TY - JOUR
AU1 - Trueblood,, Leah
AB - Abstract Referendums are regularly defined as being directly democratic. Indeed, the term ‘direct democracy’ is often used synonymously with referendums. The label ‘direct democracy’ is used to make two different types of claims about referendums: (i) descriptive claims about what referendums are; and (ii) normative claims about how their use is justified. This article challenges the treatment of referendums as devices of direct democracy both in theory and in practice. I argue instead that referendums should be theoretically and practically understood as processes that provide direction to representatives. Indeed, one consequence of my argument is the broader claim that the term ‘direct democracy’ is generally misleading. In a contemporary context, all democratic processes are—at least in part—representative. 1. Introduction Referendums1 are regularly defined as instances of direct democracy. The term ‘direct democracy’ is used to make two different types of claims about referendums. The first way that the term is used is descriptively. This use argues that, as a matter of fact, referendums are not mediated by representatives in material ways. The second way that ‘direct democracy’ is used is normatively. On this view, referendums are justified as processes that metaphorically bypass representatives. This second approach does not seek to describe referendums but, rather, justify them. The aim of this article is to challenge the idea of referendums as directly democratic in both the descriptive and the normative senses. I also put forward an alternative account of referendums as instances of both descriptive and normative representative democracy. Indeed, on my view, the term ‘direct democracy’ is misleading; democracy is itself, at least in part, representative. In arguing for the necessity of representativeness in understanding democracy, I will build on work by Urbinati,2 Plotke3 and Young,4 who all make this kind of claim (albeit to different degrees). This article begins by explaining the meaning(s) of direct democracy and the sources of dissatisfaction with the term as it applies to referendums. Direct democracy is usually defined along the lines of a process that is unmediated by representatives, but writers generally place caveats on the idea that direct democracy is unmediated. The aim of the second part of the article is to consider some different possible meanings of ‘unmediated’ democracy. The third part shows that even on different accounts of the term ‘unmediated’, the term ‘direct democracy’ is not helpful either in practice or in theory. I show first that the aggregation of political will necessarily requires strong mediation by representatives in practice, and that strong mediation is central to the normative case for referendums in theory. The fourth part of the article offers an alternative account of referendums as representative both in theory and in practice. The fifth and final section of the article applies my argument about referendums in particular to democracy in general. I argue that the very idea of direct democracy is unhelpful because it is misleading. Democracy is always mediated in some way. Further, there is no categorical difference between the mediation that takes place in elections5 and that which takes place in referendums. If there is any difference between elections and referendums, a claim about which I am very doubtful, the distinction is one of degree rather than kind. 2. Dissatisfaction with Direct Democracy Despite the ubiquity of the term ‘direct democracy’ in accounts of referendums,6 there is an undercurrent of dissatisfaction with the term. It is very common for books on referendums to begin with the caveat that the term is misleading,7 but nevertheless some still consider it to be a valuable shorthand.8 Terms such as ‘hybrid’9 or ‘quasi-direct’10 democracy are used, but none of these terms have been consistently adopted.11 Part of this dissatisfaction is driven by ambiguity in the term itself. Direct democracy is usually defined as unmediated democracy, but it is not at all clear what ‘unmediated’ means. The term is especially confusing because its use is almost always accompanied by caveats that the argument does not require democracy to be entirely unmediated. Some broader representative context,12 or institutions such as political parties, are required.13 Nevertheless, in some important respects, the label ‘direct democracy’ is used to indicate that referendums can be categorically distinguished from processes such as elections. Sometimes direct and representative democracy are placed on a spectrum,14 and sometimes they are put in different categories altogether.15 Sometimes a combination of both approaches is taken. Schiller and Setälä, for instance, argue that both direct and representative democracy are mediated, but that a distinction can be clearly drawn between these two different types of mediation.16 The subtlest approach is that taken by Uleri, who argues that referendums are reliant on representatives,17 but nevertheless that processes of direct and representative democracy can be ‘analytically and empirically distinguished’.18 I disagree. In this article, I will argue that processes of representative democracy cannot be distinguished either empirically or conceptually from direct democracy. Pace Schiller and Setälä, I will argue that the types of mediating processes in representative democracy cannot be distinguished from those of direct democracy. Empirically, they cannot be distinguished because both referendums and legislation are processes of aggregating political preferences. Further, in both the processes of elections and referendums, voters are voting for particular representatives and giving guidance to representatives. This supposed contrast between referendums and elections may prove a sticking point for my argument. It may be that those who use the label ‘direct democracy’ are prepared to accept parts of my argument and yet they think the word ‘direct’ captures something important. There really is an important distinction to be drawn between processes such as referendums and elections. I am prepared to concede that there may be differences, although I think these differences are a question of degree rather than kind. Mediation is a spectrum, and while I think that both elections and referendums must be understood as requiring strong mediation in both practice and theory, at a minimum I will argue that no sharp line can be drawn between the types of mediation that these democratic processes require.19 The claim made by using the label ‘direct democracy’ is that referendums are, in theory or in practice, unmediated by representatives in some important way. In the next two sections I will argue that referendums are mediated democratic processes in all relevant ways. First, I will show that describing referendums as unmediated democratic processes is either so demanding as to be false or so weak as to be meaningless. Secondly, I will argue that the claim that referendums bypass representatives does not capture what is normatively valuable about both referendums or democracy. 3. The Meaning of Mediation It is not clear what the ‘direct’ in direct democracy refers to. Referendums are not the only processes given this label. Presidentialism, primaries, initiatives and recall20 are also treated as processes of direct democracy.21 What connects these different processes? It seems to be the idea that they are not mediated by representatives in some material way. So, with presidentialism, for instance, the claim is that this process is direct in that the executive is chosen by voters separately rather than being drawn from a legislature. The claim that direct democracy is unmediated may be made in varying degrees of strength. It might be used in a minimal way to say that there is some privileged connection between voters and outcomes, although this minimalist view is so thin that it cannot provide much guidance on the meaning of mediation. At the other end of the spectrum, Barber takes a much stronger view. He argues that strong, direct democracy (the only real kind of democracy) requires the absence of mediation altogether.22 For Barber, writing in the Rousseauian tradition,23 to say a process is unmediated is to say that the people are engaged directly in, among other things, common decision making and common action. It is common to define direct democracy in negative terms. Instead of outlining what mediation requires substantively, the claim is instead made that direct democracy is democracy that does not involve the selection of representatives. What can be drawn from this contrast with representative democracy? There are a range of possibilities. One difference between elections and referendums might be that the content of the preferences being aggregated are different. Referendums are about substantive policy or constitutional questions. In this way, the claim is that referendums are unmediated because representatives do not determine the content of the preferences. These preferences are aggregated from voters directly. A second possibility is that referendums are directly democratic because the form of preference aggregation is different. That is to say, in referendums, preferences are aggregated in a different way, and the scope of consideration is different than when elections are held. It may be, too, that a theorist thinks that a referendum is unmediated on both of these axes. Both the form and the content of preference aggregation in referendums are qualitatively different than in elections, and consequently it is right to say that referendums are unmediated democratic processes. When offering a theoretical rationale for referendums, theorists make an even stronger claim about the meaning of direct democracy with respect to mediation. Their more demanding version of this claim is that voters instead of representatives make decisions. An even stronger version of this claim is that in these processes it is voters in their collective capacity, instead of representatives, who make decisions, and that it is the fact that voters act in this collective capacity that makes the process directly democratic.24 This is the version of direct democracy found in normative arguments for constitutional referendums such as those of Galligan25 and Tierney,26 although, crucially for those authors: (i) a broader representative context is still required for referendums; and (ii) ‘direct democracy’ is the right label only for very specific types of constitution changing referendums.27 This normative argument only holds for certain types of constitutional referendums that concern the allocation of sovereign relations in a democracy.28 Since the people are the ultimate source of authority in a democracy, it must be the case (so the argument goes) that decisions about the allocation of powers should be made by the people directly and together.29 This section has considered a range of meanings of mediation, and some of the practical and normative reasons that theorists might take in viewing referendums as unmediated democratic processes. In the next two sections, I will challenge the use of the label ‘mediation’ with respect to referendums. I will argue that the descriptive use of mediation is either so weak as to be meaningless or so strong as to be misleading. In theory, the idea of ‘unmediated’ democratic processes does not capture what makes either referendums or democracy worthwhile. The better approach, as I will argue in the penultimate section of the article, is to treat referendums both in practice and in theory as processes of representative democracy. 4. Why Referendums are Mediated in Practice While it is not exactly clear what mediation is, it must in some way refer to the aggregation of preferences. The label ‘direct democracy’ is used to identify some closer connection between voters and outcomes than is possible in practices of representative democracy. The claim is that in direct democracy, the content and/or the form of preference aggregation is different than in representative democracy. The case for the use of the label ‘direct democracy’ relies on what is actually on the ballot. The prima facie appeal of this argument is clear. There does initially seem to be something narrower or more specific on the ballot in a referendum as opposed to an election. The aim of this article is not to contest this; rather, it is to argue that this specificity should not be taken to indicate that referendums do not involve the mediation of preferences through representatives. There are a range of ways in which the term ‘mediation’ may be used, and the actual process of mediation takes a number of forms. Barreyro identifies a range of these when she asks: ‘Why are direct democratic institutions called direct if they, as law-making mechanisms[,] are mediated by elites, political parties, and mass media, and shaped by electoral law?’30 I will now consider the significance of some of these actors as mediating forces in referendums. I will argue that referendums should not be described as unmediated democratic exercises in practice for three reasons: first, the role of elites and representatives in shaping and aggregating political will; secondly, the mediation of what voters cast their ballots for in a strong sense, by the availability of choices on the ballot; and thirdly, the broader representative context that is required in order for referendums to operate at all. I argue that this broader representative context can never be decoupled from the referendum process itself, so the label ‘unmediated’ or ‘direct’ is misleading and unhelpful. A. Referendums as Mediated by Elites and Representatives A perennial worry about referendums is that they are merely tools to be manipulated by elites. Walker’s study of referendums, for instance, argues that they are used by elites to achieve their goals,31 and that they are particularly effective political tools for executives to use against legislatures.32 This is the worry expressed by Mendelsohn and Parkin too, who also think that referendums are tools elites use to fulfil their agendas.33 Even those, like Tierney, who are more optimistic about the use of referendums are clear that elites play an essential role in the agenda-setting stage of referendums.34 There is a debate in the social scientific literature, for example, about whether voters are more likely to vote for the status quo, so it matters who is voting ‘for’ or ‘against’ what.35 The role of elites plays out differently depending on the role that referendums perform in a system of constitutional and policy change. In a system where referendums are automatically used to approve constitutional reforms, representatives have a significant impact on the content of what is voted on.36 In ad hoc systems, such as that of the UK,37 the form and content of referendums are determined by the government of the day on a case-by-case basis. Referendums may be abused by representatives, but mediation or intervention by representatives in an election process is not necessarily sinister. As Tierney argues, the better view is that the agenda-setting process is inextricably shaped by representatives.38 This is unavoidable. The voting stage of the referendum process offers voters a limited number of options; representatives take the underlying decisions about what those options are. This mediation process is central to understanding how a referendum operates in practice. Further, the quality of this underlying mediation process – the question of whether proposals are helpful or divisive, for example – sets them up for success. This essential role for representatives is the first reason to challenge the helpfulness of labelling referendums as ‘unmediated’. Further, it is not persuasive to correlate narrowness or specificity with a lack of mediation. B. Ballots as Mediators of Political Will Both the form and the content of preference aggregation in referendums are determined by representatives. Part of the distinction drawn between referendums and elections is that referendums are votes for policies and elections are votes for representatives. Consequently, referendums are not mediated by representatives but elections are. This distinction is unhelpful. Votes in referendums lend credibility to representatives, and representative are elected on the basis of policies. Again, social scientific literature shows that referendum campaigns are shaped by whether voters want to support or send a message to certain political parties or elected officials.39 Those who propose and support the policy are inevitably given legitimacy by successful campaigns. My point is not that referendums are only about voting for representatives; rather, I am saying that there will be an overlap (and that there is necessarily always the potential for an overlap) between voting for propositions and voting for people. To cast the relationship between referendums and elections as voting for either policies or representatives is misleading. Once again, it is fair enough to say that what is on the ballot in a referendum is narrower or more specific than in an election. The idea of unmediated democracy does not follow from narrowness or specificity. Despite the narrowness of what is on the ballot, in practice it is the prerogative of voters that they may vote for any reason they wish (or no reason at all!). Indeed, the more specific the propositions on the ballot, the more mediated referendums are, in that more work has been done to construct political preferences, and the guidance given to representatives about how to execute those preferences is narrower too. The argument that referendums are unmediated because they offer a closer connection between voters and outcomes is also unpersuasive because, even if there could be such a connection (a claim I do not accept), this connection could only possibly hold for those who support the winning proposition. There cannot be a privileged, unmediated connection between a voter and an outcome, if he or she voted against that outcome. There is some relationship between voters who did not support the winning proposition and outcome. There may well be very good reasons to support the outcome, perhaps because of a voter’s general confidence in a democratic system. As Wollheim says, however, this reasoning can only be explained indirectly.40 Supporting an outcome that one does not agree with requires a broader appeal to the principle of democracy as a whole. As soon as this controversial move is made, there is a gap in the connection between voters and outcomes. A democratic process is doing the work of mediating the content of preferences. This claim is both descriptive and normative. It is descriptively the case that there cannot be a ‘direct’ connection between voters and outcomes when they did not support those outcomes. The claim is a normative one because this indirect connection between voters and outcomes better captures the rationale for referendums and democracy. The whole point of democracy is to manage disagreement and to ensure that some people do not get their way all of the time at the expense of others. Nevertheless, democracy gives people an indirect reason to support proposals they disagree with. The indirectness of this justification is foundational to the justification of democracy. C. Referendums as Part of a Larger Democratic Context The idea of unmediated democracy does not capture what is going on in these processes either descriptively or normatively. The whole point of referendums (and of democracy) is mediation in the strong sense. This is because the aim is necessarily to aggregate the views of a range of voters. This process of aggregation is best understood as a strong form of mediation. This is not only because of the essential features of representatives and ballots, discussed above, in determining what is aggregated. Democracy is by definition a decision-making process involving a wide range of actors. In the process of aggregation, compromises and concessions are made even for those who prevail. If there is a direct connection of any kind between one person or group’s preferences and a particular outcome, this is likely to have meant that something has gone very wrong. This is especially because our individual intentions and preferences become decoupled from votes once they are cast and all ballots are, rightly, secretly and anonymously counted. To describe referendums as unmediated is antithetical to their purpose and, I will show shortly, the idea of mediation is central to their justification. I have argued in this section that referendums should not be described as unmediated democratic processes. This is because of the roles of representatives and ballots, and the purpose of democratic processes. Even if all of my challenges to the helpfulness of the label ‘mediation’ are met, however, supporters of the label ‘direct democracy’ may still make a different kind of claim. They may argue that while mediation is necessary descriptively for referendums, it is essential for normative reasons to treat referendums as if they are direct. Put differently: the myth of direct democracy is a useful fiction. Indeed, Barreyro,41 drawing on Kelsen42 and Pitkin,43 makes the argument that representatives themselves acting together on behalf of the community in a legislative body is also myth. Barreyro is right to suggest that this must be the kind of argument that some supporters of the label ‘direct democracy’ are making. Even if Barreyro is right that these are different myths, some types of myths are more useful than others. I will now argue that the claim that referendums are unmediated exercises in democracy is not a helpful myth as it does not capture the rationale for democracy. Further, it mischaracterises deeper democratic ideas, such as voting, popular sovereignty and authority. 5. Why Referendums are Mediated in Theory When the term ‘direct democracy’ is used to make normative claims, the idea is that voters instead of representatives make decisions or, stronger still, that it is the people in their collective capacity,44 instead of representatives, who make decisions. The motivation behind this claim is that there are some decisions that cannot properly be made by representatives. Decisions about the allocation of powers in a democracy, so the argument goes, cannot properly be made by representatives because the power to do so was never transferred to representatives in the first place. It would be ultra vires for representatives to decide about questions such as secession and devolution. These different types of decisions require a different type of legitimacy, and this legitimacy can only be provided by the people directly through a referendum. On this normative defence of direct democracy, decisions cannot be mediated by representatives because the whole point of referendums is that they temporarily supplant the actions of representatives for the purposes of higher-order law making.45 This section will argue that referendums are normatively valuable and defensible precisely because they give direction to representatives. This is what provides referendums with their legitimacy and rationale; representatives cannot be supplanted, even temporarily. To see why this is the case, however, it is necessary to investigate the ideas that inform the treatment of referendums as devices of direct democracy in theory. In particular, it is necessary to challenge the view of legitimate authority, voting and popular sovereignty that informs the use of the label ‘direct democracy’. A. Voting The first reason to think that referendums must be mediated is the nature of voting and what it can achieve. Riker helpfully identifies different conceptions of voting. The first is a populist or Rousseauian view, on which referendums can coherently articulate collective political preferences. This is the approach to voting that informs the theoretical defence of referendums as instances of direct democracy. On this view, it is possible for the people to make decisions both collectively and together, and therefore to bypass representatives. This is not, however, the only way to think about the act of voting. Riker outlines another approach, which is the liberal view. On the liberal or Madisonian approach, voting is a way of limiting the power of representatives.46 The liberal approach to voting argues that voting is a relative expression of preference given a finite set of options, and that the primary aim of expressing these preferences is to limit the power of representatives While voting may well have many purposes and other sources of value, it should not be understood as an act of consent, the embodiment of political will or making a decision directly. One reason that the liberal approach to voting is to be preferred to the Rousseauian version, particularly in the democracies of interest to this thesis, which are large and pluralist, is for social choice reasons—particularly as applied by Riker.47 Riker demonstrates the many different reasons that motivate individual preferences, how quickly mandates for people and propositions can be disaggregated and how seemingly massive coalitions can quickly come apart.48 Indeed, the larger a coalition is, the more likely it is to dissolve.49 Voting is a blunt instrument; as seen above, voters’ intentions necessarily come apart from their ballots when preferences are aggregated. Voters may vote for any reason or for no reason at all. Treating votes as limits on the actions of representatives, as devices of sanction or veto, is a much more persuasive way of thinking about what voting can and should achieve. This liberal conception of voting, however, does not suggest that it is possible for referendums to act as unmediated democratic processes because the whole point of referendums on this view is to provide direction to representatives. B. Popular Sovereignty The two approaches to voting outlined above reflect different approaches to popular sovereignty. It is possible to think of referendums as expressing popular sovereignty, but another way to think of popular sovereignty is as a way of limiting or controlling the actions of representatives. This approach is drawn from the Aristotelian tradition, and Lane’s account of it is especially compelling.50 Lane explains the control thesis as popular sovereignty being the multitude’s power to control office holders.51 She also argues that, contrary to seeing this as a pessimistic way to think about democracy,52 we should see it as a real democratic achievement.53 She is not advocating a reduction of popular sovereignty either historically or normatively to the control thesis alone,54 but she does think that Aristotle offers a compelling and competing vision of what popular sovereignty requires. Loughlin is not in the tradition of Aristotle, but he too argues popular sovereignty is best understood ‘relationally’, focusing on the relationship between voters and representatives.55 Finally, Espejo argues that popular sovereignty should be understood as a process that unfolds over time, as the membership in a political community and their preferences and intentions are fluid.56 Taken together, these accounts point towards a more compelling way of thinking about popular sovereignty. Rather than a static expression of the will of the people, this is an expression of an ongoing accountability relationship between voters and representatives. These approaches avoid many of the problems associated with popular sovereignty as a static concept. These problems include who counts as the people.57 Note, too, the significance of these relationships and their purpose. As Strøm says, ‘In representative systems, the accountability of the representatives is necessary in order to maintain a link between public opinion and collective decisions (responsiveness)’.58 The accountability and responsiveness that democracy seeks to achieve require seeing popular sovereignty as an ongoing process that limits the actions of representatives. The disadvantage of my approach is seeing it not as properly democratic,59 but rather as a weaker, conservative facsimile. I do not see this as a serious problem. Representatives are always necessary in democratic decision making; as Mill argues, this was true even in an Athenian context.60 What is required is a high-quality relationship of responsiveness, where voters are really able to limit the actions of those representatives. The idea of voters as the ultimate source of authority in a democracy does not require voters to act directly and together. An account of legitimate authority that requires the people to intervene directly and together through constitutional referendums is not persuasive. C. Legitimate Authority The normative case for referendums as directly democratic relies on an unpersuasive conception of legitimate authority. Sometimes the argument draws on the Rousseauian61 tradition, sometimes on the Lockean tradition,62 but each of these two different versions of social contract theory have their own problems and limitations. The social contract does not offer the most persuasive way of thinking about legitimate authority. There have long been serious challenges to the possibility of consent as the basis of legitimate authority. Some of the most persuasive arguments against the idea of the social contract were raised by Hume.63 These concerns are familiar: ‘Do people actually consent?’, ‘Can they reasonably be said to have had other options?’, ‘What about future generations?’ and so on. It is not necessary to revisit these arguments in general here. Instead, I will focus on their application to the argument for constitutional referendums in particular. One way that supporters of the social contract as the basis for legitimate authority have avoided counterarguments like Hume’s is to say that the consent is tacit or implicit rather than explicit. On this view, the social contract is a practice rather than an event. The difficulty with making a case for constitutional referendums on the basis of the social contract, however, is that arguments for implicit or tacit consent are not available. The case made by Tierney, Galligan and Bogdanor is that referendums are instances of explicit consent. On their view, it is this explicit consent that gives constitutional referendums their normative force and justifies their use. The difficulty is that because the argument is for explicit consent, this approach to justifying referendums must respond to all of the challenges associated with explicit consent. This is particularly difficult given that, as I will now argue, voting should not be understood as an instance of explicit consent. I do not think it is helpful to characterise voting in a referendum as an instance of consent. Consent is employed by political theorists, and by Tierney in his argument, because of its significant normative force. It is used precisely because it decisively changes the normative relations between two parties.64 It may give rise to obligations. The significance of consent is a double-edged sword. If the result of voting is a normative change between parties, the process of voting must have satisfied exacting standards. There are at least three ways in which I think voting falls short of this high bar. These shortcomings reflect the three parts of consent: who consents, what have they consented to and to whom have they given their consent? Take the first question first: who is consenting? Is it the voters, the people or the majority? Tierney argues that the consenting party is the voters consenting to a process under conditions of deliberative democracy. He does not explain, however, why this subset of a community should be taken to stand for the whole. This is connected to the second problem: what is being consented to? Tierney argues that it is the process that voters agree to, rather than the outcome. I have no problem with this approach, except to say that this is not normatively or politically synonymous with saying that voters consent to an outcome. Just because I participate in a process does not mean I tacitly consent to the legitimacy of an outcome. I may be completely opposed to both choices on the ballot, albeit I find one option less objectionable than the other. I may participate in a process fully in the knowledge that it is deliberately rigged or never expecting my vote to matter. I may spoil my ballot. Tierney may well be right that there is some sort of tacit agreement to participate in the process of voting from casting a ballot, but I do not think this can be said to mean I endorse an outcome, even tacitly. Even if Tierney is right, though, and it is clear who is consenting and what they are consenting to, it is not clear who the other party is. Is it the state? The government? None of these questions are insurmountable, but in making a case for constitutional referendums as instances of explicit consent, Tierney, Bogdanor and Galligan are more vulnerable to these kinds of challenges than supporters of legitimate authority as implicit consent are. Doubtless, voting is an act of great significance, but is it an act of the type of significance that typifies consent? No. Voting is a relative expression of preference given a finite set of options. In order for someone to have consented, it is often argued in other areas of law and philosophy that they must have had some say in the circumstances under which consent is given.65 This is not true of voting in a referendum. Say, for sake of argument, though, that all of these concerns with the idea of consent could be addressed. My larger concern with basing the case for constitutional referendums on consent is that there are much better views of legitimate authority available. Most persuasive of all is the view that legitimate authority is plural or piecemeal. Indeed, this is one of the leading approaches in political and legal theory today.66 The question of whether an authority is legitimate depends on the context, and the answer varies from person to person. In the final section of this article, I make a case for constitutional referendums on the basis of a piecemeal account of legitimate authority. This enables my account to be more responsive to pluralism both in and between legal systems. Basing referendums on a piecemeal and plural account of legitimate authority offers a more nuanced and context-dependent case for their use. Are referendums necessary or helpful in a set of circumstances? It depends. Even if the social contract is the basis of legitimate authority sometimes, this will at most be only a marginal set of cases. It might be right that the case for referendums as directly democratic follows logically from an account of legitimate authority based on the social contract. The problem is that this underlying view of legitimate authority is not itself persuasive. Why should referendums be used? How can their use be justified? These are normative questions which can only be answered by reference to their representative character. This is because it better reflects three ideas that constitute referendums. The normative, representative character of referendums better captures the voting, popular sovereignty and legitimate authority. 6. Referendums as Representative, and Requiring Representatives A vote in a referendum cannot be an instance of direct democracy because a vote in a referendum process cannot itself be a freestanding decision. A decision, Raz says, requires intention and deliberation, but I will argue that voters are not the sort of group that can have such an intention. There is a need for representatives to make decisions, which voters then make choices about. Raz’s account of a fully fledged decision holds that: to decide is to form an intention; decisions are reached as a result of deliberation; decisions are taken sometime before the action; decisions are reasons.67 The first three of Raz’s requirements are very similar to Meiland’s68 and Oldenquist’s.69 On all three of these accounts, the relationship between deliberation and decision making is crucial. In some of the literature on referendums, the terms ‘deliberation’ and ‘decision making’ are treated as nearly synonymous. Meiland, Raz and Oldenquist, however, suggest that this is not the case. Raz also distinguishes between decisions and choices. He says: ‘Not every intention to perform an action is a result of a decision. It is the process of deliberation … which distinguishes decision-based intentions from other intentions.’70 This distinction between decisions and other intentional acts that are not the result of deliberation is helpful for thinking about voting. Raz, Oldenquist and Meiland all think that there is a difference between decisions and other intentional acts such as choices, and they largely agree on what that difference is. Some decisions are choices, but not all choices are decisions. Choices are a less demanding sort of act that does not necessarily require deliberation. Choices may involve deliberation, but they do not rely on deliberation in the way that decisions do. Raz and Oldenquist—in different ways—use the example of Jones and the Martini to illustrate the difference between decisions and choices. If Jones is offered a tray of different kinds of drinks and takes a martini it would be correct to say that he chose the martini but not that he decided to take the martini. If Jones decided to take the martini, then we expect that prior to the action there was deliberation, or at least preference and resolution, and that between the time he decided and the time he acted we could correctly say that he intended to take the martini.71 The example of Jones and the Martini draws out some of the differences between decisions and choices. The same is true of voting. A voter may well have deliberated and decided to vote for the ‘Wolf in Sheep’s Clothing’ party: but not necessarily. Voters need not have deliberated and come to a decision about how to vote. They might have flipped a coin, they might have chosen randomly. This is the prerogative of voters in democracies. If I do decide to vote for a particular candidate or particular party, though, I make that decision individually. How will I vote; in which box will I place an X? Even if I think I have a duty to vote in the best interests of the country, the decision I reach about how to vote is my decision about what I think is best. It is not what my community or family think is best for everyone: although those decisions might overlap, that is not what my choice is when I vote myself. Whatever other processes are happening in a democracy, which may well include a series of deliberative forums, the act of voting itself is a choice that must be decoupled from that deliberative process. Those processes may well involve deliberation about a decision, but this is not deliberation to make a decision. Those who argue that referendums are instances of collective, direct decision making are right to argue that deliberation is a necessary response to disagreement, and a necessary feature of democratic decision making. Where this argument goes astray is with the problem of large numbers and the act of voting itself. These are problems that deliberative democrats are well aware of. Tierney argues: The kind of discursive deliberation that can take place within a small group is not possible, at least not in the same way, at the macro level. But still there is the possibility that people can be encouraged to reflect on the issues, to learn more about them, and then to engage in a more attenuated national process of communicative deliberation.72 The question is what exactly Tierney means by a national process of communicative deliberation. I have no quarrel with this view if this attenuated national deliberation is about a decision, but that does not seem to be either Tierney’s view in particular or the view of deliberative democrats in general. His view seems to be that this deliberation is a necessary part of how citizens come together to make a decision. It may well be that voters individually deliberate before they cast their ballots, and that they will vote is an additional reason to deliberate it, but if that deliberation exists, it precedes a series of individual choices that are aggregated; not a collective choice. The electorate is not the right kind of group, and voting the right kind of action, for the requisite deliberation to occur for decision making. The electorate is not the right kind of group because it does not have (i) a clear membership where (ii) the meaning of that membership is uncontested. Voters might, for instance, come to the voting process with different views about whether a referendum or election is rigged. They may well participate in an attenuated process of communicative deliberation, as Tierney suggests, but not necessarily. Given that deliberation is necessary for decision making, and that it is not necessarily the case that voters participate in this collective deliberation process, I am not persuaded that a vote should be understood as an instance of collective decision making. Voting may well be an individual decision, but I do not think it can be a collective one. Instead, for large groups, representatives must deliberate on behalf of the group and supply this joint intention. The distinction between choices and decisions better reflects the options and opportunities voters have when they vote. When I vote, I cannot propose the candidate I wish were on the ballot, or answer the referendum question I think ought to have been asked. I am not deliberating from first principles; I am making a choice given the options available to me. These choices are some of the most important actions a person can take, so to call them choices is not to argue they are insignificant. To call them choices is to take seriously the limitations of the act of voting, and the impossibility of seeing it as a direct democratic process. A. Referendums as Representative Understanding referendums as representative processes requires a more expansive account of representation. It requires decoupling the idea of representativeness from that of election. As Pitkin says, despite the popularity of this approach, an expansive account of representation is the better view.73 Those who are not elected—monarchs, judges and civil servants—can stand for and act for others in a range of ways.74 They may be symbolically75 or descriptively76 representative. Representation and mediation are not synonymous; an actor may purport to represent a community without ever being responsive to others or mediating political preferences. The claim instead goes like this: referendums require two representatives to perform two types of mediating roles. The first is representation within a referendum process itself. This is the mediation discussed in the previous section about the agenda-setting stage of a referendum, the process of the referendum itself, what voters are voting for and what is on the ballot. The second type of mediation that requires representatives is bridging representation. This connects referendums to an overall democratic system. I will consider the practical importance of each of these types of representation, before turning to their normative significance. Preferences are not aggregated in a vacuum. There are three reasons why representatives are required to connect referendums with the overall process of representative democracy. The first is that representatives are required to maintain the coherence of decisions. This is to address a situation where voters choose both to eliminate tax rises in 2000 and to double government spending in 2005.77 Secondly, there is a need to implement the decisions that are taken by referendum. There is no point in having a decision unless it is executed, and this is not something that voters can do themselves. All political questions are necessarily indeterminate. This means that executing decisions requires judgment; it is not just a mechanical process. Finally, and also because of this indeterminacy, there is a need to adjudicate the meaning of voters’ choices. Judges are required to address this sort of indeterminacy. Taken together, the need for coherence between decisions, as well as the execution and implementation of decisions, mean that referendums rely on representatives to connect them to the representative democracies upon which they depend. Decisions can never be fully determinative in a referendum process. Decisions need to be interpreted and executed. Any democratic decision is an ongoing process rather than a static event. The connection between referendums and other actors, such as judges and civil servants, is not peripheral to the referendum process; it is better understood as part of the referendum process. As Tierney rightly argues, a referendum is more than the voting stage.78 It is a series of processes which include a variety of types of deliberation and disagreement. These cannot be separated from the referendum process itself because the outcomes are what people are voting for. One cannot say anything about the connection between outcomes and preferences without taking a broader view of referendums. This section has drawn on work by Pitkin to argue that the idea of representation should be understood expansively. Representation should not be taken in the narrow sense, to mean that I elect someone to act on my behalf. There are a range of ways in which actors and institutions can act on others’ behalf. This is both because they are elected, and also because they aid in the processes of aggregating, articulating, interpreting and executing political will. These processes should also be understood as mediation in a strong sense. They are essential to both the form and the content of the aggregation of political will, and they shape the connections between voters and outcomes. Above all, representatives make decisions which voters then make choices about in the voting stage of a referendum. This is not only a description of how referendums work in practice, this necessary relationship between voters and representatives captures what is good about referendums too. 7. The Necessity of Representative Democracy The language used to describe and to define referendums matters. It both reflects and informs the way that referendums are understood. To add the prefix ‘representative’ to democracy suggests that it is possible to have democracy without representatives and a system of representation. I argue that representativeness is a necessary part of the meaning of democracy itself, so the word ‘representative’ in representative democracy is misleading. The idea of unmediated democracy is unhelpful because the purpose of democracy is to mediate political will. This is the case because it is about finding ways to mediate and combine different peoples’ preferences and interests. This is no less true in a referendum than it is in other democratic processes. Contemporary democracy is by definition representative because it requires some to act on behalf of others in the service of aggregating, interpreting and executing preferences. As Plotke argues, ‘representation is crucial in constituting democratic practices. [Democracy] requires procedures for taking decisions, and there have to be ways of sustaining those decisions over time.’79 The aim of voting in a democracy is to clarify voters’ preferences; these preferences are necessarily held in the context of others’ preferences. Contemporary democracy is defined by its indirectness, even with respect to referendums. 8. Two Challenges to Referendums as Representative In this section, I will focus on two ways in which a reader might find the argument of this article unpersuasive. First is the familiar claim that referendums cannot be theorised in general terms. There is nothing that can be said about their use either practically or theoretically in general. While my argument might hold for some systems and some types of democratic processes, it is far too much to say that it holds in general. The second challenge is that while representatives are a necessary feature of all democratic processes, there nevertheless remains something direct about referendums. Democracy requires people to constantly act for each other; that is what makes it valuable. A. The Case against Generality In this article, I have made a general claim about the relationship between representative and direct democracy.80 Even more controversially, I have made a general empirical claim about the aggregation of political will in practice and the necessity of representatives in that process. Political scientists in particular may dispute this approach. They may argue that it is unhelpful to theorise about referendums in general, and it is just not possible to make claims about their relationship to representative democracy that hold universally. The argument that there is nothing that can be concluded about referendums is a familiar one, not least because the term ‘referendum’ itself is not used consistently, so it not always clear what kind of process is being referred to in the literature.81 It is right to say that there are significant differences in the uses of referendums around the world. These different uses are significant for the relationship between referendums and representative democracy. Uleri and Gallagher rightly argue, for instance, that it makes a great deal of difference to the relationship between referendums and representative democracy whether referendums are: (i) constitutionally required or optional; (ii) binding or advisory; and (iii) used to begin or conclude processes of constitutional change.82 Gallagher and Uleri are right to argue that, in some ways, the process of mediating political will is different in referendums that initiate rather than approve constitutional change. I am sympathetic, too, to the view that it is impossible to speak about the relationships between representatives and referendums in terms that capture the deep integration of referendums into the democratic framework, such as that of Switzerland, and the ad hoc use on a small set of issues, as occurs in the UK. Nevertheless, I maintain that representation (defined as strong mediation) is necessary for the use of referendums both in theory and in practice. This does not mean, or require arguing, that all uses of referendums are the same. The claim is not that the decisional mechanisms work the same in all jurisdictions. They clearly do not. These mediation processes work differently in different jurisdictions. The claim is that, in different ways, representatives are an indispensable part of the aggregation of both the form and the content of political will. That process of aggregation necessarily involves mediation in a strong sense, so it is unhelpful to characterise referendums as instances of direct democracy. There is no connection between voters and outcomes that does not run directly through representatives. B. Direct and Specific Referendums A reader might be sympathetic to my argument but nevertheless feel that there is something missing. A reader might still have the intuition that there is something direct and different about referendums as opposed to elections. I think it is better to find another word to express this intuition rather than to try to reclaim the ‘direct’ in direct democracy. This is because of the risks inherent in the use of word ‘direct’. To say that something is direct is to say that it is a shortcut or a way around something else: a direct flight or telephone line, for instance. This is because the term ‘direct democracy’ has been misused to suggest that referendums are a way around representatives. As argued above, representatives are necessary for the mediation of political will through referendums both in practice and in theory. Nevertheless, if the question on the ballot is narrowly construed, then there remains something direct about direct democracy. My alternative is to use the word ‘specific’ as a compromise instead of direct. The word ‘specific’ captures that the scope of direction given to representatives in referendums is likely, albeit not necessarily, narrower than it is in elections.83 The word is also helpful in that it captures that, while in certain circumstances the content of direction given to representatives is specific, it is nevertheless direction. 9. Conclusion: A Case against ‘Pure’ Democracy There is an unhelpful line of thought that equates direct democracy to pure democracy, and considers representative democracy as second best.84 This is unhelpful for both practical and theoretical reasons. It suggests that, as Urbinati says, mediation involves a weakening of self-government.85 The aim of this article has been to argue that, irrespective of whether the label ‘direct democracy' is used descriptively or normatively, it is misleading. In practice, the use of the label ‘unmediated’ is either so weak as to be meaningless, or so strong as to be misleading. In theory, the idea of unmediated democracy does not capture what makes referendums worthwhile in practice and in theory. The better approach is to treat referendums both in practice and in theory as processes of representative democracy. This is a significant change in approach, considering that referendums are sometimes used to typify direct democracy. Nevertheless, such a shift to seeing referendums as representative democracy is both possible and important. It will require defining representation expansively, and a greater recognition of representation as a prerequisite, and necessary conduit, for self-government too. The use of direct democracy has focused too much on the idea of ‘self’ and too little on the idea of ‘government’, which is necessary for mediation, democracy and referendums. Many thanks to Les Green, John Gardner, Stephen Tierney, Aileen Kavanagh, Nick Barber, Hayley J Hooper, Alison Young, Ron Levy, the participants in the Deliberative Governance and Law workshop at the University of Oxford, and anonymous reviewers at the OJLS for invaluable comments on previous drafts. Thanks, too, to the Pierre Elliott Trudeau Foundation and the British Academy for funding in support of this research. I am responsible for any errors that remain. Footnotes 1 The use of the term ‘referendum’ is inconsistent in the literature. Sometimes it is used as an umbrella term to cover a range of voting processes including, for instance, (i) votes initiated by the electorate and (ii) authoritarian uses of votes by dictators (which are sometimes called plebiscites instead). In one way, referendums are defined narrowly for the purposes of this article. I take Qvortrup’s orthodox approach to referendums as those processes that ‘refer questions back’ to voters (drawing on the Latin word ‘referre’, meaning ‘to refer back’); he classifies processes initiated by voters as a different democratic process altogether. Matt Qvortrup (ed), Referendums Around the World: The Continued Growth of Direct Democracy (Palgrave MacMillan 2014) 2. In another way, referendums are defined broadly in this article. No distinction is drawn between referendums and plebiscites because part of the argument of this article is that the democratic credentials of referendums vary, and it is neither possible nor helpful to draw a sharp line between those referendums that are democratic and those that are not. Authoritarian uses of referendums do, however, serve as an important reminder that, because referendums give direction to representatives, representatives can use and abuse the legitimacy of referendums to support their own ends. As Walker rightly argues, ‘political actors use referendums to achieve their goals’: Matt Walker, The Strategic Use of Referendums: Power, Legitimacy, and Democracy (Springer 2003) 1. 10 ‘Again, there doesn’t seem to be a standard word to describe the distinctive character of their innovation. So please forgive me when I call the Federalist ratifying convention an exercise in quasi-direct democracy’: Bruce Ackerman, We the People, Volume 2: Transformations (Belknap Press 1998) 83 (original emphasis). 11 Accounts of hybrid democracy focus instead on empirical accounts of specific jurisdictions. 12 ‘The referendum, properly structured, can encourage the fuller engagement of citizens and civil society in democratic processes within the broader functioning of representative democracy’: Tierney (n 6) 299; original emphasis. 13 ‘It is only realistic therefore to envisage direct democracies retaining some of the institutional features of the representative democracy from which they have evolved. These would include not only parties—which are intrinsic to modern politics—but governments and Parliaments too’: Ian Budge, The New Challenge of Direct Democracy (Polity Press 1996) 181. Budge also recognises the need for what he calls moderators (38). 14 ‘Participation and representation not as two alternative forms of democracy but on a continuum’: Urbinati (n 2) 759. 15 ‘Therefore even the categorisation of legislative referendums as an instance of direct democracy as opposed to representative democracy is perhaps something of an oversimplification. In playing a role within a broader representative system of government where the authority of these referendums is clearly subsidiary to those constitutional institutions which retain the competence to create and change hierarchical rules, they should perhaps more accurately be portrayed as part of being part of that representative system, since the effect given to the outcomes they produce is ultimately subject to the representative competence of constitutional institutions, most obviously legislatures, but also courts’: Stephen Tierney, ‘Constitutional Referendums: A Theoretical Enquiry’ (2009) 72 MLR 360, 364. 16 ‘It is possible to make a distinction between the mediation of people’s opinions in decision-making through the institutions and organizations of representative democracy on the one hand and through the institutions of direct democracy on the other’: Schiller and Setälä (n 7) 4. 17 ‘The referendums phenomenon considered as a decision-making process is not an autonomous and independent phenomenon’: Pier Uleri, ‘Introduction’ in Michael Gallagher and Pier Uleri (eds), Referendums in Europe (Macmillan 1996) 2. 18 ‘The two sets of decisional mechanisms and processes can be analytically and empirically distinguished’: ibid 2. 19 Urbinati captures this point when she says: ‘As for the character of democratic politics, focus on deliberation allows us to perceive participation and representation not as two alternative forms of democracy but as related forms constituting the continuum of political action in modern democracies’ Urbinati (n 2) 759. 20 Arend Lijphart, Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries (Yale UP 1984) 198. 21 While my argument is narrower in scope, making a case only that referendums are instances of representative democracy, I do think that my argument applies to these other processes too: that at root these processes are about mediation of political will in a strong sense, and providing direction to representatives, and consequently cannot be exercises in direct democracy on any definition. 22 ‘Strong democracy requires unmediated self-government by an engaged citizenry. It requires institutions that will involve individuals at both the neighborhood and the national level in common talk, common decision-making and political judgment, and common action’: Benjamin Barber, Strong Democracy: Participatory Politics for a New Age (University of California Press 1984) 261. 23 Barber cites Rousseau as saying ‘Once a people permits itself it be represented it is no longer free’ (ibid 261), but questions whether he is right to place himself in this tradition. Rousseau instead refers to actors called ‘agents’: Rousseau says: ‘Hence, the deputies of the people are not and cannot be its representatives; they are merely its agents, who cannot conclude anything definitively’ (Jean-Jacques Rousseau, Of the Social Contract and Other Political Writings (Penguin 2012) 92), suggesting that mediation is necessary, but that actors who perform this role should not be called representatives. 24 Madison says that what makes direct democracy direct is that The People exercise their ‘collective capacity’ to govern. ‘It is clear that the principle of representation as neither unknown to the ancient nor wholly overlooked in their political constitutions. The true distinction between these and the American governments, lies IN THE TOTAL EXCLUSION OF THE PEOPLE, IN THEIR COLLECTIVE CAPACITY, from any share in the LATTER, and not in the TOTAL EXCLUSION OF THE REPRESENTATIVES OF THE PEOPLE from the administration of the former’: Alexander Hamilton, James Madison and John Jay, The Federalist Papers (Penguin Classics 1987). 25 ‘Constitutional referendums are also fundamentally significant as a legitimating mechanism for democratic government. Through collectively deciding upon the system of government under which they love, the people express their democratic sovereignty. In so doing they constitute themselves as the sovereignty body that makes authoritative decisions about their system of government’: Brian Galligan, ‘Amending Constitutions through the Referendum Device’ in Matthew Mendelsohn and Andrew Parkin (eds), Referendum Democracy: Citizens, Elites, and Deliberation in Referendum Campaigns (Palgrave 2001) 109. 26 Referendums are a ‘means by which representation is supplanted, returning direct power to the people; in other words as some kind of reversal of the original act of transference, or at least a temporary return of power to the people’: Tierney (n 15) 367. 27 ‘Even the categorisation of legislative referendums as an instance of direct as opposed to representative democracy is perhaps something of an oversimplification. In playing a role within a broader representative system of government, where the authority of these referendums is clearly subsidiary to those constitutional institutions which retain the competence to create and change higher-order rules, they should perhaps more accurately be portrayed as being part of that representative system, since the effect given to the outcomes they produce is ultimately subject to the representative competence of constitutional institutions, most obviously legislatures, but also courts. In contrast, the people’s direct democratic capacity to act as, or at least to influence the location and distribution of, the supreme source of constitutional law within a polity distinguishes constitutional referendums as, potentially at least, true conduits of popular determination.' Tierney (n 15) 364 (original emphasis). 28 ‘There are four ways in which referendums are used today at the constitutional level: to found new states; to create or radically change constitutions; to establish complex new models of sub-state autonomy; and to transfer sovereign powers from the state to international institutions. What connects these four types of referendums is that they each address the location or distribution of ultimate lawful authority within the polity and in doing so, implicate the sovereign relations between people and government’: Tierney (n 15) 360. 29 ‘Constitutional referendums are also fundamentally significant as a legitimating mechanism for democratic government. Through collectively deciding upon the system of government under which they live, the people express their democratic sovereignty. In doing so they constitute themselves as the sovereign body that makes authoritative decisions about their system of government…Constitutional referendums, while also being instruments of direct democracy, have as their fundamental purpose the originating and changing of the system of government’: Galligan (n 25) 109–11. 30 Her answer to this question, as will be discussed in detail later in the article, is that the debate between representative and direct democracy is between two different fictions: the people in their collective capacity and representatives acting on behalf of the community. María Emilia Barreyro, ‘“Direct Democratic Institutions”: Direct and Democratic’ (2019) 10 Jurisprudence 313, 315 (original emphasis). 31 ‘Referendums are part of a bargaining process between elites who have their basis of power in different institutional settings’: Walker (n 1) 3. 32 ‘Executives can better position themselves than legislatures along a policy spectrum to win when introducing referendums’: Walker (n 1) 3. 33 Matthew Mendelsohn and Andrew Parkin, ‘Introduction’ in Matthew Mendelsohn and Andrew Parkin (eds), Referendum Democracy: Citizens, Elites and Deliberation in Referendum Campaigns (Palgrave Macmillan 2001) 3. 34 Tierney (n 6). 35 Stephen Fisher and Alan Renwick, ‘Do People Tend to Vote against Change in Referendums?’ (UCL Constitution Unit, 22 June 2016) <https://constitution-unit.com/2016/06/22/do-people-tend-to-vote-against-change-in-referendums/> accessed 1 November 2019. 36 In Australia, for instance, s 128 of the Constitution holds that changes to the Constitution must be passed by an absolute majority in both Houses of Parliament. Once passed by both Houses, it is then put to a referendum. 37 The Political Parties, Elections and Referendums Act, s 101 holds that enabling legislation is required for referendums, but the form of that enabling legislation varies on a case-by-case basis. 38 Stephen Tierney, Constitutional Referendums: The Theory and Practice of Republican Deliberation (OUP 2012). 55. 39 ‘Risk assessments, emotional reactions to EU membership, and leader image heuristics were other major proximate forces driving the choices voters made’: Harold Clarke, Matthew Goodwin and Paul Whiteley, ‘Why Britain Voted for Brexit: An Individual-Level Analysis of the 2016 Referendum Vote’ (2017) 70 Parliamentary Affairs 439, 453–4. 40 ‘Now I think it should be clear that my suggestion, if accepted, would resolve our paradox by the only mean still available to us, ie by showing that the two limbs of the conjunction are not inconsistent. For—to return to the example—“A ought to be enacted” is asserted by the citizen who has been outvoted as a direct principle, whereas “B ought to be enacted” is asserted by him as a derivation from an oblique principle, ie the principle of Democracy’: Richard Wollheim, ‘A Paradox in the Theory of Democracy’ in Donald Weiss, Peter Laslett and WG Runciman (eds), Philosophy, Politics, and Society (Blackwell 1962) 86. 41 ‘We distinguished public indirect law making, which relies on the fiction that representatives act as organs of the community, from public direct law making, which relies on the fiction that the electorate act as the community itself’: Bareyro (n 30) 333; original emphasis. 42 Hans Kelsen, General Theory of Law and State (Harvard UP 1949) 289. 43 Hanna Pitkin, The Concept of Representation (University of California Press 1967) 86. 44 Hamilton, Madison and Jay (n 24). 45 ‘The principal argument is that constitutional referendums can serve to unsettle the traditional balance between constituent power and constitutional form in the contemporary polity, substituting the people for the representational role traditionally played by the democratic constitution. The supplanting of representative constitutionalism is a very different function from that performed by ordinary referendums which, in merely replacing the legislature in ordinary law-making, carry no real implications for constitutional supremacy’: Tierney (n 15) 361. 46 William Riker, Liberalism Against Populism (Freeman and Company 1982) 12. 47 It might be countered that the concerns of social choice theorists doom liberal advocates of voting as well as populist ones. Not so. The reason that the social choice arguments do not damage the liberal arguments in the same way is that the liberal arguments make different claims about what voting can achieve and is for. For liberals, voting is an imperfect tool that is part of a democratic machine. Social choice arguments do create some problems for the liberal conception of voting, namely how votes can be used to give direction to representatives if the meaning of those votes is necessarily unclear. The difference is that by treating votes as a tool to guide representatives, the necessity for looking beyond votes to understand voters’ reasons, conducting ongoing consultations, etc is clear. If voting is an embodiment of the will of the people, then the dangers are of inflexibility and treating votes on matters of policy static and finite expressions of preference, rather than an ongoing process of providing direction to those who act on their behalf. 48 Riker (n 46) 233. 49 Riker (n 46) 214. 50 Melissa Lane, ‘Popular Sovereignty as Control of Office-Holders: Aristotle on Greek Democracy’ in Richard Bourke and Quentin Skinner (eds), Popular Sovereignty in Historical Perspective (CUP 2016). 51 ibid 72. 52 Some people think that Schumpeter's approach, for example, is cynical: ‘The democratic method is that institutional arrangement for arriving at political decisions in which individuals acquire the power to decide by means of competitive struggle for the people’s vote.’ Joseph Schumpeter, Capitalism, Socialism, and Democracy (George Allen & Unwin Ltd 1943) 269. 53 Lane (n 50) 72. 54 Lane (n 50) 71. 55 ‘Power thus resides neither in “the people” nor in the constituted authorities: it exists in the reality established between constitutional imagination and action’: Martin Loughlin, ‘On Constituent Power’ in Michael Dowdle and Michael Wilkinson (eds), Constitutionalism beyond Liberalism (CUP 2017) 171. 56 ‘I will argue that we should conceive of the people as a process rather than as an aggregation of individuals. Doing so, I argue, dissolves the problem of legitimizing rule democratically as the compositions of the people changes. The people conceived as a process can be a source of democratic legitimacy that moors state institutions but is compatible with surprise and innovation’: Paulina Espejo, The Time of Popular Sovereignty: Process and the Democratic State (Pennsylvania State University Press 2011) 3. 57 Frederick Whelan, ‘Prologue: Democratic Theory and the Boundary Problem’ (1983) 25 Nomos 13. 58 Kaare Strøm, ‘Delegation and Accountability in Parliamentary Democracies’ (2000) 37 European Journal of Political Research 261. 59 Lane (n 50). 60 In the Athenian democracy, in the time of its most complete ascendency, the popular Ecclesia could pass Psephisms (mostly decrees on single matters of policy), but so-called laws could only be made or altered by a different and less numerous body, renewed annually, called the Nomothate, whose duty it also was to review the whole body of laws, and keep them consistent with one another. JS Mill, Considerations on Representative Government (OUP 1991) 245. 61 Barber (n 22) 261. 62 Galligan (n 25); Vernon Bogdanor, The People and the Party System (CUP 1981). 63 ‘It is in vain to say, that all governments are, or should be, at first, founded on popular consent, as much as the necessity of human affairs will admit. This favours entirely my pretension. I maintain, that human affairs will never admit of this consent, seldom of the appearance of it; but that conquest or usurpation, that is, in plain terms, force, by dissolving the ancient governments, is the origin of almost all the new ones which were ever established in the world. And that in the few cases where consent may seem to have taken place, it was commonly so irregular, so confined, or so much intermixed either with fraud or violence, that it cannot have any great authority. My intention here is not to exclude the consent of the people from being one just foundation of government where it has a place. It is surely the best and most sacred of any. I only pretend, that it has very seldom had place in any degree, and never almost in its full extent; and that, therefore, some other foundation of government must also be admitted … Can we seriously say, that a poor peasant or artisan has a free choice to leave his country, when he knows no foreign language or manners, and lives, from day to day, by the small wages which he acquires? We may as well assert that a man, by remaining in a vessel, freely consents to the dominion of the master; though he was carried on board while asleep, and must leap into the ocean and perish, the moment he leaves her’: David Hume, Essays Moral, Political, and Literary (Eugene F Miller ed, Liberty 1777) Essay XII, 465. 64 ‘In the contexts in which it is appropriately given, it may—and usually does –transform the normative expectations that hold between people and groups’: John Kleinig, ‘The Nature of Consent’ in Franklin Miller and Alan Wertheimer (eds), The Ethics of Consent: Theory and Practice (OUP 2009) 2. 65 Brant McGee, ‘The Community Referendum: Participatory Democracy and the Right to Free, Prior and Informed Consent to Development’ 27 Berkeley Journal of International Law 570, 610. 66 ‘The test is … does following the authority’s instructions improve conformity with reason? For every person the question has to be asked afresh, and for every eon it has to be asked in a manner that admits of various qualifications. An expert pharmacologist may not be subject to the authority of the government in matters of the safety of drugs, an inhabitant of a little village by a river may not be subject to its authority in matters of navigation and conservation of the river by the banks of which he has spent all his life’: Joseph Raz, The Morality of Freedom (OUP 1988) 74. 67 Joseph Raz, ‘Reasons for Action, Decisions and Norms’ (1975) 84 Mind 481, 488–9. 68 ‘To decide is to form an intention to do after deliberation takes place’: Jack W Meiland, The Nature of Intention (Methuen 1970) 65. 69 Oldenquist also says that decisions are reasons in Andrew Oldenquist, ‘Choosing, Deciding, and Doing’ in Paul Edwards (ed), The Encyclopedia of Philosophy (Collier-MacMillan 1967) 99. 70 Raz (n 67) 489. 71 Raz (n 67) (original emphasis); ‘In the preceding example Jones did not decide to take a martini because decision is more closely tied to deliberation and intention than is choice’: Oldenquist (n 69) 98. 72 Tierney (n 6) 213. 73 ‘No doubt the contemporary popularity of the concept depends much upon its having become linked with the idea of democracy, as well as with ideas of liberty and justice. Yet through much of their history both the concept and the practice of representation have had little to do with democracy or liberty. Representation need not mean representative government. A king can represent a nation as can an ambassador. Any public official can sometimes represent the state. Thus, institutions are practices which embody some kind of representation are necessary in any large and articulated society, and need have nothing to do with popular self-government’: Pitkin (n 43) 2. 74 Pitkin (n 43). 75 Pitkin (n 43) 11. 76 Pitkin (n 43). 77 The challenges of the imitative system in California are a good example of this and the gridlock created by directing the legislature both to X and not to X. Peter Schrag, Paradise Lost: California’s Experience, America’s Future (University of California Press 2004) 188. 78 ‘Thus we need to envisage the referendum process as a longer and more complex democratic engagement than it is often considered to be. It can in fact be seen as a series of stages’: Tierney (n 6) 51. 79 Plotke (n 3) (original emphasis). 80 ‘Lijphart concludes: “we are forced to agree with Butler and Ranney that referendums fail to fit any clear universal pattern”’: Uleri (n 17) 2. 81 ‘The “referendum” label includes a variety of situations and usages which bear only a superficial similarity to one another”’: Gordon Smith, ‘The Functional Properties of the Referendum’ (1976) 4 European Journal of Political Research 1, 294. 82 Uleri (n 17) 6. 83 While I am prepared to make this concession, it is not the case that referendums are necessarily narrower than elections. Consider a referendum to choose a new constitution, for instance: the scope of that choice is in some respects far broader than electing a specific representative for a set period of time. 84 ‘These days, referendums are more likely to be seen as pure democratic devices’: Qvortrup (n 1) 3. 85 ‘Representation has been associated with a weakening of self-government. For democrats in particular, it has held little appeal, first because it is seen as justifying a vertical relation between citizens and the state, and second because it is seen as promoting a passive citizenry’: Urbinati (n 2) 759. 2 ‘My intention is not to put into question the normative value of direct participation but to argue for the relevance of representation’: Nadia Urbinati, ‘Representation as Advocacy: A Study of Democratic Deliberation’ (2000) 28 Political Theory 758, 759. 3 ‘Democratic politics is constituted partly through representation’: David Plotke, ‘Representation Is Democracy’ (1997) 4 Constellations 31 (original emphasis). 4 ‘Political representation is both necessary and desirable’: Iris Marion Young, ‘Deferring Group Representation’ (1997) 39 Nomos 349–376. 5 A reader might wonder why the contrast between referendums and elections is emphasised, as opposed to, for instance, that between referendums and the process of legislating. The reason is that while both the process of referendums and legislating are mediated in some of the same ways, they are also mediated in materially different ways. In legislatures, at least in theory, there are substantive opportunities for voters to modify and amend the content of what they vote on. This is not the case for either referendums or elections, where voters are only able to respond to choices put to them, so the same type of mediation is at stake. 6 Sometimes the terms are even used synonymously, as at one point in Stephen Tierney, Constitutional Referendums: The Theory and Practice of Republican Deliberation (OUP 2012) 11. 7 ‘Although widely used, the term “direct democracy” may be considered misleading because all current democracies require a system of representation, and direct democracy is not a feasible system of government in modern societies. However, “direct democracy” can be used to refer to procedures that allow citizens to raise issues on the decision-making agenda without the mediation or parliamentary actors (an initiative), or to vote on a particular political issue to be decided (a referendum)’: Theo Schiller and Maija Setälä, Referendums and Representative Democracy: Responsiveness, Accountability, and Deliberation (Routledge 2009) 4. 8 ‘Finally, one semantic note bears repeating at this point. The phrase “direct democracy” is often used as a substitute for “referendums” or for “semi-direct democracy” in political science writing. Strictly speaking this is inappropriate because direct democracy refers to ideal democratic forms in which all citizens vote on all decisions. Switzerland’s is technically a system of semi-direct democracy. However, the use of the term “direct democracy” is forgivable because there are few appropriate synonyms for “referendums” and “semi-direct democracy” is a rather cumbersome phrase”: Kris Kobach, The Referendum: Direct Democracy in Switzerland (Dartmouth Publishing Company 1993) 58. 9 ‘Most Americans live in a “hybrid democracy.” That is, they live where processes of both direct democracy and representative democracy coexist as part of state government’: Shaun Bowler and Amihai Glazer (eds), Direct Democracy’s Impact on American Political Institutions (Palgrave 2008) 1; Elizabeth Garrett, ‘Hybrid Democracy’ (2005) 73 Geo Wash L Rev 1097. © The Author(s) 2020. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Are Referendums Directly Democratic?
JF - Oxford Journal of Legal Studies
DO - 10.1093/ojls/gqaa020
DA - 2020-09-01
UR - https://www.deepdyve.com/lp/oxford-university-press/are-referendums-directly-democratic-kmh3gqHQkG
SP - 425
EP - 448
VL - 40
IS - 3
DP - DeepDyve
ER -