Introduction

The role of adjudication in the human rights context has always been contested. Human rights adjudication is generally accused of being anti-democratic, giving power to unelected judges to override the elected representatives of the people.

This paper begins by considering the democratic objection to human rights adjudication.

1. The democratic objection

The most cogent objection to justiciable human rights is that put forward by Jeremy Waldron. For Waldron, the most fundamental of rights is the right to participate on equal terms in social decisions. This right should not be confined to interstitial matters of social and economic policy, but should essentially concern issues of high principle such as those addressed by human rights.^[365] Giving judges the power to decide issues of high principle such as the content of human rights entails a profound disrespect for people in their democratic and representative capacities. ‘Disagreement about rights is not unreasonable, and people can disagree about rights while still taking rights seriously. In these circumstances, they need to adopt procedures for resolving their disagreements that respect the voices and opinions of the persons—in their millions —whose rights are at stake in these disagreements and treat them as equals in the process. At the same time, they must ensure that these procedures address, in a responsible and deliberative fashion, the tough and complex issues that rights-disagreements raise. Ordinary legislative procedures can do this,... and an additional layer of final review by courts adds little to the process except a rather insulting form of disenfranchisement and a legalistic obfuscation of the moral issues at stake in our disagreements about rights.'^[366]

Waldron's argument against judicial review in a human rights context is prem­ised on a society which fulfils a number of conditions.

The approach in this paper does not dissent from Waldron’s fundamental premise. Indeed it reiterates the key importance of participation in decision-making about the interpretation of rights. This is particularly true for the most contested of questions: how to balance rights against other public interests. However, I argue that such participation should not be confined to what Waldron calls ‘ordinary legislative procedures’. Although there may be a high level commitment to human rights, signified by the continued adherence to a Bill of Rights or human rights legislation, legislative majorities may not always evince a commitment to human rights per se. Nor can one say, in the generalized and abstract way that Waldron puts it, that contestation about rights is always about good faith disagreement amongst those who are committed to rights.

Nor should it follow from the foundational premise (namely that all important decisions should be settled by the people themselves) that Parliament should be the only forum for participation. There is little basis for asserting that the people them­selves actually do make important decisions. Decision-making is in practice skewed towards those with power in society, and even where majorities do make decisions, there is a risk that they will override the rights of minorities. The prisoners’ right to vote debate is again a good example. Given that prisoners do not have the right to vote, and are therefore not represented in Parliament, why should decisions about whether they should have such an entitlement be left to Parliamentary majorities? It could be argued that if the political system is defective in the extent to which ordinary people participate, the answer is to improve the political system rather than taking away more power from the people and giving it to the courts.

Waldron’s rejection of a legitimate role for judicial review of human rights also assumes that there is an inevitable conflict between judicial and legislative power. This is true for both ‘strong’ judicial review, which gives the courts an ultimate and non-revisable power; and ‘weak’ forms, which appear to require courts to defer to Parliament. However, a closer look at various models of human rights adjudica­tion reveal a much more fluid picture, in which courts and legislatures interact in potentially fruitful ways. This is particularly true for the UK Human Rights Act 1998, which Waldron characterizes as a form of weak judicial review. Under the HRA 1998, judges do not have the power to override legislation. Instead, judges faced with a successful human rights challenge to legislation can do one of two things. They can interpret the legislation so far as possible to comply with the courts’ understanding of the right at issue,^[371] in which case the amended legislation may be repealed or further amended by Parliament. Alternatively, they can issue a declaration of incompatibility.⁸ Such a declaration does not itself alter the law, but gives a powerful signal to the legislature that the law ought to be altered. Thus adjudication could be characterized as feeding into the political process. By making a declaration of incompatibility, the courts can reopen the political debate, as well as enriching it by the insights uniquely generated through the process of judi­cial deliberation. In this sense, the import of the declaration of incompatibility is political, not legal. There are those who have argued that legislatures would find it so difficult to gainsay the court that this amounts to a power of veto.

These approaches are not only relevant for situations in which the courts do not have the power to invalidate legislation, as is the case for the HRA 1998. Even where judges do in principle have the power to strike down legislation, this does not gen­erally mean, as Waldron suggests, that the decision is permanent and non-revisable. Constitutional courts have acknowledged that constitutions contain open-textured principles whose interpretation remains open to further discussion. Bills of rights are frequently referred to as a ‘living tree’, where changing social norms and val­ues infuse judicial interpretation in order to reinvigorate constitutional texts. The European Court of Human Rights expressly states that its decisions are responsive to the changing consensus in contracting states. An example concerns the extent to

⁸ HRA 1998,s4.

which it is discriminatory for fathers to be denied the right to paternity or parental leave. In Petrovic v Austria^{[372] [373], the Court found that a distinction on the basis of sex with respect to parental leave in the 1980s could not be regarded as a breach of the Convention, given the great disparity at that time between the legal systems of contracting states on this issue. However, in 2012, in the case of Konstantin Markin v Russia90, the Court held that society had significantly advanced, and the majority of European countries now provided parental leave for both men and women. ‘The Court cannot overlook the widespread and consistently developing views and associated legal changes to the domestic laws of Contracting States concerning this issue.'[374] [375] [376]}

Similarly, it is well known that dissenting judgments from earlier cases may in time convince the majority. A salient example is the United States Supreme Court's decision in Lawrence v Texas" holding that it was unconstitutional to criminal­ize homosexuality by prohibiting sodomy between consenting adults. In order to come to this conclusion, the court was required to hold that its earlier contrary holding in Bowers v Hardwick"’ was wrong. Indeed, Kennedy J, giving judgment for the court, held that the dissenting judgment of Stevens J in the Bowers judg­ment should have been controlling. Thus, the court held, ‘Bowers was not correct when it was decided, and it is not correct today. It ought not to remain binding precedent. Bowers v. Hardwick should be and now is overruled.'^[377] The Supreme Court of Canada, which similarly has an override power, is also willing to revise its decisions, as in the case of R v Kapp^[378], where the court held that its previous use of dignity in the context of equality claims under section 15 of the Constitution had become confusing. Instead of a legal test, which had become an obstacle in the applicant's way, dignity should be regarded as an essential value underpinning the legal principles applicable under section 15.

In any event, judges, conscious of their position relative to elected decision­makers, have often been quick to carve out areas of autonomy for decision-makers. In particular, concepts such as ‘reasonableness' permit ongoing interaction between judges and elected officials even in the context of a binding constitution. The South African Court, which also has the power of ‘strong' judicial review, has used the textual reference to ‘reasonable' measures in the socio-economic rights provisions of the Constitution to develop a strong concept of deference to elected representatives in relation to the means to deliver socio-economic rights.^[379] Nor are all constitutions as difficult to amend as the US Constitution. The Indian Constitution was amended by Parliament after a series of early cases in which the Supreme Court of India struck down social legislation in the name of the right to property.^[380] Although the court reasserted jurisdiction so that the ‘basic structure' of the Constitution could not be amended, the possibility of legislative override nevertheless acted as a constraint on the court.^[381]

2. The dialogic approach

Recent scholars have attempted a different way out of the dilemma: one which moves away from a polarization between judges and legislatures, and instead attempts to work out ways in which both can contribute to a democratic resolution of human rights disputes. An increasingly influential stream of thought characterizes the rela­tionship between courts and legislatures as one of dialogue. The advantage of this approach is to allow courts ‘to bring constitutional values... into focus in the legisla­tive forum' but leave the final decision to the legislature^[382]. Rather than courts having the final say in the matter, judicial decisions provoke a response from the legislature. Dialogic theories of adjudication, which originated in Canada, have shown how leg­islatures can respond appropriately to judicial decisions, ultimately achieving their legislative purpose without falling foul of their human rights commitments. The dialogic approach has also been applied to the HRA 1998 in the UK, which seems particularly apt for such a model, given that the court has power to issue a declara­tion of incompatibility but not to declare legislation invalid. In this paper, I build on the insights of these theories to posit a ‘deliberative' model of adjudication, which likewise regards courts as complementary rather than antagonists of Parliament.

The dialogic approach originated with a highly influential article by Hogg and Bushell,^[383] which argued that the record of decisions under the Canadian Charter demonstrated that judges did not in fact have the last word on the matter. Instead, the legislature was generally able to respond to judicial invalidation of legislation in ways that preserved the basic legislative objective. This was possible because of the particular structure of the Canadian Charter. Hogg and Bushell pointed to several opportunities for ‘dialogue' of this kind. The first was section 33 of the Charter, which expressly allows the legislature to override judicial decisions by the inclu­sion of a ‘notwithstanding' clause. Secondly, the qualified rights,^[384] the guarantee of equality in section 15, and the justification clause in section 1 of the Charter,^[385] all meant that the rights could be complied with in a number of ways. These features of the Charter ensured that ‘after a law was found to be invalid by the courts, legislatures would normally be left with a range of choices as to the design of cor­rective legislation... [and]... while the Charter would often influence the design of legislation that encroached on a guaranteed right,' legislatures ‘would usually be able to accomplish what they wanted to do while respecting the requirements of the Charter'.^[386]

Their claim was primarily a descriptive rather than a normative one. Having examined the aftermath of every case in which a law had been declared contrary to the Charter by the Supreme Court of Canada, they found that most of the cases had elicited some response from the legislature. In two thirds of the total, a new law was substituted for the old one. Apart from two cases in which the legislature defied the court by re-enacting the same law (once through section 33 and once through section 1), in all the other cases, the legislature ‘respected the judicial decision by adding some civil libertarian safeguards in the new version of the law, but maintained the legislative purpose'.^[387] The result was that in the Canadian context, judicial review did not mean that judges had the last word.

Although Hogg and Turnbull's claims were empirical, they quickly formed the basis for normative arguments in favour of a dialogic model. Thus lacobucci J, enthusiastically embraced the dialogic model in Vriend v Alberta^[388] as one which enhanced the mutual accountability of legislature and judiciary:

As I view the matter, the Charter has given rise to a more dynamic interaction among the branches of governance. This interaction has been aptly described as a ‘‘dialogue'' by some.... To my mind, a great value of judicial review and this dialogue among the branches is that each of the branches is made somewhat accountable to the other. The work of the legislature is reviewed by the courts and the work of the court in its decisions can be reacted to by the legislature in the passing of new legislation (or even overarching laws under s. 33 of the Charter). This dialogue between and accountability of each of the branches have the effect of enhancing the democratic process, not denying it.^[389]

The dialogic model is an attractive one. Rather than taking the side of either an unadulterated Parliamentary sovereignty or a robust judicial supremacy, it charac­terizes the process as collaborative. However, a closer look reveals a deep ambiguity as to the role of the judiciary. Can judges ultimately make authoritative decisions on the meaning of human rights (or other constitutional principles) or do they defer the authoritative decision to the legislature? The dialogic approach is compatible with either. Thus, Tushnet takes the view that the task of the court in constitutional adjudication should be to draw the attention of the legislature to constitutional values and then leave it to the legislature to decide how to address these.^[390] Similarly, for Roach, the courts' expertise in interpreting rights justifies their drawing ‘the atten­tion of the legislature to fundamental values that are likely to be ignored or finessed in the legislative process,' but not in their attempting to ‘end the conversation or conduct a monologue in which [their]... Charter rulings are the final word.'^[391]

By contrast, Hogg et al argue that the final authority for interpreting the Charter rests properly with the judiciary. ‘In our view, societies that have a written bill of rights will require some body or institution to have the final authority to say what the bill of rights means, otherwise there would be a possibility for “interpretive anarchy.” '^[392] Similarly, Hickman^[393] develops a model of ‘strong form' dialogue to describe the workings of the HRA 1998. A strong form of dialogue reflects a belief that ‘courts have a vital constitutional role in protecting fundamental principles from the sway of popular sentiment.'^[394] He distinguishes his model from what he calls ‘principle proposing' dialogue, according to which ‘the role of the courts under the HRA is to propose to the other branches answers to substantive questions of justice' but from which the government may legitimately deviate on compet­ing principled grounds.^[395] The strong form of dialogue nevertheless maintains a dialogic approach. Courts should not simply impose these principles, but work with the executive and legislature in ‘evolving them and in fostering their accept­ance.' In doing so courts should have the capacity for compromise in the interests of expediency, while at the same time ‘insulating fundamental principles even in the face of such compromises.'^[396] In other words, a strong form of constitutional dialogue is strong in that it permits courts to establish strong principles, which are insulated from majorities, but is dialogic in that the judge ‘engages on a wider dialogic enterprise with society that involves a certain degree of give and take, prudential management and persuasion in the evolution and protection of principle.'^[397]

There is also ambiguity in the notion of dialogue itself. Dialogue tends to suggest an ongoing process, whereby each side has the opportunity to react to the other. Yet human rights dialogue as described by these authors for the most part consists only in the legislature's reaction to courts. For Hogg and Turnbull, any leg­islative response constitutes a dialogic response, and there is no particular need for a further judicial reaction.³⁵ Indeed, neither the Canadian Charter nor the HRA 1998 has any formal mechanism for a further response by the court. In later work, Hogg et al did move beyond legislative response to examine ‘second look' cases, or cases in which a court has reviewed the validity of legislation aimed at replacing a law previously struck down as infringing Charter rights.³⁶ However, here too, the judicial response is not built in: these examples are clearly dependent on litigants returning to court. This is true too of the HRA 1998. Whereas a declaration of incompatibility may (but need not) trigger a response from the legislature, there is no in-built mechanism for the court to scrutinize the outcome. Hickman attempts to construct scenarios of different legislative reactions (no reaction, re-enactment without change, or modification). But ‘dialogue' ends there. There are now some ‘second look' cases, the prisoners' voting rights cases being a salient example.

But, as in the Canadian case, this depends entirely on new litigants turning to courts to evaluate the outcome.

It is of course possible to create mandatory mechanisms for continuing dialogue. For example, South African courts have tended to use a suspended declaration of invalidity, which gives the legislature the opportunity to frame its response while keeping the court cognizant of the matter.^[398] The Indian Supreme Court has used the device of a continuing interim mandamus.^[399] In both these cases, however, the court has the final say; an approach which is anathema to this version of dialogue.

Even if the dialogic approach could be improved to achieve a genuine to and fro discussion, it remains limited. Firstly, it is concerned only with the court and the legislature. Ideally, the process of human rights interpretation and application should be open to a much wider range of participants. Secondly, the dialogic model focuses only on the ‘output’ or the judicial decision itself. The deliberative model, as I will show, looks behind the decision to the process of decision-making in the course of human rights litigation. Judges are not simply expected to articulate ‘principles’ which, in Hickman’s model, appear to be identified in some abstract way. Instead, in recognition of the fact that the interpretation of a human right need not necessarily have a single right answer, the court itself should be required to justify itself in deliberative terms. Thirdly, the dialogic model assumes that the function of the courts is to give alternative solutions to human rights questions to that offered by the legislature. The deliberative model focuses instead on the quality of deliberation. The requirement that legislative or executive decisions be justified in a deliberative manner before a court is itself an important means of protecting human rights while at the same time reinforcing democracy. Thus, the deliberative model regards the judicial output as itself part of a process of deliberation.

3. A Deliberative response

The deliberative model begins by asking the question: what unique contribution can courts make to the protection of human rights? As we have seen, one of the main sources of the difficulty in framing an appropriate role for courts in adjudicating human rights is because of the assumption of a dichotomy between majoritarian democracy and anti-majoritarian human rights. The deliberative model addresses the problem by recognizing that majoritarian democracy does not fully describe all democratic models. This in turn requires more attention to be paid to the difference between, in Habermas’s terms, ‘interest-governed’ and ‘value-oriented’ co-ordination.^[400] It is argued here that while both have their place in a democracy, reso­lution of human rights cannot by its nature take place within the interest-regarding aspect. Deliberative solutions are necessary to prevent minority interests continually being overridden by more powerful interests. By contributing to the deliberative dimension of democracy in the human rights context, judges can enhance rather than detract from democracy.

Habermas explains his distinction as follows. Interest-governed coordination presupposes that each party comes to the bargaining table with fixed interests, the aim being to induce the other party to accept their claim. There is no background set of values against which to measure the validity of each party's interests; instead, success depends, not on the power of reasons, but on factual power. The solution is victory, surrender, or compromise; but not a change in the parties' perceptions of their own interests. A paradigm case is that of collective bargaining. Here, bar­gaining takes place in the absence of a collective agreement or legislative code and arrives at a set of outcomes based on the balance of industrial power. This contrasts with value-oriented coordination. Here parties do not come to the table with fixed interests. They enter the process aiming to justify their positions by appeal to rea­sons that all parties can accept, while at the same time being open to persuasion.^[401] Deliberation is not on the basis of fixed interests: taking the deliberative commit­ment seriously requires a willingness to revise one's own preferences and convictions. Thus, instead of factual power, such coordination is based on the ability to adduce reasons which participants sincerely expect to be persuasive to others. In place of defeat or victory, therefore, coordination takes place through rationally motivated consensus.^[402]

Drawing on these insights, deliberative democrats characterize democratic decision-making, as a situation in which citizens ‘share a commitment to the resolution of problems of collective choice through public reasoning'.^[403] Deliberative democ­racy contests the assumption that preferences are fully formed prior to the political process. Arguing that preferences are heavily influenced by social circumstances, theorists in this school conclude that the process of decision-making can itself play a central part in shaping those premises. This in turn means that initial prefer­ences may be revised in the course of deliberation in a way which encompasses not only one's own perspective, but also those of other members of an association. Deliberation enables participants to see the extent to which their preferences are adaptations to their own limited circumstances.^[404]

Nor is the aim of the process simply to reach a compromise or to aggregate those preferences as in the case of interest-bargaining. Instead, the aim is to arrive at a deci­sion which is capable of being justified by reasons which participants sincerely expect to be persuasive to others. Attempting to justify a position by resorting to self-interest alone is unlikely to be successful. In addition, taking the deliberative commitment seriously requires a willingness to revise one's own preferences and convictions.^[405] This is particularly true for preferences which a participant discovers cannot be justified. The result is to open up the possibility of resolving disagreements by reasoned persuasion. Moving from a bargaining model to a deliberative model therefore requires a substitution of interest-governed action by value-oriented action. Particularly important is the function of disciplining political representatives by requiring them to justify decisions by reference to the public interest, not to preferences (their own or voters') which could be distorted or self-seeking.^[406] At the same time, deliberative democracy is postulated as being a pluralistic association whose members have diverse preferences, convictions, and ideals. Apart from sharing a commitment to deliberative democracy itself, participants need not share a con­ception of the good.^[407] This in turn necessitates explicit discussion of what values should be the basis of community aspirations.^[408]

The deliberative democracy approach is not without its difficulties, both in principle and in practice. In particular, it is difficult to see how consensus can always be achieved through deliberation, and at what point closure is declared. One way forward is to use Sunstein's notion of incompletely theorized agreements. This involves distinguishing between different levels of agreement. People might disagree at one level of abstraction, but agree on a more particular application. This allows agreement on outcomes without agreeing on the most general theory that accounts for it. There may be no need to achieve consensus on the fundamental principle if the outcome can be agreed.^[409] For example, people might agree on a redistributive policy because they believe in charity for the poor, or because they believe that property is theft, or because they believe in equality. This may work in both directions: there may be general consensus on the principle but not on the particular application; in which case, the principle could be agreed and the speci­fication left for another time. Sunstein sets out a series of benefits of incompletely theorized agreements about constitutional principles and cases. Most important is the background value of mutual respect: even if there is no detailed agreement, people agree on the need for reciprocity, harmony, and respect. Such agreements reduce the political cost of enduring disagreements, while at the same time allowing the consensus to evolve over time.

Incompletely theorized agreement modifies deliberative democracy to the extent that, in putting forward their reasons, participants are not aiming to convince others of the soundness of their reasons, but only of their plausibility, so that the back­ground value of mutual respect permits agreement to be reached. Even in this form, however, it is clearly unrealistic to expect that all decision-making fulfils the criteria of deliberative democracy. Instead, the insights of deliberative democracy fulfil the more partial function of acting as a discipline on decision-makers. Deliberative democracy requires decision-makers to justify their decisions by reference to reasons that all can regard as sound, even if they cannot command a consensus. Thus, self-seeking or biased reasons are flushed out and regarded as unacceptable.^[410] Similarly, reasoned persuasion might still have a role in the context of a debate in which more than one reasoned solution is possible. In such a case, majority voting may be necessary. But this does not mean that the majority votes simply to advance its own personal interests. Instead, it is a way of deciding between various reasonable alternatives on the basis of the numbers of those who find one set of reasons more persuasive than another. In such cases, even if consensus cannot be achieved, it is possible to establish a reasonable set of policies which cannot be bargained away through the use of factual power.

Deliberative procedures will always co-exist alongside interest-bargaining. Habermas concedes that in complex societies, it is often the case that interests are sufficiently diverse that consensus is not possible.^[411] In such cases, resort must be had either to majority voting or to bargaining between success-oriented parties who are willing to cooperate. Similarly, Sunstein describes the legislative process as a continuum, at one pole of which interest group pressures are determinative, while at the other end, legislators engage in deliberation in which interest group pressures play little or no role. Along the continuum, outcomes depend on an amalgam of pressure, deliberation, and other factors.^[412] Even for those decisions in which interest-bargaining is unavoidable, deliberative procedures are necessary to establish the background procedural requirements. This is particularly important to ensure that all interested parties are provided with equal opportunities to influence one another.^[413]

Thus modern democracies are a complex amalgam of individual representation, interest-group bargaining, and deliberative procedures. However, human rights cannot be addressed on the basis of interest-bargaining. If they were, individuals and minorities may always be trumped by those with superior numerical, political or financial power. Even when blatant exclusionary practices are not in place, it is clear that, as Rawls acknowledged, social and economic inequalities in a modern democratic state are so large that those with greater wealth and position usually control political life and enact legislation and social policies that advance their interests.^[414] Human rights can only be addressed within a democracy through deliberative means. The power of the principle must itself be the reason for adopt­ing it, rather than the numbers of those who back it. It is here that the courts are in a position to make a unique contribution to democratic resolution of human rights issues.

Adjudicating human rights: bounded deliberation

To what extent, then, can courts contribute to a deliberative resolution of human rights issues? At first sight, they are an unlikely option. Adversarial litigation appears to be a paradigm interest-bargaining framework. Each side appears before the court in an entirely defensive or accusative mode, aiming to persuade the court from a rigid position. However, on closer inspection, it is obvious that judicial decision-making is, if correctly applied, essentially deliberative. It is not the parties' political, numerical or economic strength that persuades courts, but the strength of their reasoning. The parties' input is aimed at opening up the range of different responses to judges, on the basis of which they reach their decision. Moreover, in human rights cases, rules of standing and intervention can be adapted so that a wider range of perspectives is permitted to enter into the debate. The decision as to what the ‘principle' should be is itself a deliberative exercise and gains its legitimacy from its deliberative credentials. This also clarifies the role of parties to proceedings in cases of positive duties where there is no immediate benefit to the right-holder. Where a right is programmatic and future-oriented, a successful case would only yield a duty to carry out an approved plan, which may not provide an immediate tangible benefit to the winning party. Instead of a right-holder making a realizable claim to an individual benefit, the role of participants is to bring to the process a particular deliberative solution to the problem.

A more serious concern with applying a deliberative model to human rights adjudication concerns the role of human rights themselves. Deliberative models assume an open-ended approach, allowing the process to produce a solution with no preconditions. Human rights, by contrast, require a prior commitment to the observation of human rights which appears to belie the possibility of a deliberative approach. Indeed, any legal materials which are binding on the court potentially obstruct this approach. It is partly for this reason that Waldron argues that courts are unsuitable for human rights adjudication: far from being able to reason morally, judges are constrained by materials which are only tangentially relevant to the issue at hand. He cites as an example the difference between the legislative approach to abortion in the UK and the judicial approach to the same subject in the US. Whereas the UK Parliament was able to reason openly about the subject at hand, the US court was required to focus on tangential legal issues such as whether privacy was implied into the US Constitution and the role of precedent.

Human rights pose particularly complex questions because they are neither fully determined nor open to thoroughgoing deliberative solutions. If they were fully determined, then both courts and legislatures could simply apply formulaic responses. Both institutions would be bound by the same mandatory norms and neither would be superior. However, human rights are open to a range of interpre­tations in particular contexts. Similarly, the question of whether human rights have justifiably been limited inevitably requires a judgement. On the other hand, human rights are not simply open moral questions. They are based on a consensus which has developed over time and is universally accepted as to what the funda­mentals of being human in a political society require. It is within the framework set by this prior deliberative consensus that current decision-making must take place. Thus human rights place real constraints on both judicial and legislative decision-making, while at the same time being open to interpretation.

It is in this interpretative space that the deliberative approach functions. I argue that the primary deliberative role belongs with the legislature. Human rights deci­sions should be taken in a deliberative, rather than an interest-based mode, and the legislature is in the best position to do so. However, courts have a supervisory role, to ensure that such decisions are indeed taken deliberatively within the con­straints set by the human rights themselves. Thus the primary role of the court is to enhance the deliberative dimension to modern democracy. Judges should do more than remind legislatures of the need to consider human rights: they should insist that legislatures address interpretive disagreements on value-based rather than interest-based grounds. Decision-makers must be in a position to persuade the court that they have fulfilled their human rights obligations, account being taken of the fact that there is room for reasonable disagreement both in the interpretation and the delivery of human rights, as well as in the acceptable limits. Thus, a central function of human rights adjudication is the requirement that the state explain and justify to the court, and therefore to the litigants and the public more generally, the grounds of its decisions and the reason for the selection of particular means. A proper explanation by the state cannot be based on particular private interests, but instead requires reasons which can be publicly advocated. In other words, one of the aims of judicial intervention is to rejuvenate the deliberative process.^[415] For Habermas, this need not amount to a court overstepping the grounds of legitima­tion by substituting for the democratic process. Instead, its aim is to insist on the ‘communicative presuppositions which allow the best arguments to come into play and procedures to secure fair bargaining conditions'.^[416]

Decisions taken which affect groups without the political power to affect the decision require particular attention. Moreover, participation remains crucial. Those affected must be consulted and given an opportunity to vocalize their position, especially when silenced by majoritarian politics. As Baroness Hale has emphasized in the British Supreme Court: ‘It is a purpose of all human rights instruments to secure the protection of the essential rights of members of minority groups, even when they are unpopular with the majority. Democracy values everyone equally even if the majority does not'.^[417] Equality also serves democracy generally. In the famous words of Justice Jackson of the US Supreme Court: ‘There is no more effective practical guaranty against arbitrary and unreasonable government than to require that the principles of law which officials would impose upon a minority must be imposed generally. Conversely, nothing opens the door to arbitrary action so effectively as to allow those officials to pick and choose only a few to whom they will apply legislation and thus to escape the political retribution that might be visited upon them if larger numbers were affected. Courts can take no better measure to assure that laws will be just than to require that laws be equal in operation'.^[418]

At the same time, courts are not entitled to impose their views on open-ended moral grounds. The deliberative approach is bounded in that it operates within the

constraints of human rights, which are themselves a product of prior deliberative consensus. Waldron regards this as a reason why courts should be considered unsuit­able for human rights adjudication: the institutional setting in which judges act and the role they adopt ‘require them to address questions about rights in a particular legalistic way—indeed, in a way that, sometimes, makes it harder rather than easier for essential moral questions to be identified and addressed.' He contrasts this with ‘legislative approaches, which proceed by identifying all the issues and all the opin­ions that might be relevant to a decision, rather than artificially limiting them in the way that courts do.'^[419] However, it is paradoxical to valorize the output of legislative deliberation while at the same time regarding legal materials as irrelevant to judicial decision-making. A Bill of Rights makes judges more rather than less accountable, because they too have to justify their own decisions against a background of values reached by a process of prior consensus.

The model is illustrated well by the UK Supreme Court case of Limbuela,^[420] in which asylum seekers, living rough in car parks and on pavements, claimed that the withdrawal of all social security support, which, together with a prohibi­tion on paid work, left them to inevitable destitution, breached their rights under the European Convention on Human Rights art 3 (not to be subject to inhuman or degrading treatment or punishment). Asylum seekers have no voice in the political process; in the ‘plural bazaar' of interest-bargaining, they would certainly lose out.^[421] Deliberative accountability, however, had a very different result. Decision-makers could not resort to politically charged arguments about the need to deflect unde­serving asylum seekers. Instead, they were required to produce a transparent and reasoned account, capable of convincing others, as to why it was not inhuman and degrading to exclude the most disadvantaged from basic means of survival. Such an account was not forthcoming. The asylum seekers' claims were upheld. As Lord Brown put it: ‘It seems to me one thing to say, as the ECtHR did in Chapman, that within the contracting states there are unfortunately many homeless people and whether to provide funds for them is a political, not judicial, issue; quite another for a comparatively rich... country like the UK to single out a particular group to be left utterly destitute on the streets as a matter of policy.'^[422] The result was not a usurpation of political power, but a deliberative solution based in the pre-existing values in article 3.

It could be argued that litigation derails rather than augments democracy, since it gives a second vote to the litigant in matters already dealt with through the ballot box. However, there is no reason why the individual's only opportunity to participate in political decision-making should be through the ballot box. Where a person is actually affected by a decision, it makes full democratic sense to allow her to participate in the process of decision-making both in respect of the particulars of her own situation and by contributing to the choice of values guiding the decision-maker.^[423] More problematic is the objection that a courtroom, far from enhancing democratic participation, is an arena in which elites are able to augment their already powerful position. Litigation is sufficiently expensive, protracted, and framed in mystifying language, to make it inaccessible to most people. Thus, to the extent that litigation permits the litigants to participate in the political process, it does so in a wholly unrepresentative manner, privileging those with the resources and energy to pursue their grievance.

The current structure of many courts makes this a real criticism. However, it is not a necessary feature of justiciability. If courts are going to be taken seriously as deliberative fora, then accessibility and equality within the courtroom must be a priority. To ensure that courts properly discharge their function of holding decision-makers to account through human rights adjudication, accessibility and equality within the courtroom must be a priority. Legal aid, demystifying lan­guage, open standing rules and an accessible procedure and tone are all possible, as well as use of tribunals, human rights commissions and ombudspersons.^[424] It is only by giving an opportunity to those who cannot participate fully in the democratic process that litigation supplements democracy; not by giving a further platform to those already well represented in that process. Procedural developments in Indian courts demonstrate ways in which courts can be opened up to the most disadvan­taged in society.^[425]

On the other hand, it should not be thought that the political process pro­vides a more accessible or equal alternative. The ability of elites to manipulate and dominate is not confined to the courtroom. The political process, particularly where it involves interest bargaining, is highly vulnerable to such domination. In this respect, justiciable positive duties are capable of correcting rather than reinforc­ing inequality. A key aim of human rights is to give individuals the wherewithal to participate fully in society, including in the political process. Human rights adjudication should make it possible for even the weakest voice to be heard. At the same time, the aim is not to give any individual a veto over the broader process, but to allow all to enter into the deliberative process on an equal footing, both in terms of persuading and being persuaded.

Does the deliberative process not mean that decisions can never be made? This is problematic in a courtroom, where the parties before the court require an answer to their dispute. As Habermas argues, there is a need to reach a point of closure under pressure of time, both for the parties before the court and in creating settled expectations by which others, including state bodies, can organize their decisions and actions. How then can the paradox be resolved between the ongoing deliberative debate, which requires revisability of judicial decisions, and the dispute before the court, which requires closure in decision-making? Clearly, judges should not have the last word on the meaning of human rights values, nor should they be fossilized by non-revisable decisions. However, courts need to be in a position to enforce the decision in the case before them, as well as creating settled expectations and consistent standards.

The way forward is to regard courts' decisions as binding for the issue before them, but revisable in the long term through the dynamic forum of deliberative democracy. This, however, does not differ from the court's role in common law disputes in general. Here too, the court's decision can be revised by Parliament, but the change in the law cannot reverse the effect on the particular parties. At the point of decision-making, the duty is fixed, and the state is required to take action or is absolved from action, as the case may be. But on the broader scale, the decision remains part of a process of continuing revisability, whether through Parliament, case law or public discourse. Similarly, the court itself can, as part of the deliberative process, revise its views. But here the principle of stability requires change to be incremental, so that state officials and private individuals can order their lives according to reasonably settled rules. The cases on parental leave from the ECtHR, cited above, are a good illustration of this process.

A particularly good example of the application of a deliberative model is the litigation in the South African TAC case,^[426] in which the Constitutional Court struck down the government's decision to refuse to permit the use of Nevirapine to reduce mother-child transmission of the HIV-AIDS virus. The South African Constitution gives everyone the right of access to adequate housing,^[427] health care services, sufficient food and water, and social security. ^[428] The corresponding duty of the state is to ‘take reasonable legislative and other measures, within its available resources, to achieve the progressive realisation of this right.'⁶⁸ The case was brought by an NGO, Treatment Action Campaign, which argued that the government's decision could not be justified in the light of its constitutional commitments to take all reasonable measures, within its available resources, to provide access to health care. Notably, there was no difficulty in finding resources, as sufficient quantities of the drug had been offered free for three-to-five years by the manufacturer. However, the government had for many years refused to take effective measures to combat the AIDS pandemic on the grounds that it did not believe that the HIV virus caused AIDS. The result was that, despite the proven efficacy of nevirapine in diminishing the risks of mother-to-child transmission of the virus, the government refused to permit the drug to be administered in state hospitals outside of a handful of ‘pilot schemes'. The court upheld the challenge.

The decision was centrally based on deliberative accountability. As the judgment notes: ‘In our country the issue of HIV/AIDS has for some time been fraught with an unusual degree of political, ideological and emotional contention.'⁶⁹ By requiring the government to prove that its concerns over the safety of Nevirapine were based in evidence, the court introduced a strong requirement of accountability and transparency. It was not enough simply to produce reasons; they were also required to be reasonable. In this light, it was clear that the reasons put forward were lightweight relative to the enormous cost in human lives that the refusal entailed. But the judgment goes further than accountability, introducing a delibera­tive element into decision-making. This is because the government was required to explain its policies in a way which could convince others of their reasonableness, and on terms which were free of ideological or self-interested perspectives. Those who had previously had no voice in democratic decision-making were able to introduce their own perspectives, so that the court could create a synthesis of both. The equal­ity parameter is also evident, in that the chief victims of the policy were the poor who could not afford to buy the drug privately.

Similarly, the South African court has developed a strong participatory dimension, through its notion of meaningful engagement. Thus in the Olivia Rd case, the court required the parties to ‘engage with each other meaningfully... and in the light of the values of the Constitution, the constitutional and statutory duties of the municipality and the duties of citizens concerned' to resolve the dispute.^[429] This is not simply an instruction to parties to settle the case: instead of interest-bargaining, meaningful engagement must follow deliberative prin­ciples, by requiring parties to reach a consensus based on the constitutional rights. Most importantly, the most disadvantaged group (here squatters in der­elict inner city housing) are given a voice in a deliberative forum, when they would certainly have none outside of it. Thirdly, the duty cannot be applied in a discriminatory manner.^[430]

The South African court, however, remains hesitant in the extent to which it ensures that deliberative accountability is followed through. This is reflected in its cautious approach to remedies, and particularly to structured interdicts,^[431] its reluctance to specify minimum obligations,^[432] and specific timetables, its use of balancing rather than proportionality, and, most recently, its preparedness to take government gestures towards ‘meaningful engagement' at face value, rather than insisting on a pure deliberative forum.^[433] Moreover, the deliberative standard is an exacting one, which requires constant vigilance. It is not sufficient to accept any explanation, without demanding proper evidence.^[434]

B.