Agents of Popular Sovereignty

What Does It Mean to Say That Citizens “Act”?

The essence of an agentive understanding of popular sovereignty is that citizens can connect their agency to collective decisions about a state’s general laws and policies. As one prominent author writes, popular sovereignty so conceived involves not only a commitment to the idea that “the ultimate wielders of political power are the citizens themselves” (which, as Bodin held, is compatible with denying citizens the right to rule themselves actively) but also “that the institutions of the state are the agencies through which [citizens] act.” ⁹ Act is the keyword here, but what exactly does it refer to? I suggest that this is the first issue that must be settled. For only when we have a clear idea of what acting involves can we evaluate different conceptions of popular sovereignty according to their ability to allow citizens to engage with democratic agency in a meaningful fashion.

At the most general level, the exercise of agency requires a “me” that directs the causal order of behavioural events and is capable of critically reflecting on them. As one author puts it, “[w]hat makes us agents rather than mere subjects of behaviour . . . is our perceived capacity to interpose ourselves into the course of events in such a way that the behavioural outcome is traceable directly to us.” ¹⁰ This talk of “behavioural outcomes” typically refers to the actions of individuals, but as many authors have shown, it can plausibly be “extrapolated” to democratic life. Accordingly, to say that citizens act through the institutions of the state means that they perceive the general laws and policies as traceable to themselves. ¹¹

That citizens perceive certain political outcomes as traceable to themselves is necessary for them to be said to be acting through the institutions of the state, but is not sufficient. There must also exist some actual causal connection between their attitudes or preferences and general laws and policies. However modest the influence of a single person may be, she must be able to make a contribution to those decisions.

To say that citizens must have causal impact on laws and policies invites the familiar objection that democracy, realised in accordance with the principle of popular sovereignty, actually condemns citizens to “causal impotence.” ¹² For insofar as political power is distributed equally, and each vote (or voice) counts the same, no individual has a causal effect on the collective decision. But this objection misconceives the type of causality underlying democratic action. To exercise causal influence in a democracy is not to single-handedly decide an election, say, as implied by the objection from causal impotence, but to contribute to a political decision jointly together with others. The type of causal influence at work may be characterised as what has been called “contributory causation.” ¹³ This type of causation is difficult to understand in terms of the standard counterfactual analysis of causality, but there is nothing unfamiliar about it. Beerbohm illustrates the point as follows: “When a large majority of persons votes to humiliate a fellow citizen, is it plausible for the victim to view this decision [as one where nobody had any causal effect]? If you were the target of such abuse, would you accept the claim that none of the millions of citizens who voted to grossly mistreat you had any ‘causal effect’?” ¹⁴ Clearly it would be implausible to deny any causal connection here.

To better understand why both the perception and the causality desideratum must be satisfied, consider the following cases. Imagine first a state run by a charismatic leader who claims to speak for the people as a whole. That leader has the last word about the meaning of the general will and stands above democratic procedures and representative institutions. Suppose further that citizens perceive the leader’s decisions to be genuinely “theirs,” for instance because they identify with the leader, as in Ernesto Laclau’s model of populist democracy. ¹⁵ If so, the perception desideratum is easily satisfied. Yet, since citizens are unable to causally influence the decisions of the leader (because the leader stands above democratic procedures and representative institutions), we would be hard pressed to attribute to them the status of agents (and hence to call the state in question a properly democratic state). ¹⁶ It seems uncontroversial that the satisfaction of the perception desideratum is, by itself, insufficient.

Imagine next a polity where a causal link between citizens’ preferences and political decisions appears to exist, but citizens largely regard themselves as disempowered. One example of such a polity may be the European Union: though there is ample evidence that governments usually decide in line with domestic public opinion when acting at the EU level, being as they are constrained by domestic electoral incentives (this is the causal mechanism), citizens tend to perceive themselves as having scant influence on EU policy. ¹⁷ One’s initial inclination will be to say that the widespread perception of powerlessness poses a very real problem for the democratic credentials of that polity. This is so because our perceived capacity to interpose ourselves into the course of events is just as central to the notion of agency as the de facto capacity to shape outcomes, as established from the third-person perspective of an observer (say, a researcher). Again, it seems clear that both agency desiderata must be satisfied.

How can the citizens’ agency obtain tangible expression in a democratic polity? To what institutions and practices should one look? As I have noted, two different responses to this question are available, one that emphasises direct participation by the people as a whole (the radical view of popular sovereignty), and one that emphasises the necessity of a division of labour between citizens and political representatives (the indirect view of popular sovereignty). It is to these views that we will now turn.

Radical Popular Sovereignty

The radical view of popular sovereignty contends that the latter requires that the citizenry as a whole—the people—asserts its will directly, say in a nation-wide referendum. This view goes back to the time of the French revolution and the Rousseauian trope that democracy and representation are essentially incompatible, and parts of this legacy can still be traced in some contemporary accounts of radical popular sovereignty. While very few of these accounts would question that representation and democracy are compatible, they all assume that the people is only truly sovereign when it acts directly and in a more or less unified fashion. ¹⁸ Even if direct participation may not be the only way in which norms, laws, and policies can be made responsive to citizens’ values and preferences, goes the argument, popular sovereignty makes itself visible only in moments when the citizenry of a given political community acts as one. In these moments, we can discover the true will of the people.

Importantly, the radical view of popular sovereignty is not synonymous with the much broader notion of “direct democracy.” While the latter is typically thought to involve all sorts of more or less unmediated decision-making mechanisms that may be implemented at different levels in the political system ¹⁹ —ranging from local popular initiatives to participatory budgeting schemes—the former emphasises, as noted, the participation of the whole people in a direct decision. In practice, moreover, the radical view of popular sovereignty requires one to commit to the idea that the electorate is the people ²⁰ (and usually also to the idea that the will of the half plus one is the will of the whole). To this contestable belief proponents of direct democracy need not subscribe, as one can defend the latter simply as a palette of participatory devices and remain agnostic about whether these can afford us an insight into the true will of the people.

Now, does the radical view of popular sovereignty satisfy the two agency desiderata set out above? Instinctively one thinks that the answer must be yes. For it seems that there is simply no more effective way for citizens to determine the direction of the state than via a direct-democratic vote. It enables citizens to causally determine decisions (thus satisfying the causality desideratum), and it certainly also makes them perceive the decision as directly traceable to themselves, not least because there appear to be no intermediary bodies like parties or parliament standing between the “popular will” and the decision (thus satisfying the perception desideratum). It is for this reason that “asked to say what the exercise of popular sovereignty means, many citizens of modern democracies would point to referendums.” ²¹

But this argument overlooks that referenda are not typically designed by the people themselves but by political elites, who tend to be disconnected from the wider society and have axes to grind. This need not undermine the radical account’s capacity to satisfy the perception desideratum: citizens may still perceive the outcome to be genuinely “theirs.” But it raises doubts about its capacity to satisfy the causality desideratum. To see this, consider that the political question that is referred to the people for a direct decision in a plebiscite is hardly ever actually formulated by citizens, but by officials. As Carl Schmitt correctly observed in his discussion of the plebiscites of this time, ordinarily “the people . . . cannot pose a question, but can only answer with yes or no to a question placed before them.” ²² Without means by way of which the people can determine the issues over which a choice is being made, however, their capacity to causally shape decisions is severely circumscribed. The people are granted direct causal impact on a political decision, yes, but they have little impact on the nature and content of the decision, which is predetermined by elites. ²³ Furthermore, political elites usually also possess the resources to influence public debate so as to shape the final outcome of the referendum. Especially when there is widespread agreement among elites as to the most desirable outcome of the referendum, dissenting voices are bound to get rather little publicity. Schmitt for these reasons went so far as to suggest that, in a plebiscite, the causality of democratic decision making is actually inverted, in the sense that referenda are but a mechanism of acclamation: decisions are de facto made by political elites, while the people are relegated to head-nodding or -shaking. ²⁴

Of course, referenda are not always or necessarily manipulated by self-interested elites. There are certainly cases of referenda where citizens are granted more control over the issues that are being decided (or where the public debate prior to the referendum is inclusive). But it seems indisputably the case under contemporary political conditions that the power to decide on the substantive content of referenda rests not normally with the citizens but with their government. ²⁵ Indeed, it is difficult to think of relevant recent nation-wide referenda in established democracies where that has not been the case. It is therefore also no surprise that, if we inquire into the reasons for why certain political questions are made the subject of a nationwide majority vote, or why the exact question over which a choice is being made is formulated as it is, typically we find that, “because the people wanted a referendum on that particular question” is an dubious explanation.

For example, the much-discussed Brexit referendum in 2016 was initially suggested by the British Conservative Party in order to appease the Eurosceptic wing of the party and to avoid a flight of voters to the populist right-wing United Kingdom Independence Party (UKIP). Even if the British public has consistently been the most Eurosceptic electorate in the EU ever since the UK joined in 1973, it would therefore be misleading to say that an in-out referendum has been straightforwardly “demanded by the British people.” Similar things may be said about the Scottish Independence referendum in 2014. This had its origins not so much in the desire for the people to decide fundamental constitutional matters, but in the Labour Party’s rather clumsy response to the rise of Scottish nationalism, and the hope that a referendum victory for the Labour-backed pro-UK side would make it go away. In sum, the agenda of these momentous referenda was shaped not so much by citizens’ exercise of democratic agency as by parties who tactically sought to avoid losing electoral support. All of this is neither surprising nor uncommon in an age where many formerly powerful parties have been abandoned by their traditional electorates, and deference to political elites has almost entirely disappeared. ²⁶ But it raises serious doubts about whether radical popular sovereignty can be exercised in a way that actually gives expression to citizens’ agency.

The conclusion to take from examining these arguments is that under contemporary political conditions, the radical approach to popular sovereignty does not provide an appropriate way of enabling citizens to “act through the institutions of the state.” Though it appears prima facie empowering, on closer inspection it is prone to compromise citizens’ capacity to act by excluding the people from deciding on the terms of their empowerment.

Indirect Popular Sovereignty

The second widespread approach to popular sovereignty conceives popular sovereignty in indirect terms, appealing to the necessary division of labour between the people and their political representatives. Like the radical view of popular sovereignty, this approach has a long pedigree in the history of political thought. A figure often mentioned as its patron saint is the anti-Jacobin clergyman and writer Emmanuel Joseph Sieyès, who broke decisively with the already-mentioned Rousseauian idea that democracy cannot be representative. ²⁷ As one scholar has put it, Sieyès “saw representation as a fundamental fact of modern society, as something indelibly inscribed in the division of labour and commercial sociability, and political representation as a permanent necessity in any large and populous country in which it was virtually impossible to unite the voice of the people directly.” ²⁸

On this perspective, elections are essentially a means for creating two peoples: a group of citizens who make the laws for all and a group of citizens who obey those laws—but this reallocation of direct decision-making power does not undermine the sovereignty of the people. It only changes the mechanism through which the people can make their agency felt: rather than relying on direct rule, the “generation of a collective national will” is made “dependent upon the unity of a national representative body . . . whose representatives [have] constituted power . . . to act in the name of the people.” ²⁹ Thus the people exercise agency by way of authorising in elections representatives to act on their behalf.

This conception of representative government remains highly influential in democratic theory and practice, which tends to take for granted both the inevitability of a division of democratic labour between citizens and their representatives and the legitimacy of that division of labour. ³⁰ By making the divide between citizens and their elected representatives a constitutive feature of democratic government, then, the indirect approach to popular sovereignty “makes room for sovereignty as an inherently plural unifying process” ³¹ that involves giving collective shape to myriad individual interests, preferences, and conceptions of the common good.

Facilitating that process are political parties. Parties mediate between individual principles and projects and the good of the whole, in that they “refine and generalize particularist appeals by casting them in terms appropriate to public reason,” ³² and make these appeals the basis of general laws and norms that are created in parliamentary compromises engineered by the representatives of different (competing) parties. ³³ This allows parties to master the apparent paradox that representatives are both supposed to make laws that all citizens must obey and be responsive to the values and demands of those who elected them, such that the latter can perceive the content of laws and norms to be traceable to themselves.

Does the indirect account of popular sovereignty satisfy the two agency desiderata? The answer is that it can, but only under particular circumstances. Firstly, the indirect account can satisfy the causality desideratum if (1) citizens possess the right to an equal share in determining the political will via elections, with the incentive for representatives to get re-elected ensuring that representatives and parties are reasonably attentive to citizens’ concerns and expectations. This requirement seems not especially arduous. After all, its demands are no more exacting than those of Schumpeterian minimalist democracy. But, secondly, there are two further, more demanding requirements that must be met in order for the perception desideratum to be satisfied. Since, as one prominent theorist of representative democracy puts it, the indirect view makes “apathy, not agency, the main quality of popular sovereignty” and “citizens’ participation during the period between elections superfluous,” ³⁴ it requires (2) that the people are in principle content with non-participation between elections, and (3) that representatives continually maintain an informal sympathetic relation with their constituents, in the sense of an “interpretative or artificially created similarity” that instils in the latter the sense that they are closely connected to decision making. ³⁵

Can these requirements be met in practice? Whether requirement (1) can be met in contemporary democracies is difficult to answer. Political scientists disagree over the extent to which real-world parties and governments are actually responsive to citizens, and about the main causal factors involved. Evidence tends to point in different directions, depending on the methods used to study responsiveness and the countries studied. ³⁶ An answer to the question of whether requirements (2) and (3) can be met seems more readily available.

Let us begin with requirement (2), that which is concerned with citizens’ willingness to accept non-participation between elections. That this requirement can be met is, prima facie, doubtful. Research indicates that across established democracies, the decline of class cleavages and the concomitant rise of post-materialist and self-expression values set in motion a transformation of citizens’ participatory demands that rendered people dissatisfied with institutionalised forms of collective action that leave little or no room for individual voice. While in the age of class conflict and mass parties many citizens readily accepted non-participation between elections, not least because they felt adequately represented by parties that reliably stood up for their group interests, nowadays citizens tend to prefer different, more individualised forms of participation. ³⁷ The indirect model offers however little in the way of channels for more active, or different forms of, participation and thus risks leaving citizens feeling disconnected from the institutions through which they are supposed to act.

That representatives are capable of maintaining an informal sympathetic relation with their constituents (3) in contemporary democracies seems equally unlikely. The just-mentioned decline of class cleavages and rise of self-expression values has triggered a crisis of representation and accountability that is marked by a growing disconnect between representatives and represented. Schmitter describes the cause of that crisis in terms of the “generalized loosening of the links between interests and organizations,” understood as the unravelling of the interests that emerged out of the grand cleavages of religion and class and were mediated by Christian, social democratic, and communist parties. ³⁸ What has arisen in their place is a seemingly unaccountable and technocratic sort of politics whose main agents often appear more concerned with holding onto office and minimising the costs of losing it than with promoting principled agendas that are responsive to citizens’ political aspirations and expectations. A powerful indication of this is that in the face of declining memberships and weak partisan alignments, many established political parties have begun increasingly to rely on the state’s bureaucratic apparatus (courts, regulatory agencies, independent commissions) for their power and legitimacy. ³⁹ This has instilled in many citizens the belief that officials and representatives form an undifferentiated political class that rules in unrepresentative ways. ⁴⁰ This belief, reinforced by populist parties and movements that fervently oppose established politics, undercuts citizens’ sense of being able to exercise political agency.

In sum, there is a very real sense in which the indirect account of popular sovereignty falls short of satisfying the two agency desiderata that must be satisfied if citizens are to meaningfully act through the institutions of the state. Its exclusive reliance on mechanisms of representation, and the accompanying circumscription of popular participation, is bound to weaken citizens’ perception of being able to exercise collective political agency, especially in this age of individualisation and the decay of traditional political conflict lines.

Popular Sovereignty Through Parties

An alternative account of popular sovereignty is available, one that appears better placed to satisfy the two agency desiderata specified above. This starts from the presumption that the traditional indirect account of popular sovereignty is insufficient in terms of enabling citizens to act through the institutions of the state since it mainly regards them as passive recipients of others’ decisions. Reacting to this, it stresses the central importance of internally democratic political parties for achieving popular self-rule, arguing that such parties can connect citizens to government in a more direct and continuous fashion. Like the two more familiar accounts of popular sovereignty, this is first and foremost a normative model. But unlike them, it has rarely been realised in practice. As we shall see, however, it provides a powerful prescriptive account of how citizens can engage with collective political agency so as to rule themselves democratically.

The basic elements of this complementary model of popular sovereignty are found in the democratic theory of Hans Kelsen, who may at first appear like an unlikely point of reference. After all, Kelsen is notorious for dismissing the notion of popular sovereignty as a “political fiction” whose sole purpose it is to legitimise the division of labour between citizens and their representatives. ⁴¹ Furthermore, although he accepted the argument that a division of labour between political elites and citizens was inevitable in modern societies, he was far from unsympathetic to the Rousseauian trope that democracy and representation are essentially incompatible. ⁴² Finally, Kelsen denied that there exists such a thing as “the people” to begin with. The people, he insisted, are not a collective of citizens sharing a common history and purpose, but a purely artificial entity, “a system of individual human acts regulated by the state legal order.” ⁴³ That is, the people are that group of individuals that is granted political rights by a particular state—but no more than that, contrary to ordinary folk imagination.

However, in spite of his scepticism toward many of modern democracy’s more central concepts, Kelsen did not rule out altogether that citizens could govern themselves in a sovereign fashion. This appears in the 1929 edition of Kelsen’s treatise On the Essence and Value of Democracy—especially the section addressing the concept of the people, where a focus on the relationship between the “ideal conception of the people” and the “real people” is intended to unveil the possibilities of popular sovereignty in modern democracies. Kelsen there suggests that “among those who in fact exercise their political rights by participating in government, one would have to differentiate between the mindless masses who follow the lead of others and those few who—in accordance with the idea of democracy—decisively influence the governmental process based on independent judgment.” ⁴⁴ The latter, he continues, are those who come closest to instantiating the ideal of a self-governing people in the real world: the people as active shapers of the general will—“a ruling, and not a ruled, People.” ⁴⁵

Where may those self-governing people be found in real-world democracies? Kelsen’s answer is that they are found in particular political associations, namely in political parties, which bring “like-minded individuals together in order to secure them actual influence in shaping public affairs.” ⁴⁶ His argument is as follows. If the people are to rule actively, then a form of association that promotes and supports the ongoing pursuit of particular political goals is essential. Parties provide such an association, supplying citizens with an institutional channel through which their principles and goals can be connected to the relevant legislative and executive mechanisms. It is this unique quality of parties that makes them the main enablers of popular self-rule. Kelsen follows this thought to its logical conclusion: “the ‘people’ does not actually exist as a viable political force prior to its organisation into parties”; it is only the “integration of isolated individuals into political parties” that “unleashes social forces that can be reasonably referred to as the ‘People.’” ⁴⁷ Popular self-rule thus “rests on political parties, whose importance grows the more the democratic principle is realized in practice.” ⁴⁸

So Kelsen uses the concept of the people in two different senses. The people in the first sense is meant to refer to all those who a given state grants political rights. One may call this the “legally defined demos.” The people in the second sense are those members of the legally defined demos who actively exercise their political rights, and exercise them not only by voting in elections but by participating in political parties. It is this second conception of the people that I am interested in here. It is Kelsen’s central conceptual innovation and distinguishes his account of popular sovereignty from alternative accounts.

The people in this second sense is a group of individuals who actively participate in democratic self-rule but do not form a single collective agent, as in the radical account of popular sovereignty. Nor does this concept of the people however describe a loose collection of individual citizens who stand in no particular public relationship to each other and whose collective will only emerges through the mediation of elected representatives, as in the indirect account of popular sovereignty. Rather, Kelsen’s self-governing people is a people that divides into several different collective agents who associate as partisans so as to influence political decisions in accordance with their principles and interests. Now it is of course true that parties also figure in the indirect account of popular sovereignty. But note that while indirect accounts operationalise parties as “mere” representation-enablers run by professional politicians, the Kelsenian model conceives parties as participatory organisations that directly implicate citizens into their internal decision procedures, thus giving them the opportunity to become active shapers of common rules.

That Kelsen favoured internally democratic parties is a feature of his democratic theory that easily goes unnoticed, for he does not elaborate much on it. I interpret Kelsen to think that parties have to be internally democratic first and foremost because it is difficult to see how citizens could otherwise effectively use parties as channels for collectively exercising influence in the way he describes. But there is further support for this interpretation, namely in Kelsen’s reference to Robert Michels’s classic study Political Parties (which demonstrated that the “mass parties of integration” that became the dominant democratic forces throughout Europe in the early twentieth century were democratic only in appearance, but oligarchic in reality). ⁴⁹ Kelsen expressly agrees with Michels that “even parties pursuing a radically democratic programme” often organise the “processes of collective will-formation that occur within [them]” in “an explicitly aristocratic-autocratic” fashion, and takes issue with this in that he approvingly notes that, “anchoring political parties in the constitution provides the possibility for democratising the aspects of the governmental process that occur within the parties’ sphere of influence.” ⁵⁰

Against this interpretation, it might be argued that Kelsen’s project is not primarily or necessarily to make political parties more open and receptive to ordinary party members, but to strengthen the control of party machineries over their elected party officials. This alternative interpretation is supported by several passages in chapter four of On the Essence and Value of Democracy, where Kelsen proposes that parties (1) should be able to recall their elected representatives, and (2) that representatives should be forced to resign once they decide to leave the party. ⁵¹ These mechanisms are intended to prevent individual candidates from using parties as a taxi to get into representative institutions and then change their party affiliation based on their own personal contingent ambitions and interests, thus disfiguring the role that parties ought to play in parliamentary democracies. ⁵² Yet, however much Kelsen was concerned about these dangers, it remains the case that his understanding of popular sovereignty would make little sense without an inclusive and participatory conception of political parties. As I said, it is hard to see how citizens could use parties as vehicles for influencing laws and policies in the way Kelsen describes if they are not organised democratically. Ultimately, there may well be tensions between a participatory understanding of parties and the ambition to make representatives more accountable through such institutional mechanisms as recall. To deny altogether that Kelsen envisioned parties that allow for internal democratic participation strikes me as implausible, however.

As I conceive it, then, the party-based account of popular sovereignty expands the indirect account in important ways. Similar to the indirect account, which regards parliament as the site in which the general will is engineered, it ascribes to parliamentary deliberation and debate the function of mediating between the collective wills of the multiple different “peoples” who associate in parties. By mediating is meant the achievement of a compromise between competing partisan demands, a “synthesis” resulting from the “dialectical process within parliament.” ⁵³ (Note in this connection that Kelsen emphasised repeatedly, and consistent with this broader view of pluralist party democracy, that Proportional Representation is the only adequate way of ensuring that the multiple different “peoples” that exist in a polity can make their voices heard in the parliamentary process. ⁵⁴ ) But in contrast to the indirect account of popular sovereignty, the democratic and participatory organisation of parties that the party-based account prescribes permits citizens to assume an active role in that process, contributing to the outcome of parliamentary procedures in a continuous and coordinated fashion. Citizens who associate in parties assume this active role insofar as they (for example) jointly shape the platform of their party, or deliberate with elected officials over a particular parliamentary negotiation strategy. In this way, they can influence the nature and the scope of the compromises pursued, as well as specify positions on which no compromise is acceptable.

Because of its distinctive participatory features, the party-based account of popular sovereignty is well-placed to satisfy the two agency desiderata specified earlier. That it can satisfy the causality desideratum should be evident. The fact that those who engage in parties can actively shape the agenda the party ought to pursue, and so bring their principles and projects to bear on decisions taken in parliament, creates a tight causal link between citizens’ concerns and collective decisions. ⁵⁵ It creates a causal link that is much tighter than in the two traditional models of popular sovereignty. In the indirect account of popular sovereignty, citizens can only be said to “cause” the actions of officials insofar as the latter have a strong incentive to be re-elected (or at least be popular), and correspondingly are responsive to public opinion—but citizens cannot contribute much to more general political agendas. In the radical account, citizens produce a mandate by way of taking a direct decision on a single political issue, but as we saw, they are (usually) not involved in determining what this issue exactly involves and what courses of action must be pursued after the collective decision has been made.

The party-based account of popular sovereignty also satisfies the perception desideratum. It does so because, when parties are thoroughly democratic and permit their members to shape the party’s more general direction, those committed to the party and its wider political project will likely perceive its decisions (including negotiated bargains and agreements) as “theirs.” Of course, internal disagreements over, say, the correct interpretation of core principles, or the most appropriate electoral strategy, are likely to persist even within the most democratically organised parties, ⁵⁶ and so there will also always be party members who refuse to support their party’s decisions without reservations or protest. But so long as dissenting voices can weigh in and be heard, the notion that it is possible to link one’s action to collective decisions retains credibility, sustaining party members’ belief in the worth of engaging with political agency so as to exercise collective self-rule.

Notice that I have argued that internally democratic parties may increase the likelihood that those committed to a given party and its larger political project will perceive the party’s decisions as “theirs.” This formulation deliberately includes sympathetic non-members, who do not participate actively in partisan politics. Why? One straightforward reason is that internal party democracy can help parties reconnect with their constituencies, realigning their programmatic outlook with many principally sympathetic non-members who were alienated by the growing rift between representatives and represented that has resulted from the unravelling of the links between interests and organisations that has occurred across established democracies over the last four decades or so (see above). ⁵⁷ In short, parties that permit those in whose name they claim to speak to determine their political agenda may also significantly increase the extent to which non-members perceive the party’s decisions as traceable to themselves.

It is worth underlining in this connection that genuinely participatory and democratic parties are also likely to cater well to citizens’ shifting participatory preferences. Recall that the decline of class conflict and the rise of self-expression values has triggered a transformation of participatory demands: citizens in Western societies have grown increasingly dissatisfied with limiting their political participation to voting every couple of years, and they tend to want to have a say on a more regular basis. By providing a channel for meaningful and consequential political participation beyond periodical elections, internally democratic parties could go a long way in satisfying these demands, drawing in the renewed involvement of traditional constituencies and mobilising new, perhaps hitherto disenfranchised or silenced ones (and thus also address one of the key causes of the present crisis of party democracy). ⁵⁸ This capacity to elicit participation constitutes a significant advantage of the party-based approach to popular sovereignty relative to the indirect approach. It powerfully counteracts the risk of making citizens feel disconnected from political agency that the latter inevitably brings with it.

What role, one may wonder, does the account of popular sovereignty defended here ascribe to citizens who choose not to participate in party politics? Are those who Kelsen unflatteringly calls the “mindless masses who follow the lead of others” bound to be dominated by those who “decisively influence the governmental process”? The answer to this question is two-pronged and part of it has already been anticipated. On the one hand, there is of course a power differential between those who do participate in parties and those who don’t. That those who actively seek to harness institutional power have a greater impact on collective decisions is ultimately inescapable in democratic politics. On the other hand, however, the aforementioned capacity of the party-based approach to reconnect parties with their constituencies, together with the additional participatory incentives it engenders, can offset some of these differences in de facto influence. To the extent that those who participate in parties ensure that the party’s agenda and decisions track the preferences of those in whose name it claims to speak, internally democratic and participatory parties can therefore tether Kelsen’s “mindless masses” closer to the democratic process, situating them in the circumstances of agency.

So, to sum up, the party-based account of popular sovereignty sees the people as consisting of several collective agents—political parties—who seek to influence the general will in accordance with their interests and normative commitments. The account acknowledges that a distribution of democratic labour is inevitable in complex modern societies; it accepts the principle of representation and emphasises the ultimate importance of parliamentary deliberation aimed at compromise. What renders it unique is that it operationalises parties as participatory associations through which citizens can exercise influence on collective decisions in a more continuous and immediate fashion. This not only makes for a much stronger causal link between citizens’ preferences and collective decisions compared with the familiar radical and indirect accounts of popular sovereignty. It also delivers on the important concern that the meaningful exercise of political agency requires that citizens perceive outcomes as connected to their intentional attitudes.

Objections and conclusions

One might object at this point that the argument developed so far is marked by an imbalance ⁵⁹ between a very critical treatment of radical and indirect approaches to popular sovereignty, on the one hand, and a relatively uncritical treatment of political parties, on the other. This imbalance might be traced to how evidence about parties is deployed. While numerous failings of real-world political parties are mentioned as the source of radical and indirect populist sovereignty’s shortcomings, the discussion of the party-centred approach makes little notice of these failings, thus effectively ignoring facts about parties that are earlier highlighted in order to critique the two more familiar approaches to popular sovereignty.

This objection is understandable, but it misses the point of the argument I have been trying to make. Recall first that the party-based account of popular sovereignty is a normative account, yet unlike the other two accounts it has rarely been realised in practice. Its foremost aim is to show how we can improve on existing practice, connecting theoretical reflection and empirical observation. If we take that ambition seriously, we must also acknowledge that the party-based account is in fact far from uncritical towards parties: indeed, by suggesting that parties should be internally democratic and participatory, it allows for the diagnosis that very few existing political parties currently work as they should. In short, the argument advanced is not oblivious to the reality of political parties but rather counterposes to that reality an alternative view of political parties, the proposition being that that alternative view of parties can furnish a more promising account of how popular sovereignty can be realised in established democracies. I take this to be consistent with Kelsen’s theoretical intentions: being well aware of parties’ many empirical shortcomings, his aim was to make a more general point about what well-functioning parties can achieve and why it matters.

Even if this response is accepted as partially tempering the initial objection, the concern remains that the proposed conception of well-functioning parties is hopelessly optimistic. If it is true, as I suggested, that party leaderships are typically more concerned with holding onto power than with enhancing citizens’ ability to rule themselves, why should we believe that contemporary political parties would ever adopt an organisational form that gives members and supporters a greater say within the party? And if there is no reason to believe that parties could become more internally democratic and participatory, why should we think that “popular sovereignty through parties” is a goal worth pursuing the first place? This criticism sees the desirability of Kelsen’s understanding of popular sovereignty as closely bound up with its feasibility, and doubts its feasibility on the grounds of how parties operate under contemporary political conditions.

Admittedly, this worry is difficult to answer given what we know about how political parties tend to work. It would be foolish to deny that the resistance of party elites to power-sharing and internal democracy is a major obstacle to realising popular sovereignty through parties, one that will be hard to overcome. Yet it would be equally rash to think about these obstacles in a deterministic fashion, as if “iron laws” were at work. One straightforward reason for why is that even long-established parties have seen “revolutions from below” in which unexpected changes of leadership initiate internal democratisation and an expansion of the active membership base—an important recent case is the British Labour Party after the election of Jeremy Corbyn as party leader. ⁶⁰ Accepting this point does not require us to renounce our critical assessment of contemporary parties. We can retain the view that party elites are often self-interested and power-seeking while acknowledging that, at critical moments at least, there may be very real possibilities for organisational change and renewal—and hence also for actualising the party-based account of popular sovereignty. ⁶¹

Some readers might wonder whether this means that citizens will ultimately have to wait for the just-mentioned “critical moments” or “revolutions from below” in order for them to be able to exercise popular sovereignty? If that were the case, arguably we would again face a feasibility-related objection, albeit a different one than before. This is that, even if popular sovereignty through parties is feasible in the sense that it can in principle be realised, the likelihood that it will be realised is very small. For, although critical moments of organisational renewal and internal revolutions from below are certainly a possibility, to rely on them to happen in established democracies is simply absurd. In addition, even when parties are radically transformed, such transformations will at most occur in single parties, but certainly not at a scale sufficient to render a critical mass of parties inclusive participatory organisations. All the more reason to discard the notion of popular sovereignty through parties.

There are at least two ways to answer this objection. First, although it would be wrong to deny that feasibility and likelihood are connected in relevant ways, it is also deeply problematic to conflate feasibility with unconditional or near-unconditional likelihood. Doing so risks the kind of cynical realism that equates the real with the ideal, making superfluous any normative inquiry that goes much beyond the defence of the status quo. ⁶² Consider further how immensely difficult it is to determine in a non-arbitrary way what it means to say that a normative model’s realisation is sufficiently likely in order for it to be worth aiming at. As long as there are no in-principle unsurmountable obstacles to its realisation (which is not the case with Kelsen’s notion of popular sovereignty through parties), the judgment that it is sufficiently likely to be worth pursuing is almost entirely subjective. It depends as much on how we interpret empirical opportunity structures as on how much we want the model to become a reality. So even if it may plausibly be doubted that the instant realisation of the party-based vision of popular sovereignty is very likely, this is not sufficient to delegitimise it as a normative model.

A second possible counter-argument to the above objection is that we need not necessarily wait for “critical moments” or “revolutions from below” for popular sovereignty through parties to become a reality, but can try to take concrete measures to bring it about. This was also Kelsen’s response to the challenge of feasibility. As we saw, he argued that by “anchoring political parties in the constitution . . . the amorphous structure of the parties” due to which “the political processes that occur within them take on an explicitly aristocratic-autocratic character” could be reshaped into a manifestly democratic form, counteracting those “inner workings of the party” that undercut “democratic self-determination.” ⁶³ Though with the benefit of hindsight we may judge this particular proposal spurious, for none of the several waves of party constitutionalisation that have occurred throughout Europe since the end of World War II have palpably strengthened internal party democracy, ⁶⁴ it points to the importance of reflecting on strategies of actively promoting internally democratic and participatory parties.

Another possible way forward is certainly to form new partisan associations with an inclusive and participatory organisational structure. This means not having to reckon with the inherited organisational hierarchies and professionalised power elites that often prevent established parties from implementing democratic reforms, opening the door to experimentation with different mechanisms of internal participation. To be sure, experiments of this sort do not always succeed—the much-discussed cases of Podemos and the Movimento Cinque Stelle (M5S) serve as a warning example of how initial democratic ambitions (if they were ever there) can dissipate quickly. ⁶⁵ But even when the impulses of new partisan groupings to promote new, more democratic forms of partisan political engagement are short-lived, they can still provide a powerful check on whatever tendencies there exist among long-standing parties to refuse thinking about their relationship with the wider citizenry and what opportunities to make their voices heard they grant their members and supporters.

Fully setting this dynamic in motion requires lowering institutional barriers for political newcomers to establish themselves on the political scene. Chief amongst these are thresholds for parliamentary representation. These are characteristically low in electoral systems based on the principle of proportional representation, which, as Kelsen affirmatively noted, encourage “the formation of small, indeed the very smallest, parties.” ⁶⁶ One might likewise think of softening restrictions on access to public finance. In any case, the point to note is that correcting some of the problems generated by existing parties requires creating more opportunities for new, more democratic and participatory parties to emerge and set foot in parliament. Incidentally, this also means that the aforementioned “critical mass” of internally democratic parties that allow a maximally large number of citizens to democratically rule themselves is something that might be achieved by new parties that replace older ones, rather than by older ones reinventing themselves. This conclusion, it should be noted, is not surprising if the wider argument of this article is accurate. If anything, it affirms the proposition Kelsen put at the centre of his democratic thought: that effective popular self-rule rests on political parties.