Introduction

From a citizen perspective, democratic government bears at least three promises. First, having a say in collective decisions that result in binding law fulfils democracy’s promise of individual and collective autonomy, or effective self-government. Secondly, democracy institutionalizes the principle of equality, or equal autonomy. In a democracy, every vote counts equally and every citizen is supposed to have the same say on laws that apply to all of them equally. Thirdly, democracy is associated with a promise of rationality: citizens expect democratic decisions to be reasonable and well-justified, and most of them tend to think that democracies are overall less likely to produce decisions to be regretted in the future than autocracies are. These promises have aspirational character, meaning that to date, no political system has truly fulfilled them. In the minds of citizens, however, they constitute an ideal normative standard for the evaluation of democratic institutions and practices.

In contemporary representative democracies, the ‘democratic deficit’ (Norris 2011) between the ideal standard and citizen evaluations of existing institutions and practices appears to grow. One interpretation of the resulting crisis of democracy is that, with the inevitability of representation in modern mass democracies, republican ideals of egalitarian self-government have become unrealistic. Self-proclaimed ‘realists’ argue that respective promises cannot possibly be met and that we should give up unattainable ideals in favour of second-best alternatives such as competitive oligarchy or epistocracy (Achen/Bartels 2016; Brennan 2017). Against such sceptical perspectives, we argue that deliberative democracy still provides both an attractive normative ideal and an evaluative framework that can help us to guide and assess attempts to reconcile democracy’s republican promises of autonomy, equality and rationality with liberalism, the rule of law and necessary representation.

In our eyes, the resulting challenge for deliberative democrats and scholars of comparative government consists in developing criteria that enable us to determine if, how and under what conditions the attempt to fulfil democracy’s promises is empirically successful. Accordingly, we argue for a criterion of deliberative responsiveness that takes the fundamental condition of representation into account in specifying normative standards of democratic government. More precisely, we seek to show that deliberative responsiveness is a criterion representative democracy can and should aspire to meet and apply in the diagnosis of contemporary problems and possible solutions to them.

The following section elaborates on democracy’s promises of autonomy, equality and rationality and explains why responsiveness has become an important criterion for the assessment of the democratic qualities of representative government. We also point out that empirical demand for responsible rather than responsive government and theoretical arguments about the endogeneity and resulting volatility and manipulability of preferences to political processes have cast doubt on the adequacy of responsiveness as a criterion for successful democratic representation. We argue that a deliberative understanding of politics allows us to appreciate the endogeneity of political preferences to public discourses without giving up on responsiveness as a normative criterion for the success of representation. At the same time, as we show in section ‘Deliberative democracy, representation and institutional design’, theories of deliberative democracy have in many respects remained institutionally underdetermined. Deliberative democracy would therefore benefit from more systematically taking into account insights from comparative government and political economy in the assessment of institutional design and its consequences. Section ‘Broken Promises: How and Where Deliberative Responsiveness is Missing in Representative Democracies’ discusses three ways in which contemporary representative democracies fail to live up to their promises and to realize deliberative responsiveness: by elite deliberation being unplugged and insulated from citizen deliberation, by unequal and insufficient responsiveness to marginalized groups and by voting over party platforms failing to produce policies preferred by democratic majorities. Without aiming to provide a blueprint for an ideal representative constitution, we go on to present three possible institutional reform strategies that have potential to counteract these problems and to improve deliberative responsiveness (section ‘Institutional reform beyond “democratic innovations’). The conclusion summarizes the results and points out desiderata for future research.

Democracy’s Promises

As noted before, support for democratic government seems to be driven by three promises entailed in the ideal of democracy: the promises of autonomy, equality and rationality. What do these promises respectively entail, and to what extent can they be fulfilled?

Autonomy must be seen as the key promise of democratic government. Autonomy obtains where a person is free to choose their own goals and way of life without undue interference from others. Democracy is the only form of political rule to convincingly promise the exercise and protection of autonomy, as those who are constrained by decisions are ultimately the same ones who make these decisions. In a political context, autonomy can thus be understood as self-government. At the same time, tensions between individual and collective autonomy are inevitable where democratic rule is institutionalized as majority rule (Richardson 2002, p. 58). While autonomy is a condition to be fulfilled first and foremost at the individual level, political associations are formed precisely for the reason that individuals have goals that they cannot attain by themselves (Dahl 1998, p. 35). The ‘circumstances of politics’ are such where citizens ‘disagree about both the right and the good, yet nonetheless require a collective decision on these matters’ (Bellamy 2007, p. 5, with reference to Weale and Waldron). Contract theorists resolve the conflict between individual and collective autonomy by assuming unanimous decisions, in which every individual can safeguard their autonomy with a veto right. Majority decisions have to pass a higher justificatory threshold to be considered as realizing both individual and collective autonomy. Republican visions inspired by Rousseau’s identitarian understanding of democracy seek to lift the separation between rulers and ruled through direct self-legislation. However, the kind of direct democracy Rousseau aspired to is neither suitable for secular, pluralistic societies nor feasible under conditions of mass democracy. Neo-republicans therefore translate the promise of autonomous self-government into one of ‘non-domination’: citizens, considered as equals, must be able to jointly control government to the degree that it cannot arbitrarily interfere with their lives (Pettit 2012). Jürgen Habermas’s discourse on the theoretical interpretation of the principle of popular sovereignty (or collective autonomy) goes beyond the neo-republican standard by requiring empowered citizens to be able to view themselves as joint authors of their collective fate.Footnote 1 We will return to the joint authorship of laws as a criterion for democratic autonomy below.

Equality as a promise of democracy is closely entwined with the idea of autonomy. If autonomy is unequally distributed among citizens, that is, if some citizens have more of it than others, those who have more will necessarily be in a position to dominate those who have less. Equality, understood as institutionalized equal autonomy, has therefore been treated as the elementary principle of democracy (Dahl 1998, p. 37). In pre- and early modern times, the idea of equality was a radical one, with multiple forms of hierarchy and domination well-established and regarded as natural by most people. Early modern individualism, most notable in the writings of Hobbes and Locke, made political legitimacy dependent upon the consent of individuals possessed with equal natural rights. Although this kind of individualism continued to be used to justify other than democratic forms of rule (in Hobbes’s case, absolutist monarchy), and although women, and thus more than half of the population, long remained conceptually excluded from rights, the idea of equality was powerful in extending and spreading democratic forms of rule over time and across the globe. The promise of equality results in the expectation that citizens should, with one vote each, have an equal say on political decisions, that they be equally bound by laws and equals before the law. While the principle of equality before the law seems to be almost universally endorsed in consolidated democracies,Footnote 2 there is more controversy about the extent to which political equality requires social and economic equality or a guarantee of at least minimal social rights. T. H. Marshall argued that only social rights and a reduction of the income gap would enable all citizens to effectively use their civil and political rights in their own best interest (Marshall 1950). More recently, Näsström and Kalm have sought to demonstrate how the production of precarity in contemporary societies fosters a privatization of responsibility that corrupts the public core of democracy (Näsström and Kalm 2015). There are thus strong arguments that political equality cannot be ensured by the mere guarantee of formally equal suffrage alone.

Finally, rationality as a promise of democracy may not be as central as the promises of equality and autonomy to most theorists of democracy but it is without doubt relevant from a citizen perspective. Citizens support democracy not only for intrinsic reasons, as an end itself, but, as noted above, also as a means to achieve goals which they cannot attain by themselves. This kind of instrumentalism in support for democracy should not be viewed narrowly. A person supporting democratic rule only to achieve a specific policy decision in their own favour is not a democrat at all. However, citizens do expect democratic rule to enable decisions that are, on the whole and over time, in some sense ‘better’ than decisions taken by other than democratic means and less likely to be regretted in the future. From a citizen perspective, democracy is also an exercise in problem-solving:

The rational acceptability of results achieved in conformity with procedure follows from the institutionalization of interlinked forms of communication that, ideally speaking, ensure that all relevant questions, issues, and contributions are brought up and processed in discourses and negotiations on the basis of the best available information and arguments. (Habermas 1996, p. 170)

Taking democracy’s promise of rationality and citizens’ hopes in collective problem-solving seriously, however, does not necessarily entail an epistemic understanding of democracy, according to which the goal of democracy is to track truth or identify ‘correct’ decisions. By contrast, a deliberative perspective such as Habermas’s can regard democratic will-formation and decision-making as a constructive enterprise and reject the notion of pre-politically existing ‘solutions’ to problems waiting to be ‘found’.

Considering autonomy, equality and rationality as the promises associated with democracy in the eyes of citizens, we can better understand how and why democracy fails to meet their expectations. Modern democracy is necessarily to a considerable degree representative, which seems to make the simultaneous fulfilment of these promises difficult, leading to frustration and alienation among citizens. Apparent shortcomings of representation and resulting frustration among citizens seem to substantiate a republican perspective on representative democracy as only a second-best option to direct democracy. Republican theorists have thus advocated reforms to return from a ‘weak’ liberal and representative to a stronger, more direct and participatory form of democracy (e.g. Barber 1984). By contrast, Runciman argues that Hobbes has correctly identified representation as the fundamental condition of citizens in modern states and the only way to create unity in multitude (Runciman 2021, p. 23). In a similar train of thought, Urbinati argues that representation fulfils an important function in dividing sovereignty from government, thus ensuring the self-limitation of majority rule (Urbinati 2019, pp. 90−99). Democratic theory more broadly, and deliberative theory in particular, has since the 1990s engaged in a positive reappraisal of representation, with David Plotke pointing out the way in which it is crucial to constituting democratic practices (Plotke 1997, p. 9), and Urbinati and Warren coming to the conclusion that ‘we need to understand representation as an intrinsic part of what makes democracy possible’ (Urbinati/Warren 2008, p. 395).

If representation is not only a necessary, but even a desirable element of modern democracy, but at the same time makes it difficult to fulfil promises citizens associate with democracy, we seem to be in need of alternative and perhaps less demanding standards for democratic government. When can we say that democratic representation succeeds? Hannah Pitkin’s (1967) ‘standard account’ (Urbinati/Warren 2008, p. 389) of representation is well-established especially in the empirical literature and among scholars of government. From this standard account, responsiveness emerges as the central criterion representative government has to meet in order to qualify as democratic:

… a representative government requires that there be machinery for the expression of the wishes of the represented, and that the government respond to these wishes unless there are good reasons to the contrary. […] a representative government is one that is responsive to popular wishes when there are some. Hence there must be institutional arrangements for responsiveness to these wishes. (Pitkin 1967, pp. 232–233)


In an often somewhat selective reading of Pitkin, representation is authorized through elections and responsiveness of representation assumed to be ensured by citizens’ opportunities to hold representatives to account in future elections by sanctioning governments that have not responded to their concerns.


While responsiveness seems necessary to fulfil democracy’s promises of autonomy and equality and to ensure that government decisions are driven by citizens’ own goals and preferences, the promise of rationality may depend more strongly on prudence and expertise, to ensure that decisions are based on good reasons and evidence. This requirement is consistent with a Burkean understanding of representation as trusteeship, with representatives acting in the interests of citizens, but not necessarily in accordance with their expressed wishes and opinions:

Your representative owes you, not his industry only, but his judgement; and he betrays you instead of serving you if he sacrifices it to your opinion. (Burke 1774)


From this perspective, it could be argued that low responsiveness need not be a problem, but can also be indicative of rational government that ignores at least some wishes and preferences in favour of the common good. Moreover, there is some empirical evidence that at least in times of crisis, citizens may value expertise-driven government even more than responsive government (Bertsou/Caramani 2022).

However, a still more fundamental challenge to responsiveness as a criterion for successful representation arises from the observation of preference endogeneity. The idea of responsive government implies that citizens have interests and preferences that exist prior to and independent of political processes and that are in this sense exogenous to them. But quite obviously, preferences over specific political programmes and decisions cannot be regarded as pre-political but are the result of political communication and thus endogenous to political processes. The constructivist turn in representation theory has directed attention to reciprocal processes between representatives and citizens in which constituencies are created, salient issues evoked and political conflicts shaped (Saward 2010). On this basis, Lisa Disch has forcefully challenged the ‘bedrock norm’ of democratic theory, according to which pre-political interests should inform and ultimately drive political decision-making (Disch 2021). According to Disch, it is not so much citizens shaping representative decision-making, but competing elites shaping citizen preferences and conflicts between them. But if representatives are responsive to preferences which they themselves have helped to create, their responsiveness cannot be viewed as a means to ensure citizens’ autonomy and cannot serve as a criterion for democratic government.

If citizens value rational as much as responsive government, and if the idea of democratic responsiveness is vitiated by the observation of preference endogeneity anyway, does this mean that we have to give up on responsiveness as a standard for democratic government? Our answer is no. For citizens of representative democracies, responsiveness remains an indispensable criterion for assessing government performance and for deciding to what extent democracy meets its promises of equality and autonomy. Unless citizens can expect representatives to be responsive to preferences citizens themselves regard as autonomously formed, they are left with little reason to support democracy at all. Nevertheless, a tension between the promises of autonomy and equality on the one hand and the promise of rationality and demand for responsible government on the other hand remains. We thus need to rationally reconstruct responsiveness as a criterion for successful democratic representation from a theoretical perspective that allows us to view the exchange and assessment of reasons and the exercise of equal individual political autonomy as inextricably connected (Gaus et al. 2020, p. 338).

In our eyes, the theory of deliberative democracy offers such a perspective that, in Jon Elster’s words, views politics as ‘instrumental in purpose and public in nature’ (Elster 1997, p. 26). Contrary to constructivist theories of representation, citizens themselves view their interests and concerns as to a considerable extent pre-political, and deliberative democracy in Habermasian tradition reconstructs their ‘life-world’ (Lebenswelt) as distinct from (although not unconnected to) the political system (Habermas 1984). In this sense, pre-political concerns motivate citizens to participate in politics, which in its nature is public: reasons are exchanged and assessed, and more concrete political preferences are formed and transformed only through communicative interaction. We can thus regard preferences as endogeneous without giving up on the notion of extra- and pre-political interests and concerns and understand them as discursively constructed and autonomous at the same time. If, with Richardson, we assume that ‘we must reason together in order to rule ourselves’ (Richardson 2002, p. 18), the assessment of reasons and rational preference-formation in discursive processes of political will-formation can be seen as enabling democracy to live up to its promises of both equal autonomy and rationality.

The kind of responsiveness that is at stake here should therefore be understood as one of deliberative responsiveness. Understanding representation as a dynamic process rather than a static relationship between citizens and representatives, we argue that deliberative responsiveness obtains where citizens and representatives form informed preferences in inclusive and egalitarian discourses, leading to responsive decisions over policies and decision-making procedures which representatives explain and justify to citizens with adequate reasons. On the basis of a deliberative understanding of democracy, deliberative responsiveness can be reconstructed as a central criterion for successful democratic representation. While it is obviously not the only normative criterion than can be applied to political institutions and practices, we regard it as a necessary (but not sufficient) condition for representative democracy to fulfil the promises it holds in the eyes of its citizens.

The next section will discuss the status of the theory of deliberative democracy, arguing that it would gain from closer co-operation with comparative politics in order to overcome its institutional under-determination. We then go on to discuss how deliberative responsiveness is failed to be realized in many contemporary democracies. Assuming that deliberative responsiveness is to a significant degree a function of institutional design, we will close with some suggestions for institutional reform that have potential to improve it.

Deliberative Democracy, Representation and Institutional Design

For nearly 30 years, deliberative democracy has been the dominant paradigm in democratic theory. While many contemporary democracies seemingly have come to be characterized by practices inimical to democracy’s promises being fulfilled, the idea of deliberative politics offers a hope that these promises might still somehow be met and reconciled. In a recent paper, John Dryzek and a number of other prominent scholars in the field describe deliberation as a ‘science’ to be used in the diagnosis and therapy of democracy’s ailments (Dryzek et al. 2019). As a normative ideal, deliberative democracy seems to increasingly appeal to citizens and political elites alike, helping them to reconstruct their own practices and the institutional context which they act in as meaningful. In other words: the deliberative ideal provides them with opportunities to make sense of democracy.

Since the 1990s, the concept of deliberation has become influential throughout the social sciences, being used in disciplines such as sociology, psychology and even economics. It remains important, however, to distinguish between deliberation as a practice and method and deliberative democracy as a regime. As a practice, ‘deliberation’ describes primarily an intra-subjective process of organizing information and weighing reasons to arrive at a good decision. In group decision-making, this ‘intellectual process of identifying alternatives, gathering and evaluating information, weighting considerations, and making judgments’ (Quirk et al. 2018, p. 274) can be turned into an inter-subjective method. The Harvard concept for negotiations (Fisher et al. 1992 [1984]), townhall meetings or planning cells could be viewed as implementations of this method of deliberation avant la lettre.

Theories of deliberative democracy assign the practice and method of deliberation a central role in democratic decision-making, advocating deliberation as a central component of democratic political regimes. The first generation of deliberative democrats formulated their theories in explicit demarcation from the economic and aggregative accounts of democracy that were dominant in the 1980s, and which had reduced democratic rule to the aggregation of pre-political preferences and its success to a narrowly conceived responsiveness to these preferences.Footnote 3 By contrast, deliberative democrats highlighted the importance of discursive opinion and will-formation and the fact that political preferences are at least in part endogenous to such discursive processes.

The central idea that makes deliberative democracy so appealing is that it enables rational decisions that qualify as ‘good’ by virtue of taking everybody’s interests and all relevant arguments into account. Deliberative democracy thus combines participatory and epistemic aspirations, or the promises of egalitarian self-government and rationality, through democratic procedures: broad and equal participation in discursive processes ensures decisions that are at the same time the result of the exercise of individual and collective autonomy and more likely to be reasonable than decisions taken by other means.

While equality and broad participation are central to the idea of deliberative democracy, the theory is at its roots a theory of representative democracy. In developing their theories of deliberative democracy, scholars were influenced by the institutional orders they knew best and often tended to idealize them. The most prominent example is John Rawls (Rawls 1971, 1993), whose discussion of political justice under a just institutional order is evidently based on an idealization of the American Constitution (see Bellamy 2007, p. 100). J. M. Bessette, who was the first to use the term ‘deliberative democracy’, located deliberation in the American Congress (Bessette 1980). And the theory of democracy presented in Jürgen Habermas’s ‘Between Facts and Norms’ (Habermas 1996) is also a sociological reconstruction of (German) representative democracy as deliberative democracy (Gaus 2015).

The systemic turn the theory has taken in the last decade (Mansbridge et al. 2012) is thus in an important sense also a return to its roots. After an empirical turn at the beginning of this century (see Bächtiger et al. 2010) had resulted in many researchers becoming preoccupied with the institutionalization of deliberation in citizen conferences, mini-publics or participatory budgeting, the revaluation of the insight that legitimacy claims have to be directed to the systemic level and not to individual forums within the system was an important step. ‘Deliberative systems’, proponents of the systemic approach argue, can include both deliberative and non-deliberative forms of interaction and decision-making. What matters is that these are deliberatively justified and contribute to the deliberative and democratic qualities of the system as such.

Nonetheless, even after its ‘systemic turn’ (Parkinson and Mansbridge 2012), deliberative democracy remains institutionally under-determined. This under-determination seems due to the fact that deliberative democrats tend to focus on the effects of adding deliberative innovations, such as mini-publics, to a representative democratic system while neglecting the implications of institutional design more generally. Although deliberative democrats now seem to agree that deliberation in itself is not sufficient to take democratically authorized decisions and appreciate the need for representative and majoritarian procedures, few of them seem interested in the rules by and conditions under which these decisions are taken. In consequence, deliberative democracy after the systemic turn still predominantly fails to appreciate how representative institutions and procedures retroact on the processes that precede decision-making, and thus on deliberation itself.

More generally, deliberative democracy remains largely blind to the political economy of decision-making institutions and procedures, and thus to the fact that any institutional set-up has outcome effects that benefit some groups and interests more than others. Patterns of citizen participation (such as electoral turnout and political interest) and elite competition, as well as the mobilization of interests and cleavages are driven by political institutions. If deliberative democracy continues to neglect these implications of institutional design, the systemic turn might in fact end up promoting the conservative undercurrents in the theory that Owen and Smith (2015) have pointed out: if it ends up justifying all kinds of deficient institutions and decision-making procedures by arguing that they somehow contribute to the deliberative or democratic quality of the system, the approach gives up its emancipatory claim and loses its critical edge. To provide an adequate diagnosis of the shortcomings of existing representative institutions and use its institutional imagination to suggest innovations, deliberative democracy must therefore engage more seriously with arguments and findings from comparative politics and political economy.

Our point here is thus not to suggest specific deliberative procedures or forums for citizen participation such as large-scale deliberative assemblies to complement representative institutions. While such institutionalized deliberation has in individual cases such as Ireland certainly had positive effects on deliberation and responsiveness, creating additional opportunities for participation and citizen deliberation remains pointless where ‘deliberative uptake’ (Scudder 2020) is not ensured and where results do not have a real impact on political decision-making. Moreover, systemic biases entrenched in the design of central representative institutions—especially in the electoral system, the legislature and its relationship with the government—are unlikely to be offset by merely additive deliberative innovations. Instead of focusing on deliberative innovations, we suggest studying how institutional and societal configurations affect sources of deliberative responsiveness, as shown in Fig. 1 below:

Fig. 1
figure 1

Sources and dimensions of deliberative responsiveness

Full size image

As illustrated in Fig. 1, we assume deliberative responsiveness in a representative democracy to be derived jointly from at least three interdependent sources. First, as classical theories of representation have pointed out, responsiveness depends upon free, fair and regular elections that ensure the authorization and accountability of representatives (Pitkin 1967). The mechanisms by which authorization and accountability enable deliberative responsiveness (or not) are shaped by electoral systems and the number and positions of institutional veto points. For example, plurality-based electoral systems such as the British or US ones may increase responsiveness to territorial constituencies, but at the cost of undermining preconditions of effective deliberation, as Gerry Mackie has argued (Mackie 2018). Systems of proportional representation and consociational institutions that assign voice and veto rights to multiple groups and interests have potential for nurturing deliberation and conflict management in deeply divided societies as Ian O’Flynn has argued (O’Flynn 2010), but may come at the cost of blurring accountability. Secondly, deliberation taking place between the political system, civil society and ordinary citizens requires a vibrant public sphere that, in Habermas’s terms, ‘lays siege’ to the political system, ‘manages the pool of reasons’ that can justify decisions and that can successfully contest decisions (Habermas 1992). A critical and vibrant public sphere depends upon liberal rights of contestation (freedom of speech and association), but also upon media diversity as well as formal and informal structures that enable civic engagement and association. Thirdly, deliberative responsiveness requires direct interaction between citizens and representatives, which can happen in electoral campaign contexts, constituency work or deliberative forums, and which is essential for understanding citizen concerns, assessing the justifiability of planned policies and for the recruitment of party members and candidates. To make representation work, these three interdependent sources must feed into decision-making.

If decision-making is effectively based on these three sources, we should observe responsiveness in three dimensions. First, we should expect policy-making to be responsive to citizen preferences. That is, citizens should, over time, be more likely than not to receive the policies preferred by democratic majorities. Secondly, citizens not only have preferences over which policies are adopted, but also over how decisions are taken. As Hibbing and Theiss-Morse have argued, these process preferences are often in favour of more direct participation and less institutional democracy, and are not met with a sufficient supply on the party side (Hibbing and Theiss-Morse 2002). In line with a political rather than moral constitutionalism, however, we would argue that not only policy-making, but also institutional design must be responsive for representation to work. Procedural responsiveness obtains where institutions and decision-making procedures are adapted and revised in line with citizens’ changing process preferences. For example, the expansion of opportunities for deliberative citizen participation and forms of direct democracy since the 1980s can be viewed as a response to value changes associated with growing expectations for participation. Thirdly, responsiveness is not only about delivering policies and procedural reforms, but also about delivering explanations for them. Communicative responsiveness obtains where representatives provide answers to citizens’ concerns, arguments and questions, for example by giving justification in parliamentary speeches or press conferences, but also in letters, emails and direct communication with individuals. Once we adopt a dynamic rather than static understanding of representation, it becomes clear that congruence between citizen preferences (as voiced in opinion polls) and legislative decisions cannot be the sole indicator of responsiveness. Instead, representatives can be responsive in addressing problems and grievances voiced by constituents without implementing the specific measures preferred by them (see Severs 2010). If they justify their decision with reasons acceptable to citizens, they contribute to realizing deliberative responsiveness.

Broken Promises: How and Where Deliberative Responsiveness is Missing in Representative Democracies

Expecting that unless deliberative responsiveness in ensured, democratic representation will not work properly, we can apply it as a criterion in the assessment of existing democratic regimes. Assessing democracies for the extent to which they realize deliberative responsiveness or fail to do so can help us to better understand symptoms of its absence, such as alienation and polarization, and enable us to move towards identifying its institutional and societal determinants. While there will be numerous instances in any country where policy decisions on a given issue, procedural choices or communication and justification of decisions fail to meet this standard, we see three ways in which deliberative responsiveness tends to be systematically undermined in many contemporary democracies: by elite deliberation and decision-making becoming insulated or ‘unplugged’ from citizen deliberation, by responsiveness being not only low, but unequal with regard to the interests and preferences of different social groups and by party platforms disconnecting policy decisions from majority preferences. While insulated elite deliberation breaks the promise of autonomy and joint authorship of laws, unequal responsiveness violates the promise of democratic equality and the aggregation problem that becomes apparent in Ostrogorski’s paradox casts doubt on democracy’s promise of rationality.

Deliberation Unplugged

Assessing the quality of deliberation at the systemic level is an ambitious endeavour. The Varieties-of-Democracy (V-Dem) project tries to measure deliberation in several dimensions. However, as Fleuss and Helbig (2020) convincingly show, their measurement focuses on elite deliberation alone and fails to consider how it is connected to citizen deliberation and preference-formation. As a result, a country such as Germany, where the federal and multi-party system has increased the need for cross-party collaboration, scores very highly in the V-Dem measure. To avoid deadlock, political elites in Germany have learned to skilfully forge compromises. Yet the intensive elite-level exchange of opinions has not been accompanied by citizen involvement, and many citizens feel that politics is something that takes place behind closed doors exclusively. Where deliberation fails to connect citizens and elites because it is not ‘plugged in’ on the citizen side and where citizens feel that they have no real influence in decision-making processes, democratic autonomy is compromised.

In Habermas’s two-track model of representation, he insists that expert and citizen deliberation must feed into each other. The quality of deliberation depends on the interconnection between its two tiers. The deliberative connection between citizens and elites is important to detect new topics and put them on the agenda. While citizen deliberation might be less structured than that of professional politicians, it informs expert deliberations and safeguards against technocratic temptations. However, if the two tracks of deliberation become severed, representation is harmed. Deliberative responsiveness thus depends on a two-way communication process between representatives and citizens and cannot be achieved where deliberative uptake is muted. Lafont argues that democratic decision-making has to take the ‘long way’ of involving citizens in large-scale deliberation and self-government, instead of taking ‘shortcuts’ to decisions that effectively bypass the citizenry (Lafont 2020). The prioritization of elite and expert deliberation over public debate and citizen participation has, in Lafont’s terms, established a ‘technocratic’ short cut in political decision-making that leads to deliberation becoming unplugged at the citizens’ side. If decision-making is thus insulated from ordinary citizens, high-quality elite deliberation cannot prevent that, to citizens, reasons for decisions remain opaque and accountability is lost. In these circumstances, decision-making, whatever its other merits, is not deliberatively responsive. In sum, deliberation functioning too smoothly at the elite level may eventually undermine deliberation as a democratic regime by breaking its promise of autonomy.

Unequal Responsiveness

A large body of research shows that in many contemporary democracies, legislation is in fact little and unequally responsive to preferences citizens state in public opinion surveys (see, e.g., Gilens 2005; Elsässer et al. 2021). While congruence between opinion polls and legislation may not be a good measure of deliberative responsiveness, as opinions stated in polls are often unthinkingly voiced and less salient to respondents, inequalities found in responsiveness are more concerning. The factual exercise of equal political rights is highly unequal and biased towards privileged groups in most modern democracies, resulting in considerable and problematic inequalities in political voice. As numerous studies demonstrate, not everyone’s ‘voice’ is heard as clearly and loudly as that of the better off (Armingeon/Schädel 2015; Schlozman et al. 2018). Those with higher levels of resources vote in larger numbers, join parties more often, donate higher sums, write letters more frequently, protest more often and lobby more effectively. In addition, parliaments in rich democracies are crowded with university graduates, lawyers and business owners, while citizens without college degrees and from the working class are virtually absent (Carnes 2012; Bovens/Wille 2017; Carnes/Lupu 2015). We also know that not only policy changes demanded by the median voter frequently fail to be implemented, but that there are also considerable biases in whose interests representatives in fact represent. In general, political decisions reflect the preferences of rich and wealthy citizens much more closely than those of poor citizens (Gilens 2005; Bartels 2008; Elsässer et al. 2021). For large proportions of the citizenry, the democratic promise of equal autonomy thus remains unfulfilled.

Ostrogorski’s Paradox

The fact of representation means that, at least most of the time, citizens do not vote on policy alternatives but for parties and candidates. How do votes for party manifestos that address a large number of policy issues translate into legislative decisions over policies? Daudt and Rae have pointed out a paradox named after the Russian party researcher Moissei Ostrogorski (Daudt and Rae 1976). The paradox shows that where there are fewer parties than political issues, the winning party may be a party whose position on each of the individual issues lacks support from a majority of voters. Table 1 illustrates how this result may come about:

Table 1 The Ostrogorski Paradox
Full size table

Assume a two-party system with four voter groups. On issue 1, group 1 prefers the position party X takes, while on issues 2 and 3, group 1 prefers the position party Y takes. Taking the entire party platform into account, group 1 will thus vote for party Y. So will groups 2 and 3, both of which prefer party Y’s position on two out of three issues. While groups 1, 2 and 3 each make up 20% of the electorate, group 4 makes up 40% and prefers the position taken by party X on all three issues. As a result, party Y will win 60% of the votes, although on each individual issue, there is a majority for the position taken by party X.

Even if the implications of the Ostrogorski paradox are less severe for multi-party than for two-party systems, the fundamental problem remains: the way in which representative democracies aggregate voter preferences into policy decisions may not even ensure Pareto optimality as a minimal condition for rationality. The Ostrogorski paradox, however, not only shows how policy choices may fail to represent the majority will, it also demonstrates the importance of public deliberation and communicative accountability during the legislative period. In the example, the media and citizenry could hold party Y accountable and exert pressure on it to pass legislation that is supported by democratic majorities. If, as envisioned in Habermas’s model, party Y is under siege from a critical public sphere, it will have strong incentives to either follow the majority will or provide good justifications for not doing so. Where governing parties succeed in catering to special interests and in passing policy decisions without majority support, this is in part enabled by aggregation paradoxes, but must be treated first and foremost as a failure of deliberative government control. In consequence, democracy’s promise of rational decision-making is broken.

Institutional Reform Beyond ‘Democratic Innovations’

Having seen how contemporary democracies may fail to realize deliberative responsiveness, deliberative democrats should think beyond existing institutional orders and explore the potential of more far-reaching constitutional innovation. For ensuring deliberative responsiveness, we see less promise in the presently most popular innovation of deliberative mini-publics, which either remain merely additive innovations or lack democratic authorization and accountability. Instead, we should consider the core features of representative legacy institutions. Two of these appear particularly relevant: the electoral system and the organization of the legislature. While we do not intend to provide a blueprint for ideal deliberative and democratic institutions and while it is always important to take the interplay between individual elements in an institutional configuration into account, the following could serve as starting points for reform:

Single Transferable Vote (STV) Elections

The single transferable vote (STV) electoral system that is applied in Ireland, Malta and in sub-national elections in several other countries ensures both proportionality and strong contact between candidates/representatives and constituents (see Farrell 2011, ch. 6). STV requires a district magnitude greater than one, i.e., more than one representative is elected in any district. Ideally, the number of representatives per district should be higher than five in order to ensure proportionality. The ballot structure is preferential, meaning that voters can express their preferences over a number of candidates on the ballot paper. By use of a specific electoral formula, typically the Droop quota, it is decided which candidates pass the threshold number of votes to be elected or are dropped for having too few votes. If a candidate is dropped, votes for him or her are transferred to other candidates according to the next preference on the individual ballot papers. Thus, votes for losing candidates are not lost, but complete preference orders considered, which means that more information enters the decision about representatives.

STV is ‘quintessentially a candidate-based system’ (Farrell 2011, p. 155) in which candidates maximize their chances of being elected if they are ranked high in many people’s preference orders. This means that it will not pay to appeal to a narrow constituency or adopt radical, non-compromising positions, as a candidate needs to be liked by or at least be acceptable to as many people as possible. Contrary to majoritarian systems, STV thus tends to prevent polarization. Moreover, STV democratizes the selection of MPs, but does so without incurring the problems associated with primaries, in which only partisans have a say. It thereby creates more incentives for candidates and representatives to be responsive to relevant discourses in their constituencies, thereby improving discursive representation without sacrificing geographical representation and voter contact. STV also seems to follow the logic of deliberation better than any other electoral system: deliberation is about the weighting and ranking of options and not just about decision-making as such. In the selection of candidates, STV may thus help to prevent both ‘false positives’ and ‘false negatives’. By creating strong incentives for candidates to engage with constituents and to ensure communicative responsiveness, STV could thus counteract problems of elite deliberation becoming unplugged from citizen deliberation.

Egalitarian Elections

The unequal responsiveness to preferences of marginalized groups and the resulting loss of substantive representation of their interests can in part be accounted for by the lower participation rates of disadvantaged groups. Groups with low electoral turnout rates will feature less prominently in the development of party platforms, in within-party and government deliberation and in legislative decision-making. Measures to increase turnout and to improve egalitarian participation in elections will thus go some way to making responsiveness more equal and in helping democracy to live up to its egalitarian promises. We do not have space here to engage in detail with the controversy around compulsory voting (see Lever 2010; Birch 2016), which could undeniably result in more equal turnout rates. For a country like the United States that is troubled by severe problems and a long history of voter suppression, measures such as easing voter registration, ending felon disenfranchisement and holding elections on weekends could at least remove practical obstacles to participation. When it comes to voters’ motivation to turn out to vote, proportional systems generally do better than majoritarian ones and tend to result in less unequal participation rates. Substantive representation of disadvantaged groups and equal deliberative responsiveness could also be promoted by way of better descriptive representation. A ‘politics of presence’ (Phillips 1998) and quota rules for under-represented groups could ensure their discursive representation in decision-making processes and increase legislative responsiveness to their concerns.

Semi-Parliamentarism

An even more far-reaching reform of representative democratic systems would be the move towards what Ganghof terms ‘semi-parliamentarism’ (Ganghof 2021). The starting point of Ganghof’s argument in favour of semi-parliamentarism is that parliamentary systems are superior to presidential ones because the executive depends upon the parliament’s support. If a prime minister seems incapable of dealing with a crisis, for example, s/he could be voted out of office, which reduces their individual significance. At the same time, a parliamentary system with proportional representation leads to a multi-party legislature with a potentially high level of fragmentation. As a result, coalition governments with several parties might be necessary, which makes accountability and the identification of distinct party profiles more difficult. Hence, while proportional representation in parliamentary democracies has its democratic virtues, it is also induces costs in terms of efficiency and governability.

One of the problems with parliamentary systems is that the survival of a government depends on continuing support of all parties that have formed a coalition. If the government loses an important vote in the legislature, this often signals the end of co-operation. As a result, even policies that a majority of not only voters, but also of parliamentarians, support are not implemented if the supporters are scattered across governing and opposition parties. In an ideal setting, those policies that best reflect the preferences of citizens should be turned into law—but this will not happen if voting against the government on a particular issue threatens its survival. This is why semi-parliamentarism envisions two chambers that are jointly responsible for legislating, only one of which has the right to a vote of confidence.

The first, ‘policy’ chamber should reflect the preferences of citizens as closely as possible, which would suggest using a system of strict proportionality. Since a high effective number of parties is not detrimental to the tasks which this chamber has, there is no need to reduce the number of parties by way of electoral thresholds or majoritarian elections. By contrast, the electoral rules for the second, ‘confidence’ chamber could be such as to produce a clear majority to ensure stability. The party that receives a majority there is the natural governing party and nominates the prime minister. Semi-parliamentary government could thus avoid the problems associated with the Ostrogorski paradox and ensure that legislative decisions are at least Pareto-efficient. But it could do much more to ensure that democracy fulfils its promise of rational decision-making: freeing legislators from party discipline and the need to support a government enables them to deliberatively respond to public opinion and changes in it as well as to new evidence and arguments.

Conclusion

Democracy’s normative appeal rests on promising effective and rational self-government, in which citizens take part as equals. Since representation unavoidably creates differences between representatives and represented, democratic institutions and practices need to ensure that political elites and citizens do not disconnect. We have argued in this paper that deliberative democracy continues to constitute an attractive normative ideal that indicates how democracy’s promises of autonomy, equality and rationality can be reconciled under conditions of necessary representation. Avoiding fatalist and elitist pitfalls, it allows us to reconstruct participatory and egalitarian practices as enabling rational decision-making. However, taking the theory’s ‘systemic turn’ seriously also means that we should, if we are interested in whether and to what degree democratic representation succeeds, go beyond existing approaches in the literature by focussing neither narrowly on citizen deliberation in mini-publics nor considering only the quality of elite deliberation when assessing the success of democratic representation. What matters is that citizen and elite deliberation are interconnected, thereby enabling acts of representation that are responsive to citizens’ preferences and concerns.

Adapting Pitkin’s formulation, we thus contend that democratic representation requires acting in the interest of the represented in a deliberatively responsive manner. Accordingly, we have suggested ‘deliberative responsiveness’ as a normative standard for successful representation. Following a deliberative understanding of politics, we should neither view political preferences as given prior to and independently of public discourses, nor regard citizens as passive recipients of elite communication. Instead, our criterion of deliberative responsiveness stresses the coupling of various forums of deliberative interaction as well as the different dimensions—substantive (policy), procedural and communicative—in which representatives should be responsive to those they represent.

Under the conditions of representative democracy, deliberative responsiveness obtains in so far as political decisions are responsive to the concerns voiced in inclusive and egalitarian citizen deliberation. We believe that the degree to which requirements of deliberative responsiveness can be met depends to a large extent on institutional parameters and configurations of representative democracy and will affect citizens’ satisfaction with and support for democracy. A goal for scholars of democracy is thus to assess in how far democracies meet the standard of deliberative responsiveness and to identify its institutional determinants. In the second part of the paper, we have therefore pointed out ways in which contemporary democracies fail to ensure deliberative responsiveness and discussed institutional reforms that could potentially counteract these shortcomings. Implementing reforms to improve deliberative responsiveness would require a commitment to a democratic constitutionalism among citizens and representatives alike and should be based on processes of inclusive democratic meta-deliberation (Landwehr 2015). This paper seeks to make a contribution to such meta-deliberative discourses on institutional design for successful democratic representation.