Moral Pluralism and Constitutional Horizontality

Abstract

Despite the growing influence of constitutional rights over the regulation of horizontal (private) relations, many aspects of this trend remain under-theorized. This article criticizes four ideal-typical constitutional horizontality models for failing to accommodate moral reasons that must shape this regulatory practice: the state action model ignores basic consequentialist aspects of political morality; the direct application model ignores basic relational aspects of interpersonal morality; the strong indirect model recognizes both but subordinates the latter to the former; and the partitioned indirect model recognizes both but separates them too strongly. This article claims that a composite indirect model, which reflects basic features of the common law, can better realize constitutional rights through private law in conditions of moral pluralism: it can expose private law to constitutional rights-based and reform-oriented scrutiny without ignoring, eroding, or distorting the unique normativity of private relations and practices or their underlying values.

1. Introduction

Our galaxy, the Milky Way, is on a collision course with Andromeda, a neighbouring galaxy. ¹ When the gravitational fields of galaxies disturb one another, it can result in relatively minor changes (for example, stars might be ejected); but it can also result in more dramatic changes, like a violent merger that involves the destruction of stars. The results depend on a host of factors, like the composition, direction, and speed of the colliding galaxies.

The normative universe also oscillates between order and chaos. This article explores the collision of two normative galaxies: the constitution and private law—or, more accurately, their rights-based cores. Early signs of the collision are already visible in many legal systems: for example, tenant evictions collide with the right to private life and home; ² religiously-motivated refusals to provide services to gay customers collide with the right to equality; ³ monitoring private correspondence at the workplace collides with the right to privacy; ⁴ and certain forms of expression collide with the rights to privacy or dignity. ⁵

One of the basic themes in constitutional and legal human rights practices in the past decades is their expansion. ⁶ The entailment of positive duties—to do rather than only avoid certain actions—is one direction of expansion. Another direction is the horizontal entailment of duties for private rather than only for public agents. These directions of expansion do not necessarily go together: rights (say, against torture or slavery) can entail negative duties for private agents. However, as we will see, the more complex type of horizontality involves positive duties to protect or promote rights whose realization and enjoyment are shaped by the actions and activities of private agents. This challenge is greater because it brings public responsibilities and consequentialist considerations to bear on private actions and activities, whose regulation tends to focus on personal responsibility and nonconsequentialist reasoning and therefore inherently resists ‘external’ normative influence.

This complexity reflects the background condition of moral pluralism: the fact that we cannot systematize all moral values and that some forms of evaluation and reasoning are not only inherently incompatible but also pull in opposite directions. ⁷ The first goal of this article is to explain how this type of pluralism is reflected in the legal phenomenon of constitutional horizontality. The basic claim will be that while modern liberal legal systems have good moral reasons to recognize constitutional rights with a teleological-consequentialist structure and private law rights with a deontological-relational structure, recognizing such rights-clusters puts the constitution and private law on a collision course.

The second goal of this article is to see how violent the collision must be. While we can mostly observe galactic mergers, the understanding of normative conflicts can pave the way for their resolution or the alleviation of their negative outcomes. As Isaiah Berlin noted, while ignoring the tragic reality of moral pluralism is “either self-deceit or deliberate hypocrisy,” ⁸ “collisions … can be softened.” ⁹ I criticize, in this regard, four types of constitutional horizontality models for inadequately responding to this collision.

On one end of the spectrum is the state action model (Section 2), which reflects the common reading of the Bill of Rights of the United States. Libertarian in spirit, it ignores basic consequentialist aspects of political morality by seeing constitutional rights as constraining only the state and only when it is directly or intimately involved in harms to protected interests. On the other end of the spectrum is the direct application model (Section 3), which reflects South African and Irish constitutional principles and a basic strand in the theory and practice of the human rights duties of corporations. Utilitarian in spirit, it ignores basic relational aspects of interpersonal morality by seeing constitutional rights as imposing duties on any agent—public or private—that can effectively contribute to their realization in society.

Along the spectrum lie indirect models, which see constitutional rights as binding only public agents, but also with duties to shape private law in ways that adequately realize these rights. Such models recognize both consequentialist and relational considerations as relevant for constitutional horizontality. A strong indirect model (Section 4), reflected in European human rights law (largely following German constitutional principles), subjects state agents to positive duties to accommodate constitutional rights in private law to the extent that consequentialist reasons require (proportionality tests frame this inquiry). This model threatens to ignore, erode, or distort the unique normativity of private relations and their underlying values. A partitioned form (Section 5), adopted in the United Kingdom and Canada, subjects non-judicial decisions about private law to consequentialist scrutiny and tasks courts with its incremental application and development. This model is too unstructured and consequentialist in non-judicial settings and too rigid and relational in judicial settings.

In presenting these models, my aim is not descriptive: there are variations and combinations of them in practice. ¹⁰ My aim is more theoretical and normative in nature, and therefore I focus on ideal-typical versions of them. Relatedly, I do not explore the social, political, or legal conditions in which these models developed: for example, the influence of court or federal structures or the political culture. ¹¹ My focus is on how the models shape the legal system’s interactions with morality, given that a great part of law’s justification depends on the success of these interactions. ¹² The failures of these models teach us important lessons about how the legal system can and should rely on moral ideas to better respect, protect, and promote constitutional rights through the regulation of private relations.

Building on these lessons, the third goal of the article is to offer a composite indirect model (Section 6). It accommodates both consequentialism (contra state action) and relationality (contra direct application); applies them in their endemic normative environment (contra strong indirect effect); and allows them to interact (contra partitioned indirect effect), both as first-order reasons (to create or change private law) and as second-order reasons (about when and how to engage first-order reasons). It sees constitutional horizontality as a dynamic, evolutionary, and discursive practice in which private agents and all state branches participate. In the spirit of the common law, this practice transitions between incremental development that focuses on private relations (the normative habitat of relationality) and moments of systematic calibration (the normative habitat of consequentialism). This reflects an acceptance that a state of ongoing moral tension is not a problem to be solved but a normatively healthy feature of liberal legal systems, reflected in the practice of constitutional horizontality. Thus, while this model holds that constitutional rights normally obligate state agents to realize them in all social spheres, state agents are not obligated to subject private relations or practices (from university admissions to workplace surveillance) to constant results-oriented scrutiny: state agents must consider context-sensitive second-order reasons about how to bring the progressive aspects of constitutional rights to bear on private relations and practices without ignoring, eroding, or distorting their unique normativity or underlying values.

My claims about these models are partial and by no means settle the choice between them. The all-things-considered justification of constitutional horizontality models—including the institutional structures that support this practice and the roles they ascribe to different actors—must be contextual and take into account the features and circumstances of particular rights and legal systems. My focus is on one important cluster of considerations that has received less attention in the theoretical literature: the ability of liberal legal systems to accommodate both relational and consequentialist forms of evaluation and reasoning. ¹³ The underlying moral premise is that both moral clusters have important roles to play in the practice of constitutional horizontality: it must accommodate relational morality (reflected in basic aspects of private law) for constitutional rights to be effectively and justifiably realized in private relations and practices; and it must accommodate consequentialist morality (reflected in basic aspects of constitutional rights) for the regulation of private relations and practices to enjoy legitimacy and justification as part of a liberal constitutional order. It is in this regard that the proposed model shows great promise.

2. State Action

This model holds not only that constitutional rights can be claimed only against the state but also that their prescriptions and entailed duties address only actions or activities in which the state is directly or intimately involved: both when rights are grounded in interests particularly sensitive to state action (say, in voting or a fair trial), and when they are grounded in interests that are also sensitive to private action (say, life, liberty, equality, or expression). Whichever negative impact state inaction has on interests of the second type, it is dealt with in the political arena, in which the degree of state involvement in private relations is determined.

This model is deontological, ¹⁴ in the following sense: the prescriptions of rights focus on state actions that harm right-holders or that tie the state directly to private agents that harm right-holders. Balancing and trade-offs are rare: what the state owes to one right-holder is unaffected by what it owes to another. It is obligated not to harm or be intimately involved in harm to right-holders, but not to protect or promote their rights by intervening in their private relations with others or by creating certain desirable social conditions. ¹⁵

This makes sense given the political theory behind this veteran of constitutional horizontality models. This model, which describes the accepted reading of the Bill of Rights of the United States, ¹⁶ reflects classical liberal concerns about the state as a danger to the individual (and about the relations between central and regional units in federal systems). Constitutional rights serve the negative role of protecting the private sphere and the market from political intrusion. It is private law’s role to protect our basic interests and needs in these realms (a role that can receive constitutional protection—as in the famous Lochner case). ¹⁷

Criticizing the state action model seems almost futile, given the barrage of criticism already directed at it. ¹⁸ However, two basic points are important for our purposes, as they expose its libertarian moral myopia: the ways in which it ignores the relevance of consequentialist considerations for the practice of constitutional horizontality. This will be a first lesson out of several that will serve as the foundations of the proposed model.

Anthropomorphization

The first problem is that the conception of responsibility that underlies this model is more appropriate for private agents. This conception is “restrictive,” to use Samuel Scheffler’s term, in the sense that it limits our normative world as agents by distinguishing between what we do and what we fail to prevent, and between strangers and people with whom we have special relationships. ¹⁹ We owe to strangers mostly duties not to harm them in concrete ways (distinctions between acts and omissions, doing and allowing, or intending and foreseeing do much work here). To people with whom we have special relationships, we might owe complex duties that it is more difficult not to violate; but the ascription of duties follows an assumption of responsibility on our part. ²⁰ This reflects the priority that liberal interpersonal morality gives to our ability to control our duties to others at the expense of taking advantage of our ability to contribute to their well-being. ²¹

While there are justifications for some degree of similar state control over its responsibilities and duties to citizens, the state action model gives it nearly absolute discretion without weighing its benefits against its harms (in specific cases or over time)—a necessary justification, as the state has no autonomy or moral integrity of its own. This model takes constitutional interest in the regulation of private relations only if the state caused harm by being intimately involved in private activities (say, by subsidizing a university that racially discriminates students), or if the state assumed extended responsibilities (say, by imprisoning convicts together). For example, it was held in DeShaney that the right to equal protection does not obligate the state to protect children from their abusive custodial fathers even after several hospitalizations—as the state does not cause nor is intimately involved in such harms. ²²

This model reflects not only the common law’s hesitancy to impose positive duties, ²³ but also the libertarian tendency to apply forms of evaluation and reasoning that shape private interpersonal relations in public contexts. ²⁴ It ignores the modern state’s ability to predict and respond to harms and deprivations in society that makes aggregation and risk management basic aspects of its activity. ²⁵ Morally, public agents must not treat individuals unjustly but must also not ignore basic aspects of their well-being even if they adversely affect them indirectly or by omission. ²⁶ The state’s responsibility is not restrictive, backward-looking, and personal—it is prescriptive, forward-looking, and collective, in the sense that we care not only about what it did that ties it to privately-caused harms or deprivations but also about what it can do about them. ²⁷ This goes against the Hayekian idea that “nobody has the responsibility or the power to assure that these separate actions of many will produce a particular result for a certain person.” ²⁸ Sometimes the state has this responsibility, and sometimes it underlies constitutional rights—which gives them a society-facing and results-oriented orientation.

This more expansive conception of public responsibility is reflected in the normative structure of modern constitutional rights, which responds to factors that are external to the direct relations of the state and the citizen. ²⁹ To treat the complex values that underlie these rights (say, life, liberty, equality, privacy, or expression) as constitutionally sensitive only to harms by state action, is to ignore both the beneficiary-centered nature of these values (which are about basic individual interests and needs) and the qualitatively different nature of the state as an agent. When such values are placed at the foundations of constitutional rights, the political collective does not commit itself not to upset a socially-created status quo, but to legally create and sustain more justified social conditions: to bring about what John Rawls called ‘background justice’, against which private agents can normally interact without causing social injustice—directly or through the accumulated results of their actions. ³⁰ In the legal context, absent constitutional rights-based supervision, the public response to the unjust consequences that private law’s ongoing (agency-centered) operation brings about will be dealt with in the political domain. ³¹ This might not only result in worse outcomes in terms of social justice but also erode the legitimacy and justification of private law as part of the constitutional order.

Fetishization

This takes us to the second problem of the state action model. Even if and when state agents decide to ‘intervene’ in private relations and the door is opened for constitutional rights-based scrutiny, we end up asking the wrong questions: about the nature and extent of state involvement. For example, how much of the total funding of a university is provided by the state; how many of its employees are receiving state benefits; or whether any other elusive trail of normative breadcrumbs could lead us back to the state. ³²

This means that the nature and severity of harm to the right-holder is somewhat beside the point: it is important only if and to the extent that it points at impermissible state action. For example, if a landlord refuses to sell us an apartment or an employer refuses to hire us because of our religion or race, there is no constitutional concern about equality—as would be the case if the state showed even the slightest of preferences for or against individuals on the same grounds. ³³ The results will often be skewed, arbitrary, or just plain wrong from a moral perspective: for example, we might distinguish between permissible and impermissible discrimination of the customers of clubs or restaurants according to the ownership of the building in which they are located or of the parking they use. ³⁴ The nature of state involvement is important, but treating it as a categorical factor in all cases leads not only to arbitrary outcomes but also to the public facilitation of wrongdoing and injustice. ³⁵

This Hansel and Gretel type of constitutional scrutiny, again, both erodes the legitimacy of private law and makes the pursuit of social justice in the private sphere less effective. For example, it leaves the state leeway to drag its feet regarding private discriminatory practices that undermine the right to vote (as long as the ties to the state remain latent enough); ³⁶ or to deputize private agents to sue people involved in abortion procedures in an attempt to limit access to abortion without direct state action. ³⁷ If we care about voting or access to abortion enough to recognize them as constitutional rights, it seems that the manner in which they are harmed or restricted is of derivative and secondary importance.

By construing the scope of constitutional rights according to how the state treats us or what it does to us—as opposed how we are treated or how we do—the state action model makes questions of constitutional horizontality incoherent. The actions and activities of husbands, employers, landlords, journalists, or doctors, for example, are directly addressed only by private law, with its focus on personal responsibility and choice (more on that below). Any intervention in private law is seen as a political policy choice. Private law is seen as a domain of principle and rights that can be coherently and systematically explained and justified, while external interventions in it are seen as morally opaque and unprincipled rather than expressions of collective moral commitments that aim to realize social justice.

Summary

The state action model fails not because it is suitable only for some rights: those whose underlying values are particularly sensitive to state action—say, to vote or to a fair trial. Rather, it fails because it ignores the teleological normative structure of constitutional rights: the ways in which these rights are grounded in and draw their force from foundational values—and the sensitivity of some of these values to harm as a result of state inaction. This failure and the focus on state action rob this model of the normative tools necessary to determine which rights should apply only vertically and which rights should extend horizontally.

By ignoring the consequentialist aspects of the values that underlie rights—and by applying a private conception of responsibility to the state—this model severs the normative ties between the ongoing effort to respect, protect, and promote constitutional rights in private relations and practices and basic parts of political morality: the collective and systematic responsibilities and duties to ensure that basic aspects of well-being are realized to a sufficient extent, by legal means, in all social spheres. It is therefore not a coincidence that it was rejected by most liberal legal systems (and even in the United States it is under pressure—alleviated by the difficulty of changing its Constitution). ³⁸

3. Direct Application

If the state action model reflects a classical liberal conception of the constitution as a restraint on the state, the direct application model reflects a more progressive conception of the constitution that became popular in the past few decades and facilitated its expansion into the private sphere: the constitution is seen as an ideal of justice to be realized in society. This model sees constitutional rights not only as addressing both public and private actions but also as directly constraining and claimable against both public and private agents.

If the state action model has libertarian roots, the direct application model has utilitarian ones. ³⁹ The basic idea is that if the values that underlie constitutional rights (say, life, liberty, equality, privacy, or expression) are beneficiary-centered and agent-neutral—in the sense that the actions of any agent can affect them—it makes little sense to limit their entailment of duties to state agents. ⁴⁰ The delineation of duties is therefore determined by the ability of agents to bring about better states of affairs in terms of rights-realization in society.

In a sense, the choices of landlords, tenants, employers, employees, professors, students, sellers, or consumers, for example, are treated like choices of agents that are formally integrated in the state’s institutional apparatus: they must all constitutionally justify failures to realize rights if they can do so effectively. ⁴¹ Because some protected interests are more sensitive to state harm and because private agents are also right-holders, private duties will tend to be less cumbersome than public duties. But the important thing is that delineations of the scopes of duties are based on the potential impact of different actions or activities on rights rather than on the public or private nature of these actions or of the acting agents.

This model was adopted, to some extent, in Ireland and South Africa. ⁴² For example, the Supreme Court of Ireland held that an employer that required employees to sign contracts that obligate them to join a specific trade union violated their constitutional right to freedom of association. ⁴³ The Constitution of South Africa holds that “the Bill of Rights applies to all law, and binds the legislature, the executive, the judiciary and all organs of state” and all “natural or … juristic persons”—taking into account the nature of the right and the duties it imposes. ⁴⁴ For example, when evaluating the common law rule that placed the burden to justify the publication of potentially defamatory materials on media outlets, the Constitutional Court held that the right to freedom of expression has “direct horizontal application”; but there is also the right to human dignity—and the rule must strike a proportional balance between them (which it does by encouraging “editors and journalists to act with due care and respect for the individual interest in human dignity prior to publishing defamatory material.”) ⁴⁵

Problems do not necessarily arise for the direct model with regard to how demanding the constitutional duties of private agents will be: constitutional rights to autonomy, freedom, or liberty can prevent the imposition of disproportionate duties. Problems arise at the secondary level of figuring out which duties we have and the tertiary level of being called to account by right-holders for potential or actual violations. I will briefly present five problems in this regard that the constitutional requirement of proportionality cannot alleviate.

Complexity

The process of “human rights due diligence” includes an agent’s assessment of their direct and indirect impact on rights, taking mitigation steps, tracking their effectiveness, and remedying failures. ⁴⁶ While this kind of activity is more typical for corporations, the direct model expects individuals to engage in it as well (even if, usually, on a smaller scale). ⁴⁷ The problem is that private agents often lack the abilities and resources to figure out this impact and therefore their duties: to consider both how they can help bring about or contribute to complex social states of affairs (the reduction of poverty or promotion of workplace safety, for example) and how such benefits should be balanced against the costs, both for themselves and for others. As Rawls noted, it is often difficult for private agents to “comprehend the ramifications of their particular actions viewed collectively.” ⁴⁸

Intrusion

Even if private agents could effectively handle such demanding forms of evaluation and reasoning, allocating to them duties to do so because of their potential contribution to collective efforts to bring about desirable states of affairs—because of their abilities or resources rather than their choices, projects, or relations—undermines their autonomy and sense of authorship of their lives. ⁴⁹ Such complex duties “never sleep,” ⁵⁰ and their tentacular prescriptions reach many corners of our private lives: they are not limited to concrete actions that harm others or to cases in which we assume extended responsibilities to others. Private agents carry an ongoing burden to figure out where they stand, constitutionally—not because of certain things that they did but because of certain things that they can do.

Crowding out

Constitutionally enlisting private agents blurs the already vague boundaries between our lives and erodes our ability to see others as people with whom we interact directly and meaningfully in private relations. ⁵¹ Pursuing our shared public goals can crowd out the motivations, dispositions, attitudes, emotions, or evaluations that facilitate the forms of communication and attachment that constitute our private relations. ⁵² Take, for example, proportionality as a justification for rights infringements in private relations. ⁵³ The problem is that proportionality, in its constitutional form, operates on a social scale and addresses all potentially affected parties. To justify ourselves to private agents like that—for example, to violate our contract because doing so promotes equality in society—is to step back from our direct relations and to treat others as members of the group of beneficiaries of our actions. ⁵⁴ Such detachment would often harm our private relations and can even amount to wrongdoing (there are exceptions, but figuring out if they apply raises the previous problems). As Thomas Nagel notes, interpersonal justification has gone astray if it portrays one as a “benevolent bureaucrat” and is put in “administrative” terms, foreign to one’s direct relations with others. ⁵⁵

Scrutiny

Giving private agents the standing, authority, and discretion to call others to account as constitutional wrongdoers, in the name of the political collective, ⁵⁶ might give the practice of private rights-claiming some features of denunciation practices, under which public order is enforced through interpersonal scrutiny from the bottom up. Such heavy normative machinery is traditionally operated by public agents that are under public law constraints: this is one reason why criminal prosecutions are normally brought by state agents in the name of the political collective while tort claims are brought by private right-holders on their own behalf. ⁵⁷ But the more constraints we extend to private agents in this regard, the more we integrate them into the state’s apparatus, thereby only deepening the infiltration of the public into the private.

Authority

Burdening private agents with duties to discharge collective responsibilities involves recognizing their discretion (as we cannot specify in advance what exactly they must do to abide by their duties). ⁵⁸ If this responsibility delegation—a soft form of privatization—is not accompanied by strict administrative oversight, it raises concerns, especially given the difficulties involved in private litigation and its increasing rarity. ⁵⁹ Think, for example, about how the difficulty of calling Twitter or Facebook to account leaves them with de-facto authority to shape the market of ideas. ⁶⁰ Importantly, even if the results in terms of the realization of rights are positive, a problem remains: regulatory decisions about rights-realization should be made by agents accountable to the political collective not only through litigation but through more firm, persistent, and entrenched institutional mechanisms. ⁶¹

Summary

Versions of the problems of complexity, intrusion, crowding out, scrutiny, and authority figure in many normative debates: for example, about discrimination or privatization. ⁶² Their implications are often some distinctions between public and private. While such distinctions have their critics, they are accepted, in one form or another, in political, moral, and legal philosophy, and reflected in the architecture, principles, and norms of liberal legal systems. In our context, they justify a distinction between public and private agents as the bearers of constitutional rights-based duties.

If the state action model treats all agents as private while ignoring collective notions of responsibility, the direct model treats all agents as public while ignoring personal notions of responsibility: it ignores the ways in which the justification and value of privately-enforced rights and duties depend on their fit in the moral fabric of the relations in which they operate. The idea that the normative structure of rights-claiming and contesting imposes limits on the possible content of rights is crucial for understanding rights in general, and private legal rights in particular. ⁶³ While these limits are looser in state-citizen relations (the state action model goes wrong here), they are stricter when rights form part of private relations. It is not that some values are excluded from private relations but that the moral context limits the ways in which these values justify rights and duties: for example, it is easier to justify general duties to promote equality or protect privacy against the state than against private parties; and thus, it is no coincidence that they form parts of human rights law but take a more limited form in private law. The direct model fails to accommodate this distinction because the status of the duty-bearer is irrelevant: only the ability to realize constitutional rights matters.

It is therefore not surprising that the direct model was also rejected by most legal systems, and even its adoption in Ireland and South Africa met difficulties. ⁶⁴ Direct application can be justified in special circumstances of severe social or structural injustice, ⁶⁵ or when it comes to the operation of big business corporations in weak legal systems, whose authority is undermined. ⁶⁶ However, in most cases—more typical of modern, liberal, and only moderately unjust legal systems—the direct model will be more difficult to justify. Importantly, like the state action model, the direct model lacks the tools to distinguish between the cases: the distinctions it draws are normatively shallow and ad-hoc, given its focus on social states of affairs and the little attention it pays to relational moral considerations.

4. Strong Indirect Effect

Like state action models, indirect models see constitutional rights as imposing duties only on state agents; but like direct models, they see the content of these rights as addressing both public and private actions and activities. Interestingly, Rawls’s theory of political liberalism follows a similar path. ⁶⁷ Contra libertarianism, social justice imposes duties on public agents to ensure that the basic structure of society (crudely: the main political, social, and economic institutions) meets certain requirements; but contra utilitarianism, the principles of social justice apply directly to the basic structure and only indirectly to private agents—who are burdened with a different set of simpler duties. ⁶⁸ We therefore see complex and systematic reasons as addressing the rules, institutions, or practices that guide private agents rather than these agents directly. This is a common response to criticisms of consequentialism: we see it as a standard of large-scale evaluation rather than a method of individual reasoning. ⁶⁹

Under indirect models, private agents rely on state agents to engage large-scale and results-oriented reasons about how the simpler set of private duties should operate. ⁷⁰ This has a cost in terms of individual agency and our responsibility to reason about what we owe to others, but benefits in terms of the ability to bring more complex values to bear on private relations. ⁷¹ The state bears the direct burdens of respecting, protecting, and promoting constitutional rights, in all their complexity; and it allocates to private agents private law duties to contribute to this collective endeavour, while taking into account the problems of complexity, intrusion, crowding out, scrutiny, and authority. Private agents cannot claim constitutional rights against each other: they can only address the state with claims that it must amend the content of private law. ⁷²

The indirect model originates in the Lüth case, in which the German Constitutional Court held that the right to freedom of expression was violated by lower court orders against calls to boycott a film, which did not give the constitutional right due weight when applying private law. ⁷³ This model spread to other legal systems. In its strong version, reflected, for example, in European human rights law (largely following German constitutional principles), all state institutions are burdened with positive duties to accommodate constitutional rights in private law. ⁷⁴

This creates a complex relational structure. For example, in Evans—a dispute between former romantic partners about the destruction of embryos kept in a fertility clinic—the court identified both contractual relations between the partners and the clinic and constitutional relations between each partner and the state, which must respect conflicting aspects of the right to privacy and family life (to become or not to become a parent) and public interests. ⁷⁵ This relational structure is an important step forward; but the strong indirect effect model fails to accompany it with sufficient normative distance between the small-scale morality that shapes private relations and the large-scale morality that shapes public regulatory activities.

Public Reasoning: Distortion

The first problem is that relational moral considerations are distorted when they are translated to consequentialist considerations. Three factors pull this model towards consequentialist terrain. First, constitutional rights have beneficiary-centered and large-scale aspects: from the immediate suspects (say, rights to equality or autonomy) to more innocuous rights (say, rights against torture). These rights require certain social conditions or institutions to exist and certain duties to be allocated to different agents (say, prosecutors, police officers, and citizens). Second, because the constitution has superior legal status, its partial realization must be constitutionally justified. Third, this justification must demonstrate a form of balance or proportion, and as these rights address social states of affairs, any partial realization in ordinary law (for example, limiting certain forms of expression in libel law) must be justified by showing that it brings about better social states of affairs.

The problem in our context is not the lack of a common metric or evaluative standard: we can compare social states of affairs by making qualitative distinctions between them. ⁷⁶ It is also not that constitutional rights become the only relevant reasons for regulation: ⁷⁷ these rights are abstract and agnostic about many concrete legal arrangements, and can be limited to serve values external to them. ⁷⁸ The problem is that all reasons for regulation (except illegitimate reasons, like coercively imposing religious beliefs) are placed on a single continuum that relates to the goodness of social states of affairs. Thus, relational reasons for regulation must fit in a large-scale constitutional-consequentialist picture. But normative reasoning is path-dependent and compartmentalized, in the sense that focusing on agency leads to certain types of conclusions while a focus on beneficence leads to others. ⁷⁹ So, if our reasoning starts in the “consequentialist circle,” to borrow Paul Hurley’s term, we are drawn to engage responsibilities and duties as responses to states of affairs. ⁸⁰

This is a problem because many of the important moral considerations that underlie and shape the regulation of private relations are not about large-scale states of affairs but about what we owe to each other as private agents in relative isolation from political factors. It can be said, without going too deeply into the debate about the limits of consequentialism in terms of the accommodation of relational reasons, ⁸¹ that consequentialism is not a neutral form of reasoning: that redescribing relational reasons for regulation in consequentialist terms comes at the cost that valuable moral data is lost. ⁸² This critique is often made of economic theories of private law, which fail to capture the relational normativity of rights and duties. ⁸³ For example, even if contract law contributes, as a whole, to social welfare, the normativity of particular transactions relies on the direct relations of the parties—say, on the exercise of a normative power to assume an obligation or the invitation to rely on a promise. ⁸⁴

This raises, first, a problem of effectiveness: for example, we will fail to create the best practice of contractual agreement—and this includes justifiably realizing constitutional rights through it—if we ignore the ways in which expectations, reliance, and trust operate interpersonally. Second, there is a problem of justification. While the constitution sits at the apex of the legal system, it is not the only source of standards for its moral evaluation: there are standards that are external to the constitution—for example, and importantly for us, the relational morality of what we owe to each other. ⁸⁵ We will fail to justifiably promote equality through the regulation of private relations, for example, if we focus only on which private agents can most effectively do so, while ignoring the complex and nuanced ways in which standing, authority, responsibility, reactive attitudes, and other moral factors interact and connect together to form morally-justified private relations. ⁸⁶ Therefore, for constitutional horizontality to be effective and justified, it must work with a holistic and rich account of the normativity of private relations, which determines much of the justification and value of private rights. This cannot be done by focusing only on social states of affairs.

Private Reasoning: Contamination

Even if public agents could effectively transition from the consequentialist circle to the relational circle, doing so is likely—given private law’s normative, discursive, and institutional features—to erode or distort relational morality for private agents. The complex and tentacular prescriptions of abstract constitutional rights will gradually overshadow the structural features of relational morality—and therefore the problems of complexity, intrusion, crowding out, scrutiny, and authority will resurface.

To begin with, the processes in which public agents shape private law are relatively transparent: while a certain degree of opacity is desirable and inevitable, it cannot and must not be absolute—especially given public involvement in administrative and legislative processes and the partial fusion of law-creation and law-application in common law decisions. ⁸⁷ Given the ongoing evolution of privately-enforced legal rights through adjudication, we cannot separate them from the legal reasoning that supports them, charts a course for their future application, and ties them to neighbouring legal and moral norms and principles. ⁸⁸

A second related factor is that following legal norms is not a series of isolated technical decisions: to do so adequately, while also following norms from other sources (morality, religion, relationships, practices, and so on), we internalize legal norms, reflect on them when interacting with others, make evaluative judgements about their requirements, and self-monitor and modify our behaviour accordingly. ⁸⁹ When doing so, we reflect not only on the linguistic content of legal norms but also on their underlying public justifications. ⁹⁰ Even if this moral engagement with legal norms is less important in some areas of law (say, tax law), it is necessary for private law to possess moral authority and legitimacy: when private law recognizes rights, we want it to justifiably grant private agents the standing and authority to claim them, allocate responsibilities to duty-bearers, and invite reactive attitudes for violations. If private law is justified to a sufficient extent, and if it works effectively in the sense that it is embedded in the normative fabric of the relations it regulates, these legal actions have moral echoes that feed back into private law by shaping the ways in which private agents claim rights, abide by duties, and make other legal choices. ⁹¹

If state reasoning about private law focuses on how private agents can be conscripted to bring about better large-scale states of affairs, the problems of complexity, intrusion, crowding out, scrutiny, and authority re-emerge: not only when private agents devise strategies for litigation or receive legal advice, but more generally when they consider how to conduct themselves in their legal relations with other private agents.

On the one hand, this harms private law’s ability to respond to relational morality and the values that inhere in private relations: for example, subordinating contract law to the promotion of public policy can erode the normative robustness of duties to keep promises, both in and out of law; ⁹² and, similarly, portraying tort law as being all about public policy rather than duties to others flattens its normativity, erodes the individual motivation to comply with it, and makes liability look like a mere price for interacting with others. ⁹³ On the other hand, this harms private law’s ability to realize constitutional rights: for example, attempts to extend the wrong of discrimination to inadvertent acts that discriminate indirectly, failed to create relational normative structures of rights, duties, responsibility, standing, authority, and reactive attitudes—thereby becoming less effective. ⁹⁴ Such attempts to ground rights and duties in consequentialist considerations can also backfire by painting duties owed to others as politically grounded and therefore as owed to the political collective, which will weaken neighbouring duties: say, duties not to discriminate directly. ⁹⁵

Summary

The problem with the strong indirect model is not that it ends up collapsing to a different version of the direct model: ⁹⁶ not being accountable to each other for our constitutional transgressions and directing the constitution’s prescriptions through the state makes a great difference—both on the secondary level of figuring out which duties we have and the tertiary level of being called to account by right-holders for potential or actual violations. The problem is also not that this model provides too little regulatory guidance, by focusing on state activities and actions while ignoring private activities and actions: ⁹⁷ it does address private activities and actions, even if as parts of large-scale patterns and states of affairs.

The problem is that this model is too close to the direct model, as a result of its failure to appreciate the incompatibility and mutual resistance of consequentialist and relational reasons for regulation. It is not enough to allow relational reasons to shape the scope and content of constitutional rights or justify their proportional limitations if to do so they must address large-scale patterns of private actions and activities. In this sense, this model is not too abstract but too specific and leaves too little room for relationality. As Nagel noted, relational reasons often require “a less indirect, nonstatistical form of evaluation.” ⁹⁸ That is, they must be protected to a certain degree from consequentialization in public reasoning about the regulation of private relations. The strong indirect model therefore seems to be more at home in legal systems with constitutional courts that do not create or change private law norms. In common law legal systems, its ongoing impact over private law raises moral concerns.

5. Partitioned Indirect Effect

This indirect model offers more protection and a degree of isolation to relational morality. The state is obligated to protect or promote constitutional rights through private law, but its duties have softer edges and are more flexible: non-judicial agents are under less comprehensive positive duties, while judicial agents are constrained by values rather than rights, which means that they can normally operate within private law’s endemic structure and doctrines. This allocation of duties and authorities reflects reasons of democratic legitimacy, separation of powers, and the rule of law; ⁹⁹ but it can also accommodate moral pluralism by institutionalizing different regulatory perspectives and forms of reasoning. ¹⁰⁰

‘Upstream’ constitutional questions—including which private law regimes to adopt—are answered by non-judicial agents, who normally adopt a consequentialist perspective. But the ‘downstream’ development and application of these regimes is the responsibility of judicial agents, who rely mostly on small-scale and relational considerations, and concepts like standing, causation, foreseeability, reasonableness, or consent. ¹⁰¹ For example, constitutional rights can shape the choice between protecting employee welfare by recognizing private law rights and duties or by creating a workplace-supervising agency; but if we take the former path, constitutional rights are marginalized in the judicial decisions that develop these rights and duties through their application in concrete cases. ¹⁰²

Constitutional rights do not constrain judges in such decisions as rights. This idea originates in Lüth, mentioned above: rights operate as values whose “radiating effect” shapes the creation and application of private law. ¹⁰³ Failures to accommodate these values can harm private litigants but only as failures to apply the law fairly or adequately or to promote legal coherence and integrity in the legal system in general: that is, constitutional rights constrain judges only indirectly, as imperfect duties not owed to any right-holder in particular. ¹⁰⁴ This gives judges the flexibility to operate and reason within the doctrinal and conceptual structures of private law as it is, which rely on, express, and develop relational moral reasons for regulation. ¹⁰⁵ Constitutional rights can therefore be realized through private law but either by non-judicial large-scale decisions or when the door for their judicial realization is opened for them from within private law: when existing rules, doctrines, or principles, tie private actions, choices, or relations to these rights, as values.

A version of this model was adopted in the United Kingdom: while the Human Rights Act obligates all state agents to act compatibly with the rights listed in the European Convention on Human Rights, judges enjoy broad discretion to develop the common law incrementally while leaving extensive reform to the legislature that is under no duty to legislate. ¹⁰⁶ The adoption of this model is more explicit in Canada: its Supreme Court held that Charter rights bind legislators and administrators but not judges when they decide private law cases on the basis of the common law. ¹⁰⁷ In this context, judges are not institutionalized state agents but merely “neutral arbiters.” ¹⁰⁸ As a result: “The most that the private litigant can do is argue that the common law is inconsistent with Charter values,” which merely “provide the guidelines for any modification to the common law which the court feels is necessary.” ¹⁰⁹

Judges engage in full constitutional rights-based scrutiny of private law only to evaluate the choices of non-judicial state agents. For example, in the United Kingdom, it was held that the heterosexual definition of a spouse in the Rent Act offers inadequate protection to the right to private life and home of the homosexual spouse of a deceased statutory tenant, as it fails to protect them against eviction by a private landlord. ¹¹⁰ In Canada, it was held that the repeal of an act that facilitated collective bargaining between farm workers and their employers unfairly left this group of employees unprotected by labour law. ¹¹¹ In both cases, it was the constitutional goal of ensuring equal legal protection that did most of the work: had there been no laws at all, the judicial interventions would have been more circumscribed.

When developing the common law, judges engage constitutional rights as abstract values, in an open-ended process that leaves room for relational reasons for regulation and works from within private law’s conceptual and doctrinal structure. For example, in response to claims by a model whose pictures leaving a rehabilitation centre were displayed in a tabloid without her consent, the British House of Lords expanded the tort of breach of confidence to circumstances in which there was no pre-existing relationship between the parties rather than create a new tort of breach of privacy. ¹¹² In a similar vein, when the Canadian Supreme Court evaluated the protection of the value of freedom of expression in defamation law, it engaged a few consequentialist reasons (like the costs of trials and the state of public discourse), but had the normative leeway to focus on the impact on the defamed person and the nature of the defaming agent’s actions (their knowledge, diligence, intentions, and so on). ¹¹³

This feature of the model reflects one of the basic philosophical responses to moral pluralism: contextualism—the idea that conflicts of abstract values should be resolved while appealing to the factual and normative features of specific cases rather than in general. ¹¹⁴ The system of constitutional rights does not determine the result of common law disputes—and their resolution does not alter the system. ¹¹⁵ In a sense, judges merely locate the private counterparts of constitutional rights: private duties are grounded in the personal responsibility of private agents towards others rather than in the constitution. ¹¹⁶ For example, if employers assume authority over employees, they must respect and protect employee privacy at the workplace: the nature of employer-employee legal relations makes certain aspects of the constitutional right to privacy morally salient in this context, and therefore less disruptive of their relational normativity. This solution reduces the normative complexity of constitutional horizontality. ¹¹⁷ It also alleviates the problems of distortion and contamination that the strong indirect model creates. Accommodating constitutional rights in private law in such cases does not require state agents to consequentialize relational factors: rather, they limit their deliberation to areas in which consequentialist and relational factors happen to overlap, and to cases in which constitutional values could be smoothly integrated in private legal relations.

This model offers important insights from the perspective of moral pluralism: it protects the natural regulatory habitats of consequentialist and relational considerations while reducing their frictions and conflicts. However, as I add now, this model ignores important reasons for regulation and erects rigid barriers between conflicting reasons and the agents engaging them. Outside of court, it looks like the strong indirect model; in court, it looks like the state action model; and the outside-inside interface is too fragmented.

Upstream: Too Strong

The principle of legislative supremacy is fundamental in common law legal systems. Under the partitioned model, this principle leaves relational considerations vulnerable to being ignored, eroded, or distorted by consequentialist reasoning. While legislative agents are not under positive duties to legislate (as they are under the strong indirect model), when they choose to do so, their actions are heavily influenced by constitutional rights with consequentialist aspects: for example, they must equally protect similarly situated right-holders. Thus, while relational considerations can be shielded in common law decisions, in legislative decisions they fare pretty much as they do under the strong indirect model: as further reasons for regulation to be evaluated through a consequentialist policy prism.

This results from the fact that the response to moral pluralism under this model is a mere by-product of institutional concerns rather than a result of an appreciation of the unique normativity of private legal relations. No weight is given to the sensitivity of different private relations to public (including constitutional) exposure. Rather, the allocation of responsibilities in the practice of constitutional horizontality is made at a more systemic level, while responding (quite rigidly) to reasons grounded in the principle of separation of powers.

Downstream: Too Weak

This model creates what Jud Mathews called “strategic ambiguity”: it gives state agents (especially judges) broad discretion about how to accommodate constitutional rights in private law. ¹¹⁸ But the model seems too ambiguous in this regard: non-judicial agents are under no stringent duties to constitutionally scrutinize private law—and when they do, it is often from a consequentialist perspective; and judges must only consider the general guidelines provided by constitutional values. As importantly, the systematization of these regulatory perspectives, over time, is at risk of being partial and weak. That is, contextualism could easily end up as a form of Hayekian conservatism that lets private law take its own course while subordinating it to external organizing forces only sporadically. ¹¹⁹ The model fails to secure strong enough roles for relational reasons in non-judicial settings and consequentialist reasons in judicial settings, and to coherently integrate the regulatory outputs of different state agents.

This leads to an even bigger problem: judges are not mere neutral arbiters but state agents who also make law in the name of the political collective. Even if they should enjoy broad constitutional discretion that allows them to operate within the conceptual or doctrinal structures of private law, common law decisions must not be fully immune to constitutional rights-based scrutiny. One of the basic justifications for constitutional horizontality is that such scrutiny is necessary: it sheds light on beneficiary-centered interests and needs that private law, with its traditional focus on agency and choice, tends to neglect. ¹²⁰ The relational focus of private law could easily distract us from systematic or structural causes of social injustice and constitutional rights violations. ¹²¹ The partitioned model fails to respond to this myopia by erecting too rigid a barrier between the constitution and the common law.

Summary

While the incrementalism and contextualism of the partitioned model are justified in some circumstances, the model lacks the tools to embed these features in a system that makes the best use of them while minimizing their negative implications. In the next section, I offer a composite indirect effect model that meets these requirements. This model does not arbitrarily transition between indiscriminate legislative interventions and conservative judicial evolution. It sees public and private agents as involved in an ongoing normative dialogue about constitutional horizontality on two normative levels: about legal reform and about when and how to engage in it. This composite structure alleviates the moral concerns described above.

6. Composite Indirect Effect

While first-order reasons are about which choice to make, second-order reasons are about how and when to engage the first-order choice: second-order reasons do not outweigh first-order reasons on the merits but address their place in the choice situation. ¹²² Institutional competence, democratic legitimacy, and legal stability often operate in law as second-order reasons about when and how to engage first-order reasons for regulation. The problem with the partitioned model is that because it gives institutional second-order reasons too much weight while ignoring substantive second-order reasons, it ignores many ways in which relational and consequentialist reasons interact in non-judicial and judicial settings.

Phillipson and Williams place second-order institutional reasons at the centre of their interpretation of the British indirect model. ¹²³ However, while institutional reasons provide some protection for relational values, they are not enough to fully accommodate their unique normativity and roles as second-order reasons in the practice of constitutional horizontality, nor are they enough to ensure that judges discharge their constitutional rights-based duties (rather than just follow guidelines). The composite indirect effect model sees second-order reasons and an incremental approach to rights realization as central in the explanation and justification of the practice of constitutional horizontality.

An example can help us tease out the role of such reasons in the transition between constitutional restraint and intervention. In McKinney, the Canadian Supreme Court considered whether the constitutional right to equality limits the retirement policy of a private university. ¹²⁴ To preserve academic freedom, the Court held that despite the fact that the university performs a public function and is partly funded and regulated by the state, its public ties are not strong enough to justify applying the Charter to its activities. Because the state’s lack of control of the board of directors was a central factor, the Charter was applied in other cases to similar institutions, like community colleges, when such control existed (we see here, again, the problematic implications of the state action model). ¹²⁵

Constitutional Restraint

Another way to engage such cases is to hold that people have rights that the state prevent discrimination regardless of the status of the discriminating agent or its public ties; but when the state considers which regulatory means justifiably realize this right, it must consider the nature of the agent and its activities (and its ties to the state). The state can utilize the flexibility of rights-realization requirements (like proportionality tests) and its discretion to decide when and how to engage in legal reform: ¹²⁶ rather than constitutionally evaluate the university’s policy (direct application) or the legal rules that regulate its activities (strong indirect effect), it can focus on the second-order reasons to pull or not to pull these deliberative threads. Therefore, for example, the value of academic freedom and the normativity of academic activities can justify constitutional restraint, while leaving extensive constitutional scrutiny for another time and place. This restraint is justified not only institutionally: it forms a substantive part of a system in which restraint is balanced against other second-order reasons for reform—reasons about its urgency, complexity, and comprehensiveness.

A good example of this normative structure is Philip Pettit’s account of consequentialism in friendship-related decisions. ¹²⁷ Resorting to reasons instrumental from the perspective of particular friendships in certain times or ways can make us worse friends or even no friends at all. If a friend asks us for help, we should not weigh the costs and benefits of doing so, of being friends with them, or how friendship contributes to our well-being in general. Normally, we should follow certain predispositions out of a sense of commitment rather than try to do most good. Similarly, there are second-order reasons to isolate private relations from constitutional scrutiny, to an extent, because the type, frequency, and intensity of such scrutiny can distort or erode their endemic normativity or underlying values. Because constitutional reasoning necessarily shapes private legal reasoning, directly or indirectly, the former must exhibit a degree of restraint. This means that all state agents—not only judges—have reasons to operate contextually and incrementally in this regard: to focus on relatively isolated aspects of constitutional horizontality problems in order to prevent a flood of constitutional rights-based reasons from drowning private relations.

For example, if we want to avoid subordinating employers to a general duty to protect privacy, we can focus—when evaluating workplace surveillance—on types or degrees, reasons to engage in it, available measures, or employee knowledge and consent. It might therefore be impermissible to regularly read employee emails, ¹²⁸ but permissible to install video-surveillance in response to suspicions of theft. ¹²⁹ By making qualitative distinctions between different aspects of constitutional rights, we can isolate the aspects implicated in limited contexts, thereby avoiding full consequentialization of public reasoning. ¹³⁰

Unlike in the partitioned model, contextualism is teleologically grounded here in substantive reasons about how the normativity of private relations and their underlying values fit in a practice of constitutional horizontality that can best realize moral, constitutional, and other rights and values. Two features create the space for relational second-order reasons in this teleological scheme. First, we ascribe intrinsic value to agential involvements, commitments, and predispositions that form parts of justified private relations: for example, promising or taking adequate care not to harm others negligently. States of affairs can therefore be worse because of how they were brought about. ¹³¹ This is still a teleological framework, but what we see as best (that is, most good) is more nuanced. There might therefore be reasons not to maximize good conduct or minimize bad conduct: for example, the practice of contractual agreements will not be better if we induce private agents to make more promises that they could keep or obligate them to ensure that third parties keep their promises. ¹³²

Second, we do not evaluate isolated decisions or actions but the practice of constitutional horizontality as a whole: the ways in which many agents, institutions, and practices operate together over time. The justified realization of constitutional rights addresses not just the isolated impact of legal rules but the likely consequences of the system’s ongoing operation. There is a similarity here to Rawls’s version of distributive justice, which does not evaluate how given bundles of goods are allocated among individuals but the likely consequences of the ongoing operation of the basic structure of society. ¹³³ This does not mean that we are indifferent to particular decisions but that we focus on their impact as part of a complex system and recognize that some agential discretion is necessary to realize the system’s goals and values. This leaves room for systematically justified moments of constitutional restraint: for some parts of the system to be isolated, to some extent, from consequentialist scrutiny. To go back to the example mentioned above, this seems to be the case with universities: ongoing constitutional-consequentialist scrutiny seems at odds with the degree of isolation justified in this social sphere.

Constitutional Ascent

The justification of constitutional restraint hinges on the possibility of constitutional ascent. I borrow this term from Ronald Dworkin, who claimed that while judges should normally resolve cases by giving “local priority” to reasons closer to the factual disputes before them, “justificatory ascent” (appealing to more remote and abstract reasons) is always “on the cards.” ¹³⁴ This reflects Dworkin’s ties to the common law tradition, which includes not only incrementalism and continuity but also responsiveness and justification. ¹³⁵

This can be seen, again, in Pettit’s discussion of friendship-related decisions. Even in this relational context, we must sometimes shift gears to a more comprehensive and systematic form of reasoning that goes against our regular predispositions and commitments: first, when “red lights go on” and “signals of alarm sound,” in the sense that the bad, wrongful, or unjust consequences of our actions are serious and visible (for example, if a friend asks for help with moving not their apartment, but a dead body); and second, we must occasionally reflect on our predispositions, commitments, and behavioural patterns from a broader perspective to ensure that they meet our deep values, principles, and goals. ¹³⁶

These exceptional cases have parallels in the practice of constitutional horizontality: state agents must ensure that private law does not fall below certain levels of rights realization, and that it is progressing towards better realization justifiably and effectively. This reflects the idea that when constitutional rights are excluded from or marginalized in decisions in which restraint is shown, they do not disappear from the normative universe: while lacking weight in a particular context, such reasons keep operating “behind the scenes.” ¹³⁷

These first-order reasons spring back into action in decisions in which second-order reasons justify constitutional ascent. As Alison Young claimed, constitutional horizontality changes judicial reasoning and often brings constitutional rights from the background to the foreground. ¹³⁸ Now, some instances of the first exception—grave and visible injustice—can be accommodated by the judicial part of the partitioned model: that is, by reasoning from within private law. But there can also be instances in which the injustice is weakly tied to private law doctrines or to notions of personal responsibility and choice that allow the evolution of private law ‘from within’: when external considerations must bend private law’s trajectory more forcefully. This seems to have been the case, for example, in the Campbell case mentioned above: the court stretched the meaning of “breach of confidence” to respond to invasions of privacy among strangers; ¹³⁹ but it could have been more conceptually accurate and morally justified and effective to recognize a tort of breach of privacy.

The need in ascent is also evident in the second exceptional type of case: the occasional systematic recalibration of private law to ensure that it is going in the right direction. For example, in Grant v Torstar Corp, the Supreme Court of Canada changed the balance between privacy, reputation, and freedom of expression in defamation law by recognizing a new defence of responsible communication in matters of public interest, by engaging in open-ended proportionality analysis. ¹⁴⁰ Such cases reflect the dynamic nature of constitutional horizontality, and the normative residue left by rights-based reasons that are not fully realized because they are excluded, marginalized, or outweighed. Appreciating this inherent state of moral conflict and its tragic residues brings into focus not only corrective or remedial duties but also duties to do our best to minimize or alleviate future conflicts: to change the social conditions in which the conflict is created. ¹⁴¹ The constitution’s normative superiority can therefore manifest itself not in how we resolve particular conflicts but in the ways in which we shape the practice of constitutional horizontality and the ongoing guidance that the rights enshrined in the constitution provide for the evolution of private law: for example, in our ability to occasionally reshape defamation law to better realize constitutional rights.

This complexity also gestures towards the idea that we should not leave the systematic moments of constitutional horizontality only to non-judicial state agents. The legislative ability to express collective moral commitments and the expertise of administrative agencies and tribunals are of great value for the practice of constitutional horizontality. Importantly, judges have a unique contribution to make in this regard, which requires them to interact with non-judicial state agents—also about complex societal problems. As Seana Shiffrin claims, judges have the ability to promote legal coherence by applying a foundational conceptual framework across different cases, thus expressing local understandings of private moral relations. ¹⁴² Judges can therefore engage complex consequentialist reasons for regulation while being attuned to the normativity of private relations and their underlying values. This allows judges to contribute to the incredibly important process in which the constitution is slowly integrated into the normative DNA of private relations and practices: from the creation of legitimate expectations and the recognition of standing to claim or to blame, to holding agents accountable and granting remedies. This means that second-order institutional reasons for constitutional restraint should not be assumed to have absolute priority: the teleological orientation of the practice of constitutional horizontality entails a degree of instrumentalism and flexibility in institutional divisions of labour.

What we see here is a complex legal practice in which both public and private agents perform different deliberative roles by responding to first-order and second-order reasons for the regulation of private relations—with large-scale reasons regularly operating behind the scenes and springing into action in special cases, to ensure that private law does not fall below certain thresholds and adequately moves towards better rights-realization. This legal structure is not foreign to the common law and describes the evolution of many of its areas (think, for example, about tort law: from the rise of duties not to negligently harm strangers; through the recognition of strict liabilities and duties not to cause pure economic or emotional harms; to a modern pluralistic form, which plays an important part in the state’s regulatory efforts).

As importantly, the composite model can also accommodate the principles underlying the other models. For example, it can say when a stronger indirect effect or even forms of direct application are justified (for example, regarding large profit-oriented corporations) and when weaker forms or even a focus on state action are justified (for example, when it comes to academic or religious institutions and practices). The composite model—because of its teleological structure, which works from moral values to the best legal means to realize them—can therefore effectively and justifiably shape the scope and structure of the practice of constitutional horizontality, while responding to the basic features and limits of the other models: limits that they tend to ignore or fail to take seriously.

7. The Road Ahead

Constitutional horizontality developed, to a great extent, as a response to moral concerns and conflicts that underlie all modern liberal legal systems: between rights-based duties to respect, protect, or promote basic aspects of well-being in all social spheres, and the preservation of a certain distance between public policies and goals and private relations. The progress through the models presented above teaches us basic lessons about this normative effort: it must engage both consequentialist aspects of political morality and relational aspects of interpersonal morality; do so on their own terms while accepting their incompatibility; recognize first-order and second-order reasons for regulation; and see the progress towards rights-realization as a complex process in which both public and private agents have important roles to play.

These lessons underlie the composite version of the indirect effect model, which reflects to a great extent ideas central to the common law tradition. As Nagel noted, moral progress can be made if instead of aiming at full systematicity we will adopt more flexible approaches that progress incrementally while facilitating the imperfect resolutions of tragic conflicts. ¹⁴³ I have tried to show that this prescription is suitable for constitutional horizontality as well: we must accept a certain degree of normative myopia when evaluating private law from a constitutional perspective, while constantly keeping in mind the need to broaden our normative gaze in a timely manner; we must ensure that private law is pushed as far as it can be to accommodate constitutional rights but do so while protecting the unique normativity of private relations and practices and their underlying values.

Now, more theoretical work must be done about the composite indirect effect model. However, one of its central benefits—which reflects its common law orientation—is that it not only accepts but invites the gradual development of the practice of constitutional horizontality. Given the novelty of this practice, it is quite natural that we are still dealing with big and abstract questions. But in light of the trend towards indirect models, we are now at a stage in which more concrete questions can be dealt with, and the answers to them can start taking the form of relatively stable principles and rules that guide this practice. In justifying the composite model and criticizing the others, I tried to take a few steps in that direction, while being attentive to the values and forms of evaluation and reasoning that both legal domains utilize and to their interactions against the background condition of moral pluralism.