preferences. On the contrary, the more bound the judge is by decision rules, the more likely that he will have to reason artificially, to ignore key facts and reasons and thus be forced to choose an answer that is not the best. But as we will see, the interaction between these two requirements does not have to be one of confrontation in a zero-sum game. The feasibility and desirability of these objectives depend on a network of deeper presuppositions.
The idea of a right answer implies accepting some minimal degree of moral objectivity and the possibility of rational access to it. That is to say, that some decisions are better than others in a way that is not merely subjective and not completely dependent on intuitions.Footnote 81 An ample set of views can satisfy this requirement. For instance, one could say that normative judgments are best understood as expressions of emotion or commands so that ‘killing is wrong’ means ‘I don’t like killing’ or ‘do not kill!’, but also that given facts about human nature, the possible ways in which our emotions can be combined is limited, and that there are wise and unwise ways of doing so.Footnote 82 Alternatively, one could argue that moral truth is like truth in mathematics, that it does not reflect the outside world, but that it is the outcome of a reasonable process of construction, and that the outcome of this process determines what is right and what is wrong.Footnote 83 Or one could think that good or bad are reducible to the satisfaction of real interests from the point of view of society, a form of social, instrumental rationality.Footnote 84 What all these views have in common is that some answers to normative questions are better than others and reason can play a role in identifying them.
Methodologically, we must also assume that certain widely shared, deeply held intuitions have at least a good chance of reflecting a ‘right answer’ to a normative question. Partial reliance on intuitions is necessary because it will allow us to test the methods used to resolve conflicts of rights. If moral truth is secured only by appeal to intuitions, then it is not possible to argue with a person that has different intuitions or to prove that he is wrong. But likewise, without some trust in our intuitions it is impossible to critique methods that are internally consistent. Both intuitions and methods must achieve a sort of reflective equilibrium.Footnote 85
A final issue that must settled is whether the right answer is ‘dormant in the law’ or is it imposed from outside the law? If law is taken to mean ‘positive law’, certainly the right answer is not in the law. If conflicts between fundamental rights occur, positive law has clearly run out. Conversely, if law is taken to refer to a method of normative reasoning, in which positive law plays a role, but does not exhaust the domain, the answer may be said to be found within law. This paper espouses the second view.Footnote 86 Nevertheless, which of these two views of the extent of the domain of ‘law’ is correct might be a merely academic question of where to draw the boundaries of disciplines because both views have the same practical entailment. Whether we consider that the resolution of conflicts of rights is part and parcel of ‘law’ or not, the judge must still engage in normative reasoning in order to find the best solution to a conflict of rights.Footnote 87
The restriction of judge’s discretion that is sought does not and cannot eliminate it. Meta-rules cannot lead to a mechanical judgment; they can only exclude certain results.Footnote 88 Furthermore, restriction of discretion must be understood as a form of accountability, where the judicial behavior is contrasted with relatively clear standards set forth in advance. While after a judgment doubts may subsist on whether ‘the right answer’ was found, it should be easier to determine whether the judge followed the meta-rules or not. In this connection, the issue of indeterminacy of language needs to be taken up. Here it is submitted that language includes paradigmatic cases where language is quite determinate and areas of penumbra, where meaning is indeterminate, but this does not mean that anything goes. After all, when faced with penumbra, rather than attempting to identify a ‘true’ meaning, the judge should engage in normative reasoning,Footnote 89 which can be seen as relatively determinate given the possibility of rational access to ethical truth.
The discussion that follows also assumes that judges have some willingness to find the law, or what is best in its light, although they often fail, and that they are looking at decision rules for guidance in this process. It recognizes that judges are biased, but it assumes that biases can be substantially overcome through reflection and reasoning. If one departs from an alternative point of view that assumes that judges are thoroughly ‘political’ in the sense that they can only care to advance the political agenda that is dear to them, the discussion is definitively senseless, as all that the decision rules can be are masks, facades to occult darker motives (or unrecognized biases) under the veneer of false necessity. For many academics whose allegiance is only with critique and deconstruction, to deny this is naïve.Footnote 90 But this excessive cynicism is self-defeating. The ‘unmasking’ discourse can only be effective if there is some standard against which to criticize deceit. Without appeal to a higher purpose, critique loses its power.
Assessment
With these assumptions in place, it can be seen that hierarchy is by far the worst method, because it pits the two objectives of controlling judicial discretion and of finding the right answer against each other in a zero-sum fashion. In hard cases, either one keeps hierarchy and sacrifices intuitions, or one keeps the intuitions and sacrifices hierarchy, and it is not the case that hierarchy allows us to refine our intuitions, or to discriminate between weaker and stronger intuitions. They must all be scarified to the single consideration of the abstract importance of the rights in conflict. In contrast to this undesirable situation, what is sought after is a method that can work as an epistemological aid to finding the right answer at the same time that they constrain the judge, and therefore produce synergies between the ethical and political aims of a decision method.
Unlike hierarchy, the method of balancing (in both varieties of displacement and optimization) seems to achieve this. By specifying what sorts of reasons should be considered in solving a conflict of rights it highlights the things that should matter for a decision maker and appears to rule out considerations that are clearly not impartial; for instance, the ‘relative importance of each of the right-claiming parties in conflict’ would never be proposed as a criterion for balancing.
Even if this is true, this method leaves a lot to be desired for institutional reasons. There is reason to think that judges are relatively bad balancers. If a judge is simply asked to say whether in one particular case, privacy is more important than freedom of expression, it is hard to see how his judgment could be more reliable than that of any other citizen. This is even more so in an international environment, where intuitive reactions to the weight of certain reasons cannot pretend not to arise from culturally determined sensitivities. Balancing puts the premium in the emotions and sensibility of judges in an information-poor environment, and here the democratic alternative (if it is available) appears to be superior to adjudication in every way, up to the point that adjudication can only be justified because democracy would not have the time and resources to consider all the cases of conflicts of rights. Yet this does not mean that a better proxy to democratic decision-making that is technically feasible cannot be found. A commitment to balancing as a decision rule for solving conflicts of rights would logically entail the development of culturally varied, randomly selected international juries, which can bring more credibility to the act of weighting and balancing, without running into the purely technical problems of requiring democracies to attend very specific decisions.
Additionally, one can legitimately question if balancing really leaves something out, if it really commits judges to impartiality, and if the categories commonly provided in balancing are helpful. Consider that any sensible theory of the privacy-based limits to freedom of expression will take into account concrete factual issues that relate directly to the persons involved, such as whether they are children or public officials. If balancing rules this out, then it operates at a too high a level of abstraction to be much help. In contrast, if the consideration of the relative weight of the rights in the concrete case, allows the judge to consider that in the concrete case, one of the parties in Celebrity is a public figure, a member of royalty or something else, then it is difficult to see whether balancing has really left something out; that is, it is difficult to see it as different from deciding upon one’s own wisdom. Hence, the aim of finding a right answer is probably honoured, but judges’ discretion is not constrained: balancing as a meta-rule collapses into unrestricted primordial balancing.
Specific conflict rules as the outcome of balancing seems to bring much more determinacy to the results of the balancing operation over time. Once a specific conflict rule is set down, judges are expected to stick with it on most occasions. If balancing is seen in a rather indeterminate light, so that it allows almost any consideration to come in, and that therefore it is quite close to the act of judging according to one’s own wisdom, the addition of specific conflict rules as the outcome of balancing provides a strong element of certainty to counterbalance this. What arises here is simply a stare decisis system. The downside is that not much method is brought to bear on the initial decision that establishes the conflict rule, or on subsequent decisions that modify or reject the specific conflict rule for another one. In these situations, all the problems of balancing reappear.
Inference from an ideal is also much more constraining than balancing. For instance, in light of the ideal of a fair society of free and equal citizens, the proper measures of freedom of speech and privacy for Celebrity are much clearer. The pre-eminence often accorded to freedom of speech makes sense due to its political function, but the interference with the private life of a celebrity is not related to this function, and this undermines its claim to importance. Furthermore, the conception of citizens as free and equal makes the paparazzo’s intrusion into the celebrity’s privacy seem manifestly unreasonable. Most of us would not consent to be treated this way. This can lead the judge to infer a limiting rule that freedom of speech does not extend to the publication of matters that are essentially part of the private life of people. Likewise, Hate Speech can be readily seen in the very negative light that it commands, as attacking the foundational notion of citizens being equal.
In contrast to balancing, inference from an ideal seems to have the benefit of asking from judges something at which they are arguably good at. Here the judge functions in an information-rich environment, where he may claim to have some degree of expertise. By being an expert in the law, its motivations and history, he has a better claim to know what the ideal is, and to correctly identify the implications of this ideal for a particular case of conflicting rights. For the same reason, the level of constraint is higher. With balancing, disputes can easily go down to ‘table thumping’, where opponents simply stress the greater weight of the reasons on the other side, as it appears to them, in inference from an ideal it is much more feasible to claim that the judge has made mistakes of interpretation, as it would be obvious if it is claimed that ‘speech that is critical of a particular religion’ is not protected within the ideal of a democratic society of free and equal citizens.
The problem here is not mainly that ideals are vague and require interpretation. Laws are also vague, and except on the most radical views, this does not doom the project of interpretation to mere subjectivism. The task of judging what is a ‘better, more faithful interpretation of x’ may be a task of weighting and balancing reasons, but it is clearly cognitive, and requires factual knowledge and expertise, rather than the summing up of uninformed intuitions. The problems that arise on this view are the availability of multiple ideals and the issue of their correctness.
The process of inference from an ideal makes the most sense when there is only one dominant ideal from which to infer. When there is more than one, inference from an ideal seems to beg the question. To assert one ideal when there are other possible ones seems to be arbitrary. Alternatively, if other ideals are considered and balanced against each other, the problems of balancing are simply restated at a higher level. While there is some truth to these objections, they do not apply in full force.
With regard to the first objection, pontification from one ideal is not quite ‘table thumping’, it has much more to do with rhetoric. Ideals are compelling; they can entice and convert people with different world-views. In this sense, a judicial decision that moves forward with a contestable ideal can provide a service, not of neutrality, but of facilitating conversion to a publicly acceptable view of the good.
With regard to the second objection, it is submitted that even if there has to be a choice between multiple ideals, and that this repeats balancing at a higher level, this move to a higher level is profitable. Arguably, there are less concrete, socially justifiable ideals than possible balancing arrangements of the body of human rights. Furthermore, it is hard to say that ideals can be wholly wrong, or that they can be as fallible as particular solutions to a balancing problem. They may lead us into the wrong direction, but it will not be one that is completely devoid of virtue.
The final element that needs to be considered is the correctness of ideals. It may be argued that ideals are too messy and make no sense in contrast to comprehensive and concrete ethical theories such as certain forms of utilitarianism or libertarianism which stipulate a deductive system for judging the correctness of action founded in clear first principles and that what should be done in case of conflicts of rights is to come up with a decision procedure based around these theories. This alternative is an illusion. Comprehensive, deductive theories ask us to abandon our intuitions even in the clearest of cases, so, for instance, an extreme negative rights libertarianism would reject that we have a duty to save a drowning child even if it is at very little risk or cost to us, and extreme utilitarianism may argue that there is no reason to prefer saving the life of one’s own child if that neglects the possibility of saving two persons (of reasonably young age). In light of our intuitions, it seems quite likely that these sorts of theories are profoundly wrong, possibly because they ignore key facts about human motivation and depend on impoverished notions of what should make us happy, or of what counts as being free.Footnote 91 In contrast there are restricted ethical theories, like Hare’s utilitarianism, where the principle of utility is balanced with intuitions and rules of thumb to prevent myopic or self-serving calculations, or Rawls’ constructivism which depends on a context-specific notion of reasonableness. These theories are open to experience and for this reason, they are much more likely to be useful. Ideals are synergetic to these sorts of theories. They can bring nuance and prudence to their calculations. Finally, there are strong coordination problems of relying just on ethical theories. For instance, if everyone is a utilitarian, but chooses to calculate utility in his own way, this is likely to lead to extremely undesirable consequences. Ideals, by their social presence, counteract this tendency.
Concluding reflections
There is an assumption that international human rights law constitutes a neutral, non-evaluative framework for securing some degree of morality, or at least decency in the international sphere. If not neutral, it is at least supposed that human rights can operate through a minimal consensus. Different states, different peoples and different cultures can come to accept human rights even though they accept them for different reasons, and they cannot agree on the foundations of rights. As stated by Maritain, the spirit of the enterprise is ‘nous sommes d’accord sur ces droits, à condition qu’on ne nous demande pas pourquoi.’Footnote 92
The phenomenon of conflict of rights suggests that the agreement that rights represent is imperfect and unfinished, even from the point of view of its own limited purposes. An agreement on rights will not prevent the need for further value decisions in international society and in our sceptical, democratic era, the question will always be by what authority should judges be making those decisions? This contribution suggests that the authority for judicial resolution of conflicts of rights must arise from their commitment to rationality, expressed by the creation of defensible decision rules to solve conflicts of rights. These rules may be of hierarchy, of balancing or specific conflict rules, derived from balancing, or from an inference from an ideal. Not all these rules are equal. It seems that specific conflict rules are much more desirable than others, especially if they are backed up by appeal to ideals.
Acknowledgements
I would like to thank Jaap Hage, Menno Kamminga, Ana Maria Torres, Valentina Vadi, Wibren van der Burg and two unnamed referees of Global Constitutionalism for comments and suggestions on earlier versions of this article.
