Non-domination, “rights” and judicial power

The first objection relates to the largely distributive character of republican freedom and correspondingly, the institutional limits of judicial power. As discussed, domination is a function of “relative” powers,^{Footnote 16} and so can exist without interference as such ever occurring. Thus while freedom as non-domination responds to the apprehension and insecurity experienced by those living within asymmetrical power relationships, more generally domination is constituted, in part, in social and economic structures. In this light, judicially enforced constitutional rights may not prove very effective in protecting freedom understood in this broad sense. (Arguably, it might actively entrench domination, although I provisionally bracket this question). Freedom, in this sense, cannot be realised through a constitutional bill of rights enforced by courts against other agencies; more fundamentally, it is irreducible to any enumerated list of specific rights.^{Footnote 17} It is realised, in large part, through distributions of social, economic and political power, that judiciaries are ill equipped to effectuate or even decisively influence.

In the Rousseauan tradition, this objection is foundational and intractable, given the conceptual inseparability of freedom and collective self-rule. Insofar as political freedom requires the rule of a non-arbitrary “general” will across all areas and interests, it cannot be captured as a list of codified rights, understood as immunities or spheres of non-interference, that are defined independently of legislative process. Moreover, since a non-dominating general will must be given expression democratically and in legislative form, political freedom precludes any substantive constitutional restraints on legislative power. Since neo-Roman republicans reject Rousseau’s “authorial” concept of democracy and its emphasis on popular sovereignty,^{Footnote 18} rights-based constitutional review is not objectionable in this foundational sense; in fact Pettit, in particular, embraces “legal constitutionalism” as an important institutional buttress of non-domination, as it allows individuals to contest potentially arbitrary instances of state power, even in a legislative guise.^{Footnote 19} Nonetheless, even if the democratic objection is disregarded, it may still be impossible to capture the social breadth of freedom as non-domination within the formal and institutional constraints of judicial decision-making.^{Footnote 20} Since non-domination depends on people’s “relative” powers^{Footnote 21}—and thus, the opportunities and resources they enjoy relative to others—it requires a strong commitment to social and distributive justice that cannot be captured or realised as a set of judicially enforced individual rights, or might even be thwarted by it. If, as outlined, freedom as non-domination can be undermined simply by power disparities that make some dependent on the goodwill of others, republican social justice—and thus republican freedom—will be effectuated through, for example, social-welfare systems and wealth redistribution, domestic-violence legislation, educational provision, etc., and other measures that are not effectively pursued through constitutional law.^{Footnote 22}

Perhaps there is no reason, in principle, why freedom as non-domination should not be captured or expressed as a set of “rights”. Certainly, some republicans are ambivalent towards juridified rights-culture, insisting citizenship comprises a “continuously reflexive process, with citizens reinterpreting the basis of their collective life in new ways that correspond to their evolving needs and ideals.”^{Footnote 23} However, the republican objection can hardly be to “rights” as such. We might conceptualise rights broadly; we might say simply, for example, that there is a right not to be subject to arbitrary power. Republican scepticism, then, must be directed more specifically towards judicially enforced constitutional rights, both in terms of their efficacy and legitimacy. This objection arises in part because the adversarial format of constitutional adjudication cannot capture the diffuse and polycentric character of the power relationships on which domination is based. Additionally, it stems from the limited remedies available to courts.

Thus, precisely because domination can obtain without interference occurring, it is difficult, in particular—leaving aside institutional limitations—to envisage how complaints of domination could be framed within the form and institutional structure of constitutional rights. Rights-based constitutional review operates, for the most part, in a negative sense, through the invalidation of legislation, or other forms of public interference, that violate specific “rights”. Historically at least, it operated by asserting areas of immunity, for individuals, from state interference. On the one hand, this fails to address problems of private domination; additionally, however, it cannot effectively account for those subtle forms of insecurity, apprehension and servitude that stem from subtle, multifarious power disparities across various social and economic domains.^{Footnote 24} A constitutional-rights complaint must usually establish an “interference” in, or violation of rights. The mere existence of a power differential, the kind of which causes domination, is difficult to frame in this way, not least because there may not be a definite party against whom the complaint can be levelled, and because the complaint embraces a multitude of hypothetical future interferences. (This problem is discussed further in the final section).

It might be argued that these are historically contingent, rather than intrinsic limits on judicial power; indeed, judicial remedies for constitutional rights might be affirmative or declaratory rather than negative, while constitutionalised socio-economic rights might partly capture the distributive nature of non-domination.^{Footnote 25} Yet given the polycentric and complex nature of distributive problems, even successful models of constitutionalised socio-economic rights will not usually define such rights in relative terms—but rather in terms of rights to say, housing, education or health—and so they fail to capture the problem of power differentials as such. For example, a person might enjoy a satisfactory standard of housing and healthcare in absolute terms, but remain economically vulnerable to a capricious family and indeed, dominated in a structural/social way by virtue of patterns of class and wealth reproduction.

Non-domination, then, could certainly, in principle, be formulated as a constitutional right (“nobody shall be subject to arbitrary power”), but judicial enforcement would prove peculiarly inapt to enforce or promote the substance of non-domination, even assuming, provisionally, that it would not actively undermine it. Even judges who might be sympathetic to social and economic dimensions of freedom are limited by what are arguably intrinsic or at least intractable formal limits on judicial power, which make it unsuitable to address the distributive dimensions of domination. And importantly, this remains true without considering any more radical theories of structural or “constitutive” domination focussed on ideology or exploitation, on the control of social rationalities and value systems and so forth;^{Footnote 26} it applies to the relatively conservative account of domination embraced by the neo-Roman writers, which is focused more narrowly on the relational control of individual choice. Accordingly, it seems that the distributive dimensions of non-domination—understood as a pattern of resourcing across all social relationships—can only be realised through legislative and administrative channels, and could only ever be affected by judicial doctrines (or indeed constitutional drafting) in a piecemeal or indirect way at best.

Freedom and political constitutionalism

Leaving aside efficacy, a more fundamental republican objection centres on the political effects and the political legitimacy of strong-form judicial review. Certainly, some republican thinkers, such as Pettit, embrace strong-form judicial review as a vital check on arbitrary power and indeed, as a conduit of principled public deliberation, primarily because they maintain that the question of what constitutes non-arbitrary intervention is best determined through a non-political procedure.^{Footnote 27} Others, however, argue that rights-based constitutional review of legislation is not merely inefficacious in securing freedom as domination, but in fact, actively negates it. Specifically, these scholars argue that freedom, understood as non-domination, is inconsistent with any concept of “rights” that aims to identify areas of immunity for individuals, and thus to define, in advance, those questions that are placed beyond ordinary political contestation. From this perspective, the problem with constitutional adjudication is not merely that it is ineffective as an institutional response to social and economic domination. Rather, in attempting to depoliticise “rights” issues—that is, to take them out of the ordinary political domain and entrust them to an independent judicial sphere—it establishes a form of political domination, independently of issues of efficacy.

As already noted, the Rousseauan strand of republican thought embraced an “authorial” theory of democracy—and thus an unbridled account of popular sovereignty—that necessarily rejects any judicial constraints on legislative power in the guise of constitutional rights or otherwise.^{Footnote 28} Beyond Rousseau, some republicans argue that “rights talk” generally, or a juridified culture opposing legal rights against democratic process, tends to “depoliticise” freedom both conceptually and institutionally, depicting it as a “boundary around the individual … against the state.”^{Footnote 29} Republicans of any ilk will insist that since freedom consists of security from arbitrary power, it is constituted (rather than caused) through politics—not against or despite it.^{Footnote 30} Thus, in the Rousseauan-influenced French constitutional tradition, for example, republican scepticism towards justiciable constitutional rights historically translated to a doctrine of strong legislative supremacy.^{Footnote 31}

Arguably this doctrine leaves individuals and minorities vulnerable to societal animus; in extremis it suggests rights are “nothing but determinations of the prevailing political will”.^{Footnote 32} However, contemporary republican like Richard Bellamy have offered a more subtle defence of legislative supremacy. Bellamy defends “political” against “legal” constitutionalism not from within Rousseau’s demanding “authorial” understanding of democracy, but rather, using the familiar neo-Roman account of freedom as a resilient security against arbitrary power. While Pettit argues that freedom as non-domination requires checks on majoritarian politics, Bellamy insists, in contrast, that the non-domination ideal is best realised through a relatively unbridled and procedural version of majoritarian parliamentary democracy, unconstrained by any justiciable constitution. He argues that we cannot realise freedom, in the republican sense, by somehow settling, in advance, on a concept of rights that is to be placed beyond politics and hived off from the ordinary procedures of contestation. Inevitably, such determinations prove ad hoc and historically arbitrary: a state cannot be built on a durable consensus about “rights”. Contra Rawls, the starting point for political philosophy is the fact that reasonable citizens disagree, in a permanent and intractable way, not only on the good, but also questions of right. Therefore, freedom as non-domination is not realised by enshrining any specific concept or content of rights; rather, it is realised in the institutional process for reconciling competing rights-claims. Thus, it is realised specifically through a majoritarian parliamentary politics which, by according equal weight to each vote, enshrines “equal concern and respect” for citizens who (quite reasonably) disagree on the content of rights.^{Footnote 33} Conversely—as Waldron has equally argued—there is no reason to believe that any specific institution, judicial or otherwise, can claim a privileged perspective on the content of rights. In effect, then, Bellamy—along with Tomkins and others—argues that the indeterminacy (or uncertainty) of “rights”, combined with a principle of political equality, must yield a strongly procedural account of democracy.^{Footnote 34} To map this on to the broader republican lexicon, he effectively contends that public power is non-arbitrary (and thus non-dominating) only if it enshrines political equality (understood procedurally), and in turn that procedural political equality requires a majoritarian democracy unhindered by strong-form judicial review. Thus, “political constitutionalism” is typically defended in republican terms.^{Footnote 35}

Bellamy’s starting point is a realist one—it is not the hope of successfully capturing or enshrining a true or even a politically authoritative concept of rights; rather it lies in recognising, and accommodating the fact of disagreement concerning rights. And this gives a republican flavour to Griffith’s influential claim that “rights” are best understood as political claims rather than as objective moral propositions.^{Footnote 36} He argues that unfettered majoritarian democracy permits on-going contestation about the meaning of rights in a way that respects and enshrines citizens’ political equality. In contrast, the very features that are thought to make constitutional courts uniquely suited to protecting rights—their non-political nature non-contestatory character—undermines political freedom in this republican sense. More fundamentally, it is the very attempt to determine a-politically a substantive concept of rights, to demarcate the limits of politics using rights, that leads to political domination in the guise of arbitrary rule—as “a juridified culture of rights freezes the debate on their contents.”^{Footnote 37} In short, “depoliticisation can fuel domination.”^{Footnote 38}

If Bellamy is correct, it would seem pointless to consider how constitutional-rights doctrine might capture freedom as non-domination—say, to consider whether discrete doctrines and rulings are more or less responsive to domination—because judicially enforced constitutional rights are inconsistent with the ideal in a systemic and structural sense. Indeed while we might argue that judges ought to decide cases in particular ways—in ways, say, that are more responsive to problems of domination in social life—the point is that whatever in fact they choose to do cannot be controlled or contested through any democratic procedures.

Certainly, some republicans will dismiss any attempt to make out a “republican” analysis of an institution they perceive as being fundamentally illegitimate. They might dismiss as pointless those arguments concerning the content of constitutional rights in the abstract, preferring instead to focus on systemic questions of process and form. However, this absolutist rejection of republicanised constitutional rights—based on the political domination intrinsic to the very concept—ignores the diversity and complexity of domination in its social forms and sources and in turn, the necessarily tentative nature of specific institutional solutions aimed at optimising overall non-domination. That is to say, it ignores that domination assumes multiple sources and forms across social and political life. The fact that domination has multiple sources, public and private, means that efforts to increase non-domination in one sphere might bear costs for non-domination in another.^{Footnote 39} This may involve certain trade-offs, e.g., efforts to decrease social and economic domination may inadvertently increase bureaucratic domination. Thus as a consequentialist, Pettit argues that the republican state attempts a kind of calculus of non-domination, maximising overall non-domination.^{Footnote 40} This quality—the fact that it is optimised as a social goal, rather than guaranteed or realised in an all or nothing way—makes it difficult to crystallise non-domination as a strict set of a priori institutional prescriptions, because optimal non-domination will depend on a set of complex, context-specific variables and contingencies, including say, whether or not judicial or political culture, in a particular time of place, is capacious enough to fulfil particular goals or tasks assigned to it.

Therefore, while republicans like Bellamy might be correct in identifying strong-form judicial review as a source of domination—of political domination, specifically—this is only one kind of domination. And political domination, much like the private domination that obtains between individuals and groups, operates as a matter of degree rather than in an all-or-nothing way.^{Footnote 41} The decisive point is the following: a constitutional practice or feature that causes domination (as a matter of degree) in one dimension (e.g. political domination) might attenuate a different dimension of domination (e.g. social domination). Whether this trade-off is desirable depends on various empirical and social contingencies, or at least it is a complex question of political sociology. Crucially, the possibility that this trade-off is worthwhile holds true even if strong-form judicial review is intrinsically dominating, in a conceptual rather than a contingent way. Following Pettit’s consequentialist logic, such an institutional arrangement could never be discounted on republican terms.

Therefore, it follows that is perfectly consistent to accept that strong-form judicial review represents a source of political domination, in one dimension, yet while insisting that particular constitutional doctrines might alleviate other sources or dimensions of domination in more piecemeal or discrete ways—while bracketing the question of whether strong-form judicial review enhances non-domination overall.

Certainly, then, judicial approaches that appear to enhance non-domination in discrete domains form part of a broader system that may well represent a kind of expert technocracy and thus, political domination. Yet equally, republicans might recognise that discrete doctrines within such systems may serve to alleviate some immediate problem of domination in social life—say, by invalidating a statute that grants officials arbitrary power. Contra Bellamy, therefore, the dominating character of strong-form judicial review and indeed, the question of whether it has a net cost for non-domination, is highly context-specific. Given the consequentialist view detailed above—the idea of a “calculus” of non-domination—it follows that we can investigate the positive effects of judicial power for non-domination in one domain while admitting the possibility that it might worsen it in another dimension.

Even republicans who disagree, and who categorically reject the entire institutional edifice of rights-based constitutional adjudication, must still recognise that constitutional rights can be interpreted in ways that are more or less amenable to the republican understanding. Pragmatically, strong-form judicial review is an intractable feature of most contemporary liberal democracies. The most ardent republican opponents of judicial supremacy would have to accept that specific constitutional doctrines might relieve citizens from a particular kind of arbitrary power, say, the arbitrary power of some category of public officials, despite the mechanism itself representing a different kind of domination. And finally, political constitutionalists do not usually reject rights-based judicial review as such, but only strong-form judicial review—i.e., rights-based judicial review that is accompanied by an unqualified power of legislative invalidation.^{Footnote 42} Thus, neo-republicans should naturally be interested in how alternative systems of rights-based review—that are consistent with, or even supportive of republican aims—can effectively incorporate the non-domination ideal.

In summary, then, republicans might doubt whether judicial review is an appropriate mechanism for securing a non-dominating form of government, while acknowledging that, within a given constitutional framework, certain adjudicative approaches might be more or less sensitive to problems of domination in discrete areas of social life.

Interference without domination

Using the template of this two-stage analysis, republican theory can inform our understanding first, of what constitutes “interference” in a right prima facie, and second, what constitutes sufficient justification for such interference. On the one hand, since republicans suggest that interference as such carries no necessary “cost” for freedom, it follows that state actions which coerce the relevant choices or actions will not necessarily count as prima facie infringements of the relevant constitutional rights, of the sort that require some special justification at the second stage of analysis. For example, where a religiously motivated practice is prohibited by virtue of the general application of an otherwise legitimate law that is not itself targeted at religious practice, arguably there is no “interference” in religious freedom at the first stage of analysis.^{Footnote 49} Nor, then, does it have to be subject to any special justification or proportionality test at the second stage, provided that the prohibition meets certain criteria that qualify it as a non-dominating form of interference—say, where it is exercised pursuant to a contestatory legislative process rather than official discretion, and where it tracks interests that are shared by those it affects.^{Footnote 50} Effectively, the U.S. Supreme Court endorsed a similar conceptual schema in Employment Division v Smith. ^{Footnote 51}

This approach defines “rights” in a very particular way—not as granting immunities, or guarantees of non-interference, to privileged areas of activity (e.g. religious practice), but rather as controlling the criteria of state action in particular domains. Thus it makes constitutional rights appear less as guarantees of non-interference for individuals as such than as controls on the predicates and grounds of state action, and specifically, then, as bulwarks against arbitrariness in its republican sense. And indeed this seems attractive from the standpoint of institutional competence: using the religious-freedom example, again, judges seem better equipped to decide whether a law restricting religious practices reflects legislative animus—or spurious policy justifications—than to understand the spiritual importance, or even the theological authenticity of the practices affected (that is, to say whether or not they come within the category of activity that defines the right in question).^{Footnote 52}

This might seem a stark position. Taken to its logical conclusion, it would reject any right, however limited, for religious believers to be exempted from neutral and generally applicable legislation (for example, to use prohibited drugs for sacramental purposes).^{Footnote 53} Facially neutral laws impeding religious practices might count as “arbitrary” interference if they are non-neutral in the sense of being motivated by sectional or sectarian interest; it is less clear, however, whether laws could be overturned simply because they are indifferent or negligent to minority religious needs.^{Footnote 54}

Nonetheless, the republican perspective suggests that the role of the constitutional judge is not to define privileged spheres of activity that are to remain presumptively free of coercive interference or regulation—not least because such decisions often prove somewhat ad hoc. Instead, their role is to police the procedural and justificatory basis for administrative and legislative coercion. Of course, rights are ostensibly defined in relation to specific activities (religion, expression, association, etc.), and Pettit, for example, acknowledges that the republican state will entrench and protect only certain choices against domination (the domain of the “fundamental liberties”).^{Footnote 55} But the point remains that these rights do not protect citizens from interference as such, but only arbitrary interference, albeit within specific domains of activity, and thus they can be seen as instrumental guarantees of non-arbitrary government rather than safeguards against interference as such.^{Footnote 56}

The idea of non-domination—and particularly the republican concept of arbitrariness—can inform the legitimate limits of rights as well as their prima facie scope. Specifically, the legal-doctrinal idea of proportionate “interference” in rights roughly maps on to the theoretical understanding of interference without domination. For republicans, non-dominating interference must pursue a legitimate aim connected to a common public good—that is, an interest shared at some level by all those affected—rather than say, a factional or sectional interest. And this quintessentially republican concern is captured, at least in part, in the existing methodologies of constitutional and human-rights courts. For example, articles 8-11 of the ECHR define the criteria based on which freedom of expression, religion, assembly, etc., may legitimately be limited. These arguably echo the republican concept of arbitrariness in many respects—for example, in the requirements that interferences are “prescribed by law” (rather than based on administrative discretion)—a formula echoed in the Canadian Charter^{Footnote 57}—and that they must be “necessary in a democratic society … or for the protection of the rights and freedoms of others”. National constitutions are often less specific. The rights encompassed by the First Amendment to the United States Constitution (religion, association, expression, assembly, “petition” of government) offer no definition of limiting criteria.^{Footnote 58} The Irish Constitution uses vague formulae such as “public order and morality” (Article 44) or “the exigencies of the common good” (Article 43), while the Canadian Charter refers to the “such reasonable limits … as can be demonstrably justified in a free and democratic society” (Article 1). Yet irrespective of textual and stylistic variations, the two-stage structure of much of constitutional-rights doctrine arguably shows that the function of constitutional judges is to safeguard against arbitrary interferences in rights specifically, and thus that their role can ultimately be understood in light of freedom as non-domination rather than the more familiar liberal concept.

This idea is also reflected in the structure of proportionality doctrine.^{Footnote 59} Generally speaking, proportionality analysis requires that restrictions on rights have a legislative basis, that they pursue a legitimate aim, and that rights are impaired no further than is necessary to realise such aims.^{Footnote 60} Since limitations on rights must be “rationally connected to the objective and not be arbitrary, unfair or based on irrational considerations”,^{Footnote 61} this can be interpreted as a mechanism for ensuring that coercive state interference remains non-arbitrary in the republican sense, in that it tracks “commonly avowable” public interests.^{Footnote 62} It is true that in some variants, at least, the proportionality test will effectively require judges to make substantive, even intuitive appraisals as to the relative merits of competing claims of public and private right, and that this implicates judges in a more far-reaching normative endeavour, going beyond an assessment as to arbitrariness in its republican sense. However, the more technical aspects of the proportionality doctrine—particularly the requirement of a rational connection between the means and end of a restriction on rights—helps to show, again, that constitutional rights are not moral rights, institutionalised, but rather, controls on the predicates of state power. In turn, this makes rights-based constitutional adjudication quite consistent with a “political” concept of rights, and accommodating of the fact of “disagreement” about rights.

This understanding of constitutional rights has received support beyond the domain of republican theory. Adler, for example, argues that American constitutional law is not concerned, primarily, with the moral rights of individuals, but rather the “predicates of Government action”.^{Footnote 63} This is given expression, as already argued, in the Supreme Court’s recent jurisprudence on religious liberty, which focuses not on the religious harm caused by government actions (say, with reference to the spiritual importance of the restricted practices) but rather on whether laws restricting such practices are “neutral and generally applicable.”^{Footnote 64} Similarly, Kagan has argued that free-speech doctrine is focused primarily on the purpose rather than the effects of state action.^{Footnote 65} And in turn, if constitutional review focuses on the “pattern, purpose or context of government action” rather than “some specific aspect of one’s liberty or welfare”,^{Footnote 66} this illustrates, again, that constitutional rights are not concerned primarily with preserving, for individuals, a pre-defined sphere of non-interference, but rather with checking arbitrariness in state power.

This is reflected in standards of review other than proportionality. For example, it has been argued that the American “compelling state interest” test (“strict scrutiny”) “requires courts to determine the scope of rights with reference to the justifications government offers for limiting them”—precisely because “rights are better understood as means of realizing certain collective interests”^{Footnote 67} rather than as definite, “pre-political” spheres of non-interference for individuals. Similarly, the concept of “balancing” limits on rights against communal interests can be interpreted as an attempt to define those reasons or criteria that constitute “impermissible justifications for state actions in different spheres”—and thus, again, as a means of realising “those specific common goods the pursuit of which is the point of recognising particular constitutional rights.”^{Footnote 68} In short, then, constitutional rights “police the kinds of justifications government can act on in different spheres rather than protecting atomistic interests in autonomy, or liberty, or dignity.”^{Footnote 69} Pildes argues that if rights are understood as “channelling the kinds of reasons government can invoke when it acts in certain arenas”,^{Footnote 70} they are not, then, “trumps” on the common good—as Dworkin argues—but rather its conduits and its signposts. Again, this feature gives constitutional rights a potentially republican flavour.

Domination without interference

I have argued that, from the republican perspective, basic liberties are undermined, in a prima facie way, not by coercive interference as such, but only by powers of arbitrary interference. I have also argued that this is implicitly reflected in some familiar constitutional-rights doctrines. However, conversely the republican analysis suggests basic liberties can be infringed without any “interference” actually occurring—and this dimension of republican freedom is less well accounted for in constitutional doctrines. As explained previously, domination without interference occurs where citizens are exposed to the arbitrary power of others—in essence, to their discretion and goodwill—and therefore have to self-censor or self-invigilate in anticipation of possible interference or reprisal. This stems partly from the psychological effects of interpersonal power disparities on individuals’ life plans, and the evasive strategies dominated individuals will need to deploy to stave off interference.

If domination, and thus unfreedom can obtain without interference as such ever being suffered, how can this be accommodated within the constricted conceptual landscape of constitutional rights? While I have sketched the apparent problems previously—in particular, the fact that a complaint of domination is difficult to frame within an adversarial judicial format—there are nonetheless limited, but significant ways in which constitutional rights protect persons from the “mere” apprehension of unchecked power.

On the one hand, domination without interference underlines the need for protections against unchecked administrative discretion. Without such guarantees, citizens might suffer no interference as such, but nonetheless find themselves forced to engage in evasive strategies to ward off apprehended interference. While this dimension of non-domination will likely be protected, in the public-law sphere, through common-law principles of judicial review, it might also inform the scope of certain constitutional rights (that is, it might inform judicial review of legislation as well as of administrative action).

On the other hand, domination without interference might equally result from disparities of bargaining power in the “private” sphere. And this is more difficult to capture within the traditional strictures of constitutional law. For example, the freedoms of religion, association and expression might be undermined by the powers of invigilation or intimidation wielded by private individuals as much as by those powers of interference held by the state. Where the right to change one’s religion, for example, is conditioned by economic pressures within a family or community, this can hardly be accounted for, in full, in the doctrinal landscape of constitutional religious freedom.

Thus, again, it bears repeating that non-domination—defined by the resources and guarantees we need to resist arbitrary power in whatever form—will be realised through a “pattern of resourcing” across social relationships generally—of the sort that enables us to “walk tall and look others in the eye.”^{Footnote 71} Therefore, the polycentric nature of distributive questions—difficult to capture in an adjudicative format—means that non-domination will inevitably be effectuated at various sub-constitutional levels—for example, in the security or paths of exit that a welfare state might provide.^{Footnote 72}

Moreover, this is conceptually intractable: it applies no matter how extensively or radically constitutional rights might be re-cast. This is not to claim that say, “positive” or socio-economic rights cannot effectively be constitutionalised;^{Footnote 73} perhaps constitutional judges could usefully be empowered to protect say, rights to housing or subsistence.^{Footnote 74} Similarly, courts in some jurisdictions have allowed “horizontal” enforcement of constitutional rights against private actors, such as trade unions or private employers, as well as public and governmental agencies.^{Footnote 75} But as already argued, these innovations fail to capture the complex distributive nature of non-domination as a “status” freedom, the fact, that is, that it depends on the resources people possess in relative rather than absolute terms—such that it “consists in relating to [others] on a pattern that rules out domination [and in] command[ing] a certain standing amongst your fellows.”^{Footnote 76} Potentially significant forms of domination, for example, might occur where individuals use their economic leverage to control family members’ life choices. Such routine problems could never be captured by “rights to” specific benefits or entitlements, or whatever kind. And a generic right “not to be dominated” seems too elusive and complex, in its content, to be entrusted to judicial enforcement.

Of course, problems of private domination will be more typically addressed by criminal and private-law doctrines rather than constitutional rights, while “horizontal” constitutional rights might offer an additional safeguard for individuals affected by various kinds of private domination which have not, for one reason or another, been regulated by statute or common-law doctrines. But the broader point remains that, if private domination is constituted by patterns of distribution, such innovations will have only a piecemeal effect, at best, on an overall political project of freedom as non-domination. It is this feature, ultimately, that makes non-domination difficult to shoehorn within an adversarial judicial format. Domination obtains in the possibility or the apprehension of some indefinite future interference, based on power imbalances which are necessarily polycentric and multifaceted in nature. A dominated person will apprehend and stave off potential interference from multiple quarters, whether a capricious employer or a domineering spouse. This gives it a diffuse nature that makes it difficult to frame as a specific “rights” complaint. The fact that freedom is denied by a power imbalance as such, and thus, the possibility of a range of future hypothetical instances of interference, makes domination intrinsically difficult to frame as a cause of action in law, partly because the class of persons who may interfere is naturally indefinite. Legislation, and other more generalized state actions, seems more responsive to these features of domination.

Nonetheless, while constitutional doctrine can hardly account for all the social pressures that are brought to bear on constitutionally protected activities, the idea of non-domination might be addressed in more piecemeal ways within a realistic domain of constitutional law.

Arbitrariness, Interference, Discretion

In concrete terms, what would it mean for a court to identify a breach of constitutional rights based on a power disparity per se, that is, domination without interference? Usually, in order for a breach of a constitutional right to be recognised it is necessary that some form of “interference” has occurred, typically a prohibition or restriction in a defined sphere of activity. And generally speaking, the insecurity or fear that citizens may experience as a result of power differentials, or disparate bargaining resources, is not a cause of action in constitutional rights. As discussed, there is probably good reason for this, given the polycentric nature of distributive questions and the limited scope of judicial remedies. For example, if say, economically dependent persons hesitate to renounce their religion due to a fear of familial reaction, this is a real constraint on freedom of religion in the neo-Roman sense, but it could hardly ever be framed as a constitutional religious-freedom complaint.

Despite these intrinsic limitations on judicial power, constitutional rights might nonetheless account for domination-without-interference within a certain, limited domain. It is already accounted for, in certain modest ways, in existing constitutional doctrines. As outlined, people suffer domination where they live in apprehension of interferences that may occur, even improbably, with a change in circumstance or fortune, or in others’ goodwill. An obvious example arises where individuals live in apprehension of a statute being used to sanction particular forms of conduct, even where they have never suffered any such sanction and it is improbable ever to occur. The idea that constitutional freedoms may be undermined solely by the apprehension of interference is reflected, for example, in the law of standing (locus standi). In many instances, a litigant will not need to demonstrate that interference has in fact occurred, but only establish a possibility or fear of interference in order to challenge the constitutionality of a law. For example, in the Irish case Norris v Attorney General,^{Footnote 77} a gay senator challenged the constitutionality of a Victorian-era statute that criminalised sexual intercourse between men, but no prosecutions were taken under the law in the decades preceding the case, and Norris himself faced none. Interference as such had never occurred, and it seemed highly improbable it ever would. Nonetheless, and although Norris’ substantive arguments were rejected by a majority of the Supreme Court, it accepted that he enjoyed standing to challenge the law. It effectively recognised that constitutional rights may be undermined by the state of uncertainty or fear as to the possibility of future interference in some protected conduct, as much as by the actual experience of such interference. Henchy J (dissenting) observed:

Elsewhere, the Supreme Court stated it is obliged “not merely redress a wrong resulting from an infringement of the [constitutional] guarantees but also prevent [their] threatened or impending infringement.”^{Footnote 79}

Certainly, in some instances standing rules are drawn more narrowly. For example, in United States constitutional law some authorities have insisted that injuries suffered must be actual or “imminent”, rather than merely apprehended—which seems closer to the logic of non-interference than that of non-domination.^{Footnote 80} Yet Pildes argues that an insistence on litigants suffering direct and actual injury may reflect separations of powers concerns rather than any essential feature of the “rights” in question; it is quite consistent with the aforementioned understanding of constitutional rights as instruments of the “common good” rather than as protections for atomistic individual interests.^{Footnote 81}

And going further again, some constitutional systems will allow challenges to legislation not only from individuals who merely apprehend interference, but also from those who suffer no particular harm or prejudice at all, even in an apprehensive or anticipatory sense. For example, rules of standing might be waived where no particular citizen is especially affected by a government wrong (i.e., an unconstitutional ratification of a treaty).^{Footnote 82} Under the “overbreadth” doctrine in American constitutional law, citizens whose conduct can constitutionally be regulated can challenge a law that regulates the relevant category of conduct (e.g. speech) too broadly and thus, which violates the rights of hypothetical third parties.^{Footnote 83} And in European-style systems that practice “abstract review” of legislation via specialised tribunals, this feature is obviously present given the absence of any individual litigant.

This supports, again, the idea that constitutional justice may be concerned with the predicates of government action (and thus non-arbitrariness) rather than definite rights of non-interference as such; moreover, that citizens can challenge laws without suffering any direct prejudice themselves supports the idea that constitutional justice can be understood as a mechanism for checking government power, even where such challenges are shoe-horned in the terminologies of “rights”. It supports the idea that the starting point for “rights” is not a moral concept of the person that yields, in turn, certain immunities or spheres of non-interference, but rather a particular concept of government.^{Footnote 84} Rights, in this sense, are merely “linguistic or rhetorical tools the law deploys for pragmatic reasons and aims” and “techniques by which courts police the kinds of purposes government can offer to justify its action.”^{Footnote 85} Of course, such normative concepts of government—say, non-arbitrary rule—are not unconnected to moral understandings of individual “rights”. Yet “rights”-based constitutional adjudication does not necessarily take, as its starting point, some idea of defined areas of “non-interference” for individuals: it is consistent with an idea of rights as “political” or even as essentially indeterminate.

Perhaps constitutional rights, then, are not “subjective” rights at all,^{Footnote 86} and citizens are “rights-holders”, constitutionally speaking, only in a highly procedural and instrumental way, in that they may challenge the predicate, criteria or purpose of state action in some domain where they happen to be affected. Some constitutional provisions ostensibly relating to the structure, predicates and purpose of government—for example, the American prohibition on “establishment” of religion—have been understood being nothing more than corollaries of individual rights, such as religious freedom.^{Footnote 87} However, on a republican view, sceptical of “rights talk”, the reverse idea is more compelling: constitutional “rights” are not definite moral rights, for individuals, but signposts for government power. The fact that such provisions are, in the Anglophone world, invoked through judicial review by individuals can lead to confusion concerning their scope and moral structure. That a constitutional injunction on state power is phrased in “rights” terms may be little more, then, than linguistic and procedural accident. Again, this suggests that the institutional edifice of constitutional rights is not inconsistent with Bellamy’s insistence that “rights” in the narrow sense—of what legal immunities people enjoy—must be left open to on-going political contestation as part of the “circumstances of politics”.

Standing rules aside, there are ways in which substantive constitutional doctrine captures domination without interference. For example, constitutional rights to personal liberty may be violated where criminal offences are defined too vaguely or imprecisely.^{Footnote 88} Again this reflects the familiar republican idea that unfreedom may be experienced by those who never suffer any sanction or interference but whose conduct is self-restricted as a result of apprehension and fear. And again, this shows that constitutional rights are rarely defined in relation to a sphere of non-interference for some specified, privileged sphere activity; they are often silent as to which kinds of human activities are deserving of privileged protection, focusing instead on the arbitrariness of the means of restriction or regulation. As noted in the context of religious-freedom law, whether or not domination is suffered will not depend, primarily, on the importance of the activity restricted—or of the interests underlying the interference—but rather the process through which it is restricted.

Similarly, constitutional rights might be violated by the excessive discretion enjoyed by officials in respect of the exercise of a particular choice or activity rather than by the fact of interference in that choice as such. Most fundamentally, certain activities might permissibly be restricted by clear statutory language but not by executive, administrative or judicial discretion. Indeed, republican thought defines arbitrariness partly in relation to discretion, in the sense of powers that are exercised based on the private judgement of the decision-maker. Pettit suggests that an “arbitrary” act is one that is “chosen at the agent’s pleasure.”^{Footnote 89} However some degree of discretion is, of course, inherent in adjudication and administration alike, given the indeterminacy of legal rules.^{Footnote 90} Certainly, non-domination can only ever be achieved as a matter of degree, and so republicans might legitimately seek to minimise rather than eliminate discretion. However, a republican social politics concerned with combatting non-domination in various social domains will, correspondingly, increase the size and scope of the public administration—and thus in turn, the extent and range of discretionary powers enjoyed by administrators. Dicey argued that an expanding administrative bureaucracy would undermine the “rule of law”, by allowing state coercion to be exercised based on officials’ subjective preferences rather than enacted legal rules.^{Footnote 91}

Certainly, progressive social policies will necessarily require giving potentially extensive discretionary powers to public agencies. Arguably, a republican conundrum emerges because attempts to alleviate social and economic domination through the administrative state, by redistributing power and resources, will likely entail a corresponding increase in bureaucratic discretion and thus (arguably) increased domination. Even with the requirement of legislation authorisation, there is always a risk of administrative abuse even within the discretion permitted by the relevant statute.^{Footnote 92} Since non-domination is a social calculus and a matter of degree,^{Footnote 93} perhaps bureaucratic domination is an acceptable trade-off for enhancing non-domination in the economy and society: more intense forms of domination would arise were the state to retreat to its minimal functions.

However, the scale of the problem should not be overstated: the ground need not be ceded to Dicey and Hayek. On a closer reading, Pettit does not endorse the strict Diceyan view that regards discretionary power per se as intrinsically arbitrary. What he regards as “arbitrary” is not a decision that is made based on the subjective judgement or preference of the agent as such, but rather one made based solely on the agent’s “pleasure”. An act is arbitrary, he says, if it is “subject just to the arbitrium, the decision of judgement, of the agent”, or where an agency can act “just as their own whim … leads them.”^{Footnote 94} Arbitrary power exists where “the only check on the interference that [agents] can inflict is … their own unchecked judgment.”^{Footnote 95} Thus discretion that is exercised within a sphere of legislative authorisation or constraint entails no arbitrariness as such; only “untrammelled” or “unchecked” discretion entails domination.^{Footnote 96}

What Pettit does insist is that non-arbitrary public power must “track” interests and concerns shared by those upon whom it is exercised, i.e., public or “common” interests rather than the sectional or factional concerns of the power-wielder.^{Footnote 97} Thus “the acts of interference perpetrated by the state must be triggered by the shared interests of those affected.”^{Footnote 98} It is clear an act can be discretionary, in the Diceyan sense, while meeting this criterion; it can be based say, on policy judgment rather than legislative criteria, yet without embodying factional considerations and concerns. Pettit’s concern, then, cannot be that all administrative decisions are understood as applications of legislative intent or that they are somehow legislatively determined, but that they are subject to safeguards which ensure they track common rather than sectional concerns.

In turn, this republican concern for controlling arbitrary administrative power is reflected—at least obliquely—in a number of familiar doctrines of constitutional jurisprudence. It is well established, for example, that limitations on basic liberties must have a legislative basis,^{Footnote 99} or at least be “in accordance with law”, using the ECHR phraseology. Similarly, through the doctrine of proportionality, domestic constitutional courts have acknowledged the central role of parliamentary legislation in reconciling the exercise of constitutional rights with public interests^{Footnote 100} Of course, a requirement of legislative authorization for administrative interference provides no absolute guarantee against arbitrariness in Pettit’s sense, as legislation, using vague criteria, may accord wide, almost unlimited discretion to decision-makers. Thus, constitutional and human-rights jurisprudence does not usually insist merely that administrative interferences in rights be legislatively authorised in the bare sense. It has been recognised in ECtHR jurisprudence, for example, that laws authorising limitations on rights must indicate the extent of the executive discretion through which such limitations are imposed in particular cases.^{Footnote 101} Yet even where administrative interference is both legislatively authorised and constrained by legislative criteria, it will usually be exercised based on a degree of non-legal judgement or appraisal. The fact that a decision is determined, in part, by non-legal standards does not make it arbitrary in Pettit’s sense; conversely, however, the mere fact of legislative authorisation for discretionary decisions is not an absolute safeguard against the abuse of discretionary powers for sectional or factional purposes.

Therefore, in addition to legislative authorisation and the (weaker) requirement of legislative constraint, non-arbitrariness will require procedural constraints that can prevent statutory discretion being abused for sectional or factional purposes. Pettit insists that an act is arbitrary “by virtue of the controls—[or] specifically, the lack of controls under which it materializes [rather than] the particular consequences to which it gives rise.”^{Footnote 102} And the potential abuse of statutory powers is checked—to a limited degree at least—through the rights of procedural fairness that control the exercise of discretion in common-law systems. For example, natural-justice principles oblige decision-makers to give those affected by their decisions an opportunity to be heard, to refrain from taking irrelevant considerations into account, to provide reasons, and to avoid any perception of bias.^{Footnote 103} Thus while administrative decision-making is irreducible to legal control or determination—while it is inevitably based, to a degree, on non-legal standards and judgements—the requirement that decision-makers must hear those affected will, in practice, at least substantially reduce the arbitrariness of such powers in the republican sense. Indeed, Pettit suggests that there are “two generic constraints” that “reduce arbitrariness” in public power: on the one hand, “penalties” which prevent interference with “impunity”, and on the other, “filters”, which check the predicates of administrative action. These force the state “to track relevant interests and ideas”; they “put preconditions of action in place which make sectional interference that much more difficult.”^{Footnote 104} Examples of such “filters”, I argue, are effectively found in the quasi-constitutional requirements of natural justice that courts in common-law countries impose on power-wielders. The two fundamental requirements of natural justice are that persons affected by administrative decisions are heard (audi alteram partem), and that the decision-maker should have no vested interest in the matter (nemo iudex in causa sua). Both, effectively, serve to vindicate Pettit’s requirement that administrative interference should track interests shared by those affected, as distinct from sectional or illegitimate interests. They reflect both a substantive idea of non-arbitrariness (i.e., that the power wielder track relevant interests) and a procedural understanding (i.e., that the relevant decisions are checked by suitable procedures).^{Footnote 105} Pettit does insist, additionally, that while state power generally (presumably in its legislative guise) should be subject to an equally shared system of “popular control”, state actions should also be contestable by those affected. Thus “filters”, it seems, cannot be checks that operate independently of those affected: “it must always be possible”, Petit says, “for people in the society … to contest the assumption that the guiding interests and ideas [of state power] really are shared and, if the challenge proves sustainable, to alter the pattern of state activity.”^{Footnote 106} In practice, contestability is secured, first, by the requirement that individuals affected must be heard by decision-makers, but secondly, also, in the quasi-constitutional right to access judicial review of administrative actions.^{Footnote 107}

In summary, then, the gist of existing constitutional tradition roughly corresponds to Pettit’s neo-republican understanding of non-arbitrariness in the administrative and bureaucratic contexts.

Freedom and equality in constitutional jurisprudence

Thus far, I have focused primarily on the constitutional control of discretionary administrative powers. However, the question arises how constitutional doctrine can respond, if at all, to problems of private domination. Such domination may stem from economic disparities—which, I have argued could only ever be affected in a piecemeal way by constitutionalised socio-economic rights—while “horizontal” rights will similarly have a limited effect against abuses of private economic power. Relatedly, however, discrimination might constitute a significant source of private domination. Finally, therefore, I will argue that constitutional jurisprudence may accommodate the idea of non-domination in its approach to equality, and specifically by accounting for the interrelationship of freedom and equality. Neo-Roman republicans like Pettit have argued that equality is simply instrumental to an overarching goal of non-domination. Pettit, in particular, argues for a “sufficientarian” (as distinct from egalitarian) concept of distributive justice, which aims to redistribute resources and wealth to the extent necessary to prevent domination.^{Footnote 108} Yet similarly, equality in the narrower sense of non-discrimination may also be understood as instrumental to non-domination. While discrimination constitutes “interference” in its broad sense (where used as a means to limit choice, or as a form of agenda-setting), a licence to discriminate is, quintessentially, an instance of arbitrary power. Those who lack resilient legal protection against discrimination may be forced to engage in strategies of fawning, self-ingratiation and self-censorship—which republicans understand as salient aspects of domination—even where no such discrimination is ever suffered. Where other agents have capacity to discriminate against us, this will likely represent an “alien will” at work in our lives as we are forced to suppress or disavow the relevant characteristic, identity or choice.

Correspondingly, anti-discrimination statutes can be understood as serving the value of freedom (as non-domination), rather than promoting equality as a separate value. In constitutional jurisprudence, however, freedom and equality have often been understood not only as conceptually separate but also as competing and antagonistic values. In this understanding, strict non-discrimination principles limit and undermine individual freedoms, particularly freedom of religion and association. And in turn, constitutional freedoms trump equality where these values conflict. Thus, in constitutional doctrine, equality (in the sense of non-discrimination) usually features only as a policy-based limit on freedom (e.g., freedoms of association and religion) rather than as a freedom-based goal in its own right.^{Footnote 109}

However, freedom as non-domination challenges this traditional doctrinal stricture. Insofar as non-discrimination is understood as a safeguard for individual freedom rather than a limit on it, this might, on the one hand, support a broader interpretation and application of constitutional equality clauses that are usually applied only to state actors in the narrow sense—for example, the “equal protection of the laws” clause in the United States Constitution or the equality “before the law” clause in the Irish Constitution.^{Footnote 110} These have been interpreted as inapplicable to private bodies, even where publicly funded or where exercising public functions.^{Footnote 111} Yet the understanding of non-discrimination as a freedom-based concern might support a more generous demarcation of those actors to whom such clauses apply.

Moreover, constitutional equality law has been demarcated narrowly, in this sense, partly because discrimination by private bodies has been considered an exercise of freedom of association and of religion.^{Footnote 112} Accordingly, in jurisdictions like the United States and Ireland, equality legislation has been limited by virtue of a pervasive understanding of equality and individual freedom as antagonistic constitutional values. Some forms of discrimination, it is assumed, are an exercise of constitutionally protected freedoms—and since freedom must be understood as taking precedence over equality in the hierarchy of constitutional values, legislation prohibiting certain forms of discrimination—say, discrimination by religious bodies—may be regarded as unconstitutional. For example, the Irish Supreme Court held that a constitutional religious-equality principle could not be applied to private bodies, even where publicly funded, partly because this would “lead to a sapping and debilitation” of religious autonomy,^{Footnote 113} and that denominational exemptions from equality statutes were necessary to give “life and reality” to religious freedom.^{Footnote 114} Similarly, exemptions in equality statutes are typically justified based on an understanding of freedom and equality as separate and conflicting goals.^{Footnote 115} In United States constitutional law, first-amendment freedoms are also used to trump and negate equality guarantees.^{Footnote 116} For example, religious freedom prevents equality protections from being applied to certain categories of religious hiring (posts equivalent to “ministers” of religion).^{Footnote 117} Yet while the “ministerial exemption” frames religious freedom as a freedom of non-interference, conversely it ignores the freedom-based interest of those who are discriminated against, or who face employment insecurity: indeed this kind of interest cannot be captured by the prevalent non-interference concept. And where “freedom” takes this narrow form, correspondingly equality receives a relatively low priority.

However, insofar as equality legislation can be valued as a safeguard against domination—and thus unfreedom—in various “private” relationships, this undercuts freedom-based arguments against the extension of non-discrimination principles to these domains.^{Footnote 118} From this perspective, the dichotomy of freedom and equality overstates the degree to which discrimination constitutes an exercise of freedom, and more importantly, it overlooks how discrimination itself can represent a form of unfreedom for those who experience it. For example, those who are subject to religious discrimination in the provision of services, say in school admissions, may experience domination where they are forced to exercise discretion and self-censorship—for example, by feigning religious exercise or belief, or indeed have their children baptised, in order to secure enrolment.^{Footnote 119} In this sense, derogations from equality law work to undermine the freedom of those with relatively lesser resources and power. Those who may feel obliged to obtain a baptismal certificate, for example, in order to secure enrolment to a school, may experience domination without ever being subject to “interference” as such: it is rather the power to interfere arbitrarily—in the guise of a power to discriminate—that compels them to check and distort their life plans. In this light, any constitutionally enshrined licence to discriminate, on religious or other grounds, empowers agents to arbitrarily interfere in, and thus dominate, religious (or other) choices that others are in a position to make. And in the case of the countervailing “freedom” claim—the non-interference claim against equality law—the “interference” in question is hardly arbitrary, since it is imposed legislatively rather than on a discretionary basis, and since it tracks an interest—that of equality in access to public services—which all affected persons share.