Two concepts of constitutional legitimacy

Abstract

What legitimates constitutions? One standard answer is that constitutions are legitimate only if they represent the people they govern. This article identifies two different conceptions of representation. Representation can be grounded either in the consent or the will of the citizens or when the constitution reflects the ‘real’ identity of the members of the nation. Alternatively, it is sometimes stated that the constitution is legitimate because it promotes justice or, more generally, is grounded in reason. While constitutions are typically grounded both in claims to represent the people and in claims concerning the justness and wisdom of the constitutional provisions, we establish that there are two types of constitutions: constitutions that are primarily representational (e.g. the US Constitution) and constitutions that are primarily reason-based (e.g. the German Constitution). We also show that this distinction has important ramifications for how constitutions are drafted and ratified, and how they operate. One central implication is that the legitimacy of constitutions that make weak claims to representation – for example, constitutions that are imposed by foreign powers – can still be defended on reason-based grounds.

I. Introduction

It is common to assume that the legitimacy of constitutions hinges on the fact that they are representative (or expressive) of the people they govern. Numerous scholars have reiterated this observation and argued that constitutions should express the distinctive will, identity, character and the values of the nation they govern. ¹ Representativeness of constitutions is considered a prerequisite for their legitimacy. ² Why should we be bound by a constitution which is not ours – a constitution which does not reflect what we want, judge to be true or who we are? ³ Yet, at the same time, many theorists argue that constitutions can claim legitimacy on the grounds that they promote justice or, more generally, on the grounds that they are reason-based. ⁴ In such cases, the legitimacy of the constitution is grounded in the fact that it is just, fair, rightful, efficient and so on. While most constitutions claim legitimacy on both grounds, we establish that some constitutions are more representational while others are more reason-based, and that the distinction has important ramifications concerning the ways different constitutions evolve and operate. Before we briefly describe the two forms by which constitutions can gain legitimacy, let us concede at the outset that our discussion focuses primarily on liberal constitutions in the Global North. It is therefore necessary to qualify our description and assert that it is an investigation of two liberal concepts of constitutional legitimacy.

We use the term ‘representativeness’ in a broad way. By ‘representative constitution’, we mean a constitution that gains its legitimacy from the fact that it is a reflection of who the citizens are or what they desire or judge to be true. There is thus an intimate relationship between the citizens and the constitution, so one can argue that in essence being governed by the constitution is tantamount to being governed by the citizens themselves. Perhaps the clearest manifestation of this view is in the Preamble to the US Constitution, which states: ‘We the People …’ These three words have been interpreted to convey the conviction that the legitimacy of the US Constitution rests upon the fact that it is nothing but a reincarnation of the ‘real’ or ‘genuine’ will of the American people. ⁵

Often, although not always, representativeness is realized (or is perceived to be realized) by using real or hypothetical consent. This is why constitutions often require public acceptance in the form of referenda or elections of representatives that indicate the willingness of the people to be bound by the constitution. At other times, representativeness is a by-product of the fact that the constitution mirrors (or is perceived to mirror) the deeply entrenched cultural or moral values of the nation. In both cases, the ideal of representation of either the judgments or the preferences of the nation, or the nation’s deep values, is considered to be a necessary and often also a sufficient condition for the legitimacy of constitutions. ⁶

An alternative mode of legitimacy rests on reason rather than representation. What legitimates reason-based constitutions is reason, namely the fact that such constitutions are just and fair, and protect the right values. Thus, similarly to the Razian analysis of authority, the constitution is legitimate because it helps the polity to identify and act in accordance with right reason. ⁷ To succeed in doing so, the constitution needs to be just or fair and/or conducive to societal wellbeing. Hence the term ‘reason’, like the term ‘representation’, is used here in a very broad way, which includes any claim concerning the justness, fairness, rightness or wisdom of the constitutional provisions. ⁸

To put it provocatively, the legitimacy of a constitution need not rest on the fact that it represents the people whom it governs; it may simply rest on the belief that it is a good or just constitution and, consequently, it would be conducive to justice or to the well-being of the community which it governs. ⁹ Legitimacy is non-representational or reason-based if it rests upon the conviction that the (procedural, institutional or substantive) aspects of the constitutional order promote justice, the protection of rights, stability and/or well-being. Put differently, on the view of reason-based legitimacy, a constitution can be legitimate simply because it is ‘correct’ – that is, conforms with reason. While most constitutions gain legitimacy on both representational as well as reason-based grounds, we establish that some constitutions can be classified as more or less representational or more or less reason based.

Furthermore, ironically, in times of crisis a constitution that is judged to be good or just may gain its legitimacy precisely because it is non-representational. In these times, the polity may wish to distance itself from its own existing values and aspire to be transformed – that is, to not be governed by a representational constitution, but by a reason-based constitution. The establishment of the German Basic Law is a good example of such a constitution. The transformation of the German people was facilitated by entrenching a constitution that was not fully representational at the time it was drafted and perhaps was intended to be anti-representational; as the German nation has been corrupted and needs to be transformed, so ultimately in the future the constitution would become representational. In such cases, the constitution is legitimate not because it is representative but because it is just. The hope is that eventually it may become representative if the nation replicates the constitutional values.

Our point of departure is legitimacy. Famously, legitimacy can be understood either sociologically (descriptively) or normatively. Sociological (or descriptive) legitimacy is based on the convictions of the people that the constitution is legitimate and their willingness to subject themselves to the constitutional order. When one speaks of descriptive legitimacy, one inquires whether and why people believe that the constitution is legitimate: What makes them willing to be bound by it? Descriptive legitimacy therefore refers to public sentiments, in particular to the willingness of the public to respect the constitutional order and regard it as binding. Normative legitimacy refers to the ethical or, more broadly, normative justifiability of the constitutional order. Normatively a constitution is legitimate if there is at least a prima facie reason to obey its demands. When one speaks of normative legitimacy, one inquires whether and why people ought to believe that the constitution is legitimate: What makes them bound by it? Normative legitimacy therefore refers to the moral justifiability of the constitution. ¹⁰ This article is devoted primarily to identifying different types of descriptive legitimacy of constitutions. We establish that some constitutions are perceived to be legitimate on the grounds that they represent the people while others are perceived to be legitimate on the grounds that they conform with reason, but we also hint at some normative aspects of legitimacy of both types.

The reader may question the importance or significance of this classification of forms of descriptive legitimacy. To establish the significance of this classification, one first needs to show that different constitutional traditions rest on different modes of legitimacy and to establish that this classification has important consequences. In this article, we establish that representational or reason-based forms of legitimacy influence the ways constitutions are developed and interpreted, particularly the ways in which they are drafted and/or ratified. The classification is valuable as it can account for prominent features of the constitutional evolution in the relevant jurisdiction.

Representational constitutions are often endorsed by conducting a participatory process such as a referendum or elections. Moreover, interpretative methods such as originalism are typically grounded in representational aspirations. US legal theorists care about the original public meaning of the Constitution at the time of its ratification partly because it is the one that represents what Americans as a nation wanted or judged to be true or just. Further, the opposition to the use of foreign citations has also been dictated by the conviction that the legitimacy of the Constitution presupposes that it represents the American people. In contrast, the welcoming of foreign influence in drafting or ratifying the constitution is characteristic of reason-based constitutions. Finally, we maintain that the phenomenon of imposed constitutions – those that were forced upon the nation – can be accounted for in terms of reason-based legitimacy.

Section II describes the differences between representational and reason-based legitimacy. We differentiate between two types of representational systems: conventional representational constitutional systems that use consent or will and naturalist or ethno-national representational constitutional systems that rest on claims concerning the innate cultural values of the nation. ¹¹ We contrast both types of representational systems with reason-based systems. Section III looks at paradigmatic cases of representational and reason-based constitutional systems. It provides examples for each and establishes the impact of the difference on the constitutional evolution, the methods of interpretation and the processes of drafting and ratification. Section IV concludes the article.

II. Representational versus reason-based legitimacy

This section presents a classification of constitutional systems in accordance with the ways in which they are legitimated. We differentiate below between two forms of legitimacy: representational and reason-based. Given that our focus in this section is descriptive (or sociological) legitimacy, we rely heavily on the self-understanding of members belonging to the constitutional traditions. The classification of a constitutional tradition as falling into representational or reason-based camp rests on the self-understanding of the constitution shared by politicians, political activists and the public as a whole.

The self-understanding of members of society can be factually or historically fallacious. For instance, if a particular constitutional tradition is legitimated on the grounds that the constitution reflects (or constitutes) the will of the people, and the evidence for this is the belief that the people agreed or accepted the constitution, this latter conviction need not be historically accurate. It is the imagined or perceived realities that determine the category to which the constitutional tradition belongs, rather than what ‘really happened’.

Representational constitutionalism

According to theories of representational constitutional legitimacy, the constitution is legitimate because it is our constitution: it reflects our will or our present values or foundational convictions – most generally, it reflects who we are. The constitution is a revivification of the citizens who are governed by it. ¹² Under this view, the legitimacy of the constitution hinges on the fact that its provisions entrench what the people want, what their values are and who they really are. In John Adam’s words, a legislature ‘should be an exact portrait, in miniature, of the people at large, as it should think, feel, reason and act like them’. ¹³

Representational legitimacy is grounded in the ideal of self-governance. Under this ideal, if the constitution is based on the consent or the will or it entrenches the shared values of the people, it does not violate their autonomy; as a matter of fact, it should count as an exercise of their autonomy. Typically, this argument is being made on the grounds that the constitution enshrines government by us – the people; it is an embodiment of our values and/or collective will, and reflects who we are. ¹⁴ There are, of course, important differences between these assertions and those may have important ramifications, but they nevertheless share the view that the legitimacy of the constitution rests on an intimate bond between the constitution and the people; the people make the constitution or it reflects who they are, or even constitutes them as a polity so they are bound by it.

But what does the constitution represent? What aspects or dimensions of the people are being reflected in the constitution? Here, one may differentiate between two types of representational theories: conventionalist and naturalist. Under the conventionalist understanding, a constitution represents us because it replicates our will. Under the naturalist or essentialist understanding, a constitution represents us because it transcribes our values or ‘objective’ essence as members of a community or as a nation.

Conventionalist theories often require a ceremonial procedure in which people express their willingness to be bound by the constitution. The duty of the Jews to comply with the rules of God is traced to a verbal commitment expressed by the words recited by the Jews: ‘All the words which the Lord hath spoken will we do.’ The will to endorse the words of God is made salient in this case by making a verbal commitment. In the constitutional context, the commitment is made salient by conducting a referendum or elections.

Naturalist theories reject this view. Often the underlying premise of such theories is that the contingent or empirical will does not replicate who the people really are or what they really want; it is often superficial or manipulated. Instead, what ought to be represented are the deeply entrenched cultural or essential innate values of the nation. Under naturalist constitutionalism, the constitution is ours not because our empirical or contingent will is transcribed in it, but because our foundational ‘real’ values as a nation are replicated in it. This indicates that there is an unappreciated similarity between conventionalist representation and what we label here naturalist representation as in the latter case: ‘law is originally something held in common, something essentially ours, something indeed which only exists to the extent that it is embedded in and part of a shared way of life’. ¹⁵

To sum up, theories of representational constitutionalism rest on the belief that the constitution is legitimate because it replicates either the will of the people (conventionalist constitutionalism) or some objective features of the people – for example, their cultural or moral values (naturalist constitutionalism). In both cases, representation implies self-governance on the part of citizens. To the extent that the constitution represents the people, the people are not governed by others but by themselves, and hence they are free. Let us contrast this view with reason-based constitutionalism.

Reason-based constitutionalism

Under reason-based constitutionalism, the constitution is legitimate because its content is just, rational or grounded in reason. In this context, reason should be interpreted broadly; it means any content-based argument in favor of the constitution. For instance, the constitution protects basic rights, promotes justice or equality, guarantees stability, prosperity, or efficiency etc. To determine whether the constitution is grounded in reason, one ought to examine its content and judge whether its provisions are ones that meet high standards of morality and justice. For the sake of this article, we do not make any substantive presuppositions about what reason dictates.

In his theory of authority, Joseph Raz articulated the grounds underlying reason-based constitutionalism. In his view, law in general and constitutions in particular have authority when the law is instrumentally valuable in setting out rules and institutions that help people to comply with the demands of morality and right reason; ¹⁶ as John Gardner puts it, ‘the main point of … all law … is to help people to do what they ought to be doing anyway, quite apart from the law’. ¹⁷ Hence, the Razian account of the difference law makes in moral space is one of providing better guidance concerning the demands of right reason rather than direct (or independent) deliberation on the part of citizens.

Raz and other advocates of reason-based constitutionalism do not deny that what reason dictates may sometimes depend on people’s opinions, will, preferences, culture or values. Most likely, an effective and stable constitution is (at least as a general rule) one that respects the prevailing values of the society it governs. ¹⁸ But there is a fundamental difference between regarding the constitution as grounded in the will or the values or the identity of the people as such (as representational constitutionalists advocate) and constitutionally entrenching the will or the values of the people as a mere instrument to facilitate stability or effectiveness as (as reason-based constitutionalism advocates). Reason-based constitutions may be sensitive to the will of the people or their cultural values, but this sensitivity rests on pragmatic or instrumental considerations.

In reality, almost all constitutions gain their legitimacy in both a representational and a non-representational manner. Most famously the US Constitution is perceived to be a creation of the American people and at the same time a document designed to protect justice and freedom. ¹⁹ Yet, as we show below, some constitutions are more (or less) representational and/or more (or less) reason-based than others and this difference has important ramifications.

The dichotomy between representational and reason-based constitutional legitimacy can be analogized to the famous theological dichotomy between voluntarism and intellectualism in theology. One of the classical concerns of theologians has been to address the relations between God and goodness. Under one view, whatever God wills is good by virtue of being willed by God. This view is often labelled ‘theological voluntarism’. ²⁰ What is shared by all theological voluntarists is the view ‘that entities of some kind have at least some of their moral statuses in virtue of certain acts of divine will’. In contrast, anti-voluntarists adhere to the view that goodness precedes God’s will. The tradition of intellectualism in theology suggests that God wills the good because it is (correctly) judged by God to be good. Crudely speaking, under the former view the good is good because it is willed by God, while under the latter it is willed by God because it is good. It is precisely this classical distinction between voluntarism and the so-called intellectualism that we wish to import from theology to constitutional theory. We suggest that the legitimacy of constitutions may be grounded in voluntarism (representational constitutionalism) or intellectualism (reason-based or non-representational constitutionalism).

We turn now to illustrate the relevance of our dichotomy in the real world and, in particular, to show that some legal systems are primarily representational while other are primarily reason based. We stress the word ‘primarily’. Because the two types of legitimacy often coexist, many systems will form some combination along a spectrum between representation and reason. ²¹ There also exist what can be labelled polarized systems, where there is a controversy concerning the grounds of legitimacy and different groups use different forms of legitimacy to promote their constitutional agenda. ²² More generally, we turn to see how our proposed conceptual classification meets constitutional realities.

III. Conventionalist representation, naturalist representation, reason-based legitimacy: From constitutional theory to constitutional reality

This section uses examples to illustrate how representational and reason-based constitutionalism operate. We first provide examples to illustrate conventionalist and naturalist representational constitutions before illustrating the existence of reason-based systems.

Note that classifying a system as falling into one category or another is never an easy task. It is always the case that one can find arguments that can be classified as representational or reason based. Political and legal rhetoric never subjects itself to a neat theoretical analysis. Moreover, many constitutional systems go through different phases, and the prominent type of legitimacy they have shifts in time. The real challenge is to judge which type of legitimacy is more dominant at a particular time in a particular jurisdiction. Yet, even if the reader challenges our classification of a particular system into one of these categories, it does not undermine the significance of the classification as such.

To illustrate conventionalist representational systems, we discuss the US system and to illustrate naturalist representational systems, we use the Hungarian system. To illustrate reason-based constitutions we use the German and Bosnian examples.

Conventionalist representative constitutionalism

Conventionalist constitutionalism is voluntarist: it rests on the conviction that to be legitimate, the constitution needs to be grounded in the will of the people. Hence, many jurists who advocate conventionalist constitutionalism regard institutional mechanisms such as elections or referenda as essential for legitimacy.

The US Constitution is perhaps the main example of a constitution that is typically legitimated on conventionalist representational grounds; it is widely believed that it is the will of the American people that accounts for the normative force of the US Constitution. To illustrate the representational force of the US Constitution, Bruce Ackerman analogized the US Constitution to a non-realistic painting that may represent the deeper dimensions of one’s identity that may not be represented by a photograph. ²³ In describing the Philadelphia convention, James Wilson presented the ambition of the US Constitution:

[The Constitution] is laid before [the American people] to be judged by the natural, civil, and political rights of men. By their FIAT, it will become of value and authority; without it will never receive the character of authenticity and power. ²⁴

The US Constitution is held out by many to be the premier example of representation-based legitimacy. But it has not always been so. As Paul Kahn argues, at its founding the US Constitution was actually based on a unity of reason and will. ²⁵ The Constitution subordinated will (what we call conventionalist representation) to reason, based on a belief in rationality and what is now called political science. On this view, people should choose the proposed constitutional design precisely because it is based on reason – that is, it is the product of science and able to foster the public good, which is the opposite of private factional interests. ²⁶

Now, whereas this view prevailed during the Founding, over time representational legitimacy assumed a more dominant role. As Kahn argues, attempts to separate reason and will, science and legitimacy misconceive the original project of American constitutionalism, ²⁷ but they nevertheless succeeded – prompted by the problem of slavery, which the Constitution ‘solved’ by abdicating reason and surrendering to will, thus opening a gap between reason and will. ²⁸

The focus on will and representation can explain many features characterizing the US constitutional discourse. First it explains the importance attributed to the (alleged) participatory process by which the Constitution was drafted and ratified. Second, it explains the popularity of originalism as an interpretative method. Third, it explains the resistance of many US judges and constitutional theorists to use foreign sources – which, under this view, distort the will of the American people. Fourth, it explains the constant salience of the US Constitution in US political discourse and the extensive use of constitutional arguments in US politics. While none of these features follows inevitably from conventionalist representative constitutionalism, they are features that seem to support the hypothesis that the legitimacy of the US Constitution is grounded in the preferences and/or judgments of the people. Let us examine more thoroughly each of these features and its relations with conventionalist representative constitutionalism.

First, US constitutional discourse often rests on assumptions concerning the participatory process by which the US Constitution was drafted and ratified. ²⁹ The importance attributed to the process of drafting and ratifying characterizes many constitutions. ³⁰ The participation and the public involvement that grant legitimacy are often practised in different stages of preparing a constitution, including the drafting, consultation, deliberation, adoption and ratification. ³¹ The conventionalist representational legitimation typically hinges on real or imagined popular endorsement. In the American case, it is the 1787 Constitutional Convention and the ratification of the Constitution by the states that are considered to be particularly important. These historical events are of particular significance because they corroborate – or are believed to corroborate – the voluntary submission of the American people to the constitutional order. ³²

Second, originalism is a popular method of constitutional interpretation in the United States. Originalism is determined to identify the framers’ intent or the understanding of the people when the Constitution was ratified, independent of political science, reason or justice. History, the originalists claim, is relevant because it is only by understanding what the people believed when the Constitution was ratified that we can uphold the commitment to popular sovereignty, which in turn legitimates the Constitution. ³³ Otherwise, we are merely inserting our subjective preferences in lieu of the preferences that are embodied in the Constitution. Originalism indeed rests on various arguments, but perhaps the most influential is that it is only by being faithful to the original intent or the ‘original public meaning’ that the Constitution can truly be the product of the will of the American people and therefore be legitimate. ³⁴

Third, and relatedly, there is strong resistance in the United States to the use of foreign law in constitutional interpretation on the grounds that such use is incompatible with the sovereignty of the American people. Echoing this view, Justice Scalia has argued that:

The notion that a law of Nations, redefined to mean the consensus of states on any subject can be used by a private citizen to control a sovereign’s treatment of its own citizens within its own territory is a 20th century invention … The framers would, I am confident, be appalled by the proposition that, for example, the Americans’ people democratic adoption of the death penalty could be judicially nullified because of the disapproving views of foreigners. ³⁵

In resisting the use of foreign sources, it is often claimed that such use robs the American people of their power to rule. It is equated to a situation where others, outside the constitutional orbit, seek to determine for Americans which norms shall be binding upon them. ³⁶ Put differently, those who object to relying on foreign sources in interpreting the Constitution might also be construed as supporting conventionalist representational constitutionalism. ³⁷

Finally, the historical event in which the people endorsed (or are believed to have endorsed) the US Constitution is not sufficient to establish conventionalist representational legitimacy. To realize the ideal of conventionalist constitutionalism, the constitution needs not only to be accepted by the public at some historical moment but also to be constantly endorsed and be present in the public sphere; its meaning needs to be constantly debated and used in political debates. The US Constitution is a paradigmatic version of such a practice. It was ratified only by (some) White men, but it is nevertheless venerated by the vast majority of Americans; as such, it is continually endorsed as if it had been ratified by all. ³⁸ In the absence of constant public engagement with the constitution, constitutions would inevitably suffer from the famous intergenerational dead hand problem. ³⁹ In a famous letter, Thomas Jefferson expressed this concern and maintained that ‘the earth belongs in usufruct to the living’: the dead have neither powers nor rights over it. ⁴⁰ To overcome this problem of earlier generations governing later ones, the Constitution needs to have a persistent presence in the public sphere; each generation needs to endorse it in light of its own concerns. It therefore follows that under conventionalist representational theories, both the process by which the constitution comes into being and the ways in which it is being sustained at present in the public consciousness are crucial for gaining legitimacy.

Conventionalist representational legitimacy is regarded by many theorists as being shared by all modern constitutions. As Dennis Galligan argued, the people ‘declare the constitution to be theirs, created by them in the exercise of their authority’. ⁴¹ Under this view, the people are the ‘masters of their destiny’, and it is their participation in establishing the constitution that grants the constitution its authority. But, as we show below, conventionalist representation is only one form of representational legitimacy, which is perhaps dominant in the liberal psyche but it is far from being the only or even the most characteristic form of legitimacy.

Naturalist representational constitutionalism

Naturalist constitutions do not rest on the contingent subjective will of the people (as advocated by conventionalist theories), but on objective persistent attributes that it is claimed or believed characterize the nation. The constitution is binding because it reflects or embodies our ‘true’ or ‘real’ essence or destiny by entrenching the ‘real’ cultural and moral values of the nation.

Like conventionalist constitutionalism, naturalist constitutionalism is also grounded in the ideal of self-governance. The only difference is that according to naturalist constitutionalism, self-governance requires entrenching not the will of the people but their true and/or objective essential attributes. ⁴² Hence, unlike conventional representational legitimacy, the constitution can be imposed on us willy-nilly by virtue of our being part of a particular nation or culture, even if we are not aware of it or endorse it. This feature is what makes naturalist representational legitimacy dangerous. Yet our primary interest here is not normative legitimacy but descriptive legitimacy; hence, we merely wish to document that some existing constitutions clearly fall into this category and discuss the consequences following from this fact.

Some Eastern European countries fall into the camp that endorses primarily naturalist representational constitutionalism. The Hungarian Constitution (the Fundamental Law) is perhaps the most paradigmatic. It begins with a ‘national avowal’ that contains nationalist-cultural expressions. Thus the Preamble includes a reference to Saint Stephen who built the Hungarian state, to Christianity, to the national culture and to various significant historical events. The most revealing statements appear in the two final sections of the Preamble, which assert: ‘Our Fundamental Law shall be … a covenant among Hungarians past, present and future: a living framework which expresses the nation’s will and the form in which we want to live.’ The last provision of the Preamble asserts: ‘We, the citizens of Hungary, are ready to found the order of our country upon the common endeavors of the nation.’ While the language seems to indicate the existence of a covenant (and therefore also echoes the conventionalist representational terminology), it is not a covenant that rests exclusively on voluntary acceptance; it is determined by the objective essential natural features of the Hungarian nation and those are embedded in the Hungarian way of life. Indeed, lest there be any confusion, the Fundamental Law specifically states that there is ‘one Hungarian nation that belongs together’. ⁴³

Moreover, the entire project of the new Hungarian Fundamental Law was not essentially a legal project. As József Szájer, the head of the drafting committee of the Fundamental Law, stated:

Maybe we did not need one (a constitution – AH and AS) in a legal sense: you can sip wine from a plain drinking glass … but for this to be more than just an act of becoming inebriated, you have to do it in proper style … A constitution defines a country’s identity – it condenses what we think of our history, achievements and attainments … Hungary’s previous Constitution did not fulfil this function of defining the nation’s identity … so it perpetuated many problems and unresolved issues. ⁴⁴

Arguably, this is merely empty rhetoric, something that often characterizes constitutional preambles. Such preambles put forward a historical narrative that shapes constitutional identity, yet that narrative matters little for constitutional practice. This is a mistake for two reasons: first, preambles have interpretive consequences; ⁴⁵ and second, the Hungarian Fundamental Law draws on the concept of the Hungarian nation developed in the preamble for interpreting specific provisions. Consider two examples: first, the Fundamental Law provides that ‘The protection of the constitutional identity and Christian culture of Hungary shall be an obligation of every organ of the State’; ⁴⁶ and second, although the Fundamental Law guarantees a right to freedom of expression, it caveats that right by providing that ‘The right to freedom of expression may not be exercised with the aim of violating the dignity of the Hungarian nation. ⁴⁷

To see how the idea of the Hungarian nation and Hungarian identity, reflected in the Fundamental Law, takes shape in practice, consider the Hungarian Constitutional Court’s 2016 decision concerning the European Union’s relocation program for asylum seekers. To bolster Hungary’s position vis-à-vis the European Union, the (packed) court relied on the identity provision to hold that:

The Constitutional Court establishes that the constitutional self-identity of Hungary is a fundamental value not created by the Fundamental Law – it is merely acknowledged by the Fundamental Law. Consequently, constitutional identity cannot be waived by way of an international treaty – Hungary can only be deprived of its constitutional identity through the final termination of its sovereignty, its independent statehood. Therefore the protection of constitutional identity shall remain the duty of the Constitutional Court as long as Hungary is a sovereign State. ⁴⁸

This passage is remarkable. It explicitly acknowledges that the constitutional identity of Hungary was not created by the Fundamental Law. The latter merely recognized what already existed. Thus the will or consent, according to the court, is not especially significant. What matters is that the Fundamental Law reflects the pre-existing Hungarian identity, and by extension that of the Hungarian people, irrespective of their level of participation in crafting the new Fundamental Law.

Naturalist representation may be criticized on the grounds that it under-emphasizes the importance of the constitutional canonization of the national culture. Arguably, as many theorists have pointed out, national constitutions do not just point out or rest on the existing fixed identity of the polity, but also ‘redefine the identity of the polity and this act of re-identification is dependent on … constitutional memory, which canonizes an interpretation of the past to be remembered as the ground of the whole legal system’. ⁴⁹ Further, constitutions falling into this category entrench a particular and often sectarian vision of history and grant it salience. ⁵⁰ Naturalist constitutions thus serve to lift a snapshot of history and cement it into the mental sphere of the nation. ⁵¹ In this respect, constitutions also transform rather than merely reflect the ethno-cultural identity. ⁵² This observation, however, does not undermine our account, which captures not the realities but the perceived realities – namely not what naturalist constitutions do but what they are understood publicly to be doing or what they claim or purport to be doing. The vision underlying national constitutions is not to construct a new identity or even to revise it, but to replicate an existing one, which is what legitimizes naturalist representational constitutions.

Naturalist representational constitutionalism also has institutional implications. In contrast to conventionalist representation, the act of entrenching the Constitution is not the most significant feature for legitimizing the constitution. This observation was made explicit by the Hungarian Court when it stated that ‘the constitutional self-identity of Hungary is a fundamental value not created by the Fundamental Law – it is merely acknowledged by the Fundamental Law.’ Indeed, the constitution-making process leading up to the adoption of the new Hungarian Constitution (Fundamental Law) cannot be considered to have been inclusive, deliberative or sufficiently transparent. ⁵³ Thus, to understand what the Constitution dictates, one needs to investigate who the Hungarians really or objectively are rather than what they want or judge to be just. Hence, the Hungarian courts need not inquire into questions of original intent or original public meaning, which focuses on understandings prevalent at the time of adoption, as even the founders have no authority to make decisions concerning the values of the Hungarian nation. It is the given culture and history of the Hungarian nation that they ought to replicate, rather than the intentions or beliefs of the founders. ⁵⁴

To summarize, naturalist representational legitimacy rests on the view that constitutions are legitimated by virtue of the fact that they entrench the natural or essential characteristics of the nation. It is not the actual contingent and perhaps flimsy will of the people at any particular moment that has authority, but the objective natural historical and cultural features of the nation. Note that this does not imply that the procedure of founding the constitution has no significance; however, unlike the case of conventionalist representational constitutions, the participatory procedure of founding the constitution and the continued support and engagement of the people with the constitution have no constitutive function. The (majoritarian or democratic) procedure may be used as evidence that the constitution is representational rather than as the defining feature of representation.

Non-representational (reason-based) legitimacy

Non-representational or reason-based legitimacy rests on judgments concerning the institutional, procedural or substantive justness, correctness and usefulness of the constitution. In contrast to representational constitutionalism, reason-based constitutionalism regards the constitution not as a self-governing document, the normative force of which emanates ultimately from those who are governed by it but as an instrument to guide the governed to act in accordance with reason (however defined). The constitution is legitimate, the citizens believe, primarily because it establishes good institutions, entrenches just norms and is thus worthy of our allegiance. The constitution is not a vehicle for self-governance, a reincarnation of our will or an inner essence, but an instrument that guides us to act in accordance with reason (broadly defined). As Raz argues, the law (and, more particularly, the constitution) gain their legitimacy by correctly identifying the dictates of reason and guiding us to act accordingly. ⁵⁵

In short, under representational legitimacy, the constitution merits our allegiance because it is ours (either because we endorse it voluntarily or because it entrenches our ‘real’ cultural or national values), while under reason-based constitutionalism, it is ours because it is grounded in reason and therefore merits our allegiance. The fact of consent is therefore, at least in principle, irrelevant. The object of that consent (if such consent existed at all) – namely the content of the constitution and, in particular, whether it is good, just, wise or, more generally, whether it succeeds in guiding us to act in accordance with reason – is what legitimates the constitution. ⁵⁶

This view has one important implication: a constitution need not be the product of the nation governed by it. Imposed constitutions, namely constitutions that are forced upon the nation, cannot easily be understood as representational. ⁵⁷ Instead, imposed constitutions can be understood to rest upon reason-based assumptions; their legitimacy is grounded in the moral or political desirability (or, given our interest in descriptive legitimacy, their legitimacy is grounded in beliefs concerning the desirability) of their provisions.

Three constitutions can be used as examples of constitutions that are grounded in non-representational reason-based legitimacy. Two are the result of a foreign power imposing a constitution after winning a war (Germany and Japan); the other is a constitution that was not imposed, ⁵⁸ but rather accepted as a result of an international effort to help a country transcend its bloody past, where that effort has been deemed to be ‘total influence’ (Bosnia and Herzegovina). ⁵⁹ We discuss the cases of Germany and Bosnia and Herzegovina below.

The German Basic Law was constructed (at least partly) by elites at a time when the masses were overwhelmed by the defeat in the war. This is also true with respect to the Japanese Constitution. Both constitutions were heavily influenced by foreign forces and they cannot be regarded as an authentic creation of the nations they govern. ⁶⁰

The outlines of the German Basic Law were to a significant extent dictated by the allies at the end of World War II. The conditions set by the allies included the requirement that the Basic Law be democratic, establish a federal order and protect basic rights. But it was also understood by the allies that if the new Basic Law were to succeed, it ought to be established in cooperation with the German people. ⁶¹ Moreover, earlier German constitutions such as nineteenth-century constitutions and the Weimar Constitution also shaped the 1949 Basic Law. ⁶² As Kommers notes, ‘whether monarchical or democratic, German constitutions, both state and national, have embodied and exalted the concept of the Rechtstaat, which is Germany’s distinctive contribution to modern constitutional thought’. ⁶³ Yet, despite the participation of German jurists and the influence of the German constitutional tradition on the 1949 Basic Law, the German Basic Law cannot be considered to be an exclusive product or creation of the German people. The text of the Basic Law that ultimately was adopted was subject to the approval of the allies, who used their powers to dictate changes in earlier drafts prepared by German jurists. It eventually was approved both by the occupying powers and by the different Landers. ⁶⁴ Given these facts, it is evident that the legitimacy of the German Basic Law cannot be perceived, at least at the time of its founding, to be grounded in representation. In a rare attempt to address this concern, one of the former Justices of the German Court asserted: ‘Allied intervention did not succeed in branding the Basic Law with the stain of an instrument imposed by the occupying powers.’ ⁶⁵ This statement leaves much to be desired. Even if the German Basic Law is not an instrument imposed by the occupying power, it is also not one that was adopted freely by the German people and it cannot purport to represent them, at least not at the moment it was endorsed. ⁶⁶

To all this we add another piece of evidence indicating that the German Basic Law was not fully representational. The Basic Law was meant to be ‘preliminary’ and ‘transitional’ until the entire German people could join or accept a new constitution. Indeed, the Basic Law was conceived as a provisional partial constitution of West Germany until the population living in the Soviet Occupation Zone and the Saar Protectorate could also participate. This is made explicit in the Basic Law’s original preamble (amended post-unification), which states that, ‘[The German People] acted also on behalf of those Germans to whom participation was denied.’ This was also the principal reason why the Basic Law was not called a ‘constitution’ (Verfassung). The idea was that only after the unification of Germany could the German people be successfully represented. ⁶⁷

Thus we suggest that the German Basic Law was legitimated not by virtue of popular assent, endorsement, voluntary allegiance of the German people to the Basic Law, or even on a conviction that it reflects the German essence or nature (a view which will be rejected by most Germans), or on any other form of representational legitimacy. Instead, it is legitimated on the grounds that it is believed that its content is desirable on both substantive and institutional reasons. The German Court sided with the reason-based view when it stated:

Laws are not constitutional merely because they have been passed in conformity with procedural provisions … They must be substantively compatible with the highest values of a free and democratic order, i.e. the constitutional order of values, and must also conform to unwritten fundamental constitutional principles as well as the fundamental decisions of the Basic Law, in particular the principles of the rule of law and the social welfare state. ⁶⁸

Note, however, that as time goes by, a reason-based constitution such as the German Constitution could gradually transform so it becomes representational. The discourse concerning German ‘constitutional identity’ could perhaps be interpreted as a step indicating that the German Basic Law may have become more representational. ⁶⁹ To be clear, we do not judge whether the German Constitution has been transformed into a representational one. We merely argue that at the time of its ratification it could not have counted as a representational constitution, but it could nevertheless gain legitimacy on reason-based grounds.

Support for our hypothesis that the German Constitution was not representational at the time of its adoption can indeed be found in early scholarly work. In an article published in 1949, upon the adoption of the German Basic Law, Carl Friedrich noted:

This date (the adoption of the Basic Law – AH and AS) was chosen intentionally to remind the German people that this provisional constitution is a way-station on the road out of the chaos which the collapsing Hitler regime left behind it. Any consideration of this Basic Law should start from the fact that the charter is not the creation of a free people, and that it will have to function within limits, both territorial and functional, which severely handicap its chance of becoming a genuine constitution, securely anchored in the basic convictions of the people. ⁷⁰

Concluding his article, Friedrich ends with an apt quote from a German newspaper, published soon after the adoption of the Basic Law:

The Basic Law is groping into the future. So is Germany. After many errors and disappointments, we are facing a beginning. How much new life will sprout from the ruins depends primarily upon ourselves. ⁷¹

These statements demonstrate how representative reasoning was divorced from the German Basic Law at the founding. Yet they also express optimism regarding the future, when the Basic Law might eventually become representational depending on the will of the German people.

Like Germany (and Japan), the Bosnian constitution can be considered an imposed constitution. The Constitution is in fact Annex 4 of the Dayton Peace Agreement between representatives of Bosnia and Herzegovina, Serbia and Croatia, signed in 1995. In contrast to cases of occupation such as Germany and Japan, the Bosnian Constitution was a result of peace negotiations between the warring parties, mediated initially by the United Nations and the European Union, and then transferred to the ‘contact group’, which included representatives from the United States, Russia, the United Kingdom, France and Germany. Although the Constitution was approved by the signature (not ratification) of the parties, it was essentially drafted by the mediators in English. There was no constitutional assembly or constitutional process, and no referendum to determine its acceptance. ⁷² This led to it being called a ‘Dayton Constitution’ rather than a ‘Bosnian Constitution’. ⁷³

The question of legitimacy in all these constitutions is therefore particularly disturbing, as it cannot plausibly rest on representation. The German and Japanese Constitutions were drafted by occupying powers, affording limited space for ‘we the people’. The Bosnian case is, of course, different, but it also originated in decisions made effectively by foreign powers not by the Bosnian people. Its enactment was removed from any representative pedigree. While it was signed by the leader of Bosnia, Alija Izetbegović, the two other signatories were the leaders of Croatia and Serbia, who obviously do not represent the Bosnian people. Indeed, the Constitutional Court of Bosnia and Herzegovina has ruled that the Constitution, being an international agreement, should be interpreted like a treaty, thus demoting its status by implicitly denying that it is a document that represents the Bosnian nation. ⁷⁴

Thus one can ask: What is the source of legitimacy of these constitutions? Why are the German, Japanese or Bosnian Constitutions, which have never been properly endorsed by the German, Japanese or Bosnian people, so stable and robust? ⁷⁵

We believe that many differences between the US and German Constitutions can be explained by using our classification. Acknowledging that not all constitutions are representational explains a fundamental difference between the ways different constitutions operate. As stated earlier, the US Constitution has a central role to play in the politics of the United States. American citizens hail the Constitution as a great achievement of their own; they study its provisions in school, debate its meanings in the media and often justify their political positions on constitutional grounds. To be a representational constitution, the US Constitution needs to be known and studied by the American people as it purports to represent them. ⁷⁶ In contrast, German citizens traditionally are not engaged in debating the substantive provisions of their constitution. Instead, they defer to the expertise of the Constitutional Court and accept its determinations as binding. ⁷⁷ Research has shown that the German Constitutional Court is highly popular and enjoys broad public support. ⁷⁸ More generally, the German case illustrates that a reason-based constitution need not be one that constantly shapes and is shaped by the people, their judgments or their objective essence or nature.

Note finally that this comparison can shed light on the reasons why some constitutions are representational and others are reason-based. The US Constitution was an act of independence from the British Crown. It confirmed (or rather established) the American people as distinct and independent from the British people. Here lies the key reason for its representational character. On the other hand, the German and Japanese constitutions were not ‘new’ constitutions, as these countries had their own constitutional histories and traditions. They were not acts of independence. The particular historical moments and political purposes determined the particular ways in which each constitution is legitimated (representative: conventional or naturalist; reason-based). To be clear, we do not purport to analyse the historical contexts that give rise to representational or reason-based constitutions. Instead, we document the existence of both types and identify the implications of using one or the other.

As stated above, the distinction between representative legitimacy and reason-based legitimacy is not binary, but rather exists on a spectrum. The German Basic Law, and the Japanese and Bosnian Constitutions are three primary examples of reason-based constitutions. A possible objection would be that these examples are dated and, moreover, are examples of imposed constitutions, ⁷⁹ and as such they do not reflect the way constitutions are made today. In particular, the argument would go, the German and the Japanese examples belong to a different era and the Bosnian example is an outlier.

In some respects, this is true. Imposed constitutions in the style of a foreign occupying power writing, approving or imposing a constitution for another people, while they could serve as paradigmatic examples of reason-based, non-representational constitutions, are no longer a part the constitution-making repertoire. To begin with, wholly imposed constitutions were rare, and today they smack of colonialist odor – although, interestingly, colonial powers – namely Britain and France – were not interested in constitutional enterprises. ⁸⁰ But imposed constitutions have not disappeared completely. The Bosnian Constitution, part of the Dayton Peace Agreement, is much more current – even though its circumstances are unique, reserved to war-torn countries that rely on international mediators to overcome internal conflicts.

Further, imposed constitutions might come with a heavy normative price tag. They are not the product of collective self-determination, and therefore are incompatible with the core principle behind a constitution as an exercise of the constituent power. ⁸¹ Moreover, because they are externally imposed, they might reduce the prospects of generating a stable framework that promotes justice. ⁸² They might create a mismatch between the imported provisions and the formal and informal institutional structure they are meant to govern. ⁸³

Yet, despite these reservations, there is a larger universe of at least partially ‘imposed’ reason-based constitutions. Some of these are sometimes described as ‘borrowed’ or ‘transplanted’ because they rely on the experience of other countries and accordingly import constitutional provisions. ⁸⁴ To those we can also add constitutions that are influenced by external elements, be they countries, institutions or individuals. Indeed, although constitution-making is often thought of as an exercise in sovereignty by ‘we the people’, there is a fairly established practice of international involvement in constitution-making – very much short of imposition, but an external influence nonetheless. Such interventions are potentially vulnerable to the difficulties mentioned above, although these difficulties are significantly less serious since they do not involve the brute force or domination that characterized the ‘old’ imposed constitutions. ⁸⁵ These interventions, by UN agencies, NGOs and international human rights organizations, sometimes (but not always) at the invitation of the state, influence and shape the structure and content of new constitutions, and can bestow them with legitimacy. Such interventions which are often accompanied by pressure, are designed and can be understood to interject into constitutions the dimension of reason. ⁸⁶

Before proceeding, one objection should be examined. A possible objection is that international involvement does not amount to reason-based legitimacy because, despite external influence, the constitutions are in most cases later ratified by the people in one way or another. This objection has some force, as indeed there will be some representation-based legitimacy for these constitutions. Still, we believe international involvement in constitution-making serves as a good example of reason-based constitutional legitimacy, or at least a mixed system, for several reasons. First, as we stated above, legitimacy is not binary but exists on a spectrum. International involvement comes in a variety of forms and intensities. It can be accompanied by pressure, so that the ratification cannot be understood to be wholly authentic. ⁸⁷ Second, international involvement is viewed as challenging the traditional mode of constitution-making that is based on the work of domestic actors acting in a representative capacity on behalf of ‘we the people’ or ‘the nation’ as the exclusive mode of generating constitutional authority. ⁸⁸ Third, even though there will be public input, it will be limited. Last, international involvement often stipulates which political actors participate in the negotiations and how the process will be structured. ⁸⁹

Of course, one could argue that all constitutions incorporate ideas that originated elsewhere. ⁹⁰ There are sometimes sentiments that seem to challenge ‘constitutional borrowing’ on the grounds that it undermines the representative nature of the constitution. ⁹¹ We reserve our discussion to cases of explicit involvement of states and institutions, rather than a domestic decision to import a legal doctrine developed elsewhere, ⁹² particularly cases where the involvement was accompanied by such pressures that the endorsement is not fully voluntary.

International involvement in the making and drafting of constitutions is certainly not a new phenomenon. ⁹³ Much of that history implicates colonialism, decolonization and empire, often generating local resistance and constitutional failure. ⁹⁴ Yet some of the more recent cases of international involvement are different, focusing on advising and collaboration rather than imposition. They are derived partly from the effects of a constitutional order on neighbouring states, the desire to jointly address particular problems, the need to ensure compliance with international legal obligations ⁹⁵ and the motivation to assist state-building in post-conflict societies. ⁹⁶ Of course, power politics still plays a significant role, and powerful states and institutions exert pressure for their desired outcome. ⁹⁷ Put differently, constitution-making today is not (only) about the people getting together and giving themselves a constitution. This erstwhile conception is rooted – whether or not it is historically accurate – in French and American constitution-making. Today, however, constituent power is not embodied in one body, but rather dispersed among many sites, not all of which are domestic. ⁹⁸ We provide some examples below.

The United Nations, for example, has issued a ‘Guidance Note’ detailing the type of assistance it offers in the constitution-making process. ⁹⁹ The guiding principles carefully seek to balance UN involvement – for example, in encouraging compliance with international norms, and insisting on ‘national ownership’ and supporting ‘inclusivity, participation and transparency’. ¹⁰⁰ Despite the gentle rhetoric, this external influence cannot properly be described as representational.

Regional organizations, the most notable of which is the Venice Commission in the Council of Europe, provide assistance either to member states or to states seeking their advice. ¹⁰¹ In addition to the UN and regional organizations, constitution-making processes also may involve individual foreign experts. For example, during the constitution-making process in Timor-Leste, the Constituent Assembly secretariat included five international parliamentary experts. Four of the five members were from Portugal, the former colonial power. Consequently, much of the discussion emphasized the Portuguese Constitution, which is perhaps why, as the international advisers later noted, the Constitution that emerged did not adopt their recommendations. ¹⁰² In contrast, the Assembly was much more responsive to the comments of the UN High Commissioner of Human Rights, and revised its draft in light of his concern that the draft did not adhere to international human rights standards. ¹⁰³

The constitution-making process does not have to be confined to one external institution providing advice. For example, the constitution-making process in South Sudan was influenced by government-funded agencies, international organizations and private experts hired by the government. The National Democratic Institute, a US-based non-profit organization that works with developing countries to increase democratic capacities, has pushed to include provisions on the separation of church and state, gender-based affirmative action and a right to access information. Such provisions were included in the 2005 Interim Constitution and remained in the 2011 Transitional Constitution, still in force today. ¹⁰⁴

Constitutional advising by foreign actors (individual experts, human rights organizations, regional and professional organizations, supra-national institutions and even countries) has now become the norm in constitution-making processes, especially when those implicate countries emerging from conflict, colonial rule or the domination of a foreign power. Such involvement has taken place in varied locales such as Albania, Bangladesh, Cambodia, East Timor, Iraq, Somalia, South Sudan and Sri Lanka, among many others. ¹⁰⁵ More generally, recent decades have yielded much similarity among post-1945 constitutions, suggesting influence, transnational diffusion and convergence – so much so that scholars have referred to ‘generic rights constitutionalism’, whereby countries adopt the same (or closely similar) set of rights provisions. It is unlikely that this similarity can be attributed wholly to domestic representation. A partial explanation is that the set of constitutional advisers proceed based on certain templates that they find desirable. Thus norm diffusion and influence by states, organizations and individual experts has generated a set of norms that may be considered a type of benchmark or ‘best practices’ that countries have come to embrace, irrespective of national sentiments or revealed preferences. ¹⁰⁶ Of course, none of this means similar provisions are similarly interpreted and applied, but this is not our point. Our claim is that, insofar as the legitimacy of these provisions (and the constitutions adopting them) is discussed, it cannot wholly be addressed by representation-based arguments because those fail to take account of the content and process leading to the adoption of these provisions. The adoption of such provisions might have been the result of external influence or, in some cases, influence that came closer to pressure. ¹⁰⁷ In such cases, reason-based legitimacy may play an important role in legitimating such provisions.

As indicated above, the conventional reservations applying to any dichotomy (except perhaps in logics) should also apply here. The dichotomy between representational and reason-based legitimation is not a sharp dichotomy; it is often the case that legitimacy rests on both representational and reason-based modes of legitimation. Often the judgments that a constitution rests on representational grounds and on reasons-based grounds are interrelated. Typically, different political and social movements use both representation-based arguments and reason-based arguments to justify their allegiance to a particular constitutional order.

Moreover, the balance between these two types of legitimacy-creating mechanisms can shift over time in such a way that a constitution that has been legitimated primarily on reason-based grounds may eventually be transformed and be legitimated on representational grounds and vice versa. ¹⁰⁸ Thus the answer to the question of whether a constitution falls into one camp or another may change from time to time.

While the examples discussed in this section – for example, the United States, Hungary and Germany – are relatively clear examples of constitutions that are grounded in a distinct form of legitimacy (conventional representation, natural representation or reason-based legitimacy), there are cases where the source of legitimacy is controversial (conflictual or polarized constitutions). Such controversies may have important normative and doctrinal implications. ¹⁰⁹ However, we reserve the discussion of such cases to the future, since here we are concerned primarily with classifying imposed constitutions as reason-based, for this classification provides an account that can alleviate the concerns raised by many regarding the legitimacy of these constitutions.

IV. Conclusion

Reason-based legitimacy can explain an old theological puzzle that appears in Exodus. It is said that Moses brought the book of the Covenant and, famously, the people said ‘all that the Lord hath spoken will we do’. The traditional interpretation of the Talmud indicates that the people were committed to obeying the dictates of the covenant before they heard them. But how could this be? How can you commit yourself to obey before you know what you obey? Perhaps the answer may be grounded in reason-based legitimacy. Given that the covenant is given by God, its provisions are inevitably just. Hence, the Jewish people knew that the provisions of the Covenant were just before they even heard them. It is the belief in the justness of the provisions that grants the covenant its legitimacy.

Our analysis established that constitutional legitimacy is not always grounded exclusively or even primarily in representation. The grounds for legitimacy differ from one constitution to another, and even within the frame of a single constitutional order from one time to another. Moreover, we have shown that the different grounds of legitimacy have important ramifications for the way constitutions operate, for the interpretative methods used by courts and for the role of the constitution in the political life of the nation.

To be clear, our discussion focused on descriptive legitimacy, yet it is not difficult to see the contexts in which non-representational legitimacy can, at least prima facie, be regarded as normatively compelling. As the example of Germany illustrates, at times a country does not want to be governed by its own values, namely by a representational constitution: it aspires to be different. Representational legitimacy was, at the time, not a viable option for Germany. What was an option, though, was an attempt to base the constitution on reason and hope that in the future there would be a convergence between the nation and reason so the Constitution would thereby become representational. This, however, will not be the product of the constitution entrenching the values of the nation but of the nation entrenching the values embodied in the constitution. Indeed, this could be the story of Japan, which chose to abide by the terms of its constitution a decade after it was imposed and thus granted it some degree of representational legitimacy.

Our analysis resolves the puzzles that are often raised with respect to the German, Japanese and other non-representational constitutions. More generally, it explains the underlying normative foundations of imposed constitutions and international constitution-making. To justify these phenomena purely on the basis of representational legitimacy often requires distorting the concept of representation. Reason-based legitimacy, which rests on the justness or efficacy of the constitutional provisions, narrows the gap between constitutional theory and practice as it allows us to understand the diversity of constitutions and the diversity of the means by which constitutions can be legitimated. Importantly, constitutions that make weak claims to representation can still be defended on reason-based grounds.