Introduction
In recent years, direct popular participation in the process of constitutional change has emerged as a significant practice in many places around the world.Footnote 1 The Brexit referendum (2016) and the ill-fated and non-authorised referendum in the region of Catalonia (2017) are only the most well-known examples. In several other places, constitutional referendums are arranged when broad revisions of the political and legal system have been proposed (as in Italy 2016). However, constitutional referendums also take place when the changes proposed appear to be of minor importance (as in the Bahamas in 2016) and in nations that do not meet the minimal conditions of electoral democracy (such as Azerbaijan, Kyrgyzstan, Tajikistan in 2016 and Turkey in 2017).
The popularity of the constitutional referendum – the fact that it is practised even when not legally required – and the observation that non-democracies frequently resort to it, are presumably accounted for by a desire to frame constitutional reform as legitimate by democratic standards.Footnote 2 By summoning the electorate to the polls at times when the constitution is to be revised, ‘the people’ are arguably offered an opportunity to authorise the revisions. The paradigmatic exemplar remains that of the US Constitution, which was ratified between 1787 and 1789 through a process of ‘quasi-direct democracy’ in which each state elected delegates for the sole purpose of approving or disapproving the new constitution. As a result, a sense of popular legitimacy was conferred on the first constitution of the Union, even though it contravened the previous constitutional order.Footnote 3
Contemporary theorists are attracted in equal measure by the constitutional referendum as an instrument for democratic legitimacy. There is reportedly ‘broad agreement’ on the importance of popular ratification of constitutional revision through referendum.Footnote 4 Specifically, the idea is that ‘actual participation of citizens in constitutional politics’ is necessary for the constitution to be ‘genuinely democratic’.Footnote 5
Several objections could certainly be made against the referendum as a procedure for constitutional decision-making. For one thing, there is mixed empirical support for the notion that popular participation in the making of a constitution secures its long-term stability and acceptance. Constitutional ‘quality’ might be even more important than openness, in terms of clarity in the distribution of power between institutions.Footnote 6 In addition, it has been argued that a deliberative spirit and desire for agreement is preferable when constitutional principles are at stake. The referendum, by contrast, is typically associated with a lack of reflection and antagonism.Footnote 7
However, in this paper we focus on the claim that popular involvement is a necessary condition for the democratic legitimacy of constitutional change and the understanding that this is a requirement of popular sovereignty. The argument for popular participation in constitutional politics is thus an application of the abstract doctrine of popular sovereignty.Footnote 8
Though the abstract ideal of the sovereignty of the people lends itself to multiple interpretations, the relevant idea in this context is that the people are the sole body with the capacity to authorise public power.Footnote 9 Given that the constitution is the ‘higher law’ that determines the rules of law-making, it follows that ‘the ability to engage in constitutional change is a fundamental act of popular sovereignty’,Footnote 10 the point being that democratic legitimacy does not just require popular participation in the making of the laws; it also requires popular participation in the making of the laws that regulate the procedures for law-making. Popular participation in the process of constitutional change is to certify that the standards of democratic legitimacy apply ‘all the way down the line’. Following from this understanding, the democratic legitimacy of the constitution depends on ‘the act that created it’ and the extent to which this act complies with ‘principles of participation and inclusion’.Footnote 11 The constitutional referendum plays a central role in manifesting the people as a ‘sovereign constitutive body’.Footnote 12 I will refer to this view as the popular sovereignty argument for the constitutional referendum.
The aim of this paper is to scrutinise the relationship between the popular sovereignty argument and the constitutional referendum and to argue that the latter is often not supported by the former. The theory according to which democratic legitimacy requires popular participation in the creation of constitutional norms does not automatically apply to constitutional amendments in real world politics; ‘the constitution’, as imagined by theorists of popular sovereignty, is an ideal that only partly correlates with the codified constitutions in the real world. According to advocates of the constitutional referendum, the constitution refers to the rules that govern ‘the institutional arrangements through which [the people] are governed’.Footnote 13 In reality, though, the extent to which the rules identified as ‘constitutional’ have this function remains an open question. As argued here, the popular sovereignty argument applies to certain constitutional rules, but not all of them, and sometimes applies to legal norms that are not part of the constitution. Popular sovereignty, therefore, does not always require constitutional referendum in order for the constitutive rules of the legal and political system to be legitimate by democratic standards. And even when it does, it is unclear whether the constitutional referendum is sufficient in that respect.
The democratic importance of the constitution
Constitutional legal norms are subject to adjustment and change in many ways, sometimes through legal interpretation and shifts in social and political presuppositions.Footnote 14 Though such ‘informal’ sources of constitutional change are inevitable, advocates of popular sovereignty typically emphasise formal routes to constitutional change – alterations in the codified constitution by the rules of amendment. Only formal processes of amendment provide the possibility for popular participation; the popular sovereignty argument for the constitutional referendum is concerned with constitutional amendment understood in this way.
In order to clarify the normative gist of the popular sovereignty argument, it is helpful to invoke the doctrine of constituent power. Though there are several variants of this doctrine, they all share the basic tenet that the powers exercised by political institutions are not theirs to begin with; ultimately, they derive from powers that belong to the people. The powers vested in political institutions – the constituted powers – descend from the constituent powers of the people; ‘the constitution is […] an expression of the constituent power of the people to make and re-make the institutional arrangements through which they are governed’.Footnote 15 Proceeding from the theory that the people hold constituent power, as authority or as a matter of sociological fact,Footnote 16 it is clear why popular participation in constitutional change is critically important. Assuming that power ultimately belongs to the people, only the people are authorised to change the document that defines the terms for the exercise of power by public institutions.
However, the notion that constituent power can be exercised through referendum is open to dispute, the objection being that constituent power is not a legal power and therefore cannot be manifested through participation in legally constituted procedures. Rather, constituent power is the non-legal entity that determines the constitution while not itself being determined by it.Footnote 17 As soon as the category of ‘the people’ becomes the object of constitutional law, the people are demoted to the branch of the constituted power.Footnote 18
In an attempt to circumvent this objection, Nicholas Aroney argues that popular participation in legally constituted procedures does not necessarily deprive the people of constituent power.Footnote 19 The idea is that the constituent powers of the people can be exercised by constitutional referendum, even if the institution of referendum is part of the powers that have been constituted. Indeed, unless we are willing to accept that the people, as a category defined by the law, are able to assume the office of constituent power, at least in part, the doctrine itself would lose all relevance in contemporary settings. The contrary view, that the people as constituent power is a force that exists prior to and independently of the constituted powers of the state, is most likely incoherent.Footnote 20
This revised understanding of the constituent powers envisages the people as a source of constitutional legitimacy, but not as the creator of the legal order as such.Footnote 21 A distinction should in other words be made between the making/creation of the constitution and the authorisation/legitimation of the constitution. The constitutive powers of the people are concerned only with the conditions for constitutional legitimacy. This understanding more readily explains why popular participation is essential to democratic legitimacy and why popular participation in moments of constitutional change is necessary for the democratic legitimacy of constitutions. The constituent powers of the people are employed in the ‘secondary sense’ when we refer the people as a ‘constitutional organ’ authorised to amend the constitution.Footnote 22
The challenge facing this doctrine is why the constitution, the whole constitution and nothing but the constitution, is an object of significance needed to secure democratic legitimacy. The point is that issues framed as ‘constitutional’ in existing political systems are not necessarily constitutive in the specific sense assumed by advocates of popular sovereignty. The question is, in other words, why the constitutional referendum is necessary for the democratic legitimacy of constitutional change, when referendum is not necessary for the democratic legitimacy of legal change in other respects.
One answer is offered by Andreas Kalyvas, who emphasises the interplay between the constituent powers of the people and the ideal of popular self-determination. Following Kalyvas, the doctrine of ‘constituent power’ is a more ‘sophisticated re-statement of the old, fundamental democratic principle of self-government and self-determination, according to which the people are the authors of the laws that govern them’.Footnote 23 The constitution is a legal document that is binding on its subjects in the same sense that other legal rules of the legal system are. In order to secure ‘self-government and self-determination’, the demand for democratic legitimacy applies to constitutional law just as it applies to regular law. Popular participation in constitutional politics is a prerequisite for self-determination, since the realisation of this ideal requires that the people participate in the process of deciding the laws they abide by. Joel Colón-Riós and Alan Hutchinson adopted the same view when they argued that ‘all laws… must be permanently open to regular transformations through highly democratic procedures’.Footnote 24
However, this argument risks being too strong, as it puts the doctrine of popular sovereignty at odds with the idea of representative democracy. If constitutional change requires referendum because people are subject to constitutions, it would then follow that referendum is also required for changes to ordinary laws since people are subject to them, too. In sum, it is impossible to justify the constitutional referendum by making an appeal to the dictum that the people should be able to participate directly in the making of the laws that govern them unless it has been accepted that referendum is necessary for the legitimacy of law-making of any kind. By this reading, the doctrine of the constituent powers of the people runs counter to the idea of representative democracy when the laws are not made directly by the people but only indirectly, by those elected by the people.Footnote 25
The resulting tension between popular constituent power and representative democracy is visible, also in the way populist movements today make use of ‘the people’ for the purpose of undermining trust in democratic institutions.Footnote 26 The relevant task is therefore to evaluate the popular sovereignty argument for the constitutional referendum in the context of representative democracy. For the argument to be successful, it must explain why amendments to the constitution are special in the sense that they require democratic legitimacy of a stronger and more demanding kind than changes to ordinary law. In the following, I attempt to articulate the significance of the constitution from the vantage point of the constituent powers of the people with two well-known, yet distinct, accounts of what a constitution is.Footnote 27
The legal status conception of the constitution
One way of thinking about what a constitution is, begins with the observation that constitutions are virtually always entrenched, meaning that changes to them are subject to more or less rigorous procedural requirements.Footnote 28 Hence, it is possible to define ‘the constitution’ as a set of legal norms that is more resistant to political change than other legal norms; constitutional legal norms are identified by the fact ‘that they can only be changed in a particular way’.Footnote 29
From the vantage point of democratic legitimacy, the relative rigidity of rules that regulate constitutional change – the amendment rules – represents a potential democratic problem.Footnote 30 Given that democracy is the principle that policy and law should depend on the process of popular will formation, some justification is needed for legally restricting the extent to which majorities are able to put their mark on the constitution.
One possibility is that entrenchment is considered democratically legitimate only when lenient. The nature of the rules of amendment is accordingly crucial in determining the extent to which the constitution is consistent with the ideal of popular sovereignty.Footnote 31 Another possibility is that entrenchment is considered legitimate to the extent that it has the appropriate democratic pedigree.Footnote 32 The idea is that constitutional restrictions on legislative majorities are democratically legitimate to the extent that they were adopted by procedures that are adequately and sufficiently democratic. The democratic origins of the constitution, not its substance, are for that reason essential.Footnote 33
It appears that advocates of the constitutional referendum are reluctant to accept the idea that democratic origins are sufficient to secure democratic legitimacy. The people should not only participate in constitutional politics at the beginning; they need be continuously involved in defining the terms of public power. As argued by Alan Hutchinson and Juan Colón-Riós ‘democracy resists political closure’.Footnote 34 For this reason, they consider ‘periodic reconsideration and revision’ of the constitution to be an ‘indispensable part of any democratic compact’. No matter how democratic the process that created the rules of amendment was at some historical moment, the people need to retain the ability to participate directly in the process of constitutional law making.
Resistance to ‘political closure’ in constitutional politics does not speak in favour of greater popular participation in all circumstances. If a more open process of constitutional change requires more lenient amendment rules, one could conclude that the constitutional referendum should not be adopted. All things being equal, leniency diminishes and rigidity increases as a result of the requirement for popular ratification. Following the calculations made by David Lutz, rigidity increases substantially with a rule that requires ratification of amendments by referendum, compared to a rule that requires ratification of amendments by a vote in the legislature.Footnote 35 If more than a simple majority in referendum is required for ratification, the rigidity of the amendment rule is of course even stronger. Thus, a potential conflict exists between the constitutional referendum as condition for democratic legitimacy and the claim that democratic legitimacy requires lenient rules of amendment.
There is a further problem with the view defended by Hutchinson and Colón-Riós that ‘all arrangements’ must be ‘subject to revision’.Footnote 36 Their point is that the sovereignty of the people requires a constitution that admits revision by those subject to it at any point in time. Conversely, a constitution that denies the people opportunities for constant revision is disallowed by the ideal of popular sovereignty. But the implication of this view is that the people are not allowed to create a constitution that denies them future opportunities to amend it. In order for the constitution to remain ‘subject to revision’, the people must not create a constitution that is immune to revision, as the people do not have the power to bind themselves irrevocably.Footnote 37
Interestingly, popular sovereignty as the ideal of ‘all arrangements’ being ‘subject to revision’ is seemingly incoherent. In order to ensure that the constitution is always open to revision, the rules of amendment must not be ‘subject to revision’. Instead, they must be designed in a specific way; placing as few barriers as possible in the way of amendment of other parts of the constitution. In effect, this means that the rules of amendment must be ‘closed and solid’ in the sense of protecting popular participation in the amendment process.Footnote 38 It is now time to examine the relationship between the popular sovereignty argument for constitutional referendum and the legal status conception of the constitution.
The legal status conception and the constitutional referendum
Assuming that the popular sovereignty argument for the constitutional referendum applies to the constitution as entrenched positive law, the object of exercises of popular sovereignty should be instances of legal change that are furthermore regulated by the amendment clause (excluding the amendment clause itself). Given that this is how the argument for the constitutional referendum ought to be understood, it follows that referendum is called for whenever the constitution is to be revised.Footnote 39 The popular sovereignty argument demands ratification through referendum of amendments to the constitution, irrespective of the substantive significance of the amendment. Conversely, since the popular sovereignty argument so construed applies only to constitutional amendment, it would not require ratification through referendum of changes to ordinary law even if substantively very important.
The point is that the document named ‘the constitution’, and recognised as such by public officials, does not just regulate important matters such as the delegation of power to different political institutions and the legal rights of citizens. Constitutions also include and entrench rules of seemingly subordinate importance. The implication is that changes to the constitutional document are sometimes ‘peripheral to the changes in the constitutional regime’.Footnote 40
The extent to which the constitution refers to matters of fundamental importance depends on several factors such as, for example, the range of issues covered by the provisions included in it.Footnote 41 The more issues covered, the less likely it is that every issue is fundamentally important. An indication of this is the length of the written constitution. The longer the text, the greater its scope and the more probable it is that it also covers some rather insignificant matters. Indeed, there is considerable diversity in the length of existing constitutions. Whereas the US Constitution is no more than 7,400 words long, the constitution of India includes no less than 95,000.Footnote 42 The length of the Indian Constitution indicates that it also applies to aspects of the political system that are not fundamental and, in that sense, not ‘constitutive’ of the legal and political system. A case in point is Article 125, which devotes 199 words to the task of defining the conditions for salaries and pensions of judges.
Even the US Constitution seems to include provisions that are of less than fundamental importance. One example is the 18th amendment introduced in 1920 (and the 21st amendment that repealed it in 1933), prohibiting the ‘manufacture, sale and transportation of intoxicating liquors’. Though the regulation of alcoholic beverages might be considered fundamentally important to the health and morals of the population, it clearly does not represent an issue of constitutive significance to the political and legal system.
Another example is the 20th amendment that was introduced in 1933 in order to redefine when the terms of the president, the vice president and the Congress begin and end. The rationale for the amendment was to reduce the length of the ‘lame-duck’ period in years of presidential election, as the old rule implied a four-month break between the election in November and the beginning of the term for the new president. The amendment is thought of as having ‘no significant effect’ on the US political and legal system.Footnote 43
The point is that the popular sovereignty argument for constitutional referendum does not always cohere with the fact that constitutional amendments sometimes concern less than fundamentally important matters. The argument is that direct participation in moments of constitutional change is necessary if the people are to exercise constituent power over the conditions for the exercise of public power. The problem is that many constitutional provisions simply do not apply to the terms for the exercise of public power at all.
A more sensible position would be that the people should only be activated as the constituent power in decisions that are constitutive of the legal and political system. This would include decisions that touch upon the extent to which the state is unitary or federal, republican or monarchical, the distribution of power between the judiciary, the legislature and the executive, and matters regarding the rights and duties that regulate relations between the state and its citizens. These are constitutive aspects of the legal and political system as they regulate the ‘institutional arrangements through which they [citizens] are governed’.Footnote 44
A good example would be the 2016 referendum in Italy that proposed amendments to the constitution for the purpose of changing the composition and powers of the parliament, and the division of powers between the state, the regions and other entities. These were truly constitutive revisions of the constitution to which the popular sovereignty argument would seem to apply.
To accept that popular sovereignty is exclusively concerned with constitutive matters, as such, is to reject the view that every constitutional revision is important to the exercise of popular sovereignty. The proper focus for the doctrine of popular sovereignty is not the codified constitution, but the set of legal norms that are constitutive of the legal and political system. If this is the basis for the popular sovereignty argument, a different understanding of the constitution other than entrenched positive law is needed. For the purpose of securing the democratic legitimacy of the constitutional order, there is little point in focusing on the set of legal norms that are governed by the rules of amendment and that are in that sense of higher legal status than ordinary laws. There is little point in doing that since the extent to which the constitution as such includes the legal norms that are of utmost importance from the point of view of popular sovereignty remains an open question. The claim that popular sovereignty should be concerned with entrenched positive law should be rejected.
Before this conclusion is accepted, however, we should consider one last attempt to defend the legal status conception of the constitution. The argument is that the codified constitution merits special attention because of the systemic implications of entrenchment. Though not every provision in ‘the constitution’ is substantively important, every provision in the constitution is important for procedural reasons because of the fact of entrenchment.
Consider once again the articles of the US Constitution that regulate Congressional term limits. Even if the content of these rules is not a constitutive feature of the American political system, they remain part of the document known as the Constitution of the United States and as such impose limitations on the powers of all major political institutions. The point is that the constitutional provision stipulating when Congress should end does not just impose obligations concerning term limits; it also regulates the powers of the Congress: the provision deprives Congress of the power to regulate its term limits by any regular procedure. It then follows that constitutional rules are significant, not because of their substance, but because they are entrenched.
But this is an argument that cannot be made on the basis of the constituent powers of the people. It is one thing to say that constituent power applies to ‘the constitution defined as the provisions that are entrenched’; it is quite another to say that constituent power applies to the constitution because it is entrenched. The latter is not an idea that undergirds the popular sovereignty argument. The reason the democratic legitimacy of the constitution is important is that the constitution defines the conditions of law-making and other exercises of public power. Popular participation in the creation of the constitution is valued because it confers democratic legitimacy to aspects of the political system that are constitutive. The point is that the constitution gains significance for the sovereign people because it is substantively important, not because it is entrenched.
To see why this must be so, consider the hypothetical case of a constitution that includes only insignificant rules – traffic rules, for example. In that case, the constitution would regulate speed limits, which side to drive on etc. – and nothing else. These rules would not be constitutive of the legal and political system and hence not of fundamental importance to the exercise of public power. They would nevertheless be the constitution if they were entrenched as a result of being regulated by the rules of amendment.
Given that democratic legitimacy, as defined by the theory of popular sovereignty, requires popular participation in the making of the rules that are constitutive of that system, it follows that the traffic-constitution would be of no importance to the democratic legitimacy of the legal and political system. The constitution would consequently be irrelevant from the standpoint of the idea that the people are the sole source of legitimacy for the rules that define the conditions for the exercise of public power. And yet, if the popular sovereignty argument is coupled with the notion that the constitution refers exclusively to entrenched positive law, it follows that the traffic-constitution should be subject to democratic legitimation. Since this conclusion is inconsistent with the basic aspirations of the idea of popular sovereignty, we must conclude that the status conception of the constitution should be discarded.
The legal functions conception of the constitution
The alternative is to conceptualise the constitution independently of the characteristics of codified constitutions. Instead of saying that the constitution is important because of its status, it is important because of the constitutional functions it serves, whether or not they appear in the codified constitution. This entails defining the constitution prescriptively rather than descriptively as it identifies the constitution with a specific set of legal norms that should be identified as the constitution.Footnote 45 If so, the constitution would correspond to the set of legal norms that ‘constituted’ the legal and political system.
Given the prescriptive and functional conception of the constitution, the popular sovereignty argument for constitutional referendum can be recast in a more convincing way. The reason legitimation by the people is necessary is that the constitution should define and regulate the exercise of public power. The need for democratic legitimacy in moments of constitutional change derives from the fact that it pertains to the segments of the legal system that matter most to the people subject to it.
The resulting understanding of the constitution is further elucidated by Dicey’s classical definition according to which the constitution refers to the rules deciding ‘the distribution or the exercise of sovereign powers of the state’.Footnote 46 In similar terms, albeit more broadly, John Gardner characterises the constitution as the rules ‘without which there would be no legal system’ or as the rules that ‘make up’ the practice of law.Footnote 47
The idea of the constitution as ‘the rules that constitute the legal system’ offers a particularly inspiring explanation of the requirement that the constitution comply with the demands of democratic legitimacy. Though citizens do not typically participate directly in the creation of every law, they ought to have the opportunity to participate directly in the creation of the rules that regulate the process of law-making. Hence, it is exactly because the people participate in the creation of the rules constitutive of the legal and political system that democracy is possible – even though the people do not generally directly take part in the making of ordinary law. Democratic legitimacy derives from the set of legal norms that define and regulate every aspect of the legal and political system to the extent that they originate from genuine acts of direct popular involvement; ‘the legitimation of the constituted powers depends on the extent of the citizen’s participation in the procedure of creating a Constitution’.Footnote 48
The distinction between the rules that constitute the legal and political system and the rules that do not, can be further specified by Hart’s distinction between primary and secondary rules in legal systems. Primary rules apply to the subjects of the legal system and define permissions, rights and duties. Secondary rules apply to primary rules and define the procedures whereby primary rules are identified, made and changed.Footnote 49 Without secondary rules, there would be no means for the creation of rules and no means for changing them and, therefore, no legal system in the modern sense of the term. Based on these observations, we should define the constitution as the sub-set of legal rules that are second order rules.Footnote 50
The extraordinary demands of democratic legitimacy that follow from the theory of constituent power apply to secondary rules only because they constitute the legal and political system. From this point, it follows that the demands on democratic legitimacy apply to events of constitutional change to the extent that they entail revisions of secondary rules. In some cases, this is clearly not the case, whereas in others, it is somewhat uncertain. Both scenarios are well illustrated by the recent amendment to the Constitution of Ireland in 2015.
The parliament of the Republic of Ireland had proposed a revision of two constitutional provisions: the articles regulating the minimum age of presidential candidates and the articles regulating the legal status of same-sex marriage.Footnote 51 Both propositions were subject to referendum as the constitution stipulated that no amendment be made unless ‘duly approved by the people’.Footnote 52
Though the referendum was required by the constitution of Ireland, the extent to which the referendum was required by the ideal of popular sovereignty is of course an open question. The popular sovereignty argument that is investigated here tells us that referendum is called for only when the rules that are constitutive of the legal and political system are to be changed. Given that only secondary rules are regarded as ‘constitutive’, the question is whether the proposed changes to the Irish constitution pertained to the secondary rules of the Irish political and legal system.
Consider, first, the proposal to lower the age of candidacy for the Irish presidency. It is not immediately clear that this implied a change to the secondary rules of the Republic. Age limits for the presidency do not as such redefine the powers of that office. On the other hand, it could be argued that the age of candidacy does regulate the process of law-making and is therefore a secondary rule. After all, the president of Ireland is involved in law-making and the age of candidacy regulates who can be president. The distinction here is between two different kinds of secondary rule; those determining the powers of institutions (power-conferring rules) and those determining membership in them. The rule that mandates a particular age for candidates for the office of president is a secondary rule of the latter kind.Footnote 53 If it is accepted that rules that determine membership in public offices and institutions are secondary rules, it follows that the proposed amendment of the Irish constitution required democratic legitimacy of the kind that only constitutional referendum can provide.
Secondary rules that pertain to membership in public institutions are regularly found in ordinary legislation, and usually not in the written document that is named ‘the constitution’. Consider, for example, the conditions for the appointment of judges, including judges appointed to a Supreme Court. In the legal system of Sweden, these rules are set out by regular statute.Footnote 54 Following the view that higher standards of democratic legitimacy apply to all secondary rules, we are committed to the conclusion that the relevant sections of the Code of Judicial Procedure should be subject to the same normative standards.
The implication is that there should be a constitutional referendum on any proposed revision of secondary rules that apply to public officials, whether or not these rules are part of the codified constitution. The relevant object of democratic legitimacy from the perspective of popular sovereignty includes what is sometimes called ‘the constitution outside the constitution’.Footnote 55 As argued by Ernest Young, it is exactly because matters of constitutive importance to the legal and political system are often found in non-constitutional parts of the law that fundamental changes in the institutional regime can take place without traces in the written constitution.Footnote 56
The second part of the Irish constitutional referendum concerned the legal status of same-sex marriage. As the proposed revision was approved by a majority of voters, a new section was introduced to Article 41 of the Constitution: ‘Marriage may be contracted in accordance with law by two persons without distinction as to their sex’. The provision implies a duty on behalf of public authorities to recognise marriage ‘without distinction’, thus corresponding to a claim-right on behalf of citizens entailing that authorities allow same-sex couples access to the institution of marriage. Rules that confer claim-rights are primary rules in Hart’s sense, as they impose obligations and rights on individual subjects. Since the prescriptive account of the constitution only includes secondary rules, it follows that the amended provision regulating same-sex marriage should not be regarded as a constitutional rule. The referendum on amendment of the relevant provision was accordingly uncalled for under the doctrine of constituent power.
The argument that the constitution must be legitimate by democratic standards because it constitutes the legal and political system avoids the problem confronted by the status conception of the constitution. The latter view failed to explain why the constitution requires special attention from the point of view of popular sovereignty and constituent power. The fact that constitutions are regularly entrenched is not a sufficient basis to conclude that revisions of constitutional rules mandate direct popular involvement. The functional account of the constitution offers a more promising starting point in that regard since it is easy to see why the constitutive powers of the people should be concerned with the set of laws that are constitutive of the legal and political system.
However, the latter view has its own problems. The main problem lies not with the normative justification of the constitutional referendum but with the application of this justification to the real world of constitutional politics. First, since amendments to the codified constitution are not always concerned with secondary rules, the demand for democratic legitimacy does not always apply to amendments to the constitution. Second, since ordinary legislation frequently includes secondary rules, the demand for democratic legitimacy will also apply to many changes made outside the constitution. In sum, the result is to collapse the popular sovereignty argument for defining what is constitutional, as the very meaning of constitutional is put into question. What turns out to be important from the point of view of democratic legitimacy is not whether legal change is ‘constitutional’, but whether legal change is concerned with the secondary rules of the legal system.
Conclusions
It is a popular notion that the people should be able to participate directly in the process of constitutional amendment. According to the received view, direct participation in the process of ratifying constitutional change is necessary to secure the democratic legitimacy of the constitution. In this paper, I have shown that this view is not supported by theories of popular sovereignty and constituent power because of divergence between the ideal and actual substance of the ‘constitution’. Theories of popular sovereignty are premised on an idealised understanding according to which the constitution literally ‘constitutes’ the legal and political system. It is only because of these functions that direct participation in constitutional amendment is considered a requirement for the democratic legitimacy of the political order. The implication is that the popular sovereignty argument for the constitutional referendum applies only to decisions that modify the constitutive aspects of the legal and political system. Given that codified constitutions encompass norms without constitutive functions, it follows that changes in them do not always require ratification through referendum in order to be legitimate. In addition, norms that are constitutive of the legal system are often part of ordinary legislation, meaning that the popular sovereignty argument does seem to require direct participation in order to authorise changes to them. But this means that democratic legitimacy, as defined by theories of popular sovereignty and constituent power, remains in tension with the practices and ideals of representative democracy and only partly applies to changes to codified constitutions.
