The boundary problem is a classic issue in democratic theory. It holds that, whatever the theory of legitimacy that we believe our preferred democratic decision procedure should meet, the initial act of constituting the demos can never be met by it (Reference MillerMiller 2009; Reference WhelanWhelan 1983). The problem is that the constitution of the demos is itself a political act with wide-ranging effects. Ordinarily, political acts with wide-ranging effects are subject to a democratic legitimacy demand. So we would need a democratic procedure to legitimate the decision over who should be included in the demos—a demos-constituting procedure, if you will.Footnote 1 But a demos-constituting procedure cannot take place without first determining who ought to participate in that procedure. Given that this prior procedure (the procedure constituting who will participate in the demos-constituting procedure) is also political, with wide-ranging effects, it too needs to be legitimated by a procedure of its own ad infinitum. There is a problem of regression.
Some contemporary political theorists hold that the boundary problem can be solved by changing our conception of the democratic people from a relatively fixed and permanent body of persons grouped territorially and living under the laws of a single, unitary state to a demos including all persons bearing a certain “functional”Footnote 2 relationship to the decision that is to be taken, for instance being “affected” by that decision (e.g., Reference GoodinGoodin 2007) or being “subject to coercion” by it (e.g., Reference AbizadehAbizadeh 2008). If there is an independent standard that allows one to calculate the appropriate membership of a putative demos, then the boundary problem collapses—which persons to include and which persons to exclude from a democratic body, such views argue, can be objectively determined and therefore need not be subject to a democratic procedure to be legitimate. Let us call such sorts of “solutions” functional demos views.
Other theorists think that the boundary problem must be solved by recourse to a global democracy (e.g., Reference AgnéAgné 2010; Reference Arrhenius and FolkeArrhenius 2005: 22; Reference VerschoorVerschoor 2018: 15; cf. Reference NiliNili 2017; Reference SchafferSchaffer 2012). If all should participate equally in a global democratic process, then the question of which persons a democratic community includes, and which it excludes, is moot. All persons ought to be included in the global demos. The boundary problem again disappears. Let us call these sorts of solutions to the boundary problem global demos views.
This article criticizes both functional demos and global demos views, defending instead a conception of democracy I call pluralist democracy. This conception sees the normative value of democratic governance as historically grounded and non-ideal. Pluralist democrats demanded—and eventually won—participation rights in autocratic polities, overturning the specific wrong of their being governed autocratically. The question of the democratic legitimacy of a polity is never asked in a vacuum. This is why it is mistaken to try to imagine an ideally democratic state, as functional and global demos views advocate. In contrast, pluralist democracy recognizes that all who are subject to a polity ought to have an equal stake in authorizing the laws to which they are subjected.
Part 1 of this article opens by considering whether functional demos views solve the boundary problem as they purport to. This is an “immanent critique” in the tradition of German Ideologiekritik, in the sense that it interrogates and criticizes functional demos views on their own grounds of success, drawing the relevant normative standards internally.Footnote 3 I find that, far from solving the legitimacy puzzle presented in the boundary problem, functional demos views exacerbate it. The question of who meets a functional membership standard cannot be solved in an objective fashion. All standards would need to be transposed into policy, and the technocratic running of such policies is unavoidably political. Consequently, democratic legitimacy concerns over the constitution of the demos are, in functional demos views, merely postponed. Let us call this objection the tenacious membership objection. If the tenacious membership objection is convincing, and if we take the boundary problem seriously, we should abandon functional demos views.
Part 2 of this article recalls the argument of Sarah Song that a functional membership standard of membership in a democratic community would mean that every political decision has a largely unique set of persons that ought to be included in the democratic process. Each individual would not be part of one demos, but of many overlapping demoi—meeting the appropriate functional standard with a different set of people depending on the issue at stake—a problem partly anticipated by Reference WhelanFrederick Whelan (1983). I call this characteristic of democratic systems with functionally defined memberships their “floating” quality. While Song recognized this dilemma, she did not analyze precisely what was wrong with floating demoi from the perspective of democratic legitimacy. I expand upon her argument to fill in this gap: floating demoi make self-rule—vital for democratic conceptions of law-making—impossible. Floating demoi result in floating jurisdictions and the piecemeal enforcement of democratic rule. In short, political communities whose memberships are functionally defined cannot be democratically governed.
In Part 3, I consider the extent to which arguments for a global democracy with a global demos solve the boundary problem. Theorists who defend global demos views may do so for a variety of reasons; a global demos may be postulated specifically to solve the boundary problem (e.g., Reference AgnéAgné 2010; Reference VerschoorVerschoor 2018: 15), or it may be considered the only defensible functionally defined demos on a particular functional standard. (Particularly, the “all-affected-interests” standard is sometimes argued to require a global demos—for example in Reference GoodinGoodin 2007.)Footnote 4 Part 3 of this article argues that while global demos views are successful in solving the apparent normative quandary posed by the boundary problem, they do so at a high cost. Global democracy cannot be procedurally democratic because it cannot be held democratically accountable by its citizenry.
Finally, in Part 4 of this article I sketch out a conception of democracy I call pluralist democracy. Pluralist democracy is a non-ideal conception of democracy that recognizes democratic procedures to be historically grounded and problem-oriented. Replacing polities’ nondemocratic practices with democratic governance legitimated those polities’ political authority. Given their problem-oriented character, it does not make sense to hold democratic procedures to ideal-theoretical standards of legitimacy. In other words, the facts that the boundary problem lays out are correct but are not an obstacle to democratic legitimacy. In contrast to the functional demos and global demos views, however, considering the legitimacy of the constitution of a democratic community from this perspective does have the crucial advantage of affirming a vision of democracy that makes a constituted polity's internal democratic governance possible.
Functional Attempts at Solving the Boundary Problem
Recently, there have been several new and innovative approaches to tackling the boundary problem (e.g., Reference GoodinGoodin 2007; Reference MillerMiller 2009; Reference AgnéAgné 2010; Reference SongSong 2012; Reference VerschoorVerschoor 2018). Most of these attempts have focused on discerning a rival, defensible standard for membership to simply accepting the status quo, which is argued to be irrelevant from a moral point of view. Rival standards often relate putative members of a demos to the political decision that is to be made “functionally”; if a would-be member's relation to the decision is considered sufficiently weighty on a stipulated metric, then their participation in that procedure is justified. I label these attempts “functional membership standards” (FMS), as they focus on the functional needs and interests of persons as the locus for legitimating grounds for membership, rather than on their juridical status (a legal standard) or on the place where they are settled (a territorial standard).
One FMS, most fully worked out by Reference GoodinRobert Goodin (2007) but discussed also by Reference DahlRobert Dahl (1989), Reference ShapiroShapiro (1999), and others, claims that a demos ought to be composed of all whose interests are affected by the decisions that the demos is to take—the all-affected-interests standard (Reference DahlDahl 1989: 93–95; Reference GoodinGoodin 2007; Reference ShapiroShapiro 1999: 37). Goodin runs into the problem that it is difficult to be analytically precise about what it is to be “interested,” particularly a priori, since interests are interdependent. Is it enough, for instance, that a decision “possibly” affects a person, or must it be “likely” to affect them, or “definitely” affect them for them to be included in the demos that is to take that decision? Regardless of this difficulty, that one's interests are affected seems a plausible and reasonable functional standard for democratic inclusion at first blush.Footnote 5
Another FMS was recently defended by Reference AbizadehArash Abizadeh (2008). He proposed that all those subject to coercion ought to be included in any democratic decision procedure, rather than all those who happen to be citizens of a preexisting democratic state. To secure this conclusion, Abizadeh lays out the lines of argument familiar from Whelan's boundary problem. According to Abizadeh, the boundary problem creates a “legitimacy gap” in the constitution of the demos only if we assume that demoi are in principle “bounded”—relatively fixed in membership and territorially limited. But by theorizing the “unbounded demos,” where the right to participate in an election is based on the functional question of whether one is coerced by the state apparatus, the boundary problem is avoided.
The unbounded demos thesis has far-reaching implications: “[T]he drawing and control of boundaries must ultimately be justified to those whom the boundaries are meant to define as non-members” as well as to members (2008: 44).Footnote 6 In his article, Abizadeh applies this principle to border regimes, but his proposed metric, which he calls the “demos qua regulative principle,” is also more generally a standard on which to measure the degree to which existing democracies and their political institutions can be considered (democratically) legitimate.
While Abizadeh is usefully explicit about why his functional membership standard is democratically superior to bounded territorial states (i.e., that it supposedly solves the boundary problem), the architecture of the argument that functional standards “solve” the boundary problem can be generalized. The boundary problem argues that the constitution of a demos ought—but ordinarily cannot—be democratically legitimated, whereas functional demos views hold that the constitution of a demos need not be legitimated by a procedure; it is legitimate for all persons bearing a certain functional relationship to a given law or policy to have a stake in the demos’ creation.
The Tenacious Membership Problem
Abizadeh's unbounded demos thesis, and its FMS correlates, do not solve the boundary problem. The boundary problem suggests that the question of to whom justification is owed is itself a political question—one that ought but cannot (hence the problem) be itself subject to legitimacy procedures. But FMS do not solve this paradox. Recall, FMS posit a functional standard that is supposed to resolve the foundational question of which persons ought to be included in a democratic body. But even if, for the sake of argument, we accept a particular standard as being correct, there is still the question of the administration of this policy and the adjudication of hard cases. The forums in which these questions are to be deliberated and the institutions that may execute subsequent policy cannot arise from nothing. Abizadeh writes: “[O]f course, political decisions about which options are valuable to whom cannot legitimately proceed without the participatory input of the persons in question” (2008: 55). But deciding which persons those are is itself a membership issue that cannot be settled democratically because of the very same regressive structure of argumentation that fuels the boundary problem.
In fact, any top-down resolution of the question of who ought to participate in whatever preferred democratic procedure deciding which persons adequately meet a functional standard, or indeed even setting the agenda on such an issue, requires an agent to initiate and an unavoidably political decision on who participates in that initiation, in setting that agenda, and in deciding who is to decide which persons are relevantly affected by a given law or policy and why. Any bureaucratic organization deciding such matters will itself suffer from the same supposed legitimacy deficit that the boundary problem determines a constituted demos has.Footnote 7 This is the tenacious membership problem, and it has application to all functional standards, since any FMS raises the question of who must decide which putative members meet the relevant standard, and for what reason. An organization with this decision-making capacity cannot itself be legitimated on the preferred standard of democratic legitimacy, since such an organization cannot itself be generated and justified by the extent to which its agents meet a functional standard (which would raise a new problem of regression), nor can it be chosen by a democratic procedure (which would raise the familiar regression from the original boundary problem).
Functional Standards=Floating Demoi
If the legitimacy of a democratic decision-making procedure is indeed dependent on all those subject to coercion participating in the procedure (or all those with affected interests, or any similar FMS), then the unbounded demos thesis has wide implications. Indeed, there is no reason to suppose that the appropriate demos for any two legal or policy questions will ever be the same. Song puts it this way:
What the affected interests and coercion principles require is different demoi for different decisions. Who will be affected or coerced by any single decision will vary from decision to decision, and as a result, democratic boundaries are not fixed but constantly changing A new demos is called into being for every new decision. (Reference SongSong 2012: 41)
Song calls this the “problem of stability,” and her concern was preempted by Whelan and Dahl, who, more than any others, popularized interest in and discussion of the boundary problem. Both formulated their concerns in light of the all-affected principle, but their concerns can be extended to all FMS.
In the now-classic text “The Boundary Problem and Democratic Theory,” Whelan points out that an FMS “would require a different constituency of voters or participants for every decision” (1983: 19). Dahl, similarly, voiced his worry that “for every different set of persons affected there be a different association or decision-making unit” and asked rhetorically: “[H]ow is the citizen who is affected by so many different units of government to devote much time and energy to any of them?” (1970: 64). In sum, the demos that an FMS creates is not only deterritorialized but indeterminate and shifting; it is more appropriate in this context to speak of overlapping demoi—with a different demos corresponding to each separate policy issue or legislative proposal.
Dahl, Whelan, and Song don't go far enough in recognizing the extent to which this conception of democracy is unstable. Not only is the demos indeterminate when membership standards are functional, but it is radically indeterminate—there is no way in advance to predict the precise composition of any particular (functionally defined) demos. Of course, living in the territory of a particular polity with defined borders would expose oneself to more local policy fields than not living there, supposing that jurisdiction is still understood for the most part territorially. However, and a fortiori given the ever-more interconnected economic and social world, few legislative policies that require coercive regulation can have their “coercive field” (the scope of persons that they potentially coerce) limited precisely to a particular territory. Thus, the demoi under this conception of democracy constitute an ever-changing set of deliberative forums with considerably overlapping jurisdictions whose compositions are unpredictable. This implication, that the membership of demoi are radically indeterminate—“floating”—has grave procedural implications far beyond Dahl's concern with managing citizens’ time and energy.Footnote 8 Or so I argue in the next subsection.
Whereas the tenacious membership problem demonstrates that FMS do not solve the boundary problem, the floating characteristic of the demoi that FMS create suggests that they may exacerbate it. The legitimacy concern that the boundary problem poses is elevated from a single problem at a historically defined moment (the constitution of the democratic polity) to a diffuse and recurring issue undermining the legitimacy of each and every decision. Every democratic decision would come coupled with a need to define the demos for that decision and thus every such decision would create the putative “legitimacy gap” that the boundary problem identifies.
The Indeterminacy of Subjection to Law under Floating Demoi
I next want to draw attention to a problem beyond the question of legitimately determining which persons ought to be members of any particular demos. Let's put that problem, fundamental though it is, to one side and say for the sake of argument that we have solved it technocratically (despite my objections in Part I). It still remains a fact under such a system that for every policy and rule one will be bound to and with a different set of individuals. This itself, in my view, renders democratic government impossible.
Imagine the practicalities of such a system. Even disregarding the enormous logistical task of matching every person to the network of rights and obligations that connect them to other persons based on the shared membership of floating demoi, the enforcement of rules must also be piecemeal. Remember that the exercise of political power over an individual is supposedly legitimated by some preferred democratic legitimacy procedure, which is implemented by a set of demoi corresponding roughly to the total number of rules and policies and not corresponding in any determinate fashion to territorially defined jurisdictions. How would law enforcement look in such a system? Either enforcement organizations would enforce the rule only of a particular demos (i.e., each rule would require a specific enforcement organization), or enforcement organizations would enforce a different set of rules on each individual, corresponding to the rules and regulations that apply to them on the basis of which demoi they were a member of. In the first case (and notwithstanding the costs of such an operation), there would be overlapping jurisdictions and a near-total abandonment of the principle of a monopoly of legitimate violence. In the second, the locus of legitimacy of the enforcement agency would be difficult to identify, and a relationship of democratic accountability would be impossible to maintain.
Correlating then to the radical indeterminacy of the constitution and membership of floating demoi, subjection to law under a system of floating demoi would also be indeterminate. One would not—could not—know in advance with which persons one would be bound, and by which persons one could be legitimately coerced to follow which rules and policies. And not only would law-making bodies and procedures be unstable, but the legal jurisdictions that they generate would also be indeterminate. This would have major implications in terms of the enforceability of rules and policies across overlapping and complex demoi.
Jurisdictions being indeterminate also poses direct problems for their legitimacy. A standard account of political legitimacy focuses on the legitimate exercise of political power, where wielding political power is understood as “attempt[s] to exercise a monopoly, within a jurisdiction, in the making, application, and enforcement of laws” (Reference BuchananBuchanan 2002: 689–690). But with overlapping and indeterminate jurisdictions, this element of the monopoly of power is impossible. Indeterminate jurisdictions also present a problem with regard to the plausibility of citizens being able to identify which rules and regulations apply to them, and what their corresponding rights and obligations may be. Recall Dahl's worry that citizens may not be able to devote much time and energy to the institutions that floating demoi give rise to. Once we focus on the indeterminacy of jurisdictions and law enforcement, the problem becomes much more dramatic: under the floating demoi of FMS, jurisdictional questions will dominate the interactions between persons. Uncertainty about one's own rights and obligations and the rights and obligations of others in one's vicinity would disrupt ordinary patterns of legitimate expectation. Normal economic and civic relationships between persons are based on expectations that, to a large extent, we share many rights and obligations with the people we interact with. Under FMS with floating demoi, such expectations would be impossible. Civic and economic interactions with an individual would be accompanied by the question of which demoi one shared with them and, correspondingly, which rights one held and which obligations one owed to that particular individual (that is, the extent to which one was in overlapping jurisdictions with that individual on any salient legal, policy, and/or regulatory issue).
It is already a significant legal fiction that it is possible for individuals to be aware of all the legal rights and obligations to which they are subject in a modern, stable jurisdiction, even with substantial legal training. Where the set of rights and obligations differs for every person, this problem becomes endemic. Each person would need to be a legal specialist in their own right in order to be able to identify even very roughly what they are legally permitted and obligated to do. Together, the myriad difficulties related to the enforcement of democratic rules made in unstable jurisdictions and the legal uncertainty that citizens of such demoi would face can be labeled the floating demos objection.
A Global Democracy Solution?
One way to take the wind out of the sails of the objections I have made—the tenacious membership objection in Part 1 and the floating demos objection in Part 2—could be with recourse to the idea of a “global demos.” The general strategy here would be to argue that, for one reason or another, a particular FMS ought best to be understood as including all persons globally. A global demos would avoid the problems detailed above, as the question of membership could be settled definitively (all persons ought to be members of the global democratic body). It would also result in a stable demos with a stable jurisdiction and the possibility of one enforcement organization applying the law consistently to all.
This sort of move, however, does not come without costs—many theorists are skeptical about global democracy on a wide variety of grounds that are beyond the scope of this article (e.g., Reference ChristianoChristiano 2006; Reference Dahl, Ian and CasianoDahl 1999; Reference Miller, Keith and ShaneMiller 2010; Reference NäsströmNäsström 2007: 647; Reference NiliNili 2017). On the other hand, there are also theorists who commend global (or “cosmopolitan”) democracy for reasons entirely independent of the global demos's ability to “solve” the boundary problem (Reference ArchibugiArchibugi 2008; Reference Archibugi and DavidArchibugi and Held 1995; Reference BohmanBohman 2007; Reference GoodhartGoodhart 2008; Reference MacdonaldMacDonald 2008; Reference Marchetti, Daniele, Mathias and RaffaeleMarchetti 2011). Regardless, from the perspective of the legitimation of the constitution of the demos, the global demos seems to be a theoretically elegant solution for those who propose FMS. After showing the affinity of two important accounts for FMS—Goodin's and Abizadeh's—with global demos solutions to the boundary problem, and engaging Song's objections to global democracy, this section lays out a critique of global democracy based on the impossibility of democratic accountability in a global state.
From Functional Membership Standards to a Global Demos
Both Goodin and Abizadeh concede that their arguments for FMS (respectively, on principles of all-affected interests and all-subject-to-coercion) result in an expansionary logic that cannot be stemmed before reaching a fully global demos comprising all persons (Reference AbizadehAbizadeh 2008: 18–19; Reference GoodinGoodin 2007: 65–66). The expansionary logic has several drives. First, for any specific decision a person may be directly affected or not, but some of those that are not directly affected would be affected had the decision gone the other way (Reference GoodinGoodin 2007: 54). Second, though a specific person may not be “actually” affected by vote X, regardless of the outcome of the vote, they surely are affected by the decision-making procedure that resulted in vote X being put on the agenda rather than possible vote Y (the results of which would affect them). Goodin concludes that his “expansive conception of ‘all possibly affected interests’ causes the franchise to balloon dramatically and the scope for legitimate exclusions to shrink accordingly.” The result? That “virtually, (maybe literally) everyone in the world—and indeed everyone in all possible future worlds—should be entitled to vote on any proposal or any proposal for proposals” (2007: 55). This move requires that one cast the FMS not as affecting “precisely those people potentially coerced by the political question at hand” (or affected in the way deemed relevant by a different functional standard) but rather “at least all those people coerced [relevantly affected] by the political question at hand” (2007: 55). A system of worldwide democracy (with a global demos) would include at least every single person meeting any particular FMS while avoiding the flaws of floating demoi.
Initially, Abizadeh wants to propose a standard less inclusive than the principle Goodin discusses. He thinks that by appealing to all-subject-to-coercion rather than a version of all-affected-interests, he is making a less-demanding claim: “Notice what my argument does not say. It does not say that all those who are affected by a political regime are owed democratic justification (and hence rights to democratic participation). My argument appeals to a more restricted principle” (2008: 45). Yet, Abizadeh recognizes that the best way to theorize the unbounded demos is in terms of a global demos consisting of all (living) people.Footnote 9 For instance, he writes that “[t]he unbounded demos thesis claims that ‘the demos’ in PL [normative democratic theory's principle of legitimation] is properly glossed as ‘all persons’” (2008: 45, fn. 27). Abizadeh further specifies the global implications of his view when he reflects on how a border regime granting participation rights to all those coerced by it may look institutionally:
To be democratically legitimate, any regime of border control must either be jointly controlled by citizens and foreigners or, if it is to be under unilateral citizen control, its control must be delegated, through cosmopolitan democratic institutions giving articulation to a “global demos,” to differentiated polities on the basis of arguments addressed to all. (Reference AbizadehAbizadeh 2008: 54)Footnote 10
The Global Demos, Democratic Procedures, and Democratic Legitimacy
The global demos seems to be a theoretically elegant solution to the boundary problem. It proposes that participation rights are granted to all persons everywhere—at least for those decisions that affect them (which, as discussed above, quickly expands to include all persons everywhere). But, is such a global democracy feasible? Is it normatively desirable? Is it logically possible? This section develops a critique on the feasibility of a global democracy on procedural terms, namely that it would not be realistically possible for citizens to hold a global democratic government to account, which, given the centrality of accountability to democracy, undermines the possibility of democratic government.
One classic, liberal, response to a global state stems from the fear that centralizing all political power in one institution would lead to despotism (as the saying goes, “Power corrupts, and absolute power corrupts absolutely”). This is the position that Immanuel Kant takes:
The idea of international right presupposes the separate existence of many independent adjoining states. And such a situation is essentially a state of war, unless there is a federal union to prevent hostilities from breaking out. But in the light of the idea of reason, this state is still to be preferred to an amalgamation of the separate nations under a single power which has overruled the rest and created a universal monarchy. For the laws progressively lose their impact as the government increases its range, and a soulless despotism, after crushing the germs of goodness, will finally lapse into anarchy.
(Kant [1795] Reference Kant and H. S.1991: 113)
This liberal response seems, however, to rely on a pre-political standard of substantive goodness that politics ought to facilitate. If the final end of politics is fixed in this way, then the desirability of a particular system of government would turn on its securing that end. However, the conception of democracy I endorse is open-ended (I return to this point in the next section). A democratic legitimacy standard ought therefore not to be interrogated on substantive grounds, but rather on procedural grounds, if it is to be interrogated on its own terms.
Whatever else the merits or demerits of the global demos as an ideal, we must ask ourselves whether a global democracy can further democratic government, or whether, for procedural reasons, it brings us further from that goal. I will argue that a global democracy would in fact be suspect on procedural grounds. I object to the legitimacy of a global democracy with a global demos based on the impossibility of such a polity being accountable to its citizens.Footnote 11 If this line of argument is convincing, then the question of whether a global demos might solve the boundary problem is moot, or at least incidental.
Song has offered a criticism of proposals for a global demos that offers an excellent starting point for our investigation. She argues that democracies must have three features: first, the protection of equal rights and liberties; second, guaranteeing that these rights have equal worth by providing “equal opportunities for political influence”; and third, a minimum solidarity between citizens (2012: 44). Song stipulates that a basic set of political rights are necessary for democratic government to be possible. These rights include negative rights such as freedom of political speech, freedom of the press, and freedom of assembly, and the positive demand for equal suffrage (which Song calls simply the “right to vote”). This is the minimal procedural core of democracy. As Song puts it, “political equality is a constitutive condition of democracy; the realization of political equality depends on respecting certain equal basic liberties and ensuring equal opportunities for political influence” (2012: 47). Song goes on to challenge global conceptions of democracy on the grounds that they cannot secure sufficient solidarity among global citizens (the polity being too large and complex).Footnote 12
Recourse to solidarity, however, makes Song's argument fragile in a particular sense: solidarity is at best instrumentally necessary to democracy, and, some may say, is just another political value between which democratic procedures must arbitrate, given incommensurable differences between people's views of what is in fact valuable (i.e., value pluralism). So, despite being sympathetic to her overall argument, I choose to sketch out a line of critique based on procedural aspects of democracy that are intrinsic to democratic government. I therefore proceed by further developing an aspect of her first constitutive feature of democracy—equal rights and liberties—as it offers fertile ground for a new procedural critique of global democracy.
At first blush, a procedural objection to global democracy grounded on the need for democracies to secure equal rights and liberties is not obvious. There is no theoretical reason to suppose that the negative freedoms listed—freedom of speech, press, and assembly—could not, in principle, be guaranteed by a global democracy (no matter how unlikely it is to achieve them). The same applies to equal suffrage. However, we ought to start by asking ourselves why freedom of speech, press, and assembly are constitutive freedoms in a democracy. The answer must lie at least in part in their being prerequisites for citizens to hold governments to account; to exchange and debate their views on governmental activities; and to form their judgment on how they wish to cast their votes on election day. It is therefore a prior demand that citizens must have adequate access to knowledge and information of government action to inform those democratic (electoral) accountability mechanisms that make sense of democracy's constitutive liberties as reported by Song and Dahl.Footnote 13
Following the analytical definition of the core components of democratic accountability by Reference PhilpMark Philp (2009), I hold that three things are required for an agent to be held accountable; they must be required to inform, explain, and justify their decisions and commands. Putting aside the difficult question of precisely who is empowered to demand accountability from a political authority,Footnote 14 and how they can do so, democratic accountability requires that “members of a collectivity [are] able [to] ask for, receive, and accept (or reject) reasons and justifications for the exercise of powers” (Reference Warren, Mark, Robert E. and ThomasWarren 2014: 41).
Once we recognize the importance of adequate access to information about government activities, an accountability-based, procedural critique of global democracy starts to emerge. A global democratic government would clearly be less accountable on these terms than a national democratic government, and more likely to be oligarchic (Reference ChristianoChristiano 2006; Reference Urbinati and DanieleUrbinati 2003). Indeed, Reference Keohane, Mathias and DavidRobert Keohane (2003: 130ff) points out that sophisticated accounts of global democracy recognize the impossibility of global democracy meeting this accountability standard through electoral politics and develop alternative accountability mechanisms to try to fill this gap. But only accountability mechanisms that afford each citizen of a global demos formally equal and adequate accountability mechanisms can meet the democratic norm that a polity treats each person as an equal source of authoritative value (the “equal value demand” that I develop in the following section). Some may claim that a system of global accountability could be devised (gargantuan though the task appears) by targeting and streamlining information on the activities of the global government to those citizens concerned (Reference MacdonaldMacdonald 2008; Reference MarchettiMarchetti 2008). The problem with this move is that such a system could not make any assumptions about which citizens are or are not likely to be “affected” or “coerced” by a particular decision or government action if it is not to falter under the critique I have made of FMP toward the end of Part 1. Determining which citizens are affected in the relevant ways would be a political decision that raises the same specter of the boundary problem that global democracy purported to solve. Only citizens themselves, by “voting with their votes,” if you will, can legitimately determine whether and how they consider their interests affected by government action.
Pluralist Democracy and NonIdeal Democratic Legitimacy
So far, we have considered FMPs and global democracy as two ways of addressing the boundary problem, which holds that whatever the theory of legitimacy that we believe our preferred democratic decision procedure should meet, the initial act of constituting the demos can never be considered met by it. Yet we have seen how inadequate both solutions are. The first leads to floating demoi with radically indeterminate jurisdictions that undermine the possibility of democratic government. The second undermines the possibility of a relationship of accountability between citizens and a global democracy. This leaves us with a dilemma: is there no way for democratic theory to resolve the boundary problem?
In this last section, I want to sketch out an account of democratic legitimacy that I call “pluralist democracy.” Pluralist democracy is not, I claim, sensitive to the boundary problem. Therefore, if we accept pluralist democracy, we need not embrace the convoluted proposals of advocates of FMS, nor embark on a futile quest for a global democracy—at least, not as means to solve the putative normative dilemma posed by the boundary problem. I do not provide a first-principles defense of pluralist democracy here but show how two common and plausible postulates of democratic theory should lead us to recognize democratic legitimacy as a historically grounded and non-ideal standard.
Pluralist democracy does not resolve the boundary problem: it is still the case that the pluralist standard of inclusion in a demos—subjection to law—cannot legitimate the constitution of the demos (Reference SchafferSchaffer 2012). But pluralist democracy denies that this is a flaw. Democratic governance is justified not by its meeting an abstract, ahistorical ideal, but by reference to the illegitimacy of concrete, historically particular polities that democratic states replace. In that sense, democratic legitimacy is a comparative, not an ideal notion. So the boundary problem is a category mistake—there simply is no “initial act of constituting the demos”; demoi are made from existing polities and are justified comparatively. Relative to their nondemocratic predecessors, emerging democratic states include (more of) those subjected to the laws of the polity. They do so through formally equal participatory rights (such as the right to vote and be elected), the exercise of which collectively determines how political power in that polity is to be used. In other words, democratic processes should be considered to legitimate real and existing political communities understood in a historical and contextual (though not necessarily ethnic or national) sense. Cicero makes a similar point in On the Commonwealth when he writes “the commonwealth is the concern of a people, but a people is not any group of men assembled in any way, but an assemblage of some size associated with one another through agreement on law and community of interest” ([54 BCE] Reference Cicero and James E. G.1999: 18).
The argument is centered on the idea that the correct conception of legitimacy to be used in examining the democratic legitimacy of a polity is contextual and not universal. Whereas Abizadeh posits the core “ideal” understanding of the demos as “all persons” understood universally, pluralist democracy considers demoi to be contextually (pre)defined. Democracies do not arise from nothing; they are preceded by political institutions claiming relationships of authority over defined and determinate subjects and a definite and determinate territory. These defined and determinate subjects are the populus, and in nondemocratic states their political equality is denied. The state institutions of law-making, administration, adjudication, and coercion rule over the populus, not in its name (although many may claim to be doing so). These institutions democrats judge to be illegitimate for not living up to the standards of democratic legitimacy. This illegitimacy is the central premise in political agitation to democratic reform. Where successful (and often historically the success is piecemeal and gradual rather than comprehensive and immediate) the populus becomes the demos—a body of people united through their subjection to laws that they have an equal stake in creating.
I have chosen to name the conception of democracy outlined here “pluralist democracy” for two reasons. First, pluralist democracy will ordinarily result in a democratic polity characterized by competition between different political projects and ideologies (i.e., political pluralism). The second reason is the foundational thesis of the possibility of value pluralism. While a demand to treat persons as equal sources of value (the “equal value demand”) underlies pluralist democracy's generally democratic conception of legitimate politics, the possibility of value pluralism demands majoritarian and iterative democratic procedures. Before further exploring the equal value demand and the possibility of value pluralism, I want to briefly consider the question of how inclusive the body of citizens with suffrage is compared to all persons in the polity.
In ancient Athenian democracy, there were (roughly) four classes of persons: male citizens, female citizens, metics, and slaves. Citizens were those whose parents were Athenian. Metics lived in Athens but were not Athenian-born, and slaves were the property of citizens or of the city. Of these types, only male citizens, once they attained the age of majority and had completed their military service, could vote or hold office. Many democratic theorists now consider the demand that all adult citizens ought, in a democracy, to be able to vote and hold office central to democratic legitimacy. A polity according suffrage only to men would not, on this view, be democratic, properly speaking. Furthermore, many contemporary theorists hold that noncitizen residents (those whom Athenians would classify as metics) should also be able to vote and hold office, and perhaps even that non-resident citizens should not be able to vote and hold office (Reference López-GuerraLópez-Guerra 2014: 83–108). In contrast to the functional membership standards (such as all-those-affected and all-those coerced) discussed in Part 1 of this article, pluralist democracy demands that all those permanently subject to law ought to be included in the democratic process. This may exclude some nonresident citizens, and will include many noncitizen residents, much like the view defended by Claudio López-Guerra (Reference López Guerra2005).
Pluralist democracy argues for a (historically) highly inclusive demos—all those permanently subject to the law should be a member. This follows from two plausible claims: the normative principle of persons as equal sources of political value, and the recognition of the permanent possibility of pervasive and incommensurable value pluralism. While I do not offer a first-principles defense of the first, nor seek to prove empirically the second, I do hope to underline their normative and empirical plausibility.
It is perhaps helpful to first clarify the precise interrelation of the two foundational principles, since pluralist democracy considers them jointly necessary and individually insufficient for grounding democratic government. If only the possibility of value pluralism is correct, then we are left with no normatively salient reasons for prioritizing the majority view over others.Footnote 15 If people are not to be treated as equal sources of authoritative value, then no person ought to feel bound to accept a view other than their own as authoritative and normatively binding. If, on the other hand, only the equal value claim holds (in other words, if there can be societies constituted by permanent value monism), then there is not always a reason to turn to competitive democratic procedures to decide what is to be done—members of a society characterized by stable value monism may widely agree on the political agenda.
The possibility of value pluralism is weakly empirical, as it concerns the actual, empirical possibility of value pluralism rather than the fact of value pluralism, either in any particular context or in all human social contexts. It is in that sense a thesis that has its foundations in the moral psychology of human nature, rather than being a sociological or anthropological claim. Moving from the fact to the possibility of value pluralism gives us democratic grounds for opposing the dissolution of the institutions and procedures of democratic government through structurally democratic procedures.Footnote 16 There are other reasons to prefer an account grounded in the possibility rather than the “fact” of value pluralism. First, it allows us to avoid the arduous empirical task of determining whether and when the factual claim is actually true. Second, an account of democracy's value grounded in the fact of value pluralism seems to be importantly contingent. If we are to imagine a society that lacks this sociological characteristic, then we lose our foundation for democratic government. Equally, in countries that are not (or are less) characterized by value pluralism, the pluralist democratic defense of democratic government would be weaker or may not hold at all.
Neither the claim of the possibility of value pluralism, nor the equal value demand are metaethical claims. While it seems evident to me that the metaphysics of ethical rules or norms must be either monist or pluralist and cannot be both, I do not take a position on this question here. Neither do the possibility of value pluralism and the equal value demand make a thick claim about moral realism versus moral antirealism; pluralist democracy can remain largely agnostic about both the question of whether and how (any) ethical propositions are factually true and the question of whether value pluralism as an ethical stance is committed to moral realism or moral antirealism.
There is, however, one sense in which I cannot remain fully agnostic. Pluralist democracy is committed to the foundational principle that persons are treated as equal sources of authoritative value (the equal value demand). A full-blown moral nihilist position (or thoroughgoing moral skepticism) would deny the force or attractiveness of foundational principles in general, including the equal value demand. Accepting the equal value demand must therefore imply opposing moral nihilism and its cognates. It does, however, permit agnosticism about the relative or absolute grounding of ethical propositions. It may be, as with Aristotelian first principles, that the equal value demand is foundational in the philosophical sense and cannot be “proven.” If that is so, then perhaps it can only be comprehended in a noninferential manner and consequently corroborated by our experience. Such a conclusion would be—controversially—in line with Reference CohenG. A. Cohen's (2003) views on the fact independence of final normative principles. Perhaps, in contrast, an intrinsic or relativistic defense of the equal value demand is more convincing (Reference ChristianoChristiano 2008; Reference Invernizzi AccettiInvernizzi Accetti 2015). The challenge for such an account in light of pluralist democracy would be to mount this defense without reliance on a substantive conception of the good that would render it self-defeating.
It is important not to confuse the equal value demand with what might be called an “equal value claim.” The idea is not that, objectively, different and contrasting (perhaps incommensurable) views regarding value held by individuals have equal value. This would be an ethically ambitious and probably nonsensical idea (outside of moral antirealism, which holds that contrasting values have an equal moral value of nil; see Reference JoyceJoyce 2015). The equal value demand is much less ambitious. It holds that, politically, no one in a political community ought to be treated as a superior source of authoritative value. That is not to say that everyone's judgment over what is valuable must be respected, strictly speaking, since this would generate veto rights for every person and, consequently, that organized minorities can hold an unequal weight in political procedures. Respecting persons as equal sources of political value means no more and no less than ensuring that all have an equal stake in the political process.
Joseph Schumpeter, for example, has argued that if a particular society is, for instance, highly religious, then they should be able to exclude nonmembers of their religion from their democratic community ([1942] Reference Schumpeter2013: 244–245). It seems that such a society is not characterized by pervasive and incommensurable value pluralism. However, pluralist democracy insists on the empirical premise that, even in such a society, the possibility exists of value pluralism in the future, and for that reason all those permanently subject must be included in the demos. This idea is prefaced on democratic processes being iterative.
The demand that democratic processes are iterative stems from the equal value demand and the possibility of value pluralism. Imagine that in society A, at time T1, 90 percent are members of religious congregation Y and a vast majority of those believe that only members of their congregation ought to participate in the democratic processes. Schumpeter would argue for the legitimacy of this move, invoking (implicitly) the democratic character of this exclusion by the majority. Pluralist democracy, however, insists that the minority (the 10 percent at T1) may become the majority at T2—or that in any case the majority at T1 may, in the future, lose their status as majority at T2 (such changes in the cultural makeup of a territory's population are common, historically). If congregation Y had been able to define the demos in their image at T1, this would result in minority rule at T2. Notice, I am not saying that the disenfranchisement of the 10 percent portion of the demos at T1 would “not be democratic.” It would be a democratic decision in one important sense, namely that it would meet the descriptive condition of being a decision taken by the vast majority of electors or their representatives. It would not, however, be a democratically legitimate decision, if we accept pluralist democracy's two foundational claims—the equal value demand and the possibility of value pluralism.
There remains a question about democratic legitimacy and democratic inclusion, concerning the extent of how to judge putative democratic polities that imperfectly include all those permanently subject to the law. For instance, was Switzerland democratically legitimate before women were awarded the federal franchise by a majority (66 percent) of (only male) citizens in a referendum on February 7th 1971? This is a difficult question to answer. Denying it seems to deny the legitimacy of that vote. Pluralist democracy here shows its historically grounded and nonideal character; decisions that improve the inclusivity of a demos toward the inclusion of all those subjected to the law will be taken to constitute advances in terms of the legitimate and democratic character of a polity.
Conclusion
It is important to be clear as to the precise nature of my claim and its consequences: I do not claim that there are no pressing questions of political membership; the political independence of an ex-colony such as New Caledonia from France is an example of such a question, as is the question of the status of Northern Ireland in Great Britain, the salience of claims for West Papuan independence from Indonesia, and Catalan independence from Spain, to name but four that are live in current affairs. The claim, rather, is that abstract and universally framed considerations of “perfect” democratic communities with perfectly justified memberships have no purchase on such real-world issues, which must thus be settled by reference to other norms and values. In short, my view is that there is no boundary problem (or that it is not a problem of practical import), but there are many normative issues pertaining to boundary-setting, for instance in the context of immigration and secession. Therefore, pace Abizadeh, proposals for immigration regimes, for instance, cannot be justified via recourse to the supposed boundary-problem-solving quality of proposed arrangements.
The key is to recognize that democratic legitimacy is a non-ideal political value that must be contextualized (placed into a context) for it to have meaning. The question “is the demos legitimately constituted in this preferred theory of democratic legitimacy?” therefore has little value. Liberal democracy, far from being an ideal arrangement in universal terms (assuming, as ideal theory tends to, that persons act in full compliance with their obligations and in circumstances of perfect knowledge), is necessary, normatively, for non-ideal reasons (the possibility of value conflict grounded in value pluralism). The best justifications for democratic government are grounded in the intransigent and often incommensurable political conflict and disagreement between persons over how to handle unavoidably common affairs. To adapt a famous line from James Madison, if people were angels, no democratic government would be necessary.
Once understood in this historical, contextualized sense, the legitimacy of the constitution of the demos can be asked in relation to a necessarily determinate group of people exercising a particular form of collective control over coercive power. Indeed, pressure for reforming an existing autocratic regime follows a similar structure. There is not nor has there ever been a “state of nature,” in which free and politically independent individuals have the possibility of contracting to form an ideally conceived political association. Rather, calls for democratic reform are always made on the basis of the putative illegitimacy of an existing regime that claims authoritative status without democratic institutions. What legitimizes the democratization of a polity (or the spread of the franchise in a polity) is that its constituent members were already a collective, bound under coercive regimes that did not afford them a voice.
Of course, the above account is a simplification, not least because democratic franchise has not generally, in history, been awarded to all adult citizens at once, but rather been extended slowly, along the lines of sex, class, and property. Nevertheless, I think that when asking, for example, if the extension of the franchise from men to women was legitimate on democratic grounds, it would be absurd to say it was not, as women had no vote in the matter. Rather, the resulting constitution of the demos ought to be contrasted with what preceded it on the relevant democratic grounds (the extent to which all those subject to law are included in the demos).
This position has an interesting consequence. Take the United Kingdom's 1918 Representation of the People Act,Footnote 17 which extended suffrage to men and women without an education or property qualification. In the context of early-twentieth-century political history, the Act was a major democratic reform. In contrast with what preceded it, it extended the franchise to many more subjected to UK law, removed economic and class-based barriers, and is recognized by pluralist democracy as a breakthrough in the United Kingdom's democratic suffrage.Footnote 18 Nevertheless, if a similar measure would be taken now, it would be shockingly illegitimate on the self-same grounds, since the Act limited the franchise to women over 30, while for men the limit was 21; given the current rule in the United Kingdom granting suffrage in general (parliamentary) elections to men and women equally from the age of 18, this would constitute a major regression from the pluralist democratic standard that all those subjected to the law of a polity should have an equal stake in its authorization.
If legitimacy is understood in the way I have described, much of the sting is taken out of the boundary problem. The point that I have conceded—that the (initial or enlarged) constitution of a demos cannot be legitimated on the ordinary grounds by which the democratic legitimacy of the political order is subsequently tested—no longer seems so problematic. The legitimacy of the constitution of an existing polity is driven by a demand for more representative government by a people already “bounded” both corporately and territorially. The legitimacy of further democratic reforms is considered in this context. Do reforms extend democratic rights among those subject to the legal and political order?
While this does not solve all—or even most—of the justificatory questions pertinent to the accession to and secession from existing democratic polities (of both individuals and groups), such a non-ideal approach does allow us to theorize the democratic legitimacy of existing states in circumstances of normal politics. It also sets as a non-ideal standard a conception of democratic legitimacy that enables rather than disables the democratic government of particular, bounded polities, in stark contrast to radically indeterminate polities with functional membership standards or a global “democratic” state.
Acknowledgments
Some articles are written quickly; this is not one of them. My intellectual debts stretch back to the start of my doctoral research, when I wrote an article criticizing Abizadeh's conception of the “unbounded demos,” and which I discussed extensively with my supervisor, Justine Lacroix, as well as with Astrid von Busekist, Carlo Invernizzi Accetti, Benjamin Boudou, Amélie Ferey, and Martin Deleixhe. I presented various drafts of that piece at the Graduate Conference in Legal and Political Theory at the University of Warwick in February 2014, and at Sciences-Po's atelier de mémoires de théorie politique, the Political Theory Seminar of the Université Libre de Bruxelles, and the Tel Aviv/Sciences Po joint seminar “Boundaries and Nationhood” later that year. This article also benefited from helpful and constructive feedback from the anonymous reviewers, as well Jean-Paul Gagnon's useful suggestions and encouragement.
Open access
