Strong popular sovereignty and constitutional legitimacy

Section 1

The work of Carl Schmitt casts a heavy shadow over contemporary attempts to ground constitutional legitimacy in the constituent power of a strong popular sovereign (Böckenförde, 2017; Grimm, 2015; Kahn, 2011; Kalyvas, 2008; Loughlin, 2013). ¹ As Lars Vinx suggests, perhaps the main reason for this continued influence is that Schmitt’s account of constituent power – despite Schmitt’s own political misadventures – seems to provide the resources for a democratic theory of constitutional legitimacy (2013: 103). In light of this influence it is instructive to return to the source and examine Schmitt’s concepts of strong popular sovereignty and constituent power in order to assess whether they are capable of supporting a plausible normative theory of constitutional legitimacy. In this section, I set out the main outlines of Schmitt’s theory and demonstrate that it sustains a normative interpretation. As will become clear in the following two sections, however, the fact that Schmitt’s theory sustains a normative interpretation does not entail that it provides the resources for a cogent account of constitutional legitimacy.

In the introduction to Constitutional Theory (1928), Schmitt (2008, 2010) emphasises that his intention is to elaborate a system and develop a theory of the constitution. Consistent with this descriptive intention, the early sections of Constitutional Theory are structured around an analytic distinction between four concepts of the constitution: absolute, relative, positive and ideal. Constitution in the absolute sense refers to the constitution as a unified whole in contrast to a mere aggregate of specially constituted statutes (Schmitt, 2008: 59). This absolute concept of the constitution can itself be understood in two senses as either a concrete condition of unity and order or as the ideal unity of a closed system of norms (Schmitt, 2008: 59–66). The relative sense of the constitution refers to constitutional laws in a formalistic sense that abstracts from the status of some constitutional provisions as more fundamental in the determination of the political identity of a community (Schmitt, 2008: 67). Such a formalistic concept of the constitution culminates in the liberal bourgeois treatment of the constitution as a statute or mere aggregate of statutes. Constitution in the positive sense refers to the constitution as a complete decision over the type and form of the political unity (Schmitt, 2008: 75). The constitution in this positive sense accordingly originates from a concrete existential act of will by the constitution-making or constituent power (Schmitt, 2008: 75). Finally, constitution in the ideal sense refers to a constitution that is exemplary in containing a certain content (Schmitt, 2008: 89). An example of the constitution in the ideal sense is the bourgeois constitutional form developed in the 19th century, which privileges individual freedom by protecting individual rights and liberties and establishing a division of powers (Schmitt, 2008: 89–91).

Although it is presented as a descriptive and analytic theory of the different senses of the constitution, Schmitt’s fourfold distinction contains clear normative content. An obvious example of the normative import of Schmitt’s fourfold distinction is the association of the relative sense of the constitution with a type of legal formalism which reflects the political interests of the bourgeoisie (Schmitt, 2008: 89–93). More relevant for present purposes, however, is Schmitt’s privileging of the positive sense of the constitution and his identification of that sense with the exercise of strong popular sovereignty in acts of constituent power by the people or nation considered as a concrete unity.

The constitution-making power is defined by Schmitt (2008: 125) as ‘the political will, whose power or authority (Macht oder Autorität) is capable of making the concrete, comprehensive decision (Gesamtentscheidung) over the type and form of its political existence’. ² For Schmitt (2008: 75), the constitution-making power determines ‘the entirety of the political unity in regard to its particular form of existence through a single instance of a decision’. It is crucial to note the way in which the concept of political unity operates here at multiple levels. An act of constitution-making presupposes a prior political unity such as the people or a nation. ³ Drawing on Emmanuel-Joseph Sieyès, Schmitt (2008: 128–132) insists that the constitution-making power, which is always in the state of nature and hence inalienable, is an unmediated will that exists prior to and above every constitutional procedure. For Schmitt the state must thus be factually given as a unity prior to the establishment of the sorts of constitutional guarantees associated with the rule of law (Böckenförde, 1998). As long as a unified people have the will to continued existence, they are regarded as superior to every constitutional form and normative framework (Schmitt, 2008: 131). It is only through the concrete act and decision by an already present people or a nation for a particular constitutional form, however, that the prior political unity can be actualised on the second level as a constitutional order in the absolute sense of a whole or unity irreducible to particular constitutional provisions or statutes. As a consequence, Schmitt’s prioritisation of the positive sense of the constitution implies that the absolute sense of the constitution as a unified whole is ultimately answerable, on both descriptive and normative levels, to the pre-constitutional concrete political unity that bears the constitution-making power.

This brief sketch of the positive sense of the constitution might already raise concerns as to the normative credentials of Schmitt’s constitutional theory. Schmitt’s strong emphasis upon the existential priority of the popular sovereign and his privileging of the positive sense of the constitution as a concrete decision made by a unified people for a particular constitutional form both seem unpromising sources for the development of a thick normative conception of constitutional legitimacy. Indeed, Schmitt’s arguments suggest a voluntaristic prioritisation of the concrete and existential over the normative and a celebration of ‘the facticity of a power that overcomes every other power’ (Habermas, 1992: 130–131). This point certainly deserves closer scrutiny, for if it is simply misguided to attribute a normative conception of legitimacy to Schmitt in the first place, then it could be argued that contemporary critics of his position are setting out on the wrong path.

The question of constitutional legitimacy is directly addressed by Schmitt in section 9 of Constitutional Theory. The locus of constitutional legitimacy for Schmitt (2008: 75) is, as suggested above, the positive constitution considered as an act of constituent power which existentially determines constitutional form through ‘a single instance of decision’. But Schmitt (2008: 136) also acknowledges that ‘a constitution is legitimate not only as a factual condition’. A constitution is also recognised as ‘a just order (rechtmäßige Ordnung), when the power and authority of the constitution-making power, on whose decision it rests, is acknowledged’ (Schmitt, 2008: 136). So, for example, the legitimacy of the Weimar Constitution rested on concrete acts and decisions identifiable with the German people’s exercise of its constitution-making power in establishing that constitution and with the subsequent acknowledgement that it was a genuine expression of the German political will (Schmitt, 2008: 136).

Schmitt’s remarks regarding the recognition of a constitution as a just order are highly ambiguous. In the first instance, Schmitt distinguishes between a factual sense of legitimacy and a sense of legitimacy that seems more normatively robust insofar as it incorporates the notion of a just order. Yet both forms of legitimacy are also referred back to recognition and acknowledgement, consistent with a thin Weberian concept of legitimacy as an explanatory type that allows for a description of the belief or faith of members of a political community in the power and authority of a regime. ⁴ The attribution to Schmitt of such a normatively inert conception of legitimacy seems consistent with his denial of the possibility or relevance of a normative justification for both the state and state authority. As the political unity of a people – which precedes any form of constitutional ordering – the state for Schmitt (2008: 138) ‘exists in the sphere of the political specifically’. From Schmitt’s (2008: 138) perspective it is just as senseless for the state to advance a normative justification for itself as it would be for an ‘individual living person… [to] justify his existence normatively in the sphere of private law’. Political existence and the decision for a particular constitutional form do not call for, or even admit of, a normative justification of any kind in Schmitt’s theory (2008: 136).

It is also instructive in this context that some recent advocates of Schmitt’s theory of popular sovereignty have elected to focus upon its capacity to contribute to an understanding of contemporary constitutional practice rather than upon the normative question of constitutional legitimacy. Paul W. Kahn, for example, emphasises that his intention in engaging with Schmitt is to provide a better descriptive understanding of the political experience of Western liberal-democratic regimes (the United States in particular) rather than to pursue a normative inquiry regarding good and bad political forms. Kahn sees in Schmitt a source of insight regarding the importance of sacrifice and the irreducibility of sovereign will to rational norms for contemporary constitutional theory. Although it is doubtful whether Kahn’s (2011: 75) advocacy of a ‘jurisprudence of the will’ can be neatly separated from an account of law’s normative point, it is clear enough that his main intention is to focus upon the explanatory benefits of Schmitt’s concepts of popular sovereignty and constituent power rather than the normative question of legitimate regimes.

There are nonetheless two reasons why the attribution of a normative theory of constitutional legitimacy to Schmitt is justified. The first is that Schmitt’s frequent recourse to the notion of legitimacy at important junctures in his work during the Weimar period makes little sense except on the assumption that his concept of legitimacy has a genuine normative aspect. The second, related reason, is that a purely descriptive theory of legitimacy is of doubtful cogency in any case.

As Vinx (2013: 107) has demonstrated, Schmitt (1988: 33–50, 1996: 22–28, 2004b: 17–36, 2005: 18–21) invariably appeals to the distinction between legitimacy and legality in his arguments regarding parliamentary decision, the rule of law and judicial review throughout the Weimar period. The distinction is particularly central to Schmitt’s critique of Kelsen’s pure theory of law. According to Schmitt (2008: 64), although Kelsen’s postulation of a Grundnorm may explain legal validity, it culminates in a ‘tautology of a raw factualness’ because it abjures legitimacy considerations. As a result, Schmitt (2008: 64) argues, Kelsen is ultimately unable to explain the normative status of the constitution. ⁵ This critique appears to rest squarely on Schmitt’s own view that a constitutive existential decision for a particular constitutional form can serve as a normative foundation for legal validity (Vinx, 2013: 112). Schmitt’s polemic against Kelsen’s normativism, or reduction of legitimacy to mere legality, thus seems to presuppose a normative conception of legitimacy.

Whatever Schmitt’s own intentions, his arguments regarding constitutional form and legal validity thus suggest an approach to the problem of legitimation that is not purely descriptive. And it is in any case doubtful more generally whether an adequate treatment of political forms of legitimacy is possible without recourse to both descriptive and normative points of view. Although Weber’s (1964) descriptive account of legitimacy emphasises the belief or faith (Legitimitätsglaube) of participants in a political group with respect to power and authority, his analysis also seems to presuppose that the relevant beliefs are directed towards grounds of those beliefs, whether these are based on tradition, charisma or legal-bureaucratic rationality. Weber’s analysis thus implicitly incorporates reference to grounds of justification that would support beliefs in the descriptive sense (Beetham, 1991: 11). ⁶ The alternative is extreme: an uncompromising error theory that interprets all forms of belief in the grounds of the legitimacy of authority as inscrutable or illusory. Yet there is little basis for thinking that Schmitt was committed to such an uncompromising view given his arguments cited above.

There are thus strong grounds for attributing to Schmitt a normative theory of constitutional legitimacy grounded in a concrete constitution-making act. Additional support for this thesis is found in the fact that both critics and advocates of Schmitt’s position have offered normative readings of his theory. For Vinx, who argues that a theory of constitutional legitimacy grounded in strong popular sovereignty culminates in incoherence, Schmitt’s position is that a

constitution is itself legitimate, and thus has the power to legitimate ordinary laws enacted in accordance with its rules and constraints, if and only if it has been created by a constitution-giving act on the part of the people as constituent power and continues to enjoy the people’s tacit support. (Vinx, 2013: 102)

In sum, the core of Schmitt’s concept of constitutional legitimacy in the normative sense is the concrete exercise of constitution-making or constituent power by a prior political unity. Schmitt’s concept of constitutional legitimacy thus rests squarely on the positive concept of the constitution. In the next two sections, I will now explain why this positive conception of the constitution is ill equipped to serve in an adequate account of constitutional legitimacy. Such a conception, I claim, simply lacks the basic normative resources to discharge the required justificatory role.

Section 2

Schmitt’s theory of constitutional legitimacy, I argued in the previous section, sustains a normative interpretation. According to Schmitt, a constitution is legitimate in the normative sense when it is the outcome of an exercise of will expressed in an act of constitution-making by a politically unified people. The attribution of a normative theory of constitutional legitimacy to Schmitt has important ramifications for an assessment of the cogency of his theory. Insofar as Schmitt’s constitutional theory sustains a normative interpretation, then it must be assessed not only by reference to its description of bare facts about the establishment of political forms – and the acknowledgement of the authority of such forms – but also by reference to whether or not it satisfies the minimal explanatory constraints on a normative theory of constitutional legitimacy. For while there are a wide range of theories of constitutional legitimacy, it is nonetheless plausible that there are minimal explanatory requirements that a theory of this kind must satisfy. These requirements in turn are likely to reflect necessary constraints on theories of political legitimacy more generally. In the current section, I thus explicate a core minimal constraint on a general normative theory of political legitimacy, before turning to the specific question of constitutional legitimacy.

One particularly influential model of political legitimacy in the Anglo-American literature concentrates on the duty to obey political authorities. ⁷ On this ‘standard’ model, a political authority is legitimate only if that authority possesses a Hohfeldian moral power or right to impose on its subjects a general moral duty to obey (Green, 1988: 21–62; Raz, 1986: 23–37; Simmons, 2001: 122–157). The problem of legitimate political authority, on this model, is thus closely related to the problem of political obligation: the conditions under which members of a political community have a duty to obey the directives of political authorities. Theorists engaged in this approach to political legitimacy hence nominate and defend a privileged concept or set of concepts that would ground the relevant general duty to obey. Popular candidates to fulfil this explanatory role include fair play, consent and associative obligations.

This standard model sets a very high threshold for legitimate political authority, especially insofar as it is usually conjoined with the thesis that a duty to obey must be ‘content independent’ with respect to the merits of particular political and legal directives. Certainly one might argue that the high threshold is appropriate given that political authority entails constraints on personal autonomy, directs people to comply (or at least conform) with directives that may ‘pre-empt’ their own reasoning about what it would be prudentially or morally best to do, and is coercive. ⁸ It is one thing to set a very high threshold for legitimate political authority, however, and another to assess its instantiation on the basis of a conceptual model that necessitates a negative outcome. And the evidence to date suggests that a ‘backward’ path from political obligation to legitimate political authority has limited prospects of success. ⁹ Weak forms of ‘philosophical anarchism’, which deny the existence of general political obligation and a presumptive moral duty to support the directives of a just state, thus tend to prevail in the current legitimacy literature (Simmons, 2001: 105).

In light of this state of affairs, some recent theorists have attempted to develop alternative theories of political legitimacy which reject the thesis that ‘other things being equal, a state is legitimate only if its subjects have political obligations’ (Zhu, 2017). Alternative theories include those that focus on the state’s liberty to coerce, those which concentrate on the duty of subjects not to interfere with the directives of political authority and ‘permissive’ accounts based on the moral power of a state to change the normative situation of its subjects (Applbaum, 2010; Edmundson, 1998; Wellman, 2001). For current purposes, I will focus on Stephen Perry’s recent advocacy of an account of political legitimacy that appeals directly to the value of authority and his related arguments against the claim that it is a necessary condition for legitimate political authority that subjects have a duty to obey political and legal directives. ¹⁰

According to the standard model set out above, a moral obligation to obey the law can exist only if there are authorities that possess a moral right to impose binding legal directives. And of course on one level the converse holds: if there is no moral obligation to obey laws enacted by a de facto authority, then that authority lacks a moral right to enact those laws. Perry has nonetheless argued persuasively that if the moral right to rule is understood as a Hohfeldian power to change the normative situation of another, then that moral right can exist even in the absence of any particular directives that subjects could obey or disobey. The Hohfeldian correlate of a moral power is a liability, not a duty (Perry, 2013). In principle ‘the power can exist without ever having been exercised’ (Perry, 2013: 24). ¹¹ Hence legitimate political authorities must possess the relevant moral power before they start issuing directives. The existence of a moral right to issue legal directives must thus be established before there can be a moral obligation on subjects identifiable with a general duty to obey.

A second related objection to the standard legitimacy model is that the proof of a duty to obey is in any case insufficient to establish the legitimacy of political authority. Whilst a moral right to rule can establish a duty to obey, the reverse entailment does not apply: a duty to obey cannot establish a moral (legitimate) right to rule (Perry, 2013: 24). All plausible conditions for the existence of a duty to obey could be satisfied and the backwards path to satisfaction of the right to rule still not be traversable. Examples of such conditions are the requirement that laws provide moral reasons and the particularity requirement that a legal obligation derive from the directives of a citizen’s own state (Green, 1988: 224–229). The reason that the satisfaction of these conditions would not entail the existence of a right to rule is that the standard model is not of the right type to establish that political authority is legitimate. The right to rule is a moral power and moral powers cannot be established by reference to obligations, rather they have the capacity to establish such obligations.

These arguments suggest that an analysis of legitimate political authority is best carried out through a direct strategy. Establishing a right to rule, on this alternative model, proceeds directly through an analysis of the conditions that would need to be met to justify political authority, namely that ‘there is some good to be derived, at least potentially, from some persons having the moral power to subject other persons to obligations’ (Perry, 2013: 24). The core of legitimate political authority is thus ‘the ability of one person intentionally to change the normative situation of another, where it is sufficiently valuable or desirable that the first person possess such an ability with respect to the second’ (Perry, 2013: 5). If what is at stake with the question of political legitimacy is the status of the moral claim of political authorities to issue directives, then on this model an adequate answer to the legitimacy question must appeal directly to the value of establishing and maintaining a system of political and legal authority in the first instance. And once this assumption is granted, a clear constraint on an adequate theory of legitimate political authority will be that it provides an intelligible and plausible answer to the question ‘why have authority at all?’ In asking this question one is seeking a normative reason, value or principle that could constitute a criterion for legitimacy by providing a justificatory ground for an assessment of political deliberation and its outcomes (Murphy, 2005: 134).

This survey of indirect and direct theories of legitimate political authority based on the notion of a right to rule – although necessarily brief and impressionistic – points in the direction of a minimal constraint on an adequate normative theory of political legitimacy. The indirect and direct approaches to political legitimacy outlined both seek to provide grounds of justification – a normative criterion – for an assessment of whether instances of political authority are or are not legitimate. With respect to the indirect and direct models the pertinent difference is whether the relevant normative criterion is sought (i) indirectly through an analysis of the capacity for political concepts such as consent, associative obligation or fair play to ground a duty to obey (and correlative right to rule) and (ii) directly by an appeal to the potential value(s) that authority can serve. ¹²

An obvious rejoinder to the characterisation of the minimal constraint above is that it overlooks the prevalence of normative theories of political legitimacy that appeal primarily to procedural rather than substantive grounds of justification. Jürgen Habermas, for example, has contended that the ‘democratic procedure for the production of law… forms the only… source of legitimacy’ in a post-metaphysical age. ¹³ According to Jeremy Waldron, laws enacted through democratic procedure may be regarded as legitimate on procedural grounds, even if such laws are unable to satisfy substantive tests. ¹⁴ The influential theory of legitimacy developed by John Rawls also places a strong emphasis upon the procedural aspects of democratic decision-making (Rawls, 2001).

The prevalence of procedural approaches does not, however, undermine the validity of the minimal constraint defended here. In the first instance, it may be argued that proceduralist approaches necessarily appeal to substantive values or principles (Huo, 2009; Michelman, 1998). For the sake of brevity, I here focus on Habermas’ reconstruction of the principles that support modern liberal constitutional democracies. As Frank I. Michelman suggests, Habermas’ reliance upon hypothetical agreement seems to assume a substantive rather than purely procedural legitimation. This can be seen in Habermas’ suggestion that the liberal commitment to the rule of law sets out from the status of modern law as positive, compulsory and individualistic. Reflected in a commitment to rights, the idea of the rule of law encapsulates the liberal view that ‘the democratic self-determination of citizens can be realised only through the medium of such law, the structural properties of which ensure liberty’ (Habermas, 2001: 766). ¹⁵ Liberalism, liberty and individual rights thus serve as values and objectives that the rule of law is supposed to realise in a way that allows for an assessment of the legitimacy of constitutional arrangements in terms of their success of otherwise in attaining those ends.

The validity of the constraint is in any case independent of the success of this argument. Whether or not they ultimately rely upon substantive normative values and principles, the proceduralist approaches above provide clear normative criteria for an assessment of the legitimacy of political authority. In these proceduralist approaches, the fairness, justice or impartiality of decision-making procedures is explicitly offered as grounds of justification. Proceduralist approaches accordingly stipulate desirable procedural features as normative criteria for legitimacy assessments. In so doing, procedural approaches corroborate the thesis that it is a minimal constraint on a theory of political legitimacy that it provide grounds of justification that allow for an assessment of particular decisions or institutional forms as legitimate or illegitimate. Although the prospects for a ‘pure’ proceduralism which renounces all appeals to substantive criteria are dubious, it is unnecessary for my argument that the requisite grounds of justification are substantive rather than procedural.

In sum, it is a minimal constraint on an adequate account of political legitimacy that it provides grounds of justification capable of serving as a normative criterion for an assessment of whether a particular decision-making process or institutional form is legitimate. And this constraint clearly limits the scope of candidate explanations for what would make an instance of political authority legitimate. Because it must provide adequate grounds of justification, a theory of political legitimacy will necessarily need to draw on normative concepts, whether these be of a deontic, evaluative or procedural character. What is sought in an explanation of the conditions under which a political authority would be legitimate, that is to say, are values, principles, reasons or procedures that could serve in major premises as part of a normative justification of political authority. Consequently, brute facts about will, concrete decisions and agreement or obedience are unable to do the necessary justificatory work, at least without normative supplementation.

An example may make this point more clear. I take it that no reasonable person would see any point in establishing an organisation whose sole function is to inflict cruelty on people with red hair. This is not to deny that there may be people who wish to co-ordinate action for sadistic ends, rather to assume that the actions of such people do not withstand reflection on reasonable human conduct. Now to ask for the conditions under which such an organisation would act legitimately in creating binding obligations and imposing a correlative moral duty to obey on agents would be quite perverse, precisely insofar as such an organisation serves no reasonable individual or social goals. And if a member of the cruelty to red heads group pointed out that the organisation was in fact legitimate because a large group of people had decided to set up such an organisation, then we would surely regard this response as inadequate. At the very least, we would expect an explanation from the member of the group as to the grounds of justification for setting up such an organisation that went beyond the mere brute fact that a group had decided to set it up.

There is little reason to think that these general considerations regarding an adequate general normative theory of political legitimacy do not also apply to questions of constitutional legitimacy. Constitutional authority of course presents its own unique problems. As Michelman (1998: 67) has argued, for example, the normative authority of constitutions raises distinctive questions regarding the relationship between perceived historical facts of authorship and the authority of a constitution for successive generations. In discussing constitutional authorship, Michelman (1998: 67) nonetheless acknowledges that his concern is not the ‘bare fact’ of regarding a textual constitution as the historical product of concrete authorship. In question is the connection between such authorship and the normative authority – its ‘serious impingement’ upon our judgement regarding what is required and permitted – of fundamental constitutional law (Michelman, 1998: 67). And in pursuing this investigation, Michelman arrives at the conclusion that we must inevitably appeal to some ‘contentful’ idea of reason or right that is able to support constitutional authority. It is accordingly impossible even for a constitutional populist to make sense of the normative authority of the constitution without reference to the ‘rightness’ of the content enacted by the authors of the original constitution (Michelman, 1998: 76).

On the assumption that the legitimacy of constitutional authority should be assessed according to the same general criteria that apply to legitimate political authority, then a theory of constitutional legitimacy will be inadequate if it does not offer grounds of justification as normative criteria of the kind outlined in this section. The question of constitutional legitimacy, that is to say, must be a search for the normative condition or conditions that would establish ‘the justified character for a set of fundamental laws’ (Michelman, 1998: 88). Indeed, the status of constitutional law as fundamental law – the law which sets out the higher level norms for law-making and adjudication – would seem to entail that the requirement that grounds of justification be provided to explicate criteria of legitimacy is here more pressing than it is in the case of the first-order norms dependent on fundamental law. In the final section of the paper, I argue that attempts to ground constitutional legitimacy in the constituent power of a strong popular sovereign are unable to satisfy even this minimal constraint.

Section 3

Theories of constitutional legitimacy grounded in strong popular sovereignty and constituent power lack the basic normative resources to meet the minimal constraint on a theory of legitimacy outlined in the previous section. An appeal to a constitution-making act by a pre-existing political unity cannot, at least without supplementation, provide normative grounds of justification that could serve in an assessment of whether a particular constitutional regime is legitimate. The basic problem for strong popular sovereignty can be stated as a dilemma. On a robust interpretation, strong popular sovereignty offers a combination of voluntarism and appeals to brute factuality that render it ill suited to meet minimal constraints on a theory of constitutional legitimacy. Alternatively, a theory of popular sovereignty must import supplementary justificatory resources from elsewhere. In so doing, however, it collapses into a less robust theory of popular sovereignty grounded in substantive normative criteria irreducible to acts of constituent power.

Recent criticisms of Schmitt’s account of constitutional legitimacy have tended to emphasise its incompatibility with basic assumptions of contemporary liberal constitutionalism. Foremost among these assumptions are the need to provide a framework for authoritative legal decisions in the absence of agreement on comprehensive doctrines and a commitment to internal principles of legality commensurate with the rule of law (Dyzenhaus, 2004; Vinx, 2013). ¹⁶ It is worth examining these criticisms in more detail as they are suggestive of the more fundamental problem with strong popular sovereignty identified in this paper, albeit without always making that problem completely explicit.

Vinx’s argument against strong popular sovereignty as a theory of constitutional legitimacy rests on a substantive claim about the function of legitimate law in contemporary liberal-democratic regimes. This function ‘is to reconcile us to the heteronomy that we inevitably suffer in political community’ as a result of the need to take collective decisions that cannot satisfy everyone (Vinx, 2013: 103). ¹⁷ A theory of constitutional legitimacy based on a thick political unity expressed in a concrete act of constituent power is clearly in tension with this conception of law’s function. Insofar as the advocate of strong popular sovereignty commits to the thesis ‘laws are only justified as long as they express an antecedent shared identity’, then a convincing account of the law’s function to resolve disagreement would appear to be precluded (Vinx, 2013: 103). For, on the one hand, the notion of constitutional legitimacy will strictly speaking be redundant where an antecedent shared identity exists. And, on the other hand, where such thick identity does not obtain, it seems that an advocate of strong popular sovereignty must conclude that legitimate law is impossible in principle.

A defender of strong popular sovereignty has a potential rejoinder to Vinx’s argument. Vinx assumes a concept of legitimacy based on the arbitration of ‘differences of interest and moral opinion through the use of fair legislative and adjudicative procedures and through the enforcement of normative constraints on the exercise of political power’. The success of this argument thus seems to depend on the attribution to Schmitt of a liberal conception of legitimacy (Vinx, 2013: 107). This is problematic, however, given Schmitt’s commitment to decisionism and use of the friend/enemy distinction for the purposes of a strident critique of liberalism. Vinx (2013: 107) is aware of this concern, but argues that Schmitt’s defence of the rule of law commits him to a conception of legitimacy as involving the resolution of disagreement by implication. It would moreover seem possible to frame Vinx’s objection in a way that avoids the charge that it depends upon the attribution to Schmitt of liberal assumptions. Schmitt’s conception of strong popular sovereignty, one could argue, is so far removed from the descriptive reality of legitimacy in modern liberal states informed by value pluralism and the lack of shared thick identity that it is simply implausible.

Even this reframing of Vinx’s objection could be countered, however, by a Schmittean who is prepared to uphold a normative account of constitutional legitimacy that explicitly rejects the desirability of reconciling value pluralism through parliamentary debate, democratic deliberation and legal procedure. And it is, moreover, open to such a theorist to offer an alternative account of legitimacy grounded in the expression of a thick shared identity in the decision for a particular constitutional form. Such a conception of constitutional legitimacy may seem far removed from the circumstances of contemporary liberal-democratic regimes on a descriptive level. It may also be regarded as morally abhorrent insofar as it suggests the elimination of dissent from the political domain. Yet neither of these concerns rules out as a matter of conceptual necessity the possibility of a normative theory of constitutional legitimacy grounded in constituent power.

Similar considerations apply to David Dyzenhaus’ rejection of constituent power by reference to the internal morality of law. The notion of constituent power is a distraction for legal theory, according to Dyzenhaus, because it neglects the status of the law as a reason-giving practice. A more promising line of inquiry, Dyzenhaus (2004: 257–258) suggests, is one which deals with the authority of law as ‘a matter internal to legal order’ and which tracks the potential for the reason giving encapsulated in the idea of the rule of law to transform de facto power into legitimate authority.

Once again, however, this argument is of limited force against an unrepentant Schmittean prepared to reject wholesale liberal appeals to law’s internal morality. Even if Schmitt’s commitment to the rule of law and theory of strong popular sovereignty are ultimately irreconcilable, an advocate of strong popular sovereignty need not accept the moral implications of the rule of law. A contemporary Schmittean might, for example, have recourse to Raz’s analogy of the rule of law with a knife (which is as apt to be used for bad as good purposes) or simply dismiss law’s internal morality as a cloak for the imposition of individualistic liberal values. ¹⁸

The intention of this discussion is of course not to defend Schmitt’s theory of constitutional legitimacy. It is simply to indicate that critiques of Schmitt’s theory which rely on substantive liberal premises regarding the need to neutralise disagreement and the rule of law are vulnerable to a Schmittean who is prepared to reject the underlying assumptions of liberal constitutionalism. Dyzenhaus (2004: 259) thus points us in the direction of a more direct and powerful critique of Schmitt’s position when he suggests that the mistake made by theories of constitutional legitimacy based on strong popular sovereignty and constituent power is that they assume that ‘fiat by itself supplies an authoritative reason’ for respecting a constitution or the laws legitimated by it.

One merit of Dyzenhaus’ more direct critique of strong popular sovereignty is that it implies a constraint on a theory of constitutional legitimacy that does not presuppose a commitment to a liberal interpretation of legitimacy as a function of the need to both respect and resolve disagreement over comprehensive doctrines. In principle it is possible to reject this liberal assumption while nonetheless insisting that a requirement of an adequate account of constitutional legitimacy is that it refers to law as a reason-giving practice. One could, moreover, uphold the need for an adequate normative theory of constitutional legitimacy to appeal to justificatory reasons for establishing and maintaining constitutional authority without accepting Dyzenhaus’ further claim, i.e. that the status of law as a reason-giving practice is best understood from the internal perspective of distinctively legal reasons oriented to the value of the rule of law. What is most fundamentally at issue here is the need for an adequate theory of constitutional legitimacy to appeal to grounds of justification and normative criteria for the establishment and maintenance of a constitutional regime.

No doubt part of the motivation for Dyzenhaus’ appeal to internal principles of legality encapsulated in the idea of the rule of law is that to look outside the law for the normative force of the constitution might appear to require a commitment to a form of ‘a fictional eternal natural law’ grounded theologically or in a suspect metaphysics (Kalyvas, 2005: 229). A ‘temporal order that declares metaphysical bonds or obligations (religio) to be a private matter’, it may be argued, can ‘no longer justify law by invoking a higher being or a socially binding “truth” derived from nature and its laws’ (Stolleis, 2008: 54; C.f. Schmitt, 2008: 266). Yet these assumptions – contestable as they are – need not be contested in order to reject the view that the normative reasons that count with respect to a theory of constitutional legitimacy must be exclusively, or even primarily, reasons internal to law. Normative reasons are such ‘in as much as by endowing an action with a point or purpose they guide decision and action, and form a basis for evaluation’ (Raz, 2011: 14). And it is highly implausible that all the good reasons that there are with respect to the establishment and maintenance of a system of constitutional authority are themselves distinctively legal reasons (Essert, 2013). The good reasons that can be marshalled to support a regime of constitutional authority extend beyond the internal morality of law to incorporate ends and values such as security, communal and individual well-being, justice and the common good.

Now the defender of strong popular sovereignty could at this point double down on their rejection of liberal assumptions and simply dismiss reference to law’s reason-giving force as an illusory variant of neo-Kantianism. Such an uncompromising position, however, runs up against the difficulty of providing a normative theory of legitimacy that altogether abjures reference to rational justification. A robust version of Schmitt’s robust theory of constitutional legitimacy in terms of the positive constitution, that is to say, would appear to be unable to offer anything that can serve as the normative grounds of justification for the establishment and maintenance of a constitution, whether these normative grounds are regarded as internal or external to legality.

The reason for this impasse is that an act of constituent power by a politically unified group – considered as a concrete and contingent existential decision and expression of will – is simply the wrong kind of candidate to serve as a grounds of justification for a constitution or anything else. This can be seen by reference to the example above regarding the establishment of a group for the persecution of people with red hair. The mere decision and concrete act of setting up such an organisation does not, without the supplementation of further premises, offer a justification for anything at all. On the assumption that Schmitt’s theory of constitutional legitimacy is a normative one – as argued in section one – then it must meet the constraint that it refers to grounds of justification that would provide a criterion for an assessment of whether a particular regime is or is not legitimate. Appeal to a concrete constitution-making act that reflects a pre-existing political unity is simply unable to perform this justificatory function: in its focus upon concrete manifestations of political will, it does not supply any intelligible justificatory or normative criteria.

The normative deficit of strong popular sovereignty as just outlined explains why recent defenders of the role of constituent power have tended to import more substantive normative criteria in support of the theory from outside. Yet such attempts to supplement constituent power with concepts that are more apt to serve as grounds of justification for constitutional authority either (i) directly undermine the claim that strong popular sovereignty is sufficient to serve as a normative criterion or (ii) tend towards weaker conceptions of popular sovereignty based on values and principles such as equality, participation, fairness, self-determination, autonomy and rights. ¹⁹

In the latter stages of Constitutional Theory, Schmitt (2008: 255–267) emphasises the democratic credentials of his theory of constitutional legitimacy by appeal to the value of political equality. Schmitt (2008: 280–285) here argues that will formation grounded in a denial of natural differences is closely connected with the idea of democracy and is in fact core to the modern principle of legitimate political authority. As Vinx (2016) points out, Schmitt’s discussion of the value of political equality needs to be understood in the context of his thesis that political community is based on a distinction between friends and enemies. The principle of political equality that Schmitt (2008: 257–264) endorses is not based on a moral equality of all persons, but rather refers back to the notion of thick identity discussed above. The important point for present purposes is that Schmitt’s recourse to political equality offers justificatory support to a theory of constituent power that has insufficient normative resources to serve as a criterion for legitimate and illegitimate constitutional arrangements. If democratic constitutional authority derives primarily from its realisation of the value of equality, then the positive conception of the constitution is no longer the locus of legitimacy on Schmitt’s account.

The reading of the notion of constituent power offered by Kalyvas also demonstrates what is at stake here. As discussed in section one, Kalyvas argues that Schmitt advances constituent power as a normative criterion that allows for an assessment of the legitimacy of existing constitutions. According to Kalyvas (2008: 99), Schmitt’s theory establishes the constituent popular will as a normative benchmark ‘with which one can test and assess the legitimacy of existing constitutions and of the basic structures of society to which it gave birth’. ²⁰ The difficulty with such an approach, however, is that neither the will of the people nor a factual act (such as a revolution or constitutional convention) is apt to provide a normative criterion for anything. It is surely for this reason that Kalyvas, in his further elaboration upon Schmitt’s theory of constitutional legitimacy, links the notion of constituent power with a theory of democratic participation. In this context, Kalyvas (2005: 238) asserts that ‘the normative content of the constituent sovereign is one of participation’ and this content demands that ‘those who are subject to a constitutional order co-constitute it’. This certainly presents the Schmittean theory in a more palatable light. But Kalyvas’ emphasis upon democratic participation and the Rousseauian sentiment that those subject to the laws should also be the authors of the laws imports normative criteria that extend beyond constituent power.

The same tendencies can be seen in Dieter Grimm’s defence of the continued role of popular sovereignty on the basis of its capacity to protect democratic self-determination. ²¹ Grimm (2015: 73) squarely acknowledges that popular sovereignty is a fiction, a mere subject for the ascription of acts of public authority. At the same time, however, he regards popular sovereignty as indispensable ‘as a legitimizing principle for the rule determined by the constitution’ (Grimm, 2015: 120). Insofar as in the modern world law is no longer found, but rather made, it can no longer derive from a transcendental source, but must emerge immanently from the people (Grimm, 2015: 69). At the internal level at least, Grimm notes that ultimately every function performed by popular sovereignty can be expressed in the concept of democracy. Sovereignty still has an important legitimating role to play at the external level of international law in protecting the identity of a society and its capacity for self-determination. Yet, at the level of the legitimation of national constitutions, the concept of sovereignty derives all its normative force for Grimm from the more fundamental concept of democratic self-determination.

At this point the defender of strong popular sovereignty is thus confronted with a dilemma: either uphold a pure voluntarism with dubious normative credentials or redress the normative deficit of strong popular sovereignty based on the idea of constituent power through the introduction of substantive criteria from a theory of democratic equality, participation or self-determination. ²²

If the defender of strong popular sovereignty takes the second route, it is difficult to see how this will not culminate in a theory which weakens popular sovereignty by attenuating the normative role of concrete existential acts of constituent power. There are admittedly a spectrum of options available. Perhaps the weakest conception of popular sovereignty will be that which understands popular sovereignty as internal to a constitutional order that gives equal rights of democratic participation to the people and makes appointment to offices elective (Habermas, 1996: 271–278; Vinx, 2013: 102). A more normatively robust conception that still falls short of strong popular sovereignty in the Schmittean sense is a Rousseauian conception according to which ‘the will of the people is identified with the outcomes of a democratic process governed by constitutional law’ (Vinx, 2013: 102). As Vinx (2013: 102–103) suggests, this conception of popular sovereignty is still weak, however, since it ties the existence of a people to the continuity of its constitution and thus has little place for the idea of a value-laden thick identity informing a political unity that would subsist even through the abrogation of an existing constitution.

It is clear enough that what performs the normative heavy-lifting in such accounts is not the idea of a robust political identity expressed through a concrete act of constituent power. On a liberal-democratic position of the kind defended by Rawls and Habermas, the normative core is a post-metaphysical commitment to the priority of individual autonomy grounded in principles of fair procedure and rights. On the more civic republican conception, the normative weight of the theory derives from principles of self-determination and political participation rather than a pre-existing constitutive identity. Insofar as the defender of strong popular sovereignty leans on such theories in order to provide a convincing account of constitutional legitimacy, then they are appealing to a weaker conception of popular sovereignty than is encapsulated in a Schmittean conception.

The alternative, as suggested above, would appear to be an unrelenting voluntarism incapable of grounding any substantive normative conclusions at all. Insofar as he addresses questions of constitutional legitimacy, Kahn (2011: 131) flirts with such a voluntarism in his invocation that ‘existence precedes essence’ in the political domain and his claim that the only faculty of the sovereign is will. ²³ Kahn (2011: 131) ultimately insists that we recognise the claim of the state as legitimate only insofar as we regard it ‘as the product of our own freedom’. Yet, without appeal to substantive normative criteria allowing for an assessment of political action, our freedom to establish a democratic constitution is indistinguishable from our freedom to establish a military dictatorship.

The ultimate reason why a theory of strong popular sovereignty grounded in constituent power is unable to offer a plausible account of constitutional legitimacy, I have argued, is that it simply lacks adequate normative resources to explain the conditions under which a particular constitutional regime would be justified. And this weakness inevitably plays itself out in the fact that defences of strong popular sovereignty tend to seek supplementary normative support from elsewhere.